A World Bank Group Flagship Report JANUARY 2020 Global Economic Prospects Slow Growth, Policy Challenges JANUARY 2020 Global Economic Prospects © 2020 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved. 1 2 3 4 23 22 21 20 This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http:// creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: World Bank. 2020. Global Economic Prospects, January 2020: Slow Growth, Policy Challenges. Washington, DC: World Bank. DOI: 10.1596/978-1-4648-1468-6. License: Creative Commons Attribution CC BY 3.0 IGO. Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third-party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@worldbank.org. ISSN: 1014-8906 ISBN (paper): 978-1-4648-1468-6 ISBN (electronic): 978-1-4648-1469-3 DOI: 10.1596/ 978-1-4648-1468-6 Cover design: Bill Pragluski (Critical Stages). The cutoff date for the data used in this report was December 20, 2019. Summary of Contents Chapter 1 Global Outlook: Fragile, Handle with Care ............................................................ 1 Special Focus 1 Price Controls: Good Intentions, Bad Outcomes ...................................................49 Chapter 2 Regional Outlooks ..............................................................................................61 Special Focus 2 Low for How Much Longer? Inflation in Low-Income Countries ..........................177 Chapter 3 Fading Promise: How to Rekindle Productivity Growth .......................................191 Chapter 4 The Fourth Wave: Rapid Debt Buildup. .............................................................251 Statistical Appendix .......................................................................................................................299 Selected Topics .............................................................................................................................306 III Table of Contents Foreword....................................................................................................................................... xiii Acknowledgments .......................................................................................................................... xv Executive Summary....................................................................................................................... xvii Abbreviations ................................................................................................................................ xix Chapter 1 Global Outlook: Fragile, Handle with Care ............................................................1 Summary .............................................................................................................3 Major economies: Recent developments and outlook ............................................... 7 United States..................................................................................................7 Euro Area ......................................................................................................8 Japan.............................................................................................................8 China ............................................................................................................9 Global trends .......................................................................................................9 Global trade ................................................................................................. 10 Financial markets.......................................................................................... 11 Commodity markets ..................................................................................... 12 Emerging market and developing economies.......................................................... 13 Recent developments..................................................................................... 13 Outlook....................................................................................................... 15 Risks to the outlook ............................................................................................ 25 Summary of global outlook and risks............................................................... 25 Rising trade barriers and protracted policy uncertainty ...................................... 26 A deepening slowdown in major economies .................................................... 27 Financial stress in EMDEs ............................................................................. 29 Geopolitical and region-specific downside risks ................................................ 30 Upside risks ................................................................................................. 31 Policy challenges................................................................................................. 32 Challenges in advanced economies.................................................................. 32 Challenges in emerging market and developing economies ................................ 34 Box 1.1 Recent developments and outlook for low-income countries........................ 16 Box 1.2 Regional perspectives: Recent developments and outlook ............................ 23 References.......................................................................................................... 42 V Special Focus 1 Price Controls: Good Intentions, Bad Outcomes .................................................. 49 Introduction ...................................................................................................... 51 Use of price controls ........................................................................................... 51 Challenges of price controls ................................................................................. 55 Policy implications ............................................................................................. 57 References ......................................................................................................... 58 Chapter 2 Regional Outlooks ............................................................................................. 61 East Asia and Pacific .......................................................................................... 63 Recent developments........................................................................................ 63 Outlook.......................................................................................................... 65 Risks .............................................................................................................. 67 Box 2.1.1 Labor productivity in East Asia and Pacific: Trends and drivers .............. 69 Europe and Central Asia .................................................................................... 77 Recent developments........................................................................................ 77 Outlook.......................................................................................................... 79 Risks ............................................................................................................. 80 Box 2.2.1 Labor productivity in Europe and Central Asia: Trends and drivers ........ 84 Latin America and the Caribbean ........................................................................ 95 Recent developments........................................................................................ 95 Outlook.......................................................................................................... 97 Risks .............................................................................................................. 98 Box 2.3.1 Labor productivity in Latin America and the Caribbean: Trends and drivers ......................................................................................... 102 Middle East and North Africa ........................................................................... 111 Recent developments...................................................................................... 111 Outlook........................................................................................................ 112 Risks ............................................................................................................ 114 Box 2.4.1 Labor productivity in the Middle East and North Africa: Trends and drivers ......................................................................................... 117 South Asia ....................................................................................................... 125 Recent developments...................................................................................... 125 Outlook........................................................................................................ 127 Risks ............................................................................................................ 128 Box 2.5.1 Labor productivity in South Asia: Trends and drivers .......................... 131 VI Sub-Saharan Africa ..........................................................................................141 Recent developments...................................................................................... 141 Outlook ........................................................................................................ 143 Risks ............................................................................................................ 144 Box 2.6.1 Labor productivity in Sub-Saharan Africa: Trends and drivers .............. 148 References........................................................................................................ 159 Special Focus 2 Low for How Much Longer? Inflation in Low-Income Countries ........................ 177 Introduction.....................................................................................................179 Evolution of inflation ........................................................................................180 Factors supporting inflation developments ...........................................................181 Monetary policy challenges ................................................................................182 Policy options going forward ..............................................................................185 References ........................................................................................................187 Chapter 3 Fading Promise: How to Rekindle Productivity Growth......................................191 Introduction .................................................................................................... 193 Evolution of labor productivity growth ............................................................... 196 Labor productivity convergence.......................................................................... 204 Sources of post-crisis slowdown in labor productivity growth ................................. 205 Long-run drivers of productivity growth.............................................................. 214 Prospects for productivity growth ....................................................................... 220 Policy implications............................................................................................ 222 Box 3.1 EMDE regional productivity trends and bottlenecks ................................. 198 Box 3.2 Sectoral sources of productivity growth ................................................... 209 Box 3.3 Patterns of total factor productivity: a firm perspective .............................. 215 Box 3.4 Debt, financial crises, and productivity.................................................... 223 Annex 3.1 Challenges of productivity measurement .............................................. 231 Annex 3.2 Data and growth accounting approach................................................. 232 Annex 3.3 Drivers of productivity....................................................................... 233 Annex 3.4 Data and methodology for sectoral productivity.................................... 239 Annex 3.5 Methodology for Box 3.3 ................................................................... 240 Annex 3.6 Local projection methodology for Box 3.4............................................ 240 References........................................................................................................ 241 VII Chapter 4 The Fourth Wave: Rapid Debt Buildup ............................................................... 251 Introduction ...................................................................................................... 253 Evolution of past waves of debt ............................................................................ 256 The current wave of debt in historical context ........................................................ 259 Rapid debt accumulation episodes ........................................................................ 276 What comes next? ............................................................................................... 278 Seven lessons ...................................................................................................... 282 Policy implications.............................................................................................. 283 Box 4.1 Similarities and differences between the previous three waves ....................... 260 Box 4.2 The fourth wave ..................................................................................... 266 Box 4.3 Debt and crises ....................................................................................... 272 Annex 4.1 Event study methodology..................................................................... 286 Annex 4.2 Regression methodology ...................................................................... 287 Annex 4.3 Case studies ........................................................................................ 288 References.......................................................................................................... 289 Statistical Appendix ....................................................................................................................... 299 Selected Topics ............................................................................................................................. 306 Figures 1.1 Global growth prospects ....................................................................... 5 1.2 Global risks and policy challenges .......................................................... 6 1.3 Advanced economies ............................................................................ 8 1.4 United States ...................................................................................... 9 1.5 Euro Area ........................................................................................... 9 1.6 China ............................................................................................... 10 1.7 Global trade ...................................................................................... 11 1.8 Global finance................................................................................... 12 1.9 Commodity markets .......................................................................... 13 1.10 EMDE recent developments ............................................................... 14 1.11 EMDE commodity exporters and importers ......................................... 15 1.1.1 Recent developments in low-income countries ...................................... 17 1.1.2 Outlook for per capita GDP and risks.................................................. 18 1.12 EMDE outlook ................................................................................. 21 1.13 EMDE per capita income growth and poverty ...................................... 22 VIII 1.2.1 Regional growth ................................................................................ 24 1.14 Balance of risks.................................................................................. 26 1.15 Rising trade barriers and protracted policy uncertainty ........................... 27 1.16 A deepening slowdown in major economies .......................................... 28 1.17 Financial stress in EMDEs .................................................................. 30 1.18 Other downside risks ......................................................................... 31 1.19 Upside risks ...................................................................................... 32 1.20 Monetary and financial policies in advanced economies ......................... 32 1.21 Fiscal policy in advanced economies..................................................... 33 1.22 Structural policies in advanced economies ............................................ 34 1.23 EMDE monetary and financial policy .................................................. 36 1.24 EMDE fiscal policy............................................................................ 37 1.25 EMDE structural policies—Governance, business climate, and GVC participation....................................................................... 39 1.26 EMDE structural policies—Productivity .............................................. 40 SF1.1 Price controls .................................................................................... 53 SF1.2 Price controls on imported and exported goods ..................................... 54 2.1.1 EAP: Recent developments ................................................................ 64 2.1.2 Recent developments, China............................................................... 65 2.1.3 EAP: Outlook and risks...................................................................... 66 2.1.1.1 Productivity in EAP compared with other country groups ...................... 70 2.1.1.2 Evolution of productivity in EAP ........................................................ 71 2.1.1.3 Factors underlying productivity growth in EAP..................................... 73 2.1.1.4 Prospects for productivity growth in EAP ............................................. 74 2.2.1 ECA: Recent developments................................................................. 78 2.2.2 ECA: Outlook and risks ..................................................................... 79 2.2.1.1 Productivity in ECA compared with other regions................................. 85 2.2.1.2 Evolution of productivity in ECA........................................................ 86 2.2.1.3 Factors supporting productivity growth in ECA .................................... 88 2.2.1.4 Drivers of productivity growth in ECA ................................................ 91 2.3.1 LAC: Recent developments................................................................. 96 2.3.2 LAC: Outlook and risks ..................................................................... 98 2.3.1.1 Evolution of labor productivity growth in LAC....................................103 2.3.1.2 Sources of productivity growth in LAC ...............................................105 2.3.1.3 Sectoral productivity in LAC .............................................................106 2.3.1.4 Drivers of labor productivity growth in LAC .......................................107 IX 2.4.1 MENA: Recent developments ............................................................112 2.4.2 MENA: Outlook and risks.................................................................113 2.4.1.1 Productivity in MENA in regional comparison ....................................118 2.4.1.2 Evolution of labor productivity in MENA ...........................................119 2.4.1.3 Factors supporting productivity growth in MENA ...............................120 2.4.1.4 Policy challenges...............................................................................121 2.5.1 SAR: Recent developments ................................................................126 2.5.2 SAR: Outlook and risks.....................................................................127 2.5.1.1 Evolution of productivity growth in SAR ............................................132 2.5.1.2 Sectoral productivity and employment in SAR .....................................134 2.5.1.3 Drivers of productivity growth in SAR ................................................135 2.5.1.4 Policy options in SAR .......................................................................136 2.5.1.5 Productivity prospects in SAR ............................................................137 2.5.1.6 Constraints to productivity growth in SAR ..........................................139 2.6.1 SSA: Recent developments.................................................................142 2.6.2 SSA: Outlook and risks .....................................................................144 2.6.1.1 Productivity in SSA in regional comparison .........................................149 2.6.1.2 Evolution of labor productivity growth in SSA.....................................150 2.6.1.3 Sectoral productivity growth in SSA...................................................152 2.6.1.4 Drivers of productivity growth in SSA ................................................154 2.6.1.5 Prospects for productivity growth in SSA ............................................156 SF2.1 Inflation in low-income countries and poverty .....................................180 SF2.2 Factors supporting falling inflation in LICs..........................................181 SF2.3 Monetary policy challenges in LICs ....................................................185 3.1 Labor productivity, per capita income and poverty reduction.................194 3.2 Global productivity developments ......................................................195 3.3 Evolution of global productivity growth ..............................................197 3.1.1 Evolution of regional labor productivity ..............................................199 3.1.2 Sectoral contributions to regional productivity growth ..........................201 3.1.3 Potential bottlenecks to productivity growth........................................203 3.4 Distribution of productivity levels and convergence progress .................205 3.5 Decomposition of productivity growth................................................207 3.2.1 Agriculture, industry and services .......................................................210 3.2.2 Sectoral labor productivity .................................................................211 3.2.3 Between– and within-sector sources to productivity growth...................212 3.6 Sectoral productivity developments.....................................................214 X 3.3.1 Firm TFP and distance-to-frontier in EMDEs by industry .................... 216 3.3.2 Firm TFP by regions......................................................................... 217 3.3.3 Distance-to-frontier of TFP, firm characteristics, and regulations ........... 218 3.7 Impact of drivers on productivity growth ............................................ 220 3.8 Pre-crisis developments in productivity drivers and productivity growth . 221 3.9 Post-crisis slowdown of the drivers of productivity growth .................... 222 3.4.1 Productivity in debt accumulation episodes and financial crises.............. 224 3.10 EMDE infrastructure and education gaps............................................ 226 3.11 Developments in Fintech and Govtech ............................................... 226 3.12 Productivity growth: reform scenario .................................................. 227 3.13 Effect of governance reform spurts...................................................... 230 Annex 3.3.1 Productivity drivers in 2017, by region ............................................. 234 Annex 3.3.2 Productivity changes in productivity drivers, by region ....................... 237 4.1 Evolution of debt ............................................................................. 254 4.2 Debt in EMDEs............................................................................... 257 4.3 The fourth wave: Debt accumulation.................................................. 259 4.1.1 Comparison of previous waves ........................................................... 261 4.1.2 Changes in debt by sector and region.................................................. 263 4.1.3 GDP per capita in EMDEs during the four waves ................................ 264 4.4 The fourth wave: Vulnerabilities and use of borrowed funds.................. 265 4.2.1 The fourth wave: Debt developments ................................................. 267 4.5 Comparison of features of fourth wave and earlier waves: Debt .............. 270 4.6 Comparison of fourth wave and earlier waves: Policies and institutions... 271 4.7 Episodes of rapid debt accumulation in EMDEs .................................. 277 4.8 Crises during rapid debt accumulation episodes in EMDEs ................... 278 4.9 Macroeconomic developments during debt accumulation episodes......... 279 4.10 Fourth wave: Opportunities and risks ................................................. 280 Annex 4.1.1 Country examples of debt accumulation episodes ............................... 286 Tables 1.1 Real GDP ...........................................................................................4 1.1.1 Low-income country forecasts ............................................................. 19 1.2 Emerging market and developing economies......................................... 41 2.1.1 East Asia and Pacific forecast summary ................................................. 68 2.1.2 East Asia and Pacific country forecasts.................................................. 68 2.2.1 Europe and Central Asia forecast summary ........................................... 82 XI 2.2.2 Europe and Central Asia country forecasts ............................................83 2.3.1 Latin America and the Caribbean forecast summary ............................. 100 2.3.2 Latin America and the Caribbean country forecasts.............................. 101 2.4.1 Middle East and North Africa forecast summary.................................. 115 2.4.2 Middle East and North Africa economy forecasts................................. 116 2.5.1 South Asia forecast summary ............................................................. 129 2.5.2 South Asia country forecasts .............................................................. 130 2.6.1 Sub-Saharan Africa forecast summary ................................................. 146 2.6.2 Sub-Saharan Africa country forecasts.................................................. 147 Annex 3.3.1 Variables included in the regressions and sources ............................... 238 Annex 3.4.1 Sectoral categorization .................................................................... 239 Annex 4.1.1 Comparison of debt accumulation episodes....................................... 287 XII Foreword Following a year during which weak trade and from the reallocation of resources to more investment dragged the world economy to its productive sectors, and slowing improvements in feeblest performance since the global financial crisis, the key drivers of productivity have sapped economic growth is poised for a modest rebound momentum in this key driver of lasting growth. this year. However, for even that modest uptick to occur, many things have to go right. Additional key themes explored in this edition include price controls—which, despite good Global growth is set to rise by 2.5 percent this year, intentions, can dampen investment and growth, a small rise from an estimated 2.4 percent in 2019, worsen poverty outcomes, and lead to heavier fiscal as trade and investment gradually recover. burdens—and the drivers of the long recent period of low inflation among low-income countries and Emerging market and developing economies are necessary policies to maintain low and stable anticipated to see growth accelerate to 4.1 percent inflation. from 3.5 percent last year. However, that acceleration will not be broad-based: the pickup These messages have serious implications for the is anticipated to come largely from a handful of goals of eradicating poverty and sharing prosperity. large emerging economies stabilizing after deep Even if the recovery in emerging and developing recessions or sharp slowdowns. economy growth were to take place as expected, per capita growth would advance at a pace too slow to Even this tepid global rally could be disrupted by meet development goals. any number of threats. Trade tensions could re-escalate. A sharper-than-expected growth slow- Yet policymakers have it in their capacity to ensure down in major economies would reverberate the recovery not only stays on track, but even widely. A resurgence of financial stress in large surprises to the upside. Recent policy actions— emerging markets, an escalation of geopolitical particularly those that have mitigated trade tensions, or a series of extreme weather events could tensions—could augur a sustained reduction in all have adverse effects on economic activity. policy uncertainty. Countries could pursue decisive reforms to bolster governance and business climate, This edition of Global Economic Prospects analyzes improve tax policy, promote trade integration, and several topical themes underlying the fragile rekindle productivity growth, all while protecting outlook. vulnerable groups. Building resilient monetary and fiscal frameworks, instituting robust supervisory One is the largest, fastest, and most broad-based and regulatory regimes, and following transparent wave of debt accumulation in advanced economies debt management practices could reduce the risk of as well as in emerging and developing economies in shocks, or soften their impact, and strengthen the last 50 years. Public borrowing can be beneficial resilience against them. and spur economic development, if used to finance growth-enhancing investments. However, As a philosopher once said, one swallow does not a although currently low interest rates mitigate risks, summer make. There are signs that global growth the three previous waves of debt accumulation in skirted a rough patch and is recovering; it is up to debt have ended badly. policy makers to make sure it thrives. A second is the widespread slowdown in productivity growth over the last ten years. Growth Ceyla Pazarbasioglu in productivity—output per worker—is essential to Vice President raising living standards. However, weaker Equitable Growth, Finance, and Institutions investment and efficiency gains, dwindling gains World Bank Group XIII Acknowledgments This World Bank Group Flagship Report is a product of the Prospects Group in the Equitable Growth, Finance and Institutions (EFI) Vice Presidency. The project was managed by M. Ayhan Kose and Franziska Ohnsorge, under the general guidance of Ceyla Pazarbasioglu. Global and regional surveillance work was Gopalakrishnan, and Torie Smith. Mark coordinated by Carlos Arteta. The primary Felsenthal and Alejandra Viveros managed media authors of this report were Alistair Dieppe, relations and dissemination. Graeme Littler Justin-Damien Guénette, Jongrim Ha, Gene provided editorial support, with contributions Kindberg-Hanlon, Patrick Kirby, Peter Nagle, from Adriana Maximiliano. Rudi Steinbach, Naotaka Sugawara, Temel Taskin, Ekaterine Vashakmadze, Dana Vorisek, Regional projections and write-ups were Collette M. Wheeler, and Lei Sandy Ye. produced in coordination with country teams, country directors, and the offices of the regional Other contributors included John Baffes, Csilla chief economists. Lakatos, Alain Kabundi, Sergiy Kasyanenko, Atsushi Kawamoto, Sinem Kilic Celik, Wee The print publication was produced by Maria Chian Koh, Hideaki Matsuoka, Yoki Okawa, Hazel Macadangdang, Adriana Maximiliano, and Cedric Okou, Franz Ulrich Ruch, and Shu Yu. Quinn Sutton, in collaboration with Luiz H. Almeida, Andrew Charles Berghauser, Aziz Research assistance was provided by Vanessa Gökdemir, Michael Harrup, and Jewel Arellano Banoni, Yushu Chen, Zhuo Chen, McFadden. Khamal Antonio Clayton, Aygul Evdokimova, Awais Khuhro, Yi Li, Shihui Liu, Maria Hazel Many reviewers provided extensive advice and Macadangdang, Julia R. R. Norfleet, Vasiliki comments. The analysis also benefited from Papagianni, Jankeesh Sandhu, Shijie Shi, Xinyue comments and suggestions by staff members Wang, Jinxin Wu, Heqing Zhao, and Juncheng from World Bank Group country teams and Zhou. Modeling and data work were provided by other World Bank Group Vice Presidencies as Rajesh Kumar Danda, Julia R. R. Norfleet, and well as Executive Directors in their discussion of Shijie Shi. the report on December 17, 2019. However, both forecasts and analysis are those of the World The online publication was produced by Paul Bank Group staff and should not be attributed to Blake, Graeme Littler, Venkat Ganeshan Executive Directors or their national authorities. XV Executive Summary Global growth is projected to reach 2.5 percent in 2020, slightly faster than the post-crisis low registered last year. While growth could be stronger if reduced trade tensions lead to a sustained reduction in uncertainty, the balance of risks to the outlook is to the downside. Growth in emerging market and developing economies (EMDEs) is also expected to remain subdued, continuing a decade of disappointing outcomes. A steep and widespread productivity growth slowdown has been underway in EMDEs since the global financial crisis, despite the largest, fastest, and most broad-based accumulation of debt since the 1970s. In addition, many EMDEs, including low-income countries, face the challenge of phasing out price controls that impose heavy fiscal burdens and dampen investment. These circumstances add urgency to the need to implement measures to rebuild macroeconomic policy space and to undertake reforms to rekindle productivity growth. These efforts need to be supplemented by policies to promote inclusive long-term growth and accelerate poverty alleviation. Global Outlook: Fragile, Handle with Care. but the recovery will largely depend on a rebound Global growth is expected to recover to 2.5 in a small number of large EMDEs, some of percent in 2020—up slightly from the post-crisis which are emerging from deep recessions or sharp low of 2.4 percent registered last year amid slowdowns. weakening trade and investment—and edge up further over the forecast horizon. This projected This edition of Global Economic Prospects also recovery could be stronger if recent policy includes chapters on the productivity growth actions—particularly those that have mitigated slowdown in EMDEs since the global financial trade tensions—lead to a sustained reduction in crisis and on the rapid debt buildup in these policy uncertainty. Nevertheless, downside risks economies over the same period, and special focus predominate, including the possibility of a re- pieces on the implications of price controls in escalation of global trade tensions, sharp EMDEs and on the challenges of maintaining downturns in major economies, and financial low inflation in low-income economies (LICs). disruptions in emerging market and developing economies (EMDEs). The materialization of Fading Promise: How to Rekindle Productivity these risks would test the ability of policymakers Growth. A broad-based slowdown in labor to respond effectively to negative events. productivity growth has been underway since the Associated policy challenges are compounded by global financial crisis. In EMDEs, the slowdown high debt levels and subdued productivity has reflected weakness in investment and growth. Many EMDEs need to rebuild moderating efficiency gains as well as dwindling macroeconomic policy space to enhance resilience resource reallocation between sectors. The pace of to possible adverse developments. They also need improvements in key drivers of labor to pursue decisive reforms to bolster governance productivity—including education, urbanization, and business climates, improve tax policy, and institutions—has slowed or stagnated since promote trade integration, and rekindle the global financial crisis and is expected to productivity growth, while protecting vulnerable remain subdued. To rekindle productivity groups. These policy actions would help foster growth, a comprehensive approach is necessary: inclusive and sustainable long-term growth and facilitating investment in physical, intangible, and poverty alleviation. human capital; encouraging reallocation of resources towards more productive sectors; Regional Prospects. Growth in almost all EMDE fostering firm capabilities to reinvigorate regions has been weaker than expected, reflecting technology adoption and innovation; and downgrades to almost half of EMDEs. Activity in promoting a growth-friendly macroeconomic and most regions is expected to pick up in 2020-21, institutional environment. Specific policy XVII priorities will depend on individual country effective conduct of monetary policy. Replacing circumstances. price controls with expanded and better-targeted social safety nets, coupled with reforms to The Fourth Wave: Rapid Debt Buildup. The encourage competition and a sound regulatory global economy has experienced four waves of environment, can be both pro-poor and pro- debt accumulation over the past fifty years. The growth. Such reforms need to be carefully first three ended with financial crises in many communicated and sequenced to ensure political EMDEs. During the current wave, which started and social acceptance. Where they exist, price in 2010, the increase in debt in these economies control regimes should be transparent and has already been larger, faster, and more broad- supported by well-capitalized stabilization funds based than in any of the previous three waves. or national hedging strategies to ensure fiscal Current low interest rates—which markets expect sustainability. to be sustained into the medium term—appear to mitigate some of the risks associated with high Low for How Much Longer? Inflation in Low- debt. However, EMDEs are also confronted by Income Countries. Inflation in LICs has declined weak growth prospects, mounting vulnerabilities, sharply to a median of 3 percent in mid-2019 and elevated global risks. A menu of policy from a peak of 25 percent in 1994. The drop has options is available to reduce the likelihood of the been supported by the move to more flexible current debt wave ending in crises and, if crises exchange rate regimes, greater central bank were to take place, to alleviate their impact. independence, and a generally more benign external environment since the 1990s. However, Price Controls: Good Intentions, Bad low LIC inflation cannot be taken for granted Outcomes. The use of price controls is amid mounting fiscal pressures and the risk of widespread across EMDEs, including for food exchange rate shocks. To maintain low and stable and key imported and exported commodities. inflation, monetary and fiscal policy frameworks While sometimes used as a tool for social policy, need to be strengthened and supported by efforts price controls can dampen investment and to replace price controls with more efficient growth, worsen poverty outcomes, cause countries policies. to incur heavy fiscal burdens, and complicate the XVIII Abbreviations ACP African, Caribbean and Pacific Group of States AE advanced economy CDS credit default swap CPTPP Comprehensive and Progressive Agreement for Trans-Pacific Partnership DR-CAFTA Central America-Dominican Republic Free Trade Agreement EAP East Asia and Pacific ECA Europe and Central Asia ECB European Central Bank ECI Economic Complexity Index EMBI Emerging Market Bond Index EMDE emerging market and developing economies ERPTR exchange rate pass-through ratio EU European Union FAVAR Bayesian factor-augmented vector autoregression FCV fragility, conflict, and violence FDI foreign direct investment GCC Gulf Cooperation Council GDP gross domestic product GEP Global Economic Prospects GMM generalized methods of moments GNFS goods and nonfactor services GNI gross national income GST goods and services tax GVCs global value chains HIPC Heavily Indebted Poor Countries ICE Intercontinental Exchange ICRG International Country Risk Guide IMF International Monetary Fund IT inflation targeting LAC Latin America and the Caribbean LFPR labor force participation rate LIC low-income country LSAP Large-Scale Asset Purchase MDRI Multilateral Debt Relief Initiative MENA Middle East and North Africa MEP Maturity Extension Program MIC middle-income country XIX NEER nominal effective exchange rate NPL nonperforming loan ONI Oceanic Niño Index OPEC Organization of the Petroleum Exporting Countries PMI Purchasing Managers’ Index PPP purchasing power parity REER real effective exchange rate RHS right-hand side (in figures) RMB renminbi SAR South Asia Region SSA Sub-Saharan Africa SSE Shanghai Stock Exchange TFP total factor productivity TiVA trade in value added USMCA United States-Mexico-Canada Agreement VAT value-added tax WAEMU West African Economic and Monetary Union WGI World Governance Indicators WTO World Trade Organization XX CHAPTER 1 GLOBAL OUTLOOK Fragile, Handle with Care G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 3 Global growth is expected to recover to 2.5 percent in 2020—up slightly from the post-crisis low of 2.4 percent registered last year amid weakening trade and investment—and edge up further over the forecast horizon. This projected recovery could be stronger if recent policy actions—particularly those that have mitigated trade tensions—lead to a sustained reduction in policy uncertainty. Nevertheless, downside risks predominate, including the possibility of a re-escalation of global trade tensions, sharp downturns in major economies, and financial disruptions in emerging market and developing economies (EMDEs). The materialization of these risks would test the ability of policymakers to respond effectively to negative events. Associated policy challenges are compounded by high debt levels and subdued productivity growth. Many EMDEs need to rebuild macroeconomic policy space to enhance resilience to possible adverse developments. They also need to pursue decisive reforms to bolster governance and business climates, improve tax policy, promote trade integration, and rekindle productivity growth, while protecting vulnerable groups. These policy actions would help foster inclusive and sustainable long-term growth and poverty alleviation. Summary which has heightened policy uncertainty and weighed on international trade, confidence, and Global growth decelerated markedly in 2019, with investment. As a result of the increase of tariffs continued weakness in global trade and between the two countries over the past couple of investment (Figures 1.1.A and 1.1.B). This years, a substantially higher share of world trade weakness was widespread, affecting both advanced has become subject to protectionist measures economies—particularly the Euro Area—and (Figure 1.1.D). emerging market and developing economies Financial market sentiment improved appreciably (EMDEs). Various key indicators of economic toward the end of last year along with the activity declined in parallel, approaching their alleviation of trade tensions. That said, it had been lowest levels since the global financial crisis fragile for most of 2019. Concerns about growth (Figure 1.1.C). In particular, global trade in goods prospects triggered widespread monetary policy was in contraction for a significant part of 2019, easing by major central banks last year, as well as and manufacturing activity slowed markedly over flight to safety flows into advanced-economy bond the course of the year; recent high-frequency markets. In a context of subdued inflation, this readings suggest some tentative stabilization of pushed global yields down—in some advanced manufacturing output at weak levels. To a lesser economies, further into negative territory—for extent, services activity also moderated. A broad most of 2019. Heightened risk aversion range of economies have experienced feeble contributed to subdued EMDE capital inflows in growth, with close to 90 percent of advanced the second half of last year, as a number of economies and 60 percent of EMDEs going EMDEs faced renewed currency and equity price through varying degrees of deceleration last year. pressures. The subdued outlook led to declines in Bilateral negotiations between the United States most commodity prices, which are expected to and China since mid-October resulted in a Phase remain near current levels over the forecast period. One agreement—including a planned partial Against this international context, global growth rollback of tariffs—that has de-escalated trade weakened to an estimated 2.4 percent last year— tensions. This comes after a prolonged period of the lowest rate of expansion since the global rising trade disputes between the two countries, financial crisis. With some recent data pointing to an incipient stabilization of economic conditions, global growth is projected to edge up to 2.5 Note: This chapter was prepared by Carlos Arteta and Patrick Kirby, with contributions from Collette M. Wheeler, Justin-Damien percent in 2020, 0.2 percentage point below Guénette, Csilla Lakatos, Rudi Steinbach, and Ekaterine previous forecasts, as investment and trade Vashakmadze. Additional inputs were provided by John Baffes, gradually recover. In particular, global trade Sergiy Kasyanenko, Peter Nagle, and Franz Ulrich Ruch. Research assistance was provided by Yushu Chen, Shihui Liu, Julia Norfleet, growth—which is estimated to have slowed Vasiliki Papagianni, Shijie Shi, and Jinxin Wu. sharply from 4 percent in 2018 to 1.4 percent in 4 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 TABLE 1.1 Real GDP1 Percentage point differences (Percent change from previous year) from June 2019 projections 2017 2018 2019e 2020f 2021f 2022f 2019e 2020f 2021f World 3.2 3.0 2.4 2.5 2.6 2.7 -0.2 -0.2 -0.2 Advanced economies 2.4 2.2 1.6 1.4 1.5 1.5 -0.1 -0.1 0.0 United States 2.4 2.9 2.3 1.8 1.7 1.7 -0.2 0.1 0.1 Euro Area 2.5 1.9 1.1 1.0 1.3 1.3 -0.1 -0.4 0.0 Japan 1.9 0.8 1.1 0.7 0.6 0.4 0.3 0.0 0.0 Emerging market and developing economies 4.5 4.3 3.5 4.1 4.3 4.4 -0.5 -0.5 -0.3 Commodity-exporting EMDEs 2.2 2.0 1.5 2.6 2.9 3.0 -0.6 -0.5 -0.1 Other EMDEs 6.2 5.8 4.8 5.1 5.2 5.2 -0.4 -0.4 -0.3 Other EMDEs excluding China 5.4 5.0 3.3 4.0 4.4 4.5 -0.9 -0.8 -0.6 East Asia and Pacific 6.5 6.3 5.8 5.7 5.6 5.6 -0.1 -0.2 -0.2 China 6.8 6.6 6.1 5.9 5.8 5.7 -0.1 -0.2 -0.2 Indonesia 5.1 5.2 5.0 5.1 5.2 5.2 -0.2 -0.2 -0.1 Thailand 4.0 4.1 2.5 2.7 2.8 2.9 -1.0 -0.9 -0.9 Europe and Central Asia 4.1 3.2 2.0 2.6 2.9 2.9 0.4 -0.1 0.0 Russia 1.6 2.3 1.2 1.6 1.8 1.8 0.0 -0.2 0.0 Turkey 7.5 2.8 0.0 3.0 4.0 4.0 1.0 0.0 0.0 Poland 4.9 5.1 4.3 3.6 3.3 3.1 0.3 0.0 0.0 Latin America and the Caribbean 1.9 1.7 0.8 1.8 2.4 2.6 -0.9 -0.8 -0.3 Brazil 1.3 1.3 1.1 2.0 2.5 2.4 -0.4 -0.5 0.2 Mexico 2.1 2.1 0.0 1.2 1.8 2.3 -1.7 -0.8 -0.6 Argentina 2.7 -2.5 -3.1 -1.3 1.4 2.3 -1.9 -3.5 -1.8 Middle East and North Africa 1.1 0.8 0.1 2.4 2.7 2.8 -1.2 -0.8 0.0 Saudi Arabia -0.7 2.4 0.4 1.9 2.2 2.4 -1.3 -1.2 -0.1 Iran 3.8 -4.9 -8.7 0.0 1.0 1.0 -4.2 -0.9 0.0 Egypt2 4.2 5.3 5.6 5.8 6.0 6.0 0.1 0.0 0.0 South Asia 6.7 7.1 4.9 5.5 5.9 6.0 -2.0 -1.5 -1.2 India3 7.2 6.8 5.0 5.8 6.1 6.1 -2.5 -1.7 -1.4 Pakistan2 5.2 5.5 3.3 2.4 3.0 3.9 -0.1 -0.3 -1.0 Bangladesh2 7.3 7.9 8.1 7.2 7.3 7.3 0.8 -0.2 0.0 Sub-Saharan Africa 2.7 2.6 2.4 2.9 3.1 3.3 -0.5 -0.4 -0.4 Nigeria 0.8 1.9 2.0 2.1 2.1 2.1 -0.1 -0.1 -0.3 South Africa 1.4 0.8 0.4 0.9 1.3 1.5 -0.7 -0.6 -0.4 Angola -0.1 -1.2 -0.7 1.5 2.4 3.0 -1.7 -1.4 -0.4 Memorandum items: Real GDP1 High-income countries 2.4 2.2 1.7 1.5 1.5 1.6 -0.1 -0.1 -0.1 Developing countries 4.8 4.4 3.7 4.3 4.5 4.5 -0.4 -0.4 -0.3 Low-income countries 5.5 5.8 5.4 5.4 5.5 5.8 -0.3 -0.6 -0.6 BRICS 5.3 5.4 4.6 4.9 4.9 5.0 -0.5 -0.4 -0.4 World (2010 PPP weights) 3.9 3.7 2.9 3.2 3.3 3.4 -0.4 -0.3 -0.3 World trade volume4 5.9 4.0 1.4 1.9 2.5 2.8 -1.2 -1.3 -0.7 Commodity prices5 Oil price 23.3 29.4 -10.3 -5.4 1.9 1.9 -6.9 -3.9 1.2 Non-energy commodity price index 5.5 1.7 -4.7 0.1 1.7 1.7 -2.6 0.2 0.3 Source: World Bank. Note: PPP = purchasing power parity; e = estimate; f = forecast. World Bank forecasts are frequently updated based on new information. Consequently, projections presented here may differ from those contained in other World Bank documents, even if basic assessments of countries’ prospects do not differ at any given moment in time. Country classifications and lists of emerging market and developing economies (EMDEs) are presented in Table 1.2. BRICS include: Brazil, Russia, India, China, and South Africa. The World Bank has ceased producing a growth forecast for Venezuela and has removed Venezuela from all growth aggregates in which it was previously included. 1. Headline aggregate growth rates calculated using GDP weights at 2010 prices and market exchange rates. World growth rates based on purchasing power parity (PPP) weights attribute a greater portion of global GDP to EMDEs relative to market exchange rates due to the PPP methodology, which uses an exchange rate that is calculated from the difference in the price levels of a basket of goods and services between economies. 2. GDP growth values are on a fiscal year basis. Aggregates that include these countries are calculated using data compiled on a calendar year basis. Pakistan's growth rates are based on GDP at factor cost. The column labeled 2019 refers to FY2018/19. 3. The column labeled 2018 refers to FY2018/19. 4. World trade volume of goods and non-factor services. 5. Oil is the simple average of Brent, Dubai, and West Texas Intermediate. The non-energy index is comprised of the weighted average of 39 commodities (7 metals, 5 fertilizers, 27 agricultural commodities). For additional details, please see http://www.worldbank.org/en/research/commodity-markets. Click here to download data. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 5 2019, by far the weakest pace since the global FIGURE 1.1 Global growth prospects financial crisis—is projected to firm throughout Global growth decelerated last year to 2.4 percent—its slowest pace since 2020 and reach 1.9 percent, assuming trade the global financial crisis—amid weakening trade and investment. Key tensions do not re-escalate. In the near term, indicators deteriorated in parallel, in part reflecting heightened trade protectionism. While monetary accommodation has increased, fiscal monetary policy across the world is generally support is expected to wane. Global growth is projected to recover to 2.5 expected to remain accommodative; however, percent in 2020 and edge further up thereafter as trade and investment firm fiscal policy support is likely to fade (Figure and EMDE activity rebounds; however, per capita growth in EMDEs will remain insufficient to meet poverty alleviation goals. 1.1.E). A. Global growth B. Global trade, investment, and Near-term projections for global growth mask consumption growth different contours in advanced economies and EMDEs. Growth in advanced economies is projected to slow to 1.4 percent this year—below previous projections, in part reflecting lingering weakness in manufacturing—and improve slightly over the rest of the forecast horizon. In contrast, after decelerating to an estimated weaker-than-expected 3.5 percent last year, growth in EMDEs is projected to increase to 4.1 percent C. Global indicators of activity in 2019 D. Global trade subject to new protectionist measures in 2020—0.5 percentage point below previous forecasts, reflecting downgrades to half of EMDEs due in part to downward revisions to trade and investment growth. Nonetheless, the recovery in aggregate EMDE growth this year—which assumes continued monetary policy support in many economies, no major swings in commodity prices, and generally benign borrowing costs—is not envisioned to be broad-based: About a third of EMDEs are expected to decelerate. Instead, it is E. Stance of global fiscal and F. Per capita income growth monetary policy largely predicated on a rebound in a small number of large EMDEs, most of which are emerging from deep recessions or sharp slowdowns but remain fragile. Excluding this group of countries, there would be almost no acceleration in EMDE growth this year—and, with advanced economies slowing, global growth would actually decelerate. Going forward, EMDE growth is projected to Source: Bank for International Settlements; Consensus Economics; CPB Netherlands Bureau for stabilize at an average of 4.4 percent in 2021-22, Economic Policy Analysis; Haver Analytics; International Monetary Fund; World Bank; World Trade as trade and investment firm. In low-income Organization. Note: AEs = advanced economies; EMDEs = emerging market and developing economies. countries, growth is expected to remain little A.B.E. Shaded areas indicate forecasts. Data for 2019 are estimates. B.C. Trade measured as the average of export and import volumes. changed at 5.4 percent in 2020 and edge up to an A. Aggregate growth rates calculated using GDP weights at 2010 prices and market exchange rates. average of 5.7 percent later in the forecast horizon, B. Data for 2015-16 are simple averages. Green lines indicate average over period 1990-2018. C. Manu. = manufacturing. PMI = Purchasing Managers’ Index. PMI readings above 50 indicate boosted by increased investment in infrastructure expansion in economic activity; readings below 50 indicate contraction. Last observation is 2019Q3 for GDP, October 2019 for industrial production and goods trade, and November 2019 for PMI. and rebuilding efforts in some countries following D. Figure includes new import-restrictive measures, including tariff and non-tariff trade barriers. extreme weather-related devastation. Annual data are mid-October to mid-October. E. Aggregates calculated using nominal U.S. dollar GDP weights. Fiscal impulse is the negative change in general government cyclically adjusted primary balance. Policy rates are the December to December change. Sample includes 35 AEs and 77 EMDEs for fiscal impulse and 16 AEs and 21 Even if the recovery in EMDE growth proceeds as EMDEs for policy rates. Policy rates for 2020 use the December 2019 Consensus Forecasts report for expected, per capita growth will remain well below central bank policy rates. When these are unavailable, the change in short-term yields is used. F. EAP = East Asia and Pacific, ECA = Europe and Central Asia, LAC = Latin America and the long-term averages and far from sufficient to meet Caribbean, MNA = Middle East and North Africa, SAR = South Asia, SSA = Sub-Saharan Africa. Long-term average is calculated over the period 2000-19. Poverty rates represent latest data. Click here to download data and charts. 6 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 1.2 Global risks and policy challenges poverty alleviation goals. More specifically, Current projections represent a benign but fragile outlook given ongoing income growth will be slowest in Sub-Saharan global headwinds. Downside risks predominate and increase the likelihood Africa—the region where most low-income of much weaker-than-expected global growth. However, recent policy countries are clustered and most of the world’s actions that have reduced trade tensions could lead to a sustained mitigation of policy uncertainty and bolster investment. In advanced poor live (Figure 1.1.F). economies, the room for monetary accommodation is limited. In EMDEs, fiscal space is constrained by weak tax capacity and high debt levels, The near-term forecast for a pickup in EMDE which also hinders the ability to fund basic public services. Boosting EMDE productivity, which has been on a downward trend in recent years, growth represents a benign, but fragile, scenario is essential to foster long-term growth and poverty reduction. given ongoing global headwinds such as slowing advanced-economy growth, subdued global trade, A. Average share of EMDEs with B. Probability of global growth being 1 and moderating commodity prices (Figure 1.2.A). annual growth accelerating by more percentage point below baseline than 0.1 percentage point, 1962-2019 More generally, a deeper global downturn could result if global trade tensions re-emerge, policy uncertainty persists and becomes entrenched, or activity in major economies deteriorates significantly. Other risks include financial stress in large EMDEs, heightened geopolitical tensions, or a higher incidence of extreme weather events. Amid these downside risks, the probability that global growth in 2020 will be below baseline projections is above its historical average (Figure C. Impact of a 10-percent decrease in D. Monetary policy rate increases U.S. policy uncertainty on investment during current and previous 1.2.B). That said, the projected recovery could be growth expansions stronger than expected if recent policy actions— particularly those that have alleviated U.S.-China trade tensions—lead to a sustained reduction in policy uncertainty and bolster confidence, trade, and investment (Figure 1.2.C). Against the backdrop of a fragile outlook, the policy challenges confronting the global economy are compounded by subdued productivity growth E. Share of EMDEs with limited tax F. Productivity growth and high levels of debt (Chapters 3 and 4). In revenues to fund basic public advanced economies, the weakness of the current services expansion has made it difficult for central banks to create room for additional easing (Figure 1.2.D). Low global interest rates and the associated reduction in debt service burdens may provide some countries with additional flexibility for the implementation of structural reforms, such as investments in public infrastructure or the adoption of other growth-friendly policies. In Source: Baker, Bloom, and Davis (2016); Bloomberg; Haver Analytics; International Monetary Fund; addition, governments can create further fiscal National Bureau of Economic Research; Penn World Table; The Conference Board; World Bank. space through better tax compliance and A.F. Aggregates calculated using GDP weights at 2010 prices and market exchange rates. A. AE = advanced economies. “Subdued trade” refers to growth below 2.5 percent. “Moderating enforcement. commodity prices” refers to a year-on-year contraction in the non-energy commodity index. B. Probabilities computed from the forecast distributions of 12- and 24-month-ahead oil price futures; S&P 500 futures, and term spread forecasts. Risk factor weights are derived from the model described in Ohnsorge, Stocker, and Some (2016). Last observation is December 19, 2019. Most EMDEs are not well positioned to confront C. Figure shows median impact. See Annex SF.1B of World Bank (2017a) for methodology. negative shocks, since policy buffers generally D. U.S. expansions: 1991-2001, 2001-07, 2009-present. Euro Area expansions: 1999-2008, 2009-11, 2013-present. Calculations based on trough and peak of policy rates of each period. Last observation remain inadequate. While moderating inflation is November 2019 for the United States and 2019Q3 for the Euro Area. E. Revenue threshold needed to provide basic public services is 15 percent of GDP, per Gaspar, has allowed many EMDEs to cut policy rates to Jaramillo, and Wingender (2016). Unbalanced sample includes 70 EMDEs, of which 11 are LICs. F. Figure shows 5-year moving averages. Productivity is defined as output per worker. Sample support growth, underlying price pressures are includes 74 EMDEs and 29 advanced economies. Refer to Chapter 3 for details. Click here to download data and charts. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 7 building in some cases, and policy space would be encourage the reallocation of resources toward further eroded in the event of renewed financial more productive sectors, reinvigorate technology market pressures. Many EMDEs, including LICs, adoption and innovation among firms, and face the additional challenge of phasing out price promote a growth-friendly macroeconomic and controls and their associated distortions amid institutional environment. Within this four- moderate inflation (Special Focus 1 and Special pronged approach, specific policy priorities will Focus 2). depend on country circumstances. In addition, investments in green infrastructure can also help Although fiscal accommodation in some EMDEs achieve development goals and improve resilience may be warranted in response to adverse to climate change. developments, record-high debt levels and fragile public finance positions limit the ability to implement countercyclical policy—indeed, a large Major economies: Recent share of EMDEs, particularly LICs, do not even developments and outlook have the capacity to adequately fund basic public services (Figure 1.2.E; Chapter 4). If faced with In major economies, activity has slowed more negative shocks, authorities would need to ensure markedly than previously expected. Very weak that any fiscal support prioritizes growth- manufacturing activity has dampened growth in enhancing spending and domestic revenue advanced economies, and policy uncertainty mobilization to avoid further erosion of public associated with trade tensions has also weighed on debt sustainability. Tax policy reforms that activity in the United States and China. broaden the revenue base are needed to fund investment, which could be complemented by e growth forecast for advanced economies has measures that help reduce inequality. again been revised down as a consequence of weaker-than-expected trade and manufacturing EMDE policymakers also need to pursue decisive activity (Figure 1.3.A). Recent data show structural reforms, while protecting vulnerable particular weakness in investment and exports, groups, to promote inclusive long-term growth. particularly in the Euro Area. is, along with Policy actions that improve EMDE governance below-target in ation in many economies, has frameworks and business climates, and facilitate prompted a broad shift toward monetary policy integration in existing supply chains or spur the easing (Figure 1.3.B). Labor markets and the creation of new ones, could help counter the services sector generally remain more resilient, but adverse effects of weak global growth and subdued the latter has shown signs of moderation (Figure international trade (World Bank 2019a). Measures 1.3.C). Aggregate activity is expected to edge to improve connectivity, lower trade costs, and down in 2020, with continued softness in ensure a stable and predictable legal environment investment and trade (Figure 1.3.D). could facilitate this integration. A strong and stable multilateral trading system remains an United States important foundation for robust growth in EMDEs. Growth has decelerated amid slowing investment and exports (Figure 1.4.A). Notwithstanding the The downward trend in EMDE productivity recent trade deal with China, rising tariffs have growth in recent years complicates these policy increased trade costs, while policy uncertainty has challenges (Figure 1.2.F; Chapter 3). Measures to weighed on investment and confidence (Baker, boost EMDE productivity growth are essential to Bloom, and Davis 2016; Fajgelbaum et al. 2019). foster potential growth and ensure continued As in many other advanced economies, the U.S. progress in improving living standards and manufacturing sector has been very weak. Support alleviating poverty. To rekindle productivity from tax cuts and changes in government growth, a comprehensive approach needs to be spending is expected to fade this year and become employed involving policies that facilitate a drag on growth thereafter (Figure 1.4.B; IMF investment in physical and human capital, 2019a). 8 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 1.3 Advanced economies impacts of lingering uncertainty and a waning The growth forecast for advanced economies has been steadily revised contribution from tax cuts and government down, prompting a general shift toward monetary policy easing. Services spending, which are only partly o set by activity has so far been more resilient than investment and trade, but it has accommodative monetary policy. e forecast is also moderated. Activity is expected to edge down in 2020, with continued softness in investment and trade. predicated on tari s staying at planned levels, scal policy progressing as currently legislated, and A. Evolution of the growth forecast for B. Monetary policy in advanced the heightened degree of policy uncertainty advanced economies economies gradually dissipating. Additional progress in U.S.- China trade negotiations that leads to a further reduction in trade policy uncertainty could result in higher-than-expected U.S. growth. Euro Area Economic activity in the Euro Area has deteriorated signi cantly. Several economies were on the verge of recession at some point last year, C. Services sector expectations D. Advanced-economy trade, with particular weakness in the German industrial investment, and consumption growth sector as it struggled with falling demand from Asia and disruptions to car production (Figures 1.5.A and 1.5.B). Uncertainty concerning Brexit also weighed on growth. e ECB has provided monetary stimulus by pushing its policy rate deeper into negative territory, restarting quantitative easing, and providing inexpensive credit to banks. e overall Source: Bank for International Settlements; Haver Analytics; World Bank. scal position of the Euro Area is expected to be A. Aggregate growth rates calculated using GDP weights at 2010 prices and market exchange rates. Blue bars and orange diamonds denote forecasts in the January 2019, June 2019, and January 2020 roughly balanced over the forecast period, editions of the Global Economic Prospects report. B. Aggregate nominal policy rates calculated using moving real GDP weights at 2010 prices and providing little additional support to activity market exchange rates. Sample includes 15 advanced economies. Last observation is November 2019. despite existing space in some economies. C. Figure shows 3-month moving averages of PMI service business expectations for the year ahead. PMI readings above 50 indicate expansion in economic activity; readings below 50 indicate contraction. Last observation is November 2019. Growth is expected to slow to 1 percent in 2020, D. Trade is the average of import and export volumes. Data for 2015-16 are simple averages. Long-term average calculated over the period from 1990-2018. Shaded area indicates forecasts. 0.4 percentage point down from previous Click here to download data and charts. projections due to worse-than-expected incoming data, especially industrial production. Growth is forecast to recover modestly to an average of 1.3 Despite these headwinds, the labor market percent in 2021-22, assuming that policy support remains robust and has bene ted from a rising gains traction, the Brexit process unfolds with participation rate. Unemployment is near a ve- minimal disruption, and there is no further decade low and wage growth has been solid, escalation in trade restrictions. fueling resilient consumption. Concerns about the global outlook and persistent below-target Japan in ation have resulted in the Federal Reserve cutting its policy rates by 75 basis points since Activity in Japan declined sharply following the mid-2019. impact of Typhoon Hagibis and the increase in the value-added tax (VAT) in October last year. Growth is expected to slow over the course of the e economy is also su ering from acute weakness forecast period, from 2.3 percent in 2019 to 1.8 in manufacturing and exports, particularly those percent in 2020 and 1.7 percent in 2021-22. In to China, alongside declining consumer the near term, the slowdown re ects the negative con dence. In response, the government is G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 9 providing signi cant support. Despite recent FIGURE 1.4 United States weakness in activity, the unemployment rate Growth has decelerated, reflecting slowing investment and exports. While remains near multidecade lows, labor force the labor market remains robust, manufacturing activity has been participation continues to climb, and per capita contracting, higher tariffs have increased trade costs, and policy uncertainty has continued to weigh on investment. Support from tax cuts income growth remains healthy. and government spending is expected to fade. Growth is expected to slow from 1.1 percent in A. Selected activity indicators B. Change in the general government cyclically-adjusted primary deficit 2019 to 0.7 percent in 2020, as anticipatory purchases prior to the VAT increase in October 2019 are unwound. Growth in 2021-22 is expected to average about 0.5 percent. China Growth has decelerated more than previously expected amid cooling domestic demand and heightened trade tensions. Trade policy Source: Haver Analytics; International Monetary Fund; World Bank. uncertainty and higher tari s on trade with the A. Last observation is October 2019 for shipments of durables and exports of goods and services, and 2019Q3 for national accounts data. United States weighed on investor sentiment for B. Shaded area indicates forecasts. most of 2019. Industrial production growth has Click here to download data and charts. reached multiyear lows (Figure 1.6.A). FIGURE 1.5 Euro Area Trade ows have weakened substantially. Imports, Many economies in the region were on the verge of recession during most especially those of intermediate goods, have of 2019. The German industrial sector remains particularly weak. declined, falling more than exports, partly re ecting a deceleration in domestic demand. e A. Contributions to Euro Area growth B. Industrial production in the Euro Area and Germany contraction in exports to the United States has deepened, although shipments to the rest of the world have been somewhat more resilient. In response to the deceleration in activity, monetary policy has become more accommodative, but regulatory tightening to reduce non-bank lending has continued. e government has also stepped up some scal measures, including tax cuts and support for local Source: Haver Analytics; World Bank. A. “Other countries” includes Euro Area economies not listed. Data for 2019 are for 2019Q1-Q3 and governments for public investment spending are seasonally-adjusted annualized quarter-on-quarter growth rates. B. Industrial production excludes construction. Last observation is October 2019. (Figure 1.6.B; World Bank 2019b). Total debt has Click here to download data and charts. surpassed 260 percent of GDP, but the share of non-bank lending has continued to decline resolution of trade disputes with the United States (World Bank 2019c). that builds upon recent progress could bolster After decelerating to an estimated 6.1 percent in China’s growth prospects and reduce reliance on 2019, growth is expected to moderate to 5.9 policy support. percent in 2020 and 5.8 percent in 2021—0.2 percentage point below previous projections in Global trends both years. is is the rst time China will register a pace of expansion below 6 percent since 1990, International trade and investment have weakened amid a slowdown in labor productivity growth further, impeded by slowing global demand, as well and continued external headwinds (Chapter 3; as heightened policy uncertainty and an overall World Bank 2018a). A permanent and lasting increase in the level of tariffs despite recent de- 10 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 1.6 China nancial transactions, construction, and travel Growth has continued to decelerate amid weakening industrial activity. services, which together account for more than Imports have experienced a sharp decline. The government has also one-third of world services trade (WTO 2019a). stepped up fiscal support, including tax cuts and support to local governments for public investment spending. The slowdown in trade and manufacturing stems from a variety of factors. Weakening demand in A. Import volume and industrial B. General government gross debt production growth and decomposition of fiscal support Europe and Asia, in particular for trade-intensive measures automobiles and technology products, and the slowdown in investment growth have been important drags. Protectionist measures implemented by G20 countries since 2018 have affected over $1 trillion worth of trade flows, or nearly 7 percent of global goods trade (Figure 1.7.D; WTO 2019b). The number of regulatory restrictions affecting foreign direct investment flows has also been on the rise, increasing by more Source: Haver Analytics; International Monetary Fund; World Bank. than a third in 2018 (UNCTAD 2019a). A. Figure shows 12-month moving averages. Import data include only goods. Import volumes are calculated as import values deflated by import price deflators. Import price deflators for October and Additionally, despite recent moderation, global November are estimates. Last observation is November 2019. B. Gross debt consists of all liabilities that require payment or payments of interest and/or principal by trade policy uncertainty remains near historic the debtor to the creditor at a date or dates in the future. This includes debt liabilities in the form of SDRs, currency and deposits, debt securities, loans, insurance, pensions and standardized highs (Ahir, Bloom, and Furceri 2018; Baker, guarantee schemes, and other accounts payable. “Other” includes other net expenditures (including social security and State-Owned Enterprise funds). Fiscal support measures are World Bank staff Bloom, and Davis 2019). estimates. General government gross debt in 2019 are estimates. Click here to download data and charts. Trade tensions between the United States and China escalated throughout most of 2019, and escalation. Major central banks have loosened policy new tari s were implemented on the majority of in response, with interest rates in many advanced their bilateral trade. ese tensions, and the economies reaching unprecedented lows last year. ensuing increase in policy uncertainty, have Financial conditions in EMDEs have generally resulted in sizable aggregate losses for world trade; improved in parallel, except in economies perceived as while they have also had a positive impact on higher risk. Weak demand has pushed most some EMDEs through trade diversion, this impact commodity prices down, which has been partially has been relatively small. Trade frictions have also offset in some cases by supply restrictions. risen elsewhere, including between the United States and some of its other trading partners such Global trade as the European Union (EU), as well as between Japan and the Republic of Korea. e sharp slowdown in the trade-intensive manufacturing sector has continued to weigh on Nevertheless, negotiations between the United global trade. Global goods trade spent a signi cant States and China since mid-October resulted in a part of 2019 in contraction, with especially Phase One agreement between the two countries, pronounced weakness in advanced economies and including plans to partially roll back a subset of EMDEs such as China and the rest of East Asia U.S. tari s in exchange for Chinese commitments (Figure 1.7.A). e severe decline in the to make additional purchases of U.S. products, production of capital and intermediate goods in strengthen intellectual property protection, and G20 countries seen last year is consistent with pursue nancial services liberalization. e recent continuing weakness in trade and investment agreement, coupled with continued negotiations (Figure 1.7.B). Manufacturing export orders have and recent unilateral tari reductions by China, been contracting since late 2018 and services signals a notable de-escalation of trade tensions. export orders, while more resilient, have also Moreover, protectionist measures implemented decelerated (Figure 1.7.C). e softness in services since 2016 have been partially o set by various trade has so far been concentrated in global liberalizing measures that a ected 5 percent of G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 11 global goods trade in 2019. e U.S.-China Phase FIGURE 1.7 Global trade One agreement, as well as other positive The slowdown in global goods trade has been broad-based, with developments—such as progress in the rati cation particularly pronounced weakness in EMDEs in the East Asia and Pacific or implementation of the Africa Continental Free region. The marked decline in global capital and intermediate goods production last year highlights the weakness in trade and investment. Trade Agreement, the U.S-Japan trade agreement, Manufacturing export orders have continued to contract, and services and the United States-Mexico-Canada Agree- export orders have decelerated. Despite a recent de-escalation of trade tensions, the incidence of protectionist measures affecting global goods ment—could give a much-needed boost to trade trade has risen. growth. A. EMDE goods trade growth, B. Global production of capital and In sum, growth in global goods and services trade by region intermediate goods slowed sharply from 4 percent in 2018 to an estimated 1.4 percent last year, by far the weakest pace since the global nancial crisis, and is projected to rm throughout 2020 and reach 1.9 percent. Critically, these projections assume no further escalation or reduction of trade restrictions going forward. An additional decline in trade tensions and the associated policy uncertainty—if, for instance, ongoing U.S.-China negotiations were to result in further reductions in tari s— C. Manufacturing and services export D. Global trade subject to new orders protectionist measures could lead to a stronger-than-expected pickup in global trade growth. Financial markets Global nancing conditions eased considerably in 2019 (Figure 1.8.A). Bond yields in advanced economies fell to unprecedented lows, notwithstanding a pickup toward the end of the Source: CPB Netherlands Bureau for Economic Policy Analysis; Haver Analytics; World Trade year amid improvement in market sentiment. Organization; World Bank. Close to $12 trillion of outstanding global debt— A. Other EAP = East Asia and Pacific excl. China, ECA = Europe and Central Asia, LAC = Latin America and the Caribbean, MNA = Middle East and North Africa, SAR = South Asia, nearly a quarter of the total stock, and almost SSA = Sub-Saharan Africa. Figure shows 3-month moving averages. Trade is the average of export and import volumes. Last observation is October 2019. entirely from Western Europe and Japan—is B. Aggregate growth rates calculated using GDP weights at 2010 prices and market exchange rates. Sample includes the G20 countries for which capital goods and intermediates goods data are trading at negative interest rates. Major central available. Last observation is October 2019. banks, most notably the U.S. Federal Reserve and C. Figure shows 3-month moving average. PMI readings above 50 indicate expansion in economic activity, readings below 50 indicate contraction. Last observation is November 2019. the ECB, eased monetary policy last year in the D. Figure includes new import-restrictive measures, including tariff and non-tariff trade barriers. Annual data are mid-October to mid-October. face of softening global economic prospects, Click here to download data and charts. heightened downside risks, and persistently low in ation. Despite weak global investment, corporate debt has been rising in many countries, with particularly rapid growth in some riskier su ered from a ight to safety (Figure 1.8.B). categories, such as lending to highly leveraged Investors were particularly cautious about equity rms in the United States and the Euro Area (FSB markets in riskier EMDEs, which experienced 2019a). signi cant portfolio out ows during the period of heightened trade tensions and global growth In general, EMDE borrowing costs have fallen and concerns starting around August of last year, debt issuances have increased. Not all countries before recovering more recently (Figure 1.8.C). bene ted equally, however—EMDEs that already While equity and bond market developments in had low spreads experienced further declines, EMDEs have diverged considerably according to while economies with low sovereign credit ratings risk perception, many EMDE currencies have 12 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 1.8 Global finance rst half of 2019, with the decline being Global financing conditions have eased considerably, as major central particularly pronounced in EMDEs that had banks have provided accommodation in response to softening economic earlier experienced nancial pressures (UNCTAD prospects. However, EMDEs with low credit ratings have not benefitted 2019b). By contrast, remittances to EMDEs from the global decline in borrowing costs. Prior to their recent recovery, EMDE equity markets had been suffering significant outflows. A rising continued to grow and recently surpassed FDI share of EMDE currencies are at their lowest level against the U.S. dollar in (World Bank 2019d). a decade. Commodity markets A. Global financing conditions B. Change in EMDE bond spreads, by credit rating e prices of most commodities fell in 2019, mainly re ecting the deterioration in the growth outlook—especially that of EMDEs, which tend to have a larger income elasticity of demand for commodities (Figure 1.9.A; Ba es, Kabundi, and Nagle forthcoming). Forecasts have been revised down for most commodities in 2020 (Figure 1.9.B). C. EMDE portfolio flows D. Share of EMDE currencies at their Oil prices averaged $61/bbl in 2019, a 10 percent lowest level against the U.S. dollar since 2009 fall from 2018 and $5/bbl below previous projections. Prices were supported by production cuts by OPEC and its partners, including the December 2019 decision to remove 0.5 mb/d of production on top of previous reductions of 1.2 mb/d implemented since January 2019. Production has also been constrained in the Islamic Republic of Iran and the República Bolivariana de Venezuela by a variety of Source: Bloomberg; Haver Analytics; Institute of International Finance; International Monetary Fund; geopolitical and domestic factors. However, these J.P. Morgan; World Bank. pressures were o set by weakening oil demand, as A. Based on Goldman Sachs Financial Conditions Index for the United States, United Kingdom, Japan, Euro Area, India, Indonesia, Brazil, Mexico, Russia, and Turkey. Aggregates are calculated exempli ed by downward revisions to demand using GDP weights at 2010 prices and market exchange rates. Last observation is December 2019, which includes data through December 17, 2019. projections (Figure 1.9.C; IEA 2019). B. Figure shows change in unweighted annual averages of daily data from 2018 to 2019. Sample includes 42 EMDEs. Countries are grouped based on Fitch long-term sovereign rating. S&P ratings are used for countries not rated by Fitch (Belize, Senegal). Fitch and S&P use similar rating grades. Oil prices are forecast to decline slightly to an Bond spread shows percentage improvement in EMBI spreads versus a year ago. Last observation is December 16, 2019. average of $59/bbl in 2020 and 2021. U.S. supply C. Equity flows include Brazil, India, Indonesia, Pakistan, Philippines, Sri Lanka, South Africa, Thailand, Turkey, and Vietnam. Debt flows include Hungary, India, Indonesia, Mexico, Poland, South is expected to continue to increase in 2020 as new Africa, Thailand, and Turkey. Post-crisis average over January 1, 2010 to December 29, 2017. Last observation is December 16, 2019. pipeline capacity comes onstream. e greatest D. Figure shows 3-month moving average. To avoid excessive volatility, figure shows share of downside risk to the forecast is a further countries whose monthly average exchange rate against the U.S. dollar is within 5 percent of their most depreciated level. Sample includes 32 EMDEs. Last observation is December 2019, which deterioration in growth. Current expectations are includes data through December 17, 2019. Click here to download data and charts. for oil consumption growth to pick up to just over 1 percent in 2020, which is comparable to the pace of global oil demand seen during previous depreciated, and a growing share have fallen to global downturns (Figure 1.9.D). A critical upside their lowest exchange rate with the U.S. dollar in a risk to the forecast is the possibility of a further decade (Figure 1.8.D). signi cant reduction in trade tensions between the United States and China, which could boost oil Foreign direct investment (FDI) has continued its demand prospects. downward trend, with some of the recent weakness attributable to global policy uncertainty. Prices for most base metals weakened in the FDI weakened across all EMDE regions in the second half of 2019, primarily re ecting weaker G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 13 global growth and trade tensions. Metals prices are FIGURE 1.9 Commodity markets expected to decline further in 2020, re ecting Most commodity prices fell in 2019, and forecasts for 2020 have been subdued industrial commodity demand. As with revised down. Despite oil supply disruptions, deteriorating expectations for oil, a signi cant continued mitigation of U.S.- demand growth have put downward pressure on oil prices. A further softening in growth prospects is the key downside risk to oil demand and China trade tensions presents a key upside risk to price forecasts, while a sustained reduction of trade tensions represents a metals price projections. Agricultural prices major upside risk. declined in the second half of 2019 on improved weather conditions that ensured elevated stock A. Commodity price indexes B. Commodity price forecast revisions levels for grains. Agricultural prices are expected to stabilize in 2020, with risks to the forecast broadly balanced. Emerging market and developing economies e outlook for EMDEs has weakened signi cantly. As trade and investment rm, EMDE growth is C. Change in oil demand forecasts D. Oil demand and price growth around periods of economic downturn projected to pick up to 4.1 in 2020—0.5 percentage point below previous forecasts—and stabilize at 4.4 percent in 2021-22, with the pace of the recovery restrained by soft global demand and structural constraints, including subdued productivity growth. e near-term rebound in EMDE growth will be mainly driven by a projected pickup in a small number of large countries. Per capita income growth will remain well below long-term averages, making progress toward poverty alleviation and development Source: Energy Information Administration (EIA); International Energy Agency (IEA); Kose and Terrones (2015); Organization of Petroleum Exporting Countries (OPEC); World Bank. goals more challenging. A. Last observation is November 2019. C. Figure shows evolution of oil demand forecasts for 2019 by source. Diamonds show forecasts for oil demand in 2020. Recent developments D. Figure shows oil demand by component of global business cycle from 1971 to 2018. Over the time period, there have been four global recessions, defined as a contraction in growth, in 1975, 1982, 1991, and 2009, and three global slowdowns, defined by very low output growth, in 1998, 2001, and EMDEs have continued to experience substantial 2012 (Kose and Terrones 2015). Click here to download data and charts. weakness, with industrial production, trade flows, and investment decelerating sharply last year (Figures 1.10.A to 1.10.C). While services activity that had previously shown resilience. In all, has been appreciably more resilient than growth in about 60 percent of EMDEs is manufacturing, it has also moderated (Figure estimated to have slowed last year. In many 1.10.D). Growth has been particularly anemic in economies, subdued economic activity has been EMDEs that have experienced the lingering effects somewhat cushioned by still-resilient consumption of varying degrees of financial pressures or other and a shift toward more supportive monetary idiosyncratic factors in the past couple of years.1 policy. This weakness has also spread to other economies Growth in EMDEs that experienced recent financial or country-specific stresses remains feeble 1 These EMDEs include: (1) countries that have had an increase in (Kose and Ohnsorge 2019). To different degrees, their J.P. Morgan EMBI credit spread of at least one standard deviation above the 2010-19 average at any time since April 2018 these economies continue to face heightened (Argentina, Brazil, Egypt, Gabon, Jordan, Lebanon, Mexico, Nigeria, policy uncertainty and various domestic South Africa, Sri Lanka, Tunisia, Turkey); or (2) countries that have been subject to sanctions (Iran, Russia). Additional details about this challenges. With notable exceptions, activity has classification can be found in World Bank 2019e. started to firm somewhat; however, the recovery in 14 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 1.10 EMDE recent developments Mexico, easing monetary policy is providing some EMDEs have continued to experience substantial weakness, which has support to growth. In contrast, activity in spread to countries that, until recently, had shown resilience. Industrial Argentina has been contracting amid high policy production, trade flows, and investment have decelerated sharply. While uncertainty in the aftermath of severe financial services activity has been appreciably more resilient than manufacturing, it has also moderated. stress in mid-2019. In Iran, sanctions have been weighing significantly on growth. A. Industrial production growth B. Export and import volume growth Growth in other EMDEs has generally softened owing to global and domestic headwinds. Economies that are deeply integrated into global and regional production and trade networks— most notably in Asia and Europe—particularly suffered from global trade tensions and decelerating trade flows last year (Philippines, Thailand; World Bank 2019f, 2019g, 2019h). Tighter credit conditions in the non-banking C. Investment growth D. Manufacturing and services PMIs sector are contributing to a substantial weakening of domestic demand in India, while activity in Pakistan has decelerated in response to contractionary monetary policy intended to restore domestic and external balances. In some countries, capacity constraints are also limiting growth (Poland, Romania). Other economies have experienced temporary setbacks to construction and infrastructure projects (Costa Rica, Panama), Source: Haver Analytics; J.P. Morgan; World Bank. the effects of natural disasters (Guatemala, Papua A.-C. EMDEs under earlier pressure include: a) countries that have had an increase in their J.P. New Guinea), and the negative impact of social Morgan EMBI credit spread of at least one standard deviation above the 2010-19 average at any time since April 2018 (Argentina, Brazil, Egypt, Gabon, Jordan, Lebanon, Mexico, Nigeria, South Africa, Sri unrest (Bolivia, Chile). Lanka, Tunisia, Turkey), or b) countries that have been subject to sanctions (Iran, Russia). A. Figure shows 3-month moving averages. Dashed horizontal lines indicate the March 2006 to October 2019 averages. Industrial production growth for EMDEs under earlier pressure includes Commodity exporters those countries in the group for which data are available. Last observation is October 2019, which is estimated for Tunisia. B. Import and export data are volumes of goods and non-factor services. Aggregate growth rates calculated using GDP weights at 2010 prices and market exchange rates. “Latest” indicates 2019 full Growth in commodity exporters slowed from 2 year estimate. percent in 2018 to an estimated 1.5 percent in C. Investment is defined as gross fixed capital formation. EMDEs under earlier pressure includes those countries in the group for which data are available. “Latest” indicates 2019Q1-Q3 simple 2019, 0.6 percentage point below earlier forecasts, average. Last observation is 2019Q3. D. Figure shows 6-month moving averages. Manufacturing and services output are measured by reflecting softer-than-projected commodity prices, Purchasing Managers’ Index (PMI). PMI readings above 50 indicate expansion in economic activity; readings below 50 indicate contraction. Horizontal line indicates expansionary threshold. Last oil production cuts, decelerating investment in observation is November 2019. extractive sectors, and weakness in the largest Click here to download data and charts. countries that earlier experienced financial pressures or other country-specific stresses— most of these economies is proceeding at a particularly Argentina, Brazil, Iran, and Russia markedly slower pace than previously envisioned. (Figure 1.11.A). Weakening global demand and Some easing of lending conditions, as well as ongoing domestic challenges—including large progress on the reform agenda, are beginning to macroeconomic imbalances and domestic policy support a modest pickup in Brazil. In the Russian uncertainty—continue to discourage investment Federation, monetary policy easing and public and delay recovery in many commodity exporters infrastructure projects from the National Projects (Nigeria, South Africa; World Bank 2019h). program are buoying activity. In Turkey, activity is rebounding from earlier financial turmoil at a Despite supportive fiscal policy and stable non-oil faster-than-expected pace as domestic demand activity, difficulties in the oil sector and improves; however, the pickup remains fragile heightened geopolitical tensions are weighing on amid subdued confidence and investment. In activity in oil exporters in the MENA region G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 15 (Algeria, Iran, Saudi Arabia; World Bank 2019i). FIGURE 1.11 EMDE commodity exporters and importers In other commodity exporters with more policy Growth in both commodity exporters and importers decelerated last year. space, countercyclical policy measures have been In both groups, growth remains particularly subdued in the largest EMDEs partly offsetting the drag from weakening global that earlier experienced varying degrees of financial or country-specific stresses. demand and lower commodity prices, resulting in stable or moderately slower growth (Indonesia, A. Growth in commodity exporters B. Growth in commodity importers, Peru). excluding China Commodity importers Growth in commodity importers excluding China eased from 5 percent in 2018 to an estimated 3.3 percent in 2019—0.9 percentage point below previous projections and the slowest rate since the global financial crisis (Figure 1.11.B). This slowdown in part reflected a marked deceleration Source: World Bank. A.B. Data for 2019 are estimates. Aggregate growth rates calculated using GDP weights at 2010 in Turkey due to earlier financial stress, in Mexico prices and market exchange rates. Shaded areas indicate forecasts. Green lines indicate 2000-19 due to heightened policy uncertainty, and in India simple averages. Click here to download data and charts. due to a tightening of domestic non-bank credit conditions. Policy adjustments to address macroeconomic imbalances in Pakistan also caused by two tropical cyclones and weaker-than- weighed on aggregate growth in this group. expected coal production. For many commodity importers, momentum last Activity in other LICs, however, has been year was weaker than expected, reflecting declining somewhat more robust, reflecting improved exports and investment, only partly offset by more harvests (Malawi, Nepal), as well as continued accommodative monetary policy stances and fiscal services sector strength and solid public and support measures (Philippines, Thailand; World private investment growth (Guinea-Bissau, Bank 2019j). Nonetheless, growth in many Uganda). Nonetheless, softer external demand and commodity importers remains solid due to robust lower agricultural prices have dampened export private consumption and supportive policies in a revenues and slowed growth in some countries context of subdued inflation and resilient capital (Madagascar, Rwanda). flows (Bangladesh, Cambodia, Vietnam). Moreover, decelerating activity in some Outlook commodity importers also reflected a narrowing of positive output gaps (Poland, Romania). Growth outlook Low-income countries EMDE growth is expected to experience a moderate cyclical recovery from an estimated 3.5 The recovery in low-income countries (LICs) has percent last year to 4.1 percent in 2020—0.5 faltered amid softening external demand, weaker percentage point lower than previously projected commodity prices, political instability, and (Figure 1.12.A). Forecasts for almost all regions devastation from extreme weather events (Box 1.1; and half of EMDEs have been downgraded for Steinbach 2019; World Bank 2019e). Growth this year, largely re ecting weaker-than-expected among fragile LICs, in particular, has slowed exports and investment (Box 1.2; Chapter 2). markedly. In the Democratic Republic of Congo, EMDE growth is projected to stabilize at an falling metals prices stifled mining activity, while average rate of 4.4 percent in 2021-22, as trade the Ebola outbreak in the conflict-affected and investment rm. ese baseline projections northeastern region has persisted. Subdued growth are predicated on resilient consumption, a in Mozambique reflected widespread damage diminishing drag from earlier pressures in some 16 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 1.1 Recent developments and outlook for low-income countries Growth in low-income countries (LICs) has faltered in 2019, falling to 5.4 percent. The slowdown partly reflects global factors (softening external demand and weaker commodity prices), and idiosyncratic factors (political instability and devastation from extreme weather events). Growth is expected to firm over the forecast horizon, reaching an average of 5.7 percent in 2021-22. This pickup assumes improved stability, recovery from extreme weather events, continued investment in infrastructure, and the implementation of structural reforms and measures to strengthen business environments. Nonetheless, projected growth will be insufficient to markedly reduce poverty, particularly in LICs affected by fragility, conflict, and violence. Risks to the outlook include slower-than-expected growth in major trading partners, rising debt vulnerabilities, and growing insecurity. Recent developments demand and lower commodity prices constrained export revenues; however, sustained public investment helped Economic activity. The recovery in low-income countries offset some of this weakness. Nonetheless, activity (LICs) stalled in 2019 as global and idiosyncratic factors remained resilient, or strengthened, among some LICs. dampened activity. The global backdrop reflected Improved harvests supported rising agricultural production softening external demand and weaker commodity prices, (Malawi, Nepal), and services sector activity continued to while activity in some countries was weighed down further accelerate (Guinea-Bissau, Uganda). In Malawi, by political instability and extreme weather events. Growth agricultural production strengthened despite the impact of in LICs fell to an estimated 5.4 percent, 0.3 percentage Cyclone Idai, reflecting improved tobacco and maize point lower than previous forecasts (Figure 1.1.1.A). harvests in unaffected districts. In Ethiopia—the largest LIC economy—agricultural production slowed while The weaker-than-expected performance reflected a marked constrained hydroelectric power generation due to low slowdown in activity among fragile LICs.1 Growth in the dam levels dampened industrial activity; however, these Democratic Republic of Congo decelerated as weakening weaknesses were more than offset by continued robust external demand and lower metal prices weighed on services sector activity, particularly in travel, banking, and exports. The conflict-affected northeastern region of the telecommunications. On the demand side, activity was country is grappling with the second-largest Ebola supported by robust private consumption helped by strong outbreak on record, which began in the middle of 2018. harvests (Malawi, Nepal), and solid investment growth— In Haiti, growth is estimated to have contracted in 2019 both public and private (Guinea-Bissau, Uganda). Despite amid severe political instability, rapid exchange rate a sharp fall in aluminum prices, growth edged up in depreciation, elevated inflation, and rising food insecurity Guinea, partly due to continued infrastructure investment exacerbated by drought. Similarly, in Liberia, the in mining-related activities. In Sierra Leone, the estimated contraction in activity last year reflected the resumption of iron ore production helped boost activity. erosion of incomes from elevated inflation, weak harvests, and moderating mining production due to lower External positions. Current account balances widened commodity prices. In Mozambique—which has been on a among more than half of LICs. In some countries, larger reduced growth path since 2016—slowing growth in 2019 deficits reflected weaker exports related to softening was largely due to the devastation caused by last year’s external demand and lower international commodity cyclones alongside moderating coal production. In prices (Guinea-Bissau, Rwanda, Togo). Elsewhere, deficits addition to their heavy human toll, the cyclones have likely widened primarily due to imports of capital goods related reversed recent gains in poverty reduction in affected to large infrastructure investment projects (Mozambique, economies (Malawi, Mozambique; Baez, Caruso and Niu Togo, Uganda). Imports associated with cyclone-related 2019; World Bank 2019k). reconstruction added to existing deficits (Malawi, Activity also slowed among other LICs (Benin, Burkina Mozambique). In Ethiopia, however, the current account Faso, Madagascar, Rwanda, The Gambia, Tajikistan). In deficit narrowed amid improved services exports—largely Rwanda—one of the fastest growing economies in the transport services with Addis Ababa increasingly becoming world—growth edged down as weakening external a key regional hub—and as fiscal consolidation contributed to slower import growth. By the second half of 2019, capital flows into LICs appear to have weakened noticeably, as growing concerns over global growth Note: This box was prepared by Rudi Steinbach. Research assistance prospects and heightened trade tensions weighed on was provided by Hazel Macadangdang. investor sentiment. As a result, international reserves in the 1 Fragile LICs are those affected by fragility, conflict, and violence, median LIC have weakened somewhat and remain below according to the World Bank’s Harmonized List of Fragile Situations. the three-months-of-imports benchmark in about one- G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 17 BOX 1.1 Recent developments and outlook for low-income countries (continued) FIGURE 1.1.1 Recent developments in low-income countries Growth in low-income countries (LICs) has fallen to 5.4 percent in 2019 amid rising domestic and external headwinds. Growth is, however, expected to firm to an average of 5.7 percent in 2021-22, reflecting improved stability, recovery from extreme weather events, and continued investment in infrastructure. Fiscal deficits deteriorated sharply among LICs affected by fragility, conflict, and violence. Subdued inflation has allowed some central banks to easy policy rates. A. GDP growth B. Fiscal deficits C. Policy rates Source: Haver Analytics; World Economic Outlook, International Monetary Fund; Reserve Bank of Malawi; World Bank. Note: LICs = low-income countries. FCV LICs are LICs affected by fragility, conflict, and violence. A. Aggregate growth rates calculated using GDP weights at 2010 prices and market exchange rates. B. Unweighted averages. Sample includes 27 LICs. C. Reflects data up to December 19th, 2019. Prior to April 2017, data for Mozambique reflects the money market rate. Click here to download data and charts. quarter of countries—leaving these countries more Among fragile LICs, growth is forecast to rise to 3.7 vulnerable to negative shocks. percent in 2020, from 3.2 percent in 2019, in part due to improved political stability in some countries, Fiscal positions. LIC fiscal balances deteriorated, on strengthening business environments, and as the lingering average, in 2019 with the average deficit widening to an effects of extreme weather events wane. In Afghanistan, estimated 3 percent, from 2.6 percent in 2018 (Figure greater political stability following elections in late 2019 is 1.1.1.B). Fiscal deficits mostly widened among fragile expected to help support activity. Notable business LICs, partly reflecting low domestic revenue mobilization environment reforms in Togo will continue to bolster while public spending remained elevated. In the growth (World Bank 2020). In Chad and Mozambique, Democratic Republic of Congo, efforts to contain investment in new production capacity should spur growth spending were not sufficient to offset the decline in fiscal and boost exports, more than offsetting softer commodity revenues resulting from the weaker mining sector prices and weaker external demand. In the Democratic performance. In contrast, increased fiscal consolidation Republic of Congo, however, growth is projected to supported by greater revenue mobilization, as well as moderate further as lower metal prices—particularly for broad-ranging tax administration reforms have helped cobalt—continue to suppress mining production. deficits improve in several LICs (Burkina Faso, Ethiopia, Malawi, Mali). In other LICs, economic activity is expected to remain resilient, with growth above 6 percent over the forecast Outlook for 2020-22 horizon. In countries such as Benin and Rwanda, the Economic growth. Growth in LICs is projected to remain expansion will be supported by public investment in unchanged at 5.4 percent in 2020, before firming to an infrastructure, strong agricultural growth, and increased average of 5.7 percent in 2021-22. Forecasts for this year private sector activity as reforms continue to bolster the and next are 0.6 percentage point lower than previous business environment. Accommodative monetary policy projections, reflecting weaker external demand, lower stances amid relatively subdued inflation will further commodity prices, and policy tightening among some support activity in some countries (Malawi, Tanzania; large LICs. The expected pickup is predicated on no Special Focus 2; Figure 1.1.1.C). In Uganda, growth will further deceleration in external demand and a stabilization be boosted by public and private infrastructure of commodity prices, albeit at lower levels. investments, as well as in energy projects, as the country 18 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 1.1 Recent developments and outlook for low-income countries (continued) FIGURE 1.1.2 Outlook for per capita GDP and risks Growth in per capita incomes is expected to firm to an average of 2.9 percent in 2021-22; however, it will be markedly weaker among LICs affected by fragility, conflict, and violence. For these countries, per capita growth will be insufficient to make significant progress in poverty alleviation. Productivity in LICs is a mere 2 percent of the advanced-economy average, reflecting low productivity in comparatively larger agricultural sectors. Labor shifting to more productive sectors has been an important source of productivity growth in LICs. Debt sustainability concerns remain elevated, with a rising number of countries in debt distress. Insecurity, conflicts, and insurgencies, are leading to an increase in displaced populations. A. Per capita GDP growth B. Changes in LIC extreme poverty rates C. LIC productivity and agriculture value between 2015 and 2020 added D. Contribution to aggregate productivity E. LICs in debt distress F. Internally displaced populations in growth LICs Source: APO productivity database; Easterly and Fischer (1994); Expanded African Sector, Groningen Growth Development Center; Haver Analytics; ILOSTAT; International Monetary Fund; Penn World Table; United Nations High Commissioner for Refugees (UNHCR); World Development Indicators, World Bank. Note: Shaded area indicates forecasts. LICs = low-income countries. FCV = fragility, conflict, and violence. A. Aggregate per capita growth rates calculated by dividing the total GDP at 2010 prices and market exchange rates for each subgroup by its total population. Sample includes 25 LICs, 12 “FCV LICs”, and 13 “Other LICs”. B. The number of people living on or below the international poverty line of $1.90 per day as a share of the total population. Data for 2020 are estimates and calculated using data from World Bank. “FCV LICs” and “Other LICs” samples each include 12 and 13 countries, respectively. C. Productivity data based on 74 emerging market and developing economies (EMDEs), including 11 low-income countries (LICs). Blue bars show unweighted average output per worker during 2013-18 relative to the advanced-economy average. Whiskers indicate interquartile range relative to the advanced-economy average. Agriculture value added reflects 2018 data and is based on 132 EMDEs, including 23 LICs. Red bars show unweighted average share of agriculture in value added. D. Growth “within sector” shows the contribution to aggregate productivity growth of each sector holding employment shares fixed. The ‘between sector’ effect shows the contribution arising from changes in sectoral employment shares. Sample includes 46 EMDEs of which 8 are LICs. E. Number of LICs eligible to access the IMF’s concessional lending facilities that are either at high risk of, or in, debt distress according to the joint World Bank-IMF Debt Sustainability Framework for Low-Income Countries. The sample includes 28 LICs. F. MNA = Middle East and North Africa, SAR = South Asia, SSA = Sub-Saharan Africa. Internally Displaced Populations (IDPs) are persons or groups of persons who have been forced or obliged to flee or to leave their homes or places of habitual residence, in particular as a result of or in order to avoid the effects of armed conflict, situations of generalized violence, violations of human rights or natural or human-made disasters, and who have not crossed an internationally recognized state border. Data reflects only internally displaced populations (IDPs) who are protected or assisted by UNHCR, and country totals are not necessarily representative of the entire IDP population in that country. Sample includes 15 countries, of which 2 are in the Middle East and North Africa, 1 is in South Asia, and 12 are in Sub-Saharan Africa. Click here to download data and charts. prepares to export oil by 2023. Similarly, higher growth in growth is expected to slow due to tighter fiscal and Niger in 2022 reflects a sharp pickup in crude oil exports monetary policy stances aimed at containing inflation. as oil production is expected to quadruple from current levels. Activity in Guinea will benefit from investments in Prospects for per capita income convergence and poverty new mining production capacity. In Ethiopia, however, alleviation. Per capita GDP growth in LICs is expected to G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 19 BOX 1.1 Recent developments and outlook for low-income countries (continued) TABLE 1.1.1 Low-income country forecastsa Percentage point differences (Real GDP growth at market prices in percent, unless indicated otherwise) from June 2019 projectionsd 2017 2018 2019e 2020f 2021f 2022f 2019e 2020f 2021f Low-Income Country, GDPb 5.5 5.8 5.4 5.4 5.5 5.8 -0.3 -0.6 -0.6 Afghanistan 2.7 1.8 2.5 3.0 3.5 3.5 0.1 -0.2 -0.1 Benin 5.8 6.7 6.4 6.7 6.7 6.7 -0.1 0.2 0.2 Burkina Faso 6.3 6.8 6.0 6.0 6.0 6.0 0.0 0.0 0.0 Burundi 0.5 1.6 1.8 2.0 2.1 2.2 0.0 -0.1 0.1 Chad -3.0 2.6 3.0 5.5 4.8 4.8 -0.4 -0.1 0.0 Congo, Dem. Rep. 3.7 5.8 4.3 3.9 3.4 3.6 -1.6 -2.6 -3.4 Ethiopiac 10.0 7.9 9.0 6.3 6.4 7.1 1.1 -1.9 -1.8 Gambia, The 4.8 6.6 6.0 6.3 5.8 5.5 0.6 1.1 0.8 Guinea 10.0 5.8 5.9 6.0 6.0 6.0 0.0 0.0 0.0 Guinea-Bissau 5.9 3.8 4.6 4.9 5.0 5.0 0.3 0.1 -0.5 Haitic 1.2 1.5 -0.9 -1.4 -0.5 1.4 -1.3 -3.0 -1.8 Liberia 2.5 1.2 -1.4 1.4 3.4 4.2 -1.8 -0.2 2.1 Madagascar 4.3 5.1 4.7 5.3 4.4 5.0 -0.5 0.0 -0.7 Malawi 4.0 3.5 4.4 4.8 5.2 5.3 -0.1 0.1 0.1 Mali 5.3 4.7 5.0 5.0 4.9 4.9 0.0 0.1 0.1 Mozambique 3.7 3.4 2.0 3.7 4.2 4.4 0.0 0.2 0.0 Nepalc 8.2 6.7 7.1 6.4 6.5 6.6 0.0 0.0 0.0 Niger 4.9 6.5 6.3 6.0 5.6 11.9 -0.2 0.0 0.0 Rwanda 6.1 8.6 8.5 8.1 8.0 8.0 0.7 0.1 0.5 Sierra Leone 3.8 3.5 4.8 4.9 4.9 5.0 -0.6 -0.5 -0.3 Tajikistan 7.1 7.3 6.2 5.5 5.0 5.0 0.2 -0.5 -1.0 Tanzania 6.8 5.4 5.6 5.8 6.1 6.2 0.2 0.1 0.0 Togo 4.4 4.9 5.3 5.5 5.5 5.5 0.3 0.3 0.4 Ugandac 3.9 5.9 6.1 6.5 5.9 6.0 0.0 0.0 0.1 Source: World Bank. Note: World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of countries’ prospects do not significantly differ at any given moment in time. a. Central African Republic, the Democratic People’s Republic of Korea, Somalia, Syria, and Yemen are not forecast because of to data limitations. b. Aggregate growth rate calculated using GDP weights at 2010 prices and market exchange rates. c. GDP growth based on fiscal year data. For Nepal, the year 2019 refers to FY2018/19. d. Due to changes in the official list of countries classified as low income by the World Bank, the sample of LICs in this table is not comparable to June 2019. However, an identical sample is used for the comparison of the aggregate LIC GDP projection. Click here to download data. remain broadly unchanged at 2.5 percent in 2020, before fragile LICs—where the incidence of extreme poverty is firming to an average of 2.9 percent in 2021-22. This pace even higher—per capita GDP is expected to grow by a is insufficient to yield substantial progress in poverty mere 1 percent in 2020-22, after having contracted in 40 reduction as growth in LICs is often not inclusive and the percent of cases last year. As a result, the number of people conversion of growth into poverty reduction is therefore in LICs living below the international poverty line of low (Christiaensen, Chuhan-Pole, and Sanoh 2013; $1.90 per day will remain elevated, while continuing to Christiaensen and Hill 2018; Figure 1.1.2.A). Among rise among fragile LICs (Figure 1.1.2.B). 20 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 1.1 Recent developments and outlook for low-income countries (continued) To raise LIC growth over the medium term requires World Bank 2019m; World Bank and IMF 2018). Non– sustained improvements in labor productivity (Chapter 3). Paris Club creditors have also become a more important Labor productivity—average output per worker—in LICs source of financing over the past decade, especially in Sub- is a mere 2 percent of that in the average advanced Saharan Africa (World Bank 2015). Increased access to economy and one-tenth of the productivity level in the market-based debt may also be increasing governments’ average emerging market and developing economy exposure to interest rate and refinancing risks. Sharp (EMDE), and LIC productivity growth has been increases in debt-servicing costs would undermine much- persistently below that of EMDEs (Figure 1.1.2.C). This needed fiscal consolidation efforts and absorb revenues partly reflects LICs’ heavy reliance on agricultural sectors, that could otherwise be used for productivity-enhancing including widespread subsistence farming, as well as the investments in health care, education, and infrastructure. misallocation of resources—often caused by distortionary price controls (Special Focus 1). Raising LIC aggregate LICs’ weakening reserve buffers mean that renewed productivity will face several challenges. The reallocation episodes of financial stress, accompanied by an unexpected of labor from mostly agriculture to higher-productivity tightening of international financial conditions, could sectors such as mining and construction has been an disrupt capital inflows, fuel disorderly exchange rate important driver of LIC productivity in the pre-crisis depreciations, and raise financing costs. LICs with weaker period; however, this engine of productivity growth has macroeconomic fundamentals, higher foreign-currency- largely stalled following the collapse in global industrial denominated debt, or greater political risks would be most commodity prices (Figure 1.1.2.D). Moreover, longer- vulnerable. term prospects for commodity demand are weakening as Insecurity, conflicts, and insurgencies—particularly in the growth in China—the largest source of commodity Sahel and conflict-affected economies in the Middle East demand—slows and shifts towards less resource-intensive and North Africa—may further weigh on economic sectors (World Bank 2018b). Climate change will pose activity as well as food security in many countries if they increasing challenges to efforts to raise productivity in the were to intensify (Burkina Faso, Central African Republic, agricultural sector, with large falls in crop yields expected Chad, the Democratic Republic of Congo, Ethiopia, Mali, as global temperatures rise (Fuglie et al. 2019). Niger, Republic of Yemen, Somalia, South Sudan, Syrian Arab Republic; FAO 2019). Moreover, the large Risks. Risks to the outlook are firmly to the downside. A populations that are forcibly displaced by these conflicts faster-than-expected deceleration in growth of major world cluster in areas that often become a source of further economies and key trading partners—such as the United instability, with poverty rates being worse than in their States, the Euro Area, or China—would adversely affect places of origin (Figure 1.1.2.D; Beegle and Christiaensen export demand and investment in several LICs. Together, 2019). these three economies account for four-tenths of both LIC goods exports and foreign direct investment, and about Natural disasters related to growing climate extremes, such one-quarter of remittance inflows. Countries that depend as flooding or severe and prolonged drought episodes, on extractive industries—specifically metals producers— remain an important risk for many LICs, as agricultural would be hard-hit by a sharp slowdown in China, as it output often accounts for a high share of domestic value accounts for more than half of global metals demand added, and infrastructure is generally less resilient than in (World Bank 2018b). more developed economies (World Bank 2019e). LIC government debt reached 55 percent of GDP, on Health crises are a continuous concern. Although the pace average, in 2019—a 19 percentage point rise since 2013— of new Ebola infections in the Democratic Republic of keeping debt sustainability concerns elevated (World Bank Congo has slowed in the second half of 2019, efforts to 2019l). By November 2019, 12 out of 28 LICs were contain the second-largest outbreak in history have been regarded as being in debt distress, or at high risk thereof, complicated by conflict (Wannier et al. 2019). As under the IMF-World Bank debt sustainability evidenced by the West African Ebola outbreak of 2014-16, framework—two more than at the end of 2018 (Figure the current outbreak poses a significant risk to economic 1.1.2.C). The ratio of interest payments to GDP has activity, particularly if it were to spread to major urban doubled since 2013, in part reflecting the rising share of centers, or to neighboring countries (De la Fuente, Jacoby, non-concessional debt as commercial creditors have and Lawin 2019). become an important source of credit (Essl et al. 2019; G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 21 large economies, reduced policy uncertainty, FIGURE 1.12 EMDE outlook varying degrees of monetary policy support, EMDE growth is expected to recover moderately, reaching 4.1 percent in generally benign borrowing costs, no major swings 2020 and stabilizing at an average of 4.4 percent in 2021-22. This is a in commodity prices, no further deterioration in benign but fragile scenario given ongoing global headwinds. The recovery will not be broad-based and will instead mainly be driven by a global activity, and no new adverse shocks. ey projected pickup in a small number of large economies. Aggregate therefore represent a benign but fragile scenario, economic slack in EMDEs will persist in the near term, with actual EMDE growth expected to remain below potential. given the ongoing global headwinds of slowing advanced-economy growth, subdued global trade, and declining commodity prices (Figure 1.12.B). A. Growth outlook B. Average share of EMDEs with annual growth accelerating by more than 0.1 percentage point, 1962-2019 e expected pickup in aggregate EMDE growth is not broad-based: A third of EMDEs are projected to decelerate this year. Instead, it is largely predicated on a rebound in a small group of large EMDEs, most of which are emerging from deep recessions or sharp slowdowns caused by earlier nancial pressures or other idiosyncratic factors. Indeed, about 90 percent of the pickup in EMDE growth in 2020 is accounted for by just eight countries—Argentina, Brazil, India, Iran, C. Contributions to the change in D. EMDE growth EMDE annual growth Mexico, Russia, Saudi Arabia, and Turkey—even though they represent just a third of EMDE GDP (Figure 1.12.C). Excluding these eight countries, aggregate EMDE growth would experience almost no acceleration. More generally, aggregate economic slack in EMDEs will persist in the near term, and actual EMDE growth this year will remain below potential (Figure 1.12.D). Source: J.P. Morgan; World Bank. Projections for Argentina have been downgraded A.C. Data for 2019 are estimates. “Main drivers of pickup” includes the eight largest EMDEs that following the severe nancial market turmoil last account for 90 percent of the acceleration in EMDE growth between 2019 and 2020 (Argentina, Brazil, India, Iran, Mexico, the Russian Federation, Saudi Arabia, and Turkey). Aggregate growth year; the impact of this event is assumed to rates calculated using GDP weights at 2010 prices and market exchange rates. Shaded areas indicate forecasts. gradually diminish over the forecast horizon. In A. Green lines indicate 2000-19 simple averages. B. AE = advanced economies. “Subdued trade” refers to growth below 2.5 percent. “Moderating Brazil, Russia, and South Africa, elevated policy commodity prices” refers to a year-on-year contraction in the non-energy commodity index. uncertainty is expected to moderate; however, D. Estimates of potential growth are from a multivariate filter model of World Bank (2018a). Aggregate growth rates are calculated using GDP weights at 2010 prices and market exchange rates. Sample recovery in these countries is projected to be includes 57 EMDEs. Data for 2020 are forecasts. Click here to download data and charts. fragile due to continued challenges associated with the implementation of reforms, sanctions, or infrastructure bottlenecks. Growth in some other large economies (Egypt, India, ailand) is expected to pick up, supported by policy easing policy tightening among some large LICs (Box and gradually improving business con dence in 1.1). e expected pickup later in the forecast response to recent reforms. horizon assumes that activity among fragile LICs recovers as political stability improves Growth in LICs is projected to remain little (Afghanistan, Guinea-Bissau), investments in new changed at 5.4 percent in 2020 and edge up to an capacity o set weaker external demand (Chad, average of 5.7 percent in 2021-22. Forecasts for Mozambique), and as rebuilding e orts following this year and next are 0.6 percentage point lower last year’s cyclones boost activity (Malawi, than previous projections, re ecting weaker Mozambique; World Bank 2019h). Among other external demand, lower commodity prices, and LICs, activity is expected to remain generally 22 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 1.13 EMDE per capita income growth and EMDE productivity growth is expected to persist poverty or deepen (Chapter 3; World Bank 2018a). Going Despite significant gains in poverty alleviation over the last three decades, forward, EMDE potential growth is likely to be meeting the Sustainable Development Goals by 2030 appears out of reach dampened by the lingering e ects of past weak for many EMDEs, partly because of the recent loss of momentum in per investment and subdued investment prospects, capita income growth. In Sub-Saharan Africa, per capita growth is expected to remain below 1 percent, exacerbating the concentration of diminishing demographic dividends, and more extreme poverty. limited avenues for technological di usion, especially in the face of rising protectionism A. Sustainable Development Goals B. Per capita growth in EMDEs (World Bank 2019e). Per capita income growth and poverty e number of people living in extreme poverty— below $1.90 per day—has fallen by more than 1 billion over the past three decades, and remarkable progress has been made on several development indicators. Yet, meeting the Sustainable Development Goals (SDGs) by 2030 appears out C. Cumulative per capita income gains D. Global extreme poverty of reach for many EMDEs (Figure 1.13.A). and losses relative to 1990-2014 trend Extreme poverty rates are estimated to exceed 30 percent of the population in one-quarter of economies. Around 830 million people still live without electricity. Approximately 2 billion people do not have access to at least basic sanitation services. In LICs, child mortality rates are around triple their SDG target, while access to essential health services remains de cient. Source: United Nations; World Bank. To meet the infrastructure-related SDGs alone A. Sample includes 155 EMDEs. Orange lines indicate interquartile ranges. “Access to at least basic will require annual investment equivalent to 4.5 sanitation” and “Under-5 mortality” data reflect 2017 and 2018, respectively. B. Data for 2019 are estimates. Aggregate growth rates calculated using GDP weights at 2010 prices percent to 8.2 percent of low- and middle-income and market exchange rates. EMDE sample includes 144 countries, with 83 commodity exporters. C. EAP = East Asia and Pacific, ECA = Europe and Central Asia, LAC = Latin America and the countries’ GDP between 2015 and 2030 Caribbean, MNA = Middle East and North Africa, SAR = South Asia, SSA = Sub-Saharan Africa. (Rozenberg and Fay 2019; Vorisek and Yu, Negative bars represent the cumulative shortfalls in regional per capita income growth from 2015 to 2019 relative to the 1990-2014 average growth rate. For ECA, the average uses data for 1995-2014 forthcoming). e severity of this challenge has to exclude the immediate aftermath of the collapse of the Soviet Union. D. Data for South Asia in 2015 are estimates. been ampli ed by the loss of momentum in Click here to download data and charts. EMDE per capita income growth during recent years (Figures 1.13.B and 1.13.C). Given sustained headwinds to activity, per capita income resilient spurred by sustained public investment in growth in EMDEs is expected to stabilize around infrastructure along with greater private sector 3.2 percent over the near term—well below long- activity (Benin, Rwanda, Uganda). In some term averages. Lower income growth will also countries, more accommodative monetary policy adversely a ect poverty reduction e orts, and amid relatively subdued in ation will support there is already evidence that poverty reduction growth (Special Focus 2; Malawi, Tanzania). has started to slow (Ruch 2019a; World Bank However, in Ethiopia—the largest LIC—growth 2018c). is expected to slow due to tighter scal and monetary policy stances aimed at containing In about one-quarter of EMDEs—mostly in ation. commodity exporters—per capita growth will be inadequate to prevent income gaps from widening Longer-term growth prospects for EMDEs are also relative to advanced economies. In Sub-Saharan challenging (Ruch 2019a). In particular, the post- Africa—home to 24 of the 31 LICs and almost 60 crisis weakness in several fundamental drivers of percent of the world’s extreme poor—per capita G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 23 BOX 1.2 Regional perspectives: Recent developments and outlook Growth in almost all EMDE regions was weaker than expected in 2019, reflecting downgrades to more than half of EMDEs. Activity in most regions is expected to pick up in 2020-21, but the recovery will largely depend on a rebound in a small number of large EMDEs, some of which are emerging from deep recessions or sharp slowdowns. East Asia and Pacific. Growth in the region is projected to Mexico fades, and the recession in Argentina eases after slow from an estimated 5.8 percent in 2019 to 5.7 percent bouts of severe market stress, regional growth is projected in 2020 and moderate further to 5.6 percent in 2021-22. to rise to 1.8 percent in 2020 and about 2.4 percent in Easier financing conditions and fiscal policy support will 2021. is recovery will not be su cient to reverse the partly mitigate the lingering impact of trade tensions amid growing per capita income gap with advanced economies domestic challenges. In China, growth is expected to slow in some LAC economies. Moreover, the regional outlook gradually, from an estimated 6.1 percent in 2019, to 5.9 is subject to signi cant downside risks, including from percent in 2020, and to 5.7 percent by 2022. In the rest of market volatility and adverse market responses to weak the region, growth is expected to recover slightly to 4.9 scal conditions; deeper-than-expected spillovers from percent in 2020 and firm further to 5 percent in 2021-22. slowdowns in Argentina, China, and the United States; The balance of risks has improved, but risks to the outlook heightened social unrest; and disruptions from natural are still tilted to the downside. They include a sharp disasters and severe weather. slowdown in global trade due to renewed escalation of trade tensions amid a fragile global outlook; a sharper- Middle East and North Africa. Regional growth than-expected slowdown in major economies; and a decelerated to an estimated 0.1 percent in 2019. sudden reversal of capital flows due to an abrupt Geopolitical and policy constraints on oil sector deterioration in financing conditions, investor sentiment, production slowed growth in oil-exporting economies, or geopolitical relations. An upside risk to the forecast is despite support from public spending. Growth in oil related to stronger-than-expected recovery of regional importers remained stable, as reform progress and resilient investment and trade amid a sustained de-escalation of tourism activity were offset by structural and external trade tensions between China and the United States. headwinds. Regional growth is projected to pick up to 2.4 percent in 2020 and to about 2.8 percent in 2021-22, as Europe and Central Asia. Growth in the region infrastructure investment and business climate reforms decelerated to an estimated 2 percent in 2019, re ecting a proceed. Risks are tilted firmly to the downside— sharp slowdown in Turkey as a result of acute nancial geopolitical tensions, escalation of armed conflicts, slower- market stress in 2018, as well as in the Russian Federation than-expected pace of reforms, or weaker-than-expected amid weak demand and cuts in oil production. Regional growth in key trading partners could heavily constrain growth is projected to strengthen in 2020, to 2.6 percent, activity. as activity recovers in Turkey and Russia, and to stabilize to 2.9 percent in 2021-22. Key external risks to the South Asia. Growth in the region is estimated to have regional growth outlook include spillovers from weaker- decelerated to 4.9 percent in 2019, re ecting a sharper- than-expected activity in the Euro Area and escalation of than-expected and broad-based weakening in domestic global policy uncertainty. e region also remains demand. In India, activity was constrained by insu cient vulnerable to disorderly commodity and nancial market credit availability, as well as by subdued private developments. consumption. Regional growth is expected to pick up gradually, to 6 percent in 2022, on the assumption of a Latin America and the Caribbean. Growth in the region modest rebound in domestic demand. While growth in slowed markedly in 2019, to an estimated 0.8 percent, Bangladesh is projected to remain above 7 percent through held back by idiosyncratic factors in large economies, the forecast horizon, growth in Pakistan is projected to headwinds from slowing global trade, and social unrest in languish at 3 percent or less through 2020 as several countries. As activity in Brazil gathers pace amid macroeconomic stabilization e orts weigh on activity. improving investment conditions, policy uncertainty in Growth in India is projected to decelerate to 5 percent in FY2019/20 amid enduring nancial sector issues. Key risks to the outlook include a sharper-than-expected slowdown Note: This box was prepared by Patrick Kirby with contributions in major economies, a reescalation of regional geopolitical from Rudi Steinbach, Temel Taskin, Ekaterine Vashakmadze, Dana tensions, and a setback in reforms to address impaired Vorisek, Collette Wheeler, and Lei Ye. Research assistance was provided balance sheets in the nancial and corporate sectors. by Hazel Macadangdang. 24 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 1.2 Regional perspectives: Recent developments and outlook (continued) FIGURE 1.2.1 Regional growth Growth in almost all EMDE regions was weaker than expected in 2019, reflecting downgrades to more than half of EMDEs. Activity in most regions is expected to pick up in 2020-21, but the recovery will largely depend on a rebound in a small number of large EMDEs, some of which are emerging from deep recessions or sharp slowdowns. A. Regional growth, weighted average B. Regional growth, unweighted average C. Regional investment, weighted average D. Regional exports, weighted average E. GDP growth in ECA, weighted average F. GDP growth in LAC, weighted average Source: World Bank. A.-D. Bars denote latest forecast; diamonds correspond to January 2020 forecasts in the Global Economic Prospects report. Average for 1990-2019 is constructed depending on data availability. For Europe and Central Asia, the long-term average uses data for 1995-2019 to exclude the immediate aftermath of the collapse of the Soviet Union. A.C.D.E.F. Aggregate growth rates calculated using GDP weights at 2010 prices and market exchange rates. Since the largest economies account for about 50 percent of GDP in some regions, weighted averages predominantly reflect the developments in the largest economies in each region. Shaded areas indicate forecasts. B. Unweighted average regional growth is used to ensure broad reflection of regional trends across all countries in the region. Click here to download data and charts. Sub-Saharan Africa. Growth in the region moderated to a production, and robust growth among several exporters of slower-than-expected 2.4 percent in 2019. Activity was agricultural commodities. Nonetheless, these growth rates dampened by softening external demand, heightened will be insufficient to make significant progress in reducing global policy uncertainty, and falling commodity prices. poverty in many countries in Sub-Saharan Africa, Domestic fragilities in several countries further constrained highlighting the need for lasting improvements in labor activity. Growth is projected to firm to 2.9 percent in productivity to bolster growth over the medium term. 2020 and strengthen to 3.2 percent in 2021-22—notably Downside risks to the outlook include a sharper-than- weaker than previous projections. The growth pickup is expected deceleration in major trading partners; increased predicated on improving investor confidence in some large investor risk aversion and capital outflows triggered by economies, a strengthening cyclical recovery among elevated debt burdens; and growing insecurity. industrial commodity exporters along with a pickup in oil G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 25 income growth over the forecast horizon is Near-term projections for global growth mask expected to remain below 1 percent. In contrast, diverging contours for the forecasts for advanced per capita incomes are forecast to rise close to 5 economies and EMDEs. Aggregate growth in percent per year in East Asia and Paci c and advanced economies is expected to slow from 1.6 South Asia. As a result, the rapid declines in the percent in 2019 to 1.4 percent in 2020, primarily number of extreme poor living in these two fast- re ecting a deceleration in the United States and growing regions are likely to continue over the anemic activity elsewhere. In contrast, EMDE near term. Absent major policy e orts to lift per growth is envisioned to pick up from 3.5 percent capita growth, global extreme poverty will become in 2019 to 4.1 percent this year, mostly as a result increasingly concentrated in Sub-Saharan Africa of a pickup in a small number of large economies, (Figure 1.13.D; Beegle and Christiaensen 2019; some of which are emerging from deep recessions World Bank 2018c). or sharp slowdowns and whose outlooks are therefore fragile. Absent this group of countries, Risks to the outlook EMDE growth would be essentially stagnant and, with advanced economies decelerating, global Global growth, which weakened to an estimated 2.4 growth would actually slow. is indicates that percent in 2019, is projected to edge up to 2.5 weaker-than-expected activity in this small set of percent this year, following an expected recovery of EMDEs could derail the expected recovery in trade and investment. Despite a recent notable EMDE—and global—growth. reduction in the threat of protectionism, risks to the global outlook remain on the downside. A re- e contribution of EMDEs to the projected escalation of global trade tensions could further weigh pickup in global growth also hinges on the on world activity. Amid nancial sector weighting methodology. Using market exchange vulnerabilities, major economies could slow more rates, as is done in these baseline projections, than expected. EMDEs remain at risk of nancial yields the aforementioned tepid recovery of global stress, especially those with elevated debt, while some growth. Using purchasing power parity (PPP), EMDE regions could be a ected by geopolitical however, places greater weight on EMDEs— tensions, social unrest, large swings in commodity which are forecast to grow faster than advanced prices, or increasingly volatile weather patterns. On economies—and thus results in a somewhat more the upside, further de-escalation of trade tensions pronounced global pickup. between the United States and China could continue to mitigate global policy uncertainty and bolster As a result of the greater emphasis on the activity. contribution of EMDEs—especially large, fast- growing ones—to global activity, global growth is Summary of global outlook and risks projected at 3.2 percent in 2020 using PPP weights, compared to 2.5 percent using market In light of softening trade and manufacturing, exchange rates (Table 1.1). is is because global growth weakened to an estimated 2.4 EMDEs are expected to account for 40 percent of percent last year. is was the slowest pace of this year’s global output using market exchange expansion since the global nancial crisis—below rates but 60 percent using PPP weights. In that registered in 2012, when the Euro Area particular, China’s share of global GDP in 2020 is su ered a serious debt crisis, and in 2015-16, expected to be around 15 percent using market when many EMDE commodity exporters were exchange rates but 20 percent using PPP weights. facing large declines in commodity prices and In fact, of the 0.7 percentage point di erence in concerns about China’s economy were 2020 global growth projections between the two widespread. As international trade and investment weighting methods, China accounts for over 50 recover, global growth is projected to edge up to percent, with the three next largest contributors to 2.5 percent in 2020—0.2 percentage point below the di erence in global growth accounting for the previous forecasts—and gradually rm over the vast majority of the remainder. forecast horizon, reaching 2.7 percent by 2022. 26 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 1.14 Balance of risks e materialization of one or more of these risks Amid heightened uncertainty about the economic outlook, risks to global could lead to a more severe global downturn—a growth remain tilted to the downside. The probability of 2020 global growth situation many economies are not adequately being a full 1 percentage point or more below baseline forecasts is almost prepared to confront (Ruch 2019b). Re ecting a 20 percent and above historical averages. preponderance of downside risks, the probability that global growth in 2020 will be at least one A. Probability distribution around B. Probability of global growth being global growth forecasts 1 percentage point below current percentage point below baseline projections is baseline almost 20 percent, above historical averages. (Figure 1.14.B). Although downside risks predominate, there is also the possibility that major headwinds dissipate and the expected recovery is stronger than expected. In particular, recent policy developments—particularly those that have mitigated U.S.-China trade tensions—could lead to a sustained reduction in policy uncertainty and Source: Bloomberg; World Bank. A.B. The fan chart shows the forecast distribution of global growth using time-varying estimates of the bolster con dence, trade, and investment, which is standard deviation and skewness extracted from the forecast distribution of three underlying risk factors: Oil price futures, S&P 500 equity price futures, and term spread forecasts. Each of the risk an important upside risk to the outlook. factor’s weight is derived from the model described in Ohnsorge, Stocker, and Some (2016). Values for 2020 are computed from the forecast distribution of 12-month-ahead oil price futures, S&P 500 equity price futures, and term spread forecasts. Values for 2021 are based on 24-month-ahead Rising trade barriers and protracted forecast distributions. Last observation is December 19, 2019. Click here to download data and charts. policy uncertainty After decades of trade liberalization, protectionist measures have been implemented on a growing Regardless of the weighting scheme, baseline share of global trade (WTO 2019b). At the same projections for global growth represent a scenario time, the number of trade agreements coming into based on numerous benign assumptions. ey e ect has fallen sharply. Progress on the include no re-escalation of global trade tensions, a rati cation of important trade agreements such as mitigation in global policy uncertainty, no sharp EU-MERCOSUR has stalled. e WTO dispute slowdown in major economies, no nancial stress settlement system became deadlocked in in large EMDEs, stability in commodity prices, December, threatening a key pillar of the global and—critically—the avoidance of policy missteps. rules-based trading system. Without a well- Accordingly, there is substantial uncertainty established arbitration system, countries may use surrounding these baseline projections (Figure damaging unilateral or retaliatory trade policies to 1.14.A). resolve the increasing number of trade disputes (Figure 1.15.A). e rising number of trade On balance, risks to the outlook are on the restrictions and the associated uncertainty around downside (Ruch 2019a). e trade con ict them have contributed to the recent contraction in between the United States and China could re- global trade and the slowdown in global growth. escalate, and trade tensions involving other major e ratio of global trade-to-GDP growth has economies could emerge. Policy uncertainty could fallen below 1, far exceeding the slowdown that rise signi cantly and persistently. Some EMDEs would be expected from the ongoing maturation could su er full- edged nancial crises. of global value chains (Figure 1.15.B). Commodity markets could see disruptive swings. e United States or the Euro Area could su er Additional tari s have been imposed on the deepening slowdowns, or China could slow majority of bilateral trade between the United sharply—and the potentially large associated States and China over the past year. Despite the spillovers could substantially erode the EMDE announcement of the Phase One trade agreement outlook. Importantly, many of these risks are that resulted in the cancellation of planned tari intertwined. increases, re-escalation remains possible—many G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 27 commitments, including items related to the FIGURE 1.15 Rising trade barriers and protracted policy expansion of bilateral trade, intellectual property, uncertainty and technology transfer, may be di cult to After decades of trade liberalization, there has been a marked increase in enforce. protectionist measures and trade disputes, contributing to a slowdown in global trade growth. A re-escalation of U.S.-China trade tensions, or a deterioration in trade relations involving a broader set of countries, could e United States and China together account for substantially heighten policy uncertainty and further damage business nearly 40 percent of global GDP, nearly a quarter confidence and activity. of global trade, and an even larger share of capital goods trade (Figure 1.15.C). Accordingly, A. Trade disputes B. Ratio of global trade to GDP growth renewed disruption to U.S.-China economic ties could result in damage not only to these two economies but to the rest of the world, as its e ects would propagate through trade, nancial, and commodity linkages. ere is also the risk that trade tensions could extend to a broader set of countries. e imposition of U.S. tari s on automobiles and parts imports would impact a globally important sector that is already struggling, likely resulting in retaliation. e global C. U.S. and China share of global D. Global trade policy uncertainty and indicators, in 2018 business confidence multilateral trading system could be put at risk by a continuous rise in trade barriers stemming from many countries. In the longer run, protectionism would have serious negative consequences for the global economy, including by contributing to a further decline in the trade intensity of global growth, reducing productivity growth, and lowering real incomes (Barattieri, Cacciatore, and Ghironi Source: Ahir, Bloom, and Furceri (2018); Haver Analytics; International Monetary Fund; Organisation for Economic Co-operation and Development; World Bank; World Trade Organization. 2018). e fragmentation of global value chains A. Figure shows monthly average of active disputes. would cause e ciency losses for producers and B. Shaded area indicates forecasts. Trade measured as the average of import and export volumes. C. Trade measured as the average of goods exports and imports. Capital goods trade includes capital higher prices for consumers. Exporting rms, goods and transport equipment. D. Trade policy-related uncertainty is an index presented in Ahir, Bloom, and Furceri (2018) for 143 which tend to be more productive than exclusively countries on a quarterly basis. Business confidence data are end of period and include 7 advanced domestic rms, may need to redesign their supply economies and 5 EMDEs. Aggregate business confidence calculated using GDP weights at 2010 prices and market exchange rates. Last observation is 2019Q3 for trade policy uncertainty. Business chains using costlier inputs and bearing the cost of confidence data for 2019Q4 use October 2019. Click here to download data and charts. writing o stranded assets (Atkin, Khandelwal, and Osman 2017; Bernard and Jensen 2004). Despite recent progress in the resolution of trade baseline assumptions, policy uncertainty was to con icts, the impact of rising protectionism on rise further, the resulting impact on investment global growth has been magni ed by protracted would have critical consequences for activity in policy uncertainty and a decline in con dence both the short and long term. (Figure 1.15.D). A further increase in trade policy uncertainty could continue to be a material A deepening slowdown in major economies contributor to the softening of global growth (Caldara et al. 2019). Companies that are e United States, the Euro Area, and China are uncertain about the framework for doing business the world’s largest economies. All three su ered a in the future are reluctant to invest, often marked deceleration of activity in 2019 and face preferring to delay major, irreversible decisions downside risks (Figure 1.16.A). A deepening until the uncertainty has been resolved (Handley slowdown in any of these economies would and Limão 2015; Stokey 2016). If, in contrast to worsen economic prospects in countries around 28 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 1.16 A deepening slowdown in major the world through direct trade linkages and economies commodity prices, as well as through nancial and Activity decelerated substantially in major economies in 2019. The U.S. con dence channels. is could derail the corporate sector and the Euro Area banking sector exhibit vulnerabilities anticipated recovery in EMDE growth (World that could contribute to a deeper slowdown, which would have sizable spillovers and increase the probability of a global downturn. In China, Bank 2016a). e Latin America and the private debt as a share of GDP is well above levels observed prior to Caribbean region would be particularly impacted slowdowns in other EMDEs. Stress in the financial system could lead to by a sharp deceleration in the United States, while either a crisis or an extended period of slow growth as deleveraging drags on activity. economies in Europe and Central Asia would be disproportionately a ected by deepening weakness A. Growth in the United States, Euro B. U.S. financial vulnerability in the Euro Area. Spillovers from a slowdown in Area, and China indicators China would have sizable e ects on the country’s trading partners and in commodity producers (Ahmed et al. 2019; Stocker et al. 2018; World Bank 2016a). United States In the United States, growth is expected to decelerate as earlier tari increases, lingering C. Measures of health for Euro Area D. Probability of global downturn uncertainty, and scal policy all exert a drag on banks given U.S. or Euro Area recession activity. High corporate debt and elevated equity valuations increase the economy’s susceptibility to a more severe downturn (Figure 1.16.B). In the current environment of low rates, some high-yield borrowers have bene ted from investors’ search for yield. For example, leveraged loan issuance has increased rapidly, with borrowers bene tting from low spreads and loose lending standards. is increase has been facilitated by nancial E. Private sector debt in China F. Share of EMDEs slowing after institutions bundling many lower-rated loans into compared with peaks in other EMDEs reaching debt peaks more highly rated securities known as collateralized loan obligations (Federal Reserve Board 2019). A sudden decline in the perceived creditworthiness of borrowers could lead to a rapid fall in asset valuations and a localized credit crunch (Bank of England 2019). More generally, rising interest rates could slow activity across the entire corporate sector. Consumption has been the sole pillar supporting economic growth in recent Source: Bank for International Settlements; Center for Economic Policy and Research; Economic Cycle Research Institute; European Central Bank; Haver Analytics; Institute of International Finance; quarters, but this would be undermined if International Monetary Fund; Kose and Terrones (2015); Laeven and Valencia (2018); National Bureau of Economic Research; Shiller (2015); World Bank. tightening credit conditions and declining A. Data are seasonally adjusted for the United States and the Euro Area, and not for China. B. Last observation is December 2019 for Shiller Price-to-Earnings (P/E) ratio and 2019Q2 for debt. business con dence—for example, triggered by C. Return on equity is calculated using the average of 2008 to 2018. Euro Area aggregates calculated further increases in policy uncertainty—slowed using nominal U.S. dollar GDP weights of France, Germany, Italy, and Spain, as available. Capital adequacy ratio and non-performing loans are calculated using the average of 2009 to 2017. hiring and wage growth. Advanced economy aggregates calculated using available data for 37 advanced economies. D. Figure shows the probability of a global downturn occurring given a U.S. or Euro Area recession. Probabilities are based on annual data—the number of years with events divided by the total number Euro Area of years. U.S. recessions dated by National Bureau of Economic Research. Euro Area recessions dated by the Center for Economic and Policy Research. German recessions are used prior to the formation of the Euro Area. From 1958 to 2018, there have been four global recessions, in 1975, e Euro Area economy has already weakened 1982, 1991, and 2009, and three global slowdowns, in 1998, 2001, and 2012. E. Debt peaks defined as the highest value of private non-financial credit to GDP over the period considerably. Vulnerabilities in the banking 1960Q1 to 2019Q2. Sample includes 15 EMDEs. For China, the last observation is 2019Q2. F. Economies must have experienced a currency, systemic banking, or sovereign debt crisis within system could lead to a further slowdown, given two years after reaching the peak debt-to-GDP ratio. A slowdown is defined as a 1 percentage point or more drop in GDP growth between the two years before and the two years after peak debt-to-GDP that banks are the region’s primary source of credit ratio. Sample includes 15 EMDEs from 1960Q1 to 2019Q1. Click here to download data and charts. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 29 and—despite some recent improvement— case following periods of rapid debt accumulation continue to su er from low pro tability and (Figure 1.16.F; Kose, Sugawara, and Terrones elevated levels of non-performing loans (Figure 2019). 1.16.C). Negative interest rates in the region could further undermine bank pro tability and Financial stress in EMDEs erode nancial stability, possibly impacting sovereign borrowing costs through the “sovereign- EMDE debt burdens for both public and private bank” nexus (Arteta et al. 2016; Feyen and borrowers have grown considerably in recent years Zuccardi 2019; Molyneux, Reghezza, and Xie as part of the most recent global wave of debt forthcoming). An unexpected bank failure— (Figure 1.17.A; Chapter 4). Generally benign generated, for example, by exposure to Germany’s global nancial conditions have reduced debt- struggling industrial sector or sharp movements in service burdens for many EMDEs, but they may asset prices following Brexit—could trigger also be encouraging further debt accumulation, broader nancial stress and an associated loss of with prospect of persistently low advanced- con dence. As with the United States, a severe economy interest rates pushing some foreign slowdown in the Euro Area would substantially lenders to look for higher returns in EMDEs. In increase the probability of a more severe global some areas, debt is increasingly owing to riskier downturn (Figure 1.16.D). borrowers. Elevated debt can make economies vulnerable to large depreciations, capital out ows, China nancial stress, and abrupt policy tightening, particularly when it is nanced from abroad. In China’s primary vulnerability is its high and rising addition, solvency risks in the non-bank nancial stock of private debt in its increasingly complex sector are mounting in some large EMDEs (Arteta and interconnected nancial system (Arteta and and Kasyanenko 2019). Kasyanenko 2019; IMF and World Bank 2017). Credit to non- nancial corporates and households Recent credit booms in EMDEs have largely been as a share of GDP nearly doubled in the last used to fund consumption rather than investment decade, reaching about 210 percent in the rst (Figure 1.17.B; Chapter 4; Arteta and Kasyanenko quarter of 2019, well above the share observed 2019). is carries the risk that rising debt will not prior to previous growth slowdowns and nancial be matched by rising growth, increasing the crises in other EMDEs (Figure 1.16.E). e likelihood and impact of a loss of investor e ectiveness of credit in stimulating growth con dence. When such a loss is combined with an appears to be declining, which implies that the elevated proportion of debt denominated in bene ts of any further increase in credit would foreign currency, capital ight and depreciation diminish while risks would rise (Chen and Kang would add to existing debt sustainability concerns 2018). Rising defaults in local banks or in the and magnify the negative feedback loop (Bruno shadow banking system, a collapse in property and Shin 2018). prices, or large capital out ows alongside a sharp adjustment in asset prices could all ripple through In the past, EMDEs have been vulnerable to a the highly leveraged nancial system. is risk is broad-based strengthening of the U.S. dollar only partly mitigated by the country’s low reliance (Figure 1.17.C). Amid rapidly increasing non- on external nancing and ample capacity for scal nancial-sector debt, sharp dollar appreciation due and monetary support. to interest rate di erentials or generalized ight to safety can expose currency and maturity Alternatively, while a crisis could be avoided with mismatches and trigger widespread corporate policy support given China’s sizable policy bu ers, insolvencies (Caballero, Fernández, and Park the transition toward consumer-led and less credit- 2019; Chui, Kuruc, and Turner 2016). Large driven growth may lead to an extended period of depreciations are associated with higher borrowing subdued growth in the absence of deep structural costs, and monetary authorities are often required reforms. Moreover, private deleveraging may act as to tighten to stabilize currencies or resist the a persistent drag on activity, as is commonly the passthrough of higher import costs to domestic 30 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 1.17 Financial stress in EMDEs tightly integrated into the global nancial system. EMDE debt burdens have grown considerably in recent years for both While this has bene ts, it also facilitates the public and private borrowers; however, recent credit booms have generally contagion of global nancial shocks both to not been accompanied by rising investment. A loss in investor confidence foreign-currency and, to a lesser extent, local- could lead to an increase in bond spreads, as could a sharp U.S. dollar appreciation arising from flight to safety or other factors. currency debt markets (Agur et al. 2018; Arteta and Kasyanenko 2019; Cerutti, Claessens, and A. EMDE debt levels B. Investment surges during recent Puy 2019). e risk of contagion is further credit booms ampli ed by constrained policy room for crisis response and weaker bu ers against external shocks. Geopolitical and region-specific downside risks Downside risks to the global outlook are compounded by various geopolitical and region- C. Bond spreads and exchange rates D. Bond spreads in previous episodes speci c concerns. Geopolitical risks remain acute of stress globally and in several regions (Ruch 2019a). e disruption in Saudi oil production in mid- September highlights the potential for renewed tensions in the Middle East. In addition, if skirmishes in Eastern Europe and in South Asia escalate, there could be important consequences for growth in the associated regions. Amid geopolitical concerns, a sustained disruption Source: Bank for International Settlements; Haver Analytics; International Monetary Fund; J.P. in oil production may increase energy prices, to Morgan; Kose et al. (2017); World Bank. A. Aggregate for foreign-currency-denominated debt is calculated using moving GDP weights at 2010 the detriment of a ected suppliers and commodity prices and market exchange rates, excluding 2002-05 due to missing data. “Latest” indicates 2019Q2 importers. While commodity producers left for government debt and corporate debt, and 2018 for foreign-currency-denominated debt. B. A credit boom is defined as an episode during which the cyclical component of the nonfinancial una ected by the disruption could potentially private sector credit-to-GDP ratio (using a Hodrick-Prescott filter) is larger than 1.65 times its standard deviation in at least one year. The episode starts when the cyclical component first exceeds bene t from higher prices, these bene ts can be one standard deviation and ends in a peak year (“0”) when the nonfinancial private sector credit-to- GDP ratio declines in the following year. Consumption and investment surges are defined as periods undone if the price increase is accompanied by when the cyclical component of the consumption-to-GDP/investment-to-GDP ratio is at least one standard deviation above the HP-filtered trend. See Chapter 4 for more details. heightened volatility (van Eyden et al. 2019). C. NEER = nominal effective exchange rate. Bond spreads are represented by J.P. Morgan’s Emerging Market Bond Index (EMBI). Last observation is December 18, 2019. D. “t=0” indicates May 2013, June 2015, and March 2018. Bond spreads are represented by J.P. Alternatively, regions with a large presence of oil Morgan’s Emerging Market Bond Index (EMBI). producers, particularly MENA, would be adversely Click here to download data and charts. a ected by a sharp fall in oil prices resulting from weaker-than-expected demand amid subdued in ation (Figure 1.17.D). Similarly, large swings global growth. A sudden increase in supply— in commodity prices can potentially lead to re ecting, for instance, increased production in disruptive currency movements and balance of the United States—could also lead to a more payments di culties for vulnerable EMDEs. meaningful decline in prices. Such a decline could lead to substantial scal tightening, as was the case e risk of contagion of country-speci c nancial in 2014-16 (Figure 1.18.A; Stocker et al. 2018). distress across markets may be growing. Foreign Falls in metals or agricultural prices could follow a portfolio investors and global mutual funds are similar pattern and would also have a serious becoming more active in local bond markets, impact on economies in regions such as Sub- accounting for an increasing share of local- Saharan Africa, Latin America and the Caribbean, currency-denominated sovereign bonds. As a or Europe and Central Asia. While regions with result, EMDE nancial markets are now more large numbers of commodity importers would face G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 31 a positive terms-of-trade shock, these gains would FIGURE 1.18 Other downside risks likely be di used across many economies, only A sustained decline in the price of oil or other commodities could lead to partially o setting the relatively larger losses faced substantial fiscal tightening in commodity exporters, as was the case in by commodity exporters. 2014-16. Climate change is increasing the frequency of severe weather events and the volatility of agricultural conditions. Rising losses from severe weather events related to climate change increase the risk of Social unrest has been on the rise in a growing financial instability. number of countries in various regions, motivated by discontent about some combination of A. Change in overall fiscal balance in B. Rising frequency and costs from inequality, slow growth, governance, and oil-exporting EMDE sub-groups, from natural disasters 2014-16 economic policy. Unrest has the potential to disrupt activity and damage infrastructure. It may also make scal consolidation e orts more challenging for governments trying to ease tensions. Climate change is increasing the frequency of severe weather events and lowering agricultural productivity in some regions (IPCC 2018). As such, its impact is more detrimental for regions Source: International Monetary Fund; Munich Reinsurance Company; World Bank. A. Exchange rate classification is based on the IMF’s Annual Report on Exchange Arrangements and that have large numbers of countries with less Exchange Restrictions database, in which countries are ranked 0 (no separate legal tender) to 10 (free float). “Pegged” denotes countries ranked 1 to 6. “Floating” denotes countries ranked 7 to 10. resilient infrastructure and a larger share of Sample includes 27 oil-exporting EMDEs, based on data availability. Change in overall fiscal balance is measured from 2014-16. Above average and below average oil revenue groups are defined by agricultural production. ese countries tend to be countries above or below the sample average of oil revenues as a share of GDP based on 2014 data. poor and can ill-a ord the lost infrastructure and B. Global natural disasters and economic losses statistics from Munich Reinsurance Company including loss estimation based on Property Claim Services (PCS). The 30-year average represents income that accompanies extreme weather and 1988-2017. 5-year average represents 2014-2018. Losses adjusted to inflation based on local CPI. Click here to download data and charts. poor harvests. Similarly, regions with large coastal populations are at risk, not only from extreme weather, but also from rising sea levels. Climate recent trade agreement between the United States change also presents risks to the nancial system in and China that reverses some tari increases could some EMDEs, as the need to incorporate climate be the beginning of a constructive process leading risks into asset valuations and insurance coverage to a sustained reduction in policy uncertainty and calculations increases the risk of mispricing trade barriers. is could signi cantly improve (Figure 1.18.B). Rapid repricing is possible, for con dence and unlock pent-up demand for example, as more information becomes available investment, bolstering growth (Figure 1.19.A). about what assets are most at risk from rising sea Similarly, rapid progress on the post-Brexit trade levels or less habitable weather conditions (NGFS negotiations between the United Kingdom and 2018). the European Union could lift a cloud on Europe’s outlook. Upside risks Although downside risks predominate, there is Central banks provided signi cant accommo- also the possibility that the global recovery is dation over the course of 2019, which is expected stronger than expected. Existing headwinds to to contribute to the pickup in activity over the growth—including those related to policy near term. On a global level, falling policy rates uncertainty—could further dissipate, or additional have coincided with declining in ation, suggesting macroeconomic policy support could be deployed that there is scope for further monetary easing, in response. mainly for some EMDEs (Figure 1.19.B). In addition to the potential boost to growth from Heightened policy uncertainty exerted a notable monetary policy, some major advanced economies drag on activity throughout 2019, much of it with su cient space could choose to provide related to concerns about rising trade barriers. e additional scal support. 32 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 1.19 Upside risks Policy challenges Sustained progress in the resolution of U.S.-China trade tensions would reduce policy uncertainty, which could unlock pent-up demand for investment. A continued decline in global inflation could open the door to Challenges in advanced economies further monetary stimulus. Very low interest rates highlight the limited room A. Impact of a 10-percent decrease in B. Change in global inflation and that advanced-economy central banks have to provide U.S. policy uncertainty on investment interest rates over the last year growth additional accommodation. If persistent, they may also erode the health of nancial institutions. However, low borrowing costs have loosened some of the constraints on scal policy allowing for increased public investment or other support in countries with scal space, if needed. Fiscal positions could also be improved through better tax compliance and enforcement. Productivity growth in advanced economies has declined due to weak investment Source: Baker, Bloom, and Davis (2016); Bank for International Settlements; Bloomberg; Haver growth and aging populations. Reducing policy Analytics; World Bank. A. Figure shows median growth impact of 10 percent fall in U.S. economic policy uncertainty (EPU). uncertainty would buttress capital formation. See Annex SF.1B of World Bank (2017a) for details on the methodology. B. Calculations based on change in year-on-year global inflation and nominal interest rate between November 2018 and November 2019. Aggregate nominal interest rate calculated using GDP weights Monetary and nancial policies at 2010 prices and market exchange rates. Unbalanced samples include 35 advanced economies and 77 EMDEs, including 39 low-income countries, for nominal interest rates and include 36 advanced economies and 112 EMDEs for inflation. Last observation is November 2019. e combination of feeble growth and stubbornly Click here to download data and charts. subdued in ation in the post-crisis period has made it di cult for major central banks to remove FIGURE 1.20 Monetary and financial policies policy accommodation (Figure 1.20.A). Policy in advanced economies rates remain very low in most countries, and close Weak growth and low inflation have prevented major central banks from to their e ective lower bound, greatly limiting the removing policy accommodation in the post-crisis period. As a result, ability to further cut rates. Other policy tools, policy rates are at or close to their effective lower bounds in many such as policy guidance or quantitative easing, economies. Longer-term yields have also fallen, limiting the remaining room for other policy tools, such as forward guidance and quantitative have been used to help lower long-term interest easing. rates as short-term rates approached their lower- bound (Woodford 2012). However, the limits of A. Monetary policy rate increases B. Policy rates and 10-year sovereign these tools may also have been reached, with long- during current and previous yields expansions term yields in many economies, including Germany and Japan, now below zero (Figure 1.20.B). e downward trend in interest rates, and the associated challenges for monetary policy, appears to be a persistent phenomenon, driven in part by a fundamental weaknesses in investment demand across advanced economies (Rachel and Summers 2019; Williams 2016). Source: Bank of Japan; Bloomberg; European Central Bank; Federal Reserve System; Haver A number of ideas have been put forward to Analytics; National Bureau of Economic Research; World Bank. improve the traction of monetary policy, A. U.S. expansions: 1991-2001, 2001-07, 2009-present. Euro Area expansions: 1999-2008, 2009-11, 2013-present. Calculations based on trough and peak of policy rates of each period. Last observation including targeting price levels or nominal GDP is November 2019 for the United States and 2019Q3 for the Euro Area. B. Figure shows data as of December 18, 2019. rather than in ation, stimulating activity through Click here to download data and charts. direct transfers to households, and eliminating the lower bound by subordinating paper money to central-bank electronic money (Agarwal and Kimball 2019; Buiter 2014; Mertens and Williams 2019). ese come with their own risks G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 33 and tradeo s, including the di culty of FIGURE 1.21 Fiscal policy in advanced economies transitioning from one framework to another In many advanced economies, households are deleveraging and while maintaining the credibility and public corporate investment is weak, leaving aggregate demand unusually understanding that is essential for the e ective dependent on government borrowing. Public investment can bolster growth in both the short and long term, and increase the stock of public operation of monetary policy. capital, which has fallen in a number of economies. Aside from the constraints it places on monetary policy, an extended period of low or negative A. Change in debt over GDP since B. Public capital stock 2009, by sector interest rates may also be detrimental to the health of nancial institutions, as their interest rate margins become squeezed (Arteta et al. 2016; Brunnermeier and Koby 2019). For banks, low interest rates can reduce pro tability—and therefore resilience in the face of negative shocks—and encourage greater risk taking. Non- bank nancial institutions, which account for an increasing share of credit issuance, are also Source: Institute of International Finance; International Monetary Fund; World Bank. a ected. Pension funds and insurance companies A. Figure shows the change in the debt-to-GDP ratio since 2009. Sector aggregates are calculated using GDP weights at 2010 prices and market exchange rates. Sample includes 23 advanced often have xed future liabilities and may be economies. Last observation is 2019Q2. compelled to invest in riskier and less liquid assets B. Lines represent the ratio of general public capital stock to GDP, in billions of constant 2011 international dollars. Last observation is 2017. in order to meet their nominal return targets. Click here to download data and charts. Increased lending to over-leveraged borrowers may be sowing the seeds for future nancial stress, especially given uncertainty about non-bank positions is through better tax compliance and behavior and its impact on the nancial system enforcement (OECD 2019a). Preventing corpo- during a downturn (IMF 2019b). Regulatory rate tax avoidance through pro t shifting is one reforms have made the global nancial system way to broaden the revenue base, especially with more resilient since the global nancial crisis; respect to companies that provide digital services however, prudential authorities need to remain in a given jurisdiction without any physical vigilant to risks originating from the growing presence (World Bank 2018d). Providing tax importance of non-bank nancial institutions, and agencies with more resources to bring down tax be wary of vulnerabilities being masked by non-compliance could increase revenues while technological innovations and complex nancial helping reduce inequality (Sarin and Summers products (FSB 2019b). 2019). Should governments choose to provide scal Fiscal policy support, the focus should be on spending that has In many advanced economies, households are a high multiplier. is could include transfers to deleveraging and corporate investment is weak, low-income individuals, as well as to regional leaving aggregate demand unusually dependent on governments, whose spending tends to be more government borrowing (Figure 1.21.A). Further credit constrained and procyclical (Whalen and scal support may become necessary given the Reichling 2015). ese multipliers may be combination of slowing activity, elevated particularly large when interest rates are downside risks, and limited room for monetary constrained by their e ective lower bound, and policy accommodation. Many countries are can bene t other countries through spillovers, carrying persistent de cits, however, despite the especially if action is taken in an internat- budgetary bene ts of the global decline in interest ionally coordinated fashion (Auerbach and rates. Gorodnichenko 2013; Wieland 2010; Woodford 2011). By contrast, multipliers tend to be low One growth-friendly approach that advanced when debt levels are elevated (Huidrom et al. economies can take to improve their scal 2019). 34 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 1.22 Structural policies in advanced economies can use to help reverse this trend. Pursuing Productivity has slowed in advanced economies, primarily due to the growth-enhancing public investment, fostering decline in capital deepening, slowing gains in education and gender innovation, and increasing human capital can all equality, and lower levels of innovation associated with a shrinking working -age population. Policymakers can help reverse this trend by fostering be e ective means of boosting productivity innovation and human capital, as well as avoiding policy choices that (Chapter 3). hinder investment. e simplest option, however, is to avoid policy A. Contributions to labor productivity B. Share of advanced economies with choices that actively hinder investment. e rise of a slowdown in productivity drivers in trade protectionism and the associated uncertainty 2008-17 relative to 1998-2007 has made companies more reluctant to invest until the framework for global trade is normalized (Handley and Limão 2015; World Bank 2017a). A stable, predictable system based on a multilateral consensus about the rules governing global trade would foster investment and, thereby, strengthen potential output. Alongside weak investment, the other main drags Source: Barro and Lee (2015); Penn World Table; The Conference Board, United Nations; World Bank. on productivity in advanced economies are related A.B. Productivity defined as output per worker. Refer to Chapter 3 for details. Unbalanced sample to slowing gains in education and gender equality includes 29 advanced economies. A. Aggregate growth rates calculated using GDP weights at 2010 prices and market exchange rates. (Figure 1.22.B). In addition, the working-age B. Share of AEs where improvements in each driver of productivity were lower during 2008-2017 than in the pre-crisis period 1998-2007 or improvements were negative. Variables corresponding to each share of the population continues to shrink, which concept are: Investment = investment-to-GDP ratio, Education = years of schooling, Demography = share of working-age population, Gender equality = female average years of education minus male can slow productivity growth as younger average years. generations tend to adopt new technology more Click here to download data and charts. rapidly (Chapter 3). is trend is expected to continue in coming decades, but could be partially Public investment may be an especially e ective mitigated by allowing new migrants, who tend to form of scal support in many advanced be prime-aged, in an orderly fashion and as economies, as it can bolster growth in the short appropriate to country-speci c circumstances. term by crowding in private capital, and in the long term by increasing productivity growth and Challenges in emerging market and mitigating climate change (Bouakez, Guillard, and developing economies Roulleau-Pasdeloup 2017; Dreger and Reimers 2016; World Bank 2019j). e falling stock of While subdued in ation has allowed many EMDEs public capital as a share of GDP in some advanced to cut policy rates, a deterioration in investor economies suggests the need to ll infrastructure sentiment could require policy tightening. With the needs (Figure 1.21.B; Heintz 2010). To the extent space for scal support constrained by record-high that it boosts demand and potential output, debt, tax policy reforms are needed to broaden the tax borrowing to nance public investment may base to fund growth-enhancing and climate-friendly ultimately have a limited impact on public debt investment. Measures to improve governance and ratios (Abiad, Furceri, and Topalova 2015). business climates and phase out price controls can make institutional environments more conducive to Structural policies growth. Encouraging EMDE integration in supply chains could counterweigh the e ects of weak global Potential growth has been slowing in advanced trade. Bolstering productivity growth by encouraging economies due to a combination of demographic diversi cation and upgrading to high-value added, trends and decelerating productivity growth. e technology-intensive industries will be critical to shore latter primarily re ects the appreciably diminished up long-term growth. China’s key policy challenge is role of capital deepening as a contributor to to address lingering disruptions associated with trade growth since the global nancial crisis (Figure tensions while shifting to more balanced and 1.22.A). ere are a variety of tools policymakers sustainable growth. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 35 Policy challenges in China currency depreciation, as well as force policy interest rate hikes, exerting greater pressure on China’s authorities have provided monetary and economies still su ering the lingering e ects of scal support to mitigate the impact of higher previous nancial market stress. EMDEs with tari s on bilateral trade with the United States and large external imbalances tend to be the most weakening global demand. e central bank has vulnerable to nancial stress, including those that eased policy mainly by cutting bank reserve rely on short-term capital in ows to nance requirements. On the scal front, authorities have current accounts, borrow heavily in foreign- focused on measures to accelerate investment denominated currencies and from external lenders, spending at the subnational level. A number of and lack adequate reserve coverage levels. initiatives to improve market access for foreign investors and various reforms to improve the Many EMDEs lack bu ers to confront nancial business climate have also been implemented shocks—in nearly half of EMDEs, international (World Bank 2019c, 2020). reserves are currently below levels that would be consistent with reserve adequacy (IMF 2011; Kose China’s key policy challenge is to achieve a and Ohnsorge 2019). Among LICs, reserve permanent and lasting resolution of trade tensions coverage has fallen to a two-year low (Figure while continuing to shift to more balanced growth 1.23.D). Following the taper tantrum of 2013, and gradually reducing excessive leverage. is depreciations were less severe in countries with would require enhancing productivity by boosting larger reserves, highlighting the importance of investment in human capital; further improving restoring monetary bu ers (BIS 2019). In market access, competition, and nancial anticipation of renewed episodes of market discipline; strengthening intellectual property volatility, EMDE policymakers need to keep rights; reducing barriers to entry; continuing the expectations of longer-term in ation moderate gradual opening of China’s nancial system to and stable. is includes demonstrating a credible international investors; and fostering innovation commitment to in ation targets in economies that (Chapter 3; World Bank 2018e; World Bank and have implemented such a framework (World Bank DRC 2019). 2019n). EMDE monetary and nancial policies Since the global nancial crisis, more than two thirds of EMDEs have strengthened macro- Consistent with agging global growth, negative prudential policies to rein in the growth of credit output gaps, and moderating in ation in many to non- nancial corporations and households EMDEs, including some LICs, nearly 75 percent (Figure 1.23.E; Cerutti, Claessens, and Laeven of EMDEs have lower policy rates now than at the 2017; Koh and Yu 2019; World Bank 2019o). start of 2019, with more than half implementing Supervisory and regulatory frameworks need to be multiple cuts (Figure 1.23.A). Many EMDEs have further strengthened to confront future shocks and space to cut rates further as interest rates remain shore up nancial stability, especially in a context relatively high and in ation below target (Figures where cross-border lending has shifted from banks 1.23.B and 1.23.C). However, the e ectiveness of headquartered in advanced economies to EMDE- monetary policy in EMDEs is likely more limited headquartered banks (Figure 1.23.F). Macro- than in advanced economies, as the interest rate prudential measures, such as countercyclical channel may be weaker and the impact of external capital bu ers and limits on foreign-currency nancing conditions larger (Aoki, Benigno, and borrowing, can help contain systemic risk in Kiyotaki 2018; Choi et al. 2017). banking and corporate sectors. Additionally, carefully calibrated regulatory measures, such as Although global nancing conditions have reporting and licensing criteria, could help support generally eased, policy uncertainty and risk con dence and resilience in new platforms that aversion have tightened nancing conditions in expand the access to credit through nancial some EMDEs. An abrupt change in market technology innovations (BIS 2017). However, sentiment could reignite capital out ows and EMDEs will need to strike a careful balance when 36 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 1.23 EMDE monetary and financial policy considering the trade-o s between managing Moderating inflation and relatively high interest rates allowed many EMDEs macroprudential risk and fostering nancial to cut policy interest rates to support growth—consistent with negative development (Krishnamurti and Lee 2014). output gaps and below-target inflation. Reserve coverage sharply fell in 2019, particularly in LICs, leaving many economies unprepared to respond to financial market shocks. Strengthening regulatory frameworks in EMDEs EMDE scal policy is crucial, especially in a context where cross-border lending has shifted to EMDE-headquartered banks. Many EMDEs face narrowing scal space and may struggle to quickly rebuild bu ers, limiting A. Output gaps and policy interest B. Real interest rates their options to address a severe downturn (Figure rate actions 1.24.A; Ruch 2019b). Aggregate EMDE debt reached a historical high last year and is expected to rise further (Chapter 4). Fiscal sustainability remains a critical challenge in many EMDEs, re ecting increased spending in commodity exporters and reduced revenues in commodity importers (Figures 1.24.B and 1.24.C). Should a negative shock occur, the scope for scal accommodation may be constrained by the need C. Inflation and inflation targeters, D. Reserve coverage to ensure long-term scal stability. e case for 2019 providing scal support would be strengthened where there are clear needs, such as infrastructure gaps, and a transparent public expenditure review process. In many cases, however, the expansion of credit over the past decade has not been channeled into investment, and was instead used to fund consumption (Chapter 4; Arteta and Kasyanenko 2019). In particular, EMDE commodity exporters need E. Macroprudential policies: Use of F. Sources of cross-border bank loans financial institution-targeted to grapple with lower commodity prices, especially instruments in those oil exporters where scal breakeven prices are higher than oil prices. In many commodity exporters, scal revenues are not well diversi ed, leaving revenues highly dependent on commodity production and exposed to global commodity price volatility (Gunter et al. 2019). For scally constrained economies, building tax capacity is a crucial step towards mobilizing Source: Bank for International Settlements; Consensus Economics; Haver Analytics; International domestic resources, providing essential public Monetary Fund; World Bank. A. Output gaps aggregated using GDP weights at 2010 prices and market exchange rates and are services, pursuing appropriate redistributive estimated from a multivariate filter model of World Bank (2018a). Figure shows number of EMDEs with policy interest rates lower (higher) than start of the year. Sample includes 45 EMDEs. Countries policies to address inequality, and building scal with fixed exchange rates are excluded. Data as of December 19, 2019. bu ers (Doumbia and Lauridsen 2019). is is B. Real interest rates are nominal interest rates less expected inflation. Expected inflation is the one- year ahead forecast from Consensus Economics. Sample includes 17 EMDEs. Blue area shows particularly true in LICs, 80 percent of which lack minimum and maximum. Last observation is November 2019. C. Sample includes the 34 EMDEs with inflation targets and is based on data availability. Figure the tax revenues to provide even basic services, let shows the last observation, which is November 2019. D. Figure shows number of months of reserve coverage. Data are 6-month moving averages of the alone to meet the SDGs (Figure 1.24.D; Gaspar, sample median. Sample includes 66 EMDEs including 25 LICs. Last observation is October 2019. Jaramillo, and Wingender 2016). Overall, The Assessing Reserve Adequacy (ARA) metric is based on IMF (2011). E. Each bar represents share of EMDEs using at least one macroprudential tool that is financial policymakers need to ensure that public spending institution-targeted (for example, limits on foreign currency loans and leverage ratios). F. Sample includes 115 EMDEs, excluding China. Due to data availability, 77 EMDEs are included in is cost e ective and yields a positive growth 2018. Lending by non-BIS banks is estimated as total bank loans and deposits from the IMF Balance of Payments Statistics (excluding central banks) minus cross-border lending by BIS reporting banks. dividend, while also protecting critical social safety This difference mostly accounts for the banking flows originating from non-BIS reporting countries. nets and supporting climate-friendly measures. Click here to download data and charts. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 37 Measures that help mitigate and adapt to climate FIGURE 1.24 EMDE fiscal policy change, such as environmental tax reforms, can Fiscal deficits persist despite previous procyclical tightening in some reap a triple dividend by lowering pollution, EMDEs, as weaker-than-expected growth hindered revenue collection. raising welfare, and generating positive Weak tax capacity has contributed to fragile fiscal positions, particularly in LICs, highlighting the urgency for fiscally constrained economies to better externalities (World Bank 2019p). mobilize domestic resources or reform their tax structure to free up space to finance growth-enhancing spending. In many EMDEs, tax policy reform is a challenging process. To protect the most A. Fiscal impulses and output gaps B. Fiscal sustainability gaps in EMDEs vulnerable, adjusting income tax brackets to rising in ation can ease the tax burden and prevent the erosion of real net incomes. Harmonizing tax rates across di erent savings instruments or a well- designed earned income tax credit can support labor participation and poverty reduction without distorting the incentive to work and save (OECD 2019a). When there is a clear rationale for tax cuts, negative revenue e ects can be partly o set by measures that increase compliance—such as the C. Contribution to change in fiscal D. Share of EMDEs with limited tax balance, 2019 revenues to fund basic public introduction of a withholding mechanism or a services simpli cation of the tax structure—or that spur innovation and investment—such as tax credits on vocational education and research and development (Clavey et al. 2019; Correa and Guceri 2013; World Bank et al. 2015). Additional measures that broaden the tax base, including those that eliminate costly loopholes, can be complemented with reforms that strengthen tax administration and collection to reduce avoidance, Source: International Monetary Fund; Kose et al. (2017); World Bank. base erosion, and pro t shifting (Awasthi and A. Output gaps are estimates from a multivariate filter model of World Bank (2018a). Average of quarterly output gap data. Fiscal impulse is defined as the change in the structural fiscal deficit from Bayraktar 2014; OECD 2017; Packard et al. the previous year. A decline in structural deficit (a negative fiscal impulse) is a fiscal consolidation— countercyclical if implemented while output gaps are positive—while an increase in the structural 2019; Prichard et al. 2019; World Bank 2018d). deficit (positive fiscal impulse) is a fiscal stimulus—countercyclical if implemented while output gaps are negative. B. Fiscal sustainability gaps are measured as the difference between the primary (overall) balance EMDEs with unsustainable scal positions can and the debt-stabilizing primary (overall) balance. A negative bar indicates government debt is rising along an accelerated trajectory. also prioritize rebuilding policy space by C. Sample includes 152 EMDEs. D. Figure shows the share of EMDEs with tax revenue-to-GDP ratios that are below 15 percent, the improving spending e ciency, by shifting threshold needed to provide basic public services, as identified in Gaspar, Jaramillo, and Wingender (2016). Basic services include road infrastructure, health care, and public safety. Sample varies due spending toward growth-enhancing, climate- to data limitations. In 2017, the sample includes 70 EMDEs, of which 11 are LICs. friendly investment from unproductive current Click here to download data and charts. spending, and by strengthening governance to contain and eliminate wasteful spending (World crucial (Koh and Yu 2019; Munoz and Olaberria Bank 2017b). If public expenditure needs are 2019). high, rebalancing the tax structure can provide maneuvering room, particularly in economies with EMDE structural policies lower initial tax rates (Gunter et al. 2018, 2019). e realization of costly scal risks to public Over the long run, EMDE policymakers need to balance sheets, such as contingent liabilities, could undertake the necessary structural reforms to be stemmed through use of macroprudential buttress potential growth. Inadequate governance measures that help ensure the resilience of the and business climates need to be improved to banking sector. Building credible and transparent foster an institutional environment that is more medium-term expenditure frameworks that align conducive to growth. In a context of subdued with the strategic goals of the government is also trade growth, further integration of EMDEs into 38 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 global value chains needs to be promoted. higher incomes, and reduced poverty (World Critically, amid slowing capital deepening, Bank 2020). productivity growth—an essential driver of long- term growth and poverty reduction—needs to be Phasing out distortionary price controls rekindled. Many EMDEs, including LICs, face While introduced with the best social intentions, the added challenge of phasing out distortionary price control policies, often coupled with onerous price control policies that impede growth and subsidies, pose important obstacles to growth and development. In tackling these challenges, care development in many EMDEs, including LICs should be taken to protect vulnerable populations (Special Focus 1). e removal of these costly by improving social safety nets. controls can reduce misallocation of capital and Moreover, investment in green infrastructure and labor, spur investment, and increase competition its integration with traditional infrastructure can in sectors subject to these policies. Moreover, lower costs, help achieve development goals, and when paired with targeted social safety nets, their contribute to improving infrastructure systems’ removal can help reduce poverty and inequality resilience to climate change (Browder et al. 2019). (Verme and Araar 2017). Some of the scal Private sector nancing to meet large infrastruc- savings from the reforms can be used to fund ture investment needs and foster capital formation growth-enhancing education and infrastructure and the leveraging of digital technologies to spending. promote the inclusion, e ciency, and innovation Promoting integration into global value chains of rms in EMDEs are all crucial in boosting potential growth (World Bank 2016b). e rise in the incidence of protectionist measures over the past couple of years not only weighs on Implementing governance and business climate global trade growth but could lead to the reforms fragmentation of global supply chains and deprive Governance reforms in EMDEs have stalled, and EMDEs of a key source of growth and poverty renewed momentum is needed (World Bank reduction. Policy measures that help facilitate 2018a). e number of countries whose ranking trade in EMDEs by boosting their integration in for rule of law and control of corruption have existing supply chains and spurring the creation of signi cantly worsened in the last two decades new ones could provide a counterweight to the outnumber those whose rankings have improved global slowdown in growth and trade (World (Figure 1.25.A). Strikingly, very few large EMDEs Bank 2019a). A 10-percent increase in GVC had signi cant gains in any of the worldwide participation is estimated to boost per capita governance indicators, nor did LICs as a group. income growth by more than 10 percent, about Strengthening institutional quality and governance twice as much as standard trade (Figure 1.25.C). to protect property rights would encourage the Firms integrated in GVCs tend to be more shift from informal to more productive formal productive and capital intensive; they represent activities (World Bank 2017c). Measures that only about 15 percent of all trading rms, yet improve public sector e ciency through the account for almost 80 percent of total trade. GVC provision of high-quality and cost-e ective public participation is positively associated with foreign goods also need to be considered as they can help direct investment in EMDEs, as well as tech- raise rm productivity (Giordano et al. 2015). nology and knowledge transfers (Martínez‐Galán and Fontoura 2019; World Bank et al. 2017). Since 2009, only about a third of EMDEs increased their doing business score signi cantly, Reducing distortions to international trade can with notable regional variations (Figure 1.25.B). contribute to boosting EMDE participation in Reforms should aim to accelerate improvements in GVCs (Figure 1.25.D; OECD 2019b). e the business climate by tackling burdensome liberalization of barriers (both tari and non-tari ) regulations and enhancing the ease of doing a ecting imported intermediate inputs could business, in order to pave the way for more jobs, expand sources of supply available to EMDEs and G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 39 their ability to specialize. A one-standard-deviation FIGURE 1.25 EMDE structural policies—Governance, decrease in a country’s average manufacturing business climate, and GVC participation tari s—8 percentage points—is associated with an The number of EMDEs whose rankings for some key governance increase in the country’s backward GVC indicators have significantly worsened in the last two decades outnumber those whose rankings have improved. Since 2009, only about a third of participation (captured by the foreign value-added EMDEs increased their Doing Business score significantly. This highlights content of exports) of about 0.2 standard a critical need to foster institutional environments more conducive to deviations (Fernandes, Kee, and Winkler 2019). growth. Trade liberalization can help boost EMDEs’ participation in global value chains and contribute to rising per capita incomes. Liberalizing barriers to services trade, which are signi cantly higher than those for goods trade, is A. Change in Worldwide Governance B. Change in Doing Business scores, also important in promoting GVC growth. Indicators, 1996 to 2018 2009 to 2019 Trade facilitation policies that improve connectivity by enhancing trade and transport logistics and lower trade costs can help EMDEs better integrate into GVCs. For many goods traded in GVCs, a day’s delay has costs equivalent to a tari of 1 percent or more. Improving customs and border procedures, promoting competition in transport services, and improving port structure and governance are all strategies C. Impact of 1 percent increase in D. Impact of input tariffs on GVC that can help reduce trade costs related to time GVC participation on GDP per capita participation and uncertainty (Pathikonda and Farole 2016). Because GVCs thrive on the exible formation of networks of rms, a stable and predictable legal environment and contract enforcement are crucial (Ignatenko, Raei, and Micheva 2019). Better contract enforcement supports the supply of business services, which encourages the development of GVCs. e ability to enforce Source: World Bank. contracts relating to intellectual property is also A.B. A country significantly improved (deteriorated) if its rating increased (decreased) by two standard important for more innovative and complex value errors over the indicated periods. For Worldwide Governance Indicators, standard errors are the average between the two periods. For Doing Business, standard errors are the cross-country chains. standard deviation of changes in scores. A. Based on indicators from the Worldwide Governance Indicators (WGI) measuring aspects of governance. The four indicators are government effectiveness, regulatory quality, rule of law, and Complementary policies are also needed to ensure control of corruption. B. EAP = East Asia and Pacific, ECA = Europe and Central Asia, LAC = Latin America and the that the gains from participation in GVCs are Caribbean, MNA = Middle East and North Africa, SAR = South Asia, SSA = Sub-Saharan Africa. evenly distributed. ese include labor market C.D. Backward participation is defined as the share of foreign inputs in domestic value added. Forward participation is the share of domestic value added in exports. policies to help workers who may be hurt by C. GDP per capita increase as a result of 1 percent increase in x-axis indicators. Blue vertical lines indicate 95 percent confidence interval and red squares indicate point estimates. Estimates obtained structural change; mechanisms to ensure from a panel of standard Solow growth models augmented with measures of GVC using System Generalized Method of Moments (World Bank 2019a). Panel includes 100 countries across income compliance with labor regulations; appropriate tax groups for the period of 1990-2015. Non-GVC exports is defined as exports that neither include foreign value-added nor are exports of domestic value added that are re-exported in other countries’ policies to attract GVCs without undermining tax exports. revenues; and environmental protection measures D. Figure shows standardized beta coefficients for each variable from each of the three separate regressions listed. Results obtained from regressions using three-year lag of each determinant in (Taglioni and Winkler 2016). addition to country-year fixed effects and sectoral fixed effects. Click here to download data and charts. Fostering productivity growth EMDE productivity growth has been in a broad- boosted EMDE productivity growth prior to 2008 based downward trend in recent years (Figure are fading. Output per worker in EMDEs is, on 1.26.A; Chapter 3). is deceleration has average, less than one fth than that of advanced coincided with a slowdown in improvements in economies, and at current rates of productivity many correlates of strong productivity growth growth the average EMDE would take over 100 (Figure 1.26.B) e structural tailwinds that years to close half of the productivity gap with 40 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 1.26 EMDE structural policies—Productivity Policies to boost sectoral diversi cation are crucial, EMDE productivity growth has been in a broad-based downward trend in particularly for commodity exporters that have recent years. This deceleration has coincided with a slowdown in historically experienced low productivity improvements in many correlates of strong productivity growth. A reform growth—total factor productivity in commodity package that combines filling investment needs, boosting human capital, and improving the adoption of new technologies could lift productivity exporters has contracted by around 0.8 percent per significantly. Fostering productivity is key to alleviate poverty. year over the past four decades. Sectoral diversi cation may encourage productivity gains A. EMDE productivity growth B. Share of EMDEs with a slowdown in sectors that are less dependent on volatile in productivity drivers in 2008-17 relative to 1998-2007 commodity prices (Bahar and Santos 2018; Frankel 2010). Removing bottlenecks and barriers to investment in high value-added services sectors provides opportunities for rapid catch-up in productivity growth. Policymakers could signi cantly contribute to raising productivity in EMDEs by encouraging rms to upgrade to more high-value-added and technology-intensive subsectors (Cusolito and C. EMDE productivity reform scenario D. Productivity growth and global Maloney 2018; Syverson 2011). In addition, poverty improving the business environment fostering capital market development, and encouraging FDI could contribute to reducing cross-country sectoral productivity dispersion. Action is also needed to help reduce the vulnerability to adverse productivity shocks, such as nancial crises, disasters, and con ict (Cerra and Saxena 2008, 2017; Ray and Esteban 2017). Social safety nets play a key role in mitigating the Source: Barro and Lee (2015); Observatory of Economic Complexity; Penn World Table; Rozenberg and Fay (2019); The Conference Board; United Nations; World Bank. adverse e ects of new technologies that may Note: Productivity is defined as output per worker. Sample includes 29 advanced economies and 74 EMDEs. Refer to Chapter 3 for details. Aggregate growth rates calculated using GDP weights at 2010 initially be disruptive to employment. Policies that prices and market exchange rates. A. Figure shows 5-year moving averages. improve social insurance for unemployment are B. Econ. complexity = economic complexity. Post-crisis slowdown defined as the share of economies needed in the formal and informal sectors. Policies where improvements in each underlying driver of productivity during 2008-2018 was less than zero or the pace of improvement during the pre-crisis period 1998-2007. Unbalanced sample of 74 that incentivize adult learning, particularly for economies. Variables corresponding to each concept are (sample in parentheses): Demography =share of working-age population, Investment =investment to GDP ratio, Innovation =patents per high-order cognitive skills that complement new capita, Gender equality = ratio of female labor market participation rate to male, Urbanization = urban population (% total), Institutions = WGI Rule of Law, Income equality = (-1)*Gini coefficient, Education technologies, could help reintegrate displaced = years of schooling, ECI defined as Economic Complexity Index of Hidalgo and Hausmann (2009). Orange line indicates 50 percent. workers into the labor force (Andrews, Avitabile, C. The reform scenario assumes: (1) Fill investment needs: the investment share of GDP increases and Gatti 2019; World Bank 2018d). Measures by 4.5 percentage points as in the Rozenberg and Fay (2019) “preferred” infrastructure scenario. The increase is phased in linearly over 10 years; (2) Boost human capital: average years of education that help close the gender gap and improve female increases in each EMDE at its fastest cumulative 10-year pace during 2000-08; (3) Reinvigorate technology adoption: economic complexity (Hidalgo & Hausmann 2009) increases at the same pace labor force participation would also contribute to as its fastest 10-year rate of increase during 2000-08. D. Poverty is defined as the extreme poor living at or below $1.90 per day, in 2011 PPP terms. raising growth and productivity (Ianchovichina Click here to download data and charts. and Leipziger 2019). Overall, a reform package that combines lling investment needs, boosting human capital, and improving the adoption of new technologies could lift productivity growth by advanced economies. In addition, cyclical just over half of a percentage point over 10 years headwinds, rising protectionist measures, and (Figure 1.26.C). By bolstering productivity, these elevated policy uncertainty highlight the policies will support poverty alleviation (Figure importance of productivity-enhancing policies, 1.26.D). such as those that improve institutions, encourage investment, and promote diversi cation. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 41 TABLE 1.2 Emerging market and developing economies1 Commodity exporters2 Commodity importers3 Albania* Lao PDR Afghanistan Pakistan Algeria* Liberia Antigua and Barbuda Palau Angola* Madagascar Bahamas, The Panama Argentina Malawi Bangladesh Philippines Armenia Malaysia* Barbados Poland Azerbaijan* Mali Belarus Romania Bahrain* Mauritania Bhutan Samoa Belize Mongolia Bosnia and Herzegovina Serbia Benin Morocco Bulgaria Seychelles Bolivia* Mozambique Cabo Verde Solomon Islands Botswana Myanmar* Cambodia Sri Lanka Brazil Namibia China St. Kitts and Nevis Burkina Faso Nicaragua Comoros St. Lucia Burundi Niger Croatia St. Vincent and the Grenadines Cameroon* Nigeria* Djibouti Thailand Chad* Oman* Dominica Tonga Chile Papua New Guinea Dominican Republic Tunisia Colombia* Paraguay Egypt Turkey Congo, Dem. Rep. Peru El Salvador Tuvalu Congo, Rep.* Qatar* Eritrea Vanuatu Costa Rica Russia* Eswatini Vietnam Côte d’Ivoire Rwanda Fiji Ecuador* Saudi Arabia* Georgia Equatorial Guinea* Senegal Grenada Ethiopia Sierra Leone Haiti Gabon* South Africa Hungary Gambia, The Sudan* India Ghana* Suriname Jamaica Guatemala Tajikistan Jordan Guinea Tanzania Kiribati Guinea-Bissau Timor-Leste* Lebanon Guyana Togo Lesotho Honduras Turkmenistan* Maldives Indonesia* Uganda Marshall Islands Iran* Ukraine Mauritius Iraq* United Arab Emirates* Mexico Kazakhstan* Uruguay Micronesia, Fed. Sts. Kenya Uzbekistan Moldova, Rep. Kosovo West Bank and Gaza Montenegro Kuwait* Zambia Nepal Kyrgyz Republic Zimbabwe North Macedonia * Energy exporters. 1. Emerging market and developing economies (EMDEs) include all those that are not classified as advanced economies and for which a forecast is published for this report. Dependent territories are excluded. Advanced economies include Australia; Austria; Belgium; Canada; Cyprus; the Czech Republic; Denmark; Estonia; Finland; France; Germany; Greece; Hong Kong SAR, China; Iceland; Ireland; Israel; Italy; Japan; the Republic of Korea; Latvia; Lithuania; Luxembourg; Malta; Netherlands; New Zealand; Norway; Portugal; Singapore; the Slovak Republic; Slovenia; Spain; Sweden; Switzerland; the United Kingdom; and the United States. 2. An economy is defined as commodity exporter when, on average in 2012-14, either (i) total commodities exports accounted for 30 percent or more of total goods exports or (ii) exports of any single commodity accounted for 20 percent or more of total goods exports. Economies for which these thresholds were met as a result of re-exports were excluded. When data were not available, judgment was used. This taxonomy results in the classification of some well-diversified economies as importers, even if they are exporters of certain commodities (e.g., Mexico). 3. Commodity importers are all EMDEs that are not classified as commodity exporters. 42 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 Awasthi, R., and N. Bayraktar. 2014. “Can Tax References Simpli cation Help Lower Tax Corruption?” Policy Research Working Paper 6988, World Bank, Abiad, A., D. Furceri, and P. Topalova. 2015. “ e Washington, DC. Macroeconomic E ects of Public Investment: Evidence from Advanced Economies.” IMF Working Paper Baez, J. E., G. Caruso, and C. Niu. 2019. “Extreme 15/95, International Monetary Fund, Washington, Weather and Poverty Risk: Evidence from Multiple DC. Shocks in Mozambique.” Poverty and Equity Notes 11. World Bank, Washington, DC. Agarwal, R., and M. Kimball. 2019. “Enabling Deep Negative Rates to Fight Recessions: A Guide.” IMF Ba es, J., A. Kabundi, and P. Nagle. Forthcoming. Working Paper 19/84, International Monetary Fund, “ e Role of Income and Substitution in Global Washington, DC. Commodity Demand.” World Bank Working Paper, World Bank, Washington, DC. Agur, I., M. Chan, M. Goswami, and S. Sharma. 2018. “On International Integration of Emerging Sovereign Bahar, D., and M. A. Santos. 2018. “One More Bond Markets.” IMF Working Paper 18/18, Resource Curse: Dutch Disease and Export International Monetary Fund, Washington, DC. Concentration.” Journal of Development Economics 132 (1): 102–14. Ahir, H., N. Bloom, and D. Furceri. 2018. “ e World Uncertainty Index.” Available at https:// Baker, S. R., N. Bloom, and S. J. Davis. 2016. papers.ssrn.com/sol3/papers.cfm?abstract_id=3275033. “Measuring Economic Policy Uncertainty”. e Quarterly Journal of Economics 131 (4): 1593–1636. Ahmed, S., R. Correa, D.A Dias, N. Gornemann, J. Hoek, A. Jain, E. Liu, and A. Wong. 2019. “Global _______. 2019. “ e Extraordinary Rise in Trade Spillovers of a China Hard Landing”. International Policy Uncertainty.” Vox CEPR Policy Portal, Finance Discussion Papers 1260. Board of Governors September 17. Available at https://voxeu.org/article/ of the Federal Reserve System, Washington, DC. extraordinary-rise-trade-policy-uncertainty. Andrews, K. G., C. Avitabile, and R. V. Gatti. 2019. Bank of England. 2019. Financial Stability Report. July. “Domestic Government Spending on Human Capital: Bank of England, London. A Cross-Country Analysis of Recent Trends.” Policy Research Working Paper 9033, World Bank, Barattieri, A., M. Cacciatore, and F. Ghironi. 2018. Washington, DC. “Protectionism and the Business Cycle.” NBER Working Paper 24353, National Bureau of Economic Aoki K., G. Benigno, and N. Kiyotaki. 2018. Research, Cambridge, MA. “Monetary and Financial Policies in Emerging Markets.” Unpublished manuscript. Barro, R., and J. Lee. 2015. Education Matters: Global Schooling Gains from the 19th to the 21st Century. New Arteta, C., and S. Kasyanenko. 2019. “Financial York, NY: Oxford University Press. Market Developments.” In A Decade after the Global Recession: Lessons and Challenges for Emerging and Beegle, K., and L. Christiaensen. 2019. Accelerating Developing Economies, edited by M. A. Kose and F. Poverty Reduction in Africa. Washington, DC: World Ohnsorge. Washington, DC: World Bank. Bank. Arteta, C., M. A. Kose, M. Stocker, and T. Taskin. Bernard, A. B., and J. B. Jensen. 2004. “Exporting and 2016. “Negative Interest Rate Policies: Sources and Productivity in the USA.” Oxford Review of Economic Implications.” Policy Research Working Paper 7791, Policy 20 (3): 343-357. World Bank, Washington, DC. BIS (Bank for International Settlements). 2017. Atkin, D., A. K. Khandelwal, and A. Osman. 2017. “FinTech Credit: Market Structure, Business Models “Exporting and rm performance: Evidence from a and Financial Stability Implications.” Committee on Randomized Experiment.” e Quarterly Journal of the Global Financial System and the Financial Stability Economics 132 (2): 551-615. Board Report, Bank for International Settlements, Basel. Auerbach, A. J., and Y. Gorodnichenko. 2013. “Output Spillovers from Fiscal Policy.” American ———. 2019. Annual Economic Report. June. Basel: Economic Review 103 (3): 141-46. Bank for International Settlements. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 43 Bouakez, H., M. Guillard, and J. Roulleau- Pasdeloup. Choi, W. G., T. Kang, G.Y. Kim, and B. Lee. 2017. 2017. “Public Investment, Time to Build, and the Zero “Divergent Emerging Market Economy Responses to Lower Bound.” Review of Economic Dynamics 23 Global and Domestic Monetary Policy Shocks.” ADB (January): 60-79. Available at https://www.science economics working paper series 532, Asian direct.com/science/article/pii/S109420251630026. Development Bank, Manila, Philippines. Browder, G. J., S. Ozment, I. Rehberger Bescos, T. Christiaensen, L., P. Chuhan-Pole, and A. Sanoh. Gartner, and G. M. Lange. 2019. “Integrating Green 2013. Africa’s Growth, Poverty and Inequality Nexus– and Gray: Creating Next Generation Infrastructure.” Fostering Shared Prosperity. Washington, DC: World Washington, DC: World Bank Bank. Brunnermeier, M. K., and Y. Koby. 2019. “ e Christiaensen, L., and R. Hill. 2018. “Africa Is Not Reversal Interest Rate.” Princeton University, Poorer than Other Equally Poor Countries in Reducing Princeton, NJ. Poverty.” Background note prepared for Accelerating Bruno, V., and H. Shin. 2018. “Currency Depreciation Poverty Reduction in Africa, World Bank, and Emerging Market Corporate Distress.” BIS Washington, DC. Working Paper 753, Bank for International Chui, M., E. Kuruc, and P. Turner. 2016. “A New Settlements, Basel. Dimension to Currency Mismatches in e Emerging Buiter, W. H. 2014. “ e Simple Analytics of Markets: Non nancial Companies.” BIS Working Helicopter Money: Why it Works – Always.” Paper 550, Bank for International Settlements, Basel. Economics 8 (28). Available at https://papers.ssrn.com/ Clavey, C., J. L. Pemberton, J. Loeprick, and M. sol3/papers.cfm?abstract_id=2484853. Verhoeven. 2019. “International Tax Reform, Caballero, J., A. Fernández, and J. Park. 2019. “On Digitalization, and Developing Economies.” MTI Corporate Borrowing, Credit Spreads and Economic Discussion Paper 16, World Bank, Washington, DC. Activity in Emerging Economies: An Empirical Correa, P. G., and I. Guceri. 2013. “Tax Incentives for Investigation.” Journal of International Economics 118 Research and Development.” Innovation, Technology (C): 160-178. and Entrepreneurship Policy Note No. 4, World Bank, Caldara, D., M. Iacoviello, P. Molligo, A. Prestipino, Washington, DC. and A. Ra o. 2019. “ e Economic E ects of Trade Cusolito, A. P., and W. F. Maloney. 2018. Productivity Policy Uncertainty.” International Finance Discussion Revisited - Shifting Paradigms in Analysis and Policy. Papers 1256, Board of Governors of the Federal Washington, DC: World Bank. Reserve System, Washington, DC. Cerra, V., and S. C. Saxena. 2008. “Growth Dynamics: De la Fuente, A., H. G. Jacoby, and K. G. Lawin. e Myth of Economic Recovery.” American Economic 2019. “Impact of the West African Ebola Epidemic on Review 98 (1): 439–57. Agricultural Production and Rural Welfare Evidence from Liberia.” Policy Research Working Paper 8880, ________. 2017. “Booms, Crises, and Recoveries: A World Bank, Washington, DC. New Paradigm of the Business Cycle and Its Policy Implications.” IMF Working Paper 17/250, Doumbia, D., and M. L. Lauridsen. 2019. “Closing the International Monetary Fund, Washington, DC. SDG Financing Gap—Trends and Data.” World Bank IFC EM Compass Note 73, World Bank, Washington, Cerutti, E., S. Claessens, and L. Laeven. 2017. “ e DC. Use and E ectiveness of Macroprudential Policies: New Evidence.” Journal of Financial Stability 28 Dreger C., and H. E. Reimers. 2016. “Does Public (February): 203-224. Investment Stimulate Private Investment? Evidence for the euro area.” Economic Modelling 58 (November): Cerutti, E., S. Claessens, and D. Puy. 2019. “Push 154-158. Factors and Capital Flows to Emerging Markets: Why Knowing Your Lender Matters More an Easterly, W., and S. Fischer. 1994. “ e Soviet Fundamentals.” Journal of International Economics 119 Economic Decline: Historical and Republican Data.” (C): 133-149. NBER Working Paper 4735, National Bureau of Economic Research, Cambridge, MA. Chen, S., and J. S. Kang. 2018. “Credit Booms—Is China Di erent?” IMF Working Paper 18/2, Essl, S., S. K. Celik, P. Kirby, and A. Proite. 2019. International Monetary Fund, Washington, DC. “Debt in Low-Income Countries: Evolution, Implica- 44 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 tions, and Remedies.” Policy Research Working Paper ———. 2019. “Policy Implications of Non-linear 8950, World Bank, Washington, DC. E ects of Tax Changes on Output.” Policy Research Working Paper 8720, World Bank, Washington, DC. Fajgelbaum, P. D., P. K. Goldberg, P. J. Kennedy, and A. K. Khandelwal. 2019. “ e Return to Protec- Handley, K., and N. Limão. 2015. “Trade and tionism,” NBER Working Paper 25638, National Investment Under Policy Uncertainty: eory and Bureau of Economic Research, Cambridge, MA. Firm Evidence.” American Economic Journal: Economic Policy 7(4): 189-222. FAO (Food and Agriculture Organization of the United Nations). 2019. Crop Prospects and Food Heintz, J. 2010. “ e Impact of Public Capital on the Situation. July. Rome, Italy: FAO. US Private Economy: New Evidence and Analysis.” International Review of Applied Economics, 24 (5): Federal Reserve Board. 2019. Financial Stability Report. 619-632. November. Board of Governors of the Federal Reserve System, Washington, DC. Huidrom, R., M. A. Kose, J. J. Lim, and F. Ohnsorge. Fernandes, A., H. L. Kee, and D. Winkler. 2019. 2019. “Why Do Fiscal Multipliers Depend on Fiscal “Factors A ecting Global Value Chain Participation Positions?” Journal of Monetary Economics. Advance Across Countries.” Unpublished working paper, World online publication. https://doi.org/10.1016/j.jmoneco. Bank, Washington, DC. 2019.03.004. Feyen, E., and I. Zuccardi. 2019. “ e Sovereign-Bank Ianchovichina, E., and D. Leipziger. 2019. Nexus in EMDEs: What Is It, Is It Rising, and What “Combining Growth and Gender Diagnostics for the Are the Policy Implications?” Policy Research Working Bene t of Both.” Policy Research Working Paper 8847, Paper 8950, World Bank, Washington, DC. World Bank, Washington, DC. FSB (Financial Stability Board). 2019a. “Vulnerabilities IEA (International Energy Agency). 2019. Oil Market Associated with Leveraged Loans and Collateralised Report. September. Paris, France: IEA. Loan Obligations.” FSB, Basel. Ignatenko, A., F. Raei, and B. Mircheva. 2019. “Global ———. 2019b. “ ree Priorities for International Value Chains: What are the Bene ts and Why Do Regulatory and Supervisory Cooperation.” Speech by Countries Participate?” IMF Working Paper 19/18, FSB Secretary General Dietrich Domanski at Euro , International Monetary Fund, Washington, DC. Helsinki, September 13. IMF (International Monetary Fund). 2011. “Assessing Frankel, J. 2010. “ e Natural Resource Curse: A Reserve Adequacy.” International Monetary Fund, Survey.” NBER Working Paper 15836, National Washington, DC. Bureau of Economic Research, Cambridge, MA. ———. 2019a. Fiscal Monitor: How to Mitigate Fuglie, K., M. Gautam, A. Goyal, and W. Maloney. Climate Change. October. Washington, DC: 2019. Harvesting Prosperity: Technology and Productivity International Monetary Fund. Growth in Agriculture. World Bank, Washington, DC. ———. 2019b. Global Financial Stability Report: Gaspar, V., L. Jaramillo, and P. Wingender. 2016. Lower for Longer. October. Washington, DC: “Tax Capacity and Growth: Is ere a Tipping Point?” International Monetary Fund. IMF Working Paper 16/234, International Monetary Fund, Washington, DC. IMF and World Bank 2017. China – Financial Sector Assessment. Washington, DC: International Monetary Giordano, R., S. Lanau, P. Tommasino, and P. Fund and World Bank. Topalova. 2015. "Does Public Sector Ine ciency Constrain Firm Productivity: Evidence from Italian IPCC (Intergovernmental Panel on Climate Change). Provinces." IMF Working Paper 15/168, International 2018. Global Warming of 1.5°C., edited by V. Masson- Monetary Fund, Washington, DC. Delmotte, P. Zhai, H. O. Pörtner, D. Roberts, J. Skea, P. R. Shukla, A. Pirani, et al. Geneva: Inter- Gunter, S., D. Riera-Crichton, C. A. Vegh, and G. J. governmental Panel on Climate Change. Vuletin. 2018. “Non-Linear E ects of Tax Changes on Output: e Role of the Initial Level of Taxation.” Koh, W. C., and S. Yu. 2019. “Macroeconomic and Policy Research Working Paper 8668, World Bank, Financial Sector Policies.” In A Decade after the Global Washington, DC. Recession: Lessons and Challenges for Emerging and G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 45 Developing Economies, edited by M. A. Kose, and F. Comparative Information on OECD and Other Advanced Ohnsorge. Washington, DC: World Bank. and Emerging Economies. Paris: OECD Publishing. Kose, M. A., S. Kurlat, F. Ohnsorge, and N. Sugawara. ———. 2019a. Tax Policy Reforms 2019: OECD and 2017. “A Cross-Country Database of Fiscal Space.” Selected Partner Economies. Paris: OECD Publishing. Policy Research Working Paper 8157, World Bank, ———. 2019b. e OECD Technical Report on Progress Washington, DC. on Structural Reform under the G20 Enhanced Structural Kose, M. A., and F. Ohnsorge. 2019. A Decade after the Reform Agenda (ESRA). Paris: OECD Publishing. Global Recession: Lessons and Challenges for Emerging Ohnsorge, F., M. Stocker, and Y. M. Some. 2016. and Developing Economies. Washington, DC: World “Quantifying Uncertainties in Global Growth Bank. Forecasts.” Policy Research Working Paper 7770, Kose, M. A., N. Sugawara, and M. E. Terrones. 2019. World Bank, Washington, DC. “What Happens During Global Recessions?” In A Packard, T. G., U. Gentilini, M. E. Grosh, P. B. Decade after the Global Recession: Lessons and Challenges O’Keefe, R. J. Palacios, D. A. Robalino, and I. V. for Emerging and Developing Economies, edited by M. A. Santos. 2019. Protecting All: Risk Sharing for a Diverse Kose, and F. Ohnsorge. Washington, DC: World and Diversifying World of Work. Washington, DC: Bank. World Bank. Kose, M. A., and M. E. Terrones. 2015. Collapse and Pathikonda, V., and T. Farole. 2016. “ e Capabilities Revival: Understanding Global Recessions and Recoveries. Driving Participation in Global Value Chains.” Policy Washington, DC: International Monetary Fund. Research Paper 7804, World Bank, Washington, DC. Krishnamurti, D., and Y. C. Lee. 2014. Macro- Prichard, W., A. L. Custers, R. Dom, S. R. Davenport, prudential Policy Framework: A Practice Guide. M. A. Roscitt. 2019. “Innovations in Tax Compliance: Washington, DC: World Bank. Conceptual Framework.” Policy Research Working Paper 9032, World Bank, Washington, DC. Laeven, L., and F. Valencia. 2018. “Systemic Banking Crises Revisited.” IMF Working Paper 18/206, Rachel, L., and L. H. Summers. 2019. “On Secular International Monetary Fund, Washington, DC. Stagnation in the Industrialized World.” Brookings Papers on Economic Activity (Spring). Martínez‐Galán, E., and M. P. Fontoura. 2019. “Global Value Chains and Inward Foreign Direct Ray, D., and J. Esteban. 2017. “Con ict and Investment in the 2000s.” e World Economy 42 (1): Development.” Annual Review of Economics 9 (1): 175-196. 263–93. Mertens, T. M., and J. C. Williams. 2019. “Monetary Rozenberg, J., and M. Fay. 2019. Beyond the Gap: How Policy Frameworks and the E ective Lower Bound on Countries Can A ord the Infrastructure ey Need While Interest Rates.” Federal Reserve Bank of New York Protecting the Planet. Washington, DC: World Bank. Sta Reports. https://www.newyorkfed.org/medialibra Ruch, F. 2019a. “Prospects, Risks, and Vulnerabilities.” ry/media/research/sta _reports/sr877.pdf. In A Decade after the Global Recession: Lessons and Molyneux, P., A. Reghezza, and R. Xie. Forthcoming. Challenges for Emerging and Developing Economies, “Bank Pro ts and Margins in a World of Negative edited by M. A. Kose, and F. Ohnsorge. Washington, Rates.” Journal of Banking and Finance. DC: World Bank. Munoz, E., and E. Olaberria. 2019. “Are Budget ———. 2019b. “Policy Challenges.” In A Decade after Rigidities a Source of Fiscal Distress and a Constraint the Global Recession: Lessons and Challenges for Emerging for Fiscal Consolidation?” Policy Research Working and Developing Economies, edited by M. A. Kose, and F. Paper 8956, World Bank, Washington, DC. Ohnsorge. Washington, DC: World Bank. NGFS (Network for Greening the Financial System). Sarin, N., and L. H. Summers. 2019. “Shrinking the 2018. NGFS First Progress Report. October. Report Tax Gap: Approaches and Revenue Potential.” NBER coordinated by the NGFS Secretariat/Banque de Working Paper 26475, National Bureau of Economic France, Paris. Research, Cambridge, MA. OECD (Organisation for Economic Co-operation and Shiller, R. 2015. Irrational Exuberance. Princeton Development). 2017. Tax Administration 2017: University Press, Princeton, NJ. 46 CHAPTER 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 Steinbach, M. R. 2019. “Growth in Low-Income Trap: International eory and Evidence”, 2010 Countries: Evolution, Prospects, and Policies.” World Meeting Papers 717, Society for Economic Dynamics. Bank Policy Research Paper 8949, World Bank, Washington, DC. Williams, J. C. 2016. “Monetary Policy in a Low R-star World,” FRBSF Economic Letter, Federal Reserve Stocker, M., J. Ba es, Y. Some, D. Vorisek, and C. M. Bank of San Francisco. Wheeler. 2018. " e 2014-16 Oil Price Collapse in Retrospect: Sources and Implications." Policy Research Woodford, M. 2011. “Simple Analytics of the Working Paper 8419, World Bank, Washington, DC. Government Expenditure Multiplier.” American Economic Journal: Macroeconomics 3 (1): 1-35. Stokey, N. L. 2016. “Wait-and-See: Investment ———. 2012. “Methods of Policy Accommodation at Options Under Policy Uncertainty.” Review of the Interest-Rate Lower Bound.” Proceedings, Economic Dynamics 21 (July): 246-265. Economic Policy Symposium, Jackson Hole, Federal Syverson, C. 2011. “What Determines Productivity?” Reserve Bank of Kansas City: 185-288. Journal of Economic Literature 49 (2): 326–65. World Bank. 2015. Global Economic Prospects: e Taglioni, D., and D. Winkler. 2016. Making Global Global Economy in Transition. June. Washington, DC: Value Chains Work for Development. Washington, DC: World Bank. World Bank. ———. 2016a. Global Economic Prospects: Spillovers amid Weak Growth. January. Washington, DC: World UNCTAD (United Nations Conference on Trade and Bank. Development). 2019a. World Investment Report 2019: Special Economic Zones. Geneva, Switzerland: ———. 2016b. World Development Report 2016. UNCTAD. Digital Dividends. Washington, DC: World Bank. ———. 2019b. Global Investment Trend Monitor No. ———. 2017a. Global Economic Prospects: Weak 32. October. Geneva, Switzerland: UNCTAD. Investment in Uncertain Times. January. Washington, DC: World Bank. Van Eyden, R., M. Difeto, R. Gupta, and M. E. Wohar. 2019. “Oil Price Volatility and Economic ———. 2017b. e Distributional Impact of Taxes and Growth: Evidence from Advanced Economies using Transfers: Evidence from Eight Developing Countries. more than a Century’s Data.” Applied Energy 233-234 Washington, DC: World Bank. (January): 612-621. ———. 2017c. World Development Report 2017. Verme, P., and A. Araar. 2017. e Quest for Subsidy Governance and the Law. Washington, DC: World Reforms in the Middle East and North Africa Region: A Bank. Microsimulation Approach to Policy Making. Cham, Switzerland: Springer. ———. 2018a. Global Economic Prospects: Broad-Based Upturn, but for How Long? January. Washington, DC: Vorisek, D., and S. Yu. Forthcoming. “Understanding World Bank. the Cost of Achieving the Sustainable Development Goals: e World Bank Group’s Contributions.” ———. 2018b. Global Economic Prospects: e Turning Policy Research Working Paper, World Bank, of the Tide? June. Washington, DC: World Bank. Washington, DC. ———. 2018c. Poverty and Shared Prosperity Report 2018: Piecing Together the Poverty Puzzle. Washington, Wannier, S.R., L. Worden, N.A Ho , E. Amezcua, B. DC: World Bank. Selo, C. Sinai, M. Mossoko, et al. 2019. “Estimating the Impact of Violent Events on Transmission in Ebola ———. 2018d. World Development Report 2019: e Virus Disease Outbreak, Democratic Republic of the Changing Nature of Work. Washington, DC: World Congo, 2018–2019.” Epidemics 28 (September): 1-9. Bank. Whalen, C., and F. Reichling. 2015. “ e Fiscal ———. 2018e. China—Systematic Country Diagnostic: Multiplier and Economic Policy Analysis in the United Towards a More Inclusive and Sustainable Development. States.” Congressional Budget O ce Working Paper Washington, DC: World Bank. Series. Congressional Budget O ce, Washington, DC. ———. 2019a. World Development Report 2020: Wieland, J. 2010. “Fiscal Multipliers in the Liquidity Global Value Chains: Trading for Development in the G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 1 47 Age of Global Value Chains. Washington, DC: World ———. 2019n. In ation in Emerging and Developing Bank. Economies: Evolutions, Drivers and Policies. Washington, DC: World Bank. ———. 2019b. East Asia and Paci c Economic Update: Weathering Growing Risks. October. Washington, DC: ———. 2019o. Global Financial Development Report World Bank. 2019/2020: Bank Regulation and Supervision a Decade after the Global Financial Crisis. Washington, DC: ———. 2019c. China Economic Update: Cyclical Risks World Bank. and Structural Imperatives. December. Washington, DC: World Bank. ———. 2019p. Fiscal Policies for Development and Climate Action. Washington, DC: World Bank. ———. 2019d. “Migration and Remittances: Recent Developments and Outlook.” Migration and ———. 2020. Doing Business 2020: Comparing Development Brief 31. Washington, DC: World Bank. Business Regulation in 190 Economies. Washington, DC: World Bank. ———. 2019e. Global Economic Prospects: Heightened Tensions, Subdued Investment. June. Washington, DC: World Bank and DRC (Development Research Center World Bank. of the State Council, the People's Republic of China). 2019. Innovative China: New Drivers of Growth. ———. 2019f. South Asia Economic Focus: Making Washington, DC: World Bank. (De)centralization Work. Fall. Washington, DC: World Bank. World Bank and IMF (International Monetary Fund). 2018. Debt Vulnerabilities in IDA Countries. ———. 2019g. Europe and Central Asia Economic Washington, DC: World Bank and International Update: Migration and Brain Drain. October. Monetary Fund. Washington, DC: World Bank. World Bank et al. 2015. Options for Low Income ———. 2019h. Africa’s Pulse: An Analysis of Issues Countries E ective and E cient Use of Tax Incentives for Shaping Africa’s Economic Future. October. Investment: A Report to the G-20 Development Working Washington, DC: World Bank. Group by the IMF, OECD, UN and World Bank. ———. 2019i. Middle East and North Africa Economic Washington, DC: World Bank. Update. October. Washington, DC: World Bank. World Bank et al. 2017. Global Value Chain ———. 2019j. Fiscal Policies for Development and Development Report: Measuring and Analyzing the Climate Action. Washington, DC: World Bank. Impact of GVCs on Economic Development. A report by IDE-JETRO, OECD, UIBE, WTO, and World Bank. ———. 2019k. Malawi Economic Monitor: Charting a Washington, DC: World Bank. New Course. June. Washington, DC: World Bank. World Trade Organization (WTO). 2019a. Services ———. 2019l. International Debt Statistics 2020. Trade Barometer. Geneva: World Trade Organization. Washington, DC: World Bank. ———. 2019b. Report of e TPRB From the Director- ———. 2019m. Global Economic Prospects: Darkening General on Trade-Related Developments. Geneva: World Skies. January. Washington, DC: World Bank. Trade Organization. SPECIAL FOCUS 1 Price Controls: Good Intentions, Bad Outcomes G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 S P EC IAL FO CU S 1 51 The use of price controls is widespread across emerging markets and developing economies (EMDEs), including for food and key imported and exported commodities. While sometimes used as a tool for social policy, price controls can dampen investment and growth, worsen poverty outcomes, cause countries to incur heavy fiscal burdens, and complicate the effective conduct of monetary policy. Replacing price controls with expanded and better-targeted social safety nets, coupled with reforms to encourage competition and a sound regulatory environment, can be both pro-poor and pro-growth. Such reforms need to be carefully communicated and sequenced to ensure political and social acceptance. Where they exist, price control regimes should be transparent and supported by well-capitalized stabilization funds or national hedging strategies to ensure fiscal sustainability. Introduction • How prevalent are price controls in EMDEs? Price distortions are defined as instances “when • What challenges do they impose for growth prices and production are higher or lower than the and development and government policies? level that would usually exist in a competitive Contribution. The research adds to the literature market” (WTO 2019). One source of such on price controls in two ways. First, it presents distortions are price controls.1 Price controls can findings from a new data set. Whereas earlier work be imposed in a variety of ways. They may involve is confined to advanced economies or selected price ceilings, or price floors, imposed on selected emerging markets, this study covers an almost goods and services by the authorities.2 complete set of EMDEs.3 Second, it reviews price In emerging market and developing economies controls on a wider range of goods.4 (EMDEs), price controls on goods are often imposed to serve social and economic objectives. Use of price controls They may be part of government efforts to protect vulnerable consumers, by addressing market Widespread price controls in EMDEs. Price failures or subsidizing the cost of essential goods. controls are widely employed across advanced Or they may be intended to maintain the incomes economies and EMDEs. They tend to be more of producers, as part of a price-support program. pervasive in EMDEs than in advanced economies Alternatively, they can serve the purpose of price smoothing, especially for key commodities subject to high volatility in international markets. This 3 The data set extracts the list of products subject to price controls from the latest available Trade Policy Reviews for each can lower uncertainty about households’ real EMDE member country of the World Trade Organization. This list incomes and firms’ production costs. of products is compiled using existing legislation and additional material provided by country authorities. The data set provides a This special focus seeks to answer two questions. rough view of the prevalence of price control measures across countries, but does not include any information on the extent of these controls. 4 The micro-founded theory of price controls was developed in Note: This Special Focus was prepared by Justin-Damien part to examine the case of commodity producers in developing Guénette. countries (Stiglitz and Newbery 1979; Newbery and Stiglitz 1982). 1 Price controls have a long history with well documented More recently, for EMDEs, price controls for petroleum products examples stretching back to Revolutionary France (Morton 2001). In have been studied extensively, while those on food products have the 20th century, these policies were used extensively in several received less attention (Verme and Araar 2017; Kojima 2013; western countries during the Second World War, culminating with Devarajan 2013; Murphy et al. 2019; Shi and Sun 2017; Clements, widespread controls in the United States and the United Kingdom in Jung and Gupta 2007; Ghosh and Whalley 2004). The World the 1970s (Coyne and Coyne 2015). Price controls were also Bank’s Energy Sector Management Assistant Program (ESMAP) has ubiquitous in communist countries with planned economies, such as conducted in-depth studies of subsidy reforms for energy markets Poland (Tarr 1994). Generalized price controls fell out of favor in the across EMDEs (ESMAP 2019; Ore et al. 2018). The use of price 1980s, as inflation declined, and governments pursued deregulation. controls for pharmaceutical products, wages and rent has been widely However, controlled pricing for certain goods and services, including studied in advanced economies (e.g., Coyne and Coyne 2015; rent and pharmaceuticals, remain in use to this day (Morton 2001). Nguyen et al. 1994). Studies for individual EMDEs include China, 2 Government management of prices can also occur as a by- Indonesia and several MENA countries (Shi and Sun 2017; product of other policies. For instance, preferential exchange rates for Clements, Jung and Gupta 2007; Verme and Araar 2017). The certain goods and the imposition of non-tariff barriers can all push OECD-WBG Product Market Regulation indicators provide prices away from that which would prevail in a competitive market. summary statistics on price controls for a limited set of EMDEs. 52 S P EC IAL FO CU S 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 (OECD and World Bank 2018). And among Price controls on traded goods. EMDEs, EMDEs, they are more prevalent in LICs (Figure including LICs, apply price controls on export and SF1.1). In EMDEs that have become middle- import goods.7 Governments often impose income countries (MICs) since 2001, price controls on the domestic prices of imports to controls are somewhat less common than in the maintain real incomes of domestic consumers, average EMDE, especially in goods other than hold down costs to producers, or smooth domestic energy, food, and construction materials.5 price volatility. • Energy. Virtually all EMDEs impose price • Energy imports. In LICs, about 67 percent of controls on energy products, including energy imports—about 6 percentage points electricity and petroleum products such as more than the average for other EMDEs—are liquified petroleum gas and gasoline. potentially subject to domestic price controls (Figure SF1.2).8 • Food. Price controls are frequently applied to basic foodstuffs. This practice is more • Food imports. In both LICs and other widespread in LICs than in other EMDEs: EMDEs, only a small share of food and virtually all LICs impose price controls on beverages imports are potentially subject to some food items, compared with three- controls. quarters of other EMDEs. Products often subject to price controls include water, sugar, • Construction material imports. The largest and rice.6 Since food expenditures represent difference between LICs and other EMDEs nearly 60 percent of the consumption basket lies in the share of construction-related in LICs, compared with 42 percent in other imports that are potentially subject to price EMDEs, a larger portion of the LICs basket is controls: in LICs, they amount to one-quarter typically subject to price controls (Laborde, of imported construction materials, compared Lakatos, and Martin 2019). Virtually all LICs with almost none in other EMDEs. and other EMDEs impose price controls on petroleum products. Commodity exports. LICs often impose price controls on exportable commodities. This may • Construction materials. Nearly 20 percent of involve a monopoly marketing agency, which LICs impose price controls on construction purchases from domestic producers at a fixed materials. These include cement, reinforcing price, and resells to foreign purchasers at the world bars, and metal sheets. Beyond LICs, controls price. This arrangement implicitly taxes producers on construction materials are most common when the resale price exceeds the purchase price in the Middle East and North Africa (MNA) (Ghosh and Whalley 2004) or subsidizes and Sub-Saharan Africa (SSA). producers when the resale price falls below the purchase price. About 25 percent of EMDEs that 5 The set of LICs in 2001 that are now MICs includes Angola, rely heavily (with more than 10 percent of goods Armenia, Azerbaijan, Bangladesh, Bhutan, Côte d’Ivoire, Cameroon, exports) on a single export commodity group Republic of Congo, Comoros, Georgia, Ghana, Indonesia, India, impose price controls on it. For example, Burundi Kenya, Kyrgyz Republic, Cambodia, People’s Democratic Republic of Lao, Lesotho, Republic of Moldova, Myanmar, Mongolia, imposes controls on the price of coffee while Mauritania, Nigeria, Nicaragua, Pakistan, Sudan, Senegal, Solomon Benin imposes controls on cashew nuts. Islands, São Tomé and Príncipe, Turkmenistan, Ukraine, Uzbekistan, Vietnam, Zambia and Zimbabwe. 6 Almost all LICs, including Ethiopia, Mali, Niger, Guinea and Rwanda impose some form of price controls on petroleum products. 7 Unregulated prices depend on the world price, transport costs, As for food products, LICs such as Burkina Faso and the Democratic Republic of Congo impose price controls on sugar. Chad, Haiti and local monopoly power or other hurdles to the movement of goods, Guinea-Bissau impose controls on rice, and Benin, Ethiopia and and harvest conditions (Aksoy and Ng 2010). 8 Data on price controls on tradable goods combines the Niger impose controls on bread. Burkina Faso imposes controls on cement, reinforcing bars and metal sheets. In addition to goods, price information on controlled prices from the World Trade controls are also often imposed on public transportation services such Organization’s Trade Policies Reviews with 4-digit HS trade values as bus, train, and ship fares. from the World Bank’s World Integrated Trade Solutions database. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 S P EC IAL FO CU S 1 53 Price controls on financial services. While not FIGURE SF1.1 Price controls covered in the price control data set, the financial LICs use price controls more extensively than other EMDEs, especially for sector is also often a target of price controls. energy products such as petroleum and electricity, and basic foodstuffs Around 60 EMDEs have imposed ceilings on such as cereal products and sugar. A large portion of the LICs consumption basket is subject to price controls given the elevated share of interest rates. These measures are often motivated food in the consumption bundle. Across EMDEs more broadly, price by a desire to provide targeted support to strategic controls are most prevalent in MENA and Sub-Saharan Africa and least prevalent in South Asia. industries or to shield consumers from financial exploitation. For example, in the case of Zambia, A. Economies with price controls B. Goods most frequently subject controls were implemented from 2012 to 2015 to Price Controls in LICs to reduce the perceived risk of over indebtedness and broaden access to credit (Maimbo and Gallegos 2014). Decline in price controls. Starting in the 1980s, several EMDEs reduced the scope of price controls, opting instead to strengthen their competition policies and regulation (WTO 2000- 2019). In some cases, the liberalization of prices was supported and encouraged by policy lending C. Share of food in total consumption D. Economies with price controls programs and debt relief efforts in highly indebted expenditure by sub-region poor countries (HIPC). The removal of controls often become more feasible following an easing of the conditions that led to their imposition. For example, after 2011, as food prices declined from cyclical highs, some countries eliminated controls. EMDEs such as Mexico, Rwanda, and Côte d’Ivoire took advantage of the sharp decline in oil prices in 2014-16 to reduce petroleum subsidies (Baffes et al. 2018; Stocker et al. 2015). Source: World Bank; World Trade Organization. Note: EMDEs = emerging markets and developing economies; LICs = low-income countries. A. B. D. Listed price control policies are retrieved from the latest (2003-19) country Trade Policy Reforms in MENA. Under pressure from social Review publication. A. C. D. Unweighted averages. tensions during the Arab Spring, some countries A. Sample includes 21 low-income countries, 23 LICs turned middle-income countries (MICs) since 2001, and 56 other EMDEs. in the region introduced or tightened food price B. Sample includes 21 low-income countries. controls in 2011 (Ianchovichina, Loening and C. Sample includes 23 low-income countries and 67 other EMDEs. D. Sample includes 21 low-income countries and 79 other EMDEs. EAP = East Asia and Pacific, Wood 2014). Conversely, however, high oil prices ECA = Europe and Central Asia, LAC = Latin America and the Caribbean, MNA = Middle East and North Africa, SAR = South Asia, SSA = Sub-Saharan Africa. and fiscal pressures encouraged a few MENA Click here to download data and charts. countries, including the Arab Republic of Egypt, Morocco, and Tunisia, to reform price controls and related subsidies on energy between 2010 and 2014 (Verme and Araar 2017).9 The reforms were programs, however, differed substantially in their associated with improvements in the ease of doing scope, and speed of implementation. They also business. Within two years of the reform, varied with respect to compensatory transfers to enterprises in all three countries reported easier disadvantaged population groups. Morocco access to electricity (Figure SF1.2.D). The reduced the fiscal burden of petroleum subsidies, while at the same time avoiding severe adverse consequences for poverty and inequality. Egypt, 9 Djibouti, Egypt, Jordan, Libya, Morocco, Tunisia, and Yemen however, took a sequential, gradual, approach to designed and implemented subsidy reform programs. These cases reform especially for products such as liquified contrast with some other countries in the region, where social tensions during the Arab Spring, caused an increased use of food petroleum gas (LPG), which account for a price controls in 2011 (Ianchovichina, Loening, and Wood 2014). disproportionately large expense for the poor. 54 S P EC IAL FO CU S 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE SF1.2 Price controls on imported and exported measures for the poor, including expanding goods food subsidies (Verme and Araar 2017). The shares of imports and exports potentially covered by price controls are Moreover, the government used a share of the higher in LICs than in other EMDEs. Reforms of price controls on energy proceeds from the reforms to increase products in Egypt, Morocco and Tunisia were associated with expenditures on health care and education improvements in an index of the ease of getting electricity within two years of energy price reforms. provision (ESMAP 2017a). However, attempts to communicate to the affected A. Share of total imports subject to B. Share of 2-digit HS category public that they might eventually benefit from price controls imports subject to price controls the diversion of energy subsidies to more equitable uses failed, largely because the country does not have the social security net to implement an effective system of cash compensation (Verme and Araar 2017). • Morocco. Starting in 2013, the government first transitioned to price indexation for petroleum products, and gradually moved to fully liberalize most energy products. In C. Share of countries with price con- D. Ease of getting electricity trols on export goods August 2014, prices of household utilities jumped as part of a multiyear effort to liberalize electricity prices. The reforms were implemented without triggering social unrest despite the absence of cash transfers to households. The fiscal savings from the reform were instead used to fund other reforms. Source: World Bank; World Trade Organization. • Tunisia. The fiscal cost of Tunisia’s energy Note: EMDEs = emerging markets and developing economies; LICs = low-income countries. subsidies had risen to unsustainable levels (7 A.-C. 2017 data. Listed price control policies are retrieved from the latest (2003-19) country Trade Policy Review publication. percent of GDP in 2013), and in response the A. B. Sample includes 12 low-income countries and 63 other EMDEs. government gradually reduced them B. Share of 4-digit Harmonized System (HS) category subject to controlled prices in high-level groupings of 2-digit HS categories. Construction materials aggregate includes HS68 and HS73, beginning in late 2012 in tandem with Energy aggregate includes HS27, Food and Beverage aggregate includes HS01 to HS22. Other aggregate includes all other imports. reforms to social benefits. Petroleum and C. Countries that rely heavily on a single export defined as a country in which exports of one or more 4-digit HS category represents 10 percent or more of its total exports in 2017. Chart shows the share electricity prices were increased over 2012-13 of all LICs and other EMDEs that relying heavily on a single export whose price is subject to price controls. Sample includes 12 low-income countries and 61 other EMDEs. and an automatic price formula was D. Chart shows the World Bank’s index for ease of getting electricity in year before (t-1) and the two introduced for gasoline in 2014. In 2016, the years after (t+1, t+2) energy subsidy reform (World Bank 2019b). Time t=0 refers to 2014 for Egypt and Morocco and 2012 for Tunisia. government agreed to further reduce subsidies Click here to download data and charts. as part of a reform program supported by IMF lending. Energy prices were increased several times since then, with the goal of fully • Egypt. In July 2014, comprehensive reforms to eliminating energy subsidies by 2022. Over fuel and electricity prices resulted in a the years, measures were implemented to significant rise in gasoline, natural gas, diesel, cushion the impact of reforms on vulnerable and electricity prices which contributed to a households, including expanded social spurt of headline inflation. Initial price housing and higher income tax deductions. adjustments were followed by stepwise gradual increases to fully eliminate energy subsidies Reforms in other regions. In Ukraine in 2015-16, over a five-year period. While the initial price the government raised the price of natural gas, increases themselves are estimated to have which had been heavily subsidized for decades. raised the poverty rate and inequality, the These reforms were coupled with a strong public government has put in place some mitigating communication campaign highlighting social G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 S P EC IAL FO CU S 1 55 assistance mechanisms targeted to cushion the they can discourage foreign investment in impact on low-income households. The reforms those sectors by increasing the country risk were successful in allowing public utilities to premium facing global firms (Sabal 2005). In achieve cost recovery, with the targeted support the opposite case, where the controlled price is measures estimated to have reduced the poverty above that required for a competitive return to rate (ESMAP 2017b). In India starting in 2012, investment, its maintenance requires barriers the government reformed its subsidy regime for to entry or costly government stockpiling of liquified petroleum gas (LPG). LPG subsidies to excess supply (a common occurrence with households encouraged the formation of black price support schemes in agriculture). Price- markets where subsidized LPG distributed to support controls can depress competition and households was diverted to the commercial sector. sustain high producer margins (e.g., Rwanda’s The government gradually increased the price of transportation sector; Teravaninthorn and LPG for households while implementing a large- Raballand 2009). scale targeted cash transfer mechanism. The program successfully eliminated distortions in the • Lower productivity. Price control regimes may LPG market, with limited adverse consequences tilt the allocation of resources towards the for the poor, and the fiscal savings obtained from subsidized sector. In LICs, this is often most the reduction in subsidies fully offset the costs of visible in the agricultural sector where output the targeted cash transfer (ESMAP 2016). price controls have been complemented by input (especially fertilizer) subsidies. Yet, such Challenges of price controls policies can end up reducing productivity, and worsening income inequality (Goyal and Nash While they may be introduced with the best 2017). They may lead to inefficient use of intentions to improve social outcomes, price subsidized inputs (Jayne, Mason, Burke and controls often undermine growth and Ariga 2016). They can also adversely affect development, impose fiscal burdens and can incentives to adopt productivity-raising new weaken the effectiveness of monetary policy. At technologies. Empirical evidence suggests that least in part, this is because price controls cause a market-oriented structural reforms, including shift in consumption towards the subsidized good, the reduction of price controls and their and away from other non-subsidized goods. related subsidies, are strongly associated with Moreover, when there are trend increases in improved firm-level productivity in EMDEs international prices, or when they interact with (Kouame and Tapsoba 2018). Conversely, in barriers to entry, price control measures frequently the case of petroleum products in the Middle morph into distortive subsidy regimes. Important East and North Africa, high subsidies that social, fiscal and environmental costs are likely to underpin price controls appear to be follow, as well as adverse consequences for associated with lower per capita output investment and employment, and productivity growth (Mundaca 2017). growth. • Increased informality. Price controls that Growth challenges. The use of price controls can distort consumption towards price-controlled have adverse consequences for growth for several goods, can cause chronic shortages of these reasons: goods, the formation of parallel markets with higher prices, and substitution towards lower- • Stifled competition and reduced investment. quality alternatives (Weitzman 1991; Patel Price ceilings can depress producer margins and Villar 2016; Fengler 2012; Winkler and discourage domestic investment and 2015). Similarly, producers of price-controlled entrepreneurial activity, as in Zimbabwe’s goods may turn to black markets which have transportation sector (Newfarmer and Pierola elevated transaction costs and lack basic 2015). If margins depend on subsidies to local regulation (Murphy, Pierru and Smeers businesses to compensate for price controls, 2019). In addition, the situation encourages 56 S P EC IAL FO CU S 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 production to shift to firms in the informal subject to price controls, such as petroleum, can sector, which avoid regulation (De Soto 2000; be a large portion of government expenditures, in World Bank 2019a). some cases exceeding 10 percent of GDP (Algeria, Iran; World Bank 2014b). • Distorted financial markets. Price controls in the financial sector, such as ceilings on interest Monetary policy challenges. In all advanced rates can distort financial markets (Maimbo economies, and in many EMDEs, monetary policy and Gallegos 2014). These measures reduce has played a major role in reducing inflation to a the supply of credit to safer borrowers and low, stable rate, often in the context of an explicit small and medium-sized enterprises, increase inflation-targeting regime. The key has been a the level of non-performing loans, reduce transparent strategy aimed at the medium and competition and innovation in lending longer term. This has largely stabilized longer-run markets, and increase informal lending. expectations of inflation, in line with central bank Moreover, they can exacerbate inequality by objectives. In these circumstances, the one-off limiting the poor’s access to lending. impact on the inflation rate of the removal of price controls can be handled with the help of careful • Increased vulnerability to climate change. Price communication from policymakers as to the controls and subsidies on energy products may strategy they will employ to get inflation back on heighten vulnerability to climate change and track. In LICs, however, the monetary policy inhibit the transition to a climate-resilient, challenges go deeper. First, the wider use of price low-carbon economy. controls complicates the choice of inflation target by weakening the usefulness of the overall CPI as a Social policy and political economy challenges. measure of underlying inflation pressures (Patel The use of price controls combined with large and Villar 2016).10 Second, it can raise inflation subsidies is an inefficient tool for redistributing because the authorities tend to respond domestic income (Devarajan 2013; Coyne and asymmetrically when faced with cost increases, as Coyne 2015). These policies tend to be is often the case in response to food price shocks inequitable, as wealthier segments of the (De Mello 2008; Ianchovichina, Loening and population, usually urban consumers, benefit Wood 2012). Third, it can increase the stickiness disproportionately given their greater consump- of the inflation process as changes in controlled tion of the price-controlled good compared to prices often involve a lengthy regulatory process rural consumers and producers. For example, (Springer de Freitas and Bugarin 2007). Fourth, subsidies and below-market prices for gasoline and one-off changes in controlled prices can have liquid natural gas have proven highly regressive, persistent effects on inflation in LICs, where with only a small share of the subsidy benefiting inflation expectations are less well anchored (Ha, the poorest segments of the population (Baffes et Kose, and Ohnsorge 2019a; BIS 2003). Lastly, al. 2015; IEG 2008; Coady et al. 2006). price controls in the financial sector, including ceilings on interest rates can reduce the ability of Fiscal challenges. Price controls impose an explicit monetary policy to affect financial conditions. or implicit set of taxes and subsidies that varies over time, and their enforcement may require Price controls in times of hyperinflation. The use additional regulations to constrain consumption of price controls has often coincided with and production. Typically, a system of price historical episodes of hyperinflation. In Brazil in controls on goods ends up as a growing burden on the 1980s, for example, the use of price controls either the fiscal budget and public debt or the profitability of producers (Alleyne 2013; World Bank 2014a). Potential or implicit fiscal costs from price controls can be particularly high in 10 In addition, volatility in headline CPI inflation is amplified by the high proportion of food in the LIC consumer basket. Food prices LICs due to their more widespread use of these are liable to frequent large fluctuations from variations in local policies. Even in EMDEs, subsidies for products harvests, and in international supply and demand. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 S P EC IAL FO CU S 1 57 has proved ineffective at addressing hyperinflation Comprehensive reforms of price control policies in Brazil (Cardoso 1991). More recently, in the and related subsidies. Replacing price controls case of Zimbabwe, widespread shortages of goods with expanded and better-targeted social safety in part due to excessively accommodative nets, coupled with structural reforms, can be both monetary policy were accompanied by extensive pro-poor and pro-growth. Indeed, policies to price controls (Munoz 2006; Coomer and lower subsidies that underpin price controls Gstraunthaler 2011). Similarly, high inflation in appear to be associated with higher per capita República Bolivariana de Venezuela was output growth, in part because savings generated accompanied by highly restrictive price controls by lower subsidies can fund productivity- (Vera 2017; Contreras and Guarata 2013). enhancing education and infrastructure (Mundaca 2017). The removal of price controls needs to be Collateral damage from foreign price controls. coupled with targeted support for those segments LICs are also more vulnerable to the collateral of the population that might be adversely affected damage from other countries’ price controls on (World Bank 2014a).11 In India, for example, the food and energy, because of the high share of food removal of price controls was accompanied by and energy in their consumption baskets and targeted cash transfers and in Brazil by targeted trade. Policies by individual countries to contain assistance to low-income households for energy the effects of spikes in global commodity process conservation (Deichmann and Zhang 2013).The in their local markets have been shown to have different prongs of reforms, however, need to be had the perverse effect of raising global prices carefully sequenced and communicated. (Laborde, Lakatos, and Martin 2019). Export restrictions in major commodity producers Enhanced competition. Improving the compet- exacerbate global shortages, thus contributing to itive environment can be a more effective means higher prices on the international market. In the of lowering costs to consumers and producers than case of the 2007-08 surge in food prices, a the use of price controls. Carefully-designed and majority of EMDEs put in place policies to properly enforced antitrust laws and consumer insulate domestic markets from the rise in protection legislation, are essential components of international prices (World Bank 2009). institutional frameworks that support market mechanisms. A sound legal and regulatory Policy implications framework favoring competitive markets provides a more effective response to many of the problems Price controls have been used to mitigate the that price controls attempt to address (Kovasic impact of commodity price volatility on the most 1995). For example, the removal of price controls vulnerable members of society. For instance, the and barriers to entry in the transportation sector use of temporary stabilization funds, as introduced significantly increased competition and lowered in Chile and Peru, or national hedging strategies, transportation costs in Rwanda (Teravaninthorn as introduced in Mexico, have been used to and Raballand 2009). Even in the case where protect domestic consumers and firms from spikes incumbent firms maintained outsized market in the prices of basic commodities on international shares, the presence of competition, and the markets (Kojima 2013; Ma and Valencia 2018). potential for new entrants, significantly lowered However, most governments have had difficulty their markups (World Bank 2006). designing frameworks that deliver lasting benefits. Over time, price stabilization policies often result in costly and distortionary subsidies, posing important challenges to growth, development, and macroeconomic policy suggesting that other policy instruments may be more effective in achieving 11 Despite the regressive nature of price controls and subsidies, poor households spend a higher share of their income on products social protection objectives. subject to price controls and are liable to suffer distressful real income losses when price restrictions are lifted (World Bank 2014a). 58 S P EC IAL FO CU S 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 De Soto, H.. 2000. The Mystery of Capital: Why References Capitalism Triumphs in the West and Fails Everywhere Else. New York: Basic Books. Aaron, H. 1966. “Rent Controls and Urban Development: A Case Study of Mexico City.” Social Devarajan, S., and M. Giugale. 2013. “The Case for and Economic Studies 15 (4): 314-328. Direct Transfers of Resources Revenues in Africa”. Aksoy, A., and F. Ng. 2010. Food Prices and Rural Working Paper 333, Center for Global Development, Poverty. Washington, DC: World Bank. Washington, DC. Alleyne, T. 2013. Energy Subsidy Reform in Sub- ESMAP (Energy Sector Management Assistance Saharan Africa: Experiences and Lessons. Washington, Program). 2017a. “Ukraine.” Energy Subsidy Reform DC: International Monetary Fund. Facility Ukraine Country Brief, World Bank, Washington, DC. Baffes, J., M. A. Kose, F. Ohnsorge, and M. Stocker. 2015. “The Great Plunge in Oil Prices: Causes, ———. 2017b. “Egypt.” Energy Subsidy Reform Consequences, and Policy Responses.” World Bank Facility Ukraine Country Brief, World Bank, Policy Research Note 15(01). Washington, DC. Bank for International Settlements (BIS). 2003. Fiscal ———. 2019. “Lifting the Burden of Electricity Issues and Central Banking in Emerging Economics. BIS Subsidies, While Expanding Access.” Energy Subsidy Paper 20, Bank for International Settlements, Basel. Reform Facility Rwanda Country Brief, World Bank, Washington, DC. Cardoso, E. 1991. “Deficit Finance and Monetary Dynamics in Brazil and Mexico.” Journal of Fengler, W. 2012. “How Price Controls Can Lead to Development Economics 37 (1-2): 173-197. Higher Prices.” World Bank Blogs, World Bank, Washington, DC. Clements, B., H.-S. Jung, and S. Gupta. 2007. “Real and Distributive Effects of Petroleum Price Ghosh, M., and J. Whalley. 2004. “Are Price Controls Liberalization: The Case of Indonesia.” The Developing Necessarily Bad? The Case of Rice in Vietnam.” Journal Economies XLV (2): 220-237. of Development Economics 73(1): 215-232. Coady, D., M. El-Said, R. Gillingham, K. Kpodar, P. Goyal, A., and J. Nash. 2017. Reaping Richer Returns: Medas, and D. Newhouse. 2006. “The Magnitude and Public Spending Priorities for African Agriculture Distribution of Fuel Subsidies: Evidence from Bolivia, Productivity Growth. Africa Development Forum. Ghana, Jordan, Mali, and Sri Lanka.” IMF Working Washington, DC: World Bank. Paper 06/247, International Monetary Fund, Washington, DC. Ha, J., M. A. Kose, and F. Ohnsorge, eds. 2019. Inflation in Emerging and Developing Economies: Contreras, J., and N. Guarata. 2013. “Inflacion y Evolution, Drivers, and Policies. Washington, DC: Variacion de Precios Relativos en Venezuela.” Economia World Bank. 38 (36) : 85-122. Ianchovichina, E., Loening, J. and Wood, C. Coomer, J., and T. Gstraunthaler. 2011. “The 2012. “How Vulnerable are Arab Countries to Global Hyperinflation in Zimbabwe.” The Quarterly Journal of Food Price Shocks?” Policy Research Working Paper Austrian Economics 14(3): 311-346. 6018, World Bank, Washington, DC. Coyne, C. and R. Coyne, eds. 2015. Flaws and Ceilings IEG (Independent Evaluation Group). 2008. “Climate Price Controls and the Damage They Cause. London: Change and the World Bank Group. Phase 1—An Institute of Economic Affairs. Evaluation of World Bank Win-Win Energy Policy De Mello. L., ed. 2008. Monetary Policies and Inflation Reforms.” World Bank, Washington, DC. Targeting in Emerging Economies. Paris: Organisation Jayne, T. S., N. M. Mason, W. J. Burke, and J. Ariga. for Economic Co-operation and Development. 2016. “Agricultural Input Subsidy Programs in Africa: Deichmann, U., and F. Zhang. 2013. Growing Green An Assessment of Recent Evidence.” Working Paper the Economic Benefits of Climate Action. Washington, 245892, Department of Agricultural, Food, and DC: World Bank. Resource Economics, Michigan State University. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 S P EC IAL FO CU S 1 59 Kojima, M. 2013. “Petroleum Product Pricing and Random Prices: A Diagrammatic Analysis.” The Complementary Policies.” World Bank Policy Research Quarterly Journal of Economics 97 (1): 1-26. Working paper 6396, World Bank, Washington, DC. Nguyen, A. D. M., J. Dridi, F. Unsal, and O. H. Kouame, W., and S. J.-A. Tapsoba. “Structural Williams. 2017. “On the Drivers of Inflation in Sub- Reforms and Firms’ Productivity: Evidence from Saharan Africa.” International Economics 151 Developing Countries”. Policy Research Working (October): 71-84. Paper 8308, World Bank, Washington, DC. Ore, M. A. H., L. A. Sanchez, L. D. Sousa, and L. Kovasic, W. E. 1995. “Designing and Implementing Tornarolli, eds. 2018. Fiscal and Welfare Impacts of Competition and Consumer Protection Reforms in Electricity Subsidies in Central America. Washington Transitional Economies: Perspectives from Mongolia, DC: World Bank. Nepal, Ukraine and Zimbabwe.” DePaul Law Review 44 (4): 1197-1224. OECD and World Bank. 2018. Markets and Competition OECD-WBG PMR Indictors for Selected Laborde, D., C. Lakatos, and W. Martin. 2019. Non-OECD Countries (2013-2018). Washington, DC: “Poverty Impact of Food Price Shocks and Policies.” World Bank. Policy Research Working Paper 8724. World Bank, Patel, N., and A. Villar. 2016. “Measuring Inflation.” Washington, DC. BIS Paper 89, Bank of International Settlements, Basel. Ma, C., and F. Valencia. 2018. “Welfare Gains from Pena Nelz, M. 2016. “LPG Subsidy Reform in India Market insurance: The Case of Mexican Oil Price Put the Right Systems in Place First.” Energy Sector Risk.” IMF Working Paper 18/35, International Management Assistance Program, World Bank, Monetary Fund, Washington, DC. Washington, DC. Maimbo, S. M. and C. A. Henriquez Gallegos. 2014. Shi, X., and S. Sun. 2017. “Energy Price, Regulatory “Interest Rate Caps around the World Still Popular, Price Distortion and Economic Growth: A Case Study but a Blunt Instrument.” Policy Research Working of China.” Energy Economics 63 (March): 261-271. Paper 7070, World Bank, Washington, DC. Silveira, R., and S. Malpezzi. 1991. “Welfare analysis of Morton, F. M. S. 2001. “The Problems of Price rent control in Brazil: the case of Rio de Janeiro.” Controls.” Regulation 24(1): 50-54. Discussion Paper INU 83, World Bank, Washington, DC. Mpapalika, J., and C. Malikane. 2019. “The Determinants of Sovereign Risk Premium in African Springer de Freitas, P., and M. Bugarin. 2007. “A Countries.” Journal of Risk and Financial Management Study on Administered Prices and Optimal Monetary 12 (1): 29-49. Policy: The Brazilian Case.” Centro de Estudios Monetarios Latinoamericanos, Mexico City. Mundaca, G. 2017. “Energy Subsidies, Public Investment and Endogenous Growth.” Energy Policy Stiglitz, J. E., and D. M. G. Newbery. 1979. “The 110 (C): 693-709. Theory of Commodity price Stabilization Rules: Welfare Impacts and Supply Responses.” The Economic Munoz, S. 2006. “Suppressed Inflation and Money Journal 89 (December): 799-817. Demand in Zimbabwe.” IMF Working Paper 06/15, International Monetary Fund, Washington, DC. Stocker, M., J. Baffes, Y. M. Some, D. Vorisek, and C. M. Wheeler. 2018. “The 2014-16 Oil Price Collapse in Murphy, F., A. Pierru, and Y. Smeers. 2019. Retrospect.” World Bank Policy Research Working “Measuring the Effects of Price Controls Using Mixed Paper 8419, World Bank, Washington, DC. Complementarity Models.” European Journal of Operational Research 275 (2): 666-676. Tarr, D. 1994. “The Welfare Costs of Price Controls for Cars and Color Televisions in Poland: Contrasting Newfarmer, R., and M. D. Pierola. 2015. Trade in Estimates of Rent-Seeking from Recent Experience.” Zimbabwe Changing Incentives to Enhance World Bank Economic Review 8 (3): 415-443. Competitiveness. Washington DC: World Bank. Teravaninthorn, S., and G. Raballand. 2009. Transport Newbery, D. M. G., and J. E. Stiglitz. 1982. “Risk Prices and Costs in Africa: A Review of the International Aversion, Supply Response, and the Optimality of Corridors. Washington, DC: World Bank. 60 S P EC IAL FO CU S 1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 Vera, L. 2017. “In Search of Stabilization and Subsidies that Encourage Wasteful Consumption.” Recovery: Macro Policy and Reforms in Venezuela.” Contribution by the World Bank to G20 Finance Journal of Post Keynesian Economics 40 (1): 9-26. Ministers and Central Bank Governors. Washington, DC: World Bank (September). Verme, P., and A. Araar, eds. 2017. The Quest for Subsidy Reforms in the Middle East and North Africa ———. 2014b. MENA Economic Monitor: Corrosive Region. Washington DC: World Bank. Subsidies. October. Washington, DC: World Bank. Weitzman, M. 1991. “Price Distortion and Shortage ———. 2018. A New Economy for the Middle East and Deformation, or What Happened to the Soap?”. North Africa. MENA Economic Monitor. Washington, American Economic Review 81 (3): 401-414. DC: World Bank. Winkler, R. 2015. “Feast or Famine: The Welfare ———. 2019. Global Economic Prospects: Darkening Impact of Food Price Controls in Nazi Germany.” Skies. January. Washington, DC: World Bank. Discussion Papers in Economic and Social History 136, University of Oxford. ———. 2019b. Doing Business 2020: Sustaining the World Bank. 2006. Review of Selected Railway Pace of Reforms. Washington DC: World Bank. Concessions in Sub-Saharan Africa. Washington, DC: World Trade Organization. 2000-2019. Trade Policy World Bank. Review. World Trade Organization, Switzerland. ———. 2009. Global Economic Prospects: Commodities at Crossroads. January. Washington, DC: World Bank. ———. 2019. “Distortion.” WTO Glossary. Geneva, Switzerland: World Trade Organization, Geneva. ———. 2014a. “Transitional Policies to Assist the Poor While Phasing Out Inefficient Fossil Fuel CHAPTER 2 REGIONAL OUTLOOKS Growth in the East Asia and Pacific (EAP) region is projected to slow from an estimated 5.8 percent in 2019 to 5.7 percent in 2020 and moderate further to 5.6 percent in 2021-22. Easier financing conditions and fiscal policy support will partly mitigate the lingering impact of trade tensions amid domestic challenges. Despite the recent slowdown, EAP remains the region with the fastest labor productivity growth. Nevertheless, productivity levels remain below the EMDE average in most EAP economies. In China, growth is expected to slow gradually, from an estimated 6.1 percent in 2019, to 5.9 percent in 2020, and to 5.7 percent by 2022. In the rest of the region, growth is expected to recover slightly to 4.9 percent in 2020 and firm further to 5 percent in 2021-22. The balance of risks to the outlook has improved, but is still tilted to the downside. Downside risks include a sharp slowdown in global trade due to a re-escalation of trade tensions; a sharper-than-expected slowdown in major economies; and a sudden reversal of capital flows due to an abrupt deterioration in financing conditions, investor sentiment, or geopolitical relations. An upside risk to the forecast is that the recent trade agreement between China and the United States leads to a sustained reduction in trade uncertainty, resulting in a stronger-than-expected recovery of regional investment and trade. Recent developments decelerated sharply from its 2017-18 peak. Imports have also moderated, reflecting a The East Asia and Pacific (EAP) region has been drawdown of inventories and a slowdown in experiencing a continued cooling of domestic investment growth due to deteriorated business demand in China alongside sizable external sentiment amid delays in certain major public headwinds (Figure 2.1.1). Global demand has infrastructure projects (China, Malaysia, Thailand, weakened, and trade policy uncertainty related to the Philippines). trade disputes between China and the United States was elevated prior to the recent bilateral In China, weakening exports have compounded agreement. In addition, trade tensions between the impact on GDP of the ongoing slowdown of Japan and the Republic of Korea, a maturing domestic demand (Figure 2.1.2; World Bank electronics cycle, and disruptions caused by rapid 2019c). Regulatory tightening aimed at curbing shifts in technological and emission standards, non-bank lending has contributed to further have also weighed on regional manufacturing cooling of domestic demand. Policy uncertainty activity and trade (World Bank 2019a, 2019b). and higher tariffs on exports to the United States, have dampened manufacturing activity, weighed The global trade slowdown and heightened trade on investor sentiment, and dented private policy uncertainty have affected regional growth investment. Despite this, net exports have been through three main channels: weaker total exports; contributing to growth. Imports, especially disruptions in cross-border supply chains; and intermediate goods imports, have contracted, declining private investment amid low business partly reflecting high base effect, drawdown of confidence (China, Indonesia, Malaysia, Thailand, inventories, disruptions in global and regional the Philippines). Regional export growth has supply chains, an onshoring of foreign manufacturing operations, and a weaker renminbi. The negative shock to exports and output from Note: This section was prepared by Ekaterine Vashakmadze. trade tensions with the United States has been Research assistance was provided by Juncheng Zhou and Yushu Chen. partly offset by currency depreciation, price 64 CHAPTER 2.1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 2.1.1 EAP: Recent developments adjustments, some reshoring of production, and Growth continues to slow in China and has moderated in the rest of the the redirection of exports to other countries. As a region. Regional export growth has decelerated sharply from the 2017-18 result, the current account surplus has widened. peak. In the region excluding China, exports are showing incipient signs of recovery. Monetary policies have been eased across the region amid subdued inflation. Net capital outflows from China have resumed in 2019. In the rest of the region, some commodity In the rest of the region, capital flows have been essentially balanced. importers operating at or above capacity have Bond spreads have generally declined. experienced a cyclical moderation of activity (Cambodia, the Philippines, Thailand, Vietnam). A. Growth, 2019 B. Export growth Weak export growth has added to the slowdown, especially in the economies that are deeply integrated into global and regional production networks (Thailand, the Philippines). Although Thailand and Vietnam have benefited somewhat from the diversion of U.S. demand away from China, the trade diversion only partially offset the decline in their exports to Asia arising from the overall negative impact of global trade tensions C. Consumer price inflation and cooling global demand, including in China, D. Nominal policy rates and change in inflation-adjusted policy rates on the region. In commodity exporters, which have only recently recovered from the effects of the earlier fall in commodity prices, the pace and composition of growth continues to reflect country-specific factors. In larger and more diversified economies, where past terms-of-trade shocks were less acute and macroeconomic fundamentals are solid, steady growth has continued at rates of around E. Net capital flows F. EMBI spreads 4.5-5 percent per year (Indonesia, Malaysia). In Indonesia, growth has been supported by private consumption and a positive contribution from net exports amid import compression. In Malaysia, weak investment growth has been offset by robust consumption growth supported by tight labor markets. The negative impact of slowing regional trade has Source: Haver Analytics; World Bank. so far been partly mitigated by monetary and fiscal A. Aggregate growth rates are calculated using GDP weights at 2010 prices and market exchange rates. Investment indicates fixed asset investment. Import and export data are volumes of goods and policy support in major regional economies. non-factor services. Investment, export, and import data for East Asia and Pacific region excl. China include Cambodia, Indonesia, Lao PDR, Malaysia, Mongolia, Philippines, Solomon Islands, Thailand, Monetary policy in many countries has become Vanuatu, and Vietnam. GDP data for East Asian and Pacific Countries excl. China include Cambodia, more accommodative in response to slowing Fiji, Indonesia, Kiribati, Lao PDR, Malaysia, Marshall Islands, Micronesia Fed. Sets., Mongolia, Myanmar, Palau, Papua New Guinea, Philippines, Samoa, Solomon Islands, Thailand, Timor-Leste, activity amid subdued inflation (Malaysia, the Tonga, Tuvalu, Vanuatu, and Vietnam. Bars indicate 2019 which are estimates. B. Export volumes. Data include only goods. 6-month moving average. Regional aggregate excludes Philippines, Thailand, Vietnam). In China, the Cambodia, Fiji, Lao PDR, Mongolia, Myanmar, Solomon Islands, Papua New Guinea, Timor-Leste, Vanuatu, and Vietnam due to data limitations. October-November export price deflators for China are central bank has eased policy mainly by cutting estimates. Last observation is November 2019 for China and October 2019 for EAP excl. China. bank reserve requirements, including the 0.5 C. Average year-on-year consumer price inflation. Mid-point of inflation target for Indonesia, Philippines, and Thailand. Inflation target for China and Vietnam. For Malaysia, the mid-point of Bank percentage point cut implemented in early January Negara’s official forecast range of 0.7-1.7 percent in 2019. Last observation is November 2019. D. Latest rate refers to Malaysia’s overnight policy rate, Indonesia’s 7-day reverse repo rate, China’s 2020. In Indonesia, continued capital inflows, a loan prime rate, Thailand’s one-day repurchase rate, Vietnam’s discount rate and Philippines’ overnight reverse repo rate. Change refers to the difference in real interest rate between November stable exchange rate, and low inflation have 2019 and January 2019. Last observation is November 2019. E. Net capital flows are estimates. Net capital inflows include net capital and financial account provided the necessary space for Bank Indonesia balance, errors and omissions. Last observation is 2019Q3. to continue policy easing. Several countries have F. J.P. Moran Emerging Market Bond Index (EMBI) spread. Last observation is December 2019. Click here to download data and charts. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 EAST ASIA AND PACIFIC 65 also provided fiscal support (China, Malaysia, FIGURE 2.1.2 Recent developments, China Thailand). Thailand announced a broad range of In China, growth has further decelerated amid continued cooling of stimulus measures, including a support package domestic demand and heightened trade tensions. The negative shock to for farmers, SMEs, and low-income households exports from higher tariffs on trade with the United States has been partly offset by currency depreciation. Imports from the United States have (World Bank 2019a). China has introduced plummeted, while imports from other regions have also weakened. The reductions in taxes and government fees, and a current account surplus has widened. The government has stepped up its fiscal support measures, with a focus on tax- and non-tax-revenue cuts, higher limit for local government on-budget and support for public investment spending through higher quotas for local borrowing. The consolidated fiscal and monetary government bonds. The stock of debt has stabilized, reflecting the decline policy support package implemented is, however, of non-bank lending. significantly smaller than the one adopted in the A. GDP growth B. Export growth wake of the global financial crisis, and somewhat smaller than the one deployed in 2016. Other factors—still-robust private consumption across much of the region and import compression—have dampened the impact of weakening manufacturing activity and exports on growth (Table 2.1.1). The recent de-escalation of China-U.S. trade tensions has buoyed asset prices and business confidence, and contributed to C. Import growth D. Balance of payments supportive external financing conditions. Bond spreads have narrowed, and net capital inflows have generally risen, despite sporadic episodes of market pressures. Outlook After moderating from an estimated 5.8 percent in 2019 to a projected 5.7 percent in 2020, regional growth is expected to ease further to 5.6 percent in E. General government debt and F. GDP growth and total debt 2021-22 (Tables 2.1.1 and 2.1.2). This mainly fiscal balance reflects a further moderate slowdown in China to 5.9 percent in 2020 amid continued domestic and external headwinds, including the lingering impact of trade tensions (Figure 2.1.3). This outlook is predicated on no re-escalation of trade tensions between China and the United States going forward and a gradual stabilization in global trade. It also assumes that authorities in Source: Haver Analytics; International Monetary Fund; National Bureau of Statistics of China; World Bank. China continue to implement monetary and fiscal A. Investment refers to gross capital formation, which includes change in inventories. Last policies to offset the negative impact of weak observation is 2019Q3. B.C. Data include only goods. Years cover January-November period. Last observation is November exports. The baseline projections embody a 2019. Asia includes both advanced and emerging Market and developing economies. ROW = all trading partners excluding the United States and Asia. Last observation for Asia and for ROW is weakened global outlook relative to June, partly October 2019. Export and import values. Data for total export growth in 2019 is -0.34 percent. D. Net capital flows and change in reserves are estimates. Net capital inflows include net capital and reflecting a much weaker-than-expected outlook financial account balance, errors and omissions. Last observation is 2019Q3. for global trade, manufacturing, and investment. E. Gross debt consists of all liabilities that require payment or payments of interest and/or principal by the debtor to the creditor at a date or dates in the future. Other includes other net expenditure (incl. social security and State-Owned Enterprise funds). Fiscal support measures are World Bank staff estimates. General government gross debt in 2019 are estimates. Despite recent de-escalation in China-U.S. F. Total debt is defined as a sum of domestic and external debt. Aggregate growth rates are calculated using GDP weights at 2010 prices and market exchange rates. Last observation for total bilateral trade tensions, heightened uncertainty credit and GDP growth is 2019Q3. surrounding the external environment is likely to Click here to download data and charts. 66 CHAPTER 2.1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 2.1.3 EAP: Outlook and risks persist, amid a fragile global outlook, com- EAP growth is projected to gradually decline, mainly reflecting slower pounding the trade weakness over the near term. growth in China. Growth in the rest of the region is expected to stabilize by 2020, with notable cross-country heterogeneity reflecting country specific Regional growth excluding China is projected to conditions. The long-term investment outlook is for broad-based deceleration. The region is characterized by deep global integration, which slightly recover to 4.9 percent in 2020 and firm makes countries vulnerable to external trade or financial shocks. further to 5 percent in 2021-22—toward its potential—assuming that weakness in A. GDP growth B. Output and potential growth manufacturing and export activity does not spill over to consumption and services. Domestic demand will continue to benefit from generally supportive financing conditions, amid low inflation and robust capital flows (Cambodia, the Philippines, Thailand, Vietnam). Some countries will benefit from large public infrastructure projects coming onstream (Thailand, the Philippines). C. Contribution to productivity growth D. Ten-year ahead investment forecasts Growth in the economies that are deeply integrated into global and regional production networks is expected to moderate further in 2021- 22, and to adjust a bit faster than expected toward potential, reflecting capacity constraints and subdued external demand. In particular, in Malaysia, growth is expected to inch down to 4.5 percent in 2020-21, with weak export growth partly offset by strong domestic demand, underpinned by favorable financing conditions, a E. Export growth, 2013-18 F. Growth impact of 1 percentage point slower growth in China, Japan, rebound in investment, stable labor market or other G7 countries conditions, and low inflation. In Indonesia, which depends less on exports than other regional economies, growth is projected to fluctuate around 5 percent throughout the forecast horizon. This forecast is predicated on a continued support from private consumption, a pickup in investment, solid growth of the working-age population, and improving labor markets. In small Source: International Monetary Fund; Penn World Tables; The Consensus Forecasts; World Bank. A. EAP excl. China = Cambodia, Indonesia, Lao PDR, Malaysia, Mongolia, Myanmar, Philippines, commodity exporters, growth is expected to Thailand, and Vietnam. Pacific Island excl. PNG includes Fiji, Kiribati, Marshall Islands, Micronesia, decelerate, but remain strong (Mongolia), or Palau, Samoa, Solomon Islands, Timor-Leste, Tonga, Tuvalu, and Vanuatu. 1990-2019 average for EAP excl. China excludes Myanmar and 1990-2019 for Pacific Island excl. PNG excludes Marshall rebound in 2022 (Papua New Guinea), supported Islands, Micronesia, Palau, Timor-Leste, and Tuvalu. Aggregate growth rates are calculated using GDP weights at 2010 prices and market exchange rates. Data in shaded areas are forecasts. by investment in infrastructure and mining. B. Potential growth estimates are from a multivariate filter model of WB (2018a). Aggregate growth rates are calculated using GDP weights at 2010 prices and market exchange rates. Includes China, Indonesia, Malaysia, Mongolia, the Philippines, and Thailand. Output growth in 2020 is a forecast. While growth in the region is projected to remain C. Productivity defined as output per worker in 2010 U.S. dollars at 2010 prices and exchange rates. D. 10-year-ahead forecasts surveyed in indicated year. Constant 2010 U.S. dollar investment- robust in the near term, underlying potential weighted averages. Sample includes China, Indonesia, Malaysia, the Philippines, and Thailand. E. EA = East Asia. PI = Pacific Islands. EA1 = Brunei Darussalam, Cambodia, Malaysia, Mongolia, growth is likely to continue to decline over the Thailand, and Vietnam; EA2 = Indonesia, Lao PDR, Myanmar and Philippines. PI1 = Kiribati, Marshall long term (Chapter 3; World Bank 2018a, Islands, Micronesia, Timor-Leste, Tonga, and Tuvalu; PI2 = Palau and Vanuatu; PI3 comprises Fiji, Papua New Guinea, Samoa, and Solomon Islands. 2018b). The slowdown is expected to be broad- F. Median cumulative responses after two years. Based on a Bayesian structural VAR model. Data coverage is 1998Q1-2018Q2 and is shorter for some countries. The endogenous variables include, in based, reflecting deteriorating demographic trends, this Cholesky ordering: growth in G7 excluding Japan, EMBI, Japan’s growth, China’s growth, Korea’s growth, and three variables for each shock-recipient country: real commodity price index, growth, and especially in China, Thailand, and Vietnam, real effective exchange rates. Global spillovers refer to spillovers from growth shocks in the G7 excluding Japan. The model includes a dummy that captures the global financial crisis of 2008-09. combined with a projected slowdown in capital Vertical lines represent the 33-66 percent confidence bands. Click here to download data and charts. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 EAST ASIA AND PACIFIC 67 accumulation and lower total factor productivity and the disruption of global value chains, as well in China as credit growth is reined in (Box 2.1.1.; as through financial, investment, commodity, and Chapter 3). Investment growth in the rest of the confidence channels (Chapter 1; World Bank EAP region is also expected to be subdued and 2016a; World Bank 2019d). below historical averages, as the long-term prospects for investment growth in the region Risks of a sharper-than-expected slowdown in remain weak and have been persistently China stem from domestic challenges, as well as downgraded since 2010. from the difficult external environment. The total leverage of the economy—measured as the ratio of total credit (general government and non-financial Risks private sector) to GDP—has surpassed 260 The balance of risks has improved, but remains percent of GDP in 2019, although the share of tilted to the downside. Downside risks include a non-bank lending continued to decline due to sharp slowdown in global trade due to a re- regulatory tightening. High corporate escalation of global trade tensions; a sharper-than- indebtedness in sectors with weak profitability is expected slowdown in major economies; and a of concern (World Bank 2019e). A sizable portion sudden reversal of capital flows due to an abrupt of recent support measures has taken the form of deterioration in global financing conditions, expanding local government special bond quotas. investor sentiment, or geopolitical relations. The growing debt burden on local authorities may increase their vulnerability to shocks. A renewed spike in trade policy uncertainty could cause a deterioration in confidence, investment, Most of the EAP region weathered the and trade (Caldara et al. 2019; Freund et al. deterioration of external conditions in 2019 well, 2018). Failure by China and the United States to relying on exchange rate flexibility and monetary reach a long-term, comprehensive, and durable and fiscal stimulus. A further deterioration would agreement could lead to renewed trade tensions, test the resilience of the region’s economies. Even with broad-ranging global and regional conse- though most large countries have generally sound quences. economic fundamentals—track record of solid growth, fast labor productivity growth, large An upside risk to the forecast is the possibility of a consumer bases, diversified economies, sound sustained de-escalation of trade tensions between policy frameworks, and strong policy buffers—the China and the United States. The recent trade region remains vulnerable to risks related to agreement that reverses some tariff increases could abrupt changes in global financial conditions. be the beginning of a constructive process leading to a sustained reduction in policy uncertainty and Many countries have pockets of vulnerabilities. trade barriers. This could significantly improve These include elevated debt (Lao People’s confidence and unlock pent-up demand for Democratic Republic, Malaysia, Mongolia, investment, bolstering growth. Vietnam); sizable fiscal deficits (Lao PDR, Vietnam); or heavy reliance on volatile capital In the baseline scenario, the impact of slower flows (Cambodia, Indonesia). Renewed episodes global growth and external demand on the region of financial market stress could have pronounced is offset by more supportive financing conditions and widespread effects on countries with high and stronger policy stimulus. However, a sharper- indebtedness (Chapter 1). Vulnerabilities among than-baseline deceleration of activity in large some EAP countries could amplify the impact of economies—the Euro Area, China, or the United external shocks, such as a sudden stop in capital States—could have adverse repercussions across flows or a rise in borrowing costs. the region through weaker demand for exports 68 CHAPTER 2.1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 TABLE 2.1.1 East Asia and Pacific forecast summary Percentage point differences (Real GDP growth at market prices in percent, unless indicated otherwise) from June 2019 projections 2017 2018 2019e 2020f 2021f 2022f 2019e 2020f 2021f EMDE EAP, GDP 1 6.5 6.3 5.8 5.7 5.6 5.6 -0.1 -0.2 -0.2 (Average including countries with full national accounts and balance of payments data only) 2 EMDE EAP, GDP 2 6.5 6.3 5.8 5.7 5.6 5.6 -0.1 -0.2 -0.2 GDP per capita (U.S. dollars) 5.8 5.6 5.2 5.1 5.1 5.1 -0.2 -0.2 -0.2 PPP GDP 6.4 6.3 5.8 5.7 5.6 5.6 -0.1 -0.2 -0.2 Private consumption 6.1 8.4 6.8 6.9 6.6 6.6 -0.2 -0.1 -0.4 Public consumption 8.9 8.7 7.7 7.6 7.5 7.5 0.1 0.1 0.1 Fixed investment 4.7 5.1 4.1 4.6 4.7 4.6 -1.0 -0.5 -0.2 Exports, GNFS 3 9.5 5.0 1.4 1.3 2.0 2.4 -1.9 -2.6 -2.3 Imports, GNFS 3 8.4 8.4 -0.3 2.0 2.5 2.9 -5.0 -3.0 -3.2 Net exports, contribution to growth 0.4 -0.9 0.5 -0.2 -0.1 -0.1 0.9 0.2 0.4 Memo items: GDP East Asia excluding China 5.4 5.2 4.8 4.9 5.0 5.0 -0.3 -0.3 -0.2 China 6.8 6.6 6.1 5.9 5.8 5.7 -0.1 -0.2 -0.2 Indonesia 5.1 5.2 5.0 5.1 5.2 5.2 -0.2 -0.2 -0.1 Thailand 4.0 4.1 2.5 2.7 2.8 2.9 -1.0 -0.9 -0.9 Source: World Bank. Note: e = estimate; f = forecast. EMDE = emerging market and developing economies. World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of countries’ prospects do not differ at any given moment in time. 1. GDP and expenditure components are measured in 2010 prices and market exchange rates. Excludes Democratic People’s Republic of Korea and dependent territories. 2. Sub-region aggregate excludes Democratic People’s Republic of Korea, dependent territories, Fiji, Kiribati, the Marshall Islands, the Federated States of Micronesia, Myanmar, Nauru, Palau, Papua New Guinea, Samoa, Timor-Leste, Tonga, and Tuvalu, for which data limitations prevent the forecasting of GDP components. 3. Exports and imports of goods and non-factor services (GNFS). Click here to download data. TABLE 2.1.2 East Asia and Pacific country forecasts1 Percentage point differences (Real GDP growth at market prices in percent, unless indicated otherwise) from June 2019 projections 2017 2018 2019e 2020f 2021f 2022f 2019e 2020f 2021f Cambodia 7.0 7.5 7.0 6.8 6.8 6.8 0.0 -0.1 0.0 China 6.8 6.6 6.1 5.9 5.8 5.7 -0.1 -0.2 -0.2 Fiji 5.2 4.2 1.0 1.7 2.9 3.0 -2.4 -1.6 -0.4 Indonesia 5.1 5.2 5.0 5.1 5.2 5.2 -0.2 -0.2 -0.1 Lao PDR 6.9 6.3 5.2 5.8 5.7 5.6 -1.4 -0.9 -0.9 Malaysia 5.7 4.7 4.6 4.5 4.5 4.5 0.0 -0.1 -0.1 Mongolia 5.3 7.2 5.7 5.5 5.2 5.5 -1.5 -1.4 -1.0 Myanmar 6.8 6.5 6.6 6.7 6.8 6.8 0.1 0.1 0.0 Papua New Guinea 3.5 -0.8 5.6 2.9 2.9 3.0 0.0 -0.2 -0.6 Philippines 6.7 6.2 5.8 6.1 6.2 6.2 -0.6 -0.4 -0.3 Solomon Islands 3.0 3.5 2.9 2.8 2.8 2.7 0.0 0.0 0.1 Thailand 4.0 4.1 2.5 2.7 2.8 2.9 -1.0 -0.9 -0.9 Timor-Leste -3.5 -1.1 4.2 4.6 4.9 5.0 0.3 0.0 -0.1 Vietnam 6.8 7.1 6.8 6.5 6.5 6.4 0.2 0.0 0.0 Source: World Bank. Note: e = estimate; f = forecast. World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of countries’ prospects do not significantly differ at any given moment in time. 1. GDP and expenditure components are measured in 2010 prices and market exchange rates. Click here to download data. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 EAST ASIA AND PACIFIC 69 BOX 2.1.1 Labor productivity in East Asia and Pacific: Trends and drivers East Asia and Pacific (EAP) remains the region with the fastest productivity growth, averaging 6.3 percent a year in 2013-18, notwithstanding the second-steepest post-crisis slowdown among emerging market and developing economy (EMDE) regions. Nevertheless, productivity levels remain below the EMDE average in most EAP economies. While factor reallocation toward more productive sectors, high investment, and trade integration with product upgrading have promoted above-average productivity growth, most of these drivers are expected to become less favorable in the future. A comprehensive set of reforms to liberalize services sectors, improve corporate management, level the playing field for private firms, enhance human capital, facilitate urban development, and foster innovation is needed to reverse the recent productivity growth slowdown. Introduction how product and labor market reforms have increased output and productivity.4 Growth of labor productivity, defined as output (GDP) per worker, averaged 6.3 percent a year in the East Asia Against this backdrop and drawing on these studies, this and Pacific (EAP) region in 2013-18 (Figure 2.1.1.1).1 box compares productivity developments in EAP with While this pace remained the fastest among emerging other EMDE regions. In particular, it discusses the market and developing economy (EMDE) regions, it was following questions: almost 3 percentage points below EAP’s pre-crisis (2003- 08) average after the second-steepest post-crisis decline in • How has productivity evolved in the region? labor productivity growth among EMDE regions. The post-crisis slowdown in productivity growth has been • What factors have been associated with productivity broad-based, affecting 60 percent of EMDEs in EAP. growth in the region? At 12 percent of the advanced-economy average in 2013- • What policy options are available to boost regional 18, average productivity in EAP remains below the EMDE productivity growth? average.2 Labor productivity levels in EAP are more homogeneous than in other EMDE regions. Similarly, This box considers labor productivity, defined as real GDP productivity growth is more homogeneous across EAP than per worker (at 2010 prices and market exchange rates). across other EMDE regions, possibly reflecting particularly The data are available for sixteen countries: Cambodia, close regional integration, including through regional China, Fiji, Indonesia, Lao People’s Democratic Republic, supply chains. Malaysia, Mongolia, Myanmar, Papua New Guinea, the Philippines, Samoa, the Solomon Islands, Thailand, This box builds on a considerable literature that examines Tonga, Vanuatu, and Vietnam. productivity growth in EAP. Earlier studies have documented the recent productivity growth slowdown in Evolution of regional productivity EAP using country-level and firm-level data.3 Others have identified education, innovation, market efficiency, Rapid productivity growth. Labor productivity growth in institutions, and physical infrastructure as the main drivers EAP averaged 6.4 percent a year between the early 1980s of productivity improvements in EAP (Kim and Loayza and 2018—the highest growth rate of all EMDE regions, 2019). Another set of studies has empirically documented mainly reflecting rapid growth in China. EAP labor productivity growth rose from 4.3 percent a year in the 1980s to 6.3 percent a year in the 1990s, and peaked at 8.9 percent a year in 2003-08 (Figure 2.1.1.2). Since the Note: This section was prepared by Ekaterine Vashakmadze, building upon analysis in Chapter 3. Research assistance was provided by global financial crisis, EAP productivity growth has slowed Juncheng Zhou and Shijie Shi. to 6.3 percent a year on average during 2013-18. This 1 Unless otherwise specified, productivity is defined as labor post-crisis slowdown is also accounted for largely by productivity, that is, output per worker. China, in particular its policy-guided move towards more 2 EAP averages are heavily influenced by China, which accounts for 80 sustainable growth after a period of exceptionally rapid percent of EAP output in 2013-18. That said, even the median productivity level in EAP is below that of the median EMDE region. expansion of fixed investment and exports; in the region’s 3 For studies using country-level data, see APO (2018); IMF (2006), (2017); World Bank (2018a), and World Bank (2019a). For studies using firm-level data, see Di Mauro et al. (2018); de Nicola, Kehayova, 4 See Adler et al. (2017); Bouis, Duval and Eugster (2016); Chen and Nguyen (2018); OECD (2016); and World Bank and DRCSC (2002); Nicoletti and Scarpetta (2005); Timmer and Szirmai (2000). (2019). 70 CHAPTER 2.1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.1.1 Labor productivity in East Asia and Pacific: Trends and drivers (continued) other major economies, productivity growth has been FIGURE 2.1.1.1 Productivity in EAP broadly stable. Around two-thirds of EAP economies in compared with other country groups 2013-18 were still experiencing labor productivity growth above their long-run average. EAP has remained the region with the fastest productivity growth, at 6.3 percent a year in 2013-18, notwithstanding the second-steepest post-crisis Within-region heterogeneity. While productivity growth slowdown among EMDE regions. Nevertheless, exceeded the EMDE average during 2013-18 in most EAP productivity levels remain below the EMDE average in economies (the exceptions being some Pacific Islands), most EAP economies. there was some cross-country heterogeneity. Productivity growth was particularly fast in China, followed by several A. Average annual growth in EMDE regions large Association of Southeast Asian Nations (ASEAN) economies, including Vietnam. These countries were among the ten percent of EMDE economies with the fastest productivity growth in the period. They benefited from improvements in human capital, and trade openness, technology transfer and adaptation, high investment rates, and an industrial base that was rapidly becoming more sophisticated (Andrews et al. 2015). Productivity growth was slowest among EAP economies in some Pacific Islands, including Solomon Islands, partly reflecting political tensions. Low productivity levels. Notwithstanding rapid productivity growth, average productivity levels in EAP (12 percent of the advanced-economy average in 2013-18), B. Productivity levels in 2013-18 and annual convergence including China, remained below the EMDE average rates in EMDE regions (which is close to 20 percent of the advanced-economy average; APO 2018; Di Mauro et al. 2018). Malaysia, the EAP economy with the highest productivity level (25 percent of the advanced-economy average), has benefited from several decades of sustained high growth rates reflecting its diversified production and export base and sound macroeconomic policies (Munoz et al. 2016). Labor productivity convergence. Whereas convergence of productivity toward advanced-economy levels in most other EMDE regions has slowed since the financial crisis, it has remained robust in EAP reflecting macroeconomic stability, strong fundamentals, still high investment rates, and diversified and competitive production bases in the Source: Penn World Table; The Conference Board; World Bank (World region’s major economies (Chapter 3). Assuming recent Development Indicators). productivity growth can be sustained, at least 50 percent of Note: Unless otherwise specified, productivity refers to labor productivity, defined as output per worker. Sample comprises 35 advanced economies economies in the region are on course to halve their and 127 EMDEs, of which 16 are in East Asia and the Pacific (EAP), 21 are in in Eastern Europe and Central Asia (ECA), 25 are in in Latin America and productivity gap relative to advanced-economy averages the Caribbean (LAC), 14 are in Middle East and North Africa (MNA), 7 are in over the next 40 years. History shows how successful South Asia (SAR), and 44 are in Sub-Saharan Africa (SSA). A. Blue bars denote range across GDP-weighted averages for 6 EMDE productivity convergence by such economies as Singapore regions. Yellow lines denote simple average of the 6 EMDE regional averages. Red dots denote simple average of 16 EMDEs in EAP. and the Republic of Korea, which were reclassified as B. Rate of convergence calculated as the difference in productivity growth advanced economies in the 1990s, required high and rates with the average advanced economy (AE) divided by the log difference in productivity levels with the average advanced economy. Regional rate of sustained productivity growth differentials relative to convergence is the GDP-weighted average of EMDE members of each established advanced economies over several decades region. "Level" of productivity refers to the GDP weighted average of regional productivity as a share of the average advanced economy during 2013-2018. (Chapter 3). Click here to download data and charts. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 EAST ASIA AND PACIFIC 71 BOX 2.1.1 Labor productivity in East Asia and Pacific: Trends and drivers (continued) FIGURE 2.1.1.2 Evolution of productivity in EAP To a larger extent than in the average EMDE, the post-crisis slowdown in EAP’s productivity growth has reflected slowing total factor productivity growth, especially in China. In EAP, slowing TFP growth accounted for two-thirds of the post-crisis slowdown in labor productivity growth, compared to about half in the average EMDE. Notwithstanding rapid productivity growth, average productivity levels in EAP—12 percent of the advanced-economy average—remain below the EMDE average. A. Annual productivity growth in EAP B. Share of economies with productivity C. Contributions to annual productivity growth in 2013-18 below long-run and growth pre-crisis averages D. Contributions to annual productivity E. Productivity levels relative to F. Annual labor force growth growth advanced-economy average Source: Barro and Lee (2015); Haver Analytics; International Monetary Fund; Penn World Tables; United Nations; Wittgenstein Centre for Demography and Global Human Capital; World Bank (World Development Indicators). Note: Unless otherwise specified, productivity refers to labor productivity, defined as output per worker in 2010 U.S. dollars at market exchange rates. A. Average growth rates calculated using 2010 U.S. dollars at market exchange rates. B. Share of countries for which productivity growth average over 2013-18 is lower compared to a long-run (1992-2018) and pre-crisis (2003-08) average. Yellow line denotes 50-percent line. C.D. Aggregate growth rates calculated using GDP weights at 2010 prices and market exchange rates. Samples comprise 92 EMDEs and 16 EAP economies. E. Aggregate growth rates calculated using GDP weights at 2010 prices and market exchange rates. Samples comprise 35 advanced economies, 127 EMDEs and 16 EAP economies. F. Derived using data from International Labour Organization, ILOSTAT database and World Bank population estimates. Labor force data retrieved in September 2019. Click here to download data and charts. Sources of productivity growth. Productivity growth can (Mason and Shetty 2019; Tuan, Ng, and Zhao 2009; Xu, be decomposed into its sources: factor accumulation Xinpeng and Sheng 2012). These reforms were (human or physical capital) and increases in the efficiency accompanied by improvements in macroeconomic of factor use (total factor productivity, or TFP). In EAP, policies, strengthening institutions, and higher investment slowing TFP growth accounted for two-thirds of the post- in infrastructure and human capital in several countries crisis slowdown in labor productivity growth, compared to (China, Indonesia, Malaysia, the Philippines, Vietnam). about half in the average EMDE. This followed a decade The post-crisis slowdown in the region’s TFP growth of surging TFP growth in EAP, as China’s World Trade partly reflects a moderation in the pace of global Organization accession in 2001 was followed by rapid integration (Ruta, Constantinescu, and Mattoo 2017). trade integration, large foreign direct investment (FDI) Weaker investment accounted for another one-third of the inflows into the region, and rapid technological adaptation slowdown in labor productivity growth in EAP, as 72 CHAPTER 2.1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.1.1 Labor productivity in East Asia and Pacific: Trends and drivers (continued) investment booms before the global financial crisis and in the Philippines (World Bank 2018c). In Vietnam, its immediate wake subsided, especially in response to intersectoral reallocation has continued to account for policy guided moderation in China (Kose and Ohnsorge approximately half of labor productivity growth, with no 2019). sign of deceleration (World Bank and MPIV 2016). Heterogeneity in productivity growth. EAP’s high average Productivity growth in the manufacturing sector has been productivity growth masks some divergence between a major driving force behind overall productivity growth China and the rest of EAP. Whereas TFP growth and in most EAP countries (APO 2018; Figure 2.1.1.3). Since capital deepening slowed in China between 2003-08 and 2000s, the contribution of services to productivity growth 2013-18 amid a policy-guided investment slowdown, they has increased, albeit from a low base, as innovations in this accelerated in the rest of EAP and especially in some sector took hold.6 For example, e-commerce has ASEAN countries (the Philippines and Vietnam) reflecting accelerated sharply in China, with e-commerce firms significant FDI inflows and high rates of investment having 30 percent higher productivity, as well as being spending. The decline in China’s TFP growth has been more export-oriented than other firms (IMF 2019). attributed not only to the slowdown in investment growth, Recent advances in information and communication with its associated embodied technical progress, but also to technology have bolstered productivity growth in fading gains from global trade integration and institutional wholesale and retail trade, hotels, and restaurants; reforms.5 transport, storage, and communications; and finance, real estate, and business activities. It is likely that the growth in Sources of productivity growth value-added generated by intangible services is underestimated to the extent they are incorporated in the Productivity growth through sectoral reallocation. Strong production of manufactured goods (ADB 2019). pre-crisis productivity growth in EAP was supported by policies that encouraged resource reallocation from low- to In contrast to most other EMDE regions, within-sector high-productivity sectors, as well as within-sector upgrades productivity growth accelerated in many EAP economies (IMF 2006). Following the crisis, however, and as in other in the post-crisis period. China was an exception: there, EMDE regions, gains from factor reallocation toward within-sector productivity growth slowed amid increased more productive sectors slowed sharply, as the pace of overcapacity, declining firm dynamism, and increasing urbanization slowed (in most cases well before reaching financial constraints, including as a result of rising leverage Organisation for Economic Co-operation and (IMF 2018a). This is notwithstanding considerable in- Development—average levels) and overcapacity in China house research and development, and technology transfers weighed on the efficiency of investment. During 2013-15, both domestically and from abroad (Hu, Jefferson, and sectoral reallocation is estimated to have accounted for Jinchang 2005). under one-fifth of EAP productivity growth, less than half of its share during 2003-08 (two-fifths; Figure 2.1.1.3). Drivers of productivity. Fundamental drivers of productivity have improved more rapidly in EAP than in In East Asia, structural transformation, in the form of the the average EMDE (Figure 2.1.1.3). In general, movement of people and capital from agriculture to productivity in economies with favorable initial conditions manufacturing and services, has been a key driver of have grown by up to 0.8 percentage point per year faster productivity growth as countries have risen from low- to than other economies (Chapter 3), which partly explains middle-income status. Once countries have reached faster productivity growth in countries with strong human middle-income levels, within-sector productivity gains capital, including China, Malaysia, and Vietnam. have become a more important driver of productivity Compared to many other EMDEs, productivity growth in growth and cross-sectoral shifts less important (de Nicola, EAP economies have benefited from high investment Kehayova and Nguyen, 2018; Mason and Shetty, 2019). (IMF 2006; World Bank 2019b). Other factors However, there has been considerable heterogeneity across contributing to relatively high productivity growth in the the region in this respect: thus in recent years sectoral EAP region include trade integration, including through reallocation has stalled in Thailand, proceeded slowly in global supply chains; foreign investment, which supported Malaysia, and continued apace in Indonesia, Vietnam, and rapid technology adoption from abroad; and progress 5 See World Bank and DRCSC (2014); World Bank (2019a); Baldwin 6 See APO (2018); ADB (2019); Cirera and Maloney (2017); and (2013), and Subramanian and Kessler (2013). Kinda (2019). G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 EAST ASIA AND PACIFIC 73 BOX 2.1.1 Labor productivity in East Asia and Pacific: Trends and drivers (continued) FIGURE 2.1.1.3 Factors underlying productivity growth in EAP Factor reallocation toward more productive sectors, high investment, trade integration with product upgrading, and rapid innovation have all contributed to above-EMDE-average productivity growth in EAP. Productivity growth in the manufacturing sector has been a major driving force behind overall productivity growth in most EAP countries. Fundamental drivers of productivity have improved more rapidly in EAP than in the average EMDE. A. Contributions to annual productivity B. Annual sectoral productivity growth C. Contributions to annual productivity growth growth D. Composition of sectoral value-added E. Drivers index F. Level of drivers across regions, 2017 Source: APO productivity database; Expanded African Sector Database; Groningen Growth Development Center Database; Haver Analytics; ILOSTAT; OECD STAN; United Nations; World KLEMS. A.B. Productivity refers to labor productivity, defined as output per worker. Medians of county-specific contributions. Sample comprises 9 EAP economies and 46 EMDEs. A. Within-sector contribution shows the contribution to overall productivity growth of initial real value added-weighted sectoral productivity growth; between-sector contribution shows the contribution of intersectoral changes in employment shares. C. Median of the country groups. Sample comprises 9 EAP economies. D. Values are calculated using 2010 U.S. dollars at 2010 market exchange rates. E. For each country, index is a weighted average—weighted by the normalized coefficients shown in Annex 3.5—of the normalized value of each driver of productivity. Drivers include the ICRG rule of law index, patents per capita, share of non-tropical area, investment in percent of GDP, ratio of female average years of education to male average years, share of population in urban area, Economic Complexity Index, years of schooling, and share of working-age population, and inflation. See Annex 3.5 for details. Regional and EMDE indexes are GDP-weighted averages. Samples comprise 7 economies in EAP. F. Unweighted average levels of drivers, normalized as average of AEs as 100 and standard deviation of EMDEs as 10. Orange diamond represents average within EAP economies in 2017. Blue bar represents range of the average drivers for six regions in 2017. Variables corresponding to the concepts are follows: Education = years of education, Urbanization = share of population living in urban area, Investment = share of investment to GDP, Institution = Government Effectiveness, Econ. Complexity = Economic complexity index+, Geography = share of land area which are outside of tropical region, Gender Equality = Share of the year of schooling for female to male, Demography=share of population under 14, Innovation=Log patent per capita, Trade = Export + Import/GDP. Click here to download data and charts. toward more complex products with higher value-added encouraged investment, while trade and investment (World Bank 2019d).7 Macroeconomic stability has openness and above-EMDE-average research and development have supported innovation (Kim and Loayza 7 EAP is characterized by an above-average share of larger and 2019). exporting firms (Chapter 3). In EAP, 35 percent of firms are large (compared with 25 percent in the average EMDE) and 16 percent of firms are exporters (compared with 12 percent in the average EMDE). managerial practices that help them make better decisions regarding More productive firms tend to self-select into exporting firms which have investment, input selection, and production process (Hallward- higher productivity, as they are exposed to frontier knowledge and best Driemeier, Iarossi, and Sokoloff 2002). 74 CHAPTER 2.1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.1.1 Labor productivity in East Asia and Pacific: Trends and drivers (continued) FIGURE 2.1.1.4 Prospects for productivity growth in EAP Being less able to rely on export growth than in the past, EAP countries need to unleash domestic sources of productivity growth. Priority areas include reforms to enhance human capital, address informality, foster innovation, and facilitate urban development. In addition, achieving long-term sustainable development calls for debt overhangs to be addressed and excessive leverage to be avoided. A. Contribution of export growth to B. Human capital index and annual C. Informal economies annual GDP growth productivity growth D. Research and development E. Urbanization F. Debt and labor productivity expenditure Source: Elgin et al. (forthcoming); Haver Analytics; World Bank (World Development Indicators). A.B.F. Productivity refers to labor productivity, defined as output per worker. A. Growth of volume of exports of goods and non-factor services. B. The HCI calculates the contributions of health and education to worker productivity. The final index score ranges from zero to one and measures the productivity as a future worker of a child born today relative to the benchmark of a child with full health care and complete education. HCI data are for 2017. Labor productivity growth data are for 2018. C. Blue bars show the share of informal output in total output based on the Dynamic General Equilibrium (DGE) model. The diamonds show the share of informal employment in total employment. D. Data are not available for all featured economies. E. Urbanization levels denote share of urban population in total population. F. Total debt comprises bank credit to households, non-financial corporations, and general government debt (broad definition). Click here to download data and charts. That said, the factors supporting post-crisis productivity previously helped to spur EAP productivity growth have growth have differed somewhat across EAP economies. also deteriorated since the crisis. For example, the trend Growth of the drivers most strongly associated with toward broadening production to a more diverse range of productivity growth, including labor force growth and products at more upstream stages of the value chain slowed investment, has slowed in EAP since 2008. Investment partly because of a stagnation in global value chains after growth in many EAP economies has slowed, led by a 2008 (World Bank 2019b). policy-led moderation of investment rates to reduce credit expansion. In addition, earlier favorable demographic Prospects for productivity growth. Productivity gaps are trends in China, Thailand, and Vietnam have waned as still substantial between advanced economies and EAP populations have started to age. Other factors that had countries, suggesting potential for further significant G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 EAST ASIA AND PACIFIC 75 BOX 2.1.1 Labor productivity in East Asia and Pacific: Trends and drivers (continued) productivity gains. However, although EAP productivity transparency could boost productive public investment growth remains solid relative to long-run historical rates, it (World Bank 2018b). is likely to soften further over the near future, as trends in a number of fundamental drivers of productivity become Remove obstacles to private investment. Private less favorable. Thus, trade and investment growth are investment could be spurred by higher FDI inflows that expected to continue to ease in an environment of could offer knowledge and technology transfers, deeper weakening global demand, heightened global policy regional trade integration and better institutional uncertainty, and a continued policy-guided slowdown in environments (World Bank 2018b, 2019b). In China, investment growth in China (Figure 2.1.1.4). Slowing private investment could be lifted by improved market global trade growth may also lower incentives to innovate access, increased competition, policies that provide a more or upgrade products and processes (World Bank 2019b). level playing field relative to state-owned enterprises Structural declines in working-age populations in major (SOEs), greater financial discipline, stronger intellectual economies will also weaken growth momentum (World property rights, lower barriers to entry, and a gradual Bank 2016b, 2018a). opening of China’s financial system to international investors (World Bank 2018a, 2018d; World Bank and Policy implications DRCSC 2019). Other major economies in the region, including Indonesia, Malaysia, Thailand, and Vietnam, A comprehensive set of policy efforts can help countries in could boost private investment by increasing private sector the region improve their investment and productivity participation in major infrastructure projects and by growth and speed up their income convergence with the changing their funding policies to provide more advanced economies. These policies fall into four broad opportunities for international and domestic private categories: improving factors of production, including investors. through human capital development; encouraging productivity at the firm level, including by leveling the Increase human capital. Children born in the EAP region playing field for private relative to state-owned firms and today will, at age 18, be only 53 percent as productive as improving corporate governance; removing obstacles to they could be if they benefited from best practices in between-sector reallocation, including through continued education and health (World Bank 2019b). Several EAP urban development; and fostering a productivity-friendly economies have below-average educational attainment business environment. Specific policies within these four (Cambodia, Lao PDR). In general, reforms that augment broad categories depend on country specific circumstances human capital, through initiatives to strengthen the (World Bank 2018b; Kim and Loayza 2017; Munoz et al. quality and flexibility of education systems and improve 2016). education outcomes, are critical to achieving and sustaining high productivity growth. Improving factors of production Boosting firm productivity Slowing capital deepening has contributed to the post- crisis productivity growth slowdown in several EAP While within-sector productivity growth has been resilient countries, while outside China the contribution of human in EAP, especially outside China, there is room for capital gains to productivity growth has stalled. To boost generating additional productivity gains. Factor productivity growth, policies are needed to improve public misallocation, although it has declined, remains sizable investment, lift private investment, and improve human (World Bank 2019b). In the current weak external capital. environment that allows limited productivity gains through knowledge and technology spillovers from trade, Improve public investment. A wide range of policy efforts this is likely to be a critical source of productivity gains for are needed to improve the investment outlook, especially the region. Policy measures can include levelling the in countries with particularly large investment needs playing field for private and state-owned firms, improving (Cambodia, Indonesia, Lao PDR, Myanmar; World Bank firm capabilities, streamlining regulations to encourage 2018a). Access to adequate infrastructure in EAP remains informal enterprises to grow into more productive firms in fragmented, particularly in water and sanitation and the formal economy, and fostering innovation. transport, and in several lower-middle-income economies (World Bank 2018a). In these countries, strengthening the Reduce market distortions and level the playing field for efficiency of public investment management and fiscal private firms. A gradual transfer from public to private 76 CHAPTER 2.1 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.1.1 Labor productivity in East Asia and Pacific: Trends and drivers (continued) firm ownership in many cases, and greater involvement of Encouraging sectoral reallocation international firms, as well as reforms to lower entry costs and encourage fair competition, including in trade and Productivity gains from sectoral reallocation have slowed innovation, can help level the playing field for private in EAP. Policy measures to accelerate the process of firms and state owned enterprises. Curbing preferential reallocation again include reforms to allow the services lending agreements with state-owned enterprises and sector to thrive and absorb labor and measures to sustain easing the access of private firms to long-term funding can rapid urbanization. improve the allocative efficiency of capital and raise Liberalize service markets and shift out of agriculture. A productivity. Greater product market competition would gradual liberalization of service sectors, including spur innovation (Cusolito and Maloney 2018). education, health care, the financial sector, communications, transport, and utilities, could encourage Encourage innovation. Effective policies to promote job creation in these sectors (Beverelli, Fiorini, and innovation begin with strengthening managerial and Hoekman 2017). It could also boost manufacturing organizational practices (Cirera and Maloney 2017). In productivity, as services sectors provide important inputs addition, strengthening the effectiveness of research and into manufacturing. development (R&D) spending and measures to raise productivity in the services sectors are key (World Bank Encourage urbanization. The reallocation of factors, 2016c). Fiscal incentives for R&D are already in place in especially labor, from low-productivity agricultural some EAP countries (China, Malaysia), but in many other activities to higher-productivity manufacturing and cases R&D spending is small relative to GDP (Figure services can accelerate the convergence of EAP to the 2.1.1.4). Strengthening intellectual property rights regimes productivity frontier. Clarification of land ownership while avoiding undue limitations on competition could rights and transferable social benefits could encourage such also encourage R&D, as could competition for research labor movement (Fuglie et al 2019). Urban planning can grants. These reforms could be complemented by efforts encourage a reallocation of labor towards more productive that facilitate moving up the value chain through sectors by improving access to jobs, affordable housing, innovation, especially in R&D-intensive sectors, and public transportation, health care, education, and other enabling new business processes, including through services (World Bank 2015a). Road congestion, which is a digitization, and higher energy efficiency. major problem in many large cities may discourage job switching (World Bank 2018e, 2019f). Accelerated Address informality. The share of informal output in the productivity growth will also require improved EAP region is below the EMDE average while the share of management of country and regional transportation, informal employment is above average (World Bank telecommunications, and utility infrastructure in 2019f). Within the region, informality is higher in lower- metropolitan areas. income countries. However, even higher-income economies in EAP have urban informality (China, Creating a growth-friendly environment Malaysia, Thailand). To address challenges associated with Safeguard macroeconomic stability. Over the longer term, informality higher-income countries can prioritize urban strong and sustained productivity gains require financial planning along with the provision of essential social stability (Chapter 3; Box 3.4). Elevated corporate debt, protection to informal workers. Lower-income countries especially in China, weighs on investment and can focus on policies that encourage investment and productivity in exposed corporations. Policy measures to reduce costs of regulatory compliance. rein in financial risks are therefore critical. Growth in Europe and Central Asia decelerated to an estimated 2 percent in 2019, reflecting a sharp slowdown in Turkey as a result of acute financial market stress in 2018, as well as in the Russian Federation amid weak demand and cuts in oil production. Regional growth is projected to strengthen in 2020, to 2.6 percent, as activity recovers in Turkey and Russia, and to stabilize to 2.9 percent in 2021-22. Key external risks to the regional growth outlook include spillovers from weaker-than-expected activity in the Euro Area and escalation of global policy uncertainty. The region also remains vulnerable to disorderly commodity and financial market developments. A comprehensive reform agenda is needed to boost productivity, increase investment in physical and human capital, address continuing demographic pressures, and raise innovation. Recent developments further policy rate cuts, with some economies tightening policy to rein in inflation (Georgia, Growth in Europe and Central Asia (ECA) is Kazakhstan). Core inflation is also rising in some estimated to have decelerated markedly in 2019, economies, especially those with increasing wages to a three-year low of 2 percent (Table 2.2.1). The as a result of labor shortages and other capacity weak regional performance predominantly reflects constraints (Poland, Romania; Figure 2.2.1.D). slowdowns in the region’s two largest economies, In Russia, softer-than-expected investment and Russia and Turkey (Figure 2.2.1.A). trade, together with a continuation of A sustained weakness in exports growth has international economic sanctions, resulted in a continued amid slowing manufacturing activity growth slowdown to an estimated 1.2 percent. and investment. Sluggish new export orders in Industrial activity also softened, as oil production recent months suggest that export growth will cuts agreed with OPEC took effect and pipeline- continue to be weak in the near term, especially in related disruptions occurred. Retail sales volumes economies with deep trade and financial linkages weakened substantially following a VAT hike, to the Euro Area, such as those in Central Europe while consumer confidence remained low. The (Figure 2.2.1.B). central bank reversed a previous tightening stance, cutting the key policy rate five times since June. Headline inflation in ECA has eased, as the impact from the value-added tax (VAT) hike in In Turkey, industrial production and Russia and earlier currency depreciation in Turkey manufacturing data suggest that the economy faded. This, combined with weakening growth began to stabilize in late 2019, following the momentum, has allowed Russia and Turkey, as disruptions from acute financial market pressures well as other ECA economies, to pause or reverse in the previous year. Still, growth slowed sharply previous monetary policy tightening (Romania, for the year, falling 2.8 percentage points to near- Ukraine; Figure 2.2.1.C). Inflation remains above nil. Elevated inflation and associated pressures on or near target, however, limiting the scope for real incomes, as well as rising unemployment, dampened consumption. Investment contracted deeply, to rates comparable with the global Note: This section was prepared by Collette M. Wheeler. Research financial crisis, partly reflecting lingering policy assistance was provided by Vasiliki Papagianni and Julia Norfleet. uncertainty (Figure 2.2.1.E). Although the 78 CHAPTER 2.2 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 2.2.1 ECA: Recent developments contribution of net exports to growth was positive, Europe and Central Asia faced substantial headwinds in 2019 amid a this was due in large part to import compression. sharp slowdown in major economies, such as Turkey and Russia. Export In the second half of 2019, the central bank growth weakened significantly, particularly in Central Europe, which is sharply reversed its policy stance by cutting the tightly connected to the Euro Area through value chains. Headline inflation moderated in the region’s major economies, allowing for substantial policy policy rate in half, to 12 percent, despite above- rate cuts to support growth. Capacity constraints and a slowing Euro Area target inflation. weighed on activity in Central Europe. In Central Europe, the boost to private A. Contribution to regional GDP B. Export volume growth, growth by subregion consumption in early 2019 from rising real wages and government transfers helped to sustain above- potential growth. This impact dissipated by the end of the year, however, contributing to a slowdown in growth to an estimated 4.2 percent, despite an investment-led construction sector pickup in some economies (Hungary, Romania). The slowdown in the Euro Area weighed on exports in some cases (Bulgaria, Romania; Figure 2.2.1.F). C. Real interest rates and bond D. Core inflation and capacity spreads in ECA utilization in Central Europe In the Western Balkans, a deceleration in public investment (Kosovo), manufacturing (Serbia), and export growth (Albania, Serbia) contributed to a moderation in growth to an estimated 3.2 percent in 2019. Temporary factors related to weather and energy production dampened activity in Albania, while strong import demand for public investment projects led to negative contribution of net exports in Montenegro. In Eastern Europe, industrial production growth has softened, reflecting marked E. GDP and investment growth in F. Industrial production and export Turkey volume growth in Central Europe weakness in manufacturing amid slowing export growth, particularly in Belarus. Ukraine, however, benefited from a bumper crop harvest in the first half of 2019. Firming growth in the South Caucasus, to an estimated 3.7 percent in 2019, was supported by private consumption, and on the supply side by strong manufacturing growth, as well as by a recovery in mining production in Armenia. Source: Consensus Economics; Haver Analytics; J.P. Morgan; Organisation for Economic Co- operation and Development; World Bank. Expanding natural gas production and steady A.B. ECA = Europe and Central Asia. Aggregate growth rates calculated using GDP weights at 2010 growth in non-energy sectors supported prices and market exchange rates. Data for 2019 are estimated. Refer to Table 2.2.1 for further details. Azerbaijan’s economy in the first half of 2019. In B. South Cauc. = South Caucasus. C. Real interest rates calculated using the policy interest rate less the Consensus Economics Georgia, growth strengthened despite the forecast for inflation. Bond spreads are from the J.P. Morgan Emerging Market Bond Index (EMBI). Sample includes Hungary, Poland, Russia, Turkey, and Ukraine, due to data availability. Last imposition of travel restrictions by Russia. In observation is December 16, 2019 for the bond spread and November 2019 for the real interest rate. Central Asia, the cyclical expansion moderated, D. Aggregates calculated using GDP weights at 2010 prices and market exchange rates. The sample includes Hungary, Poland, and Romania. Last observation is 2019Q4 for capacity utilization and yet growth was still robust at 4.5 percent in 2019. 2019Q3 for core inflation. E. Figure uses the annualized 4-quarter on 4-quarter average. In Kazakhstan, the largest subregional economy, F. Dashed lines represent the 2000-18 average. Aggregate industrial production calculated using production weights at 2010 prices and market exchange rates; aggregate export volume growth slowing exports from lower oil prices were offset calculated using GDP weights at 2010 prices and market exchange rates. by fiscal expansion. Click here to download data and charts. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 E URO PE AND CE NTRAL AS IA 79 FIGURE 2.2.2 ECA: Outlook and risks Outlook Growth in Europe and Central Asia is projected to firm to 2.6 percent in Regional growth is expected to firm over the 2020, as activity recovers in Turkey and Russia, and to stabilize at 2.9 percent in 2021-22. Weaker-than-expected growth in the Euro Area or forecast horizon, to 2.6 percent in 2020 and 2.9 China could dent activity in tightly connected subregions. Heightened percent in 2021-22, on the assumptions that key policy uncertainty in the broader region could impact portfolio flows to commodity prices and growth in the Euro Area ECA, while the future of funding options in Central Europe remains uncertain. The ability to confront growth headwinds is reduced by limited stabilize, and that Turkey’s economy recovers fiscal space. from earlier financial pressures and Russia firms on the back of policy support (Figure 2.2.2.A). A. Growth B. Share of exports by destination, 2017 Considerable variation across economies is expected to continue. Economies in Central Europe are anticipated to slow as fiscal policy support wanes and demographic pressures persist, while those in Central Asia are projected to continue growing at a robust pace, and more rapidly than previously envisaged, on the back of structural reform progress (World Bank 2019g). The baseline projection for regional growth also C. Turkey: Credit to firms and D. Gross portfolio outflows in ECA non-performing loan ratio and Euro Area economic policy assumes that trade tensions between the United uncertainty States and China will not re-escalate; the United Kingdom’s exit from the European Union will be orderly; and that fiscal and monetary policy avert further financial market turbulence in Turkey as the country moves past acute financial stress. In Russia, growth is projected to firm moderately, to 1.8 percent by 2021 (Table 2.2.2). Despite OPEC and its partners recently announcing deeper cuts until March 2020, oil production is E. European Union structural fund F. Gross government and external payments to Central Europe, 2019 debt, by subregion expected to remain stable in Russia due to an exemption on gas condensates (IEA 2019). National Projects—which are partly funded by the 2019 VAT hike and include investment in infrastructure and human capital—are expected to buoy growth over the forecast horizon. Nevertheless, private investment remains tepid in the projection, due to policy uncertainty and slowing potential growth over the longer term as demographic pressures increase, and as structural Source: Baker, Bloom, and Davis (2016); Bank for International Settlements; European Commission; Haver Analytics; Institute of International Finance; International Monetary Fund; World Bank. problems, such as the lack of competition, A. Aggregate growth rates calculated using GDP weights at 2010 prices and market exchange rates. Shaded areas indicate forecasts. Data for 2019 are estimated. Yellow diamonds correspond to accumulate. forecasts from the June 2019 Global Economic Prospects report. B. The “Within region” data refer to within ECA for all stacked bars except the Euro Area, which refers to within the Euro Area. Shares are calculated from exports in millions of U.S. dollars. Growth is projected to recover in Turkey, to 3 C. Last observation is November 2019 for non-performing loans and 2019Q2 for credit to non-financial corporations. percent in 2020, as investment recovers from a D. EPU = economic policy uncertainty. The Euro Area economic policy uncertainty is calculated by Baker, Bloom, and Davis (2016), which is based on the frequency of words in domestic newspapers deep contraction in 2019. Gradual improvement mentioning economic policy uncertainty. Figure shows the 3-month moving average. Sample for portfolio outflows includes Hungary, Poland, and Turkey, due to data availability. Last observation is in domestic demand is expected to support growth December 20, 2019 for portfolio outflows and November 2019 for economic policy uncertainty. over the forecast horizon. This outlook assumes E. Note: Figure shows the cumulated payments to EMDEs in Central Europe for the 2014-20 program period. Data on the “Share of country allocation” reflect the amount of EU structural fund that fiscal and monetary policy remain steady, that allocations paid out to each economy as a share of its total EU structural fund allocation as of 2019Q2, due to data availability. F. ECA = Europe and Central Asia, South Cauc. = South Caucasus, Gov. = government, Ext. = external. Aggregates calculated using the median. Sample includes 24 economies. Click here to download data and charts. 80 CHAPTER 2.2 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 remain steady, that the currency does not come capital to boost the quality of education and under pressure, and that corporate debt reduce skills mismatches. restructurings proceed smoothly. Growth is expected to firm over the forecast Central Europe is forecast to sharply decelerate horizon in Eastern Europe and stabilize in Central over the forecast horizon, to 3.4 percent GDP Asia, but growth in both subregions is subject to growth in 2020 and 3 percent by 2022. Fiscal considerable policy uncertainty. These subregions support, and the resulting private consumption face a challenging external environment as growth boost, will begin to fade in some of the subregion’s remains tepid in key trading partners, including largest economies by 2020-21, with limited fiscal the Euro Area and Russia (for Eastern Europe) space available to fully offset potential adverse and China (for Central Asia). Ukraine, which is spillovers from the Euro Area (Poland, Romania). the largest economy in Eastern Europe, recently Shrinking working-age populations, partly reached a preliminary agreement with the IMF for reflecting emigration to Western Europe in recent a $5.5 billion program, which should help years, limits growth prospects. Progress on advance structural reforms and foster growth over structural reforms is key to support private the forecast horizon. In Central Asia, growth is investment growth over the medium term. expected to slightly moderate this year following Growth in the subregion is highly dependent on agreed-upon production cuts by a non-OPEC the continued absorption of EU structural funds, partner (Kazakhstan). Activity in Kazakhstan will with the current cycle expected to end in 2020. likely be dampened by the waning effect of earlier fiscal stimulus, modest or slowing growth in key Growth is projected to firm to 3.8 percent by trading partners (Russia, China), and low 2021 in the Western Balkans, assuming political productivity. In Eastern Europe and Central Asia, instability and policy uncertainty remain the pace of growth depends on the successful contained. Rising fiscal liabilities in the subregion, implementation of structural reforms to improve in some cases due to large public sector wage the business environment, achieve debt increases, social transfers, or higher-than-expected sustainability, and restructure state-owned costs for infrastructure projects, could reduce enterprises to improve competition (EBRD 2017; space for future countercyclical fiscal stimulus and Funke, Isakova, and Ivanyna 2017). weaken the business climate (Kosovo, Montenegro, North Macedonia). Additionally, Risks recent earthquakes in the Western Balkans— primarily affecting Albania—took a heavy toll on The regional outlook remains subject to human life and physical infrastructure. The significant downside risks, including slowing outlook for the subregion remains challenging as growth in major trading partners, geopolitical falling business confidence and heightened turbulence, heightened policy uncertainty, uncertainty coincide with a worsening external exposure to disorderly financial market environment (World Bank 2019h). developments, as well as weakening productivity growth over the long run (Box 2.2.1). A sharper- The South Caucasus is estimated to grow 3.7 than-expected slowdown in the Euro Area, ECA’s percent in 2019, and to decelerate to 3.1 percent most important trading partner, could generate over the remaining forecast horizon. In negative spillovers in economies with tightly Azerbaijan, activity is expected to be dampened by linked trade and financial ties (Figure 2.2.2.B; the effects of subdued oil prices and lingering Elekdag, Muir, and Wu 2015). Slowing growth in structural rigidities in the non-oil sector. Longer- China could be propagated through trade and term growth depends on continuation of domestic commodity price channels to Central Asia, as well reforms to enhance private sector development as metals exporters in the ECA region, which are and address fragilities emanating from the increasingly reliant on China as an export financial sector, as well as investment in human destination. The region’s energy exporters— G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 E URO PE AND CE NTRAL AS IA 81 Azerbaijan, Kazakhstan, and Russia—remain regional economies, particularly energy and metals vulnerable to large swings in global commodity exporters. Additional sanctions on Russia could prices, particularly when accompanied by have a negative impact on the region, particularly heightened volatility (van Eyden et al. 2019). in economies where domestic demand relies on remittances (Central Asia, Eastern Europe, the In many economies in Central Europe, the policy South Caucasus). space to confront negative shocks is limited by persistent budget deficits. Fiscal support has Although policy uncertainty surrounding the contributed to these growing imbalances— United Kingdom’s exit from the European Union increasing public sector wages, rising government has dissipated somewhat, the process remains transfers, and low tax capacity, have widened fiscal vulnerable to disruption until the end of the deficits, with the latter approaching or exceeding 3 transition period, currently scheduled for the end percent of GDP—the upper limit of the EU of the year (Bank of England 2018; H.M. threshold, particularly in Romania. Across ECA, Government 2018). The future program of EU public sector debt relative to GDP is higher than structural funds after 2020 must also be prior to the global financial crisis, with the largest determined, with the potential redirection of EU increases observed in Eastern Europe and the funds to advanced economies in Southern Europe South Caucasus. In Turkey, recent policies to limiting funding options for Central Europe. In support growth through credit expansion run the several countries, structural fund payments risk of worsened external imbalances (Figure represented 5 percent or more of GDP over the 2.2.2.C). last program period from 2014-20 (Figure 2.2.2.E). Historically, when the absorption of EU Following military disagreements with the North funding was low, activity also decelerated Atlantic Treaty Organization, Turkey faces a new substantially, as was the case in Poland in 2016. round of U.S. economic sanctions. Renewed involvement in conflicts in the Syrian Arab An unexpected tightening of global financing Republic or Ukraine, could trigger additional conditions could generate financial market sanctions against large economies in the region. pressures in ECA, renewing capital outflows and currency volatility, particularly in economies with More generally, a pervasive rise in policy large external financing needs (Chapter 1; World uncertainty could undermine business and Bank 2019f; EBRD 2019a). Many regional investor sentiment (Figure 2.2.2.D). The re- economies have relied on short-term capital escalation of trade tensions, and a resulting inflows to finance large current account deficits. slowdown in global demand, could weaken Low foreign-currency reserves leave these exports and commodity prices for the region, economies all the more vulnerable to capital flight presenting challenges to growth and fiscal and constrain the capacity of central banks to planning. Renewed trade tensions between the buffer the impact of negative external shocks. A United States and Europe, particularly with fall in incomes in the region’s largest economies respect to vehicle and auto part tariffs, could also would dent remittance inflows to Eastern Europe sideswipe the ECA region, especially for and Central Asia (World Bank 2016d). On the economies in Central Europe that are tightly domestic front, increased public spending and low integrated into European value chains. Similarly, a tax capacity have contributed to historically high triggering of trade tensions between the United public debt levels, which limit fiscal policy space States and China could adversely affect some (Figure 2.2.2.F). 82 CHAPTER 2.2 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 TABLE 2.2.1 Europe and Central Asia forecast summary Percentage point differences (Real GDP growth at market prices in percent, unless indicated otherwise) from June 2019 projections 2017 2018 2019e 2020f 2021f 2022f 2019e 2020f 2021f EMDE ECA, GDP1 4.1 3.2 2.0 2.6 2.9 2.9 0.4 -0.1 0.0 EMDE ECA, GDP excl. Turkey 3.0 3.3 2.6 2.5 2.6 2.6 0.2 -0.1 0.0 (Average including countries with full national accounts and balance of payments data only)2 EMDE ECA, GDP2 4.1 3.1 1.8 2.5 2.8 2.8 0.4 -0.1 -0.1 GDP per capita (U.S. dollars) 3.6 2.7 1.5 2.3 2.6 2.7 0.4 -0.1 -0.1 PPP GDP 4.0 3.1 1.9 2.6 2.8 2.8 0.4 0.0 -0.1 Private consumption 5.0 2.6 1.5 2.6 2.6 2.6 0.2 0.0 -0.1 Public consumption 3.4 2.0 1.6 1.7 1.8 1.8 0.0 0.1 0.0 Fixed investment 6.7 3.1 -0.6 4.3 4.2 4.2 0.5 1.0 0.6 Exports, GNFS 3 7.3 5.6 2.8 2.3 3.4 3.4 -1.2 -2.0 -0.6 Imports, GNFS3 11.5 3.2 1.3 4.1 4.9 5.0 -1.7 -1.4 -0.9 Net exports, contribution to growth -0.8 1.0 0.6 -0.5 -0.3 -0.4 0.1 -0.4 0.1 Memo items: GDP Commodity exporters4 2.1 2.6 1.9 2.1 2.3 2.3 0.1 -0.1 0.0 Commodity importers 5 6.1 3.7 2.1 3.1 3.5 3.4 0.7 0.0 0.0 Central Europe6 5.1 4.7 4.2 3.4 3.1 3.0 0.5 0.1 0.0 Western Balkans 7 2.6 4.0 3.2 3.6 3.8 3.8 -0.3 -0.2 -0.1 Eastern Europe8 2.6 3.3 2.8 2.9 3.1 3.1 0.4 0.2 0.1 South Caucasus 9 1.7 2.6 3.7 3.1 3.1 3.1 0.0 -0.8 -1.1 Central Asia10 4.6 4.7 4.5 4.4 4.6 4.5 0.3 0.4 0.5 Russia 1.6 2.3 1.2 1.6 1.8 1.8 0.0 -0.2 0.0 Turkey 7.5 2.8 0.0 3.0 4.0 4.0 1.0 0.0 0.0 Poland 4.9 5.1 4.3 3.6 3.3 3.1 0.3 0.0 0.0 Source: World Bank. Note: e = estimate; f = forecast. EMDE = emerging market and developing economies. World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of countries’ prospects do not differ at any given moment in time. 1. GDP and expenditure components are measured in 2010 prices and market exchange rates. 2. Aggregates presented here exclude Bosnia and Herzegovina, Kazakhstan, Kosovo, Montenegro, Serbia, Tajikistan, and Turkmenistan, for which data limitations prevent the forecasting of GDP components. 3. Exports and imports of goods and non-factor services (GNFS). 4. Includes Albania, Armenia, Azerbaijan, Kazakhstan, the Kyrgyz Republic, Kosovo, Russia, Tajikistan, Turkmenistan, Ukraine, and Uzbekistan. 5. Includes Belarus, Bosnia and Herzegovina, Bulgaria, Croatia, Georgia, Hungary, Moldova, Montenegro, North Macedonia, Poland, Romania, Serbia, and Turkey. 6. Includes Bulgaria, Croatia, Hungary, Poland, and Romania. 7. Includes Albania, Bosnia and Herzegovina, Kosovo, Montenegro, North Macedonia, and Serbia. 8. Includes Belarus, Moldova, and Ukraine. 9. Includes Armenia, Azerbaijan, and Georgia. 10. Includes Kazakhstan, the Kyrgyz Republic, Tajikistan, Turkmenistan, and Uzbekistan. Click here to download data. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 E URO PE AND CE NTRAL AS IA 83 TABLE 2.2.2 Europe and Central Asia country forecasts1 Percentage point differences (Real GDP growth at market prices in percent, unless indicated otherwise) from June 2019 projections 2017 2018 2019e 2020f 2021f 2022f 2019e 2020f 2021f Albania 3.8 4.1 2.9 3.4 3.6 3.5 -0.8 -0.3 -0.2 Armenia 7.5 5.2 6.9 5.1 5.2 5.2 2.7 0.2 0.0 Azerbaijan -0.3 1.4 2.5 2.3 2.1 2.1 -0.8 -1.2 -1.6 Belarus 2.5 3.0 1.0 0.9 0.5 0.5 -0.8 -0.4 -0.7 Bosnia and Herzegovina2 3.2 3.6 3.1 3.4 3.9 3.9 -0.3 -0.5 -0.1 Bulgaria 3.5 3.1 3.6 3.0 3.1 3.1 0.6 0.2 0.3 Croatia 3.1 2.7 2.9 2.6 2.4 2.4 0.4 0.1 0.0 Georgia 4.8 4.8 5.2 4.3 4.5 4.5 0.6 -0.5 -0.5 Hungary 4.3 5.1 4.9 3.0 2.6 2.6 1.1 0.2 0.0 Kazakhstan 4.1 4.1 4.0 3.7 3.9 3.7 0.5 0.5 0.7 Kosovo 4.2 3.8 4.0 4.2 4.1 4.0 -0.4 -0.3 -0.4 Kyrgyz Republic 4.7 3.5 4.2 4.0 4.0 4.2 -0.1 0.0 -0.1 Moldova 4.7 4.0 3.6 3.6 3.8 3.8 0.2 0.0 0.0 Montenegro 4.7 5.1 3.0 3.1 2.8 3.2 0.1 0.7 0.5 North Macedonia 0.2 2.9 3.1 3.2 3.3 3.1 0.2 0.0 -0.3 Poland 4.9 5.1 4.3 3.6 3.3 3.1 0.3 0.0 0.0 Romania 7.1 4.0 3.9 3.4 3.1 3.1 0.3 0.1 0.0 Russia 1.6 2.3 1.2 1.6 1.8 1.8 0.0 -0.2 0.0 Serbia 2.0 4.4 3.3 3.9 4.0 4.0 -0.2 -0.1 0.0 Tajikistan 7.1 7.3 6.2 5.5 5.0 5.0 0.2 -0.5 -1.0 Turkey 7.5 2.8 0.0 3.0 4.0 4.0 1.0 0.0 0.0 Turkmenistan 6.5 6.2 5.0 5.2 5.5 5.5 -0.6 0.1 0.6 Ukraine 2.5 3.3 3.6 3.7 4.2 4.2 0.9 0.3 0.4 Uzbekistan 4.5 5.1 5.5 5.7 6.0 6.0 0.2 0.2 0.0 Source: World Bank. Note: e = estimate; f = forecast. World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of countries’ prospects do not significantly differ at any given moment in time. 1. GDP and expenditure components are measured in 2010 prices and market exchange rates, unless indicated otherwise. 2. GDP growth rate at constant prices is based on production approach. Click here to download data. 84 CHAPTER 2.2 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.2.1 Labor productivity in Europe and Central Asia: Trends and drivers Productivity growth in Europe and Central Asia (ECA) has fallen from an above-EMDE-average pre-crisis rate of 5.5 percent to a below-EMDE-average post-crisis rate of 1.6 percent—the steepest decline of any EMDE region. There has been wide heterogeneity within the region, however, with productivity growth near zero since 2013 in the Western Balkans and above 2.5 percent in Central Europe. In the Western Balkans and the Russian Federation, investment weakness has weighed on productivity growth. The productivity slowdown in ECA has predominantly reflected weaker within-sector productivity growth, with a particularly sharp decline in the growth of services productivity, and weaker total factor productivity growth in Eastern Europe, the South Caucasus, and the Western Balkans. Sectoral reallocation has also slowed in the post-crisis period, reflecting headwinds that have limited the ability of firms with higher productivity to continue to absorb additional labor from less productive sectors. A comprehensive reform agenda is needed to boost investment in physical and human capital, address continuing demographic pressures, and raise innovation. Such reforms are also needed to improve business climates and governance, reduce the role of the state in the economy, and promote the diversification of commodity-dependent economies. Introduction • What have been the factors associated with productivity growth in the region? Productivity growth in the Europe and Central Asia (ECA) region has fallen from an above-EMDE-average • What policy options are available to boost regional pre-crisis (2003-08) rate of 5.5 percent to a below-average productivity growth? post-crisis (2013-18) rate of 1.6 percent—the steepest decline of any EMDE region (Figure 2.2.1.1). For the purposes of this box, productivity, unless otherwise Productivity levels in ECA in 2018 were one-half above indicated, refers to labor productivity, defined as real GDP the EMDE average, but only 30 percent of the advanced- (at 2010 prices and market exchange rates) per economy average. The sharp post-crisis slowdown in worker. The data refer to a sample of 21 ECA economies: productivity growth has significantly reduced the pace of Kosovo, Turkmenistan, and Uzbekistan are excluded in ECA’s convergence with advanced economies. some analysis due to limited data availability.1 Within the ECA region, there is wide heterogeneity across Evolution of regional productivity economies. Productivity growth in Central Europe has been solid in the post-crisis period, at 2.6 percent, while it Sharp post-crisis productivity growth slowdown. In the has been near zero in Russia and the Western Balkans. The steepest post-crisis decline of any EMDE region, average region’s agricultural commodity exporters, most of which productivity growth in ECA fell to 1.6 percent in 2013- are in Central Asia (excluding Kazakhstan) and Eastern 18, below the EMDE average, from the above-average rate Europe, have ECA’s lowest productivity levels, at 3 and 10 of 5.5 percent in 2003-08. This slowdown was percent of the advanced-economy average, respectively. In broad-based across the region, affecting nearly all contrast, Poland and Turkey have productivity levels over economies, with post-crisis productivity growth below 35 percent of the advanced-economy average, reflecting longer-term (1992-2018) averages in roughly two-thirds of their integration into global value chains and roles as the region’s economies (Figure 2.2.1.2). regional financial centers. Central Europe, whose economies are members of the European Union (EU), is Within-region heterogeneity. There has been wide deeply embedded in Western European supply chains and heterogeneity within the region. The productivity growth has the highest productivity of the ECA subregions, at 34 slowdown was particularly steep in the South Caucasus percent of the advanced-economy average. and Russia, as well as in the Western Balkans, the latter of which was hit by the Euro Area crisis of 2010-12 amid Against this backdrop, this box addresses the following already elevated unemployment rates. In contrast, the questions. deceleration was milder in Central Europe, which is better • How has productivity growth evolved in the ECA region? 1 Central Europe includes Bulgaria, Croatia, Hungary, Poland, and Romania. Western Balkans includes Albania, Bosnia and Herzegovina, Kosovo, Montenegro, North Macedonia, and Serbia. Eastern Europe Note: This box was prepared by Collette M. Wheeler, building upon includes Belarus, Moldova, and Ukraine. South Caucasus includes Arme- analysis in Chapter 3. Research assistance was provided by Vasiliki nia, Azerbaijan, and Georgia. Central Asia includes Kazakhstan, the Papagianni and Shijie Shi. Kyrgyz Republic, Tajikistan, Turkmenistan, and Uzbekistan. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 E URO PE AND CE NTRAL AS IA 85 BOX 2.2.1 Labor productivity in Europe and Central Asia: Trends and drivers (continued) integrated into global supply chains, and Central Asia, FIGURE 2.2.1.1 Productivity in ECA which has growing economic ties with China. compared with other regions Productivity growth in Europe and Central Asia (ECA) • South Caucasus. The post-crisis decline in productivity has fallen from an above-EMDE-average pre-crisis rate growth was most pronounced, at 14 percentage of 5.5 percent to a below-EMDE-average post-crisis rate points, in the South Caucasus. After reaching double- of 1.6 percent—the steepest decline of any EMDE digit annual productivity growth pre-crisis, the region. Convergence toward advanced economies slowed in the post-crisis period, after having been the subregion suffered several post-crisis shocks, including fastest among EMDE regions in the pre-crisis period. conflict (Georgia), bouts of violence (Armenia), and a Productivity levels in ECA, while above the EMDE plunge in commodity prices (Armenia, Azerbaijan, average, are still one-third of those in advanced Georgia). economies. • Western Balkans. Productivity growth also markedly A. Average annual productivity growth in EMDE regions declined in the Western Balkans (by 5.4 percentage points), where the Euro Area crisis disrupted financial intermediation, including foreign bank retrenchment, and progress on structural reforms stalled. Since 2013, productivity growth in this subregion has been near zero. • Russia. Amid international sanctions and the 2014-16 oil price collapse, Russia’s productivity growth was, on average, near zero in 2013-18—a sharp decline from 6.0 percent in 2003-08. • Turkey. Productivity growth more than halved relative to its pre-crisis average, slowing to 2.1 percent in 2013-18, as the economy faced political and B. Productivity levels and convergence in EMDE regions economic shocks. • Central Europe. Annual productivity growth slowed by just over 1 percentage point, from 3.7 percent to 2.6 percent, in Central Europe in the post-crisis period, in tandem with the modest slowdown in the Euro Area. This partly reflects the close integration of this subregion with Western European supply chains. Notwithstanding anemic Euro Area growth since the global financial crisis, Central Europe achieved the second highest productivity growth of any ECA subregion in 2013-18, after only Central Asia. This, in part, reflected buoyed investment, which was supported by the absorption of EU structural funds. Source: Penn World Table; The Conference Board; World Development Indicators, World Bank. Note: ECA = Europe and Central Asia, EMDE = emerging market and • Eastern Europe. Annual productivity growth in developing economies. Aggregate regional growth rates calculated using Eastern Europe slowed by about 4.5 percentage points GDP weights at 2010 prices and market exchange rates. Unless otherwise specified, productivity refers to labor productivity, defined as output per from pre-crisis rates, to 2.2 percent during 2013-18. worker. Sample includes 127 EMDEs, of which 21 are ECA economies. Productivity growth averaged only 1.4 percent in A. Blue bars denote the range across six (GDP-weighted) averages for EMDE regions. Yellow bars denote the simple average of the six EMDE 2013-16, which reflected the dual shocks of conflict regional averages. B. Rate of convergence calculated as the difference in productivity growth in Ukraine and a commodity price plunge, but picked rates with the average advanced economy divided by the log difference in up in the next two years. productivity levels with the average advanced economy. Regional rate of convergence is the GDP-weighted average of EMDE economies of each region. “Level” of productivity refers to the GDP-weighted average of regional • Central Asia. Central Asia insulated itself somewhat productivity as a share of the average advanced economy during 2013-18. Advanced-economy sample includes 35 advanced economies. from the impact of the oil price slump of 2014-16 Click here to download data and charts. 86 CHAPTER 2.2 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.2.1 Labor productivity in Europe and Central Asia: Trends and drivers (continued) FIGURE 2.2.1.2 Evolution of productivity in ECA The post-crisis slowdown in productivity growth has affected nearly all the economies in ECA. There is wide heterogeneity within the region, however, with productivity growth near-zero since 2013 in Russia and the Western Balkans but above 2.5 percent in Central Asia and Central Europe. The post-crisis productivity growth slowdown has reflected a sharp deceleration in total factor productivity growth in Eastern Europe, the South Caucasus, and the Western Balkans but investment weakness in Russia and Central Europe. A. Productivity growth in ECA B. Productivity growth in Central Asia, C. Productivity levels relative to South Caucasus, and Western Balkans advanced-economy average, 2018 D. Share of economies with productivity E. Contribution to productivity growth F. Contribution to productivity growth, by growth below long-run and pre-crisis ECA subregion averages, 2013-18 Source: Barro and Lee (2015); Haver Analytics; International Monetary Fund; Penn World Table; United Nations; Wittgenstein Centre for Demography and Global Human Capital; World Development Indicators, World Bank. Note: CA=Central Asia, CE=Central Europe, KAZ=Kazakhstan, SC=South Caucasus, WBK=Western Balkans. Unless otherwise specified, productivity refers to labor productivity, defined as output per worker. A.-F. Aggregate growth rates calculated using GDP weights at 2010 prices and market exchange rates. C. Figure shows 2018 subregional productivity levels as a share of 2018 advanced-economy weighted average. Sample includes 35 advanced economies and 21 ECA economies. D. Figure shows the share of economies for which average productivity growth in 2013-18 was lower than a long-run (1992-2018) and the pre-crisis (2003-08) average. Sample includes 127 EMDEs, of which 21 are ECA economies. E.F. Productivity defined as output per worker in 2010 U.S. dollars. Samples are unbalanced due to data availability, and include up to 21 ECA economies and 92 EMDEs. Click here to download data and charts. and recession in Russia during 2015-16 by pivoting High productivity levels relative to EMDEs, but with its exports toward China. By 2018, China had wide range. Partly as a result of rapid productivity growth become the second largest export market for Central in 2003-08, the average productivity level in ECA in Asia after the Euro Area, accounting for 20 percent of 2013-18 was 30 percent of the advanced-economy exports. As a result, the subregion’s productivity average—roughly one-half above the EMDE average. The growth slowed mildly in comparison to the rest of the ECA average, however, masks wide divergences across region, by 2.3 percentage points to 3.4 percent in ECA subregions, from 3 and 10 percent of the advanced- 2013-18—the fastest productivity growth in ECA in economy average in predominantly agricultural the period. commodity-exporting Central Asia (excluding Kazakhstan) G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 E URO PE AND CE NTRAL AS IA 87 BOX 2.2.1 Labor productivity in Europe and Central Asia: Trends and drivers (continued) and Eastern Europe, respectively, to 34 percent of the facing notable challenges with corporate over-indebtedness advanced-economy average in Central Europe, which is and market concentration (EBRD 2018a). In both Turkey deeply embedded into Euro Area supply chains and which and Central Asia, the sources of the productivity has benefited from the absorption of EU structural funds. deceleration were broad-based, reflecting a slowdown in Poland and Turkey had productivity levels above 35 physical capital deepening and human capital percent of the advanced-economy average, partly reflecting improvements, as well as in TFP growth, particularly in their openness to trade and positions as regional financial Kazakhstan. In Central Asia, Eastern Europe, and the centers (World Bank 2014; World Bank 2019i). Since the Western Balkans, reform momentum has also slowed, with global financial crisis, the pace of convergence to advanced many of these economies falling short of completing the -economy productivity levels in the ECA region as a whole transition to competitive and inclusive markets. has slowed sharply, to average less than 1 percent per year over 2013-18—one-fifth of its rate in 2003-08. Sources of regional productivity growth Sources of productivity growth. Labor productivity Post-crisis slowdown across all sectors. Pre-crisis growth can be decomposed into its sources: Factor productivity growth in ECA was mostly driven by shifts of accumulation (human or physical capital) and advances in resources from agriculture and industry to higher- the efficiency of factor use (total factor productivity, or productivity services sectors, partly as a result of continued TFP). Two-thirds of the post-crisis slowdown in reforms to address resource misallocation inherited from productivity growth in ECA is estimated to have been due central planning (World Bank 2008). The post-crisis to slowing capital accumulation—partly reflecting weak period, however, was marked by weakness of growth across investment amid lower foreign direct investment (FDI) all sectors as a slowdown in manufacturing, exacerbated by inflows and declining commodity prices—and one-third dwindling global trade growth and a collapse in to slowing TFP growth, compared with about equal commodity prices, spilled over to services (Figure 2.2.1.3; contributions of these sources in the average EMDE. Orlic, Hashi, and Hisarciklilar 2018). In contrast to the EMDE average, the contribution of services to In Russia and Central Europe, particularly Bulgaria and productivity growth in 2013-15 was negative in ECA, Romania, weakening capital services deepening accounted likely reflecting, in part, spillovers from the Euro Area debt for most (three-quarters) of the slowdown in productivity crisis and the continued migration of skilled labor to growth in the post-crisis period. In Russia, international Western Europe. sanctions, combined with the 2014-16 oil price plunge, deterred investment, which was further dampened by the Sectoral reallocation as a source of productivity growth in weak business environment (Russell 2018). Although EU ECA. Resource reallocation toward more productive structural funds have buoyed overall investment in Central sectors accounted for almost half of ECA’s productivity Europe, they have not fully offset weakness in machinery growth in the 1990s, as output of the region’s services and equipment investment, which has been due partly to sectors increased by nearly 15 percentage points of GDP reduced commercial credit supply (Gradzewicz et al. 2018; (World Bank 2008; Arnold, Javorcik, and Mattoo 2011; Levenko, Oja, and Staehr 2019). World Bank 2015b). In contrast, the surge in productivity growth of 2003-08 mostly reflected within-sector growth, In contrast, reduced TFP growth has been the main source as firms in Central Europe became integrated into Euro (accounting for three-quarters) of the productivity growth Area supply chains, technology transfer accelerated, and slowdown in Eastern Europe, the Western Balkans, and the services sectors became more liberalized.2 After the the South Caucasus. This has partly reflected pockets of global financial crisis, however, within-sector productivity conflict and violence (Armenia, Georgia, Ukraine). However, private and public investment has also been weak in the post-crisis period, contributing to reduced 2 As economies in Central Europe initiated the process to join the TFP growth. As a result of weak investment, these European Union, structural policies that boosted competition and subregions face large infrastructure gaps, particularly in facilitated integration with global value chains helped spur within-sector transport and telecommunications networks, which limits growth, particularly within services. Thus, liberalization of services sectors the capacity to promote regional integration and, for is likely to have increased the average productivity of incumbent firms and facilitated the entry of new and more innovative firms. Please refer to energy exporters, diversification (IMF 2014). Obstacles to Bartelsman and Scarpetta (2007); Brown and Earle (2007); Georgiev, private sector development also constrain TFP in these Nagy-Mohacsi, and Plekhanov (2017); Shepotylo and Vakhitov (2015); subregions, with certain economies in the Western Balkans and World Bank (2008) for further detail. 88 CHAPTER 2.2 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.2.1 Labor productivity in Europe and Central Asia: Trends and drivers (continued) FIGURE 2.2.1.3 Factors supporting productivity growth in ECA Within-sector productivity growth—the main driver of pre-crisis productivity growth in ECA—fell sharply in the post-crisis period, and productivity gains from sectoral reallocation halved as economies moved to services sectors with relatively low productivity levels. The deceleration of productivity reflected slower improvements in a broad range of its fundamental drivers. A. Contribution to productivity growth B. Sectoral productivity levels C. Sectoral contribution to productivity growth D. Sectoral composition of GDP E. Share of EMDEs with a post-crisis F. Drivers of productivity, 2017 slowdown in growth of underlying drivers of productivity Source: APO productivity database; Expanded African Sector Database; Groningen Growth Development Center Database; Haver Analytics; International Country Risk Guide; ILOSTAT; Observatory of Economic Complexity; Organisation of Economic Co-operation and Development STAN; Penn World Table; United Nations; World Bank; World KLEMS. Note: Unless otherwise specified, productivity refers to labor productivity, defined as output per worker. A.-D. The sample includes 6 ECA economies and 46 EMDEs. A.D. Aggregates calculated using GDP weights at 2010 prices and market exchange rates. A. Growth “within sector” shows the contribution to aggregate productivity growth of each sector holding employment shares fixed. The “between sector” effect shows the contribution arising from changes in sectoral employment shares. B. Figure shows the median of country groups. C. Figure shows median values. “Manufacturing” includes manufacturing, mining and utilities; “Services” includes financial and business services, government and personal services, trade services, and transport services. D. Figure shows the share of total value added within each sector. “Manufacturing” includes mining and utilities; “Finance” includes business services. E. Post-crisis slowdown defined as a decline in the growth of each variable during 2008-17 compared to growth in the pre-crisis period, defined as 1998-2007. The blue bars represent share of 21 economies in Europe and Central Asia economies where improvements in each driver of productivity were lower during 2008-17 than in the pre-crisis period 1998-2007 or changes in 2008-17 were below zero. Orange diamond is the corresponding values for EMDE countries. Variables corresponding to each concept and their sample sizes are: Institutions=government effectiveness (20 ECAs; 126 EMDEs), Innovation=patents per capita (15 ECAs; 43 EMDEs), Investment=investment to GDP ratio (21 ECAs; 109 EMDEs), Income equality=(-1)*Gini (21 ECAs; 121 EMDEs), Urbanization=urban population percentage (21 ECAs; 127 EMDEs), Complexity = Hidalgo and Hausmann (2009)'s Economic Complexity Index (17 ECAs; 79 EMDEs), Education=years of schooling (17 ECAs; 103 EMDEs), Demography=share of working-age population (21 ECAs; 127 EMDEs), Gender equality= female average years of education divided by male average years (17 ECAs; 102 EMDEs). Green horizontal line indicates 50 percent. F. Figure shows the unweighted average levels of drivers normalized as an average of AEs as 100 and standard deviation of 10. Blue bars represent average within Europe and Central Asia economies in 2017. Orange whiskers represent the range of the average drivers for six regions in 2017. Variables corresponding to the concepts are as follows: Education = years of education, Urbanization = share of population living in urban area, Investment = share of investment to GDP, Institution= rules of law, Complexity=Economic complexity index, Geography=share of land area which are not in tropical region, Gender equality= Share of the year of schooling for female to male, Demography=share of population under 14, Innovation=Log patent per capita, Trade=Export+Import/GDP, Price stability=(-1)*log inflation rate. Sample includes 21 ECA economies. Click here to download data and charts. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 E URO PE AND CE NTRAL AS IA 89 BOX 2.2.1 Labor productivity in Europe and Central Asia: Trends and drivers (continued) growth collapsed, falling to near zero in 2013-15. This skilled sectors such as construction or trade and less than may have reflected falling investment in physical capital, one-sixth of productivity in high-skilled services such as particularly in commodity exporters amid the collapse of finance, which accounts for 9 percent of GDP. commodity prices, as well as stalling structural reforms to improve business environments (EBRD 2018b; Georgiev, Reform momentum Nagy-Mohacsi, and Plekhanov 2017). Two waves of reform spurred pre-crisis productivity Between-sector shifts in resources to productivity growth growth in ECA. In the first wave, in the 1990s, the region also declined in ECA after the crisis: In 2013-15 it was half transitioned toward market from centrally planned its pre-crisis average. The fall may partly have reflected a economies. In the early 2000s, the second wave of reforms shift out of agriculture into lower-productivity sectors post in ECA was associated with the initiation by countries in -crisis (trade services) than pre-crisis (manufacturing), such Central Europe and the Western Balkans of their EU as was the case in Kazakhstan (World Bank 2019j). In accession process. Romania, reallocation towards more efficient firms was First wave. In the wake of the collapse of the Soviet Union more important within the manufacturing sector, as other in the early 1990s, productivity plunged as transition sectors, including services, were less exposed to foreign economies fell into deep recessions caused by the rupture competition (Iootty, Pena, and De Rosa 2019). More of trade and financial links with the Soviet Union, the broadly, spillovers from the Euro Area debt crisis, slowing emigration of skilled labor, and armed conflict in parts of global trade growth, and the oil price plunge dampened the region. Central planning was dismantled and replaced growth in sectors with higher levels of productivity— by more market-based approaches (Falcetti, Lysenko, and including in finance, manufacturing, and mining— Sanfey 2006). ECA economies were opened up to limiting their ability to continue to absorb additional labor international trade and capital markets, prices and interest from other sectors with lower productivity (ILO 2017). rates were liberalized, and state-owned enterprises were Indeed, unemployment grew and labor participation fell in privatized to a degree (Georgiev, Nagy-Mohacsi, and the region, particularly in the Western Balkans and South Plekhanov 2017). These reforms helped boost productivity Caucasus. growth in the mid-1990s, particularly in Central Asia and Widely varying sectoral productivity levels across ECA. the South Caucasus (World Bank 2018f). Although most sectors in ECA have productivity levels above EMDE averages, aggregate regional numbers mask Second wave. In the early 2000s, accession to the EU by significant variations. For example, although productivity the countries of Central Europe accelerated their has improved in most sectors, agricultural productivity has international economic integration, and drove institutional declined in the economies of Central Asia since their improvements, further privatization, and deepening of transition to market economies in the 1990s amid their capital markets (Bruszt and Campos 2016). FDI and disruptions to markets and trade (Gharleghi and Popov private investment surged as reforms were anchored 2018). While there are a few exceptions in Central Asia externally, with many ECA economies rapidly becoming where productivity in the production of specific integrated into global value chains with Western Europe, commodities has improved—mainly grains in Uzbekistan accelerating the adoption of new technologies and and oil seeds in Kazakhstan—reallocating labor and capital practices (Aiyar et al. 2013, EBRD 2014). The growing from less competitive agricultural subsectors to more international integration of financial and banking systems productive sectors continues to have the potential to boost helped deepen capital markets, particularly in Central economy-wide productivity. Europe. It was subsequently accompanied by a credit boom (de Haas and van Lelyveld 2006).3 More broadly in ECA, cross-sectoral productivity differentials continue to imply scope for further overall Post-crisis reform momentum. Post-crisis, ECA has faced productivity gains from resource reallocation. In sectors multiple headwinds, including the legacy of the global such as agriculture, mining and utilities, ECA’s productivity lags about 50-70 percent behind advanced- 3 The rise in foreign currency borrowing by households and firms, economy averages, and in mining and utilities it lags however, left sectors exposed to external vulnerabilities, such as capital behind even EMDE averages. On average in ECA, flow reversals, and deepened the recession following the global financial productivity in agriculture (which accounts for 18 percent crisis as economies faced a credit crunch and a period of deleveraging of GDP) is about one-third of productivity in other low- (Zettelmeyer et al. 2010; de Haas et al. 2015). 90 CHAPTER 2.2 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.2.1 Labor productivity in Europe and Central Asia: Trends and drivers (continued) financial crisis, the collapse of oil prices in 2014-16, capital consists of its natural resources such as oil, metals, heightened geopolitical tensions, and international and agricultural land, and is particularly relevant to ECA sanctions on Russia. Meanwhile, reform momentum has given the presence of large commodity exporters. During slowed, with parts of the region witnessing reform the pre-crisis commodity price boom and the reversals, leaving a need for substantial reform progress, accompanying boom in resource exploration and especially in Central Asia and Eastern Europe—which are development, the increased extraction of natural capital not anchored to an EU accession process—and the lifted productivity growth in ECA (Khan et al. 2016). The Western Balkans.4 Many of the commodity exporters in rate of natural capital extraction declined in some the region also suffer from structural constraints, including economies following the boom and as commodity prices a lack of export diversification, large state presence in fell, dampening TFP growth. firms, unfavorable business environments, and weak international competitiveness (Azerbaijan, Kazakhstan, Urbanization. Urbanization tends to be associated with Russia, Ukraine; EBRD 2017; Funke, Isakova, and productivity gains because it encourages more rapid Ivanyna 2017). dissemination of knowledge and technologies and facilitates the reallocation of resources from lower to Post-crisis slowdown in drivers of productivity. There has higher-productivity sectors: such reallocation can be been a broad-based slowdown in the growth of most key constrained by limited urban development, as has been the drivers of labor productivity in ECA in the post-crisis case in the Kyrgyz Republic (World Bank 2018g). The fact period. Demographic pressures have been intensifying, that ECA’s population density is lower than in other particularly in the past decade, in nearly all ECA EMDE regions with similar GDP per capita levels (such as economies. Growth in working-age populations in the Latin America and Caribbean) indicates scope for further region has long lagged the average for EMDEs as a result productivity gains from urbanization. Yet, relative to the of significant migration to western European countries in rest of the world, economies in ECA—particularly those in the EU and to Russia and sharp declines in fertility rates. Central Europe, the Western Balkans, and Eastern Additionally, more than three quarters of the economies in Europe—have recently experienced declines in urban, as ECA have experienced post-crisis slowdowns in investment well as total, populations amid decades of below- rates, reflecting adverse shifts in investor sentiment amid replacement fertility and net emigration (World Bank conflicts and financial pressures in the region, as well as 2017a). weak external economic growth, including in the Euro Area, and adverse shocks to Russia. Low innovation Policy options rates—which partly stem from weak competitiveness, inadequate control of corruption, and a high presence of Across nearly all ECA economies, productivity growth has state-owned enterprises—also continue to dampen the slowed since the financial crisis. To reinvigorate business environment and hinder investment in the region, productivity growth, a four-pronged, comprehensive particularly in the absence of progress with other reforms policy approach is needed to improve the provision and (EBRD 2018a; EBRD 2019b). quality of factors of production, boost firm productivity, promote productivity-enhancing sectoral reallocation, and Other factors affecting productivity in the region establish a more growth-friendly business environment. Some of these policies offer the prospect of relatively short- Natural resource extraction. Standard productivity growth term productivity gains, such as changes in state-owned decompositions fold the extraction of natural capital into enterprise ownership and improvements in the investment total factor productivity growth and, to a lesser extent, climate, while others are more likely to lay the foundation physical capital growth (Brandt, Schreyer, and Zipperer for longer-term productivity gains, such as efforts to 2017; Calderón and Cantu 2019). An economy’s natural improve human capital or adjust migration policies. Within these broad categories, specific policy priorities need to be tailored to country-specific circumstances, 4 A reversal of structural reforms remains a key risk in these regions, especially given the region’s wide heterogeneity. and the pace of growth will depend partly on the successful implementation of structural reforms to enhance the business Improving factors of production environment, achieve debt sustainability, and restructure state-owned enterprises to improve competition. Please refer to EBRD (2013); Lehne, Mo, and Plekhanov (2014); Georgiev, Nagy-Mohacsi, and Plekhanov In ECA, roughly two-thirds of the post-crisis slowdown in (2017); Rovo (2019); and World Bank (2019g) for further detail. productivity growth has reflected slower physical capital G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 E URO PE AND CE NTRAL AS IA 91 BOX 2.2.1 Labor productivity in Europe and Central Asia: Trends and drivers (continued) FIGURE 2.2.1.4 Drivers of productivity growth in ECA Investment growth across the region has fallen in the post-crisis period, reflecting external headwinds—including a commodity price plunge—and idiosyncratic factors—including conflict in pockets of the region and financial pressures in large economies. The workforce is continuing to age, and the working-age population share is declining. Learning gaps are sizable in parts of the region, the control of corruption indicator and business climates remain weak, and the role of the state has remained large. A. Actual and Consensus forecasts for B. Learning gaps, 2017 C. Share of regional GDP accounted for investment growth by economies with growing working-age populations D. Doing Business indicators E. Assessment of transition to a F. Control of corruption, 2017 competitive market economy, 2019 Source: Consensus Economics; European Bank for Reconstruction and Development; Kraay (2018); United Nations; World Bank. Note: Unless otherwise specified, productivity refers to labor productivity, defined as output per worker. A. Blue bars denote actual investment growth, where investment is measured as gross fixed capital formation. Actual growth aggregate calculated using GDP weights at 2010 prices and market exchange rates. Consensus forecasts aggregate calculated as a simple average of surveys for periods indicated based on data availability. Unbalanced sample includes 8 ECA economies, due to data availability. B. CE = Central Europe, CA = Central Asia, EE = Eastern Europe, SC = South Caucasus, and WBK = Western Balkans. The learning gap is the difference between expected years of schooling and learning-adjusted years of schooling, as in Kraay (2018). The sample includes 21 ECA EMDEs. C. The working-age population is defined as people aged 15-64. Unbalanced sample including 23 ECA economies. D. AEs=advanced economies. Figure shows the median ECA value for 2010 and 2019, and the median advanced economy for 2019. The full names of the Doing Business reform areas given on the x-axis are: Making it easier to start a business, making it easier to deal with construction permits, making it easier to get electricity, making it easier to register property, making it easier to get credit, making it easier to protect minority investors, making it easier to pay taxes, making it easier to trade across borders, making it easier to enforce contracts, and making it easier to resolve insolvency. Sample includes 33 advanced economies and 22 ECA economies. E. Figure shows the distance to the frontier for achieving a full transition to a competitive market economy, as measured by EBRD (2019b). Economies with higher index levels are closer to the frontier, where scores range from 1 to 10, with 10 denoting the synthetic frontier. The sample includes 24 ECA economies. F. CE = Central Europe, CA = Central Asia, EE = Eastern Europe, SC = South Caucasus, and WBK = Western Balkans. The indicator reflects perceptions of the extent to which public power is exercised for private gain, including both petty and grand forms of corruption, as well as “capture” of the state by elites and private interests, as measured by the World Governance Indicators. Sample includes 23 ECA economies and 150 EMDEs. Click here to download data and charts. accumulation, with investment weakness particularly accumulation has contributed less to productivity growth notable in Central Europe, Russia, and, more recently, in ECA than in other EMDEs, suggesting there is also a Turkey, which together account for over 85 percent of need to improve this source of productivity growth. ECA’s GDP. In some parts of the region, infrastructure investment gaps are sizable. There is thus need for a Addressing investment and infrastructure gaps. renewed push to close infrastructure gaps as well as to Investment growth has fallen sharply in ECA in the post- boost private investment. Meanwhile, human capital crisis period, reflecting a commodity price plunge as well 92 CHAPTER 2.2 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.2.1 Labor productivity in Europe and Central Asia: Trends and drivers (continued) as weakening external economic growth and investor wider than the global average in most Western Balkan sentiment, amid conflict, international sanctions, and economies—particularly in Kosovo and North financial pressures (Figure 2.2.1.4). Across the region, Macedonia—as well as in a few economies in Eastern reforms to boost private sector development, transition to Europe (Moldova), Central Asia (the Kyrgyz Republic), competitive and inclusive markets, and regional and the South Caucasus (Georgia). While most economies integration are needed to attract private investment and in Central Europe have smaller gaps than ECA as a whole, capital flows, particularly to economies outside the EU Romania is an exception. Turkey also has a larger gap than that lack access to financing sources, such as EU structural ECA, in addition to low education attainment in the work funds (EBRD 2018a; World Bank 2019g). force, large gender gaps in education, and an inadequacy of skills, which is often cited as a constraint for doing In certain subregions, particularly Central Asia, removing business and a bottleneck to innovation in the country key bottlenecks to private sector development, such as (World Bank 2019d). Some economies in ECA with large inadequate infrastructure, is especially important to learning gaps, including Georgia, have taken measures to support productivity growth. In some pockets of the reform the education sector and its funding (Kraay 2018). region, improved connectivity could accelerate the absorption of technology and speed convergence with Although ECA has the lowest rate of extreme poverty of all advanced economies (Gould 2018). Infrastructure needs EMDE regions, the share of school-age children not remain large in ECA, particularly in transport and enrolled in school is higher than in both East Asia and electricity. Unreliable electricity supply hinders activity in Pacific and Latin America and Caribbean (World Bank parts of the region: while the percentage of firms 2018h). Economies in the Western Balkans (Albania, experiencing electrical outages is lower in ECA than in any Montenegro, Serbia), Eastern Europe (Moldova), Central other EMDE region, related losses for affected firms in Asia (the Kyrgyz Republic, Uzbekistan), and Turkey have Central Asia can exceed 9 percent of annual sales (Blimpo elevated out-of-school rates relative to the ECA average for and Cosgrove-Davies 2019; IMF 2019a). In surveyed secondary education (UNICEF 2019). The diversity of manufacturing firms in Uzbekistan, for instance, smaller situations in ECA—with human capital investment quite firms report more interruptions of electricity, gas, and high in some economies but lagging in others (such as in water supply, than larger firms, as well as a lack of territory Central Asia)—indicates a need for policies to be tailored or high lease rates on land as impediments to expanding to countries’ specific needs (Kraay 2018). Education policy output production (Trushin, E. 2018). Appropriate land and training programs can also be redesigned to adapt the use planning and urbanization policies can substantially skills of aging populations to changing needs and new reduce the cost of meeting transport needs while technologies (Hallward-Driemeier and Nayyar 2018; minimizing carbon footprints (ITF 2018; Rozenberg and World Bank 2018a). Fay 2019). Counteracting unfavorable demographic trends. In ECA, Raising human capital. In a few economies in ECA, the workforce is continuing to age, and the working-age particularly in Central Asia, inadequate investment in population share is declining, with many young and skilled human capital has left parts of the workforce poorly workers having emigrated. These developments are likely equipped with the skills required for the future, and to weaken productivity growth and highlight the need for unprepared for rapid technological change (Flabbi and education to help workers adapt to new job requirements Gatti 2018). Boosting human capital investment— and technologies (Aiyar, Ebeke, and Shao 2016). including through education and health—could help Generating stronger productivity growth will require remove bottlenecks to productivity growth. How measures to mitigate the decline in skilled workforces. education systems adapt to evolving skill needs will be a Implementing more flexible immigration policies could key determinant of the productivity and distributional help relieve skilled labor shortages by attracting skilled effects of technological change (Barro and Lee 2015). foreign workers in an orderly way (Delogu, Docquier, and Machado 2014; World Bank 2019g). In several economies in ECA, educational attainment and the acquisition of needed skills have been lower than Boosting firm productivity expected given the level of school enrollment and the average years of schooling (Altinok, Angrist, and Patrinos Within-sector productivity gains stalled in ECA in 2013- 2018). The learning gap (the difference between years 15, consistent with slowing reallocation of resources spent in schools and educational assessment outcomes) is between firms and slowing productivity growth within G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 E URO PE AND CE NTRAL AS IA 93 BOX 2.2.1 Labor productivity in Europe and Central Asia: Trends and drivers (continued) firms. This highlights the need to boost firm productivity more widespread adoption of these technologies could also in the region, including by completing the transition to help expand tax bases through the fiscalization of informal competitive and inclusive markets, which could strengthen sector transactions (World Bank 2019a). Increasing SMEs’ the environment for private investment and innovation access to finance could help these firms increase their (World Bank 2019a). Policy options include measures to average size and reduce their reliance on retained earnings level the playing field for private and state-owned firms to fund investment, which in turn would support job and expanding access to finance to a wider range of firms. creation (Ayyagari, Demirgüç-Kunt, and Maksimovic 2017; Ayyagari et al. 2016). Leveling the playing field. State-owned enterprises tend to be less efficient than those in the private sector (World Encouraging sectoral reallocation Bank 1995). In Eastern Europe and Central Asia, and to some extent Russia, the state’s presence in the economy Between-sector productivity gains from sectoral remains large, with state ownership accounting for more reallocation have slowed in ECA since the global financial than 10 percent of firms surveyed in some cases, and with crisis. Also, some ECA economies remain undiversified: ECA ranking second overall among EMDE regions, after renewed efforts to diversify commodity-based economies Sub-Saharan Africa (World Bank 2019k). In Ukraine, could generate new opportunities for labor to move toward firms with at least partial state presence account for more productive employment. roughly 20 percent of total turnover by firms and over 25 percent of firms’ assets (Balabushko et al. 2018). State- Diversifying economies. Energy-exporting economies, owned enterprises also have a large presence in Moldova, including those in ECA, are characterized by generally low accounting for one-third of GDP (World Bank 2019l). levels of economic diversification, in terms of both exports Restructuring or privatizing state-owned enterprises and fiscal revenue (Grigoli, Herman, and Swiston 2017).5 therefore still presents an opportunity to raise economy- Energy sector production tends to be capital-intensive, wide productivity in several countries across the region, if with relatively high labor productivity (Aslam et al. 2016; it is accompanied by effective regulation and Danforth, Medas, and Salins 2016; Stocker et al. 2018). improvements in management, corporate governance, and Productivity growth, however, has been more tepid in the business environment (Brown, Earle, and Telegdy ECA’s energy-exporting countries than in the region 2006; EBRD 2019b; Funke, Isakova, and Ivanyna 2017). overall, with post-crisis (2013-18) growth at 1.2 percent Additionally, there are a number of economies, including versus 1.6 percent, reflecting weaker TFP growth. in Eastern Europe, where price controls remain in place for Diversification therefore presents an opportunity to boost particular goods, tending to constrain competition and TFP and productivity growth, as well as macroeconomic lower productivity. stability (Brenton, Newfarmer, and Walkenhorst 2009; Papageorgiou and Spatafora 2012). Diversification of Financial market development and financial inclusion. resource-based economies can be promoted by reforms Small and medium-sized enterprises (SMEs) have the that increase capital and skill accumulation, innovation, largest potential for productivity catch-up with advanced and reduce transaction costs.6 economies. Their growth continues to be hindered by many factors, including insufficient access to finance and Enhancing a growth-friendly environment regulatory barriers (Ayyagari, Demirgüç-Kunt, and Maksimovic 2017; Cusolito, Safadi, and Taglioni 2017; Several ECA economies have severe institutional Wang 2016). The largest gaps in financial inclusion for weaknesses that continue to erode incentives for SMEs in ECA are in Central Asia and the South Caucasus innovation and investment. Addressing these weaknesses (excluding Georgia), where access to financial services is requires reforms to improve governance and business nearly as limited as in the Middle East and North Africa, climates. South Asia, and Sub-Saharan Africa (IMF 2019b). Policies that promote more widespread adoption of digital 5 On the budget front, Russia has made strides in anchoring fiscal technologies, including in the delivery of financial and policy by implementing a fiscal rule that targets a primary balance of zero public sector services, could bolster financial inclusion and at the benchmark oil price of $40 per barrel. Any excess fiscal reserves that are generated from higher oil prices are saved in the National boost productivity by helping spread innovation and Welfare Fund. improving private sector and government efficiency 6 Please refer to Beck (2018); Gylfason (2018); Lederman and Maloney (Baldwin 2019). In economies with large informal sectors, (2007); Hesse (2008); and IMF (2016a) for further detail. 94 CHAPTER 2.2 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.2.1 Labor productivity in Europe and Central Asia: Trends and drivers (continued) Growth-friendly governance. Over the long term, bottlenecks in the region, hindering the ability to attract institutional quality is one of the most important foreign direct investment and private investment in some determinants of productivity growth (Chapter 3). In ECA, economies (Kazakhstan, Russia, Ukraine; World Bank productivity catch-up to advanced economies was 2016b; Shepotylo and Vakhitov 2015). Over the past particularly pronounced in Central Europe during the pre- decade, several ECA economies made significant strides in crisis period, reflecting the anchoring of structural and improving their business environments. As a result, in institutional reforms to the EU accession process several countries in Central Europe, the Western Balkans, (Rodríguez-Pose and Ketterer 2019). Overall, however, the and the South Caucasus, business environment indexes region continues to face governance challenges with nearly have recently approached the levels in advanced EU 75 percent of ECA EMDEs falling below the global economies (World Bank 2018a). average for the control of corruption, including almost all of the economies of Central Europe, Eastern Europe, and Notwithstanding these improvements, business climates in the South Caucasus (Kaufmann, Kraay, and Mastruzzi Eastern Europe and Central Asia lag the ECA average, 2010). Because progress in confronting perceived and with the latter trailing the EMDE average in access to actual corruption has been slow, continuing efforts have electricity and the ease of trading across borders (World reinforced the perception that the control of corruption is Bank 2019j). For example, in Ukraine, the largest higher than in other EMDEs (Transparency International economy in Eastern Europe, the average worker takes one 2019). year to produce the same amount that the average worker in Germany produces in 17 days (World Bank 2019m). At Structural reforms to improve governance can lead to current growth trends, Ukraine is unlikely to converge to sizable productivity gains, particularly in countries that are Poland’s per capita income, despite having had similar farthest from best practices (Chapter 3; Acemoglu, income levels in 1990; this partly reflects Ukraine’s Johnson, and Robinson 2005; Cusolito and Maloney relatively low ratio of capital stock to GDP. Removing 2018). Major governance and business reforms in EMDEs market distortions and improving resource allocation have in the past been associated with higher growth rates could triple manufacturing productivity and help improve in output, total factor productivity, and investment prospects in Ukraine (Ryzhenkov 2016). The Western (Hodge et al. 2011; Divanbeigi and Ramalho 2015; World Balkans also struggle to attract FDI notwithstanding Bank 2018a). The detrimental effects of corruption on reforms to improve business climates and further firm productivity can be exacerbated by excess or complex integration into regional and global markets (Jirasavetakul regulation (Amin and Ulku 2019). Anticorruption and Rahman 2018; World Bank 2019h; World Bank campaigns, as well as reductions in the number of 2019n). Although Turkey has high productivity levels, it regulations and tax complexity, have helped some lags well behind the ECA average for resolving insolvency, economies tackle corruption (IMF 2019c). which could dampen overall productivity as less productive firms are more likely to remain in the market (World Bank Growth-friendly business climates. Lack of exposure to 2019d). To address this, Turkey has recently introduced a international competition—including from non-tariff more streamlined procedure that focuses on business barriers and complex trade rules—as well as restrictive continuation instead of liquidation. product market and services regulation, remain structural Growth in Latin American and the Caribbean slowed markedly in 2019, to an estimated 0.8 percent, held back by idiosyncratic factors in large economies, headwinds from slowing global trade, and social unrest in several countries. As activity in Brazil gathers pace amid improving investment conditions, policy uncertainty in Mexico fades, and the recession in Argentina eases after bouts of severe market stress, regional growth is projected to rise to 1.8 percent in 2020 and about 2.4 percent in 2021. This recovery will not be sufficient to reverse the growing per capita income gap with advanced economies in some LAC economies. Moreover, the regional outlook is subject to significant downside risks, including from market volatility and adverse market responses to weak fiscal conditions; deeper-than-expected spillovers from slowdowns in Argentina, China, and the United States; heightened social unrest; and disruptions from natural disasters and severe weather. Brazil, Chile, Mexico, Peru), though conditions in Recent developments most countries improved later in the year. Within Growth in Latin America and the Caribbean the industrial sector, mining activity contracted (LAC) decelerated markedly in 2019, to an sharply following an iron ore mining dam disaster estimated 0.8 percent. The slowdown was broad- in Brazil, continued oil production declines in based across economies and sectors. All three of Mexico, and temporary mining disruptions in the largest economies in the region—Brazil, Chile, while policy uncertainty contributed to a Mexico, and Argentina—grew significantly less sharp contraction in mining and construction than projected in June. Brazil experienced a larger- activity (Figure 2.3.1.B). than-expected impact from a major mining Sluggish investment and private consumption held accident, as well as slowing exports to China, and back regional growth in 2019. Investment relatively sluggish improvements in labor market contracted as policy uncertainty lingered, investor conditions. Growth in Mexico was hindered by an sentiment worsened, and governments retrenched. uncertain investment climate, tight monetary In Mexico, the government cancelled public policy, and public spending cuts, while infrastructure projects. Several other countries cut Argentina’s economy was held back by the effects public spending (Argentina, Ecuador, Haiti, of renewed financial stress. In Colombia, however, Panama, Paraguay). growth accelerated as private consumption and investment picked in the context of Regional export growth has slowed along with accommodative monetary policy and fiscal global trade activity. Yet export trends among the incentives to support investment. large regional economies are not uniform. Bilateral The regional growth slowdown was generally more tariff hikes between China and the United States acute in industrial sectors than in services (Figure gave an initial boost to Brazil’s soybean exports, 2.3.1.A). Industrial activity was stagnant or which has since faded as demand slowed and the contracting in five of the region’s six largest soybean price differential between Brazil and the economies in the first half of 2019 (Argentina, rest of the world narrowed. Exports from Brazil, Chile, and Peru to China—the largest export destination of all three countries—plateaued or Note: This section was prepared by Dana Vorisek. Research slowed in the second half of 2019 (Figure assistance was provided by Vanessa Arellano Banoni. The regional aggregate statistics presented in this section do not include Venezuela. 2.3.1.C). However, exports from Mexico, 80 96 CHAPTER 2.3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 2.3.1 LAC: Recent developments percent of which go to the United States, Services sector growth has slowed in LAC, mirroring a much stronger continued to grow. There is evidence that Mexico slowdown in the industrial sector. Within the industrial sector, mining has benefitted from trade diversion to the United production contracted sharply in 2019. Economies highly reliant on China States as a result of the U.S.-China trade conflict as a trade destination have seen their exports plateau or fall after tariff hikes between China and the United States, while Mexico’s exports to the (UNCTAD 2019; World Bank 2019o). In the United States have continued to grow. A sharp recession in Argentina has first half of 2019, the electrical machinery, impacted Bolivia, Brazil, and Paraguay through trade and remittance channels. With output gaps becoming more negative, and inflation at the transport equipment, and agriculture and food low end of target ranges, monetary policy is easing in numerous countries. sectors benefited the most, in value terms. Late 2019 was marked by the emergence of social A. Services and industrial sector B. Industrial production growth, growth by subsector tensions in Bolivia, Chile, and Ecuador, related to economic policy decisions and elections. The events contributed to a downgrade of estimated growth in all of these countries. The recent developments follow similar events in Haiti and Nicaragua, which contributed to deteriorating economic conditions in both countries. Colombia, as well, experienced protests in late 2019. Economic and social conditions in Venezuela C. Exports to China and the United D. Exports to and remittances from continue to be dire. The population is States Argentina experiencing frequent electricity outages and water shortages; widespread scarcity of basic goods; and a sharp rise in preventable diseases, malnutrition, and mortality rates. More than 4.7 million people have left. Several countries have imposed entry restrictions on Venezuelans as the provision of services to migrants becomes more fiscally burdensome and social tensions rise. However, these restrictions are not expected to halt outward E. Regional output gap F. Inflation migration from Venezuela, and migration may have growing policy implications elsewhere. In Colombia, for instance, the fiscal council has allowed additional spending related to migrants. In the medium to long term, host countries could benefit. The Colombian government estimates that the net impact of migration on growth, after accounting for the fiscal cost, will be 0.1-0.5 percentage point between 2018 and 2021 relative Source: Central Bank of Bolivia; Central Bank of Costa Rica; Central Bank of the Dominican Repub- to a no-migration scenario, primarily via the lic; Central Bank of Guatemala; Central Bank of Paraguay; Central Bank of Uruguay; Haver Analyt- ics; International Monetary Fund (World Economic Outlook); World Bank. consumption channel (Colombia Departamento A. Lines show GDP-weighted averages of Argentina, Brazil, Chile, Colombia, Mexico, and Peru (90 Nacional de Planeación 2018). percent of regional GDP). Last observation is 2019Q3. B. Lines show industrial production-weighted averages of Brazil, Chile, Colombia, Mexico, and Peru. Last observation is September 2019. In Argentina, following a sharp currency C. Index based on exports (in value) from Brazil, Chile and Peru to China, and from Mexico to the United States. Gray area begins when China and the United States began to increase bilateral tariffs, depreciation in the wake of the primary election in July 2018. Last observation is October 2019. D. Bars show average year-on-year growth of monthly flows during the indicated period. Bars for results in August, the government implemented 2019 are constructed using monthly data through November (exports) and September (remittances). capital controls and imposed a maturity extension E. Includes data for Argentina, Belize, Bolivia, Brazil, Chile, Colombia, Costa Rica, the Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama, Paraguay, on part of its short-term debt. The bouts of Peru, and Uruguay. Shaded areas around center line indicate confidence intervals. Last observation is 2019Q2. financial stress in Argentina since early 2018 have F. Blue boxes show central inflation targets; vertical lines show target bands. affected neighboring countries through trade Click here to download data and charts. (lower exports to Argentina from Brazil and G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 LATIN AME RIC A AN D THE C ARIBBE AN 97 Paraguay, in particular), remittances (sharply Colombia, is still envisioned to boost the outlook lower from Argentina to Bolivia and Paraguay), for the region (Figure 2.3.2.A). Regional growth is and tourism (downturn in spending and arrivals forecast to increase to 1.8 percent in 2020, and to by Argentines in Uruguay; Figure 2.3.1.D). 2.4 percent in 2021 (Tables 2.3.1 and 2.3.2). Under this projection, growth is not expected to With a small number of exceptions (Argentina, exceed the average during the past three decades. Ecuador, Venezuela), bond yields in the region Moreover, the projected performance will not be have been broadly stable in recent months. Some sufficient to reverse the widening per capita currencies have depreciated against the U.S. dollar income relative to advanced economies since 2014 (Argentina, Brazil, Chile, Colombia, Paraguay, in some countries in the region (Figure 2.3.2.B). Uruguay). Capital inflows to the region, which come predominantly from the United States and Metals and agriculture prices are projected to be the Euro Area, have slowed. flat due to weak global demand, providing little incremental support for exporters of these The output gap in nearly all economies has commodities. Likewise, lower global demand for become steadily more negative, after the regional oil, together with expanding production in the output gap nearly closed in late 2018 (Figure United States, will put downward pressure on oil 2.3.1.E). With growth and inflation expectations prices in the short term. broadly moderating and inflation at the low end of target ranges among most inflation-targeting Fiscal space is limited or absent in most of the countries, a growing number of central banks are region, leaving little capacity to pursue easing monetary policy (Figure 2.3.1.F). Policy expansionary spending to support growth. One interest rates in Brazil, Chile, Costa Rica, the exception is Chile, which is planning a fiscal Dominican Republic, Jamaica, Mexico, Paraguay, stimulus that will boost public investment and and Peru were lowered in the second half of 2019, support small and medium enterprises. in most cases multiple times. Policymakers in some other large economies (Brazil, Mexico) remain committed to spending Several major policy developments have occurred restraint over the forecast horizon despite sluggish in the region. Following revisions to the previously growth, in order to improve medium-term fiscal negotiated United-States-Mexico-Canada Agree- sustainability and retain investor confidence. In ment, among others on labor standard Argentina, continued fiscal consolidation will be a enforcement mechanisms, the agreement moved necessary component of the budget strategy. closer to ratification by the United States with the passage by the House of Representatives in In the baseline outlook, growth in the largest December. Mexico has ratified the agreement. In economies will pick up and domestic demand—in Brazil, a long-awaited pension reform was passed particular, investment—will strengthen. This by Congress in October. Policymakers have begun outlook is also contingent on an acceleration of to work on tax reform, the next key item on the exports, after weakness in 2019 (Figure 2.3.2.C). reform agenda. Mercosur (Argentina, Brazil, Paraguay, and Uruguay) reached a trade In Brazil, a boost to investor confidence following agreement with the European Union in June that progress on major reforms, a moderate easing of was two decades in the making. The agreement lending conditions, and a gradual improvement in will now need to be ratified and implemented at labor market conditions are slated to support a the country level. pickup in investment and private consumption, helping push growth to 2 percent in 2020 and 2.5 percent in 2021. Investment in Mexico is also Outlook expected to pick up as investor sentiment Although growth projections have been revised improves and the private sector is more involved down since June, an expected easing of domestic in infrastructure projects, while easing monetary constraints in the three largest economies, together policy will provide modest support to private with a continued growth acceleration in consumption. Growth is forecast to rise to a still 98 CHAPTER 2.3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 2.3.2 LAC: Outlook and risks subdued 1.2 percent in 2020 and 1.8 percent in Growth in LAC is projected to firm during the forecast horizon, supported 2021. predominantly by an easing of domestic constraints in the three largest economies in the region. The recovery will be subdued, however, and will In the near-term projection, investment and not be sufficient to offset a growing per capita income gap with advanced consumption in Argentina will continue economies in some LAC economies. Faster regional growth is contingent on an upturn in investment, which has been repeatedly downgraded contracting, though at a slower pace, while import during the past year, and on a pickup in weak export growth. Weak fiscal compression will recede. Ultimately, the economy positions are a risk for financial stability and growth in the region, while high levels of inequality could spark further social unrest and result in is expected to experience three years of economic disruptions. contraction, and growth to revert to a positive rate only in 2021. A. Growth B. Per capita income in LAC relative to advanced economies In Colombia, investment is expected to accelerate as planned infrastructure projects are carried out. Favorable financing conditions are envisioned to support domestic demand more broadly. These factors will support a rise in growth to 3.6 percent in 2020 and about 3.9 percent in 2021-22. Growth in Chile is projected to recover after interruptions from social unrest in late 2019, to 2.5 percent in 2020 and 3.0 percent in 2021. This C. New export orders D. Investment growth forecast assumes a higher volume of copper exports after mine disruptions in 2019, improved private sector sentiment as business sector reforms are rolled out, and a boost from fiscal stimulus. Aggregate growth in Central America is projected to firm over the forecast horizon. Easing credit conditions (especially in Costa Rica) and an unwinding of temporary setbacks to construction and infrastructure projects (Panama) will help E. Fiscal balance and government F. Income inequality debt boost output. The subregion is also expected to benefit from trade and business environment reforms in recent years, including an expanded customs union between Guatemala and Honduras. Growth in the Caribbean is forecast to accelerate in the near term, predominantly due to major offshore oil production developments in Guyana, while growth in the largest Caribbean economy, Source: Haver Analytics; International Monetary Fund (World Economic Outlook); World Bank. (PovcalNet). the Dominican Republic, is projected to be stable B. Per capita GDP is calculated as the sum of GDP in the countries in the indicated groups divided by the sum of the population in the same country groups. at about 5 percent as the tourism sector stabilizes C. Last observation is November 2019. following disruptions in 2019 linked to health D. Lines show GDP-weighted averages of all LAC economies (20 in total) for which expenditure components of GDP are available. concerns. E. Sample includes 32 countries. F. Bottom 40 percent refers to the bottom four deciles of consumption or income share, middle 50 percent to the middle five deciles, and top 10 percent to the top decile. Data are for latest available year from 2010 to 2017. EMDEs = emerging market and developing economies. AEs = advanced Risks economies. LAC, EMDE, and AE bars show simple averages of 19, 111, and 31 countries, respectively. Click here to download data and charts. LAC continues to face predominantly downside risks to growth. External risks, particularly those linked to trade and finance, are elevated. A further G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 LATIN AME RIC A AN D THE C ARIBBE AN 99 growth slowdown in China, should the bilateral more politically challenging to implement U.S.-China trade dispute reescalate, could expose structural reforms, which are key to boosting LAC to additional negative spillovers through longstanding low productivity (Box 2.3.1). trade, commodity price, and confidence channels. This risk is particularly acute for countries highly Social tensions in several countries in late 2019 reliant on China as an export destination (Brazil, could become more widespread, with negative Chile, Peru, and Uruguay). Likewise, sluggish economic repercussions. Discontent about lack of U.S. growth could be a greater-than-expected opportunities is a significant underlying risk for hindrance for Mexico and other countries reliant social stability, and ultimately for growth, in the on the United States. Continued weak export region. Although inequality in many LAC growth would contribute to a rise in already-large countries has fallen in recent years, due in large current account deficits in some countries part to gains at the low end of the income (Bolivia, Colombia, El Salvador, Honduras, and distribution, it remains high relative to other Panama, among others). The financing of large regions (World Bank 2016e, 2020; Messina and external imbalances could become more Silva 2019). The share of income going to the challenging should countries experience severe bottom 40 percent of households in LAC currency pressures or an unexpected rise in economies is lower, on average, than in all borrowing costs. EMDEs and in advanced economies, while the share going to the top 10 percent is higher, at 35 Adverse market responses to domestic market percent, versus 25 percent in advanced economies conditions within the region, including weak fiscal and 31 percent in EMDEs (Figure 2.3.2.F). profiles, could dent capital inflows and investment. Investment growth is projected to Disruptions related to natural disasters, including firm in the baseline outlook but has been the heightened frequency, duration, and force of repeatedly buffeted by unanticipated climate events, are a persistent and growing developments during the past year and could downside risk for a host of LAC economies. The continue to be hindered by policy uncertainty human and economic toll of Hurricane Dorian in (Figures 2.3.2.D). The Bahamas in September 2019 is the latest example of the Caribbean’s vulnerability to Adverse intraregional spillovers from the market hurricanes, and illustrates the devastating volatility and sharp recession in Argentina could consequences of natural disasters in the region. In take a further toll if restoring the country’s Brazil, the large-scale fires in the Amazon economic stability takes longer than expected. rainforest last year have had widespread Negative repercussions are especially a risk for environmental consequences. They have also Bolivia, Paraguay, and Uruguay. Moreover, for presented a policy risk, in that the authorities’ Argentina, another bout of severe financial market sluggish response has complicated the political stress could further inhibit debt sustainability and task of completing the EU-Mercosur trade set back an already protracted economic recovery. agreement. Should easing monetary policy be insufficient to e EU-Mercosur agreement has the potential to counter weak growth, commitments to public signi cantly boost the depth of global trade spending prudence in the region could come integration in LAC if it passes in its current form, under pressure. Though a return to an especially as the region has long been less open to expansionary fiscal stance would provide a trade than most other EMDE regions (World temporary boost to growth, it could have negative Bank 2019p). Completion of the agreement in its consequences for financial stability and fiscal current form is an upside risk for the outlook. sustainability. Fiscal positions are already on track Deeper trade linkages and participation in global to deteriorate somewhat during the forecast value chains have the potential to stimulate horizon as a result of sustained budget deficits productivity through increased investment and (Figure 2.3.2.E). Weak growth could make it deeper participation in global value chains. 100 CHAPTER 2.3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 TABLE 2.3.1 Latin America and the Caribbean forecast summary (Real GDP growth at market prices in percent, unless indicated otherwise) Percentage point differences from June 2019 projections 2017 2018 2019e 2020f 2021f 2022f 2019e 2020f 2021f EMDE LAC, GDP1 1.9 1.7 0.8 1.8 2.4 2.6 -0.9 -0.8 -0.3 (Average including countries with full national accounts and balance of payments data only)2 EMDE LAC, GDP2 1.9 1.7 0.8 1.8 2.4 2.6 -0.9 -0.7 -0.3 GDP per capita (U.S. dollars) 0.7 0.6 -0.3 0.8 1.5 1.8 -1.1 -0.8 -0.2 PPP GDP 2.0 1.7 0.8 1.8 2.5 2.7 -1.0 -0.8 -0.2 Private consumption 2.7 2.1 1.1 2.1 2.7 2.9 -0.7 -1.1 -0.2 Public consumption 0.7 0.9 -0.3 0.9 1.0 1.2 -0.1 -0.3 -0.2 Fixed investment -0.2 2.1 -0.3 2.6 4.0 3.6 -1.6 -0.5 -0.4 Exports, GNFS3 3.8 4.2 1.1 2.8 3.1 3.5 -3.0 -0.9 -0.7 Imports, GNFS 3 6.3 5.5 0.4 3.3 3.8 4.0 -2.6 -1.4 -0.8 Net exports, contribution to growth -0.5 -0.3 0.1 -0.1 -0.2 -0.1 0.1 0.1 0.0 Memo items: GDP South America4 1.6 1.4 0.9 1.9 2.6 2.7 -0.7 -0.7 -0.1 Central America5 3.8 2.7 2.5 3.0 3.3 3.4 -0.6 -0.4 -0.3 Caribbean6 3.3 5.0 3.8 5.6 3.9 4.3 0.0 1.1 -0.5 Brazil 1.3 1.3 1.1 2.0 2.5 2.4 -0.4 -0.5 0.2 Mexico 2.1 2.1 0.0 1.2 1.8 2.3 -1.7 -0.8 -0.6 Argentina 2.7 -2.5 -3.1 -1.3 1.4 2.3 -1.9 -3.5 -1.8 Source: World Bank. Note: e = estimate; f = forecast. EMDE = emerging market and developing economies. World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of countries’ prospects do not differ at any given moment in time. The World Bank has ceased producing a growth forecast for Venezuela and has removed Venezuela from all growth aggregates in which it was previously included. 1. GDP and expenditure components are measured in 2010 prices and market exchange rates. 2. Aggregate includes all countries in Table 2.3.2 except Dominica, Grenada, Guyana, Haiti, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines, and Suriname. 3. Exports and imports of goods and non-factor services (GNFS). 4. Includes Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador, Paraguay, Peru, and Uruguay. 5. Includes Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, and Panama. 6. Includes Antigua and Barbuda, The Bahamas, Barbados, Belize, Dominica, Dominican Republic, Grenada, Guyana, Haiti, Jamaica, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines, and Suriname. Click here to download data. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 LATIN AME RIC A AN D THE C ARIBBE AN 101 TABLE 2.3.2 Latin America and the Caribbean country forecasts1 Percentage point differences (Real GDP growth at market prices in percent, unless indicated otherwise) from June 2019 projections 2017 2018 2019e 2020f 2021f 2022f 2019e 2020f 2021f Argentina 2.7 -2.5 -3.1 -1.3 1.4 2.3 -1.9 -3.5 -1.8 Belize 1.9 2.1 2.7 2.1 1.8 1.8 0.4 0.0 -0.1 Bolivia 4.2 4.2 2.2 3.0 3.2 3.4 -1.8 -0.6 -0.2 Brazil 1.3 1.3 1.1 2.0 2.5 2.4 -0.4 -0.5 0.2 Chile 1.3 4.0 1.3 2.5 3.0 3.0 -2.2 -0.6 0.0 Colombia 1.4 2.6 3.3 3.6 3.9 3.9 -0.2 -0.1 0.2 Costa Rica 3.4 2.6 2.0 2.5 3.0 3.2 -1.0 -0.6 -0.4 Dominican Republic 4.7 7.0 5.3 5.0 5.0 5.0 0.1 0.0 0.0 Ecuador 2.4 1.4 -0.3 0.2 0.8 1.2 -0.3 -0.2 0.0 El Salvador 2.3 2.5 2.4 2.5 2.5 2.5 -0.2 0.0 0.1 Grenada 4.4 4.2 3.5 2.9 2.9 3.2 -0.4 -0.8 -0.8 Guatemala 2.8 3.1 3.4 3.0 3.2 3.2 0.1 0.3 0.2 Guyana 2.1 4.1 4.5 86.7 10.5 14.6 -0.1 53.2 -12.4 Haiti2 1.2 1.5 -0.9 -1.4 -0.5 1.4 -1.3 -3.0 -1.8 Honduras 4.8 3.7 3.3 3.5 3.5 3.5 -0.3 -0.3 -0.4 Jamaica 1.0 1.9 1.0 1.1 1.2 2.0 -0.6 -0.6 -0.7 Mexico 2.1 2.1 0.0 1.2 1.8 2.3 -1.7 -0.8 -0.6 Nicaragua 4.7 -3.8 -5.0 -0.5 0.6 1.0 0.0 -1.6 -0.7 Panama 5.6 3.7 3.5 4.2 4.6 4.8 -1.5 -1.2 -0.6 Paraguay 5.0 3.7 0.7 3.1 3.9 3.8 -2.6 -0.9 -0.1 Peru 2.5 4.0 2.6 3.2 3.5 3.6 -1.2 -0.7 -0.5 St. Lucia 2.6 0.9 1.8 3.2 3.0 2.4 -1.6 -0.3 0.6 St. Vincent and the Grenadines 1.0 2.2 2.3 2.3 2.3 2.3 0.2 0.0 0.0 Suriname 1.8 2.6 2.2 2.5 2.1 2.1 0.2 0.4 0.0 Uruguay 2.6 1.6 0.5 2.5 3.5 3.2 -1.0 0.2 1.0 Source: World Bank. Note: e = estimate; f = forecast. World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of countries’ prospects do not significantly differ at any given moment in time. 1. GDP and expenditure components are measured in 2010 prices and market exchange rates. 2. GDP is based on fiscal year, which runs from October to September of next year. Click here to download data. 102 CHAPTER 2.3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.3.1 Labor productivity in Latin America and the Caribbean: Trends and drivers Labor productivity growth in Latin America and the Caribbean (LAC) slowed to near zero during 2013-18, among the lowest of the six emerging market and developing economy (EMDE) regions. This rate is well below the 1.7 percent average during the pre- crisis period (2003-08) and a return to the average during the preceding four decades. In two-fifths of LAC economies, productivity growth was negative during 2013-18. Sluggish productivity growth during 2013-18 mainly reflects negative total factor productivity (TFP) growth in some large LAC economies, as the commodity price slump and intensifying market distortions allowed unproductive firms to continue operating. Despite anemic productivity growth, the level of productivity remains higher than the EMDE average, albeit still less than one-quarter of the level in advanced economies. Many countries in the region would benefit from reforms to improve competition and innovation, deepen trade linkages, improve the quality of education, reduce labor market inefficiencies, strengthen institutional quality, and increase the volume and efficiency of infrastructure investment. Introduction • What policy options are available to boost productivity growth? For decades, productivity growth in Latin America and the Caribbean (LAC) has been anemic (Fernández-Arias and This Box defines productivity as labor productivity, Rodríguez-Apolinar 2016). After a brief pre-crisis burst, represented by real GDP per person employed (at 2010 productivity growth fizzled out again after the global prices and exchange rates). This definition deviates from financial crisis. Relative to a pre-crisis (2003-08) average of some previous work on productivity in the region, which 1.7 percent, productivity growth in the region dropped to focused on total factor productivity (TFP). Labor 0.4 percent during 2013-18—a slowdown broadly in line productivity data used in this Box are available for nine with the emerging market and developing economy EMDEs in South America (Argentina, Bolivia, Brazil, (EMDE) average but from lower starting rates (Figure Chile, Colombia, Ecuador, Paraguay, Peru, and Uruguay), 2.3.1.1.A). Although the level of productivity in LAC is seven EMDEs in North and Central America (Costa Rica, still higher than in most other EMDE regions, sluggish El Salvador, Guatemala, Honduras, Mexico, Nicaragua, productivity growth in the post-crisis period has slowed and Panama), and nine EMDEs in the Caribbean the region’s progress toward the level of productivity in (Barbados, Belize, the Dominican Republic, Guyana, advanced economies (Figure 2.3.1.1.B). The productivity Haiti, Jamaica, St. Lucia, St. Vincent and the Grenadines, slowdown during 2013-18 was broad based, affecting three and Suriname). Data availability further restricts the -fifths of LAC countries (Figure 2.3.1.1.C). sample in the two decomposition exercises below. Within LAC, productivity growth has been heterogeneous Evolution of regional productivity across the three geographical subregions. South America, which was hard hit by the 2011-16 commodity price slide, Post-crisis productivity growth slowdown to near zero. political uncertainty, and challenging macroeconomic Like other EMDE regions, productivity growth in LAC conditions in the largest economies, had the lowest has slowed since the global financial crisis. At 0.4 percent productivity growth during 2013-18, at an average of just during 2013-18, the post-crisis average returned 0.1 percent per year, and the Caribbean had the highest, at productivity growth to its long-term average of near zero 2.5 percent. Productivity growth in the Mexico and (0.3 percent; Figure 2.3.1.1.D). This rate is well below Central America subregion was 1.2 percent during 2013- average post-crisis productivity growth in EMDEs (2.6 18, higher than in 2003-08, although this occurred in the percent). Negative productivity growth occurred 9 of 25 context of weak long-term productivity growth in Mexico. countries, nearly all of which are in South America and the Caribbean, in 2013-18. In most cases, productivity growth Against this backdrop, this Box addresses the following was also lower than both the pre-crisis and long-term questions: averages, as major economies in the region struggled with poor business climates, political tensions, regulatory • How has productivity growth evolved in the region? burdens, and plunging commodity prices. Over the course • What factors have been associated with productivity of the past four decades, troughs in productivity growth growth in the region? have broadly coincided with major adverse economic events, including a series of severe debt crises in the 1980s that spawned the region’s “lost decade,” the global Note: This box was prepared by Dana Vorisek, building upon analysis financial crisis, and periodic commodity price slumps. A in Chapter 3. Research assistance was provided by Vanessa Arellano Banoni and Shijie Shi. brief pre-crisis burst in productivity growth to 1.7 percent G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 LATIN AME RIC A AN D THE C ARIBBE AN 103 BOX 2.3.1 Labor productivity in Latin America and the Caribbean: Trends and drivers (continued) FIGURE 2.3.1.1 Evolution of labor productivity growth in LAC Productivity growth in LAC fell from 1.7 percent in 2003-08 to 0.4 percent in 2013-18. The level of productivity in LAC is still higher than that in other EMDE regions, yet sluggish productivity growth in the post-crisis period has resulted in the region’s losing ground in converging toward the level of productivity in advanced economies. Despite weak aggregate productivity growth in the region, some countries, including Bolivia, Costa Rica, the Dominican Republic, and Paraguay, achieved productivity growth in line with the EMDE average during 2013-18. A. Productivity growth B. Level of productivity and rate of C. Share of economies with 2013-18 convergence productivity growth below previous averages D. Productivity growth E. Productivity growth, by country F. Productivity levels, 2018 Source: Conference Board; Penn World Tables; World Bank (World Development Indicators). A.-E. Productivity is defined as labor productivity (real GDP per person employed). Country group aggregates for a given year are calculated using constant 2010 U.S. dollar GDP weights. Data for multiyear spans shows simple averages of the annual data. Sample includes 25 LAC countries and 127 EMDEs. A. Blue bars show the range of average productivity across the six EMDE regions: East Asia Pacific (EAP), Europe and Central Asia (ECA), Latin America and the Caribbean (LAC), the Middle East and North Africa (MENA), South Asia (SAR), and Sub-Saharan Africa (SSA). Orange dashes show the average of the six regional aggregates. B. Rate of convergence is calculated as the difference in productivity growth rates over the log difference in productivity levels between LAC and advanced economies (AEs). Blue bars and orange dashes show the range and average of the six EMDE regional aggregates. “Level” of productivity refers to the GDP-weighted average of regional productivity as a share of the average advanced economy during 2013-18. C. Orange line represents a 50 percent threshold. D. Dotted lines show 1981-2018 averages. E. Data for multiyear spans shows simple averages of the annual data. DOM = the Dominican Republic, PRY = Paraguay, BOL = Bolivia, CRI = Costa Rica, PAN = Panama, PER = Peru, COL = Colombia, GTM = Guatemala, URY = Uruguay, MEX = Mexico, SLV = El Salvador, NIC = Nicaragua, CHL = Chile, HND = Honduras, BRA = Brazil, HTI = Haiti, JAM = Jamaica, BRB = Barbados, ARG = Argentina, ECU = Ecuador, and SUR = Suriname. F. Productivity is measured in 2010 U.S. dollars. Country group aggregates are calculated using 2010 U.S. dollar GDP weights. Sample includes 25 LAC economies (9 in South America, 7 in Mexico and Central America, and 9 in the Caribbean) and 127 EMDEs. Click here to download data and charts. during 2003-08 comprised LAC’s second-longest period of highest labor productivity growth in the region, measuring positive productivity growth since 1980. well above pre-crisis and long-term averages (Figure 2.3.1.1.E). The improvement in the Dominican Republic Within-region heterogeneity of labor productivity reflects greater contribution from capital deepening and growth. Notwithstanding weak labor productivity growth higher TFP growth; this arose from increased foreign at the aggregate level in LAC during 2013-18, there was direct investment (FDI) inflows, which were encouraged considerable heterogeneity across countries. Bolivia, Costa by the reforms that opened most sectors to foreign Rica, the Dominican Republic, and Paraguay featured the 104 CHAPTER 2.3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.3.1 Labor productivity in Latin America and the Caribbean: Trends and drivers (continued) investment and by tax incentives for foreign investment economies (Figure 2.3.1.2.B). TFP growth in South (World Bank 2018i). Bolivia and Paraguay benefited from America was continually negative during 2013-18, in population migration from rural to urban areas, which part reflecting growing directed credit in Brazil (Dutz coincided with a shrinking share of agriculture as a share of 2018; Calice, Ribiero, and Byskov 2018). It also employment (IMF 2016b; World Bank 2018j). In Costa reflected intensifying economic distortions (such as Rica, the rise in productivity growth was broad-based trade restrictions and price controls) in Argentina across sectors, supported by continued policy reforms and during the early part of the period, which allowed positive spillovers from FDI inflows (OECD 2018a). In unproductive firms to survive. four of the six LAC economies with the highest productivity growth during 2013-18 (Bolivia, the • Mexico and Central America. In Mexico and Central Dominican Republic, Panama, and Peru), productivity America, the early impacts of the global financial crisis growth benefited from the steepest declines in the share of in 2007 and 2008 weighed on TFP in Mexico during informal activity in the region during the decade to 2016 2003-08. Although post-crisis TFP growth was (World Bank 2019f). subdued, and capital deepening weakened during this period in the context of the repeated bouts of policy High productivity levels relative to EMDEs but slowing uncertainty, the removal of the crisis effects in Mexico convergence with advanced economies. Despite low allowed higher productivity growth in the subregion productivity growth in the region over an extended period, during 2013-18. the level of productivity in LAC (22 percent of the advanced-economy average) is above the EMDE average • The Caribbean. In the Caribbean, TFP growth (19 percent of the advanced-economy average; Figure accelerated during the post-crisis period, largely 2.3.1.1.F). High productivity levels in LAC are a legacy of reflecting capital deepening in the largest economy in the mid-20th century. Since the 1980s, labor productivity the subregion, the Dominican Republic. in LAC relative to the level in advanced economies has fallen (Ferreira, de Abreu Pessôa, and Veloso 2013; Post-crisis productivity growth slowdown across sectors. Fernández-Arias and Rodríguez-Apolinar 2016). The pre- As in the average EMDE, manufacturing made the largest crisis rise in productivity growth halted this divergence sectoral contribution to productivity growth in LAC only briefly. This is in stark contrast to the narrowing during the 1990s and the pre-crisis period. Relative to the labor productivity gap between the broader group of pre-crisis period, the post-crisis period in LAC was marked EMDEs and advanced economies since the 1990s. by a broad-based slowdown in productivity growth across sectors, particularly in manufacturing, trade, and finance. Sources of regional productivity growth Stalling within-sector labor productivity growth. For Labor productivity can be decomposed into three countries with available sectoral data, the within-sector sources—human capital accumulation, physical capital contribution to productivity growth has historically been accumulation, and TFP, or the efficiency with which labor greater than the between-sector contribution from labor and capital are used during production. The post-crisis reallocation from low-productivity to higher-productivity productivity growth slowdown predominantly reflected a sectors (Figure 2.3.1.3.A). This is consistent with other return to negative TFP growth rates, as had prevailed in studies of the region (Brown et al. 2016; Diao, McMillan, LAC during the 1990s (Figure 2.3.1.2.A; Busso, Madrigal, and Rodrik 2017). During the 1990s, a substantial part of Pagés 2013). However, the post-crisis (2013-18) average labor productivity growth was due to within-sector growth disguises a steep slowdown in investment growth during as LAC countries liberalized trade policy in the second half 2016-18, as Brazil struggled to exit a deep recession, the of the 1980s and the early 1990s (Rodrik 2016a). The effects of the commodity price slump rippled through the 1990s and early 2000s were a period of significant change region’s many commodity-reliant economies, and in LAC’s manufacturing industry. Faced with increasing numerous economies experienced bouts of policy foreign competition as the result of globalization, domestic uncertainty. manufacturing firms implemented more efficient processes that required less labor, and uncompetitive firms ceased • South America. The post-crisis labor productivity operating. As workers were displaced from manufacturing, slowdown was most pronounced in South America. they shifted toward lower-productivity services and which was deeply impacted by the commodity price informal activities (Pagés-Serra 2010; McMillan, Rodrik, slump and country-specific constraints in large and Verduzco-Gallo 2014). In Argentina and Brazil, two G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 LATIN AME RIC A AN D THE C ARIBBE AN 105 BOX 2.3.1 Labor productivity in Latin America and the Caribbean: Trends and drivers (continued) of the largest economies in the region, labor shifted in the 1990s from manufacturing into less-productive non- FIGURE 2.3.1.2 Sources of productivity tradable sectors, such as personal services and wholesale growth in LAC and retail trade, limiting between-sector productivity Sluggish productivity growth in LAC during the post- growth. Of the six LAC countries with available sectoral crisis period predominantly reflected a negative data, only Costa Rica and Mexico have consistently contribution from total factor productivity (TFP). The TFP contraction was especially pronounced in South experienced positive between-sector productivity growth, America. In recent years, capital deepening has made a and even in those cases the within-sector contribution has slowing contribution to productivity growth. been smaller than the between-sector contribution. As the manufacturing sector in LAC transformed during A. Contributions to productivity growth the 1990s and early 2000s, the agricultural sector became more productive relative to other sectors, with a shrinking share of agricultural employment accounting for a stable share of output between 1995 and 2008 (Figures 2.3.1.3.B and 2.3.1.3.C). The government sector, however, became less productive, accounting for a growing share of employment and the same share of output. Since 2013, between-sector productivity gains have stalled in several large economies (Argentina, Brazil, Colombia). Within-sector productivity growth has collapsed to near zero as multiple structural constraints (e.g., inefficient provision of credit in Brazil and trade restrictions and price controls in Argentina) were compounded by an inability to adjust to adverse events, including unfavorable policy choices, a commodity price collapse, and financial stress B. Contributions to productivity growth, by subregion episodes. Sectoral productivity levels in LAC relative to EMDEs. In most sectors, and particularly in mining, productivity levels in LAC are higher than the EMDE average. However, LAC lags notably in trade and finance. Removing productivity barriers in these sectors would benefit aggregate regional productivity. Key drivers of productivity. LAC has long lagged other EMDE regions in several key drivers of productivity— investment, innovation, and trade—and performs only about average in other drivers (Figure 2.3.1.4.A). Over time, the drivers of productivity in LAC have improved but the improvement has not kept pace with that in Source: Barro and Lee (2015); International Monetary Fund; Penn World Tables; United Nations (Human Development Reports), Wittgenstein Centre EMDEs (Figure 2.3.1.4.B). Cyclical factors, such as weak for Demography and Global Human Capital; World Bank. investment in large economies in the region and gyrations A-B. Country groups aggregated using constant 2010 U.S. dollar GDP weights. in global commodity price trends, are also linked to weak A. Samples include 25 LAC economies and 92 EMDEs. productivity growth in LAC. Investment growth has B. Samples include 9 economies in South America, 7 economies in Mexico and Central America, and 9 economies in the Caribbean. weakened substantially in the post-crisis period (Figure Click here to download data and charts. 2.3.1.4.C). Limited innovation and technology adoption. Innovation, achieved through dedicating resources to research and development (R&D) or introducing new processes or 106 CHAPTER 2.3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.3.1 Labor productivity in Latin America and the Caribbean: Trends and drivers (continued) FIGURE 2.3.1.3 Sectoral productivity in LAC Within-sector productivity growth, the main driver of productivity growth in LAC during the pre-crisis period, was much lower during the post-crisis period in several large economies, while between-sector productivity growth slowed in all economies with available sectoral data. A. Within-sector and between-sector B. Composition of employment, by sector C. Composition of GDP, by sector contributions to productivity growth Source: Groningen Growth Development Center database, Haver Analytics, ILOSTAT, OECD STAN, United Nations, World KLEMS, World Bank. A. The within-sector productivity contribution shows the initial real value added-weighted productivity growth; the between-sector contribution measures the productivity growth from a cross-sectoral shift of employment. ARG = Argentina, BRA = Brazil, CHL = Chile, COL = Colombia, CRI = Costa Rica, and MEX = Mexico. B.C. “Other industry” includes construction, mining, and utilities; “finance” includes business services; “government” includes personal services. Samples include 6 LAC economies (Argentina, Brazil, Chile, Colombia, Costa Rica, and Mexico) and 46 EMDEs. Click here to download data and charts. products, has been a key driver of labor and firm harmonization of rules of origin and non-tariff measures productivity in LAC (Crespi and Zuniga 2011; Grazzi and across agreements, and there is no region-wide trade Jung 2016). Likewise, adoption of new technologies can agreement. These characteristics result in fragmentation of reduce information costs and facilitate market access, trading priorities and, together with weak diversification of thereby increasing productivity and expanding output in traded goods in many countries, limit the development of the region (Dutz, Almeida, and Packard 2018). LAC is intraregional global value chains. Rules of origin imposed missing key opportunities to raise productivity through under preferential trade agreements in the region are these channels. R&D expenditure as a share of GDP is low estimated to negate more than 15 percent of the positive in LAC relative to that in comparator EMDEs, as is the trade effect of the agreements, while the costs of non-tariff likelihood of firms in LAC introducing product measures imposed by LAC countries are estimated to innovations (Lederman et al. 2014; Figure 2.3.1.4.D). equate to a 15 percent tariff for intermediate goods (Cadestin, Gourdon, and Kowalski 2016). Weak trade linkages. In three large economies in the region (Argentina, Chile, Mexico), deeper participation in Poor-quality education and labor market constraints. At a global value chains is associated with positive effects on median of 9.2 years in 2018, the duration of schooling in firm productivity (Montalbano, Nenci, and Pietrobelli LAC compares favorably with 7.7 years in the average 2018). Yet nearly all LAC economies trade less (as a share EMDE. In addition, the gap between the median years of of their GDP) than EMDEs overall, and global value chain schooling in LAC and advanced economies narrowed participation is lower than in the East Asia and Pacific during the past decade, from 3.5 years in 2008 to 2.9 years region and in Europe and Central Asia (Figure 2.3.1.4.E). in 2018. However, learning outcomes in LAC fall short of Even the LAC countries most integrated in global value their potential, as indicated by international standardized chains (Chile, Costa Rica, and Mexico) are not among the test results and high dropout rates at the tertiary level most integrated EMDEs (OECD 2018b). The (World Bank 2017b). Moreover, in most LAC countries, opportunity for regional productivity gains through trade education outcomes are highly correlated with is further hindered by the structure of intra- and socioeconomic conditions, a scenario reinforced by extraregional trade relationships. Although LAC countries persistently elevated income inequality (World Bank are party to numerous trade agreements, there is little 2018k). Ultimately, skills deficiencies and mismatches and G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 LATIN AME RIC A AN D THE C ARIBBE AN 107 BOX 2.3.1 Labor productivity in Latin America and the Caribbean: Trends and drivers (continued) FIGURE 2.3.1.4 Drivers of labor productivity growth in LAC Multiple structural constraints contribute to low productivity growth in LAC. The region performs particularly poorly relative to other EMDE regions in measures of investment, innovation, and trade. In other drivers, LAC is a mediocre performer relative to other regions. The drivers of productivity growth have become more supportive over time but at a slower pace than the EMDE average. A. Drivers of productivity growth, 2017 B. Index of productivity drivers C. Investment growth D. R&D spending E. Trade F. Firms indicating inadequately educated workers as their biggest obstacle Source: Freedom House; Haver Analytics; International Country Risk Guide; Organisation for Economic Co-operation and Development; Observatory of Economic Complexity; Penn World Tables; United Nations Educational, Scientific, and Cultural Organization (Institute for Statistics); United Nations Population Prospects; World Integrated Trade Solution; World Bank (Doing Business, Enterprise Surveys, and Global Financial Development Database). A. Unweighted average levels of drivers, normalized as average of AEs (index =100) and standard deviation of EMDEs as 10. Blue bars represent average of LAC economies in 2017. Orange whiskers represent range of averages for the six EMDE regions in 2017. Variables are defined as: Education=years of education, Urbanization=share of population living in urban areas, Investment=share of investment to GDP, Institutions=government effectiveness, Economic complexity=Economic Complexity Index of Hidalgo and Hausmann (2009), Gender equality=share of years of schooling for females to males, Demography=share of population under age 14, Innovation=log patents per capita, and Trade=(exports+imports)/GDP. Samples include 17-31 LAC economies, depending on the driver, and 63-150 EMDEs. B. For each country, index is a weighted average—weighted by the normalized coefficients shown in Annex 3.3—of the normalized value of each driver of productivity. Drivers include the ICRG rule of law index, patents per capita, non-tropical share of land area, investment in percent of GDP, ratio of female average years of education to male average years, and share of population in urban area, Economic Complexity Index, years of schooling, working-age share of population, and inflation. Regional and EMDE indexes are GDP-weighted averages for single years and simple averages for time periods. Samples includes 17 LAC economies and 77 EMDEs. C. Bars show investment-weighted averages. Last observation is 2019Q3. Data for Mexico for Q3 is estimated. D. Sample includes 16 economies for LAC and 94 for EMDEs. E. Bars show 2015-17 average of exports plus imports as a share of GDP. BRA = Brazil, ARG = Argentina, COL = Colombia, URY = Uruguay, DOM = the Dominican Republic, ECU = Ecuador, PAN = Panama, PER = Peru, GTM = Guatemala, JAM = Jamaica, CRI = Costa Rica, CHL = Chile, BOL = Bolivia, PRY = Paraguay, SLV = El Salvador, HND = Honduras, MEX = Mexico, NIC = Nicaragua. Sample includes 96 EMDEs. F. Sample includes 30 LAC economies and 113 EMDEs. Click here to download data and charts. low-quality education have negative implications for labor indicate that 7 percent of firms in LAC perceive an productivity and the functioning of labor markets. The inadequately educated workforce as their biggest obstacle, incidence of youth who are neither in school nor working more than double the share in all EMDEs (Figure is high (de Hoyos, Rogers, and Székely 2016). An 2.3.1.4.F). The poor functioning of labor markets due to estimated half of firms are unable to find local workers skills deficiencies are compounded by longstanding with the skills they need, and consequently turn to foreign regulatory rigidities that prevent efficient allocation and labor (OECD 2018b). Firm-level survey data for 2013-18 mobility of workers (Kaplan 2009). 108 CHAPTER 2.3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.3.1 Labor productivity in Latin America and the Caribbean: Trends and drivers (continued) High informality. The informal sector averages slightly through development of secure digital payment systems more than one-third of GDP, higher than in all other and fintech regulatory frameworks (World Bank 2017c). EMDE regions except Sub-Saharan Africa (World Bank Improving the speed of uptake of new technologies in 2019f). In LAC, high informality has been associated with LAC, where firms adopt new technologies with a lower aggregate and firm-level productivity (Loayza, significant lag relative to the United States, would also Servén, and Sugawara 2010; de Paula and Sheinkman boost productivity (Eden and Nguyen 2016). 2011; Chong, Galdo, and Saavedra 2008). In Paraguay, informal firms are not only less productive than formal Deepen trade linkages and reduce trade barriers. Trade firms, but have negative spillovers on formal firms’ relationships can boost productivity by facilitating productivity (Vargas 2015). knowledge exchange and innovation for the participating firms (Bown et al. 2017). Significant productivity gains Policy options could be made by reducing barriers to trade in LAC. The landmark European Union-Mercosur trade agreement, A range of options, targeted to country experiences, can be finalized by negotiators in June 2019 but not yet ratified, pursued to boost productivity in LAC and put the region holds significant promise for decreasing trade barriers and on a path toward closing the productivity gap with deepening trade flows between Latin America and Europe. advanced economies. Productivity in the region stands to In addition, there have been some recent efforts to reduce benefit most from policy reforms to boost TFP, rather trade barriers within the region; for instance, the Pacific than to improve factors of production. Alliance eliminated tariffs among its members (Chile, Colombia, Mexico, and Peru) in May 2016. Improving factors of production Boost quality of education and implement labor market Increase the volume and efficiency of infrastructure reforms. With the working-age share of the population in investment. Relative to the pre-crisis period, capital the region now at a peak and on track to begin the long- deepening has been the main source of productivity term downward trajectory that East Asia and Pacific and growth in large parts of the region during the post-crisis Europe and Central Asia have already begun, the period. However, it has slowed sharply in the past three contribution of additional labor to productivity growth in years, and large infrastructure gaps remain. Access to water LAC will fall in the years ahead. Advancing human capital and electricity in LAC is high relative to all EMDEs; through education and skills development will become however, the region underperforms in transportation and increasingly important. For many countries in the region, sanitation (Fay et al. 2017). To address this, transportation including Brazil, adapting labor markets to shifting development is underway in several countries. Colombia, economic opportunities in the strongly integrated global for instance, is implementing 4G, a major public road economy will require revision of dated labor market infrastructure program. In addition, across the region, regulation (Dutz 2018). For firms, additional use of on- there is significant capacity to reduce infrastructure gaps by the-job training is an important element of boosting the improving infrastructure spending efficiency—in productivity of their workers, especially in the context of particular, through improvements at the appraisal and rapidly changing technologies. Implementing programs evaluation stages of public investment projects and in that engage youth who are neither working nor studying is public procurement systems. a critical policy concern in the region (Almeida and Packard 2018). Skills training programs such as Jovenes en Boosting firm productivity Acción in Colombia and ProJoven in Peru have had Pursue well-targeted competition and innovation positive impacts on employment and productivity among policies. Reducing barriers to entry for firms and the the target populations and could be replicated elsewhere rigidity of labor regulations, on which LAC performs (Attanasio et al. 2015; Diaz and Rosas 2016). poorly compared to other EMDE regions and which Apprenticeship programs, which have been successful in encourages informal operation, is critical for promoting several advanced economies, could also be explored. entrepreneurship and productivity. In Peru, for example, Reducing labor market rigidities (such as restrictions on the elimination of subnational barriers to entry is found to use of term contracts, restrictions on working hours, use of have boosted firm productivity (Schiffbauer and Sampi minimum wages above market equilibrium, and 2019). Boosting low R&D spending and low technology- imposition of high costs and penalties for redundancy) can related innovations can also improve financial inclusion boost productivity. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 LATIN AME RIC A AN D THE C ARIBBE AN 109 BOX 2.3.1 Labor productivity in Latin America and the Caribbean: Trends and drivers (continued) Encouraging sectoral reallocation productivity over the long term. For instance, fair contract enforcement, straightforward and transparent legal Given that within-sector productivity gains in several large processes, and contained political risk have all been shown economies in LAC have stalled since the global financial to support productivity gains (Acemoglu et al. 2019; crisis, countries in the region should rekindle efforts to Rodrik 1999; Rodrik, Subramanian, and Trebbi 2004). implement policies that reallocate capital and labor Relative to other regions, however, LAC is a mediocre towards more productive firms within the sectors. Policies performer on measures of governance (Figure 2.3.1.4.A). could aim to strengthen competition, including through Moreover, the region’s performance has deteriorated trade, and reform labor markets to facilitate the movement during the post-crisis period in measures of government and productivity of labor. At the same time, the effectiveness, control of corruption, and regulatory quality longstanding weakness in the region’s between-sector (Kaufmann, Kraay, and Mastruzzi 2010). Especially when productivity growth in the region calls for policies that the burden of regulation is high, as it tends to be in LAC, reduce misallocation of capital and labor toward sectors corruption is detrimental for productivity (Amin and Ulku with low productivity. In particular, with limited 2019). On measures of doing business, no country in LAC opportunity for further industrialization, LAC countries is among the top 50 performers in the world (World Bank should target lack of competition in services industries, 2020). Business environment reforms can also help reduce including transport, finance, trade, and information and the size of the informal sector, where productivity is lower communications technology, and ensure that workers have than in the formal sector. The process of institutional sufficiently strong skills to thrive in occupations being reforms could be spearheaded through productivity transformed by technology (Araujo, Vostroknutova, and commissions such as those created in Chile, Colombia, Wacker 2017; World Bank forthcoming). and Mexico. Colombia, for example, is implementing a Creating a business-friendly environment series of structural reforms as part of its Productive Development Policy 2016-2025. Implement supportive governance and business climate reforms. Institutional quality is a key driver of Regional growth in the Middle East and North Africa decelerated to an estimated 0.1 percent in 2019. Geopolitical and policy constraints on oil sector production slowed growth in oil-exporting economies, despite support from public spending. Growth in oil importers remained stable, as reform progress and resilient tourism activity were offset by structural and external headwinds. Regional growth is projected to pick up to 2.4 percent in 2020 and to about 2.8 percent in 2021-22, as infrastructure investment and business climate reforms proceed. Risks are tilted firmly to the downside—geopolitical tensions, escalation of armed conflicts, slower-than-expected pace of reforms, or weaker-than-expected growth in key trading partners could heavily constrain activity. Sustained proliferation of these risks could also hamper long-term productivity prospects. Recent developments to less supportive global demand, commitments to the oil production-cut agreement of the Growth in the Middle East and North Africa Organization of the Petroleum Exporting (MENA) slowed to an estimated 0.1 percent in Countries and other signatory countries (OPEC+) 2019, down from 0.8 percent the previous year and regional geopolitical events further (Table 2.4.1; Figure 2.4.1.A).1 The slowdown constrained the oil sector. largely reflected the sharp growth contraction in Among oil importers, growth has been more the Islamic Republic of Iran, following the stable. In Egypt, the subregion’s largest economy, tightening of U.S. sanctions, geopolitical tensions net exports as well as investment, partly supported in the Strait of Hormuz, and diplomatic setbacks. by more accommodative monetary stance, Weakened global growth weighed on demand for continued to support growth. The maturity of its oil and other exports, further hindering activity in external debt has also shifted towards long-term the region generally (Figure 2.4.1.B). instruments (Figure 2.4.1.D). Favorable tourism Public spending has been robust in some oil activity continues to support growth in oil exporters, including those in the Gulf importers, such as Morocco and Tunisia. Cooperation Council (GCC). Non-oil activity has However, agricultural production has become less also shown supportive signs (Figure 2.4.1.C). favorable and weighed on activity in Morocco. However, these developments were insufficient to Export growth potential in oil importers was offset weak activity in the oil sector. In addition weighed by weakened global demand, including from the Euro Area. Inflation in the region generally eased. In GCC Note: is section was prepared by Lei Sandy Ye. Research economies in 2019, it registered less than 1 assistance was provided by Vanessa Arellano Banoni. 1 e World Bank’s Middle East and North Africa aggregate percent on average (Figure 2.4.1.E). Inflation in includes 16 economies and is grouped into three subregions. Bahrain, Egypt subsided substantially in the second half of Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates the year, allowing the central bank to cut interest comprise the Gulf Cooperation Council (GCC); all are oil exporters. Other oil exporters in the region are Algeria, the Islamic Republic of rates three times since August. In smaller oil Iran, and Iraq. Oil importers in the region are Djibouti, the Arab importers (e.g., Jordan), inflation has also Republic of Egypt, Jordan, Lebanon, Morocco, Tunisia, and West moderated generally. In Iran, however, inflation Bank and Gaza. Syrian Arab Republic, the Republic of Yemen, and Libya are excluded from regional growth aggregates due to data rose sharply to more than 50 percent in mid-2019, limitations. partly reflecting the earlier depreciation of the rial 112 CHAPTER 2.4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 2.4.1 MENA: Recent developments in the parallel market, although inflation has Growth in the MENA region fell in 2019, for the third consecutive year, to an subsided in late 2019 to below 30 percent. estimated 0.1 percent. In the large oil-exporting economies, oil production cuts, weak global economic momentum, and U.S. sanctions on Iran Financial sector conditions in MENA have been weighed on activity, despite signs of non-oil activity improvement. Activity supportive to activity. Banking systems in the among oil importers was supported by improved conditions in Egypt. Inflation rose sharply in Iran, while remaining generally low elsewhere. GCC economies remain broadly resilient, with Easier financing conditions in advanced economies have supported capital adequacy ratios generally sound and non- international capital raising. performing loan ratios contained. Benign global financing conditions associated with more A. Growth B. Global oil demand growth and prices accommodative advanced economies’ monetary policy have supported equity flows in the region and encouraged investor risk appetite in the large economies (e.g., GCC and Egypt). In the GCC, new bonds were issued in international capital markets in both the corporate and sovereign sectors, and bank credit growth has shown improvement (Figure 2.4.1.F). However, access to finance elsewhere remains a major obstacle to investment, especially for small and medium-sized C. Composite PMI D. Egypt: Inflation, policy rate, and enterprises (SMEs; Ghassibe, Appendino, and external debt maturity Mahmoudi 2019). Outlook Growth in the region is projected to accelerate in 2020 to 2.4 percent, supported by higher investment, promoted by both infrastructure initiatives and stronger business climates. The forecasted stabilization in Iran assumes that the E. Inflation F. International debt securities impact of sanctions tapers somewhat (Table outstanding 2.4.2). Regional growth is expected to remain stable over 2021-22, at about 2.8 percent. Continued reform efforts and strengthening domestic demand in key economies should provide support to activity. Despite the projected growth acceleration, long-standing challenges, such as high unemployment rates among youth and women and high poverty rates in some countries, will remain. In particular, for economies Source: Bank for International Settlements; Haver Analytics; International Energy Agency; Interna- affected by fragility, conflict, and violence, armed tional Monetary Fund; World Bank. A. Weighted average growth rates of real GDP. Gray shaded area denotes forecasts. conflicts imposed further setbacks to poverty via B. Left panel denotes year-on-year growth of period average global oil demand in millions of barrels a day. Right panel denotes average oil price during the periods shown. Oil price denotes average of lower provision of public services and social safety Brent, Dubai, and WTI. 2019H2 denotes latest data as of Dec 19. 2020 denote World Bank forecast nets. More sustained growth will be needed to for oil price and IEA forecast for global oil demand. 2019 global oil demand data are estimates. C. Above 50 denotes expansion; below 50 contraction. 2019Q4 denotes average of October and resolve these challenges. November. D. Averages over the period denoted. Policy rate refers to the overnight lending rate. Inflation refers to CPI inflation. 2019Q4 data for inflation and policy rate denote average of October and November. Among oil exporters, growth is expected to pick E. CPI inflation (year-on-year monthly rate). Last observation is November 2019 for Iran and oil importers and October for the GCC. GCC include 6 economies. Oil importers include 4 economies. up to 2 percent in 2020. Infrastructure F. Includes 5 GCC and 6 non-GCC economies. Sum of international debt securities outstanding. investment, along with an improved regulatory “Corporates” include non-financial and financial corporations. Click here to download data and charts. environment backed by business climate reforms, are expected to support activity in the GCC G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 MID D LE E AS T AN D NO RTH AFRIC A 113 (World Bank 2019q, Figure 2.4.2.A). Iran’s FIGURE 2.4.2 MENA: Outlook and risks economy is expected to stagnate at a lower base, as Stronger momentum in the non-oil sector in the GCC, aided by business the initial intensive impact of sanctions on oil climate reforms, is expected to support activity. Oil importers’ growth production and exports is assumed to taper prospects are also supported by policy reforms but are challenged by high debt levels and structural issues. Geopolitical risks are acute and have somewhat. Algeria’s growth is expected to pick up prolonged the refugee crisis in fragile areas. Political instability hampers modestly, as policy uncertainty abates somewhat reform progress and poses a major constraint to productivity. Lower-than- expected growth in the Euro Area would constrain external demand for the and investment improves. Investment associated region, especially oil importers. with reconstruction and fiscal easing is expected to support Iraq’s growth. Facilities and capacity A. Improvement in business climate: B. Public debt in MENA expansion in oil and gas sectors is also expected to 2018-19 support activity in many oil exporters. Over the medium term, growth in GCC economies is expected to remain steady, underpinned by planned diversification programs, longer-term infrastructure programs, and measures to ease foreign investment restrictions. Growth in oil importers is expected to rise slightly in 2020, to 4.4 percent, led by improvements in larger economies. Growth in oil importers is C. Labor market competitiveness D. Syrian refugees’ intention to return contingent upon the materialization of reform plans and no escalation of political risks. Tourism, aided by government promotion initiatives and improved security, is expected to continue supporting activity in Egypt, Morocco, and Tunisia. However, for smaller oil importers, banking sector fragility and high public debt are significant constraints on growth (Figure 2.4.2.B). Moreover, the sustainability of debt or external E. Political instability as biggest F. Euro Area growth forecasts position in these economies often depends on the obstacle to firm operations materialization of expected multilateral and bilateral financing flows or on the strength of sovereign credit; and are vulnerable to sudden shifts in market confidence. Modest growth in smaller oil importers weighs further on the high budgetary financing pressures of these economies and the sustainability of their high debt. Medium-term growth prospects for the MENA Source: Bank for International Settlements,; Haver Analytics; International Monetary Fund; United region are contingent on an attenuation of armed Nations; World Bank; World Economic Forum. conflicts, and on limiting their regional spillovers. A. Includes 6 GCC and 9 non-GCC economies. Unweighted average of each economy’s change in Distance to Frontier Score in the denoted measures between 2018-19 (2020 DoingBusiness edition). Structural reforms, such as those to provide B. Unweighted averages. 2019 data are estimates. C. Index of labor market competitiveness based on the Global Competitiveness Index. Index stronger fiscal management and to enhance the constructed based on data on labor market entry/exit, wage flexibility and skills match. Unweighted averages. AE denotes advanced economies. Based on 2019 data edition. investment climate, are underway in many GCC D. Based on United Nation’s Annual Surveys on Syrians’ Refugees’ Perceptions and Intentions to and non-GCC economies. New financial reforms, Return to Syria. Survey respondents include Syrian refugees in Egypt, Iraq, Lebanon, and Jordan. X- axis denotes two questions to survey respondents on whether they “hope to return to Syria one day” such as investment law and stronger minority and whether they “intend to return to Syria in the next 12 months (“Yes”, “No”, “Do not know”). 2018 data denote survey conducted between Nov 2018 and Feb 2019. investor protection in Egypt; the relaxation of E. Percent of firms citing political instability as biggest obstacle to business operations, based on the World Bank’s Enterprise Surveys data. Unweighted averages across economies. Data for latest foreign investment restrictions across 13 sectors available year across 9 MENA economies. and in SME licensing in the United Arab F. Legend denotes month-year for which World Bank forecast is published. Columns denote the growth forecast year. Emirates; and a new secured transactions law in Click here to download data and charts. 114 CHAPTER 2.4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 Jordan have been adopted. They are expected to government (Mansour, Maseeh, and Celiku help relieve financial constraints in the corporate 2019). Such uncertainties and delays could hinder sector, support investor confidence, and raise productivity and private sector development – foreign direct investment (FDI). Structural survey evidence shows that political instability is reforms of this nature could help raise the by far a bigger obstacle to firm operations in historically weak long-term productivity MENA than any other EMDE regions (Figure performance in these countries (Arezki et al. 2.4.2.E; World Bank 2016f). 2019a,b; Youssef et al. 2019; Box 2.4.1). Nonetheless, the scope for improvements in many Geopolitical factors related to U.S.-Iran tensions, areas remains large – for instance, limited churn of as well as the recent attack on Saudi Aramco’s oil firms, barriers to competition, and labor market facilities, have raised volatility in oil prices. This inefficiencies hinder MENA firms’ ability to volatility may rise further. A sharp rise in oil price generate private sector jobs (Figure 2.4.2C; Arezki volatility may complicate or stall fiscal adjustments et al. 2019a). in both oil exporters and importers. It could also set back investment programs in oil exporters and cause difficulties for subsidy reforms in oil Risks importers by increasing the uncertainty associated Risks are firmly tilted to the downside. These with future revenue and income streams. include the long-standing risks from geopolitical Renewed escalation of global trade tensions may conflicts, political uncertainty, and volatility in oil further weaken growth prospects in advanced prices as well as more recent risks associated with economies and several large EMDEs. This may reescalation of global trade tensions. translate into further setback to growth in the Geopolitical risks have increased substantially. Euro Area, to which the MENA region and Syria and surrounding countries remain filled especially the Maghreb region have significant with high uncertainties and diverse intra– and trade exposure (Figure 2.4.2.F). Oil importers are interregional developments. Armed conflicts in subject to risks from the GCC, a significant source Syria have held back refugees’ short-term intention of remittances and FDI flows. Global trade to return, despite greater desire to ultimately tensions may also affect the MENA region resettle in their home country (Figure 2.4.2.D). In through the oil price channel (IEA 2019). Sharp Yemen, the near-term prospects remain highly oil price declines via weaker global oil demand uncertain due to the active conflict, now in its would significantly affect activity in MENA oil fifth year. Yemen’s socioeconomic outlook exporters. depends critically on a cessation of hostilities and a renewed political vision for the country. An Volatility in external financing conditions could escalation of U.S.-Iran tensions would pose destabilize MENA’s financial markets. For difficulties for other regional economies as well as example, higher uncertainty about the path of Iran itself. advanced economy’s monetary easing stance could present a downside risk to capital flows to GCC Political uncertainty also clouds MENA’s growth economies, which have low debt levels relative to prospects, particularly in non-GCC economies. oil importers but rising exposure to international While political impasse and some previously financial markets. Moreover, it could raise their delayed reforms have been partly resolved, policy difficulties in financing contingent liabilities in uncertainty in Algeria remains significant. public spending projects through large bond Reconstruction in Iraq had already experienced issuances. For oil importers, volatility in global some delays, and a lack of political consensus on interest rates could raise the debt service costs of economic reforms continues to challenge the their high levels of public debt. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 MID D LE E AS T AN D NO RTH AFRIC A 115 TABLE 2.4.1 Middle East and North Africa forecast summary Percentage point differences (Real GDP growth at market prices in percent, unless indicated otherwise) from June 2019 projections 2017 2018 2019e 2020f 2021f 2022f 2019e 2020f 2021f EMDE MENA, GDP1 1.1 0.8 0.1 2.4 2.7 2.8 -1.2 -0.8 0.0 (Average including countries with full national accounts and balance of payments data only) 2 EMDE MENA, GDP2 1.4 0.9 -0.4 2.3 2.7 2.8 -1.5 -0.6 0.0 GDP per capita (U.S. dollars) -0.4 -0.8 -2.0 0.7 1.2 1.4 -1.6 -0.8 -0.2 PPP GDP 1.7 0.9 -0.4 2.4 2.9 2.9 -1.5 -0.6 0.1 Private consumption 2.6 0.7 1.4 1.9 2.2 2.2 0.1 0.0 0.1 Public consumption 4.9 2.9 0.4 2.1 2.3 2.4 -1.0 0.7 1.3 Fixed investment 1.7 0.2 2.4 5.2 5.7 6.1 -2.0 -0.5 -0.7 Exports, GNFS3 4.5 2.4 -1.7 3.0 3.6 3.7 -2.1 -0.9 0.0 Imports, GNFS 3 7.7 -2.0 1.1 3.4 4.0 4.0 -0.8 0.1 0.2 Net exports, contribution to growth -0.5 2.0 -1.3 0.3 0.4 0.4 -0.8 -0.5 0.0 Memo items: GDP Oil exporters4 0.6 0.1 -0.8 2.0 2.3 2.3 -1.5 -0.9 0.1 GCC countries5 -0.3 2.0 0.8 2.2 2.6 2.7 -1.3 -1.0 -0.1 Saudi Arabia -0.7 2.4 0.4 1.9 2.2 2.4 -1.3 -1.2 -0.1 Iran 3.8 -4.9 -8.7 0.0 1.0 1.0 -4.2 -0.9 0.0 Oil importers6 3.8 3.9 4.0 4.4 4.6 4.6 -0.1 -0.1 -0.1 Egypt 4.8 5.5 5.7 5.9 6.0 6.0 0.0 0.0 0.0 Fiscal year basis7 4.2 5.3 5.6 5.8 6.0 6.0 0.1 0.0 0.0 Source: World Bank. Note: e = estimate; f = forecast. EMDE = emerging market and developing economies. World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of countries’ prospects do not differ at any given moment in time. 1. GDP and expenditure components are measured in 2010 prices and market exchange rates. Excludes Libya, Syria, and Yemen due to data limitations. 2. Aggregate includes all countries in notes 4 and 6 except Djibouti, Iraq, Qatar, and West Bank and Gaza, for which data limitations prevent the forecasting of GDP components. 3. Exports and imports of goods and non-factor services (GNFS). 4. Oil exporters include Algeria, Bahrain, Iran, Iraq, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates. 5. The Gulf Cooperation Council (GCC) includes Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates. 6. Oil importers include Djibouti, Egypt, Jordan, Lebanon, Morocco, Tunisia, and West Bank and Gaza. 7. The fiscal year runs from July 1 to June 30 in Egypt; the column labeled 2018 reflects the fiscal year ended June 30, 2018. Click here to download data. 116 CHAPTER 2.4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 TABLE 2.4.2 Middle East and North Africa economy forecasts1 Percentage point differences (Real GDP growth at market prices in percent, unless indicated otherwise) from June 2019 projections 2017 2018 2019e 2020f 2021f 2022f 2019e 2020f 2021f Algeria 1.3 1.4 1.3 1.9 2.2 2.2 -0.6 0.2 0.8 Bahrain 3.8 2.2 2.0 2.1 2.4 2.4 0.0 -0.1 -0.4 Djibouti 5.1 5.5 7.2 7.5 8.0 8.4 0.2 0.0 0.0 Egypt 4.8 5.5 5.7 5.9 6.0 6.0 0.0 0.0 0.0 Fiscal year basis 2 4.2 5.3 5.6 5.8 6.0 6.0 0.1 0.0 0.0 Iran 3.8 -4.9 -8.7 0.0 1.0 1.0 -4.2 -0.9 0.0 Iraq -2.5 -0.6 4.8 5.1 2.7 2.5 2.0 -3.0 0.4 Jordan 2.1 1.9 2.0 2.2 2.4 2.5 -0.2 -0.2 -0.2 Kuwait -3.5 1.2 0.4 2.2 2.0 2.0 -1.2 -0.8 -0.9 Lebanon 0.6 0.2 -0.2 0.3 0.4 0.5 -1.1 -1.0 -1.1 Morocco 4.2 3.0 2.7 3.5 3.6 3.8 -0.2 0.0 0.0 Oman 0.3 1.8 0.0 3.7 4.3 4.3 -1.2 -2.3 1.5 Qatar 1.6 1.5 0.5 1.5 3.2 3.2 -2.5 -1.7 -0.2 Saudi Arabia -0.7 2.4 0.4 1.9 2.2 2.4 -1.3 -1.2 -0.1 Tunisia 1.8 2.5 1.6 2.2 2.6 2.6 -1.1 -1.0 -0.9 United Arab Emirates 0.5 1.7 1.8 2.6 3.0 3.0 -0.8 -0.4 -0.2 West Bank and Gaza 3.1 0.9 0.5 2.5 2.6 2.7 0.0 1.5 1.0 Source: World Bank. Note: e = estimate; f = forecast. World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of economies’ prospects do not significantly differ at any given moment in time. 1. GDP at market prices and expenditure components are measured in 2010 prices and market exchange rates. Excludes Libya, Syria, and Yemen due to data limitations. 2. The fiscal year runs from July 1 to June 30 in Egypt; the column labeled 2018 reflects the fiscal year ended June 30, 2018. Click here to download data. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 MID D LE E AS T AN D NO RTH AFRIC A 117 BOX 2.4.1 Labor productivity in the Middle East and North Africa: Trends and drivers Labor productivity growth in the Middle East and North Africa (MENA) has been the weakest among emerging market and developing economy (EMDE) regions, both pre-crisis and post-crisis. It averaged 0.3 percent between 2013-18, although with wide heterogeneity. Weak productivity growth had widened the productivity gap between advanced economies and MENA EMDEs. Large public sectors, underdeveloped private sectors, and lack of economic diversification hold back productivity growth, although recent reform initiatives in many countries in the region are promising. Introduction about 0.3 percent during 2013-18. This slowdown affected about half of EMDEs in the region, especially Labor productivity growth in the Middle East and North energy exporters (Figure 2.4.1.2). Weak post-crisis Africa (MENA) has been the weakest among emerging productivity growth in the region continues a long- market and developing economy (EMDE) regions, standing trend that featured productivity growth below the averaging 0.3 percent during 2013-18 (Figure 2.4.1.1). EMDE average for the past two decades. There is wide heterogeneity across the region in productivity growth, but on average, the productivity gap Within-region heterogeneity. Productivity trends in the between MENA EMDEs and advanced economies has MENA region differ considerably by country. Among widened. In energy exporters, labor productivity growth energy exporters, productivity growth averaged about 0 has been severely constrained by weak investment, while in percent in 2013-18 amid a 50 percent oil price collapse energy importers, it has stagnated below the EMDE from its mid-2014 peak. The oil price collapse also did not average rate. Moreover, the continuing importance of greatly benefit energy importers in the region – commodity exports in many economies means that they productivity growth remained flat at about 1.5 percent have not experienced the diversification or expansion of during both 2003-08 and 2013-18, well below the EMDE other sectors that helped drive high productivity growth in average. regions like East Asia and the Pacific. Wide dispersion in labor productivity levels. At nearly Against this backdrop, this box addresses the following half of advanced-economy productivity, MENA has the questions for the MENA region: highest productivity level of any EMDE region. However, productivity levels in MENA differ widely within region, • How has productivity growth evolved? with substantially higher levels in the Gulf Cooperation Council (GCC) economies than in energy importers. This • What factors have been associated with productivity disparity reflects the variation in natural resource growth? endowments between lower-middle-income energy importers such as Egypt, Morocco, and Tunisia, and high- • What policy options are available to boost income energy exporters such as Saudi Arabia and United productivity growth? Arab Emirates. MENA’s convergence towards advanced Unless otherwise noted, discussion of productivity in this economy productivity levels has decelerated further from box refers to labor productivity, measured as output per the 2003-08 to 2013-18 periods due to weak productivity worker. The primary sample under which regional labor growth. productivity trends are discussed is based on 14 MENA Sources of labor productivity growth. In the two decades economies: Algeria, Bahrain, Egypt, Iran, Iraq, Jordan, prior to the oil price collapse of 2014-16, labor Kuwait, Lebanon, Morocco, Oman, Qatar, Saudi Arabia, productivity growth in the region was primarily supported Tunisia, and the United Arab Emirates. by capital deepening, driven by capital investment by Evolution of regional productivity energy exporters (IMF 2012, 2015; Malik and Masood 2018). In an alternative labor productivity decomposition Low labor productivity growth. From already weak pre- that also incorporates natural resources (Brandt, Schreyer crisis rates (1.3 percent during 2003-08), labor and Zipperer 2017), natural resource activity appears to productivity growth in MENA decelerated further, to drive MENA productivity growth significantly. Its average contribution to productivity growth shrank from about 1.2 percentage points during 2003-08 to 0.2 percentage Note: This box was prepared by Lei Sandy Ye, building upon analysis point during 2013-14, the last year for which natural in Chapter 3. Research assistance was provided by Vanessa Arellano Banoni and Shijie Shi. resources data are available (Figure 2.4.1.2). 118 CHAPTER 2.4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.4.1 Labor productivity in the Middle East and North Africa: Trends and drivers (continued) The commodity sector is capital-intensive. As a result, oil FIGURE 2.4.1.1 Productivity in MENA in prices and capital expenditures are closely linked in the regional comparison MENA region (IMF 2018b; Albino-War et al. 2014). Labor productivity growth in the Middle East and North Foreign direct investment is also highly undiversified and Africa (MENA) has been the weakest among emerging heavily concentrated in the commodity sector (World market and developing economy (EMDE) regions, both Bank 2003). After the global financial crisis, investment pre-crisis and post-crisis. It averaged 0.3 percent between 2013-18. Despite high average productivity growth in the region slowed sharply. Among energy level relative to other EMDE regions, weak productivity exporters, this slower growth has been attributed to tight growth has recently widened its productivity gap with financial constraints associated with lower oil prices. advanced economies. Among energy importers, the legacies of the Arab Spring movements led many economies to increase investment on A. Average productivity growth defense at the expense of infrastructure and other productivity-enhancing projects and initiatives (Baffes et al. 2015; Ianchovichina 2017). Pre-crisis capital deepening was partly offset by contractionary total factor productivity (TFP) growth, the weakness of which has been widely documented for the region over the past three decades.1 The inverse relationship between capital accumulation and TFP growth suggests inefficient investment, and may be attributed to two factors. First, predominantly public investment combined with the large economic role of state-owned enterprises crowds out private investment and job creation. Second, fiscal policy tends to be procyclical— just like public investment—as countries often pursue B. Productivity levels and convergence expansionary fiscal policy during oil price booms (Abdih et al. 2010). During periods of high capital investment and oil price booms, technology-enhancing-oriented reform momentum tends to be weaker, weighing on TFP growth. Negative TFP growth in MENA before the global financial crisis stands in sharp contrast to the robust pre- crisis TFP growth in the broader group of EMDEs. TFP growth started to pick up as oil prices bottomed out in 2016, although it remained low at 1 percent on average during 2016-18.2 Heterogeneity in sources of labor productivity growth. While labor productivity growth in the MENA region as a Source: Penn World Table; The Conference Board; World Bank. whole has long been anemic and continues to be weak, Note: Productivity is defined as labor productivity (real GDP per person there has been wide divergence within the region in its employed). Sample includes 35 advanced economies and 127 EMDEs: 16 in East Asia and the Pacific (EAP), 21 in Eastern Europe and Central Asia (ECA), 25 in Latin America and the Caribbean (LAC), 14 in Middle East and North Africa (MNA), 7 in South Asia (SAR), and 44 in Sub-Saharan Africa (SSA). The 14 MNA economies in the sample are 1 Weak or negative TFP growth is found to be a prevalent feature in Algeria, Bahrain, Egypt, Iran, Iraq, Jordan, Kuwait, Lebanon, Morocco, Oman, Qatar, Saudi Arabia, Tunisia, and the United Arab Emirates. the MENA region during the past three decades. For regional and A. Aggregate growth rates in 2010 U.S. dollars at 2010 prices and exchange country-specific studies that highlight TFP growth in MENA, see Baier, rates. Dwyer, and Tamura (2006); Bisat, El-Erjan, and T. Helbling (1997); B. Rate of convergence is calculated as the difference in productivity growth Callen et al. (2014); IMF (2012); Keller and Nabli (2002); Malik and rates over the log difference in productivity levels between MNA and advanced economies (AE). Blue bars and orange dashes show the range Masood (2018); World Bank (2017d); and Yousef (2004). and average of the six EMDE regional aggregates. “Level” of productivity 2 TFP growth can also be affected by non-technology factors, such as refers to the GDP weighted average of regional productivity as a share of the average advanced economy during 2013-2018. capital and labor utilization. Hence, TFP growth estimates may over- or Click here to download data and charts. understate the true change in the influence of technology on productivity (Dieppe, Kindberg-Hanlon, and Kiliç elik, forthcoming). G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 MID D LE E AS T AN D NO RTH AFRIC A 119 BOX 2.4.1 Labor productivity in the Middle East and North Africa: Trends and drivers (continued) FIGURE 2.4.1.2 Evolution of labor productivity growth in MENA The post-crisis productivity growth slowdown was concentrated in energy exporters and affected about half of the region’s economies. During 2013-18, average productivity growth was around zero percent in energy exporters and about 1.5 percent (still below the EMDE average) in energy importers. Productivity growth has been largely driven by declining capital stock amid weak TFP growth, especially in energy exporters. Productivity levels in energy exporters are much higher than in energy importers. The contribution of natural resources to productivity growth fell significantly from the 2003-08 to 2013-18 periods. A. Productivity growth in MENA B. Share of economies with productivity C. Productivity relative to advanced growth below long-run and pre-crisis economies averages D. Contributions to regional productivity E. Contributions to productivity growth F. Role of natural resources growth Source: Barro and Lee (2015); Haver Analytics; International Monetary Fund; Penn World Tables; United Nations (Human Development Reports), Wittgenstein Centre for Demography and Global Human Capital; World Bank. Note: Productivity is defined as labor productivity (real GDP per person employed). Aggregate growth rates calculated using GDP weights at 2010 prices and market exchange rates. A-C. The sample includes 14 MNA economies: Algeria, Bahrain, Egypt, Iran, Iraq, Jordan, Kuwait, Lebanon, Morocco, Oman, Qatar, Saudi Arabia, Tunisia, and the United Arab Emirates. Includes 127 EMDEs. A. Dashed lines indicate the average long-term labor productivity growth (1981-2018). B. Share of countries for which productivity growth average over 2013-18 is lower compared to a long-run (1992-2018) and pre-crisis (2003-08) average. D.E. MNA Sample in decomposition is the same as in A but excludes Algeria and UAE due to data availability. Includes 92 EMDEs in D. F. The sample includes 10 MNA economies with available data on natural resources capital: Bahrain, Egypt, Jordan, Kuwait, Lebanon, Morocco, Oman, Qatar, Saudi Arabia, and Tunisia. Click here to download data and charts. driving forces. For energy exporters, productivity growth Sources of regional labor productivity growth decelerated markedly from 2003-08 to the post-crisis period of 2013-18 due to sharply declining investment High barriers to factor reallocation. Factor reallocation activity. For energy importers, productivity growth toward more productive activity has played only a limited improved modestly from a weak base, largely due to the role in driving productivity growth in MENA. is muted recovery from negative average TFP growth rates during in uence has re ected high barriers to entry and 2003-08 to slightly above zero percent during 2013-18. distortions such as the lack of competitive markets (Arezki 120 CHAPTER 2.4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.4.1 Labor productivity in the Middle East and North Africa: Trends and drivers (continued) FIGURE 2.4.1.3 Factors supporting productivity growth in MENA Productivity levels relative to advanced economies are the highest in MENA for the capital-intensive industrial sector, while employment is concentrated in the services sector. Evidence for Egypt and Morocco suggests that productivity growth in North Africa has been limited to within-sector productivity gains. A. Sectoral productivity B. Sectoral employment C. Within versus between sector contribution to productivity growth Source: Groningen Growth Development Center Database; Haver Analytics; International Labour Organization; Penn World Tables; World Bank. Note: Productivity is defined as labor productivity (real GDP or value-added per person employed). A.B. Medians across economies in each sector. Includes 12 MENA economies. Panel A based on 2017 data. AE average denotes weighted average across advanced economies. C. The within-sector productivity contribution shows the initial real value added-weighted productivity growth; the between-sector contribution measures the productivity growth from a cross-sectoral shift of employment. Based on nine-sector decomposition. Click here to download data and charts. et al. 2019a). Small exporting rms are hesitant to scale up • Large public sector. On average, about one-fifth of the their operations and bene t little from global value chain region’s workforce is employed in the public sector, integration (World Bank 2016f). For the North Africa and public-private sector wage gaps are among the region, evidence from Egypt and Morocco suggests that highest in the world (Purfield et al. 2018; Tamirisa within-sector productivity gains were the main source of and Duenwald 2018). The education system is productivity growth for their economies (Figure 2.4.1.3). targeted towards government employment, with few In Saudi Arabia, employment appears to have moved high-quality private sector jobs (World Bank 2018l). towards sectors with relatively low productivity in the past These dynamics hold back the adoption of technology (Fayad and Rasmussen 2012). These trends imply from abroad (Mitra et al. 2016; Raggl 2015; distortions in the economy exist that prevent more Samargandi 2018). In the Gulf Cooperation Council, efficient reallocation of resources across sectors. High weak productivity growth has been associated with capital intensity of the commodity sector accounted for low mobility of high-skilled foreign workers (Callen et high average productivity levels in MENA, and scope for al. 2014). productivity improvement in the private sector remains large. Moreover, the majority of employment is • Restrictive business climate. Poor governance quality, concentrated in the services sector, reflecting an large informal sectors, and cumbersome tax policy and exceptionally high proportion of the workforce (about administration hampered the reallocation of resources one-fifth) employed in the public sector (Tamirisa and from low-productivity to higher-productivity firms Duenwald 2018). (Nabli 2007; World Bank 2016f). Non-GCC economies in MENA rank especially low in the World Other drivers of labor productivity growth. Weak Bank’s Worldwide Governance Indicators, such as productivity in the MENA region has been associated with regulatory quality and government effectiveness. underdevelopment of the private sector, overreliance on Private firms often face challenges in access to finance; the public sector, and lack of economic diversification yet, providing access to formal finance is associated (Devarajan and Mottaghi 2015). with labor productivity growth being 2 percentage points higher in MENA firms (Blancher et al. 2019). G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 MID D LE E AS T AN D NO RTH AFRIC A 121 BOX 2.4.1 Labor productivity in the Middle East and North Africa: Trends and drivers (continued) FIGURE 2.4.1.4 Policy challenges Multipronged and sustainable reforms that improve governance and boost private sector development are crucial in MENA. Reforms could lift the potential of its young population and relieve constraints to firm productivity, such as access to finance. A. Access to finance as an obstacle to B. Youth unemployed or not in education C. Governance: Non-GCC economies productivity Source: World Bank. A. Percent of firms citing access to finance as a major obstacle to firm operations. Based on World Bank’s Enterprise Surveys. Latest available survey year for each economy denoted. Non-GCC MENA denotes average of all economies shown in the figure. B. Share of youth not in education, employment, or training, as a percent of youth population. UAE stands for United Arab Emirates. Latest available data since 2015. C. Includes 10 non-GCC economies. Unweighted averages. Based on 2018 data (or latest available year). Index, based on Worldwide Governance Indicators, ranges from -2.5 to 2.5. A lower index denotes worse rating. Click here to download data and charts. • Anemic private sector. Firm productivity in MENA has may have begun to support productivity growth. In the been restricted by low firm turnover and creation. GCC, a series of plans include measures to improve Only six limited liability companies were created productivity and diversify away from the energy sector. annually for every 10,000 working-age people in Efforts to boost small and medium-sized enterprise (SME) MENA during 2009-12—considerably less than in growth and encourage private-sector development include other EMDEs (Schiffbauer et al. 2015). the establishment of an SME agency in Saudi Arabia and SME delicensing in the United Arab Emirates. Among • Lack of diversification. Trade openness and export energy importers, measures to improve the business and diversification in MENA remain low among EMDE private sector climate have been enacted in Egypt, regions. This lack of diversification is partly the result Morocco, and Tunisia (World Bank 2019r). Initial market of exchange rate misalignments associated with high responses to these developments suggest that efficiency reliance on extractive industries or low technological gains have been generated. For instance, Saudi Arabia was content of exports (Benhassine et al. 2009). In the included in the MSCI Emerging Markets Index recently, large EMDEs of the region, low export diversification and many GCC economies established policies to relax has been found to hinder productivity growth.3 foreign investment restrictions (e.g., UAE’s relaxation of Research and development, as measured by the restriction in 13 sectors in 2019). These changes have been number of patent applications per capita, has been associated with foreign investment inflows, which in above the EMDE average. However, it remains well EMDEs often catalyze productivity-enhancing private below advanced-economy averages and has held back investment (Henry 2007). These policies have also made it productivity growth and diversification (Samargandi easier to raise international capital, which has already 2018, Rahmati and Pilehvari 2017). helped finance fiscal and balance-of-payments needs in MENA (IMF 2019d). Egypt’s macroeconomic reforms Recent reforms. A number of large economies in the since 2016 include the liberalization of the exchange rate, region have adopted reform plans in the past five years that business climate reforms, and energy subsidy reforms. These reforms have been positively perceived by investors 3 See IMF (2013, 2015); Morsey, Levy, and Sanchez (2014); and may have raised the country’s export and investment Samargandi (2018). prospects (Youssef et al. 2019). 122 CHAPTER 2.4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.4.1 Labor productivity in the Middle East and North Africa: Trends and drivers (continued) Prospects for labor productivity growth. Recent broad- Raise human capital. The contribution of human capital based reform commitments across the region are to labor productivity growth has been modest in the past promising for labor productivity growth. However, many two decades, amounting to only about half a percentage reforms are subject to high risk of delays in point. The region’s human capital challenge is to improve implementation, especially in non-GCC economies where educational access for youth and women, improve the political fragmentation and budget irresolution have connection between educational attainment and private frequently held back multiyear reform plans. In some non- sector jobs, and to shift its bias in educational training GCC economies, recent protests related to social tensions away from the public sector (World Bank 2018l). These and political developments underscore the fragility measures would help the productivity potential of its large associated with reform progress. Armed conflicts in youth population. More educational programs to improve economies like Yemen continue to challenge the peace that the skills match between workers and employers can these economies need in order to work toward higher enhance the quality of jobs in MENA (Gatti et al. 2013). productivity. Boost firm productivity Policy options Disincentives for innovation and factor reallocation Concerted and multipronged efforts are required to between firms discourages labor productivity in MENA. reliably raise productivity growth. Policies need to be directed at raising the quality of human capital and Improve access to finance. Access to finance is a large boosting private sector investment, increasing firm obstacle for firms in MENA, particularly for non-GCC productivity, removing obstacles to sectoral reallocation, economies, as lack of financing hinders their ability to and creating business-friendly environments. Within these invest and innovate (Figure 2.4.1.4). Better access to broad themes, specific policies need to be tailored to a credit, supported by broader credit bureau coverage and country’s specific circumstances.4 stronger insolvency resolution regimes, appears to yield sizable benefits to productivity growth in MENA The effectiveness of reform in practice is contingent on the (Ghassibe, Appendino, and Mahmoudi 2019). New health of each economy and the timing of political events insolvency resolution laws adopted in Djibouti, Egypt, (Alesina, et al. 2019). Under some circumstances, a Saudi Arabia, and Jordan are promising for facilitating targeted approach that leverages synergies may be debt resolution between creditors and debtors. New warranted. Deep institutional reforms to raise market minority investor protection regulation in Egypt helps contestability, for example, may also bring a variety of improve corporate governance and investor confidence by collateral benefits like higher technological progress requiring shareholder approval in issuing new shares. (Arezki et al. 2019a). Similarly, well-designed deployment of FinTech could help garner broad-based support for Address informality. Informality, although low by average institutional reforms (World Bank 2019r). EMDE standards, presents a challenge to businesses in non-GCC economies. Competition from the informal Improving factors of production sector is a major obstacle for formal sector businesses in several large economies (Morocco, Tunisia), and a higher Boosting private investment. While capital deepening has share of informal workers in SMEs is associated with lower been a main driver of productivity growth in MENA, it wages and more limited export potential (Elbadawi and has been primarily supported by large public spending (for Loayza 2008). Aligning tax systems to international best example, in the commodity sector in the GCC; IMF practices (e.g., harmonized electronic filing systems in 2018b). This suggests large scope to boost private Morocco) and reducing regulatory hurdles for firms can investment. A wide range of reforms is needed to help attract informal firms to more productive formal encourage private investment, including expanding access activity while raising revenue collection. to finance, improving business climates and governance, reducing the wage premium of government employment, Encouraging efficient resource reallocation and leveling the playing field with state-controlled enterprises (Arezki, et al. 2019a). Reallocation towards more productive private sector activities has made limited contributions to productivity 4 Higher labor productivity gains in the region could in turn help growth in MENA. In energy exporters, policies to reduce external imbalances in the region (Arezki et al. 2019b). encourage diversification of exports and output can G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 MID D LE E AS T AN D NO RTH AFRIC A 123 BOX 2.4.1 Labor productivity in the Middle East and North Africa: Trends and drivers (continued) generate new opportunities for labor to move into more taxation systems) and stronger entrepreneurship activities productive private sector opportunities. In energy (e.g., lower cost to start a business). In MENA, reforms importers, such as Egypt and Morocco, expanding that move an economy one unit higher in the Global exporters’ global market reach and improving the quality Competitiveness Index have been estimated to raise of exports could help improve productivity (World Bank productivity growth significantly (Mitra et al. 2016). 2016f). Many MENA economies have adopted broad-based business climate reforms recently, including improved Diversification through trade. Reforms in investment, electricity connection in Bahrain, enhanced electronic tax trade, and tariff policies will help MENA economies move filing in Jordan, and easier property registration in Kuwait. up the export value chain and encourage greater product variety, which currently lags behind international Improve governance. Governance quality in MENA, benchmarks. Regional integration efforts (e.g., Compact especially non-GCC economies, lags behind other EMDEs with Africa) could provide an avenue to promote and has exhibited little improvement over the past decade diversification and raise productivity. (Figure 2.4.1.4). Weak governance has discouraged private Diversification from commodity dependence. For energy sector activity and investment (Nabli 2007). Governance exporters, including the GCC, stronger fiscal management reforms, such as streamlining public service delivery and could help promote diversification by broadening the strengthening legal frameworks in areas like procurement revenue base (Diop and Marotta 2012; World Bank laws can increase productivity growth by encouraging 2019q). For energy importers, options for diversification more efficient allocation of resources. They can also may include investment in renewable energies via public- increase investment prospects through improved investor private partnerships (e.g., Egypt; Vagliasindi 2013), or confidence. Reforms for state-owned enterprises in initiatives to boost the private services sector (e.g., tourism telecom industries can also enhance productivity via higher initiatives in oil importers). Efforts to expand the reach of efficiency (Arezki et al. 2019b). firms to the global market can also help boost productivity Improve gender equality. Women comprise only about growth (World Bank 2016f). one-fifth of the labor force in MENA. Bridging the gender Creating a growth friendly environment gap in a number of areas, including workforce development and access to digital and financial services, is Improve business climates. Business climate reforms, such especially relevant for MENA. Closing these gaps can raise as the reduction of regulatory hurdles to start businesses or productivity growth through more vibrant the removal of particularly distortionary taxes, can help entrepreneurship and private sector participation. boost private investment and productivity. They can also Legislation to reduce economic discrimination against provide firms easier access to critical inputs, such as women in Tunisia is an example of recent reform in this improved electricity supply. They can support productivity area. through better allocation of resources (e.g., more efficient Growth in South Asia is estimated to have decelerated to 4.9 percent in 2019, reflecting a sharper-than- expected and broad-based weakening in domestic demand. In India, activity was constrained by insufficient credit availability, as well as by subdued private consumption. Regional growth is expected to pick up gradually, to 6 percent in 2022, on the assumption of a modest rebound in domestic demand. While growth in Bangladesh is projected to remain above 7 percent through the forecast horizon, growth in Pakistan is projected to languish at 3 percent or less through 2020 as macroeconomic stabilization efforts weigh on activity. Growth in India is projected to decelerate to 5 percent in FY2019/20 amid enduring financial sector issues. Policy measures such as enhancing foreign direct investment inflows and competitiveness, promoting access to finance for small enterprises, and improving infrastructure can deliver productivity gains and lift growth in the region. Key risks to the outlook include a sharper-than-expected slowdown in major economies, a reescalation of regional geopolitical tensions, and a setback in reforms to address impaired balance sheets in the financial and corporate sectors. Recent developments most pronounced in the manufacturing and agriculture sectors, whereas government-related South Asia’s growth is estimated to have services subsectors received signi cant support decelerated to 4.9 percent in 2019, substantially from public spending. GDP growth decelerated to weaker than 7.1 percent in the previous year 5 percent and 4.5 percent (y/y) in the April-June (Figure 2.5.1.A). e deceleration was and July-September quarters of 2019, pronounced in the two largest economies, India respectively—the lowest readings since 2013. and Pakistan. Weak con dence, liquidity issues in Sharp slowdowns in household consumption and the nancial sector (India), and monetary investment o set the rise in government spending tightening (Pakistan) caused a sharp slowdown in (Figure 2.5.1.B). High-frequency data suggest that xed investment and a considerable softening in activity continued to be weak for the rest of 2019 private consumption. Export and import growth (Figure 2.5.1.C). for the region as a whole moderated, in line with a continued slowdown in global trade and industrial In Pakistan, growth decelerated to an estimated activity (World Bank 2019s). Business con dence 3.3 percent in FY2018/19, re ecting a broad- was hampered by subdued consumer demand in based weakening in domestic demand. Signi cant India and security challenges in Sri Lanka. depreciation of the Pakistani rupee (the nominal e ective exchange rate depreciated about 20 Demand faltered amid credit tightening, re ecting percent over the past year) resulted in in ationary structurally high non-performing assets (e.g., pressures (SBP 2019). Monetary policy tightening Bangladesh, India, Pakistan), liquidity shortages in in response to elevated in ation restricted access to the non-bank nancial sector in India, and credit. e government retrenched, curtailing tightening policies in Pakistan. In India, activity public investment, to deal with large twin de cits slowed substantially in 2019, with the deceleration and low international reserves. Bangladesh, the third-largest economy in the Note: This section was prepared by Temel Taskin. Research region, fared better than India and Pakistan, with assistance was provided by Jankeesh Sandhu. growth o cially estimated at 8.1 percent in 126 CHAPTER 2.5 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 2.5.1 SAR: Recent developments FY2018/19. A moderation in domestic demand Regional growth is estimated to have decelerated to 4.9 percent in 2019. In was more than o set by a pickup in exports, partly India, the combination of funding issues in non-banking financial as a result of trade diversion following bilateral companies (NBFC) and uncertainty weighed on growth. Industrial tari increases between China and the United production points to continuing weakness in activity. While regional exports softened in aggregate, Bangladesh’s export growth accelerated, partly States. Bangladesh’s exports showed signs of reflecting trade diversion amid trade tensions between major economies. softening in recent months, after a substantial Monetary policy was broadly accommodative amid weak activity and subdued inflation. Current account deficits narrowed with weakening increase in exports to major trade partners in the imports. last scal year (Figure 2.5.1.D). A. Growth B. Private consumption and Growth in Sri Lanka continued to soften in 2019, investment in India to an estimated 2.7 percent, as tourist arrivals collapsed following terror attacks in April (World Bank 2019t). e Central Bank of Sri Lanka eased its policy stance with cuts in interest rates and reserve requirements in response to subdued economic activity. In Afghanistan, growth recovered to an estimated 2.5 percent in 2019, bene ting from a pickup in agriculture thanks to benign weather conditions. However, political C. Industrial production growth D. Bangladesh: Goods exports growth uncertainty and security challenges weighed on the manufacturing and services sectors. Growth in Nepal is estimated at 7.1 percent in FY2018/19, the third consecutive year of over 6 percent growth. Activity was underpinned by solid remittance in ows, buoyant tourist arrivals, and good monsoons. In Bhutan, activity remained subdued as underlying drivers—hydropower and tourism—have not picked up signi cantly in FY2018/19, resulting in 3.9 percent GDP growth. E. Inflation F. Monetary policy rates While tourist arrivals increased, tourism receipts declined re ecting lower average spending by tourists. Growth in Maldives moderated, despite an increase in tourism, amid softening construction activity, which partly re ected the completion of infrastructure projects and delays in the implementation of new ones. Accordingly, activity is estimated to have expanded by 5.2 percent in 2019. Source: Bangladesh Bureau of Export Promotion; Haver Analytics; World Bank. A. SAR = South Asia region. Aggregate growth rates calculated using constant 2010 U.S. dollar GDP weights. Data for 2019 are estimates. In ation has been mostly stable in the region on B. Last observation is 2019Q3. the back of weak domestic demand and broadly C. Last observation is October 2019. D. Exports data are merchandise exports and in current U.S. dollars. stable currency markets, with the notable E. Bangladesh, India, and Pakistan data reflects fiscal years. 2019 data reflects November for Bangladesh, India, Pakistan, and Sri Lanka. exception of Pakistan (Figure 2.5.1.E). Central F. Data represent monetary policy rates of Reserve Bank of India, State Bank of Pakistan, banks in other major economies were able to cut Bangladesh Bank, and Central Bank of Sri Lanka. Last observation is December 2019. Click here to download data and charts. policy rates several times amid negative output gaps and persistently below-target in ation (India, Sri Lanka; Figure 2.5.1.F). Progress in scal consolidation has broadly weakened. Pakistan’s budget de cit rose more G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 SOUTH ASIA 127 sharply than expected. Contributing factors were a FIGURE 2.5.2 SAR: Outlook and risks shortfall in revenue collection, combined with a Growth is projected to increase gradually, reflecting a modest rebound in sizable increase in interest payments. In Sri Lanka, domestic demand. The regional outlook for 2020 has deteriorated recently, weaker-than-expected tax revenues and increased and risks are tilted to the downside. Financial sector weakness will likely weigh on activity unless balance sheet vulnerabilities are addressed. public spending resulted in a widening budget NBFCs represent a significant share of total loans, and their linkages with de cit. Current account de cits have generally the banking sector imply broad-based contagion risks in India. Lack of progress in reforms to improve tax collection could exacerbate fiscal narrowed over the past year, on the back of deficits. softening imports in the region. A. SAR: Growth contributions B. SAR: Growth forecasts Outlook Growth in South Asia is projected to gradually pick up over the forecast period, from 4.9 percent in 2019 to 6 percent in 2022 (Figure 2.5.2.A; Tables 2.5.1, 2.5.2). is projection assumes a modest rebound in domestic demand. e weak global trade outlook will continue to weigh on regional export growth in the near term. Regional economic activity is expected to bene t from C. Non-performing assets D. India: Non-bank financial system assets, 2018 policy accommodation (India, Sri Lanka), improvement in business con dence and support from infrastructure investments (Afghanistan, Bangladesh, Pakistan). In India, weakness in credit from non-bank nancial companies is expected to linger. Although a gradual growth recovery is expected in the second half of the scal year, the challenges faced by the economy over the rst half should E. Fiscal balances F. Current account balances contribute to a third consecutive year of slowing growth in FY2019/20 (April 2019-March 2020). ereafter, growth is expected to gradually recover, to 6.1 percent in FY2021/22. is forecast is predicated on the monetary policy stance remaining accommodative. It also assumes that stimulative scal and structural measures already taken—including corporate tax cuts, income transfers to farmers, spending on rural development, support measures to the automobile Source: Haver Analytics; Consensus Economics; Reserve Bank of India; World Bank. industry, and further liberalization of foreign A. SAR= South Asia region. Aggregate growth rates are calculated using constant 2010 U.S. dollar GDP-weights. Data for 2019 are estimates. Shaded areas are forecasts. direct investment (FDI)—will begin to pay-o . B. Blue bars represent World Bank forecasts. Last observation is December 2019. C. Last observation is 2019Q2 for Afghanistan, Bhutan, India, Pakistan, and Maldives, and 2019Q1 e scope for more proactive support from scal for Sri Lanka. Bangladesh observation is for 2018. and monetary policies is limited, as in ation has D. Data obtained from RBI (2019) and represent December 2018. E. Shaded areas indicate forecasts. Data for 2019 are estimates. The data refer to fiscal years of recently crossed the midpoint of the target range, countries except for Sri Lanka, as described in Table 2.5.1. F. Data for 2019 are estimates. The data refer to fiscal years of countries except for Sri Lanka, as and weaker-than-expected tax revenues are being described in Table 2.5.1. accompanied by increased public spending. Click here to download data and charts. Macroeconomic adjustment in Pakistan, including a continuation of tight monetary policy and scal consolidation, is expected to continue. Growth is 128 CHAPTER 2.5 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 projected to bottom-out at 2.4 percent in both within region and across major FY2019/20 (July 2019-June 2020). ereafter, as destinations—is much larger (World Bank macroeconomic conditions improve and structural 2019v). Greater participation in global and reforms support investment, growth is projected to regional value chains would lift growth, convey steadily advance, reaching 3.9 percent by positive productivity and technology spillovers, FY2021/22. and narrow current account de cits in the region. Growth in Bangladesh is projected to remain Risks above 7 percent throughout the forecast horizon. A solid macroeconomic framework, political South Asia’s growth outlook has deteriorated stability, implementation of planned public considerably over the past six months. Private infrastructure projects, and ongoing reforms to consumption and investment weakened sharply improve the business environment underlie this amid challenges in the nancial sector, which projection (World Bank 2019u). hampered con dence (Figure 2.5.2.B). Risks to the growth outlook remain tilted to the downside Sri Lanka’s growth is projected to advance to 3.3 and relate primarily to nancial sector percent in 2020. e acceleration afterwards will vulnerabilities, geopolitical tensions, and lack of be supported by recovering investment and progress on reforms. Although recent tensions exports, as long as the security challenges and between India and Pakistan have abated, a political uncertainty of 2019 dissipate. Growth is reescalation would damage con dence and weigh seen to stabilize around 3.7 percent over the rest of on investment in the region. the forecast horizon, in line with potential growth. Non-performing assets in the nancial sector In Afghanistan, activity is expected to continue remain high amid weakening regional growth accelerating, assuming a stable political transition (Figure 2.5.2.C). Further deterioration of balance after elections and a subsequent improvement in sheets of banks and corporates would threaten the business con dence. Nepal’s economy is projected funding of productive investments (Behera and to grow at about 6.5 percent through 2022, Sharma 2019). Failure to close the infrastructure supported by strong services and construction gaps would hold back output and employment sector activities, amid buoyant tourist arrivals and (World Bank 2020). Announced initiatives, such rising public spending. Growth in Bhutan and as the recapitalization and consolidation of public Maldives will continue to be underpinned by sector banks and measures to foster FDI in ows, tourism and infrastructure projects, over the are expected to support activity. Insu cient forecast horizon averaging 6.5 percent and 5.6 progress in implementing these reforms would set percent, respectively. back growth in the region. Over the medium term, regional growth is e non-bank nancial system in India remains expected to rise toward potential. Trends in vulnerable to stress. A major idiosyncratic default urbanization, progress in human capital could trigger a broader liquidity shortage in the accumulation, and demographic developments sector, as it did over the past year (RBI 2019). will support potential growth and productivity. Non-banks represent a signi cant share of total Policy measures such as enhancing FDI in ows loans, and their linkages with the banking sector and competitiveness, promoting access to nance imply that contagion risks are material (Figure for small enterprises, and improving infrastructure 2.5.2.D). can deliver productivity gains in the region (World Bank 2019d; Kochhar et al. 2006; Box Lack of progress in reforms to improve tax 2.5.1). South Asia’s participation in international collection could result in more acute revenue trade remains substantially below that of other shortfalls (Bangladesh, Sri Lanka) and put further regions. While both imports and exports as a share pressure on elevated scal de cits (Pakistan; Figure of GDP in South Asian countries are below levels 2.5.2.E). is could have negative consequences of comparable economies, the gap in exports— for infrastructure investment, and hence for G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 SOUTH ASIA 129 projected growth, as well as for the scal space with strong trade links to these economies available to respond to a future cyclical downturn. (Chapter 1). For countries with elevated debt levels and large current account de cits (Figure With respect to external risks, a sharper-than- 2.5.2.F; Pakistan, Sri Lanka), an unexpected expected slowdown in major external markets tightening in global nancing conditions could such as the United States and the Euro Area, sharply raise borrowing costs and lead to stops in would a ect South Asia through trade, nancial, capital in ows (Sengupta and Gupta 2015). and con dence channels, especially for countries TABLE 2.5.1 South Asia forecast summary Percentage point differences (Real GDP growth at market prices in percent, unless indicated otherwise) from June 2019 projections 2017 2018 2019e 2020f 2021f 2022f 2019e 2020f 2021f EMDE South Asia, GDP 1, 2 6.7 7.1 4.9 5.5 5.9 6.0 -2.0 -1.5 -1.2 (Average including countries with full national accounts and balance of payments data only)3 EMDE South Asia, GDP3 6.7 7.1 4.9 5.5 5.9 6.0 -2.0 -1.5 -1.2 GDP per capita (U.S. dollars) 5.5 5.9 3.7 4.3 4.8 4.8 -2.0 -1.5 -1.1 PPP GDP 6.7 7.1 4.9 5.5 5.9 5.9 -2.0 -1.5 -1.2 Private consumption 6.7 7.9 4.4 5.8 6.3 6.8 -2.6 -1.1 -0.7 Public consumption 12.4 10.3 10.2 7.8 7.6 7.6 2.6 0.9 0.5 Fixed investment 8.5 10.3 4.0 6.4 6.5 6.5 -4.3 -1.4 -1.4 Exports, GNFS4 5.5 10.3 4.9 5.2 5.9 6.0 -0.5 0.0 0.4 Imports, GNFS4 15.5 15.1 3.4 4.8 6.1 6.2 -2.8 -1.0 0.0 Net exports, contribution to growth -2.7 -1.9 0.1 -0.3 -0.5 -0.5 0.8 0.3 0.1 Memo items: GDP2 16/17 17/18 18/19e 19/20f 20/21f 21/22f 18/19e 19/20f 20/21f South Asia excluding India 5.8 6.0 5.6 4.8 4.7 5.1 0.1 0.0 -0.3 India 8.2 7.2 6.8 5.0 5.8 6.1 -0.4 -2.5 -1.7 Pakistan (factor cost) 5.2 5.5 3.3 2.4 3.0 3.9 -0.1 -0.3 -1.0 Bangladesh 7.3 7.9 8.1 7.2 7.3 7.3 0.8 -0.2 0.0 Source: World Bank. Note: e = estimate; f = forecast. EMDE = emerging market and developing economies. World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of countries’ prospects do not differ at any given moment in time. 1. GDP and expenditure components are measured in 2010 prices and market exchange rates. 2. National income and product account data refer to fiscal years (FY) for the South Asian countries, while aggregates are presented in calendar year (CY) terms. The fiscal year runs from July 1 through June 30 in Bangladesh, Bhutan, and Pakistan, from July 16 through July 15 in Nepal, and April 1 through March 31 in India. 3. Subregion aggregate excludes Afghanistan, Bhutan, and Maldives, for which data limitations prevent the forecasting of GDP components. 4. Exports and imports of goods and non-factor services (GNFS). Click here to download data. 130 CHAPTER 2.5 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 TABLE 2.5.2 South Asia country forecasts Percentage point differences (Real GDP growth at market prices in percent, unless indicated otherwise) from June 2019 projections 2017 2018 2019e 2020f 2021f 2022f 2019e 2020f 2021f Calendar year basis 1 Afghanistan 2.7 1.8 2.5 3.0 3.5 3.5 0.1 -0.2 -0.1 Maldives 6.9 6.7 5.2 5.5 5.6 5.6 -0.5 0.3 0.3 Sri Lanka 3.4 3.2 2.7 3.3 3.7 3.7 -0.8 -0.3 0.0 Fiscal year basis1 16/17 17/18 18/19e 19/20f 20/21f 21/22f 18/19e 19/20f 20/21f Bangladesh 7.3 7.9 8.1 7.2 7.3 7.3 0.8 -0.2 0.0 Bhutan 6.3 3.8 3.9 5.6 7.6 6.2 -1.5 0.2 2.4 India 8.2 7.2 6.8 5.0 5.8 6.1 -0.4 -2.5 -1.7 Nepal 8.2 6.7 7.1 6.4 6.5 6.6 0.0 0.0 0.0 Pakistan (factor cost) 5.2 5.5 3.3 2.4 3.0 3.9 -0.1 -0.3 -1.0 Source: World Bank. Note: e = estimate; f = forecast. World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of countries’ prospects do not significantly differ at any given moment in time. 1. Historical data is reported on a market price basis. National income and product account data refer to fiscal years (FY) for the South Asian countries with the exception of Afghanistan, Maldives, and Sri Lanka, which report in calendar year. The fiscal year runs from July 1 through June 30 in Bangladesh, Bhutan, and Pakistan, from July 16 through July 15 in Nepal, and April 1 through March 31 in India. Click here to download data. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 SOUTH ASIA 131 BOX 2.5.1 Labor productivity in South Asia: Trends and drivers In contrast to other emerging market and developing (EMDE) regions, labor productivity growth in South Asia (SAR) has slowed only mildly since the global financial crisis. In 2013-18, SAR productivity growth remained the second fastest (after East Asia and Pacific) among EMDE regions, at 5.3 percent a year. Rapid growth has helped reduce the region’s wide productivity gap with the advanced-economy average. But the level of productivity in SAR remains the lowest among EMDE regions, in part reflecting widespread informal economic activity and struggling manufacturing sectors. Low human capital, poor business environments, inefficient resource allocation, and limited exposure to foreign firms and foreign investment weigh on productivity. Opening up SAR economies by enhancing foreign direct investment inflows and participation in global and regional value chains could support technology and information transfer to the region. Promoting access to finance and improving infrastructure could unlock growth bottlenecks for firms and lift productivity in the region. Introduction because of limited data availability: India and Sri Lanka for growth accounting decompositions, and India, Pakistan, In contrast to other emerging market and developing and Sri Lanka for sectoral analysis using nine-sector data. (EMDE) regions, productivity growth in South Asia (SAR) slowed only mildly after the global financial crisis Evolution of regional productivity from pre-crisis rates, to 5.3 percent a year during 2013-18 (Figure 2.5.1.1.A). This followed a steady rise from anemic Robust productivity growth. Productivity growth in SAR rates in the mid-1980s when heavily state-directed remained robust, at 5.3 percent a year, in 2013-18, only economic policy strategies dampened investment and narrowly below the pre-crisis average of 6.4 percent in innovation. As a result, the region’s convergence towards 2003-08 (Figure 2.5.1.1.C). In the post-crisis period, a advanced-economy productivity levels was the second- slight moderation in India’s productivity growth, and fastest over 2013-18 (after East Asia and the Pacific). larger declines in the smaller economies of Afghanistan, Despite this, the region had the lowest average Bhutan and Sri Lanka, were partially offset by pickups in productivity level of any EMDE region, at 5 percent of the Bangladesh and Pakistan (Figure 2.5.1.1.D). The region’s advanced-economy average in the post-crisis period resilience reflected three main elements: (1) SAR’s limited (Figure 2.5.1.1.B). exposure to external headwinds, (2) continued rapid urbanization, and (3) an improving business environment Against this backdrop, this box will discuss the following that supported productivity gains from the continuing questions about the evolution of productivity growth in shift away from agriculture toward more productive the SAR region: services sectors (World Bank 2016a; APO 2018). As a result, in the post-crisis period, the share of economies • How has productivity evolved in the region? with productivity growth below long-run and pre-crisis • What have been the factors associated with averages weas lower than in other EMDEs (Figure productivity growth? 2.5.1.1.E). • What policy options are available to boost • In India, disruptions to economic activity due to cash productivity growth? shortages in 2016 and transitional costs related to the introduction of the new Goods and Services Tax This box defines productivity as labor productivity, (GST) system in 2017 contributed to a slowing of measured as real GDP per worker at constant (2010) local productivity growth to 5.6 percent a year during 2013 currency prices. Cross-country comparisons of labor -18, from the 2003-08 average of 7.1 percent a year. productivity levels use average 2010 market exchange rates. Nevertheless, India’s post-crisis productivity growth Data for labor productivity at the national level, as well as remained in the highest decile among EMDEs. It was for the three main production sectors (agriculture, supported by investments in the energy and transport manufacturing and mining, and services) are available for sectors, improvements in the ease of doing business, all EMDEs in SAR: Afghanistan, Bangladesh, Bhutan, and ongoing structural reforms. India, Maldives, Nepal, Pakistan, and Sri Lanka. However, the analysis in some cases uses only limited samples • In Pakistan, annual productivity growth picked up from a pre-crisis average of 2.4 percent to 3.1 percent during 2013-18, slightly below the EMDE average of Note: This section was prepared by Temel Taskin, building upon 3.4 percent. During the post-crisis period, analysis in Chapter 3. Research assistance was provided by Jankeesh productivity growth benefited from strong foreign Sandhu and Shijie Shi. 132 CHAPTER 2.5 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.5.1 Labor productivity in South Asia: Trends and drivers (continued) FIGURE 2.5.1.1 Evolution of productivity growth in SAR On average labor productivity expanded by 5.3 percent a year over the last ten years, significantly higher than the EMDE average. The catch-up to advanced economies starts from a low base, as productivity levels in SAR are just one-quarter of levels in the average EMDE. The increasing trend in productivity growth was broad-based in larger economies of the region. However, there is significant dispersion in the level of productivity across the region. A. Productivity growth B. Productivity gap and convergence C. Productivity growth in SAR and EMDEs D. SAR: Productivity growth distribution E. Share of economies with productivity F. Relative productivity levels in SAR growth below long-run and pre-crisis averages, 2013-18 Source: Haver Analytics; Penn World Tables; World Bank. Note: SAR = South Asia region. EMDE = emerging and developing economy. AE = advanced economy. Productivity refers to labor productivity unless otherwise indicated. Sample includes 127 EMDEs and 7 SAR economies unless otherwise indicated. A.B. Range indicates interquartile range of country-level productivity distribution. Rate of convergence calculated as the difference in productivity growth rates with the average advanced economy divided by the log difference in productivity levels with the average advanced economy. C.E.F. Aggregate growth rates calculated using U.S. dollar GDP weights at 2010 prices and exchange rates. D. The year brackets refer to the average growth within the corresponding periods. Click here to download data and charts. direct investment (FDI) inflows and infrastructure with the global trend (Chapter 3). The factors behind projects which supported private sector activity. the slowdown included natural disasters, macro- economic and political instability, and weaker growth • In Bangladesh, post-crisis productivity growth of global trade and manufacturing activity. benefited from improved macroeconomic and political stability which supported both public and SAR’s robust productivity growth through the 2000s is in private fixed investment. As a result, productivity stark contrast to its weakness during the 1980s and 1990s, growth in Bangladesh was robust during 2013-18 at even though in those decades also it was mostly stronger 5.1 percent, slightly above the pre-crisis average of 4.7 than in other EMDEs. In the 1980s, India’s state-directed percent and in the top decile of EMDEs. economy generated minimal productivity growth as heavy regulation and widespread corruption (the “license raj”) • Productivity growth in the rest of the region either stifled manufacturing, investment, and technology stalled or declined in the post-crisis episode in line adoption. In the wake of India’s 1991 balance of payments G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 SOUTH ASIA 133 BOX 2.5.1 Labor productivity in South Asia: Trends and drivers (continued) crisis, major reforms reduced restrictions on product and trade in recent years has weighed further on investment as factor markets and allowed more trade, catalyzing a surge well as exports. The slowdown of investment growth was in productivity growth (Rodrik and Subramanian 2004; from high pre-crisis rates that were fueled partly by large Virmani and Hashim 2011). In Pakistan, productivity foreign direct investment inflows after financial growth was limited by macroeconomic instability (Lopez- liberalization reforms in the 1990s (Fujimori and Sato Calix et al. 2012; Amjad and Awais 2016). 2015; Park 2010). Low productivity levels. Despite its strong growth over the Sources of regional productivity growth past three decades, labor productivity in SAR in 2013-18 was still only 5 percent of the advanced economy average, The slight moderation in SAR’s post-crisis productivity the lowest among EMDE regions and significantly below growth was accounted for mainly by India, and mainly by the EMDE average, which was around 20 percent of the weaker growth in the industrial sector, as in other EMDEs. advanced-economy average. In contrast to other EMDE The median productivity level of the industrial sector in regions, however, the pace of convergence has picked up SAR is just slightly more than one half of the EMDE since the global financial crisis. At the recent rate of median (Figure 2.5.1.2.B). Poor manufacturing convergence, about half of economies in South Asia would productivity in part reflects limited integration into halve their productivity gap with advanced economies over international trade networks and global value chains, the next 40 years. which has constrained the region’s interaction with more productive foreign firms and reduced opportunities to Within-region dispersion in productivity levels. benefit from technology transfer from other countries Productivity differences across countries are very large in (Figure 2.5.1.2.C). This said, post-crisis productivity SAR. Afghanistan and Nepal have the lowest productivity growth in this sector remained higher than the EMDE levels, at around 7 percent of the EMDE average, partly average reflecting improvements in the business reflecting political instability, including prolonged armed environment as well as ongoing public investment in conflict in Afghanistan, and natural disasters. Bhutan, transportation and energy infrastructure. Maldives, and Sri Lanka have higher productivity levels, in the range of 32-85 percent of the EMDE average, Growing productivity gains from between-sector reflecting the benefit of relatively large service sectors, in reallocation. Factor reallocation from low- productivity to particular tourism activity. Productivity levels in the three high-productivity sectors and firms has historically not largest economies of SAR—India, Bangladesh, and been an important source of productivity gains in SAR Pakistan—are lower, ranging between 14 and 27 percent (World Bank 2017e; Mallick 2017; Doughtery et al. 2009; of the EMDE average, reflecting their relatively large Goretti, Kihara, and Salgado 2019). However, since the informal sectors, low urbanization rates, and weak global financial crisis, between-sector reallocation financial development (Figure 2.5.1.1.F). accounted for more than one-third of productivity growth in 2013-15, up from one-tenth in 2003-08. Meanwhile, Slowing contribution from capital deepening. Labor within-sector productivity growth slowed sharply, by more productivity growth can be decomposed into contributions than one-third from pre-crisis rates (Figure 2.5.1.2.D). from increases in other factors of production (human and physical capital) and advances in the effectiveness of their Most of the post-crisis productivity gains from sectoral use (total factor productivity, TFP). Estimates for this reallocation reflected a shift from agriculture, which decomposition are available for India and Sri Lanka. In accounted for about 10 percent of SAR GDP in 2015 but these economies, a slowdown in investment growth almost half of employment, into services, which accounted accounted for all of the post-crisis slowdown in for about half of GDP but roughly one-third of productivity growth, and thus for more than the average employment (Figure 2.5.1.2.E). Agriculture, the region’s contribution of investment to productivity slowdowns in lowest-productivity sector (with average productivity 22 all EMDEs. The contributions to labor productivity percent of the advanced-economy average), has roughly growth of total factor productivity growth and human one-tenth the productivity of financial services (68 percent capital growth remained the same as in the pre-crisis of the advanced-economy average) which is the region’s period (Figure 2.5.1.2.A). The weakening of investment most productive sector. In the post-crisis episode, the growth in part reflected the economic disruptions in India contribution of services sectors to economy-wide around the currency exchange of 2016 and the productivity in SAR has declined while that of agriculture introduction of the GST in 2017. Slower growth of global has increased, as in other EMDEs (Figure 2.5.1.2.F). 134 CHAPTER 2.5 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.5.1 Labor productivity in South Asia: Trends and drivers (continued) FIGURE 2.5.1.2 Sectoral productivity and employment in SAR The gains in productivity of the region are mostly accounted for by improvements in TFP growth and capital deepening. Productivity levels in industry and services sectors are much higher relative to the agriculture sector, and have grown significantly over the past three decades. Progress in within-sector productivity growth has played a much larger role in South Asia relative to other EMDEs. The share of employment in trade and financial services increased over time as workers have shifted away from low-productivity agricultural production to these sectors. A. Productivity decomposition B. Sectoral productivity trends in SAR C. Sectoral productivity levels, 2015 D. Within- and between-sector E. Sectoral employment shares F. Sectoral contribution to productivity contributions to productivity growth Sources: APO productivity database; Expanded African Sector; Groningen Growth Development Center database; ILOSTAT; OECD STAN; United Nations; World KLEMS. Note: SAR = South Asia region, EMDE = emerging and developing economy. Productivity refers to labor productivity unless otherwise indicated. A. SAR sample includes India and Sri Lanka. EMDE sample includes 92 countries. B. The year brackets refer to the average growth within the corresponding periods. SAR sample includes Afghanistan, Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan, and Sri Lanka. EMDE sample includes 127 EMDEs. C.-F. EMDE sample includes 46 countries. SAR sample includes 3 countries: India, Pakistan, Sri Lanka. D. Growth within sector shows the contribution of initial real value added-weighted productivity growth rate of each sector, holding employment shares fixed, and ‘between sector’ effect shows the contribution arising from changes in sectoral employment shares. Median of the county-specific contributions. E.-F. “Other” includes transport services and government services. “Manufacturing” includes mining and utilities; “Finance” includes business services. Click here to download data and charts. Other drivers of productivity. In SAR, the contributions of improvement in several of the long-run determinants of of most of the long-run drivers of productivity to productivity slowed, including average years of schooling, productivity growth have remained low compared to other labor force participation, investment, urbanization and EMDEs and advanced economies despite substantial economic complexity. Nonetheless, improvements in these progress since the early 1990s in a range of these variables drivers did continue. Despite a slowdown in the post-crisis (Figure 2.5.1.3.A). Measures of gender equality and trade episode, investment continued to contribute to openness are below other EMDE regions, as demonstrated productivity growth more than in other EMDEs and by very low female participation rates and weak integration advanced economies (Figure 2.5.1.3.B). By contrast, with global value chains. In the post-crisis period, the pace limited global integration, weakness in control of G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 SOUTH ASIA 135 BOX 2.5.1 Labor productivity in South Asia: Trends and drivers (continued) FIGURE 2.5.1.3 Drivers of productivity growth in SAR Many of the drivers remain at the low end of the EMDE regional range suggesting scope for further improvements. While investment consistently supports economic activity, research and development lag significantly behind other regions. A. Level of drivers of productivity in SAR B. Investment C. Research and development, 2017 Source: Haver Analytics; United Nations; World Bank. Note: EMDE = emerging and developing economy. AE = advanced economy. SAR = South Asia region. A. Unweighted average levels of drivers normalized as average of AEs as 100 and standard deviation is 10. Blue bars represent average within SAR economies in 2018. Orange lines represent range of the average drivers for six regions in 2018. Variables corresponding to the concepts are as follows: Education = years of education, Urbanization = share of population living in urban area, Investment = share of investment to GDP, Institution = WGI Government Effectiveness Index, Econ. Complexity = Economic complexity index, Geography=share of land area which are not in tropical region, Gender equality=female average years of education minus male average years, Demography = share of population under 14, Innovation = Log patent per capita, Trade = (Export+Import)/GDP, Price stability = (-1)*log inflation rate. Numbers of countries are 7 for SAR. See Annex 3.3 for details. B. Investment growth: growth in gross fixed capital formation; investment share: change in gross fixed capital formation as a share of GDP. C. R&D exp: research and development expenditures. Aggregates are calculated using constant 2010 U.S. dollar GDP weights. Click here to download data and charts. corruption, low research and development activity, and • Firm characteristics. Heavy regulatory restrictions pervasive informality continued to weigh on SAR’s have deterred rm growth and prevented rms from productivity growth (Figure 2.5.1.3.C). becoming more productive, including through productivity-improving investment (Cirera and • Limited global integration. Export-oriented rms Cusolito 2019; Kanwar and Sperlich 2019). have been more productive than non-exporters in Complicated tax systems, labor regulations, and SAR (Figure 2.5.1.4.A). However, SAR’s largest licensing requirements have been factors containing economies are less open to trade than the average the productivity of smaller rms (Figure 2.5.1.4.F). EMDE or advanced economy (Figure 2.5.1.4.B). Such factors have encouraged widespread informality, Similarly, while FDI in ows have grown, they remain with the informal sector accounting for roughly one- below the EMDE average (Figure 2.5.1.4.C). SAR’s third of GDP and self-employment accounting for 70 limited contacts with more productive foreign rms percent of total employment in SAR (World Bank reduce the potential for technology and information 2019f). e potential for productivity gains in SAR transfer (Figure 2.5.1.4.D; Maiti 2019; Fujimori and from resource reallocation from less productive to Sato 2015; Topalova and Khandelwal 2011). more productive rms has been estimated to be large • Lack of supporting infrastructure. Many rms cite (Lall, Shalizi and Deichmann 2003).1 infrastructure gaps as important obstacles to their business activities. Firms that cited infrastructure obstacles were found to be less productive in Pakistan and Bangladesh (Grainger and Zhang 2017; 1 For example, equalizing the efficiency of capital and labor allocation Fernandes 2008). e environment has also been less across firms to the level of United States could have increased TFP in supportive in terms of access to nance (Figure India as much as 50 percent in the 1990s (Hsieh and Klenow 2009). Similarly, a one-standard deviation decrease in the misallocation of land 2.5.1.4.E), with state-owned banks dominating and buildings in India was estimated to have improved labor productivity banking system assets (e.g. roughly 70 percent in by 25 percent between 1989 and 2010 (Duranton et al. 2015). Direct India) and their balance sheets encumbered by and indirect contribution of services to the total value added of elevated nonperforming loan ratios (usually around 10 manufacturing sector varies between 33 percent and 50 percent as of 2017 in South Asia (Mercer-Blackman and Ablaza 2018). percent). 136 CHAPTER 2.5 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.5.1 Labor productivity in South Asia: Trends and drivers (continued) FIGURE 2.5.1.4 Policy options in SAR Low trade openness remains a major constraint for productivity growth in SAR. Continued urbanization in the region can bring agglomeration benefits and enhance productivity if it is accompanied with doing-business reforms given that firms in larger cities tend to be more productive. Low FDI inflows to SAR, compared to other EMDEs, hold back positive spillovers from productive foreign firms. Given their low productivity, state banks weigh on financial sector productivity. Small firms face more severe obstacles to access to finance and their TFP is lower than large firms in South Asia. A. Export status, location scale, and TFP B. Trade openness C. FDI inflows in SAR D. Ownership status and TFP in SAR E. Access to finance F. Firm size and TFP in SAR Source: World Bank. Note: Firm-level TFP is computed using a Cobb-Douglas production function, assuming elasticities of output with respect to inputs are the same across countries in a given income group. See Chapter 3 Appendix 3.3 for a detailed description of calculation and sample coverage. SAR = South Asia region. EMDE = emerging and developing economy. AE = advanced economy. A.D.E.F. Calculations are based on World Bank Enterprise Surveys. TFPR = Log Total Factor Productivity based on Revenues. TFPVA = Log Total Factor Productivity based on Value Added. The bars represent estimated coefficients of dummy variables for “exporter”, “located in a city with population larger than 1 million”, “foreign owner”, and “public enterprise” in a regression where dependent variable is log TFP and independent variables are the aforementioned dummy variable (large, exporter, etc.), country dummy variables, and year dummy variables. Survey weights are used in all calculations. Sample includes 15,248 firms in 109 EMDEs, including 20 LICs, for the period 2007-17. B. Trade openness index is described as the ratio of imports and exports to GDP. Aggregates are calculated using constant 2010 U.S. dollar GDP weights. Sample includes 155 EMDEs and 35 AEs. C. FDI = Foreign direct investment. Aggregates are calculated using constant 2010 U.S. dollar GDP weights. Sample includes 155 EMDEs and 35 AEs. E. The vertical axis shows the percentage of responses which indicate “access to finance” as a moderate/major/very severe obstacle. Click here to download data and charts. • Weak human capital. SAR has lagged most EMDE and progress in this area is mixed across the region regions in educational enrolment and attainment, as (Goretti, Kihara, and Salgado 2019). Gender gaps in well as in mortality indicators. In addition, poor workforce participation, education, and financial operations and human resource management quality inclusion restrain the region’s long-term growth has reduced the productivity of firms (Bloom et al. potential (Khera 2018 ). 2012). Robust productivity outlook. Looking ahead, the fact that • Gender gaps. South Asia’s female labor force many of the drivers of productivity have remained at the participation rate is far below comparable economies, low end of the EMDE range indicates scope for substantial G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 SOUTH ASIA 137 BOX 2.5.1 Labor productivity in South Asia: Trends and drivers (continued) FIGURE 2.5.1.5 Productivity prospects in SAR Rising working-age population shares, educational attainment and life expectancy will improve human capital. Increasing urbanization, accompanied by sectoral reallocation, could support productivity in the region. On the other hand, the region is highly vulnerable to natural disasters, environmental deterioration and climate change risks. A. School enrollment projections B. Urbanization projections C. Damage from natural disasters Source: Centre for Research on the Epidemiology of Disasters; United Nations; World Bank. Note: SAR = South Asia Region. EMDE = Emerging and Developing Economy. A.-C. Aggregates are calculated using constant 2010 U.S. dollar GDP weights. A. Last observation is 2018. B. SAR sample includes 8 South Asian countries. EMDE sample includes 159 countries. Last projection year is 2050. C. Simple average of aggregate regional damages per year. Click here to download data and charts. improvements. Increasing rates of school enrolment would regions. Against the backdrop of improving human capital, lift human capital and improve productivity (Figure and continued urbanization, this increase in the labor force 2.5.1.5.A). Low urbanization rates compared to other is expected to contribute to productivity growth in the EMDEs limit the benefits from agglomeration in SAR in years ahead (Annex 3.3). the near term, but longer-term trends may be expected to raise the contribution of urbanization to productivity Policy options growth (Figure 2.5.1.5.B). Recent reforms, such as the new GST system in India and the Inland Revenue Act in Many drivers of productivity are still much lower in SAR Sri Lanka are expected to broaden the tax base and make than in advanced economies and other EMDE regions, resources available for human capital and infrastructure indicating significant room for policy reforms that reduce investments (World Bank 2018m). Business climates have obstacles to faster productivity growth. Such policies need improved significantly in recent years, as shown for to be directed at improving the quality as well as quantity example by shortening approval times for trademarks and of human and physical capital, increasing firm patents, lowering restrictions on foreign direct investment, productivity, encouraging efficient sectoral reallocation, and accelerating investment in energy and transport and creating business-friendly environments. infrastructure (World Bank 2017f). On the other hand, the region is highly vulnerable to natural disasters, and Improving factors of production environmental deterioration and climate change risks Support physical capital accumulation, especially weigh on the productivity growth outlook (Figure infrastructure investment. The post-crisis slowdown in 2.5.1.5.C). An improved productivity outlook will require SAR productivity growth mostly reflected weaker capital the resolution of financial sector issues to unlock credit for accumulation. Many firms cite infrastructure gaps as investment along with further improvements in the ease of important obstacles to their business activities (Figure doing business. 2.5.1.6.A). Moreover, firms facing infrastructure obstacles The working-age share of the population is expected to have been found to be less productive than others in increase in SAR until 2045, providing a larger and more Pakistan and Bangladesh (Grainger and Zhang 2017; prolonged demographic dividend than in most other Fernandes 2008). Improved infrastructure in the energy 138 CHAPTER 2.5 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.5.1 Labor productivity in South Asia: Trends and drivers (continued) and transportation sectors, as well as technology-oriented standards and quality. Similarly, Bangladesh’s duty-free capital accumulation, can promote productivity growth access to the European Union (EU) from 2001 boosted and boost international competitiveness (Calderón, Moral- knitwear exports to the EU between 2000 and 2004, Benito, and Serven 2015). enhanced the productivity of producers, and helped them expand to other export markets (World Bank 2019d). Strengthen investment in human capital. While the region has benefited from raising life expectancy, reducing Improve corporate management practices. Lack of mortality, and expanding access to education over the past information and training on best management practices three decades, there is still significant room for further seems to limit progress in productivity at the firm level. human capital development (Figure 2.5.1.6.B). With the Governments can help improve the quality of increasing working-age share of the population in the management in the region by organizing training region, delivering strong output growth and improvements programs and workshops to disseminate information on in human capital will be key to progress in productivity best management practices. In India, for example, growth (Goretti, Kihara, and Salgado 2019). A better productivity in firms that provided management training educated and healthier workforce can have better and increased by 17 percent in the first year of the intervention more stable jobs and be more productive (World Bank (Bloom et al. 2013). The low number of patents granted 2018a). Policies to expand school attendance and support and the limited number of staff engaged in research and nutrition programs for early childhood development can development in South Asian firms have also been in part boost educational outcomes in SAR (Beteille 2019; Torlese attributed to limited management capacity (Cirera and and Raju 2018; World Bank 2018n). Maloney 2017). Policies that ensure property rights and create technology hubs can increase firm participation in Tackle gender gaps. Addressing constraints on economic product innovation and expand their business in foreign opportunities for women can provide significant gains in markets (Cirera and Maloney 2017). long-term growth (Khera 2018). Key policies such as increasing access to childcare, improving financial Address informality. Self-employment accounts for inclusion, and ensuring public safety and sanitation can around 70 percent of employment in SAR (Figure promote gender equality and boost productivity in SAR 2.5.1.6.C). The level of output informality (DGE and (Sharafudheen 2017; World Bank 2016g). MIMIC) and some obstacles related to business operations are comparable to other EMDEs (Figure 2.5.1.6.D). This Enhancing firm productivity sector is associated with lower productivity and weaker access to finance, a barrier to productive investment and a Increase the region’s integration into the global economy. constraint on firms. Encouraging participation in global SAR’s participation in international trade remains value chains and enhancing a business-friendly regulatory substantially less than that of other regions (Gould, Tan, and tax environment can promote resource reallocation and Emamgholi 2013). While both imports and exports in from less productive informal activities to more productive SAR, relative to GDP, are lower than in comparable formal ones in SAR (Artuc et al. 2019; Amin, Ohnsorge, economies, the gap in exports—both within and outside and Okou 2019). the region—is much larger than that in imports (World Bank 2019v). The empirical evidence on positive With sizable rural populations employed informally in productivity spillovers from international trade and FDI agriculture and large shares of self-employment in the inflows indicates that measures to foster FDI and workforce, productivity in the region could benefit participation in global and regional value chains can lift significantly from improvements in the productivity of the productivity in SAR. SAR may benefit from shifting FDI informal sector. Policies to promote such improvements flows in the context of recent shifts in global could include efforts to by improve labor force skills and manufacturing activity. enhance the functioning of agricultural markets (Goretti, Kihara, and Salgado 2019). Bangladesh’s apparel sector benefited substantially from tailored policies during the 1990s and 2000s, which lifted Promoting efficient sectoral reallocation of barriers to international trade and investment and resources enhanced participation in global value chains. The interaction with foreign firms lifted productivity of local Promote productivity-enhancing sectoral reallocation and suppliers through the demand for inputs with higher improvements in within-sector allocation of resources. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 SOUTH ASIA 139 BOX 2.5.1 Labor productivity in South Asia: Trends and drivers (continued) FIGURE 2.5.1.6 Constraints to productivity growth in SAR Many firms experience obstacles in their operations due to infrastructure gaps and political instability. The region is behind other EMDEs in terms of some doing business indicators, as well as human capital development, limiting opportunities to improve productivity. Financial development is also weaker compared to other EMDEs, which is reflected in low credit to GDP ratios. Many of these obstacles to doing business contribute to the high levels of informality in the region. A. Biggest obstacles in SAR B. Human capital C. Informality D. Obstacles related to regulations E. Financial development F. Doing business, distance to frontier Source: Elgin et al. (2012); United Nations; World Bank. Note: SAR = South Asia region. EMDE = emerging and developing economy. AE = advanced economy. A. Calculations are based on World Bank Enterprise Surveys. Survey weights are used in calculations. Left section represents the responses to “How much of an obstacle?” question in World Bank Enterprise Survey. The vertical axis shows the percentage of responses which indicate moderate/major/very severe obstacle. Right section represents the responses to “What is the biggest obstacle affecting the operations of this establishment?” question. Vertical axis shows the percentage of responses. Others include: Access to land, business licensing and permits, corruption, courts, crime/theft/disorder, customs and trade regulations, inadequately educated workforce, labor regulations, practices of competitors in the informal sector, tax administration, tax rates. B. HCI = Human Capital Index. Range reflects the minimum and maximum of the distribution across countries. Higher values of the index reflect better human capital development. See World Bank (2018a) for details of the methodology. Aggregates are calculated using U.S. dollar GDP weights at 2010 prices and exchange rates. C. DGE = dynamic general equilibrium model. MIMIC = multiple indicators multiple causes model. Both DGE and MIMIC estimates measure the informal output in percent of official GDP. D. Calculations are based on World Bank Enterprise Surveys and represents the responses to “How much of an obstacle?” question. The vertical axis shows the percentage of responses which indicate moderate/major/very severe obstacle. F. SAR sample includes 8 South Asian countries. EMDE sample includes 159 countries. The orange whiskers indicate interquartile range of EMDEs. Click here to download data and charts. SAR has received a welcome boost to productivity from The contribution of within-sector productivity growth has intersectoral reallocation of resources since the global weakened substantially since the global financial crisis. financial crisis. A policy challenge will be to maintain this This calls for a renewed effort to promote the reallocation momentum. The productivity gains from sectoral of capital and labor to more productive firms within reallocation from agriculture to more productive sectors sectors. By one estimate, such interfirm reallocation could can be increased if accompanied by improved local services unlock 40-60 percent productivity gains in India (Hsieh and urban planning (Ellis and Roberts 2016; World Bank and Klenow 2009). Productivity-enhancing interfirm 2019s). Such policies should be complemented by reallocation could be encouraged by policies to foster measures to increase the productivity of the agriculture competition and by reducing regulatory burdens that sector (Cusolito and Maloney 2018). discourage firm growth (Duranton et al. 2016). 140 CHAPTER 2.5 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.5.1 Labor productivity in South Asia: Trends and drivers (continued) Encourage intersectoral linkages. Intersectoral linkages business-friendly EMDE regions, reflected in distance-to- play an important role in improving productivity through frontier scores in doing business statistics (2.5.1.6.F). value chains in South Asia. For instance, progress in India’s economic reforms during the early 1990s enhanced information and communication technologies provides openness and eased regulatory burdens in the services positive productivity spillovers to broader services sectors sector, and these were followed by a significant expansion (Krishna et al. 2016). Reducing barriers to trade and in domestic and foreign investment. In India also, the encouraging intersectoral and regional linkages can lift entry of foreign service providers was associated with more productivity through technology spillovers. For example, competitive business services, which supported in India, Bangladesh, and Sri Lanka the creation of special productivity gains in the manufacturing sector (Arnold et economic zones has helped expand exports and product al. 2016). diversification (Aggarwal, Hoppe, and Walkenhorst 2019). Ensure macroeconomic and political stability. Economic Creating a growth-friendly environment and financial crises have proven to hold back productivity in the region, as observed after the global financial crisis Unlock access to finance. Infrastructure spending in and in economic downturns in India and Pakistan in the recent years has eased supply-side bottlenecks in SAR. 1990s. Political instability seems to be a more severe However, poor access to finance remains a hindrance for obstacle to the operations of South Asian firms than in the region, particularly given the weaknesses on corporate other EMDE regions (World Bank 2013b, 2013c). and financial sector balance sheets. Weak access to finance Strengthening economic policy institutions, improving constrains small and medium-sized firms—especially monetary and fiscal policy frameworks, and enhancing women-owned businesses—and holds back firm-level financial regulation and supervision can help to provide a productivity gains in India (Figure 2.5.1.6.E; World Bank stable macroeconomic framework for firms, reduce 2013a; Schiantarelli and Srivastava 1997). uncertainty, and boost productivity. Improve the ease of doing business. Despite improvements in recent years, SAR is still among the least Growth in Sub-Saharan Africa moderated to a slower-than-expected 2.4 percent in 2019. Activity was dampened by softening external demand, heightened global policy uncertainty, and falling commodity prices. Domestic fragilities in several countries further constrained activity. Growth is projected to firm to 2.9 percent in 2020 and strengthen to 3.2 percent in 2021-22—notably weaker than previous projections. The growth pickup is predicated on improving investor confidence in some large economies, a strengthening cyclical recovery among industrial commodity exporters along with a pickup in oil production, and robust growth among several exporters of agricultural commodities. Nonetheless, these growth rates will be insufficient to make significant progress in reducing poverty in many countries in the region, highlighting the need for lasting improvements in labor productivity to bolster growth over the medium term. Downside risks to the outlook include a sharper- than-expected deceleration in major trading partners; increased investor risk aversion and capital outflows triggered by elevated debt burdens; and growing insecurity. Recent developments unreliable electricity supply constrained manufacturing activity. Some of this weakness The feeble economic recovery in Sub-Saharan was, however, offset by increased oil production. Africa has lost momentum, with growth in 2019 estimated to have edged down to 2.4 percent, In South Africa, growth remained anemic in 2019 from 2.6 percent in 2018. This was a weaker pace as it fell to an estimated 0.4 percent. Weak growth than anticipated in June (Figure 2.6.1.A). momentum has reflected an array of overlapping Intensifying global headwinds such as decelerating constraints. These include persistent policy activity in major trading partners, elevated policy uncertainty, constrained fiscal space, subdued uncertainty, and falling commodity prices, have business confidence, infrastructure bottlenecks— been compounded by domestic fragilities in especially in electricity supply—and weakening several countries. external demand, particularly from the Euro Area and China. In addition, financial stresses at the In Angola, Nigeria, and South Africa—the three public energy utility have worsened the largest economies in the region—growth was government budget balance and raised debt subdued in 2019, remaining well below historical sustainability concerns, weighing further on averages and contracting for a fifth consecutive sentiment (Figure 2.6.1.B). year on a per capita basis. Activity in Nigeria was lackluster, as both macroeconomic policy and the Activity in Angola is estimated to have contracted business environment remain unconducive to by 0.7 percent in 2019, as oil output declined for strong domestic demand. Growth in 2019 is the fourth consecutive year due to lower yields estimated to have remained broadly unchanged at from aging fields and postponed investment in 2 percent, as the agricultural sector continued to new capacity. Nonetheless, growth in the non-oil underperform due to lingering insurgency in the sector strengthened further as several key reforms Northeast and farmers-herdsmen clashes, while continued to improve the business environment. In Sudan, the fourth largest economy in the Note: is section was prepared by Rudi Steinbach. Research region, political instability, alongside an ongoing assistance was provided by Jankeesh Sandhu. currency crisis, has caused activity to contract 142 CHAPTER 2.6 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 2.6.1. SSA: Recent developments sharply. However, the formation of a three-year The recovery in Sub-Saharan Africa has stalled, as intensifying global interim government to oversee the country’s headwinds have compounded domestic weakness in several economies. transition to democracy helped improve stability In South Africa, power cuts and financial stress constrained growth and in the second half of last year. worsened fiscal deficits. More broadly, lower commodity prices are weighing on activity in commodity exporters and contributing to deteriorating current account balances. Inflation has been mostly Beyond the large economies, growth deteriorated subdued—helped in part by lower oil prices. Persistent budget deficits in several industrial commodity exporters in 2019 have partly reflected weaker commodity revenues and growing interest burdens. as weaker prices and softer demand dampened activity in extractives sectors (Democratic A. Growth B. South Africa budget deficit and Republic of Congo, Liberia, Namibia; Figure support for Eskom 2.6.1.C). In contrast, growth accelerated in some countries as investments in new oil and mining capacity boosted activity (Ghana, Guinea, Mauritania). Among exporters of agricultural commodities, growth rates have been more robust, notwithstanding some mild slowdowns. Estimates for 2019 indicate that growth averaged in excess of 5 percent, as sustained public investment in C. Commodity price changes D. Current account balances infrastructure continued to support activity (Togo, Uganda). Yet, growth softened in some other countries as decelerating external demand and lower commodity prices constrained export revenues (Madagascar, Rwanda). In others, agricultural production suffered from severe drought (Senegal, Zimbabwe), or late rains (Kenya). Zimbabwe also suffered a sharp rise in inflation that continued to squeeze real incomes, resulting in a large contraction in economic E. Inflation, annual rate F. Fiscal balances activity, estimated at 7.5 percent. Activity has been further constrained by persistent shortages of food, fuel, electricity, and foreign exchange. Current account deficits are estimated to have widened, on average, across the region (Figure 2.6.1.D). In several countries, capital imports related to large infrastructure projects underpinned deficits (Mauritania, Mozambique, Niger, Uganda). In others, weaker export Source: Haver Analytics; National Treasury, Republic of South Africa; World Economic Outlook, International Monetary Fund; World Bank; World Bank Pink Sheet; Zimbabwe National Statistics. performances, due to softening external demand Note: “Industrial-commodity exporters” represents oil and metal exporting countries. “Other SSA” includes agricultural commodity exporting and commodity importing countries. and lower commodity prices, were responsible for A. Aggregate growth rates calculated using GDP weights at 2010 prices and market exchange rates. “Industrial-commodity exporters” excludes Angola, Nigeria, and South Africa. larger external balances (Angola, Chad, Republic B. Eskom is the South African public energy utility. Years represent fiscal years; for example, the year of Congo). In some countries, current account 2017 is the 2017/18 fiscal year. C. Bars represent the percentage change in the November 2019 monthly price relative to January balances improved as a result of import 2018. “High” and “Low” represent the respective peaks and troughs of price changes, in percent, since January 2018. compression due to weak domestic demand D. Unweighted averages of country groupings. (Namibia, South Africa, Zambia). In others, E. AGO = Angola, GHA = Ghana, ZAF = South Africa, ZMB = Zambia, ETH = Ethiopia, ZWE = Zimbabwe. 2019Q4 reflects the average of October and November. infrastructure improvements and reforms in F. Unweighted averages. Click here to download data and charts. export-oriented industries led to increased exports and an improved trade balance (Burkina Faso, Côte d’Ivoire). Current account financing was G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 SUB-SAHARAN AFRICA 143 more challenging for most of 2019, as growing firming to an average of 0.7 percent in 2021-22. concerns over global growth prospects and In the projection, per capita incomes rise by more heightened trade tensions weighed on investor than 4 percent per year in several countries that, sentiment and capital inflows. Eurobond issuances together, account for one-tenth of the region’s during the first ten months of 2019 were down by poor (e.g., Côte d’Ivoire, Ethiopia, Rwanda, one-third compared to the same period in 2018. Senegal). However, per capita incomes contract among some of the largest economies that account Inflation continued to moderate in most of the for one-third of the region’s poor (Angola, region last year, partly reflecting lower oil prices as Nigeria, Sudan). Projected per capita growth for well as earlier monetary policy tightening in some the region is insufficient to yield significant countries (Figure 2.6.1.E). This allowed progress in poverty alleviation. Lasting authorities in several countries to adopt more improvements in labor productivity are needed to accommodative monetary policy stances (Angola, bolster growth over the medium term (Box 2.6.1; Botswana, Kenya, Mauritius, Mozambique, Figure 2.6.2.B). Namibia, South Africa). In some countries, however, inflation accelerated amid rising food Growth in Nigeria is expected to remain subdued. prices (Ethiopia, Zambia, Zimbabwe) and The macroeconomic framework—characterized by exchange rate pressures (Zambia, Zimbabwe). multiple exchange rates, foreign exchange restrictions, high persistent inflation, and a central Large and persistent budget deficits have reflected bank targeting manifold objectives—does not growing interest burdens, as well as weaker provide a firm anchor for confidence. Growing commodity revenues among industrial-commodity uncertainty about the direction of government exporters and sustained public investment among policies is expected to further dampen the outlook. exporters of agricultural commodities (Figure Growth is projected to remain broadly unchanged, 2.6.1.F). In several countries, budget balances rising only to an average of 2.1 percent in 2020- have improved due to a combination of fiscal 22. This is weaker than previous projections, discipline, more efficient domestic resource reflecting softer external demand, lower oil mobilization, tax administration reforms, and prices, and a slower-than-previously-expected reforms of energy subsidies (Benin, Cabo Verde, improvement in oil production in view of the lack Cote d’Ivoire, Gabon, Mali, Sierra Leone). of much-needed reforms. Outlook In South Africa, growth is expected to firm to 0.9 percent in 2020, before strengthening to an Growth in the region is expected to firm to 2.9 average of 1.4 percent in 2021-22. This assumes percent in 2020, and accelerate further to an that the new administration’s structural reform average of 3.2 percent in 2021-22 (Figure agenda gathers pace, that policy uncertainty 2.6.2.A). The pickup assumes that investor wanes, and that investment—both public and confidence improves in some large economies, private—gradually recovers. The outlook is, that energy bottlenecks ease, that a pickup in oil however, markedly weaker than previous production contributes to a cyclical recovery projections. Increasingly binding infrastructure among industrial commodity exporters, and that constraints—notably in electricity supply—are robust growth continues among exporters of expected to inhibit domestic growth, while export agricultural commodities. However, the forecast momentum will be hindered by weak external for 2020-22 is 0.4 percentage point lower than demand. previously projected, reflecting weaker demand from key trading partners, lower commodity Growth in Angola is projected to rise to 1.5 prices, and adverse domestic developments in percent in 2020 and to average 2.7 percent in several countries. 2021-22. This projection assumes that ongoing structural reforms—supported by prudent On a per capita basis, the outlook translates into monetary policy and fiscal consolidation—provide Sub-Saharan Africa growth of 0.3 percent in 2020, greater macroeconomic stability, continue to 144 CHAPTER 2.6 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 2.6.2 SSA: Outlook and risks improve the business environment and bolster Growth in the region is projected to firm somewhat as investor confidence private investment. In particular, recently in some of the large economies improves and oil production in major oil announced oil-sector reforms are expected to exporters picks up, while activity among exporters of agricultural support a recovery in oil production. commodities remains solid. Per capita growth, however, will remain below 1 percent. Several downside risks could materialize, including slower-than- expected growth in major trading partners, episodes of financial stress Elsewhere in the region, growth is forecast to given rising debt vulnerabilities, and disruptions to activity amid increased strengthen, stabilizing just below 5 percent in displacement of populations and growing climate risks. 2021-22. In the West African Economic and Monetary Union (WAEMU), growth is expected A. GDP growth B. GDP growth per capita to average 6.7 percent. Among the region’s exporters of agricultural commodities, sustained strong public infrastructure spending, combined with increased private sector activity (Madagascar, Rwanda, Uganda), or continued reforms to raise the productivity and competitiveness of export- oriented sectors (Burkina Faso, Côte d’Ivoire), will continue to support output. In Kenya, growth is expected to remain solid, but soften somewhat as accommodative monetary policy does not fully C. Cumulative revisions to 2020 D. Government debt in SSA growth in key trading partners offset the impact of a fiscal tightening. In contrast, the ongoing cyclical recovery among oil and metals exporters will be more sluggish, reflecting weaker external demand and softer commodity prices. In some countries, growth is projected to moderate somewhat over the forecast, in part due to slowing resource production (Democratic Republic of Congo, Ghana). Activity in Ghana—the region’s fifth largest economy—is expected to soften from the 7 percent growth of E. Internally displaced populations F. Extreme weather events in SSA and countries requiring external 2019 partly due to slowing oil production as assistance for food much-needed maintenance on various oil fields is carried out to ensure their long-term viability. Longer-term growth prospects will, however, be supported by the improved strength of the financial sector following much-needed reforms implemented during 2018-19. Despite the global headwinds, investments in new oil and mining capacity are expected to support faster growth in several oil and metals exporters (Botswana, Source: The Emergency Events Database; Université Catholique de Louvain; United Nations Food and Agricultural Organization; United Nations High Commissioner for Refugees (UNHCR); World Cameroon, Chad, Guinea, Mozambique, Bank; World Economic Outlook, International Monetary Fund. A.-B. Aggregate growth rates calculated using GDP weights at 2010 prices and market exchange Namibia). In Sudan, the business climate is rates. “Industrial-commodity exporters” excludes Angola, Nigeria, and South Africa. expected to improve if tensions continue their C. Cumulative revisions to the 2020 growth forecasts since the June 2018 Global Economic Prospects report. recent easing during the 3-year political transition. D. “2020 SSA median” reflects the median of 47 countries. E. “Food insecurity” reflects countries in SSA requiring external assistance for food. These countries are expected to lack the resources to deal with reported critical problems of food insecurity. The sample includes countries with a lack of food availability, widespread lack of access to food, or severe but localized problems. “Displaced population” reflects only internally displaced populations Risks (IDPs) who are protected or assisted by UNHCR. These are also not necessarily representative of the entire IDP population in a given country. F. Data reflect annual averages of extreme weather events in SSA as of October 31, 2019. The balance of risks for Sub-Saharan Africa is Click here to download data and charts. firmly to the downside. A sharper-than-expected deceleration in major trading partners such as G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 SUB-SAHARAN AFRICA 145 China, the Euro Area, or the United States, would Bogoev 2018; Bova et al. 2016). In addition to substantially lower export revenues and raising fiscal sustainability concerns, economic investment. Together these economies account for activity can be directly affected by potential 40 percent of the region’s goods exports and one- disruptions at state-owned enterprises, particularly third of FDI inflows, and their growth prospects if they provide essentials such as electricity. Some continue to be downgraded (Figure 2.6.2.C). countries are, however, implementing reforms to China, in particular, accounts for one-half of improve the functioning of state-owned global metals demand and one-quarter of global enterprises and to alleviate their government’s oil demand (World Bank 2018o). A faster-than- exposure to contingent liabilities (Ethiopia, expected slowdown in China would cause a sharp Ghana, The Gambia). fall in commodity prices and, given Sub-Saharan Africa’s heavy reliance on extractive sectors for Insecurity, conflicts, and insurgencies— export and fiscal revenues, weigh heavily on particularly in the Sahel—would weigh on regional activity. economic activity and food security in several economies (Burkina Faso, Chad, Ethiopia, Mali, Government debt in the region is expected to Niger, Nigeria), if they were to intensify further or reach 62 percent of GDP, on average, in 2020, up spread geographically (Figure 2.6.2.E; FAO 2019; from its trough of 39 percent of GDP in 2011. UNHCR 2019). Moreover, the large populations This broad-based rise in government debt has led that are forcibly displaced by these conflicts cluster to sharp increases in interest burdens, crowding in areas that often become a source of further out non-interest expenditure and raising concerns instability, with poverty rates being worse than in about debt sustainability. Countries with elevated their places of origin (Beegle and Christiaensen debt burdens are susceptible to sudden increases in 2019). investor risk aversion (Angola, Ghana, Mozambique, Namibia, South Africa, Zambia; Extreme weather events are becoming more Figure 2.6.2.D). This can lead to sizable currency frequent as the climate changes, posing a depreciations, capital outflows, and increases in significant downside risk to activity due to the borrowing costs as risk premia rise sharply. Where disproportionate role played by agriculture in debt is largely denominated in foreign currency, many economies in the region (Figure 2.6.2.F). sharp currency depreciations would make The devastation caused by the tropical cyclones servicing debt more challenging. that hit low-income countries in East and Southern Africa in 2019 bear testimony to this, as Ballooning debt burdens of state-owned do persistent drought conditions, particularly in enterprises represent substantial contingent the Sahel and Southern Africa. As droughts liability risks in several countries (Ethiopia, continue to suppress agricultural output, they Ghana, South Africa, The Gambia); increase food insecurity, raise food price inflation, materialization of these risks could damage exacerbate poverty levels, and often contribute to already-fragile fiscal outlooks (Bachmair and forced displacement of populations (IPCC 2019). 146 CHAPTER 2.6 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 TABLE 2.6.1 Sub-Saharan Africa forecast summary Percentage point differences (Real GDP growth at market prices in percent, unless indicated otherwise) from June 2019 projections 2017 2018 2019e 2020f 2021f 2022f 2019e 2020f 2021f EMDE SSA, GDP1 2.7 2.6 2.4 2.9 3.1 3.3 -0.5 -0.4 -0.4 (Average including countries with full national accounts and balance of payments data only) 2 EMDE SSA, GDP2,3 2.7 2.6 2.5 2.9 3.1 3.3 -0.4 -0.4 -0.4 GDP per capita (U.S. dollars) -0.1 -0.1 -0.2 0.3 0.5 0.7 -0.4 -0.4 -0.4 PPP GDP 2.9 2.8 2.7 3.1 3.3 3.5 -0.4 -0.4 -0.4 Private consumption 2.5 2.4 2.1 2.6 2.8 2.8 -0.1 -0.1 0.0 Public consumption 1.2 2.5 2.7 2.4 2.5 2.6 0.1 -0.1 -0.2 Fixed investment 4.5 6.2 3.1 3.1 4.1 5.5 -2.8 -3.0 -2.6 Exports, GNFS 4 6.1 3.1 1.6 1.5 2.6 3.0 -0.7 -1.6 -0.4 Imports, GNFS4 1.0 5.9 2.4 2.5 3.1 3.6 -0.6 -0.9 -0.6 Net exports, contribution to growth 1.5 -0.7 -0.2 -0.3 -0.1 -0.1 0.0 -0.2 0.1 Memo items: GDP SSA excluding Nigeria, South Africa, 4.8 4.4 4.1 4.6 4.7 4.8 -0.5 -0.3 -0.3 and Angola Oil exporters5 1.5 1.4 1.7 2.3 2.3 2.4 -0.4 -0.2 -0.3 CFA countries6 3.5 4.2 4.6 5.1 5.2 5.2 -0.4 0.1 0.1 CEMAC 0.0 1.4 2.3 3.3 3.4 2.9 -0.8 0.2 0.1 WAEMU 6.6 6.6 6.4 6.4 6.5 6.9 -0.2 -0.1 0.0 SSA3 1.0 1.2 1.1 1.5 1.8 2.0 -0.5 -0.5 -0.4 Nigeria 0.8 1.9 2.0 2.1 2.1 2.1 -0.1 -0.1 -0.3 South Africa 1.4 0.8 0.4 0.9 1.3 1.5 -0.7 -0.6 -0.4 Angola -0.1 -1.2 -0.7 1.5 2.4 3.0 -1.7 -1.4 -0.4 Source: World Bank. Note: e = estimate; f = forecast. EMDE = emerging market and developing economies. World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of countries’ prospects do not differ at any given moment in time. 1. GDP and expenditure components are measured in 2010 prices and market exchange rates. Excludes Central African Republic, São Tomé and Príncipe, Somalia, and South Sudan. 2. Subregion aggregate excludes Central African Republic, São Tomé and Príncipe, Somalia, and South Sudan, for which data limitations prevent the forecasting of GDP components. 3. Subregion growth rates may differ from the most recent edition of Africa's Pulse (https://www.worldbank.org/en/region/afr/publication/africas-pulse) due to data revisions and the inclusion of the Central African Republic and São Tomé and Principe in the subregion aggregate of that publication. 4. Exports and imports of goods and non-factor services (GNFS). 5. Includes Angola, Cameroon, Chad, Republic of Congo, Gabon, Ghana, Nigeria, and Sudan. 6. Includes Benin, Burkina Faso, Cameroon, Central African Republic, Chad, Republic of Congo, Côte d’Ivoire, Equatorial Guinea, Gabon, Mali, Niger, Senegal, and Togo. Click here to download data. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 SUB-SAHARAN AFRICA 147 TABLE 2.6.2 Sub-Saharan Africa country forecasts1 Percentage point differences (Real GDP growth at market prices in percent, unless indicated otherwise) from June 2019 projections 2017 2018 2019e 2020f 2021f 2022f 2019e 2020f 2021f Angola -0.1 -1.2 -0.7 1.5 2.4 3.0 -1.7 -1.4 -0.4 Benin 5.8 6.7 6.4 6.7 6.7 6.7 -0.1 0.2 0.2 Botswana 2.9 4.5 4.0 4.1 4.2 4.2 -0.2 0.2 0.2 Burkina Faso 6.3 6.8 6.0 6.0 6.0 6.0 0.0 0.0 0.0 Burundi 0.5 1.6 1.8 2.0 2.1 2.2 0.0 -0.1 0.1 Cabo Verde 3.7 5.1 5.0 5.0 5.0 5.0 0.6 0.4 0.3 Cameroon 3.5 4.1 4.0 4.2 4.3 4.5 -0.2 -0.2 -0.3 Chad -3.0 2.6 3.0 5.5 4.8 4.8 -0.4 -0.1 0.0 Comoros 3.8 3.4 1.7 4.8 3.7 3.6 -1.4 1.6 0.5 Congo, Dem. Rep. 3.7 5.8 4.3 3.9 3.4 3.6 -1.6 -2.6 -3.4 Congo, Rep. -1.8 1.6 2.2 4.6 1.9 2.4 -3.2 3.1 0.0 Côte d’Ivoire 7.7 7.4 7.3 7.0 7.1 7.1 -0.1 -0.3 -0.2 Equatorial Guinea -4.7 -6.1 -4.3 -2.3 1.0 -4.8 -2.1 -0.4 2.8 Eswatini 2.0 2.4 1.3 2.6 2.5 2.5 0.2 1.0 0.8 Ethiopia2 10.0 7.9 9.0 6.3 6.4 7.1 1.1 -1.9 -1.8 Gabon 0.5 0.8 2.9 3.0 3.2 3.3 0.1 -0.7 -0.7 Gambia, The 4.8 6.6 6.0 6.3 5.8 5.5 0.6 1.1 0.8 Ghana 8.1 6.3 7.0 6.8 5.2 4.6 -0.6 -0.2 -0.6 Guinea 10.0 5.8 5.9 6.0 6.0 6.0 0.0 0.0 0.0 Guinea-Bissau 5.9 3.8 4.6 4.9 5.0 5.0 0.3 0.1 -0.5 Kenya 4.9 6.3 5.8 6.0 5.8 5.8 0.1 0.1 -0.2 Lesotho -0.4 1.5 2.6 0.7 2.1 2.8 1.1 0.3 -2.0 Liberia 2.5 1.2 -1.4 1.4 3.4 4.2 -1.8 -0.2 2.1 Madagascar 4.3 5.1 4.7 5.3 4.4 5.0 -0.5 0.0 -0.7 Malawi 4.0 3.5 4.4 4.8 5.2 5.3 -0.1 0.1 0.1 Mali 5.3 4.7 5.0 5.0 4.9 4.9 0.0 0.1 0.1 Mauritania 3.0 3.6 6.4 5.7 5.8 8.7 -0.3 -0.1 -0.2 Mauritius 3.8 3.8 3.9 3.9 4.0 4.0 0.0 0.0 0.5 Mozambique 3.7 3.4 2.0 3.7 4.2 4.4 0.0 0.2 0.0 Namibia -0.9 -0.1 -0.5 0.9 1.7 1.9 -1.4 -0.6 -0.2 Niger 4.9 6.5 6.3 6.0 5.6 11.9 -0.2 0.0 0.0 Nigeria 0.8 1.9 2.0 2.1 2.1 2.1 -0.1 -0.1 -0.3 Rwanda 6.1 8.6 8.5 8.1 8.0 8.0 0.7 0.1 0.5 Senegal 7.1 6.8 6.3 6.8 7.0 7.0 -0.5 -0.2 0.0 Seychelles 4.3 4.1 3.5 3.3 3.3 3.4 0.1 0.3 0.1 Sierra Leone 3.8 3.5 4.8 4.9 4.9 5.0 -0.6 -0.5 -0.3 South Africa 1.4 0.8 0.4 0.9 1.3 1.5 -0.7 -0.6 -0.4 Sudan 4.3 -2.3 -2.6 -1.4 -0.6 0.2 -0.7 -0.1 0.2 Tanzania 6.8 5.4 5.6 5.8 6.1 6.2 0.2 0.1 0.0 Togo 4.4 4.9 5.3 5.5 5.5 5.5 0.3 0.3 0.4 Uganda2 3.9 5.9 6.1 6.5 5.9 6.0 0.0 0.0 0.1 Zambia 4.1 3.1 1.8 2.6 2.6 4.0 -0.7 -0.2 -0.2 Zimbabwe 4.7 3.5 -7.5 2.7 2.5 2.8 -4.4 -0.8 -2.4 Source: World Bank. Note: e = estimate; f = forecast. World Bank forecasts are frequently updated based on new information and changing (global) circumstances. Consequently, projections presented here may differ from those contained in other Bank documents, even if basic assessments of countries’ prospects do not significantly differ at any given moment in time. 1. GDP and expenditure components are measured in 2010 prices and market exchange rates. Excludes Central African Republic, São Tomé and Príncipe, Somalia, and South Sudan. 2. Fiscal-year based numbers. Click here to download data. 148 CHAPTER 2.6 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.6.1 Labor productivity in Sub-Saharan Africa: Trends and drivers Since 2013, Sub-Saharan Africa has experienced a broad-based slowdown in labor productivity growth. Productivity growth has all but stalled amid falling commodity prices, weakening external demand, and growing domestic fragilities. In the decade prior to the global financial crisis, productivity growth benefited from strengthening institutions, stronger investment, infrastructure development, improving human capital, and better macroeconomic policy frameworks, but the pace of improvement has stagnated. Productivity in the region is still only one-half of that in EMDEs and roughly one-tenth of that in advanced economies. Ambitious policy efforts will be needed to generate the productivity growth required for per capita incomes in Sub-Saharan Africa to reach those of its EMDE peers, let alone those of advanced economies. To stimulate labor productivity growth, the region needs to implement policies that boost agricultural productivity, increase resilience to climate change, broaden economic diversification, and continue human capital development. Introduction stagnation offers dim prospects for the nearly 60 percent of the global extreme poor that currently reside in SSA. In one of the steepest declines of any emerging market and developing economy (EMDE) region, labor productivity Against this backdrop, this box addresses the following growth has slowed sharply in Sub-Saharan Africa (SSA) questions: since the global financial crisis, from about 2.9 percent during the pre-crisis period of 2003-2008 to 0.5 percent 1. How has productivity evolved in the region? during 2013-18 (Figure 2.6.1.1.A). The slowdown was particularly sharp among industrial commodity 2. What are the factors associated with productivity exporters—exporters of oil and metals account for roughly growth in the region? 80 percent of the region’s GDP—whereas productivity growth continued to accelerate among several agricultural 3. What policy options are available to boost commodity exporters.1 This deceleration returns productivity growth? productivity growth to near its 1990s average (-0.4 This box defines productivity as labor productivity, percent) and ends a period of solid growth of 2-3 percent represented by real GDP per person employed (at 2010 throughout the pre-crisis period, when it was supported by prices and exchange rates). Growth in labor productivity is a favorable external environment, strengthening decomposed into the contributions made by changes in institutions, improving human capital, and better the standard factor inputs (human and physical capital per macroeconomic policy frameworks. worker) and the effective use of these inputs, as captured SSA’s productivity levels are low, at around one-half of the by total factor productivity, assuming a Cobb-Douglas EMDE average and 11 percent of the advanced-economy production function. Cross-country comparisons of labor average in 2018 (Figure 2.6.1.1.B). However, if a few high productivity use market exchange rates in 2010 to convert -productivity countries are excluded, SSA’s productivity national currency units into U.S. dollars. Data are levels are far lower, at a mere 3 percent of the advanced- available for 44 EMDEs in SSA, of which 21 are oil or economy average. At near-nil productivity growth, SSA’s metals exporters, 19 are exporters of agricultural productivity levels have now started to further diverge commodities, and 5 are commodity importers.2 from advanced-economy averages. Among EMDE regions, only the Middle East and North Africa has a slower pace Evolution of regional productivity of convergence, but starting from productivity levels that Robust pre-crisis productivity growth. Productivity average about four times those of Sub-Saharan Africa. growth in SSA started improving in the mid-1990s, as the Absent major policy efforts to lift productivity growth, its region recovered from some of the adverse factors that had weighed heavily on activity in the 1980s and early 1990s.3 Note: This box was prepared by Rudi Steinbach, with contributions Prior to the crisis, productivity growth rose sharply, to 2.9 from Sinem Kilic Celik, and builds upon analysis in Chapter 3. Research assistance was provided by Jankeesh Sandhu and Shijie Shi. 1 An economy is defined as a commodity exporter when, on average 2 One country, Chad, is classified as both an oil and an agricultural- in 2012-14, either (1) total commodities exports accounted for 30 commodity exporter. percent or more of total goods exports or (2) exports of any single 3 Adverse developments in the 1980s and early 1990s included a commodity accounted for 20 percent or more of total goods exports. multitude of sovereign debt, banking, and currency crises, debt overhang, Economies for which these thresholds are met as a result of reexports are low commodity prices, weak investment, and severe conflicts and political excluded. Commodity importers are economies not classified as instability in several countries (Calderón and Boreux 2016; Reinhart and commodity exporters. Rogoff 2009; Straus 2012). G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 SUB-SAHARAN AFRICA 149 BOX 2.6.1 Labor productivity in Sub-Saharan Africa: Trends and drivers (continued) percent, on average during 2003-08. Growth was supported by a favorable external environment, including a FIGURE 2.6.1.1 Productivity in SSA in commodity price boom between 2001-11 that fueled an regional comparison inflow of foreign capital and unprecedented investment Productivity growth in Sub-Saharan Africa (SSA) rose and benefited many of the region’s low-income countries sharply in the pre-crisis period, reflecting a favorable (Figure 2.6.1.2.A; Khan et al. 2016; Steinbach 2019; external environment and improvements in key drivers of World Bank 2019a). Faster productivity growth was also productivity. Stronger productivity growth also allowed a supported by improvements in education, health care, large productivity gap between advanced economies and SSA EMDEs to narrow slightly over this period. infrastructure, financial access, and trade openness Since then, productivity growth in the region has slowed (Calderón and Servén 2010; Cole and Neumayer 2006; sharply. At near-zero productivity growth, the region’s Shiferaw et al. 2015; World Bank 2018k, 2019z). productivity levels have, on average, diverged from advanced economy levels during the post-crisis period. In the 2000s, productivity growth in the region’s industrial commodity-exporting countries picked up sooner and A. Productivity growth more sharply than in agricultural commodity exporters and commodity importers. In addition to the higher export revenues brought about by rising commodity prices, oil and metal exporting countries benefited from substantial investments in commodity production and exploration (Khan et al. 2016; Schodde 2013). The productivity growth pick-up in industrial-commodity exporters was also driven by country-specific developments. In South Africa—the region’s largest metal exporter—productivity growth accelerated sharply after the country’s transition to democracy in 1994, thanks in part to improving policy frameworks, increased trade openness and foreign capital inflows (Arora 2005; Du Plessis and Smit 2007). By the mid-2000s, the more than 20 percent decline in productivity during the final decade B. Productivity gap and convergence of Apartheid had been fully reversed. Stalling post-crisis productivity. Since the global financial crisis, productivity growth has fallen sharply in SSA, to near-nil (0.5 percent) on average during the post-crisis period (2013-18). Productivity growth slowed in a broad range of economies, with post-crisis productivity growth falling below its pre-crisis average in over 60 percent of countries. Oil- and metal-exporting countries experienced the steepest slowdowns amid the commodity price slump of 2014-16, as productivity growth fell to 0 percent in the post-crisis period, from 3.2 percent growth pre-crisis. Post-crisis productivity growth in agricultural commodity- Source: Penn World Table; The Conference Board; World Development exporters and commodity importers was more resilient, Indicators, World Bank. particularly among the former for whom it strengthened to Note: Unless specified otherwise, productivity is defined as labor productivity, (real GDP per person employed). 2.3 percent. Despite the sharp fall in agricultural A. Sample includes range and simple average for the 127 EMDEs and simple commodity prices during the commodity price slump— average for 44 Sub-Saharan Africa countries. B. Sample includes 35 advanced economies (AE) and 127 EMDEs. Rate of albeit less severe than the drop in industrial commodity convergence is calculated as the difference in productivity growth rates over prices—sustained productivity growth was supported by the log difference in productivity levels between SSA and advanced economies. Blue bars and orange dashes show the range and average of the improving macroeconomic policy frameworks, investment six EMDE regional aggregates. “Level” of productivity refers to the GDP- weighted average of regional productivity as a share of the average in infrastructure, and continuous efforts to improve advanced economy during 2013-2018. business environments. Doing Business rankings improved Click here to download data and charts. 150 CHAPTER 2.6 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.6.1 Labor productivity in Sub-Saharan Africa: Trends and drivers (continued) FIGURE 2.6.1.2 Evolution of labor productivity growth in SSA The sharp slowdown in SSA’s productivity growth relative to the pre-crisis period is concentrated among exporters of industrial commodities, in part reflecting the commodity-price slump of 2014-16. Excluding five high-productivity countries, productivity levels in the region are, on average, 3 percent that of advanced economies. Rapid productivity growth between the 1990s and 2008 reflected improvements in human capital, the deepening of physical capital, as well as a rise in total factor productivity (TFP). Following the commodity price slump, TFP slowed sharply among industrial-commodity exporters. Among exporters of agricultural commodities, capital deepening has reflected continued investment in infrastructure. TFP has contracted in recent years, mostly among industrial-commodity exporters. However, the fall in TFP was likely less severe when the contribution from slowing extraction of natural capital is accounted for. A. SSA and EMDE labor productivity B. Productivity relative to advanced C. Contributions to productivity growth growth economies D. Contributions to productivity growth, E. Contributions to productivity growth F. Contribution to productivity growth, by by export composition in Nigeria natural capital Source: Penn World Table; Wealth Accounting, World Bank. Note: Unless specified otherwise, productivity is defined as labor productivity (real GDP per person employed). A. Aggregate growth rates calculated using GDP weights at 2010 prices and market exchange rates. Dashed lines indicate average long-term labor productivity growth (1981-2018 for SSA; 1990-2018 for EMDEs excl. China). Samples include 44 Sub-Saharan African economies and 126 EMDEs. “Other SSA” includes agriculture exporters and commodity importers. B. GDP-weighted averages calculated using GDP weights at 2010 prices and market exchange rates. Sample includes 127 EMDEs and 44 Sub-Saharan African economies. “SSA high productivity” includes Equatorial Guinea, Gabon, Mauritius, Seychelles, and South Africa. C.-F. Aggregate growth rates calculated using GDP weights at 2010 prices and market exchange rates. Samples include 26 Sub-Saharan African economies and 92 EMDEs. D. “Industrial-commodity exporters” includes metals and oil exporters. “Other SSA” includes agricultural commodity exporters and commodity importers. F. For comparability, the sample for both the natural and standard decomposition includes 22 countries. Click here to download data and charts. by three positions in the median agricultural commodity- the business environment, to attract private investment exporter between the pre- and post-crisis periods, (World Bank 2019w). In 2018, the country led SSA in its compared to a median deterioration of seven positions ease of doing business, ranking 29th globally. In Côte among industrial commodity exporters. Several country- d’Ivoire, a return to stability following the end of decade- specific reasons also helped lift productivity among long civil strife in 2011 has since enabled a sharp rise in agricultural commodity exporters. In Rwanda, productivity, amid increased public investment, recovering productivity growth was boosted by continued reforms to foreign direct investment (FDI) inflows, an improving strengthen institutions and governance, upgrade business environment and rising export activity (Klapper, infrastructure, increase access to education, and improve Richmond, and Tran 2013; World Bank 2015c). G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 SUB-SAHARAN AFRICA 151 BOX 2.6.1 Labor productivity in Sub-Saharan Africa: Trends and drivers (continued) Low productivity levels. Productivity in SSA is the in investment, FDI inflows, and exports, compounded by second-lowest of all EMDE regions, after South Asia. somewhat weaker business environments.5 In Liberia and However, if the five most productive economies are Sierra Leone, the post-crisis fall in TFP was exacerbated by excluded (Equatorial Guinea, Gabon, Mauritius, the devastating Ebola outbreak of 2014-16 (World Bank Seychelles, South Africa), SSA has the lowest productivity 2019x). of all EMDE regions, at 3 percent of the advanced- In contrast, TFP has remained resilient, or even economy average in 2018 (Figure 2.6.1.2.B). Higher strengthened, among some exporters of agricultural productivity levels in these five economies—at 24 percent commodities and commodity importers (Côte d’Ivoire, of the advanced-economy average—is roughly one-quarter Kenya, Mauritius, Togo). Agricultural commodity prices above the EMDE average. It exceeds productivity in other fell less steeply, on average, than industrial commodity SSA economies, in part due to significant oil wealth prices during the 2011-16 commodity price slump, and (Equatorial Guinea, Gabon), dominant tourism sectors in beneficial terms of trade supported activity among island states (Mauritius, Seychelles), and a considerably commodity importers. Faster TFP growth in these higher capital stock combined with mineral wealth (South economies was also underpinned by sustained public Africa). The post-crisis slowdown in productivity growth investment in infrastructure, continued efforts to improve has dimmed prospects for SSA’s continued convergence business environments, and more robust macroeconomic with advanced economies and other EMDEs. If recent policy frameworks. rates of productivity growth persist, less than 5 percent of economies in SSA are on course to halve their productivity Post-crisis acceleration of capital deepening. The gap with advanced economies over the next 40 years. contraction in TFP growth offset the post-crisis boost to Post-crisis total factor productivity decline. The post- productivity growth generated from capital deepening. crisis slowdown in SSA’s productivity growth reflected less Labor productivity in agricultural commodity exporters effective use of factor inputs, as captured by total factor benefited from heavy public investment.6 In Nigeria, productivity (TFP; Figure 2.6.1.2.C).4 TFP growth, which investment was fueled by large FDI inflows into the accounted for the majority (three-fifths) of productivity energy, banking, manufacturing, and telecommunications growth pre-crisis, plunged from 1.4 percent pre-crisis to sectors (although investment slowed sharply after 2014 as -0.9 percent post-crisis in the sharpest deterioration of any oil prices collapsed; Figure 2.6.1.2.E; World Bank 2019y). EMDE region. Rapid pre-crisis TFP growth, especially in In contrast, investment has fallen sharply in other industrial commodity exporters, reflected heavy resource industrial commodity exporters in SSA—by 7 percentage investment and exploration during the commodity boom, points of GDP in the median economy—following the large FDI inflows, communication infrastructure 2014-16 commodity price slump, compounding the improvements (including the increased use of mobile already slowing TFP growth. phones), expanded access to finance, and better business Impact of natural resource extraction on productivity climates (Figure 2.6.1.2.D; Aker and Mbiti 2010; measurement. Natural capital accounts for an economy’s Goedhuys, Janz, and Mohnen 2008; Keefer and Knack natural resources, such as oil, metals, and agricultural land, 2007; Wamboye, Tochkov, and Sergi 2015). The sharp and is particularly relevant given SSA’s commodity post-crisis decline in TFP was most pronounced in reliance. Standard productivity decompositions fold the industrial commodity exporters, following the commodity extraction of natural capital into total factor productivity price collapse of 2014-16 and the accompanying collapse and, to a lesser extent, physical capital, biasing their estimated contributions to productivity growth (Brandt, 4 The standard productivity growth decomposition does not explicitly Schreyer and Zipperer 2017; Calderón and Cantu 2019; account for the contribution of natural capital as a factor of production. World Bank 2019z). During the pre-crisis commodity As a result, the TFP estimates produced here are potentially biased as they implicitly include the productivity contribution from natural capital. From a longer-term perspective, World Bank (2019z) finds that the 5 TFP declines have been most severe in oil-exporting Angola, significant difference between productivity in SSA and that of the productivity frontier (United States) largely reflected weak factor Nigeria, and Chad, as well as in metal-exporting countries such as accumulation between 1960 and the 1990s, as the index of human capital Botswana, Mozambique, Sierra Leone and South Africa. in SSA relative to that of the United States declined sharply from 1960 to 6 Greater fiscal space, partly due to the Multilateral Debt Relief 1980, while the relative accumulation of physical capital remained Initiative (MDRI) and Heavily-Indebted Poor Countries (HIPC) subdued. In contrast, from 2000, the gap in efficiency (or TFP) became initiative, supported increased investment in infrastructure and human the major contributor to difference in productivity between SSA and the capital which resulted in an 18-percentage-point rise in average secondary frontier. This TFP gap widened further from 2010 onwards. school enrollment rates from 33 percent in 2000 to 51 percent in 2014. 152 CHAPTER 2.6 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.6.1 Labor productivity in Sub-Saharan Africa: Trends and drivers (continued) FIGURE 2.6.1.3 Sectoral productivity growth in SSA The sectoral reallocation of labor in Sub-Saharan Africa has been an important driver of regional productivity growth; however, its contribution has dwindled more recently. Agriculture in SSA has the lowest productivity, while productivity is highest in mining and finance. Low aggregate productivity in the region is partly explained by the agricultural sector’s significant contribution to value added, combined with the disproportionate share of employment devoted to the sector. A. Within-sector and structural B. Sectoral productivity, 2015 C. Employment by sector contributions to productivity growth Source: APO productivity database; de Vries, Timmer, and de Vries 2013; Expanded Africa Sector Database; Groningen Growth Development Center database; Haver Analytics; ILOSTAT; Mensah and Szirmai (2018); Mensah et al. (2018); OECD STAN; United Nations; World Bank; World KLEMS. Note: Unless specified otherwise, productivity is defined as labor productivity (real GDP per person employed). A. Growth within sector shows the contribution of initial real value added-weighted productivity growth rate of each sector and ‘between sector’ effect shows the contribution arising from changes in sectoral employment shares. Median of the county-specific contributions. Sample includes 19 Sub-Saharan African economies and 46 EMDEs. B. Figure shows the median of country groups. The sample includes 19 SSA economies and 46 EMDEs. C. Sample includes 19 SSA countries and 46 EMDEs. Click here to download data and charts. price boom and the accompanying boom in resource post-crisis slowdown in productivity growth from pre- exploration and development, the increased extraction of crisis rates reflects slowing gains brought by the natural capital lifted productivity growth in SSA (Figure reallocation of labor from low-productivity sectors (mostly 2.6.1.2.F; Khan et al. 2016). However, as the boom ended agriculture) to higher-productivity sectors. In contrast, and commodity prices began to fall, natural capital within-sector productivity growth has continued apace extraction declined accordingly, and its contribution (Figure 2.6.1.3.A).8 detracted from overall productivity growth. Data for natural capital is available until 2014, the year the Productivity has differed widely across sectors in SSA commodity price slide intensified, but well before prices (Figure 2.6.1.3.B). Productivity in agriculture—the least reached their early-2016 troughs. Even during these early productive sector that employs more than half of the years (2013-14), it appears that the post-crisis fall in TFP workforce and accounts for 18 percent of GDP—is was likely less severe than the standard decomposition between 4 and 7 percent of the productivity in mining and suggests, as the decline in natural capital potentially finance, the two most productive sectors at the nine-sector accounted for a large share of the slowdown in TFP level (Figure 2.6.1.3.C).9 Relative to the wider EMDE growth from pre-crisis years.7 sample, agricultural productivity in SSA is about three times lower, on average. Low agricultural productivity in Sources of regional productivity growth SSA reflects the prevalence of subsistence farming, sub- optimal crop selection, poor land quality amid unfavorable Productivity growth through sectoral reallocation. The climates, limited uptake of modern technologies and production methods to improve yields, and small farm 7 Direct comparisons between the standard decomposition and that sizes (Adamopoulos and Restuccia 2014, 2018; Caselli including natural capital are complicated by the smaller country sample 2005; Sinha and Xi 2018). Moreover, the use of price in the natural capital decomposition, as it includes 22 countries (72 percent of SSA GDP) compared to 26 countries (83 percent of SSA 8 Sectoral productivity data are available for only about half the SSA GDP) in the standard decomposition. Furthermore, the decline in natural capital may capture a lower valuation of the stock of natural economies with data for aggregate productivity. 9 The sample includes 19 SSA economies at the nine-sector level. capital. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 SUB-SAHARAN AFRICA 153 BOX 2.6.1 Labor productivity in Sub-Saharan Africa: Trends and drivers (continued) controls—a widespread practice across particularly low- caused by price controls, have not only constrained income countries in the region—often distort the productivity by distorting the efficient allocation of allocation of resources and inputs in agricultural sectors resources, but have also deterred private sector investment and weigh further on productivity by adversely affecting (Cirera, Fattal Jaef, and Maemir 2017; World Bank incentives to invest in human capital or adopt modern 2019z). technologies and production methods (Special Focus 1; Chen 2017; Chen and Restuccia 2018; World Bank Integration with the global economy. Between the mid- 2019z). The agricultural sector’s significant contribution 1990s and 2008, the region’s openness to trade—that is, to value added, combined with the disproportionate share the sum of imports and exports relative to the size of the of employment devoted to the sector, helps explain SSA’s economy—rose 16 percentage points to 81 percent of low aggregate productivity relative to other EMDE GDP, helping to boost productivity. However, alongside regions. falling commodity prices and slowing external demand, particularly from China and the Euro Area (the region’s Pre-crisis, sectoral reallocation accounted for more than two largest trading partners), trade integration has partially half of aggregate productivity growth as labor moved from unwound in the post-crisis period, with openness falling to agriculture to services sectors and, to a lesser extent, 74 percent of GDP by 2017. The region’s heavy manufacturing (Chapter 3; Enache, Ghani, and O’Connell dependence on commodity extraction sectors manifests in 2016; Haile 2018; Rodrik 2016b). This process was a smaller share of exporting firms compared to the EMDE facilitated by rapid urbanization as the urban share of average (Figure 2.6.1.4.E). Although the share of foreign- population rose by 5 percentage points, to 39 percent, owned firms—which are generally more productive than between 2000 and 2010. Since the crisis, however, the their domestically owned counterparts—is high, such firms sectoral reallocation of labor to more productive sectors tend to cluster in extractives sectors with limited links to has slowed. As growth in commodity-exporting economies other sectors (Figure 2.6.1.4.F; Liu and Steenbergen 2019; fell sharply during the commodity price slump of 2014-16, World Bank 2018p). SSA’s participation in global value construction stalled, consumption eased, and credit chains is mostly limited to exports of raw agricultural contracted. Real-income losses in industrial sectors spilled commodities and natural resources used as inputs in other over to weaker demand in the broader economy. As a countries’ exports (World Bank 2019d). Greater result, services sectors were no longer able to absorb as manufacturing sector participation in international trade much labor as they did pre-crisis. and global value chains has been constrained by the sector’s relative lack of international competitiveness, in Other drivers of productivity growth. Rapid part due to high productivity-adjusted labor costs (Gelb et improvements in the key drivers of productivity during the al. 2017) and an array of non-tariff barriers, including the pre-crisis period supported productivity growth until the region’s disadvantageous geography (Christ and Ferrantino global financial crisis; however, the pace of improvement 2011; Raballand et al. 2012). has since lost momentum. Productivity drivers with particularly prominent slowdowns in improvements Prospects for productivity growth slowdown. Although include innovation, gender equality, education, health, wide sectoral productivity differentials offer ample trade openness, institutional quality, and investment productivity growth potential through sectoral reallocation (Figure 2.6.1.4.A and 2.6.1.4.B). Moreover, SSA away from the agriculture sector, headwinds to continues to lag well behind other EMDEs in most drivers productivity growth are substantial and expected to persist. of productivity (Figure 2.6.1.4.C). • Weather-related shocks. Given agriculture’s Institutional quality and the business environment. prominence in economic activity in SSA, climate Although various aspects of governance and institutional change presents severe challenges to productivity quality improved in the region from the late 1990s into growth prospects in agricultural sectors as mean the pre-crisis period, this progress has mostly stalled, and temperatures continue to rise and extreme weather even deteriorated in some instances. On average, business events occur more frequently (IPCC 2014; Steinbach climates have also regressed during the post-crisis period; 2019; World Bank 2019a, 2019f). today, almost two-thirds of SSA countries still rank in the lowest quartile of countries by business climates, and one- • Constraints to public investment. Government half do so for poor governance (Figure 2.6.1.4.D). Poor indebtedness in SSA has increased sharply since 2013, business climates and governance, as well as distortions rising by 20 percentage points, on average, to 60 154 CHAPTER 2.6 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.6.1 Labor productivity in Sub-Saharan Africa: Trends and drivers (continued) FIGURE 2.6.1.4 Drivers of productivity growth in SSA Despite significant improvements, key productivity drivers remain significantly below those of advanced economies and EMDEs. Moreover, their pace of improvement has slowed in recent years. On average, business environments in Sub- Saharan Africa are more challenging than in other countries. While the region boasts the largest share of higher-productivity foreign-owned firms, its firms export less than their counterparts in other EMDEs. A. Index of productivity growth drivers B. Share of SSA economies with slower C. Levels of drivers across regions, 2018 improvements in drivers 2013-18 relative to 2003-08 D. Obstacles to doing business E. Share of exporting firms F. Ownership status Source: Penn World Table; United Nations (2015); World Bank (Enterprise Surveys, Wealth Accounting, and World Development Indicators). Note: Unless specified otherwise, productivity is defined as labor productivity (real GDP per person employed). A. For each country, index is a weighted average (the normalized coefficients appearing in Annex 3.3) of the normalized value of each driver of productivity. Drivers include the International Country Risk Guide rule of law index, patents per capita, share of non-tropical area, investment as a percent of GDP, ratio of female average years of education to male average years, share of population in urban areas, Economic Complexity Index, years of schooling, share of working-age population, and inflation. See Chapter 3 (Annex 3.3) for details. Regional and EMDE indexes are GDP-weighted averages. Samples include 54 EMDEs and 11 economies in SSA. B. Blue bars represent share of 48 economies in Sub-Saharan African economies where improvements in each driver of productivity were lower during 2008-17 than in the pre-crisis period 1998-2007, or changes in 2008-17 were below zero. Orange diamond is the corresponding values for 152 EMDE countries. Variables are defined as: Institutions = Government effectiveness; Innovation = patents per capita; Investment = investment to GDP ratio; Income equality = (-1) * Gini; Urbanization = urban population percentage; Economic complexity = Hidalgo and Hausmann (2009)'s Economic Complexity Index; Education = years of schooling; Demography = share of working-age population; and Gender equality = female average years of education divided by male average years. Samples include 26-48 SSA economies, depending on the driver, and 98-151 EMDEs. C. Unweighted average levels of drivers, normalized as average of advanced economies as 100. Blue bar represents average within SSA. Orange lines represent range of the average drivers for six regions in 2017. Variables corresponding to the concepts are follows: Education = years of education; Urbanization = share of population living in urban area; Investment = share of investment to GDP; Institution = Government Effectiveness; Economic Complexity = Economic Complexity Index+; Geography = share of land area which are outside of tropical region; Gender Equality = Share of the year of schooling for female to male; Demography = share of population under 14; Innovation = Log patent per capita; Trade = Exports + Imports/GDP; and Price stability = (-1)* inflation rate. D. Unweighted averages. Variables corresponding to the concepts are follows: Corruption = percent of firms identifying corruption as a major constraint; Electricity = Percent of firms identifying electricity as a major constraint; Financial access = percent of firms identifying access to finance as a major constraint; Informal sector competition = percent of firms identifying practices of competitors in the informal sector as a major constraint; Tax system is the average of tax rates (percent of firms identifying tax rates as a major constraint) and tax administration (percent of firms identifying tax administration as a major constraint); Trade regulations = percent of firms identifying customs and trade regulations as a major constraint; Crime = percent of firms identifying crime, theft and disorder as a major constraint. E. Share of exporting firms. Firms classified as high, medium, and low export more than 75 percent, between 50 and 75, and up to 25 percent of their sales, respectively. F. Share of firms with foreign ownership. Click here to download data and charts. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 SUB-SAHARAN AFRICA 155 BOX 2.6.1 Labor productivity in Sub-Saharan Africa: Trends and drivers (continued) percent of GDP in 2019. Reduced fiscal space could falling dependency ratios requires policies that support weigh on future productivity growth as it will likely female empowerment, including education, health care, constrain investment in productivity-enhancing and greater labor market access for women (Figure infrastructure, health, and education as well as 2.6.1.5.B; Bloom, Kuhn and Prettner 2017; Groth and research and development. It can also make countries May 2017; Kalemli-Ozcan 2003). As the ratio of the more vulnerable to financial crises (Box 3.4). young dependent population to the working-age population declines in SSA, resources could be freed up to • Commodity-reliance. Growth prospects for commodity invest in the health and education of the young, boosting sectors that could encourage capital deepening are the productivity of the future labor force and spurring per dim. Long-term commodity demand growth is capita growth (Ashraf, Weil and Wilde 2013). expected to moderate as growth in China—the largest source of commodity demand—slows and shifts Narrowing the gender gap. Despite some improvements, toward less resource-intensive sectors (World Bank gender gaps remain large in SSA (World Bank 2012). 2018o). Although the gender gap in labor force participation has been narrowing, on average, significant gaps in earnings of • High informality. High informality in the region— women relative to men persist. This reflects gender around 40 percent of official GDP and 90 percent of disparity in secondary and tertiary education, differing total employment—may inhibit faster aggregate occupations, and greater time devoted by women to productivity growth, as productivity among informal housework and childcare (World Bank 2019aa). firms are only one-seventh of that in their formal Moreover, improvements in the ratio of average years of counterparts (La Porta and Shleifer 2014; World education of females to males have been slowing in the Bank 2019f). In addition, much-needed productivity- post-crisis period. This is reflected by lower productivity of enhancing government spending is constrained females in agriculture, as well as female entrepreneurs— because informal firms do not pay taxes. crops tended by women yield one-third less per hectare than those of men, and a similar margin applies to profits Policy options earned by female entrepreneurs (Figure 2.6.1.5.C; O’Sullivan et al. 2014; Campos et al. 2019). Policies to Coordinated policy efforts are required to achieve stronger empower women and boost their productivity include productivity growth, notable reductions in extreme those promoting skills building beyond traditional training poverty, and a narrowing of the significant income gap programs, such as a greater focus on developing an with the rest of the world. There are four strands of policy entrepreneurial mindset; this approach has been found to options that emerge from the findings of this box. lift sales and profits in Togo (Campos et al. 2017, World Improving factors of production Bank 2019aa). Other policies include relieving capital constraints faced by females due to lower asset holdings Boosting human capital and leveraging demographic offering limited collateral; and addressing social norms that dividends. Improving human capital has been an constrain women’s economic opportunities and earnings, important source of productivity growth in SSA. such as perceptions about the type of work that is suitable Continued investment and increased spending on health to men or women. care, including greater provision of treatment for highly prevalent conditions such as malaria and HIV/AIDS, Closing infrastructure gaps. Although capital deepening could raise productivity of the labor force and life has continued apace among the region’s agricultural expectancy in general (Figure 2.6.1.5.A; Asiki et al. 2016; commodity exporters and commodity importers, it has Barofsky, Anekwe and Chase 2015; Ferreira, Pessôa and slowed considerably among most industrial commodity Dos Santos 2011). Increased life expectancy due to exporters, and severe infrastructure deficiencies remain improved health care also generates incentives to invest in throughout the region. Meeting the infrastructure-related education (Cervellati and Sunde 2011). In Ethiopia, a Sustainable Development Goals in 2030 will require rapid decline in fertility rates between 1995-2015, rising additional investment spending between 2015-30 of incomes, and falling poverty rates reflected an approach roughly 7 percent of GDP per year in SSA (excluding combining improvements in education and health, family maintenance spending)—the highest of all EMDE regions planning, and increased economic opportunity (World (Figure 2.6.1.5.D; Rozenberg and Fay 2019). Stronger Bank 2019aa). Harnessing the region’s potential productivity growth—through both capital-deepening demographic dividend from declining fertility rates and investment and improved TFP—is contingent on 156 CHAPTER 2.6 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.6.1 Labor productivity in Sub-Saharan Africa: Trends and drivers (continued) FIGURE 2.6.1.5 Prospects for productivity growth in SSA Continued improvements in health care could raise life expectancy and the overall productivity of the labor force, as increased life expectancy also generates incentives to invest in education. Sub-Saharan Africa could harness a significant demographic dividend, as falling fertility rates lead to a lower dependency ratio. Owing to limited access to resources and training, crops tended by women yield one-third less per hectare than those of men; a similar margin applies to profits earned by female entrepreneurs. To meet the SDGs by 2030 will require investment spending of about 7 percent of GDP per year. Reducing trade costs in SSA will help accelerate regional and global integration. Conflicts have been rising in the region, particularly acts of violence against civilians. A. Human capital development B. Dependency ratios and fertility rates C. Shortfalls in profits and agricultural output of females relative to males D. Infrastructure spending needs E. Import and export compliance costs F. Conflict events Source: Armed Conflict Location and Event Data Project database; Campos et al. (2019); O’Sullivan et al. (2014); Rozenberg and Fay (2019); World Bank Doing Business 2020; United Nations. Note: Unless specified otherwise, productivity is defined as labor productivity (real GDP per person employed). A. Unweighted averages. “Mortality rate” refers to under-five mortality. B. The dependency ratio is calculated as the ratio of the population at ages 0–14 plus the population aged 65+ to the population at ages 15–64. C. Bars for “Entrepreneur profits” show the extent to which profits for male-owned firms exceed those of female-owned firms using data from impact evaluations. Bars for “Agricultural output per hectare” show the extent to which agricultural output per hectare on male-managed plots exceeds that of female-managed plots. Entrepreneur profits in Ghana reflect the average of both the Grants for Micro-Enterprises Survey and the Tailoring Survey; Entrepreneur profits in Nigeria reflect the average of both the Growth and Employment Survey and the Business Plan Competition Survey. Agricultural output per hectare accounts for differences in plot size and geographic factors. Agricultural output in Nigeria reflects a simple average of gaps for northern Nigeria (46 percent) and southern Nigeria (17 percent). D.E. Bars show average annual spending needs during 2015-30. Estimates are generated using policy assumptions that cap investment needs at 4.5 percent of LMICs’ GDP per year. SSA=Sub-Saharan Africa, SAR=South Asia, MNA=Middle East and North Africa, EAP=East Asia and Pacific, LAC=Latin America and the Caribbean. E. Unweighted averages. Sample includes 156 EMDEs and 47 SSA economies. EMDE average excludes SSA. F. Sample includes 30 SSA economies. Last observation is November 9, 2019. Click here to download data and charts. infrastructure deficiencies being addressed. Access to resilience of existing infrastructure are needed to limit electricity is a critical obstacle to achieving development frequent disruptions, particularly in power, water and goals in SSA, and reforms to improve access in a sanitation, transport, and telecommunications (World sustainable manner need to strike a balance between Bank 2019ab). To ensure public investment is efficient in affordable provision for consumers, particularly the poor, boosting growth and productivity, it should be supported and cost recovery for utilities (Blimpo and Cosgrove- by adequate public investment management frameworks Davies 2019; Vorisek and Yu (forthcoming). In addition that encompass strong cash management and procurement to closing infrastructure gaps, improvements to the processes. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 SUB-SAHARAN AFRICA 157 BOX 2.6.1 Labor productivity in Sub-Saharan Africa: Trends and drivers (continued) Boosting firm productivity digital technologies—more so than other regions (Choi, Dutz, and Usman 2019; Hjort and Poulsen 2019). SSA’s Boosting productivity in agriculture. Given the large comparatively low levels of human capital and high degree share of activity and employment accounted for by of informality are ideally suited for the adoption and agriculture, measures to raise agricultural productivity at development of productivity-enhancing, low-skill-biased the farm level—especially in staple crops—can yield digital technologies in the agriculture, manufacturing and significant development gains (Beegle and Christiaensen services sectors. In some countries, the use of digital 2019). These include ensuring secure land tenures, better technologies has been found to boost firm productivity by access to markets and finance, better crop choices, more facilitating process and product innovation (Democratic effective and increased use of fertilizers, improved Republic of Congo, Tanzania; Cirera, Lage, and Sabetti irrigation, diffusion and adoption of new technologies, as 2016). Digital technologies can also help in banking the well as targeted trainings to help small farmers reap the unbanked and transform lending in SSA. Kenya’s mobile benefits of cutting-edge knowledge and practices specific money service, M-Pesa, boosted the financial savings of to the area and product (Chen 2017; Fuglie et al. 2019; female-headed households and enabled women to move Sinha and Xi 2018; World Bank 2019aa). For example, out of agriculture into more productive sectors (Suri and text messages providing advice and reminders to sugarcane Jack 2016). Digital loans offered through mobile money farmers in Kenya helped boost fertilizer use and crop yields platforms are also growing in popularity and may grant (Casaburi et al 2014; Fuglie et al. 2019). Ensuring gender financial inclusion to individuals without credit scores or equality in access to resources could further boost sufficient collateral, as digital loan providers use alternative agricultural productivity; giving women in Malawi and credit scores based on telecommunications data (Cook and Ghana the same access to fertilizers and other inputs as McKay 2015; Francis, Blumenstock, and Robinson 2017; men could boost maize yields by one-sixth (World Bank World Bank 2019aa). However, the use of digital credit 2012). Gains from faster productivity growth in has so far been largely concentrated in urban areas, at short agriculture will free up workers to transition to other, maturities, and not as investment loans by the rural poor more productive, sectors. (Björkegren and Grissen 2018). Addressing informality. Although informality is higher in SSA than in other EMDE regions, informal firms often Accelerating trade openness and global integration. The brim with potential—more formal firms in SSA started as African Continental Free Trade Area (AfCFTA) has the informal firms, and this period of transition is found to be potential to boost regional trade and bolster firm shorter than in other EMDEs (World Bank 2019f). productivity by facilitating investment, international Policies to unlock informal firms’ potential include competitiveness, the transfer of technology and new upgrading skills of workers, ensuring better access to innovations, and participation in regional and global value inputs and resources like financial services, transport and chains (Berg and Krueger 2003; Calderon and Cantú communications connectivity, health services, land and 2019; Del Prete, Giovannetti, and Marvasi 2017; Laget et property rights, and product markets (Oosthuizen et al. al. 2018; World Bank 2019d). To maximize the potential 2016). Removing barriers to enter the formal sector can productivity gains from the free trade area, infrastructure further accelerate the transition out of informality: needs to be expanded—particularly transport networks— lowering registration costs by half could double the share and business climates improved. In addition, gains from of formal enterprises through formalization of informal AfCFTA depend on the implementation of trade firms and new entrants (Nguimkeu 2015; World Bank facilitation measures and addressing of significant non- 2019aa). Regulatory and institutional reforms to build tariff barriers to trade—trade costs in SSA, such as border public trust can strengthen incentives for firms to operate and documentary compliance costs, are roughly one-half formally. Policies aimed directly at the youth can bolster higher than those of other EMDE regions (Figure the prospects of the future workforce and help alleviate 2.6.1.5.E; World Bank 2019d). Currently, most regional youth unemployment. In Rwanda, entrepreneurship has trade in SSA takes place among countries within existing regional economic communities, as high tariffs and non- been introduced as a secondary school subject to help tariff barriers limit trade between countries of different prepare the youth to be successful entrepreneurs or to compete in the formal labor market (Choi, Dutz, and groupings. Usman 2019). Encouraging sectoral reallocation Leveraging digital technologies. Firm productivity in SSA could also benefit significantly from the proliferation of Enabling factor mobility. Productivity gains from sectoral 158 CHAPTER 2.6 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 2.6.1 Labor productivity in Sub-Saharan Africa: Trends and drivers (continued) reallocation of labor in the region—a major driver of pre- capacity in policy implementation, boosting access to crisis productivity growth—can be reignited by policies adaptation financing, and raising public awareness of aimed at reducing the barriers to factor mobility. These climate change (Adenle et al. 2017; World Bank 2019ac). barriers include low human capital of the labor force, weak infrastructure (such as inadequate transport systems in Stability. SSA has historically witnessed many conflicts, urban areas), low access to finance, and disadvantageous particularly between the 1970s and early 2000s, that not trade policies. In Nigeria, tariff structures have been shown only took heavy human tolls, but also shook the stability of to reduce incentives for sectoral reallocation to higher- the affected countries by weakening institutions and productivity sectors, as the tariffs systematically boosted severely damaging or destroying infrastructure. Conflicts in profitability of the least productive sectors but not that of Burundi, the Democratic Republic of Congo, Liberia, higher-productivity sectors (World Bank 2017g). Rwanda, and Sierra Leone inflicted losses of human life equivalent to between 1 and 10 percent of their Diversification. Countries with highly diversified populations (Steinbach 2019; World Bank 2019a). More economic activity across a broad range of sectors tend to recently, rising incidence of conflict—particularly acts of have higher productivity levels (Chapter 3). SSA, however, violence against civilians—has increasingly weighed on remains heavily dependent on extractives sectors, activity in several countries and forcibly displaced large particularly for export and fiscal revenues, with the latter populations (Figure 2.6.1.5.F). Efforts to achieve lasting dependence often a cause of procyclical fiscal policies. peace can strengthen economic activity and boost Policy measures aimed at broadening the production base productivity through stronger investment and increased toward a wider and more complex array of export goods, TFP (Chen, Loayza, and Reynal-Querol 2008). across a range of manufacturing and services sectors, will enable greater participation in value chains and help Strengthening institutional quality and business insulate economic activity from the destabilizing effects of environments. Business environments stand to benefit large international commodity price swings. In Côte from improved infrastructure; limited access to reliable d’Ivoire—the world’s largest supplier of cocoa beans— electricity and poor transport infrastructure are often cited diversification along the cocoa value chain through the as key constraints to business in SSA. In addition, high expansion of domestic grinding and processing facilities non-infrastructure-related costs, such as high prices to has allowed the country to also produce a diverse array of transport goods within countries and across borders, tend value-added cocoa products and to overtake the to exacerbate the burden of weak infrastructure. In many Netherlands as the world’s leading cocoa-processing instances, high road-transport costs reflect excessive market country (World Bank 2016h). AfCFTA could contribute power of trucking companies. Competition-enhancing to economic diversification if it leads to the establishment deregulation can help alleviate this business constraint and of regional value chains. However, successful economic boost productivity. For example, in landlocked Rwanda, diversification requires several supporting measures, deregulation in the transport sector led to an abrupt fall in including improved human capital, better infrastructure, transport costs (Barrett et al. 2017) Business environment stronger governance, and deeper financial markets with deficiencies can further be addressed by increasing access to increased access to credit (Fosu and Abass 2019). finance, simplifying tax systems, reducing regulatory burdens and compliance requirements, improving judicial Creating a growth-friendly environment systems to address corruption and strengthen enforcement, and liberalizing labor and product markets (Bah and Fang Protection from climate change. Some of the adverse 2015; World Bank 2019f). Strengthening institutional effects of climate change can be mitigated through quality by improving judicial systems can help address appropriate land-use planning and investment in climate- corruption—a leading obstacle to doing business—and smart infrastructure (Collier, Conway and Venables 2008; strengthen contract enforcement. Such structural reforms World Bank 2019a). Effective social protection policies, can bolster firm productivity (Kouamé and Tapsoba possibly financed with energy taxes or the removal of fuel 2018). Reforms aimed at improving the business subsidies, could provide resources to support livelihoods environment can also help lower the size of the informal during extreme events (Hallegatte et al 2015). Climate sector, which tends to have lower productivity than the adaptation policies can be strengthened by building formal economy. G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 C H A P TE R 2 159 Aiyar, S., B. Augustyniak, C. Ebeke, E. Ebrahimy, S. References Elekdag, N. Klein, S. Lall, H. Zhao and D. Muir. Abdih, Y., P. Lopez-Murphy, A. Roitman, and R. 2013. “German-Central European Supply Chain – Sahay. 2010. “The Cyclicality of Fiscal Policy in the Cluster Report”. IMF Country Report 13/263, Middle East and Central Asia: Is the Current Crisis International Monetary Fund, Washington, DC. Different?” IMF Working Paper 10/68, International Aiyar, S., C. Ebeke, X. Shao. 2016. “The Impact of Monetary Fund, Washington, DC. Workforce Aging on European Productivity.” IMF Acemoglu, D., S. Johnson, and J. Robinson. 2005. Working Paper 16/238, International Monetary Fund, “Institutions as a Fundamental Cause of Long-Run Washington, DC. Growth.” In Handbook of Economic Growth 1A, edited Aker, J. C., and I. M. Mbiti. 2010. “Mobile Phones by P. Aghion and S. Durlauf, 385–472. Amsterdam: and Economic Development in Africa.” Journal of North-Holland. Economic Perspectives 24 (3): 207-32. Acemoglu, D., S. Naidu, P. Restrepo, and J. Robinson. Albino-War, M., S. Cerovic, F. Grigoli, J. C. Flores, J. 2019. “Democracy Does Cause Growth.” Journal of Kapsoli, H. Qu, Y. Said, B. Shukurov, M. Sommer, Political Economy 127 (1): 47–100. and S. Yoon. 2014. “Making the Most of Public Investment in MENA and CCA Oil Exporting Adamopoulos, T., and D. Restuccia. 2014. “The Size Countries.” IMF Staff Discussion Note 14/10, Distribution of Farms and International Productivity International Monetary Fund, Washington, DC. Differences.” American Economic Review 107 (10): 1667-97. Alesina, A., D. Furceri, J. Ostry, C. Papageorgiou, and D. Quinn. 2019. “Structural Reforms and Election: Adamopoulos, T., and D. Restuccia. 2018. “Geography Evidence from a World-wide New Dataset.” Harvard and Agricultural Productivity: Cross-Country Evidence University Working Paper, Cambridge, MA. from Micro Plot-Level Data.” NBER Working Paper 24532, National Bureau of Economic Research, Almeida, R., and T. Packard. 2018. Competências e Cambridge, MA. Empregos: uma Agenda para a Juventude. Washington, DC: Grupo Banco Mundial. ADB (Asian Development Bank). 2019. The Servicification of Manufacturing in Asia: Redefining the Altinok, N., N. Angrist, and H. A. Patrinos. 2018. Sources of Labor Productivity. Manila: Asian “Global Data Set on Education Quality (1965– Development Bank. 2015).” Policy Research Working Paper 8314, World Bank, Washington, DC. Adenle, A. A., J. D. Ford, J. Morton, S. Twomlow, K. Alverson, A. Cattaneo, R. Cervigni, et al. 2017. Amin, M., and H. Ulku. 2019. “Corruption, “Managing Climate Change Risks in Africa—A Global Regulatory Burden and Firm Productivity.” Policy Perspective.” Ecological Economics 141 (November): 190 Research Working Paper 8911, World Bank, -201. Washington, DC. Amjad, R., and N. Awais. 2016. "Pakistan’s Adler, G., R. Duval, D. Furceri, S. K. Çelik, K. Productivity Performance and TFP Trends, 1980- Koloskova, M. Poplawski-Ribeiro. 2017. “Gone with 2015: Cause for Real Concern." Lahore Journal of the Headwinds: Global Productivity.” IMF Staff Economics 21 (SE): 33-63. Discussion Note 17/04, International Monetary Fund, Washington, DC. Andrews, D., C. Criscuolo, and P. N. Gal. 2015. “Frontier Firms, Technology Diffusion and Public Aggarwal, A., M. Hoppe, and P. Walkenhorst. 2019. Policy: Micro Evidence from OECD Countries.” In “Special Economic Zones in South Asia: Industrial The Future of Productivity. Main Background Papers. Islands or Vehicles for Diversification?” Washington, Paris: Organization for Economic Co-operation and DC: World Bank. Development. Ahluwalia, I. J., S. Chaudhuri, and S. Chaudri. 2003. APO (Asian Productivity Organization). 2017. Asian "Trade Liberalization, Productivity, and Export Productivity Databook. Tokyo: Asian Productivity Performance: A Study of the Indian Manufacturing Organization. Sector in the 1990s." South Asia Region Internal Discussion Paper. Report IDP-189. Washington, DC: ———. 2018. APO Productivity Databook. Tokyo: World Bank. Asian Productivity Organization. 160 C H A P TE R 2 G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 Araujo, J. T., E. Vostroknutova, and K. Wacker. 2017. Working Paper 21390, National Bureau of Economic “Productivity Growth in Latin America and the Research, Cambridge, MA. Caribbean: Exploring the Macro-Micro Linkages.” MFM Global Practice Discussion Paper 19, World Ayyagari, M., A. Demirgüç-Kunt, and V. Maksimovic. Bank, Washington, DC. 2017. “SME Finance.” Policy Research Working Paper 8241, World Bank, Washington, DC. Arezki, R., M. Ait Ali Slimane, A. Barone, K. Decker, D. Detter, R. Fan, H. Nguyen, G. Miralles, L. Senbet, Ayyagari, M., P. F. Juarros, M. S. Martinez Peria, and 2019a. “Reaching New Heights: Promoting Fair S. Singh. 2016. “Access to Finance and Job Growth: Competition in the Middle East and North Africa.” Firm-level Evidence Across Developing Countries.” MENA Economic Update. October. Washington, DC: Policy Research Working Paper 7604, World Bank, World Bank. Washington, DC. Arezki, R., D. Lederman, A. Harb, R. Fan, and H. Bachmair, F. and J. Bogoev. 2018. “Assessment of Nguyen. 2019b. “Reforms and External Imbalances: Contingent Liabilities and Their Impact on Debt The Labor-Productivity Connectivity Connection in Dynamics in South Africa.” MTI Discussion Paper 1, the Middle East and North Africa.” MENA Economic Macroeconomics, Trade and Investment Global Update. April. Washington, DC: World Bank. Practice, The World Bank, Washington, DC. Arnold, J. M., B. S. Javorcik, M. Lipscomb, and A. Baffes, J., M. A. Kose, F. Ohnsorge, and M. Stocker. Mattoo. 2016. “Services Reform and Manufacturing 2015. “The Great Plunge in Oil Prices: Causes, Performance: Evidence from India.” Economic Journal Consequences, and Policy Responses.” Policy Research 126 (590): 1–39. Note 1, World Bank, Washington DC. Arnold, J. M., B. S. Javorcik, and A. Mattoo. 2011. Bah, E., and L. Fang. 2015. “Impact of the Business “Does Services Liberalization Benefit Manufacturing Environment on Output and Productivity in Africa.” Firms? Evidence from the Czech Republic.” Journal of Journal of Development Economics 114 (May): 159-171. International Economics 85 (1): 136-146. Baier, S. L., G. P. Dwyer, and R. Tamura. 2006. “How Arora, V. 2005. “Economic Growth in Post-Apartheid Important Are Capital and Total Factor Productivity South Africa: A Growth-Accounting Analysis.” In Post- for Economic Growth.” Economic Inquiry 44 (1): apartheid South Africa: The First Ten Years, edited by L. A. Ricci and M. Nowak, 13-22. Washington, DC: 23-49. International Monetary Fund. Baker, S. R., N. Bloom, and S. J. Davis. 2016. Artuc, E., G. Lopez-Acevedo, R. Robertson, and D. “Measuring Economic Policy Uncertainty.” The Samaan. 2019. "Export to Jobs: Boosting the Gains Quarterly Journal of Economics 131 (4): 1593–1636. from Trade in South Asia." South Asia Development Balabushko, O., O. Betliy, V. Movchan, R. Forum. Washington, DC: World Bank. Piontkivsky, and M. Ryzhenkov. 2018. “Crony Ashraf, Q. H., D. N. Weil, and J. Wilde. 2013. “The Capitalism in Ukraine: Relationship between Political Effect of Fertility Reduction on Economic Growth.” Connectedness and Firms' Performance.” Policy Population and Development Review 39 (1): 97-130. Research Paper 8471, World Bank, Washington DC. Asiki, G., G. Reniers, R. Newton, K. Baisley, J. Baldwin, R. 2019. The Globotics Upheaval: Nakiyingi-Miiro, E. Slaymaker, I. Kasamba et al. 2016. Globalization, Robotics and the Future of Work. New “Adult Life Expectancy Trends in the Era of York: Oxford University Press. Antiretroviral Treatment in Rural Uganda (1991– 2012).” Aids 30 (3): 487-493. Baldwin, R., and J. Lopez-Gonzalez. 2013. “Supply- Chain Trade: A Portrait of Global Patterns and Several Aslam, A., S. Beidas-Strom, R. Bems, O. Celasun, S. K. Testable Hypotheses.” The World Economy 38 (11): Celik, and Z. Koczan. 2016. “Trading on Their Terms? 1682-1721. Commodity Exporters in the Aftermath of the Commodity Boom.” IMF Working Paper 16/27, Bank of England. 2018. EU Withdrawal Scenarios and International Monetary Fund, Washington, DC. Monetary and Financial Stability: A Response to the House of Commons Treasury Committee. November. Attanasio, O., A. Guarin, C. Medina, and C. Meghir. London: Bank of England. 2015. “Long Term Impacts of Vouchers for Vocational Training: Experimental Evidence in Colombia.” NBER Barofsky, J., T. D. Anekwe, and C. Chase. 2015. G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 C H A P TE R 2 161 “Malaria Eradication and Economic Outcomes in Sub- Björkegren, D., and D. Grissen. 2018. “The Potential Saharan Africa: Evidence from Uganda.” Journal of of Digital Credit to Bank the Poor.” AEA Papers and Health Economics 44 (December): 118-136. Proceedings 108 (May): 68-71. Barrett, C. B., L. Christiaensen, M. Sheahan, and A. Blancher, N., A. Bibolov, and A. Fouejieu. 2019. Shimeles. 2017. “On the Structural Transformation of “Financial Inclusion of Small and Medium-Sized Rural Africa.” Policy Research Working Paper 7938, Enterprises in the Middle East and Central Asia.” IMF World Bank, Washington, DC. Working Paper 19/02, International Monetary Fund, Washington, DC. Barro, R. J., and J.-W. Lee. 2015. Education Matters: Global Schooling Gains from the 19th to the 21st Blimpo, M. P., and M. Cosgrove-Davies. Century. New York: Oxford University Press. 2019. Electricity Access in Sub-Saharan Africa: Uptake, Reliability, and Complementary Factors for Economic Bartelsman, E. J., and S. Scarpetta. 2007. “Reallocation Impact. Washington, DC: World Bank. and Productivity in Transition and Integration Phases.” Policy Research Working Paper 6572, World Bank, Bloom, D. E., M. Kuhn, and K. Prettner. 2017 Washington, DC. “Africa’s Prospects for Enjoying a Demographic Dividend.” Journal of Demographic Economics 83 (1): Beck, T. 2018. “Finance and Resource Booms: A 63-76. Complicated Relationship. Two Proposals.” In Rethinking the Macroeconomics of Resource-Rich Bloom, N., B. Eifert, A. Mahajan, D. McKenzie, and J. Countries, edited by Arezki, A, R. Boucekkine, J. Roberts. 2013. “Does Management Matter? Evidence Frankel, M. Laksaci, and R. van der Ploeg. CEPR’s from India.” Quarterly Journal of Economics 128 (1): Policy Portal. https://voxeu.org/system/files/epubli 1-51. cation/Resource_Rich_Countries_eBook.pdf. Bouis, R., R. Duval, and J. Eugster. 2016. “Product Beegle, K., and L. Christiaensen. 2019. Accelerating Market Deregulation and Growth: New Country- Poverty Reduction in Africa. Washington, DC: World Industry-Level Evidence.” IMF Working Paper 16/114, Bank. International Monetary Fund, Washington, DC. Behera, H., and S. Sharma. 2019. “Does Financial Bova, E., M. Ruiz-Arranz, F. Toscani, and H. E. Ture. Cycle Exist in India?” RBI Working Paper 03, Reserve 2016. “The Fiscal Costs of Contingent Liabilities; A Bank of India, Mumbai. New Dataset.” IMF Working Paper 16/14, International Monetary Fund, Washington, DC. Benhassine, N., S. H. Youssef, P. Keefer, A. Stone, and S. N. Wahba. 2009. “From Privilege to Competition: Bown, C., D. Lederman, S. Pienknagura, and R. Unlocking Private-led Growth in the Middle East Robertson. 2017. Better Neighbors: Toward a Renewal of and North Africa.” MENA Development Report. Integration in Latin America. Washington, DC: World Washington, DC: World Bank. Bank. Berg, A., and A. O. Krueger. 2003. “Trade, Growth, Brandt, N., P. Schreyer, and V. Zipperer. 2017. and Poverty: A Selective Survey.” IMF Working Paper “Productivity Measurement with Natural Capital.” 03/30, International Monetary Fund, Washington, Review of Income and Wealth 63 (S1): S7-S21. DC. Brenton, P., R. Newfarmer, and P. Walkenhorst. 2009. Beteille, T. 2019. Ready to Learn, Ready to Thrive: “Avenues for Export Diversification: Issues for Low- Before School, In School and Beyond School in South Asia. Income Countries.” MPRA Paper 22758, University Washington, DC: World Bank. Library of Munich. Beverelli, C., M. Fiorini, and B. Hoekman. 2017. Brown, J. D., G. A. Crespi, L. Iacovone, and L. “Services Trade Policy and Manufacturing Marcolin. 2016. “Productivity Convergence at Firm Productivity: The Role of Institutions.” Journal of Level: New Evidence from the Americas.” In International Economics 104 (January): 166–182. Understanding the Income and Efficiency Gap in Latin America and the Caribbean, edited by J. T. Araujo, E. Bisat, A., M. A. El-Erian, and T. Helbling. 1997. Vostroknutova, K. Wacker, and M. Clavijo. “Growth, Investment, and Saving in the Arab Washington, DC: World Bank. Economies.” IMF Working Paper 97/85, International Monetary Fund, Washington, DC. Brown, J. D., and J. S. Earle. 2007. “Firm-Level 162 C H A P TE R 2 G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 Components of Aggregate Productivity Growth in GCC: Past, Present, and Future.” IMF Staff Discussion ECA Economies.” Background Paper, World Bank, Note, International Monetary Fund, Washington, DC. Washington, DC. Campos, F., R. Coleman, A. Conconi, A. Donald, M. Brown, J. D., J. S. Earle, and Á. Telegdy, “The Gassier, M. Goldstein, Z. Chavez, et al. 2019. Profiting Productivity Effects of Privatization: Longitudinal from Parity: Unlocking the Potential of Women’s Business Estimates from Hungary, Romania, Russia, and in Africa. Washington, DC: World Bank. Ukraine.” Journal of Political Economy 114 (1): 61-99. Campos, F., M. Frese, M. Goldstein, L. Iacovone, H. Bruszt, L., and N. Campos. 2016. “Deep Economic Johnson, D. McKenzie, and M. Mensmann. 2017. Integration and State Capacity: The Case of the Eastern “Teaching Personal Initiative Beats Traditional Enlargement of the European Union.” ADP Working Training in Boosting Small Business in West Africa.” Paper, Asian Development Bank, Manilla. Science 357 (6357): 1287–90. Busso, M., L. Madrigal, and C. Pagés. 2013. Casaburi, L., M. Kremer, S. Mullainathan, and R. “Productivity and Resource Misallocation in Latin Ramrattan. 2014. “Harnessing ICT to Increase America.” B. E. Journal of Macroeconomics 13 (1): 903– Agricultural Production: Evidence from Kenya.” 32. Harvard University, Cambridge, MA. Cadestin, C., J. Gourdon, and P. Kowalski. 2016. Caselli, F. 2005. “Accounting for Cross-Country “Participation in Global Value Chains in Latin Income Differences.” In Handbook of Economic Growth America: Implications for Trade and Trade-Related 1, edited by P. Aghion and S. N. Durlauf, 679-741. Policy.” Trade Policy Paper No. 192, Organisation for Amsterdam: Elsevier. Economic Co-operation and Development, Paris. Cervellati, M., and U. Sunde. 2011. “Life Expectancy Caldara, D., M. Iacoviello, P. Molligo, A. Prestipino, and Economic Growth: The Role of the Demographic and A. Raffo. 2019. “The Economic Effects of Trade Transition.” Journal of Economic Growth 16 (2): Policy Uncertainty.” International Finance Discussion 99-133. Papers 1256, Board of Governors of the Federal Chen, C. 2017. “Technology Adoption, Capital Reserve System, Washington, DC. Deepening, and International Productivity Differ- Calderón, C., and S. Boreux. 2016. “Citius, Altius, ences.” Department of Economics Working Paper 584, Fortius: Is Growth in Sub-Saharan Africa More University of Toronto. Resilient?” Journal of African Economies 25 (4): 502- Chen, C. and D. Restuccia. 2018. “Agricultural 528. Productivity Growth in Africa.” Background paper Calderón, C., and C. Cantú. 2019. “Trade Integration prepared for the AFRCE project Boosting Productivity and Growth: Evidence from Sub-Saharan Africa.” in Sub-Saharan Africa, World Bank, Washington, DC. Policy Research Working Paper 8859, World Bank, Chen, E. 2002. “The Total Factor Productivity Debate: Washington, DC. Determinants of Economic Growth in East Asia.” Asian Pacific Economic Literature 40 (1):18-38. Calderón, C., and L. Servén. 2010. “Infrastructure and Economic Development in Sub-Saharan Africa.” Chen, S., N. V. Loayza, and M. Reynal-Querol. 2008 Journal of African Economies 19 (S1): i13-i87. “The Aftermath of Civil War.” The World Bank Economic Review 22 (1): 63-85. Calderón, C. Moral-Benito, E. and L. Serven. 2015. Is Infrastructure Capital Productive? A Dynamic Choi, J., M. Dutz, and Z. Usman. 2019. The Future of Heterogeneous Approach. Journal of Applied Work in Africa: Harnessing the Potential of Digital Econometrics, Vol 30(2): 177-198. Technologies for All. Washington, DC: World Bank. Calice, P., E. P. Ribiero, and S. Byskov. 2018. Chong, A., J. Galdo, and J. Saavedra. 2008. “Efficient Financial Allocation and Productivity “Informality and Productivity in the Labor Market in Growth in Brazil.” Policy Research Working Paper Peru.” Journal of Economic Policy Reform 11 (4): 229– 8479, World Bank, Washington, DC. 45. Callen, T., R. Cherif, F. Hasanov, A. Hegazy, and P. Christ, N., and M. J. Ferrantino. 2011. “Land Khandelwal. 2014. “Economic Diversification in the Transport for Export: The Effects of Cost, Time, and G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 C H A P TE R 2 163 Uncertainty in Sub-Saharan Africa.” World Develop- and the Organization for Economic Co-operation and ment 39 (10): 1749–59. Development. Chuah, L. L., Loayza, N. V. Nguyen, and H. Minh. Danforth, J., P. A. Medas, and V. Salins. 2016. How to 2018. “Resource Misallocation and Productivity Gaps Adjust to a Large Fall in Commodity Prices. Washington, in Malaysia.” Policy Research Working Paper 8368, DC: International Monetary Fund. World Bank, Washington, DC. De Haas, R., and I. Van Lelyveld. “Foreign Banks and Cirera, X., and A. P. Cusolito. 2019. “Innovation Credit Stability in Central and Eastern Europe. A Panel Patterns and Their Effects on Firm-Level Productivity Data Analysis.” Journal of Banking & Finance 30 (7): in South Asia.” Policy Research Working Paper 8876, 1927-1952. World Bank, Washington, DC. de Hoyos, R., H. Rogers, and M. Székely. 2016. Out of Cirera, X., R. N. Fattal Jaef, and H. B. Maemir. 2017. School and Out of Work: Risk and Opportunities for “Taxing the Good? Distortions, Misallocation, and Latin America’s Ninis. Washington, DC: World Bank. Productivity in Sub-Saharan Africa.” Policy Research Working Paper 7949, World Bank, Washington, DC. De Nicola, F., V. V. Kehayova, and H. M. Nguyen. 2018. “On the Allocation of Resources in Developing Cirera, X., F. Lage, and L. Sabetti. 2016. “ICT Use, East Asia and Pacific.” Policy Research Working Paper Innovation, and Productivity: Evidence from Sub- 8634, Washington, DC: World Bank. Saharan Africa.” Policy Research Working Paper 7868, World Bank, Washington, DC. de Paula, A., and J. A. Scheinkman. 2011. “The Informal Sector: An Equilibrium Model and Some Cirera, X., and W. F. Maloney. 2017. “The Innovation Empirical Evidence from Brazil.” Review of Income and Paradox: Developing-Country Capabilities and the Wealth 57 (s1): s8-s26. Unrealized Promise of Technological Catch-Up.” World Bank, Washington, DC. de Vries, G.J., M.P. Timmer, and K. de Vries 2013. “Structural Transformation in Africa: Static Gains, Cole, M. A., and E. Neumayer. 2006. “The Impact of Dynamic Losses.” GGDC Research Memorandum Poor Health on Total Factor Productivity.” The Journal 136, Groningen Growth and Development Centre, of Development Studies 42 (6): 918-938. University of Groningen, Netherlands. Collier, P., G. Conway, and T. Venables. 2008. Del Prete, D., G. Giovannetti, and E. Marvasi. 2017. “Climate Change and Africa.” Oxford Review of “Global Value Chains Participation and Productivity Economic Policy 24 (2): 337-353. Gains for North African firms.” Review of World Colombia Departamento Nacional de Planeación. Economics 153 (4): 675-701. 2018. “Estrategia para la atención de la migración desde Delogu, M., F. Docquier, and J. Machado. 2014. “The Venezuela.” Colombia Departamento Nacional de Dynamic Implications of Liberalizing Global Planeación, Bogotá. Migration.” CESifo Working Paper Series 4596, Cook, T., and C. McKay. 2015. “How M-Shwari CESifo Group, Munich. Works: The Story So Far.” Consultative Group to Devarajan, S., and L. Mottaghi. 2015. MENA Assist the Poor (CGAP) and Financial Sector Economic Monitor. Towards a New Social Contract. Deepening (FSD) Kenya. April. Washington, DC: World Bank. Crespi, G., and P. Zuniga. 2011. “Innovation and Productivity: Evidence from Six Latin American Di Mauro, F., D. Hoang, A. Feng, S. J. Ong, and J. Pang. 2018. “Productivity Was Not that Sluggish in Countries.” World Development 40 (2): 273-90. Developing Asia, Afterall. A Firm Level Perspective Cusolito, A. P., W., and Maloney 2018. Productivity Using a Novel Dataset.” Productivity Research Revisited: Shifting Paradigms in Analysis and Policy. Network, National University of Singapore mimeo. Washington, DC: World Bank. Diao, X., M. McMillan, and D. Rodrik. 2017. “The Cusolito, A. P., R. Safadi, and D. Taglioni. 2017. Recent Growth Boom in Developing Economies: A “Inclusive Global Value Chains: Policy Options for Structural Change Perspective.” NBER Working Paper Small and Medium Enterprises and Low-Income 23132, National Bureau of Economic Research, Countries.” A co-publication of the World Bank Group Cambridge, MA. 164 C H A P TE R 2 G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 Diaz, J. J., and D. Rosas. 2016. “Impact Evalation of the Western Balkans Investment Summit, hosted by the the Job Youth Training Program Projoven.” Working EBRD, February 26, 2018, European Bank for Paper 693, Inter-American Development Bank, Reconstruction and Development, London. Washington, DC. ———. 2018b. Transition Report 2018-19: Work in Dieppe, A., G. Kindberg-Hanlon, and S. Kiliç elik. Transition. London: European Bank for Reconstruction Forthcoming. “Global Productivity Trends: Looking and Development. Through the Turbulence.” In Productivity in Emerging and Developing Economies. Washington, DC: World ———. 2019a. Regional Economic Prospects in the Bank. EBRD Regions: Stalling Engines of Growth. London: European Bank for Reconstruction and Development. Diop, N., and D. Marotta. 2012. Natural Resource Abundance, Growth, and Diversification in the Middle ———. 2019b. Transition Report 2019-20: Better East and North Africa: The Effects of Natural Resources Governance, Better Economies. London: European Bank and the Role of Policies. Washington, DC: World Bank. for Reconstruction and Development. Divanbeigi, R., and R. Ramalho. 2015. “Business Eden, M., and H. Nguyen. 2016. “Reconciling Micro- Regulations and Growth.” Policy Research Working and Macro-Based Estimates of Technology Adoption Paper 7299, World Bank, Washington, DC. Lags in a Model of Endogenous Technology Adop- tion.” In Understanding the Income and Efficiency Gap in Du Plessis, S., and B. Smit. 2007. “South Africa's Latin America and the Caribbean, edited by J. T. Growth Revival after 1994.” Journal of African Araujo, E. Vostroknutova, K. Wacker, and M. Clavijo. Economies 16 (5): 668-704. Washington, DC: World Bank. Duranton, G., Ghani. E., Goswami, A. G., and W. Elbadawi, I., and N. Loayza. 2008. “Informality, Kerr. 2016. “A Detailed Anatomy of Factor Employment and Economic Development in the Arab Misallocation in India.” Policy Research Working World.” Journal of Development and Economic Policies Paper 7547, World Bank, Washington, DC. 10 (2): 27-75. Dutz, M. A. 2018. Jobs and Growth: Brazil’s Elekdag, S., D. Muir, and Y. Wu. 2015. “Trade Productivity Agenda. Washington, DC: World Bank. Linkages, Balance Sheets, and Spillovers: The Germany- Central European Supply Chain.” Journal of Policy Dutz, M. A., R. K. Almeida, and T. G. Packard. 2018. Modeling 37 (2): 374-387. The Jobs of Tomorrow: Technology, Productivity, and Prosperity in Latin America and the Caribbean. Elgin, C., and O. Oztunali. 2012. “Shadow Economies Washington, DC: World Bank. Around the World: Model Based Estimates.” Working Paper 2012/05, Department of Economics, Boğaziçi Dutz, M. A., and S. D. O’Connell. 2013. University, Istanbul. “Productivity, Innovation and Growth in Sri Lanka: An Empirical Investigation.” Policy Research Working Ellis, P., and M. Roberts. 2016. Leveraging Paper 6354, World Bank, Washington, DC. Urbanization in South Asia: Managing Spatial Transformation for Prosperity and Livability. South Asia Easterly, W., and S. Fischer. 1994. “The Soviet Development Matters. Washington, DC: World Bank. Economic Decline Revisited.” Policy Research Working Paper 1284, World Bank, Washington, DC. Enache, M., E. Ghani, and S. O'Connell. 2016. “Structural Transformation in Africa: A Historical EBRD (European Bank for Reconstruction and View.” Policy Research Working Paper 7743, World Development). 2014. Transition Report 2013: Stuck in Bank, Washington, DC. Transition? London: European Bank for Reconstruction and Development. Falcetti, E., T. Lysenko, and P. Sanfey. 2006. “Reforms and Growth in Transition: Re-examining the ———. Transition Report 2017-18: Sustaining Growth. Evidence.” Journal of Comparative Economics 34 (3): London: European Bank for Reconstruction and 421-445. Development. FAO (Food and Agriculture Organization of the United ———. 2018a. “The Western Balkans in Transition: Nations). 2019. Crop Prospects and Food Situation. July. Diagnosing the Constraints on the Path to a Food and Agriculture Organization of the United Sustainable Market Economy.” Background paper for Nations: Rome. G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 C H A P TE R 2 165 Fay, M., L. Andres, C. Fox, U. Narloch, S. Straub, and Funke, N., A. Isakova, and M. Ivanyna. 2017. M. Slawson. 2017. Rethinking Infrastructure in Latin “Identifying Structural Reform Gaps in Emerging America and the Caribbean: Spending Better to Achieve Europe, the Caucasus, and Central Asia.” IMF More. Washington, DC: World Bank. Working Paper 17/82, International Monetary Fund, Washington, DC. Fayad, G., and T. Rasmussen. 2012. “Realizing Growth Objectives: Transitioning from Factor Gatti, R., M. Morgandi, E. Grun, S. Brodmann, D. Accumulation to Productivity Improvement.” In Saudi Angel-Urdinola, J.M. Moreno, D. Marotta, M. Arabia: Selected Issues, IMF Country Report 12/272, Schiffbauer, and E. Mata Lorenzo. 2013. Jobs for International Monetary Fund, Washington, DC. Shared Prosperity: Time for Action in the Middle East and North Africa. Washington, DC: World Bank. Fernandes, A. M. 2008. “Firm Productivity in Bangladesh Manufacturing Industries.” World Develop- Gautam, M., and A. Mansur. 2018. “Too Small to be ment 36 (10): 1725-1744. Beautiful? The Farm Size and Productivity Relationship in Bangladesh.” Food Policy 84 (April): 165-175. Fernández-Arias, E., and S. Rodríguez-Apolinar. 2016. “The Productivity Gap in Latin America: Lessons from Gelb, A., C. J. Meyer, V. Ramachandran, and D. 50 Years of Development.” Working Paper IDB-WP- Wadhwa. 2017. “Can Africa Be a Manufacturing 692, Inter-American Development Bank, Washington, Destination? Labor Costs in Comparative Perspective.” DC. CGD Working Paper 466, Center for Global Development, Washington, DC. Ferreira, P. C., S. de Abreu Pessôa, and F. Veloso. 2013. “On the Evolution of Total Factor Productivity Georgiev, Y., P. Nagy-Mohacsi, and A. Plekhanov. in Latin America.” Economic Inquiry 51 (1): 16–30. 2017. “Structural Reform and Productivity Growth in Emerging Europe and Central Asia.” ADB Economics Ferreira, P. C., S. Pessôa, and M. R. Dos Santos. 2011. Working Paper Series 532, Asian Development Bank, “The Impact of AIDS on Income and Human Manila. Capital.” Economic Inquiry 49 (4): 1104-1116. Ghani, E., W. R. Kerr, and S. D. O’Connell. 2013. Flabbi, L., and R. Gatti. 2018. “A Primer on Human “Input Usage and Productivity in Indian Capital.” Policy Research Working Paper 8309, World Manufacturing Plants.” Policy Research Working Paper Bank, Washington, DC. 6656. World Bank, Washington, DC. Fosu, A. K., and A. F. Abass. 2019. “Domestic Credit Gharleghi, B., and V. Popov. 2018. “Farewell to and Export Diversification: Africa from a Global Agriculture? Productivity Trends and the Competi- Perspective.” Journal of African Business 20 (2): 160-179. tiveness of Agriculture in Central Asia.” MPRA Paper 89520, University Library of Munich, Germany. Francis, E., J. Blumenstock, and J. Robinson. 2017. “Digital Credit in Emerging Markets: A Snapshot of Ghassibe, M., M. Appendino, and S.E. Mahmoudi. the Current Landscape and Open Research 2019. “SME Financial Inclusion for Sustained Growth Questions.” Digital Credit Observatory, Center for in the Middle East and Central Asia.” IMF Working Effective Global Action, University of California, Paper 19/209, International Monetary Fund, Berkeley. Washington, DC. Freund, C., M. J. Ferrantino, M. Maliszewska, and M. Goedhuys, M., N. Janz, and P. Mohnen. 2008. “What Ruta. 2018. “Impacts on Global Trade and Income of Drives Productivity in Tanzanian Manufacturing Current Trade Disputes.” MTI Practice Note 2, World Firms: Technology or Business Environment?” The Bank, Washington, DC. European Journal of Development Research 20 (2): 199- 218. Fuglie, K., M. Gautam, A. Goyal, and W. Maloney. 2019. Harvesting Prosperity: Technology and Productivity Goretti, M., D. Kihara, R. Salgado, and A. M. Gulde. Growth in Agriculture. Washington, DC: World Bank. 2019. “Is South Asia Ready for Take Off?” Asia and Pacific Department Paper 19/18, International Fujimori, A., and T. Sato. 2015. “Productivity and Monetary Fund, Washington, DC. Technology Diffusion in India: The Spillover Effects from Foreign Direct Investment.” Journal of Policy Gould, D. 2018. Critical Connections: Promoting Modeling 37 (4): 630-651. Economic Growth and Resilience in Europe and Central 166 C H A P TE R 2 G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 Asia. Europe and Central Asia Studies. Washington, DC: Vogt-Schilb. 2015. Shock Waves: Managing the Impacts World Bank. of Climate Change on Poverty. Washington, DC: World Bank. Gould, D. M., C. Tan, and A. S. S. Emamgholy. 2013. “Attracting Foreign Direct Investment: What Can Hallward-Driemeier M., G. Iarossi, and K. L. Sokoloff. South Asia’s Lack of Success Teach Other Developing 2002. “Exports and Manufacturing Productivity in East Countries?” Policy Research Working Paper 6696, Asia: A Comparative Analysis with Firm-Level Data.” World Bank, Washington, DC. NBER Working Paper 8894, National Bureau of Economic Research, Cambridge, MA. Gradzewicz, M., J. Growiec, M. Kolasa, Ł. Postek, and P. Strzelecki. 2018. “Poland’s Uninterrupted Growth Hallward-Driemeier, M., and G. Nayyar. 2018. Performance: New Growth Accounting Evidence.” Trouble in the Making? The Future of Manufacturing- Post-Communist Economies 30 (2): 238-272. Led Development. Washington, DC: World Bank. Grainger, C. A., and F. Zhang. 2017. “The Impact of Helble, M., T. Long, and T. Le. 2019. “Sectoral and Electricity Shortages on Firm Productivity: Evidence Skill Contributions to Labor Productivity in Asia.” from Pakistan.” Policy Research Working Paper 8130, ADBI Working Paper 929, Asian Development Bank World Bank, Washington, DC. Institute, Tokyo. Grazzi, M., and J. Jung. 2016. “ICT, Innovation, and Henry, P. 2007. “Capital Account Liberalization: Productivity.” In Firm Innovation and Productivity in Theory, Evidence, and Speculation.” Journal of Latin America and the Caribbean: The Engine of Economic Literature 45 (4): 887-935. Economic Development, edited by M. Grazzi and C. Hesse, H. 2008. “Export Diversification and Economic Pietrobelli, 103-36. Washington, DC: Inter-American Growth.” Working Paper 21, Commission on Growth Development Bank. and Development, World Bank, Washington, DC. Grigoli, F., A. Herman, A. J. Swiston. 2017. “A Crude Hjort, J., and J. Poulsen. 2019. “The Arrival of Fast Shock : Explaining the Impact of the 2014-16 Oil Price Internet and Employment in Africa.” American Decline Across Exporters.” IMF Working Paper Economic Review 109 (3): 1032-79. 17/160, International Monetary Fund, Washington, DC. Hodge, A., S. Shankar, D. S. Prasada Rao, and A. Duhs. 2011. “Exploring the Links between Corruption Groth, H., and J. F. May, eds. 2017. Africa’s and Growth.” Review of Development Economics 15 (3): Population: In Search of a Demographic Dividend. 474-90. Cham. Switzerland: Springer International. Hsieh, C., and P. J. Klenow. 2009. “Misallocation and Gylfason, T. 2018. “From Economic Diversification to Manufacturing TFP in China and India.” Quarterly Growth. Two Proposals.” In Rethinking the Journal of Economics 124 (4): 1403-1448. Macroeconomics of Resource-Rich Countries, edited by Arezki, A, R. Boucekkine, J. Frankel, M. Laksaci, and Hu A. G. Z., G. H. Jefferson and Q. Jinchang. 2005. R. van der Ploeg. CEPR’s Policy Portal. https:// “R&D and Technology Transfer: Firm-Level Evidence voxeu.org/system/files/epublication/Resource_Rich_Co from Chinese Industry.” Review of Economics and untries_eBook.pdf. Statistics 87 (4): p.780-786. H. M. Government. 2018. “EU Exit: Long-Term Ianchovichina, E. 2017. Eruptions of Popular Anger: The Economic Analysis.” November. HM Government, Economics of the Arab Spring and Its Aftermath. MENA London. Development Report. Washington, DC: World Bank. Ha, J., M. A. Kose, and F. Ohnsorge. 2019. Inflation in IDB (Inter-American Development Bank). 2016. Time Emerging and Developing Economies: Evolution, Drivers, to Act: Latin America and the Caribbean Facing Strong and Policies. Washington, DC: World Bank. Challenges. Washington, DC: Inter-American Development Bank. Haile, F. 2018. “Structural Change in West Africa: A Tale of Gain and Loss.” Policy Research Working IEA (International Energy Agency). 2019. Oil Market Paper 8336, World Bank, Washington, DC. Report. Paris: IEA. Hallegatte, S., M. Bangalore, L. Bonzanigo, M. Fay, T. ILO (International Labour Office). 2017. What Future Kane, U. Narloch, J. Rozenberg, D. Treguer, and A. for Decent Work in Europe and Central Asia: G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 C H A P TE R 2 167 Opportunities and Challenges. Geneva: International ———. 2019c. Fiscal Monitor: Curbing Corruption. Labour Office. April. Washington, DC: International Monetary Fund. IMF (International Monetary Fund). 2006. World ———. 2019d. Middle East and Central Asia Economic Economic Outlook. Financial Systems and Economic Outlook. October. Washington, DC: International Cycles. Chapter 3. Asia Rising: Patterns of Economic Monetary Fund. Development and growth. Iootty, M., J. Pena, and D. De Rosa. 2019. ———. 2012. “Saudi Arabia Article IV Consultation.” “Productivity Growth in Romania: A Firm-Level IMF Country Report 12/271, International Monetary Analysis.” Policy Research Working Paper 9043, World Fund, Washington, DC. Bank, Washington, DC. ———. 2013. “Labor Market Reforms to Boost IPCC (Intergovernmental Panel on Climate Change). Employment and Productivity in the GCC.” Annual 2014. Climate Change 2014: Impacts, Adaptation, and Meeting of the Gulf Cooperation Council Ministers of Vulnerability. IPCC Working Group II. Geneva: Finance and Central Bank Governors, Riyadh, October Intergovernmental Panel on Climate Change. 5. International Monetary Fund, Washington, DC. ———. 2019. Climate Change and Land: an IPCC ———. 2014. The Caucasus and Central Asia: Special Report on Climate Change, Desertification, Land Transitioning to Emerging Markets. Washington, DC: Degradation, Sustainable Land Management, Food Security, and Greenhouse Gas Fluxes in Terrestrial International Monetary Fund. Ecosystems. Geneva: Intergovernmental Panel on ———. 2015. “Islamic Republic of Iran Climate Change. Consultation.” IMF Country Report No. 15/349, ITF (International Transport Forum). 2018. “The International Monetary Fund, Washington, DC. Billion Dollar Question: How Much Will It Cost to ———. 2016a. “Economic Diversification in Oil- Decarbonise Cities’ Transport Systems?” Discussion Exporting Arab Countries.” Annual Meeting of Arab Paper, Paris. Ministers of Finance. Jirasavetakul, L. B. F., and J. Rahman. “Foreign Direct ———. 2016b. “Bolivia: 2016 Article IV Staff Investment in New Member States of the EU and Report.” IMF Country Report 16/387, International Western Balkans: Taking Stock and Assessing Prospects.” IMF Working Paper 18/187, International Monetary Fund, Washington, DC. Monetary Fund, Washington, DC. ———. 2017. Asia and Pacific: Preparing for Choppy Kalemli-Ozcan, S. 2003. “A Stochastic Model of Seas. Regional Economic Outlook series. May. Mortality, Fertility, and Human Capital Investment.” Washington, DC: International Monetary Fund. Journal of Development Economics 70 (1): 103-118. ———. 2018a. “Productivity Growth in Asia: Kanwar, S., and S. Sperlich. 2019. "Innovation, Boosting Firm Dynamism and Weeding out the Productivity and Intellectual Property Reform in an Zombies.” In Asia at the Forefront: Growth Challenges Emerging Market Economy: Evidence from India." for the Next Decade and Beyond. Regional Economic Empirical Economics. https://doi.org/10.1007/s00181- Outlook series. October. Washington, DC: 019-01707-3. International Monetary Fund. Kaplan, D. 2009. “Job Creation and Labor Reform in ———. 2018b. Middle East and Central Asia. Regional Latin America.” Journal of Comparative Economics 37 Economic Outlook series. November. Washington, (1): 91-105. DC: International Monetary Fund. Kaufmann, D., A. Kraay, and M. Mastruzzi. 2010. ———. 2019a. Promoting Inclusive Growth in the “The Worldwide Governance Indicators: Methodology Caucasus and Central Asia. Washington, DC: and Analytica Issues.” Policy Research Working Paper International Monetary Fund. 5430, World Bank, Washington, DC. ———. 2019b. “Financial Inclusion of Small and Kaufmann, D., A. Kraay, and M. Mastruzzi. 2010. Medium-Sized Enterprises in the Middle East and "Response to ‘What Do the Worldwide Governance Central Asia.” IMF Departmental Paper 19/02, Indicators Measure?’" The European Journal of International Monetary Fund, Washington, DC. Development Research 22 (1): 55-58. 168 C H A P TE R 2 G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 Keefer, P., and S. Knack. 2007. "Boondoggles, Rent- Developing Countries.” Policy Research Working seeking, and Political Checks and Balances: Public Paper 8308, World Bank, Washington, DC. Investment under Unaccountable Governments." The Review of Economics and Statistics 89 (3): 566-572. Krishna, K. L., D. K. Das, A. A. Erumdan, S. Aggarwal, and P. C. Das. 2016. “Productivity Dynamics in Keller, J., and M. K. Nabli. 2002. “The India’s Service Sector: An Industry-Level Perspective.” Macroeconomics of Labor Market Outcomes in Working Paper 261. Center for Development MENA over the 1990s: How Growth Has Failed to Economics, Delhi School of Economics, Delhi. Keep Pace with a Burgeoning Labor Market.” Working Paper 71, Egyptian Center for Economic Studies, Kraay, A. 2018. “Methodology for a World Bank Cairo. Human Capital Index.” Policy Research Working Paper 8593, World Bank, Washington, DC. Khan, T., T. Nguyen, F. Ohnsorge, and R. Schodde. 2016. “From Commodity Discovery to Production.” La Porta, R., and A. Scheifler. 2014. “Informality and Policy Research Working Paper 7823, World Bank, Development.” Journal of Economic Perspectives 28 (3): Washington, DC. 109-126. Khera, P. 2018. “Closing Gender Gaps in India: Does Laget, E., A. Osnago, N. Rocha, and M. Ruta. 2018. Increasing Women’s Access to Finance Help?” IMF “Deep Trade Agreements and Global Value Chains.” Working Paper 18/212, International Monetary Fund, Policy Research Working Paper 8491, World Bank, Washington, DC. Washington, DC. Kim, Y. E., and N. V. Loayza. 2017. “Productivity Lall, S., Z. Shalizi, and U. Deichmann. 2001. Determinants: Innovation, Education, Efficiency, “Agglomeration Economies and Productivity in Indian Infrastructure, and Institutions.” World Bank, Industry.” Journal of Development Economics 73 (2): Washington, DC. 643-673. Kim, Y. E., and N. V. Loayza. 2019. “Productivity Growth: Patterns and Determinants Across the Lederman, D., and W. F. Maloney. 2007. Natural World.” Policy Research Working Paper 8852, World Resources: Neither Curse nor Destiny. Palo Alto, CA: Bank, Washington, DC. Stanford University Press. Kim, Y. E., N. V. Loayza, and C. Meza-Cuadra. Lederman, D., J. Messina, S. Pienknagura, and J. 2016. “Productivity as the Key to Economic Growth Rigolini. 2014. Latin American Entrepreneurs: Many and Development.” Research & Policy Brief 3, World Firms but Little Innovation. Washington, DC: World Bank, Washington, DC. Bank. Kinda, T. 2019. “E-commerce as a Potential New Lehne, J., J. Mo, and A. Plekhanov. 2014. “What Engine for Growth in Asia.” IMF Working Paper Determines the Quality of Economic Institutions? 19/135, International Monetary Fund, Washington, Cross-Country Evidence.” EBRD Working Paper 171, DC. European Bank for Reconstruction and Development, London. Klapper, L., C. Richmond, and T. Tran. 2013. “Civil Conflict and Firm Performance: Evidence from Cote Levenko, N., K. Oja, and K. Staehr. 2019. “Total d'Ivoire.” Policy Research Working Paper 6640, World Factor Productivity Growth in Central and Eastern Bank, Washington, DC. Europe Before, During and After the Global Financial Crisis.” Post-Communist Economies 31 (2): 137-160. Kochhar, K., K. Utsav, R. Rajan, A. Subramanian, and I. Tokatlidis. 2006. “India’s Pattern of Development: Liu, Y., and V. Steenbergen. 2019. “The Role of FDI What Happened, What Follows?” Journal of Monetary in Global Value Chains (GVCs): Implications for Economics 53 (5): 981-1019. SubSaharan Africa.” Unpublished working paper, Kose, M. A., and F. Ohnsorge. 2019. A Decade since the World Bank, Washington, DC. Global Recession: Lessons and Challenges for Emerging Loayza, N. V., L. Servén, and N. Sugawara. 2010. and Developing Economies. Washington, DC: World “Informality in Latin America and the Caribbean.” In Bank. Business Regulation and Economic Performance, edited by Kouamé, W., and S. Tapsoba. 2018. “Structural N. Loayza and L. Servén, 157–196. Washington, DC: Reforms and Firms' Productivity: Evidence from World Bank. G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 C H A P TE R 2 169 Lopez-Calix, J., T. G. Srinivasan, and M. Waheed. Mercer-Blackman, V., and C. Ablaza. 2018. “The 2012. "What Do We Know About Growth Patterns in Servicification of Manufacturing in Asia: Redefining Pakistan?" Policy Paper Series on Pakistan (PK 05/12), the Sources of Labor Productivity.” ADBI Working World Bank, Washington, DC. Paper 902, Asian Development Bank Institute, Tokyo. Lovász, A. 2016 “Childcare Expansion and Mothers’ Messina, J., and J. Silva. 2019. “Twenty Years of Wage Employment in Post-Socialist Countries.” IZA World Inequality in Latin America.” Policy Research Working of Labor, Bonn, Germany. https://wol.iza.org/articles/ Paper 8995, World Bank, Washington, DC. childcare-expansion-and-mothers-employment-in-post- socialist-countries/long. Mitra, A. 2016. “Productivity Growth in India: Determinants and Policy Initiatives Based on the Lovász, A., and A. Szabó-Morvai 2013. “Does Existing Literature.” United Nations Working Paper Childcare Matter for Maternal Labor Supply? Pushing 16/08, United Nations, New York. the Limits of the Regression Discontinuity Framework.” Budapest Working Papers on the Labour Mitra, P., A. Hosny, G. Minasyan, M. Fischer, and Market 1313, Hungarian Academy of Sciences, G. Abajyan. 2016. “Avoiding the New Mediocre: Budapest. Raising Long-Term Growth in the Middle East and Central Asia.” Middle East and Central Asia Maiti, D. 2019. "Trade, Labor Share, and Productivity Departmental Paper 16/1, International Monetary in India’s Industries." ADBI Working Paper 926, Asian Fund, Washington, DC. Development Bank Institute, Tokyo. Montalbano, P., S. Nenci, and C. Pietrobelli. 2018. Malik, M., and T. Masood. 2018. “Economic Growth, “Opening and Linking Up: Firms, GVCs, and Productivity and Convergence of the Middle East and Productivity in Latin America.” Small Business North African Countries.” MPRA Paper 87882, Economics 50 (4): 917–35. Munich Personal RePEc Archive, Munich. Morsy, H., A. Levy, and C. Sanchez. 2014. “Growing Mallick, J. 2017. “Structural Change and Productivity Without Changing: A Tale of Egypt’s Weak Growth in India and the People’s Republic of China.” Productivity Growth.” EBRD Working Paper 172, ADBI Working Paper 656, Asian Development Bank European Bank for Reconstruction and Development, Institute, Tokyo. London. Mansour, W., A. N. Maseeh, and B. Celiku. 2019. Munoz, M. R., V. Perotti, J. Revilla, N. V. Loayza, T. “Iraq Economic Monitor: Turning the Corner - Sharifuddin, S. Binti, J. Kunicova, et al. 2016. “The Sustaining Growth and Creating Opportunities for Quest for Productivity Growth.” Malaysia Economic Iraq’s Youth.” World Bank, Washington, DC. Monitor. Washington, DC: World Bank. Mason, A. D., and S. Shetty. 2019. A Resurgent East Nabli, M. 2007. Breaking the Barriers to Higher Asia: Navigating a Changing World. World Bank East Economic Growth: Better Governance and Deeper Asia and Pacific Regional Report. Washington, DC: Reforms in the Middle East and North Africa. World Bank. Washington, DC: World Bank. McMillan, M., D. Rodrik, and I. Verduzco-Gallo. Nguimkeu, P. E. 2015. “An Estimated Model of 2014. “Globalization, Structural Change, and Informality with Constrained Entrepreneurship.” Productivity Growth, with and Update on Africa.” Working Paper, Georgia State University. World Development 63 (C): 11-32. Mensah, E. B., S. Owusu, N. Foster-McGregor, and A. Nicoletti, G., and S. Scarpetta. 2005. "Regulation and Szirmai. 2018. “Structural Change, Productivity Economic Performance: Product Market Reforms and Growth and Labor Market Turbulence in Africa.” Productivity in the OECD," OECD Economics UNU-MERIT Working Paper 2018-25, University in Department Working Paper 460, Organisation for Maastricht, Netherlands. Economic Co-operation and Development, Paris. Mensah, E. B., and A. Szirmai. 2018. “Africa Sector O’Sullivan, M., A. Rao, R. Banerjee, K. Gulati, and M. Database (ASD): Expansion and Update”. UNU- Vinez. 2014. “Levelling the Field: Improving MERIT Working Paper 2018-20, University in Opportunities for Women Farmers in Africa.” World Maastricht, Netherlands. Bank and ONE Campaign, Washington, DC. 170 C H A P TE R 2 G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 OECD (Organisation for Economic Co-operation and Raggl, A. 2015. “Determinants of Total Factor Development). 2016. “The Best Versus the Rest: The Productivity in the Middle East and North Africa.” Global Productivity Slowdown, Divergence Across Review of Middle East Economic and Finance 11 (2): Firms and the Role of Public Policy.” OECD 119-143. Productivity Working Paper 5, Organisation for Economic Co-operation and Development, Paris. Rahmati, M., and A. Pilehvari. 2017. “The Productivity Trend in Iran: Evidence from ———. 2018a. OECD Economic Survey of Costa Rica: Manufacturing Firms.” Economics of Transition and Research Findings on Productivity. Paris: Organisation Institutional Change 27 (1): 395-408. for Economic Co-operation and Development. RBI (Reserve Bank of India). 2019. Financial Stability ———. 2018b. Boosting Productivity and Inclusive Report. July 2019. Mumbai: Reserve Bank of India. Growth in Latin America. Paris: Organisation for Economic Co-operation and Development. Reinhart, C. M., and K. S. Rogoff. 2009. This Time is Different: Eight Centuries of Financial Folly. Princeton, Oosthuizen, M., K. Lilenstein, F. Steenkamp, and A. NJ: Princeton University Press. Cassim. 2016. “Informality and Inclusive Growth in Sub-Saharan Africa.” ELLA Regional Evidence Paper, Rodríguez-Pose, A., and T. Ketterer. 2019. ELLA Network, Lima. “Institutional Change and the Development of Lagging Regions in Europe.” Regional Studies. https:// Orlic, E., I. Hashi, and M. Hisarciklilar. 2018. “Cross doi.org/10.1080/00343404.2019.1608356. Sectoral FDI Spillovers and Their Impact on Manufacturing Productivity.” International Business Rodrik, D. 1999. “Where Did All the Growth Go? Review 27 (4): 777-796. External Shocks, Social Conflict, and Growth Collapses.” Journal of Economic Growth 4 (4): Page, J. 1994. "The East Asian Miracle: Four Lessons 385–412. for Development Policy." NBER Macroeconomics Annual 1994 9: 219-282. ———. 2016a. “Premature Deindustrialization.” Pagés-Serra, C., ed. 2010. The Age of Productivity: Journal of Economic Growth 21 (1): 1–33. Transforming Economies from the Bottom Up. ———. 2016b. “An African Growth Miracle?” Journal Washington, DC: Inter-American Development Bank. of African Economies 27 (1): 10-27. Papageorgiou, C., and N. Spatafora. 2012. “Economic Rodrik, D. and A. Subramanian. 2004. "From 'Hindu Diversification in LICs: Stylized Facts and Growth' to Productivity Surge: The Mystery of the Macroeconomic Implications.” IMF Staff Discussion Indian Growth Transition." IMF Working Paper Note 12/13, International Monetary Fund, 04/77, International Monetary Fund, Washington, Washington, DC. DC. Park, J. 2010. "Projection of Long-Term Total Factor Rodrik, D., A. Subramanian, and F. Trebbi. 2004. Productivity Growth for 12 Asian Economies." ADB “Institutions Rule: The Primacy of Institutions Over Economics Working Paper 227, Asian Development Geography and Integration in Economic Bank, Manila. Development.” Journal of Economic Growth 9 (2): 131– Peralta-Alva, A., and A. Roitman. 2008. “Technology 65. and the Future of Work.” IMF Working Paper 19/207, Rovo, N. 2019. "Structural Reforms to Set the Growth International Monetary Fund, Washington, DC. Ambition: New Growth Agenda.” World Bank, Purfield C., H. Finger, K. Ongley, B. Baduel, C. Washington DC. Castellanos, G. Pierre, V. Stepanyan, and E. Roos. 2018. “Opportunity for All: Promoting Growth and Rozenberg, J., and M. Fay, eds. 2019. Beyond the Gap: Inclusiveness in the Middle East and North Africa.” How Countries Can Afford the Infrastructure They Need IMF Departmental Paper 18/11, International While Protecting the Planet. Washington, DC: World Monetary Fund, Washington, DC. Bank. Raballand, G., S. Refas, M. Beuran, and G. Isik. 2012. Russell, M. 2018. “Seven Economic Challenges for Why Does Cargo Spend Weeks in Sub-Saharan African Russia: Breaking Out of Stagnation?” European Ports? Lessons from Six Countries. Washington, DC: Parliament Think Tank In-Depth Analysis Paper, World Bank. European Parliament, Brussels. G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 C H A P TE R 2 171 Ruta, M., I. C. Constantinescu, and A. Mattoo. 2017. Sharafudheen, T. 2017. South Asia Gender Initiative. “Does Vertical Specialization Increase Productivity?” Washington, DC: World Bank. Policy Research Working Paper 7978, World Bank, Washington DC. Shepotylo, O., and V. Vakhitov. 2015. “Services Liberalization and Productivity of Manufacturing Ryzhenkov, M. 2016. “Resource Misallocation and Firms: Evidence from Ukraine.” Economics of Manufacturing Productivity: The Case of Ukraine,” Transition 23 (1): 1-44. Journal of Comparative Economics 44 (1): 41-55. Shiferaw, A., M. Söderbom, E. Siba, and G. Alemu. Saleem, H., M. Shahzad, M. B. Khan, and B. A. 2015. “Road Infrastructure and Enterprise Dynamics in Khiliji. 2019. “Innovation, Total Factor Productivity Ethiopia.” The Journal of Development Studies 51 (11): and Economic Growth in Pakistan: A Policy 1541-1558. Perspective.” Journal of Economic Structures. 8 (1): 1-18. Sinha, R., and X. Xi. 2018. “Agronomic Endowment, Samargandi, N. 2018. “Determinants of Labor Crop Choice and Agricultural Productivity.” Back- Productivity in MENA Countries.” Emerging Markets ground paper prepared for the AFRCE project Finance and Trade 54 (5): 1063–81. Boosting Productivity in Sub-Saharan Africa, World Bank, Washington, DC. Satpathy, L., B. Chatterjee, and J. Mahakud. 2017. “Financial Constraints and Total Factor Productivity: Steinbach, M. R. 2019. “Growth in Low-Income Evidence from Indian Manufacturing Companies.” Countries: Evolution, Prospects, and Policies.” Policy Journal of Management Research 17 (3): 146-162. Research Working Paper 8949, World Bank, Washington, DC. SBP (State Bank of Pakistan). 2019. The State of Pakistan’s Economy—Third Quarterly Report 2018- Stocker, M., J. Baffes, Y. Some, D. Vorisek, and C. 2019. Islamabad: State Bank of Pakistan. Wheeler. 2018. “The 2014-16 Oil Price Collapse in Retrospect: Sources and Implications.” Policy Research Schiantarelli, F., and V. Srivastava. 1997. “Debt Working Paper 8419, World Bank, Washington, DC. Maturity and Firm Performance: A Panel Study of Indian Companies.” Policy Research Working Paper Straus, S. 2012. “Wars Do End! Changing Patterns of 1724, World Bank, Washington, DC. Political Violence in Sub-Saharan Africa.” African Affairs 111 (443): 179-201. Schiffbauer, M., and J. Sampi. 2019. “Enforcing Competition and Firm Productivity: Evidence from Subramanian, A., and M. Kessler. 2013. “The 1,800 Peruvian Municipalities.” Policy Research Hyperglobalization of Trade and Its Future,” Working Working Paper 8714, World Bank, Washington, DC. Paper 13-6, Peterson Institute for International Economics, Washington. Schiffbauer, M., A. Sy, S. Hussain, H. Sahnoun, and P. Keefer. 2015. “Jobs or Privileges: Unleashing the Suri, T., and W. Jack. 2016. “The Long-Run Poverty Employment Potential of the Middle East and North and Gender Impacts of Mobile Money.” Science 354 Africa.” MENA Development Report. Washington, DC: (6317): 1288–92. World Bank. Tamirisa, N., and C. Duenwald. 2018. “Public Wage Schneider, F., A. Buehn, and C. Montenegro. 2010. Bills in the Middle East and North Africa.” Middle “Shadow Economies All over the World: New East and Central Asia Department, International Estimates for 162 Countries from 1999 to 2007.” Monetary Fund, Washington, DC. Policy Research Working Paper 5356, World Bank, Washington, DC. Timmer, M. P., and A. Szirmai. 2000. “Productivity Growth in Asian Manufacturing: The Structural Bonus Schodde, R. 2013. “The Impact of Commodity Prices Hypothesis Examined.” Structural Change and and Other Factors on the Level of Exploration.” Minex Economic Dynamics 11 (4): 371-392. Consulting seminar series, South Yarra, Australia. Topalova, P., and A. Khandelwal. 2011. “Trade Sengupta, R., and A. S. Gupta. 2015. "Capital Flows Liberalization and Firm Productivity: The Case of and Capital Account Management in Selected Asian India.” The Review of Economics and Statistics 93 (3): Economies.” New Thinking and the New G20 series. 995-1009. No. 9. Waterloo, Canada: The Centre for International Governance Innovation. Torlese, H., and D. Raju. 2018. “Feeding of Infants 172 C H A P TE R 2 G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 and Young Children in South Asia.” Policy Research Goals: The World Bank Group’s Contributions.” Working Paper 8655, World Bank, Washington, DC. Policy Research Working Paper, World Bank, Washington, DC. Transparency International. 2019. Corruption Perceptions Index 2018. Accessed July 23. https:// Wamboye, E., K. Tochkov, and B. S. Sergi. 2015. www.transparency.org/cpi2018. “Technology Adoption and Growth in Sub-Saharan African Countries.” Comparative Economic Studies 57 Trushin, E. 2018. “Growth and Job Creation in (1): 136-167. Uzbekistan: An In-Depth Diagnostic.” World Bank, Washington, DC. Wang, Y. 2016. “What Are the Biggest Obstacles to Growth of SMEs in Developing Countries?—An Tuan, T., L. F. Y. Ng, and B. Zhao. 2009. “China’s Empirical Evidence from an Enterprise Survey.” Borsa Post-Economic Reform Growth: The Role of FDI and Istanbul Review 16 (3): 167-176. Productivity Progress.” Journal of Asian Economics 20 (3): 280-293. World Bank. 1995. Bureaucrats in Business: The Economics and Politics of Government Ownership. UNCTAD (United Nations Conference on Trade and Washington, DC: World Bank. Development). 2019. “Trade and Trade Diversion Effects of United States Tariffs on China.” Research ———. 2003. Trade, Investment and Development in Paper 37, Geneva, UNCTAD. the Middle East and North Africa—Engaging with the World. Washington, DC: World Bank. Unel, B. 2003. "Productivity Trends in India’s Manufacturing Sectors in the Last Two Decades." IMF ———. 2008. Unleashing Prosperity: Productivity Working Paper 03/22, International Monetary Fund, Growth in Eastern Europe and the Former Soviet Union. Washington, DC. Washington, DC: World Bank. UNHCR (United Nations High Commissioner for ———. 2012. World Development Report 2012: Gender Refugees). 2019. Global Trends: Forced Displacement in Equality and Development. Washington, DC: World 2018. Geneva: United Nations High Commissioner for Bank. Refugees. ———. 2013a. Micro, Small, and Medium Enterprise UNICEF (United Nations Children’s Fund). 2019. Finance: Women-Owned Business in India. Washington, Primary Education. Accessed November 12, 2019. DC: World Bank. https://data.unicef.org/topic/education/primary-educati on/. ———. 2013b. Enterprise Surveys: Nepal Country Highlights 2013. Enterprise Surveys Country Highlights. Vagliasindi, M. 2013. Revisiting Public-Private Washington, DC: World Bank. Partnerships in the Power Sector. Washington, DC: World Bank. ———. 2013c. Enterprise Surveys: Bangladesh Country Highlights 2013. Enterprise Surveys Country Highlights. Van Eyden, R., M. Difeto, R. Gupta, and M. E. Washington, DC: World Bank. Wohar. 2019. “Oil Price Volatility and Economic Growth: Evidence from Advanced Economies Using ———. 2014. Turkey’s Transition. Washington, DC: More than a Century’s Data.” Applied Energy 233-234 World Bank. (January): 612-621. ———. 2015a. East Asia’s Changing Urban Landscape: Vargas, M. 2015. “Informality in Paraguay: Macro- Measuring a Decade of Spatial Growth. Washington, Micro Evidence and Policy Implications.” IMF DC: World Bank. Working Paper 15/245, International Monetary Fund, Washington, DC. ———. 2015b. Golden Aging: Prospects for Healthy, Active, and Prosperous Aging in Europe and Central Asia. Virmani, A., and D. Hashim. 2011. “J-Curve of Europe and Central Asia Studies. June. Washington Productivity and Growth: Indian Manufacturing Post- DC: World Bank. Liberalization.” IMF Working Paper 11/163, International Monetary Fund, Washington, DC. ———. 2015c. Côte d’Ivoire Systematic Country Diagnostic: From Crisis to Sustained Growth—Priorities Vorisek, D., and S. Yu. Forthcoming. “Understanding for Ending Poverty and Boosting Shared Prosperity. the Cost of Achieving the Sustainable Development Washington, DC: World Bank. G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 C H A P TE R 2 173 ———. 2016a. Global Economic Prospects: Spillovers ———. 2018b. East Asia and Pacific Economic Update: amid Weak Growth. January. Washington, DC: World Enhancing Potential. April. Washington, DC: World Bank. Bank. ———. 2016b. Thailand Economic Monitor. Aging ———. 2018c. Growth and Productivity in the Society and Economy. Washington, DC: World Bank. Philippines: Winning the Future. Washington, DC: World Bank Group. ———. 2016c. World Development Report 2016: Digital Dividends. Washington, DC: World Bank. ———. 2018d. China—Systematic Country Diagnostic: Towards a More Inclusive and Sustainable Development. ———. 2016d. Russian Federation—Systematic Washington, DC: World Bank. Country Diagnostic: Pathways to Inclusive Growth. Washington, DC: World Bank. ———. 2018e. “Productivity Unplugged: The Challenges of Malaysia’s Transition into a High- ———. 2016e. Poverty and Shared Prosperity 2016: Income Country.” The Malaysian Development Taking on Inequality. Washington, DC: World Bank. Experience Series. Washington, DC: World Bank ———. 2016f. What’s Holding Back the Private Sector Group. in MENA? Lessons from the Enterprise Survey. ———. 2018f. Europe and Central Asia Economic Washington, DC: World Bank. Update: Cryptocurrencies and Blockchain. May. ———. 2016g. South Asia Regional Gender Action Plan Washington, DC: World Bank. (RGAP): FY16-FY21—June 2016. Washington, DC: ———. 2018g. Jobs in the Kyrgyz Republic. World Bank. Washington, DC: World Bank. ———. 2016h. Africa’s Pulse: An Analysis of Issues ———. 2018h. Poverty and Shared Prosperity 2018: Shaping Africa’s Economic Future. October. Piecing Together the Poverty Puzzle. Washington, DC: Washington, DC: World Bank. World Bank. ———. 2017a. “Cities in Eastern Europe and Central ———. 2018i. “Dominican Republic: Systemic Asia: A Story of Urban Growth and Decline.” World Country Diagnostic.” World Bank, Washington, DC. Bank, Washington, DC. ———. 2018j. “Paraguay Systemic Country ———. 2017b. At a Crossroads: Higher Education in Diagnostic.” World Bank, Washington, DC. Latin America and the Caribbean. Washington, DC: World Bank. ———. 2018k. World Development Report 2018: Learning to Realize Education’s Promise. Washington, ———. 2017c. The Global Findex Database 2017: DC: World Bank. Measuring Financial Inclusion and the Fintech Revolution. Washington, DC: World Bank. ———. 2018l. Expectations and Aspirations: A New Framework for Education in the Middle East and North ———. 2017d. Gulf Economic Monitor. Sustaining Africa. Washington, DC: World Bank. Fiscal Reforms in the Long Term. June. Washington, DC: Washington DC. ———. 2018m. World Development Report 2019: The Changing Nature of Work. Washington, DC: World ———. 2017e. South Asia’s Turn: Policies to Boost Bank. Competitiveness and Create the Next Export Powerhouse. South Asia Development Matters. Washington, DC: ———. 2018n. India Gender Portfolio Review 2018. World Bank. Part-2: Sectoral Findings Vol. 2. Washington, DC: World Bank. ———. 2017f. Doing Business 2018: Reforming to Create Jobs. Washington, DC: World Bank. ———. 2018o. Global Economic Prospects: The Turning of the Tide? June. World Bank: Washington, ———. 2017g. Nigeria Bi-Annual Economic Update: DC. Fragile Recovery. April. Washington, DC: World Bank. ———. 2018p. Global Investment Competitiveness ———. 2018a. Global Economic Prospects: Broad-Based Report 2017/2018: Foreign Investor Perspectives and Upturn, but for How Long? January. Washington, DC: Policy Implications. Washington, DC: World Bank. World Bank. 174 C H A P TE R 2 G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 ———. 2019a. Global Economic Prospects: Heightened ———. 2019q. Gulf Economic Update: Economic Tensions, Subdued Investment. June. Washington, DC: Diversification for a Sustainable and Resilient GCC. World Bank. World Bank, Washington, DC. ———. 2019b. East Asia and Pacific Economic Update: ———. 2019r. The Middle East and North Africa: Weathering Growing Risks. October. Washington, DC: From Transition to Transformation. Washington, DC: World Bank. World Bank. ———. 2019c. China Economic Update. December. ———. 2019s. South Asia Economic Focus. Making Washington, DC: World Bank. (De)centralization Work. Fall. Washington, DC: World Bank. ———. 2019d. World Development Report 2020: Trading for Development in the Age of Global Value ———. 2019t. Sri Lanka Development Update. Chains. Washington DC: World Bank. October. Washington, DC: World Bank. ———. 2019e. A Decade Since the Global Recession: ———. 2019u. Bangladesh Development Update. Lessons and Challenges for Emerging and Developing October. Washington, DC: World Bank. Economies. Washington, DC: World Bank. ———. 2019v. South Asia Economic Focus. Exports ———. 2019f. Global Economic Prospects: Darkening Wanted. Spring. Washington, DC: World Bank. Skies. January. Washington DC: World Bank. ———. 2019w. Rwanda Systematic Country Diagnostic. ———. 2019g. Europe and Central Asia Economic Washington, DC: World Bank. Update, Fall 2019: Migration and Brain Drain. October. Washington, DC: World Bank. ———. 2019x. Liberia Growth and Economic Diversification Agenda Productivity-driven Growth and ———. 2019h. Western Balkans Regular Economic Diversification. Washington, DC: World Bank. Report: Rising Uncertainty. Fall. October. Washington, DC: World Bank. ———. 2019y. Nigeria Bi-Annual Economic Update: Jumpstarting Inclusive Growth—Unlocking the ———. 2019i. Country Economic Memorandum: Firm Productive Potential of Nigeria’s People and Resource Productivity and Economic Growth in Turkey. Endowments. Fall. Washington, DC: World Bank. Washington, DC: World Bank. ———. 2019z. Boosting Productivity in Sub-Saharan ———. 2019j. Kazakhstan: Reversing Productivity Stag- Africa. Washington, DC: World Bank. nation-Country Economic Memorandum. Washington, DC: World Bank. ———. 2019aa. Africa’s Pulse: An Analysis of Issues Shaping Africa’s Economic Future. October. ———. 2019k. Enterprise Surveys. Accessed Washington, DC: World Bank. November 4, 2019. https://www.enterprisesurveys.org/ en/enterprisesurveys. ———. 2019ab. Lifelines: The Resilient Infrastructure Opportunity. Washington, DC: World Bank. ———. 2019l. Moldova: Rekindling Economic Dynamism. Washington, DC: World Bank. ———. 2019ac. The World Bank Group’s Action Plan on Climate Change Adaptation and Resilience. ———. 2019m. Ukraine Growth: Faster, Lasting and Washington, DC: World Bank. Kinder. Washington, DC: World Bank. ———. 2020. Doing Business 2020. Comparing ———. 2019n. Western Balkans Regular Economic Business Regulation in 190 Economies. Washington, DC: Report: Reform Momentum Needed. Spring. Washington World Bank. DC: World Bank. ———. Forthcoming. Economic Transformation and ———. 2019o. “Global Trade Watch 2018: Trade the Future of Work in LAC. Washington, DC: World amid Tensions.” World Bank, Washington, DC. Bank. ———. 2019p. Trade Integration as a Pathway to World Bank and DRCSC (Development Research Development?: LAC Semiannual Report. Washington, Center of the State Council, the People’s Republic of DC: World Bank. China). 2014. Urban China: Toward Efficient, Inclusive, G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 C H A P TE R 2 175 and Sustainable Urbanization. Washington, DC: World Yousef, T. 2004. “Development, Growth and Policy Bank. Reforms in the Middle East and North Africa.” Journal of Economic Perspectives 18 (3): 91-116. World Bank and DRCSC (Development Research Center of the State Council, the People's Republic of Youssef, H., S. Alnashar, J. Erian, A. Elshawarby, and China). 2019. Innovative China: New Drivers of C. Zaki. 2019. Egypt Economic Monitor: From Floating Growth. Washington, DC: World Bank. to Thriving—Taking Egypt’s Exports to New Levels. Washington, DC: World Bank. World Bank and MPIV (Ministry of Planning and Investment of Vietnam). 2016. Vietnam 2035: Toward Zettelmeyer, J., E. Berglöf, L. Bruynooghe, H. Prosperity, Creativity, Equity, and Democracy. Harmgart, P. Sanfey, and H. Schweiger. 2010. Washington, DC: World Bank. “European Transition at Twenty: Assessing Progress in Countries and Sectors.” Working Paper 91, United Xu, X., and Y. Sheng. 2012. “Productivity Spillovers Nations University—World Institute for Development from Foreign Direct Investment: Firm-Level Evidence Economics Research, Helsinki. from China.” World Development 40 (1): 62-74. SPECIAL FOCUS 2 Low for How Much Longer? Inflation in Low-Income Countries G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 S P EC IAL FO CU S 2 179 Inflation in low-income countries (LICs) has declined sharply to a median of 3 percent in mid-2019 from a peak of 25 percent in 1994. The drop has been supported by the move to more flexible exchange rate regimes, greater central bank independence, and a generally more benign external environment since the 1990s. However, low LIC inflation cannot be taken for granted amid mounting fiscal pressures and the risk of exchange rate shocks. To maintain low and stable inflation, monetary and fiscal policy frameworks need to be strengthened and supported by efforts to replace price controls with more efficient policies. Introduction changes more apparent, provides confidence for long-term savers and investors, protects the The number of low-income countries (LICs) has purchasing power of household income and more than halved since 2001. As of 2019, 31 wealth, and enhances financial stability (Easterly countries are classified as “low income” according 2019; Ha, Kose, and Ohnsorge 2019a).2 By to the World Bank definition, down from 64 in contrast, economies that have experienced high 2001, following the graduation of 35 mostly inflation have suffered significantly lower econom- metals-exporting and transition economies to ic growth (Kremer, Bick, and Nautz 2013). middle-income status.1 Today, LICs are predominantly agriculture-based, small, and Low and stable inflation is especially important for fragile, and they tend to have weak institutions LICs, where a large number of the world’s poor (World Bank 2015). All but six are in Sub- reside. Those most at risk are the “near poor”— Saharan Africa. those living on incomes just above $1.90 per day, the World Bank’s threshold for extreme poverty. LICs have made large strides in price stabilization (The very poorest households hold few nominal over the past five decades, with sharp declines in assets or incomes that would be affected by inflation levels and volatility (Figure SF2.1). That inflation.) Poorer households—which are more said, the level and volatility of inflation in LICs prevalent in LICs than in the EMDEs—may has remained higher than in advanced economies suffer greater welfare losses from inflation than and other emerging market and developing wealthier households because they are less able to economies (EMDEs) over the past two decades protect the real value of their income and assets (Ha, Ivanova et al. 2019a). Reasons include from the impact of inflation (Ha, Ivanova et al. monetary policy challenges that arise in LICs due 2019b). An erosion of their real incomes and to their volatile economies, pervasive use of assets through inflation could tip these households administered pricing, conflicts among central into extreme poverty.3 In addition, by stabilizing bank policy objectives, weaknesses in monetary output fluctuations that disproportionally hurt the policy transmission, and limited analytical poor, the adoption of a credible monetary policy capacity at central banks. The disinflation in regime that maintains low and stable inflation today’s LICs was also considerably less may help reduce poverty and inequality (Romer pronounced than in the (larger number of) and Romer 1999). EMDEs that were classified as LICs in 2000 but have since achieved middle-income status. 2 Several policy outcomes have improved considerably since the 1990s, including lower inflation, smaller black market premiums, Low inflation has typically been associated with and lesser currency overvaluation (Easterly 2019). more stable output and employment, higher 3 Although the evidence of a positive correlation between output growth and investment, and falling poverty inflation and inequality or poverty is mixed at the aggregate level, the links are better established at the household level (Ha, Ivanova et al. rates. Low and stable inflation makes relative price 2019b). For example, single-country studies on EMDEs, such as India (Datt and Ravallion 1998), the Philippines (Blejer and Guerrero 1990), and Brazil (Ferreira and Litchfield 2001), find that higher inflation is associated with a lower share of income held by the Note: This Special Focus was prepared by Jongrim Ha and poor or higher inequality. Using panel data of 24 developed and 66 Franziska Ohnsorge. developing countries over 1990–2014, Siami-Namini and Hudson 1 In addition, there are two countries (South Sudan and Syrian (2019) similarly find bi-directional Granger causality between Arab Republic) that are newly grouped as LICs in 2019. inflation and income inequality in both groups. 180 S P EC IAL FO CU S 2 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE SF2.1 Inflation in low-income countries and Against this backdrop, this Special Focus delves poverty into the characteristics of LIC inflation, quantifies Inflation and inflation volatility in LICs have declined since 1970, broadly in its drivers, and examines related monetary policy line with other EMDEs. The decline has been broad-based across challenges. Specifically, it discusses the following countries, as well as across components of inflation. Those that have grown to middle-income status have had faster declines in inflation. The questions: remaining LICs feature higher poverty than EMDEs. Those just above the extreme poor level are at risk of being tipped back into poverty when • How has inflation evolved in LICs? inflation erodes the real value of their assets and incomes. • What factors have supported inflation A. Inflation B. Number of LICs by inflation bracket developments in LICs? • What policy challenges do LIC central banks face in managing inflation? Evolution of inflation Among LICs, median inflation has fallen by two- thirds since 1970, to 3.2 percent in mid-2019— broadly in line with inflation developments in C. Inflation D. Inflation volatility other EMDEs). The inflation decline has been broad-based across countries as well as inflation components (e.g., food, energy). As a result, the wide heterogeneity of inflation among LICs in the 1990s has narrowed sharply. 1970s to 1990s. Median inflation among LICs was 9-10 percent over this period. Although broadly in line with inflation in other EMDEs, LIC inflation underwent several spikes (up to 25 E. Inflation in former and current LICs F. Poverty percent), especially in the early 1990s, amid currency crises. In half the years between 1970 and 2000, the majority of LICs had double-digit inflation. Post-2000. Median inflation in LICs has fallen rapidly—to 3.2 percent in mid-2019 from a peak of 25.2 percent in 1994 (Figure SF2.1.A). This decline was broad-based and narrowed some of the wide heterogeneity in inflation among LICs. In Source: Haver Analytics; International Monetary Fund; World Bank. Note: Data for 26 low-income countries and 99 other EMDEs. Inflation refers to year-on-year inflation. one-third of LICs, inflation in mid-2019 was less EMDEs = emerging market and developing economies; LICs = low-income countries. than one-third of its level in 1970. In an even A. Blue lines are cross-country medians of inflation; dashed lines indicate the interquartile range across 26 LICs. 2019 inflation rates are based on year-on-year inflation during the first half of 2019 in larger number (63 percent) of LICs, inflation in 19 LICs. B. Number of LICs in which inflation was in the bracket indicated. Data for 2019 are not yet available mid-2019 was less than one-third of its 1994 level. for some LICs and was not included. C.D. Cross-country medians of inflation (C) or standard deviations of inflation (D). The differences By 2008, hyperinflation episodes in LICs across sample periods are all statistically significant. (inflation in excess of 1,000 percent) had also E. Median inflation across countries. “LICs turned MICs” indicates 33 countries classified as low- income countries in 2000 but classified as middle-income countries as of 2019. “Current LICs” subsided.4 In mid-2019, inflation was in the single indicates 29 low-income countries as of 2019. F. Median share of population in extreme poverty (living on less than $1.90 per day) and near-poverty (living on $1.90-$3.20 per day) in 27 LICs and 109 other EMDEs. Click here to download data and charts. 4 In the 1990s Democratic Republic of Congo and Tajikistan experienced inflation over 1,000 percent. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 S P EC IAL FO CU S 2 181 digits in more than three-quarters of LICs, FIGURE SF2.2 Factors supporting falling inflation in compared with less than one-fifth in 1994 (Figure LICs SF2.1.B). The decline in LIC inflation has been supported by the move to more flexible exchange rate regimes, greater central bank independence, lower Since 1970, core, food price, and energy price government debt, and a more benign external environment. inflation have also declined, as has inflation volatility (Figures SF2.1.C and SF2.1.D). A. Central bank transparency index B. Number of LICs, by exchange rate regime Inflation in non-LIC EMDEs. Although the inflation decline in LICs has been broadly in line with developments in other EMDEs, its level remains well above its counterparts.5 Disinflation in today’s LICs has also fallen short of that among (the larger number of) EMDEs that used to be LICs in 2000 but that have since achieved middle- income status, even though these countries started with lower levels of inflation (Figure SF2.1.E). C. Government debt D. Financial and trade openness Factors supporting inflation developments Since 2000, improvements in LIC policies and a benign global macroeconomic environment have supported the decline in LIC inflation. That said, policy frameworks in the median LIC remain generally weaker than those in other EMDEs. E. Inflation, by country characteristics F. Exchange rate volatility Improved policies. The adoption of more resilient monetary, exchange rate, and fiscal policy frameworks has facilitated more effective control of inflation (Hammond, Kanbur, and Prasad 2009; Taylor 2014). Inflation has tended to be lower in LICs with higher degrees of central bank independence and transparency, lower central bank head turnover, and lower public debt ratios (Easterly 2019; Ha, Kose, and Ohnsorge 2019a). Source: Dincer, Eichengreen, and Geraats (2019); Dreher, Stum, and De Haan (2010); Haver Since 1970, monetary policy frameworks have Analytics; International Monetary Fund (IMF); Shambaugh (2004); World Bank. strengthened in LICs. For example, the index of Note: Data for 28 low-income countries and 96 other EMDEs. EMDEs = emerging markets and developing economies; GDP = gross domestic product; LICs = low-income countries. central bank transparency by Dincer, Eichengreen, A.C. Unweighted averages. A. Central bank transparency index as defined in Dincer, Eichengreen, and Geraats (2019). Data for and Geraats (2019) (available for 9 LICs) doubled 9 LICs and 83 other EMDEs. between 1998, when the series starts, and 2015, B. Exchange rate regime as defined in Shambaugh (2004). C. Data for 2019 are based on IMF (2019). when the series ends (Figure SF2.2.A). In 1970, all D. Median trade openness (measured by trade-to-GDP ratio) and financial openness (international asset and liabilities to GDP) across countries. but three LICs had pegged exchange rates whereas, E. Median year-on-year inflation in LICs during 1998-2018, by country characteristics. “High” indicates above-median financial openness, central bank transparency, and turn-over rate of central bank in 2019, less than half (14 of 29 LICs with governors. “Low” indicates below-median financial openness, central bank transparency, and turn-over rate. F. Exchange rate volatility is the cross-country average of the standard deviation of nominal effective appreciation in 28 low-income countries during each time period. Click here to download data and charts. 5 For instance, inflation remains in double-digits in Ethiopia, mainly due to recent currency depreciation and surging food prices after road disruptions and a drought. 182 S P EC IAL FO CU S 2 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 available data) did (Figure SF2.2.B).6 In addition, disinflation around the global financial crisis and fiscal pressures on monetary policy also appear to oil price plunges in 2014-16 may have added have eased. In part as a result of debt relief downward pressure to inflation in LICs. initiatives, government debt has declined from a peak of 121 percent of GDP in 2000, on average, Monetary policy challenges to 53 percent of GDP in 2019— broadly in line with the average non-LIC EMDE (Figure The level and volatility of inflation in LICs have SF2.2.C).7 remained higher than in advanced economies and other EMDEs over the past two decades. This More benign external environment. LIC difference may reflect monetary policy challenges economies, on average, have become more open to particular to LICs arising from higher economic trade and finance since the 1970s, although they volatility and pervasive use of administered remain less open than other EMDEs (Figure pricing, conflicts among central bank policy SF2.2.D; IMF 2011a). Higher capital account objectives, weaknesses in monetary policy openness, in particular, has been associated with transmission, and limited analytical capacity at lower inflation, whereas there appears to be little central banks (Ha, Ivanova, et al. 2019a). difference between LICs that have been highly open to trade and those that have not (Figure Volatile economies. Policymakers in LICs must SF2.2.E). Despite a growing number of LICs contend with greater economic volatility than switching to floating exchange rate regimes, their counterparts in other countries. This in part exchange rates have been considerably more stable reflects the greater frequency of supply shocks and since 1998 than in the preceding two decades the poorer anchoring of inflation expectations that (Figure SF2.2.F). This has helped lower LIC allow exchange rate fluctuations to spill over into inflation volatility and inflation. inflation. Global inflation cycle. LICs are now more • Supply shocks. LIC economies are particularly integrated into the global economy. As a result, vulnerable to supply shocks, especially LIC inflation has become increasingly weather-related ones. Agriculture sectors tend synchronized with the global inflation cycle. What to be large, poor transport links prevent risk was once a negligible contribution to LIC sharing, and food comprises a large share of inflation, global inflation’s impact on domestic household consumption (Bleaney and inflation has become sizeable, especially since Francisco 2018; Cachia 2014). As a result, 2000 (Ha, Kose, and Ohnsorge 2019b; Parker rainfall appears to have the most pronounced 2018).8 Over the past decade, the global effect on economic growth in EMDEs in Sub- Saharan Africa (Barrios, Bertinelli, and Strobl 2010). 6 Several Sub-Saharan African LICs (as well as some recent low- and middle-income countries) belong to monetary unions (e.g., the • Exchange rate volatility. Exchange rates in West African Economic and Monetary Union, and the Central African Economic and Monetary Community). Many of these LICs LICs tend to be more volatile than those in have also experienced low levels of inflation over the recent decades other EMDEs, in part reflecting their greater (Ha, Kose and Ohnsorge 2019a). frequency of supply shocks. With inflation 7 In addition, the relationship between fiscal position and inflation appears to be non-linear: in a low-inflation environment, expectations poorly anchored, exchange rate fiscal deficits tend to be less inflationary (Catão and Terrones 2005; pass-through also tends to be higher in LICs Lin and Chu 2013). As a result, the current low-inflation than in other EMDEs (Ha, Ivanova et al. environment may help further mute the pressures from fiscal dominance on inflation in LICs. 2019a). 8 Using a dynamic factor model for 99 countries (including 16 LICs), Ha, Kose, and Ohnsorge (2019b) find that the contribution Conflicts among policy objectives. LICs of global inflation factor to domestic inflation variation increased to 17 percent in 2001-17 from a 3-4 percent in 1970s to 1990s. Parker frequently have multiple monetary policy (2018) similarly finds that global inflation accounted for around a objectives, with inflation being only one among quarter of inflation variation in LICs over 2001-2012, compared to several. This in part reflects challenges in its contribution (10-20 percent) in the earlier periods. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 S P EC IAL FO CU S 2 183 formulating an appropriate numerical inflation the exchange rate may be an important policy target for LICs. The threshold at which inflation objective (Buffie et al. 2004; IMF 2015; has clear negative effects on output is significantly Mishkin and Savastano 2001; Taylor 2001). A higher for EMDEs than for advanced economies declared strategy of stabilizing the exchange and varies widely depending on country rate against currencies of trading partners with characteristics (Khan and Senhadji 2001). A a track record of low and stable inflation may survey of low- and lower-middle-income countries well be compatible with achieving domestic that listed price stability as a central bank price stability and the limited international objective, found that most countries did not have financial integration of many LICs may still a numerical inflation target, and those that had afford some room for active monetary policy such a target simply tended to align it with the (Ostry, Ghosh, and Chamon 2012). bank’s inflation forecast (IMF 2015). LICs central However, when currency exposures are high, banks are thus likely to have a broader set of exchange rate pressures may prevent central objectives; the exchange rate is more likely to be a banks from acting to preserve low and stable separate and important policy objective (Berg and inflation. Miao 2010; Rodrik 2018).9 Other objectives may include supporting activity or fiscal sustainability. • Conflicts between inflation and fiscal objectives. For LIC governments with weak revenue- • Conflicts between inflation and output raising capabilities and an absence of well- objectives. To lower inflation after a history of functioning capital markets, inflation may high inflation, the central bank must be become an important source of financing willing to tolerate weak activity perhaps for an fiscal deficits (Baldacci, Hillman, and Kojo extended period. A commitment to lowering 2004). The presence of large fiscal deficits or inflation from a history of high inflation will high government debt in LICs can cause fiscal require the central bank to be willing to policy to rely on accommodative monetary tolerate weak activity perhaps for an extended policy to ensure fiscal sustainability (Baldini period (Kasa 2001; Gemayel, Jahan, and Peter and Poplawski-Ribeiro 2011; Weidmann 2011). However, frequent supply shocks in 2013). In almost every year between 1992 and LIC, for example from the effects of weather 2002, two-thirds of LICs had higher debt-to- events on agricultural production, may raise GDP ratios than the one-third of non-LIC inflation while depressing output (Frankel EMDEs with the highest debt levels. In half 2011).10 Stabilizing inflation in response to the years between 1995 and 2017, the median such supply shocks may thus require failing to fiscal deficit in LICs was above that in non- maintain output (Adam 2011; Bashar 2011; LIC EMDEs. Weak institutions (Bleaney, Nguyen et al. 2017). Morozumi, and Mumuni 2016) and political instability (Aisen and Veiga 2006) may • Conflicts between inflation and exchange rate reinforce the negative association between objectives. In LICs (as in some other EMDEs) budget deficits and price stability. Central banks in LICs are therefore more likely to face conflicts between price stability and pressures 9 Using a heterogeneous structural vector autoregressive model for to maintain low interest rates or provide 105 countries, Ha, Ivanova et al. (2019a) find that core inflation in outright fiscal financing (Mas 1995; Prasad LICs with a floating exchange rate regime is less robust in the face of 2010). external shocks than in countries that fixed exchange rates. In advanced economies and other EMDEs, shocks to global core inflation account for a much larger fraction of the variance of Widespread price controls. Price controls— domestic core inflation in fixed regimes than in floating regimes. typically imposed to protect vulnerable groups— 10 For instance, a poor harvest will tend to increase inflation in the short term while depressing economic activity. Supply shocks are more common in LICs than in other EMDEs thus push inflation and output growth in opposite directions, giving (Special Focus 1). The most frequently used price rise to a conflict between monetary policy’s primary objective of controls in LICs are on basic food stuffs and stabilizing prices and its secondary objectives of supporting growth and maintaining a narrow output gap. petroleum. Since food expenditures represent 184 S P EC IAL FO CU S 2 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 nearly 60 percent of the consumption basket in in LICs are half the share of GDP of other LICs, compared with 42 percent in other EMDEs, EMDEs (Figure SF2.3.A). Only one-third of a significant portion of the basket is therefore adults have a bank account in LICs, compared subject to administered pricing (Laborde, Lakatos with 57 percent in other EMDEs (Figure and Martin 2019).11 Price controls can SF2.3.B). As a result, the financial system has temporarily contribute to price stabilization in only weak links to overall economic activity. LICs, especially for key commodities subject to Around 80 percent of investment in LICs is perceived excessive volatility in international financed internally and three-quarters of firms markets.12 However, this poses monetary policy do not tap banks to finance investment challenges, as well as fiscal and growth challenges (Figures SF2.3.C and SF2.3.D). that can heighten conflicts between monetary policy objectives. • Weak institutions. The institutional and legal environment in LICs—including property Weaknesses in the instruments and transmission rights, accounting and disclosure standards, mechanism of monetary policy. In advanced and contract enforcement—tends to be weak economies and many EMDEs, the key monetary (Beck, Demirgüç-Kunt, and Levine 2004). policy instrument is a short-term interest rate, This makes financial intermediation from most often an interbank rate. An advanced- private savers to private borrowers costly and economy central bank can guide the interbank rate risky, inducing banks to limit this activity and through bank reserves and standing facilities. In to hold safer government securities. LICs, however, interbank markets are typically absent, as are liquid secondary markets in • Preponderance of large firms. Productive government securities, which the central bank activity in LICs is often characterized by a few could seek to influence through open-market large, well-established firms and many very operations (Mishra, Montiel, and Spilimbergo small, opaque, and often unstable ones. The 2012). The government securities market in LICs marginal cost of bank lending to large firms tends to be a primary market in which tends to be lower than that of extending credit counterparties are commercial banks that buy and to small firms. As a result, the volume of hold government securities. Thus, the central lending to large firms may be very insensitive bank often conducts monetary policy by directly to fluctuations in bank funding costs induced lending to and borrowing from the commercial by monetary policy (Mishra and Montiel banking system. However, even the bank lending 2013; Mishra et al. 2014). channel can be impaired in LICs. • Widespread informality. The informal sector • Limited financial inclusion. LICs tend to have accounts for about almost two-fifths of GDP large informal sectors but small formal and 90 percent of employment in the average financial sectors (World Bank 2019). Broad LIC, in part reflecting large agricultural money and domestic credit by financial sector sectors and a high share of unskilled workers (World Bank 2019). Firms in the informal sector have limited access to credit from the banking sector and capital markets, and thus 11 In addition, LICs suffer collateral damage from other countries’ administered prices on food and energy because of the high share of have limited interactions with the formal food and energy in LIC consumption baskets and trade. Volatility in financial sector. This dampens monetary global food and energy commodity prices is amplified when other policy transmission through the formal countries respond to rising global commodity prices by imposing price and other controls to suppress prices in local markets (Laborde, financial system. Lakatos, and Martin 2019). The resulting higher volatility of import prices in LICs complicates central banks’ efforts to maintain low and • Other factors. In addition, the strength of stable inflation. 12 Median food and headline inflation were lower in the 52 monetary transmission in LICs has proven EMDEs with food price controls than in the 23 EMDEs without difficult to estimate because of data such controls. However, the relative price distortions introduced by limitations (Li et al. 2016). What empirical highly restricted food price controls have been associated with high inflation in LICs (Special Focus 1). evidence has been estimated suggests that the G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 S P EC IAL FO CU S 2 185 transmission is weak for several reasons: credit FIGURE SF2.3 Monetary policy challenges in LICs and other financial markets tend to be Financial systems are small and have narrow reach in LICs, and this limits shallow; contract enforceability is limited; monetary policy transmission through the financial sector. Broad money information asymmetries are pervasive; and and domestic credit by the financial sector in LICs are half the share of GDP than in other EMDEs. Only a third of adults have bank accounts in many LICs retain elements of financial LICs, compared with 60 percent in other EMDEs. Around 80 percent of repression in the form of interest rate investment in LICs is financed internally, while less than 20 percent is financed by the banking sector. controls.13 For example, while changes in policy rates tended to be transmitted almost one-for-one into retail bank lending rates in A. Broad money and domestic credit B. Fraction of adults with bank accounts advanced economies, pass-through in EMDEs was only in the range of 30-45 percent (Abuka et al. 2015; Saborowski and Weber 2013). Shortcomings in the analytical capacity of central banks. Because monetary policy affects the economy with lags, an important component of any monetary policy regime is the ability of the central bank to accurately forecast its target variables on the assumption of unchanged policies C. Share of investment that is D. Share of firms that approach banks as well as to assess the effects of policy changes on internally financed to finance investment those variables. Few LIC central banks have the structural models with proven track records required for such forecasts (IMF 2015). This reflects in part lack of relevant historical data, insufficient knowledge about the macroeconomic structure of the economies concerned, rapid structural change in the economy, and shortages of research expertise (Gemayel, Jahan, and Peter 2011; IMF 2015). Source: Enterprise Survey; Global Findex Database; World Development Indicators. Note: EMDEs = emerging markets and developing economies; LICs = low-income countries. Unweighted averages across countries. Complications introduced by globalization. A. Broad monetary and domestic credit provided by financial sector (both percent of GDP) in 2017, based on 20 LICs and 110 other EMDEs. Globalization is likely to alter the monetary B. Proportion of adults (age over 15) holding a bank account in 2017. Survey based on 23 LICs and 86 other EMDEs. transmission mechanism in complicated ways C. Proportion of investment financed internally. Enterprise survey based on 15 LICs and 47 other (Abuka et. al. 2015; Montiel and Pedroni 2018). EMDEs. D. Proportion of firms using banks to finance investments. Enterprise survey based on 15 LICs and It increases the economy’s exposure to external 47 other EMDEs. Click here to download data and charts. shocks, in the form of exogenous changes in the foreign-currency prices of traded goods, remittance flows, and capital flows. It may also alter the trade-offs between different central bank benign or fiscal pressures mount, the ability of objectives. central banks in LICs to maintain low inflation may be tested. Since 2013, government debt has risen rapidly, by almost 15 percentage points of Policy options going GDP in the median LIC; about half of LIC debt forward is external and, hence, predominantly foreign- currency-denominated (World Bank 2019). This Going forward, the achievements of low and stable increases LIC governments’ vulnerability to inflation in many LICs cannot be taken for financial market disruptions that raise borrowing granted. If the external environment turns less costs. Mounting fiscal pressures could heighten tensions between the multiple objectives of LIC 13 For details, see Mishra, Montiel, and Spilimbergo (2012); IMF central banks. Separately, because of poorly (2015); and Mishra and Montiel (2013). anchored inflation expectations, exchange rate 186 S P EC IAL FO CU S 2 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 depreciations following financial market stress that reduce the economy’s vulnerability to could raise LIC inflation. Broader policy efforts shocks, strengthen automatic fiscal stabilizers, aimed at strengthening fiscal and monetary policy increase the flexibility and effectiveness of frameworks, and improving debt management, are discretionary fiscal policy, and increase the therefore required in LICs to safeguard low and flexibility of labor markets. Institutional stable inflation. changes could include entrusting respon- sibility for financial stability to a separate Many of the monetary policy challenges facing supervisory and regulatory authority, LICs are related to their level of economic and associated with a well-capitalized deposit financial development. Addressing these challenges insurance agency. requires a broader development process and includes: the development of financial markets to • Expanding central bank tools. The central bank provide the central bank with more effective could develop or strengthen instruments policy instruments; the improvement of systems separate from monetary policy to address its compiling economic statistics; and capacity objective of financial stability, including development in central banks and economic capital flow management measures and ministries, including strengthening economic macroprudential policies. expertise. • Considering best suitable nominal anchors for • Strengthening central bank independence. monetary policy. Although inflation targeting, Central bank independence has increased with its usual focus on the CPI, has been the among LICs since the early 1990s, partly as a most popular among advanced economies and means to allow central banks to give primacy larger EMDEs, other EMDEs and LICs could to price stability over other objectives and consider alternative nominal anchors for enhance their credibility (Dincer and monetary policy that best suit their economic Eichengreen 2014; Garriga 2016). However, structures. For example, countries that central bank independence of lower- and produce commodities that are subject to middle-income countries remains less than in volatile global commodity prices, and have other EMDEs and advanced economies, and procyclical access to global capital markets, de jure independence does not necessarily could target export prices or producer prices. translate into de facto independence (IMF These targets may stabilize output better than 2019). CPI targeting in the presence of frequent terms of trade or financial shocks (Frankel • Clarifying priorities in central banks’ objectives. 2011). A transparent prioritization of central bank objectives in the event of conflicts between • Building and maintaining central bank different objectives could help central banks credibility. The central bank could strengthen achieve their primary targets. Other policy its efforts to convince the public of the options could be developed to help achieving primacy it gives to the low-inflation objective central banks’ secondary objectives— (Mishkin 1997). Declaration of a specific including for output or financial stability—of inflation target could serve this purpose, but monetary policy. Such policies could include this strategy may not yet suit LICs with weak the judicious use of budgetary policy when and uncertain monetary transmission, data there is fiscal space, and structural reforms deficiencies, and limited analytical capacity. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 S P EC IAL FO CU S 2 187 Blejer, M., and I. Guerrero. 1990. “The Impact of References Macroeconomic Policies on Income Distribution: An Empirical Study of the Philippines.” Review of Abuka, C., R. Alinda, C. Minoui, J. L. Peydro, and A. Economics and Statistics 72 (3): 414-23. Presbitero. 2015. “The Bank Lending Channel in a Frontier Economy: Evidence from Loan-Level Data.” Buffie, E., C. Adam, S. O’Connell, and C. Pattillo. https://www.rieti.go.jp/jp/events/15061501/pdf/7-2_pr 2004. “Exchange Rate Policy and the Management of esbitero.pdf. Official and Private Capital Flows in Africa.” IMF Staff Papers 51 (s1): 126-60. Adam, C. 2011. “On the Macroeconomic Management of Food Price Shocks in Low-Income Cachia, F. 2014. “Regional Food Price Inflation Countries.” Journal of African Economies 20 (1): i63- Transmission.” FAO Working Paper ESS/14-01, Food i99. and Agriculture Organization of the United Nations, Rome. Aisen, A., and F. J. Veiga. 2006. “Does Political Instability Lead to Higher Inflation? A Panel Data Catão, L., and M. Terrones. 2001. “Fiscal Deficits and Analysis.” Journal of Money, Credit and Banking 38 (5): Inflation—A New Look at the Emerging Market 1379-89. Evidence.” IMF Working Paper 01/74, International Monetary Fund, Washington, DC. Baldacci, E., A. L. Hillman, and N. C. Kojo. 2004. “Growth, Governance, and Fiscal Policy Transmission Datt, G., and M. Ravallion. 1998. “Why Have Some Channels in Low-Income Countries.” European Journal Indian States Done Better Than Others at Reducing of Political Economy 20 (3): 517-49. Rural Poverty?” Economica 65 (257): 17-38. Baldini, A., and M. Poplawski-Ribeiro. 2011. “Fiscal Dincer, N., and B. Eichengreen. 2014. “Central Bank and Monetary Determinants of Inflation in Low- Transparency and Independence: Updates and New Income Countries: Theory and Evidence from Sub- Measures.” International Journal of Central Banking 10 Saharan Africa.” Journal of African Economies 20 (3): (1): 189-259. 419-62. Dincer, N., B. Eichengreen, and P. Geraats. 2019. Barrios, S., L. Bertinelli, and E. Strobl. 2010. “Trends "Transparency of Monetary Policy in the Post Crisis in Rainfall and Economic Growth in Africa: A World." In Oxford Handbook of the Economics of Neglected Cause of the African Growth Tragedy.” Central Banking, edited by D. Mayes, P. Siklos, and J.- Review of Economics and Statistics 92 (2): 350-66. E. Sturm. New York: Oxford University Press. Bashar, O. H. M. N. 2011. “The Role of Aggregate Dreher, A., J. Sturm, and J. de Haan. 2010. “When Is a Demand and Supply Shocks in a Low-Income Central Bank Governor Replaced? Evidence Based on a Country: Evidence from Bangladesh.” Journal of New Data Set.” Journal of Macroeconomics 32 (3) 766- Developing Areas 44 (2): 243-64. 781. Beck, T., A. Demirgüç-Kunt, and R. Levine. 2004. Easterly, W. 2019. “In Search of Reforms for Growth: “Finance, Inequality and Poverty: Cross-Country New Stylized Facts on Policy and Growth Outcomes.” Evidence.” Policy Research Working Paper 3338, NBER Working Paper 26318, National Bureau of World Bank, Washington, DC. Economic Research, Cambridge, MA. Berg, A., and Y. Miao. 2010. “The Real Exchange Rate Frankel, J. A. 2011. “Monetary Policy in Emerging and Growth Revisited: The Washington Consensus Markets: A Survey.” In Handbook of Monetary Strikes Back?” IMF Working Paper 10/58, Economics 3B, edited by B. M. Friedman and M. International Monetary Fund, Washington, DC. Woodford. New York: Elsevier. Bleaney, M., and M. Francisco. 2018. “Is the Phillips Garriga, A. C. 2016. “Central Bank Independence in Curve Different in Poor Countries?” Bulletin of the World: A New Data Set.” International Interactions Economic Research 70 (1): E17-E28. 42 (5): 849-68. Bleaney, M., A. Morozumi, and Z. Mumuni. 2016. Ha, J., A. Ivanova, P. Montiel, and P. Pedroni. 2019a. “Fiscal Deficits and Inflation: Institutions Matter.” “Inflation in Low-Income Countries.” Policy Research University of Nottingham, U.K. Working Paper 8934. World Bank, Washington, DC. 188 S P EC IAL FO CU S 2 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 Ha, J., A. Ivanova, F. Ohnsorge, and D. F. Unsal. Mas, I. 1995. “Central Bank Independence: A Critical 2019b. “Inflation: Concepts, Evolution, and View from a Developing Country Perspective.” World Correlates.” Policy Research Working Paper 8738, Development 23 (10): 1639-52. World Bank, Washington, D.C. Mishkin, F. 1997. “Strategies for Controlling Ha, J., M. A. Kose, and F. Ohnsorge, eds. Inflation.” NBER Working Paper 6122, National 2019a. Inflation in Emerging and Developing Economies: Bureau of Economic Research, Cambridge, MA. Evolution, Drivers, and Policies. Washington, DC: World Bank. Mishkin, F. S., and M. A. Savastano. 2001. “Monetary Policy Strategies for Latin America.” Journal of ———. 2019b. “Global Inflation Synchronization.” Development Economics 66 (2): 415-444. Policy Research Working Paper 8768. World Bank, Washington, DC. Mishra, P., P. Montiel, and A. Spilimbergo. 2012. Hammond, G., R. Kanbur, and E. Prasad. 2009. “Monetary Transmission in Low-Income Countries: “Monetary Policy Challenges for Emerging Market Effectiveness and Policy Implications.” IMF Economic Economies.” In New Monetary Policy Frameworks for Review 60 (2): 270-302. Emerging Markets: Coping with the Challenges of Financial Globalization, edited by G. Hammond, R. Mishra, P., and P. Montiel. 2013. “How Effective Is Kanbur, and E. Prasad. Cheltenham, U.K.: Edward Monetary Transmission in Low-Income Countries? A Elgar Publishing. Survey of the Empirical Evidence.” Economic Systems 37 (2): 187-216. IMF (International Monetary Fund). 2015. Evolving Monetary Policy Frameworks in Low-Income and Other Mishra, P., P. J. Montiel, P. Pedroni, and A. Developing Countries. Washington, DC: International Spilimbergo. 2014. “Monetary Policy and Bank Monetary Fund. Lending Rates in Low-Income Countries: Heterogeneous Panel Estimates.” Journal of ———. 2019. World Economic Outlook: Growth Development Economics 111 (November): 117-31. Slowdown, Precarious Recovery. April. Washington, DC: International Monetary Fund. Nguyen, A. D. M., J. Dridi, F. Unsal, and O. H. Williams. 2017. “On the Drivers of Inflation in Sub- Kasa, K. 2001. “Will Inflation Targeting Work in Saharan Africa.” International Economics 151 Developing Countries?” FRBSF Economic Letter 2001 (October): 71-84. (1), Federal Reserve Bank of San Francisco, San Francisco, CA. Ostry, J. D., A. R. Ghosh, and M. Chamon. 2012. Khan, M., and A. Senhadji. 2001. “Threshold Effects “Two Targets, Two Instruments: Monetary and in the Relationship between Inflation and Growth.” Exchange Rate Policies in Emerging Economies.” Staff IMF Staff Papers 48 (1): 1-21. Discussion Note 12/01, International Monetary Fund, Washington, DC. Kremer, S., A. Bick, and D. Nautz. 2013. “Inflation and Growth: New Evidence from a Dynamic Panel Parker, M. 2018. “How Global Is ‘Global Inflation’?” Threshold Analysis.” Empirical Economics 44 (2): 861- Journal of Macroeconomics 58 (December): 174-97. 78. Prasad, E. S. 2010. “Financial Sector Regulation and Laborde, D., C. Lakatos, and W. Martin. 2019. Reforms in Emerging Markets: An Overview.” NBER “Poverty Impact of Food Price Shocks and Policies.” Working Paper 16428, National Bureau of Economic Policy Research Working Paper 8724. World Bank, Research, Cambridge, MA. Washington, DC. Rodrik, D. 2018. “The Real Exchange Rate and Li, B. G., S. A. O’Connell, C. Adam, A. Berg, and P. Economic Growth.” Brookings Papers on Economic Montiel. 2016. “VAR Meets DSGE: Uncovering the Activity (Fall): 365-439. Monetary Transmission Mechanism in Low-Income Countries.” IMF Working Paper 16/90, International Saborowski, C., and S. Weber. 2013. “Assessing the Monetary Fund, Washington, DC. Determinants of Interest Rate Transmission through Lin, H.-Y., and H.-P. Chu. 2013. “Are Fiscal Deficits Conditional Impulse Response Functions.” IMF Inflationary?” Journal of International Money and Working Paper 13/23, International Monetary Fund, Finance 32 (C): 214-33. Washington, DC. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 S P EC IAL FO CU S 2 189 Shambaugh, J. C. 2004. “The Effect of Fixed Exchange South Africa and the Challenge of a Changing Rates on Monetary Policy.” Quarterly Journal of Mandate, South African Reserve Bank, Pretoria, Economics 119 (1): 301-52. December 19. Siami-Namini, S., and D. Hudson. 2019, “Inflation Weidmann, J. 2013. “Who Calls the Shots? The and Income Inequality in Developed and Developing Problem of Fiscal Dominance.” Speech at the BdF-BBk Countries.” Journal of Economic Studies 46 (3): 611- Macroeconomics and Finance Conference, Deutsche 632. Bundesbank Eurosystem, Paris. Taylor, J. 2001. “The Roles of Exchange Rates in World Bank. 2015. Global Economic Prospects: Having Monetary-Policy Rules.” American Economic Review 91 Fiscal Space and Using It. January. Washington, DC: (2): 263-67. World Bank. ———. 2014. “Inflation Targeting in Emerging ———. 2019. Global Economic Prospects: Darkening Markets: Le Global Experience.” Speech at the Skies. January. Washington, DC: World Bank. Conference on Fourteen Years of Inflation Targeting in CHAPTER 3 FADING PROMISE How to Rekindle Productivity Growth G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 193 A broad-based slowdown in labor productivity growth has been underway since the global financial crisis. In emerging market and developing economies (EMDEs), the slowdown has reflected weakness in investment and moderating efficiency gains as well as dwindling resource reallocation between sectors. The pace of improvements in key drivers of labor productivity—including education, urbanization, and institutions—has slowed or stagnated since the global financial crisis and is expected to remain subdued. To rekindle productivity growth, a comprehensive approach is necessary: facilitating investment in physical, intangible, and human capital; encouraging reallocation of resources towards more productive sectors; fostering firm capabilities to reinvigorate technology adoption and innovation; and promoting a growth-friendly macroeconomic and institutional environment. Specific policy priorities will depend on individual country circumstances. Introduction • How has the pace of productivity convergence changed? Productivity growth is the primary source of lasting income growth, which in turn is the main • What are the underlying factors associated driver of poverty reduction. Most cross-country with productivity growth? differences in income per capita have been attributed to differences in productivity (Figure • What policy options are available to boost 3.1).1 Whereas the one-quarter of emerging productivity growth? market and developing economies (EMDEs) with Contribution and framework. The chapter makes the fastest productivity growth have reduced their several contributions to the literature and policy extreme poverty rates by an average of more than debate on labor productivity. The framework of 1 percentage point per year since 1981, poverty the analysis in this chapter is as follows: rates rose in EMDEs with productivity growth in the lowest quartile. • EMDE focus. Thus far, the literature has The broad-based slowdown in labor productivity focused on trends in subsets of countries such growth over the past decade has raised concerns as advanced economies, OECD countries or about progress in achieving development goals. In specific regions.2 The chapter is the first to EMDEs, the slowdown puts at risk hard-won provide both an overarching global and in- gains in productivity catch-up to advanced depth EMDE view of productivity trends economies prior to the 2007-09 global financial alongside detailed regional analysis. To crisis. Labor productivity gaps with advanced achieve this, it utilizes a comprehensive economies remain substantial, with workers in the dataset of multiple measures of productivity average EMDE producing less than one-fifth of growth for up to 29 advanced economies and the output of those in advanced economies. 74 EMDEs during 1981-2018. Against this backdrop, this chapter presents a • Multiple approaches. The chapter synthesizes comprehensive examination of the evolution of findings from empirical exercises using productivity, the correlates of productivity macroeconomic, sectoral, and firm-level data improvements, and policy options to rekindle on productivity. Previous studies have productivity growth. Specifically, the chapter typically analyzed productivity using data for addresses the following questions: only one of these three dimensions.3 This • How has productivity growth evolved over the last four decades? 2 For details, see Fernald (2012), Adler et al. (2017), OECD (2015), ADB (2017), Dabla-Norris et al. (2015), Cusolito and Note: This chapter was prepared by Alistair Dieppe and Gene Maloney (2018), World Bank (2018a). 3 For macroeconomic analysis, see Adler et al. (2017) and Kim Kindberg-Hanlon, with contributions from Atsushi Kawamoto, Sinem Kilic Celik, Hideaki Matsuoka, Yoki Okawa, and Cedric and Loayza (2019). For sectoral analysis, see McMillan, Rodrik, and Okou. Research assistance was provided by Khamal Clayton, Aygul Verduzco-Gallo (2014); and McMillan, Rodrik, and Sepulveda Evdokimova, Awais Khuhro, Xinyue Wang, and Heqing Zhao. (2017). For firm-level analysis, see Cirera and Maloney (2017); 1 See for details Caselli (2005) and Hall and Jones (1999). Cusolito and Maloney (2018); and Fuglie et al. (2019). 194 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 3.1 Labor productivity, per capita income and 2016 and since then has remained low, at 1.9 poverty reduction percent in 2018. The post-crisis slowdown has Cross-country differences in labor productivity explain most of the variation been broad-based, affecting nearly 70 percent in income per capita. Poverty declined by more than 1 percentage point on of advanced economies and EMDEs and over average per year in the one-quarter of EMDEs with the highest productivity growth during 1981-2015, while poverty rose in EMDEs with the lowest 80 percent of the global extreme poor and has productivity growth. affected all EMDE regions (Figure 3.2). In advanced economies, the slowdown continues A. Labor productivity and per capita B. Annual change in the poverty rate a trend that has been underway since the late income in EMDEs, by productivity growth 1990s. In EMDEs, which have a history of recurring multi-year productivity growth surges and setbacks, the productivity growth slowdown from peak (6.6 percent in 2007) to trough (3.2 percent in 2015) has been the steepest, longest, and broadest yet. Com- modity-exporting EMDEs—which account for almost two-thirds of EMDEs—have been the worst affected.6 Source: PovcalNet; World Bank. Note: Sample includes 29 advanced economies and 74 EMDEs. A. Income per capita and output per worker measured in US dollars at 2010 prices and exchange • Large labor productivity gaps, slow convergence rates. B. Unweighted averages using annual data during 1981-2015. Fastest-growing EMDEs are those in in EMDEs. Average output per worker in the top quartile by productivity growth; slowest-growing EMDEs are those in the bottom quartile of EMDEs is less than one-fifth of that in the labor productivity growth. Poverty rate defined as the share of the population living on less than $1.90 a day (2011 PPP). average advanced economy, and just 2 percent Click here to download data and charts. in LICs. Although EMDE productivity convergence improved ahead of the global chapter combines these approaches and financial crisis, it is now progressing at rates includes a thorough review of the literature in that would require over a century to halve the each area. current productivity gap with the average • Comprehensive assessment of correlates of advanced economy. However, the pace of productivity growth. The chapter reviews a convergence differs across regions: more than large body of literature on the correlates of half of EMDEs in East Asia and Pacific (EAP) productivity growth. It undertakes an are on course to halve their productivity gap empirical exercise that expands upon previous in less than 40 years, while fewer than 20 work, whose data typically use either a shorter percent of economies in the Middle East and sample or a narrower set of correlates.4 The North Africa (MNA), Latin America and the chapter also quantifies the damage that Caribbean (LAC), and Sub-Saharan Africa financial crises inflict on productivity growth.5 (SSA) will likely achieve the same reduction over this timeframe. Main findings. The following findings emerge • Accounting for the slowdown. Slower capital from the chapter. deepening has accounted for the lion’s share of the post-crisis (2013-18) slowdown in • Broad-based post-crisis decline in labor productivity growth in advanced economies productivity growth. Global labor productivity from pre-crisis averages (2003-08). In growth slowed from its pre-crisis peak of 2.7 EMDEs, subdued investment and slowing percent in 2007 to a trough of 1.5 percent in total factor productivity (TFP) growth have 4 Durlauf, Kourtellos, and Tan (2008); Kim and Loayza (2019); Adler et al. (2017). 6 In commodity-exporting EMDEs, productivity growth slowed by 5 This complements earlier work documenting damage from 4.1 percentage points between 2007 and 2015 to around 0, financial crises to the level of potential output (Cerra and Saxena compared with 3.5 percentage points in commodity-importing 2008) and to potential growth (Furceri and Mourougane 2012a). EMDEs. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 195 accounted, in approximately equal measure, FIGURE 3.2 Global productivity developments for the post-crisis productivity growth A broad-based slowdown in productivity growth has been underway, slowdown. About one-half of the slowdown in affecting the majority of advanced economies and EMDEs. In EMDEs, productivity growth slowed from its most recent peak of 6.6 percent in EMDEs reflects fading gains from the 2007 to 3.2 percent in 2015, the steepest, longest, and broadest slowdown reallocation of resources towards more in 40 years. Productivity levels in EMDEs are less than 20 percent of the productive sectors. Reallocation previously advanced-economy average, and just 2 percent in LICs. The productivity slowdown has coincided with lower gains from sectoral reallocation and a drove more than one-third of pre-crisis slowdown in improvements in many drivers of productivity growth. productivity growth in EMDEs, and three- quarters in LICs. A. Global, advanced-economy, B. Share of economies and global and EMDE productivity growth poor with 2013-18 productivity growth below historical averages • Challenging prospects for labor productivity growth. Since the global financial crisis, improvements in many key correlates of productivity growth in EMDEs have slowed or gone into reverse. Working-age population growth has slowed, educational attainment has stabilized, and the pace of expansion into more diverse and complex forms of production has lost momentum as the growth D. EMDE productivity levels, 2013-18 of global value chains stalled. At the firm C. Magnitude and extent of multi-year productivity slowdowns and level, EMDE firms that are large and export- recoveries oriented are closest to the productivity frontier, suggesting that continued global trade weakness and slower global production integration could be particularly damaging to productivity growth in EMDEs. In addition, the global financial crisis dented productivity growth and momentum has yet to be rebuilt. • Policy priorities. The broad-based nature of the E. Within and between sector F. Share of EMDEs with a post-crisis labor productivity growth slowdown can be contributions to productivity growth slowdown in the growth of underlying drivers of productivity addressed with a comprehensive set of policies. Policies can lift labor productivity economy-wide by stimulating private and public investment, and improving human capital; fostering firm productivity, including by upgrading workforce skills; exposing firms to trade and foreign investment; facilitating the reallocation of resources towards more productive and a more diversified set of Source: World Bank (full sources in subsequent figures). sectors; and creating a generally growth- Note: Productivity is defined as output per worker. Unless otherwise indicated, data are from a sample of 29 advanced economies (AEs) and 74 emerging market and developing economies friendly macroeconomic and institutional (EMDEs). Aggregates are GDP-weighted at constant 2010 prices and exchange rates. B. Percent of economies, or share of global extreme poor (population living on less than $1.90 per environment. day), with productivity growth in 2013-18 below pre-crisis (2003-08) or long-term (1981-2018) average productivity growth. Grey line indicates 50 percent. C. “Magnitude of slowdown” is the cumulative decline in EMDE productivity growth from the peak of Concepts. Throughout this chapter, productivity the episode to the trough for episodes lasting more than two years. “Magnitude of rebound” is the cumulative increase in EMDE productivity growth from the trough (end) of the episode to three years is defined as output (GDP) per input of a unit of later. “Affected EMDEs” is the share of EMDEs that experienced a slowdown. labor. To ensure as large and comparable a sample D. Blue bars show unweighted average output per worker during 2013-18 relative to the advanced- economy average. Whiskers indicate interquartile range relative to the advanced-economy average. as possible over time and across countries, this E. Sample includes 80 economies, including 46 EMDEs (of which 8 are LICs), using data for 1995- 2015. Growth “within sector” shows the contribution to aggregate productivity growth of each sector chapter uses the number of people employed holding employment shares fixed. The ‘between sector’ effect shows the contribution arising from changes in sectoral employment shares. rather than the number of hours worked as the F. Post-crisis slowdown defined as the share of economies where improvements in each underlying driver of productivity during 2008-2017 was less than zero or the pace of improvement during the pre- crisis period 1998-2007. Variables definitions in Chart 3.9.A. Click here to download data and charts. 196 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 measure of labor input.7 A second measure, total product market regulations in parts of Europe.8 factor productivity (TFP), is also featured in the During the global financial crisis, productivity chapter. TFP measures the efficiency with which growth in advanced economies plunged and never factor inputs are combined and is often used to recovered to pre-crisis levels. At 0.8 percent on proxy technological progress (Annex 3.2). average during 2013-18, it was one-half its long- term average and 0.4 percentage points below its Evolution of labor pre-crisis average. This slowdown relative to long- run averages affected nearly 90 percent of productivity growth advanced economies. Since 2007, a broad-based slowdown in labor EMDEs. Productivity growth in EMDEs has productivity growth has been underway that has slowed sharply from its 2007 peak of 6.6 percent reached the majority of advanced economies and to a low of 3.2 percent in 2015 and, since then, EMDEs. For EMDEs, this has partly reversed a pre- has inched up to 3.6 percent in 2018. The post- crisis productivity growth surge, although crisis slowdown from pre-crisis averages affected productivity growth remains above the very weak nearly 70 percent of EMDEs and, in around half rates of the 1980s and 1990s. Some low-income of EMDEs, productivity growth has now fallen countries have escaped the productivity growth below its long-term (1981-2018) average. The slowdown but productivity growth has regressed in slowdown has been particularly pronounced in some fragile and conflict-afflicted low-income China, where a policy-guided decline in public countries. investment growth has been underway for several years, and in commodity exporters, which have Global productivity. From its peak in 2007, been hit hard by the commodity price plunge of global productivity growth has slowed by 0.8 2014-16. Weak post-crisis productivity growth percentage point, to 1.9 percent in 2018. The follows on the heels of a major productivity surge post-crisis (2013-18) average of 1.8 percent was during 2003-08 when EMDE productivity 0.5 percentage point below the pre-crisis (2003- growth more than doubled from 1990s averages, 08) average and slightly below the long-term in part reflecting a strong cyclical rebound from (1981-2018) average (Figure 3.3). This post-crisis the 1997-98 Asian financial crisis. slowdown from pre-crisis averages was broad- based, affecting two-thirds of economies, both Since 1980, EMDE productivity growth has gone advanced economies and EMDEs. Those through three multi-year surges and setbacks in economies with slower post-crisis productivity productivity growth. Previous multi-year growth than during the pre-crisis period account slowdowns—in 1986-1990 and 1995-1998— for over 80 percent of global GDP and the preceded global recessions (1991) or global extreme poor. slowdowns and EMDE crises (1998). However, the slowdown since 2007 has been the most Advanced economies. The post-crisis slowdown in prolonged, steepest and broadest-based yet.9 In advanced-economy productivity growth continues a trend that has been underway since the late 1990s, following a brief resurgence from an even 8 For a summary of the effects of the ICT slowdown on U.S. longer-running negative trend. The slowdown has productivity in the 2000s, see Duval, Hong, and Timmer (2017), Jorgenson, Ho, and Stiroh (2008), and Fernald (2012). In Europe, been attributed to a declining contribution from the trend decline in productivity has been ascribed to sectoral information and communication technology misallocation due to cheap credit in southern Europe (Gopinath et al. 2017), a failure to adopt ICT and associated technology to the same (ICT) intensive sectors in the United States, and extent as the United States (van Ark, O’Mahony, and Timmer 2008), slow adoption of ICT technologies, and restrictive and restrictive product market regulations (Haltiwanger, Scarpetta, and Schweiger 2014). 9 The most recent slowdown in productivity growth has lasted 7 Number of people engaged includes employees and self- eight years—compared with the four years of 1986-90 and the three employed. Alternative measures might better capture labor input but years of 1995-98—and, from peak to trough, has been around 50 have insufficient coverage for EMDEs (Annex 3.1). In countries with percent steeper than the slowdowns in the late 1980s and the late large informal sectors, both employment and output may be subject 1990s. It has reached 64 percent of EMDEs, slightly more than the to sizable measurement error (World Bank 2019a, Annex 3.1). slowdown in the 1990s (59 percent) and 1980s (57 percent). G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 197 contrast to previous episodes, the current FIGURE 3.3 Evolution of global productivity growth productivity slowdown has yet to be marked by a In EMDEs, productivity growth has declined from pre-crisis levels, although strong rebound. it remains strong relative to longer-run averages in half of EMDEs. At 0.6 percent, EMDE commodity exporters have had the weakest average productivity growth since 2013. Productivity growth in EMDE commodity EMDE productivity growth remains slightly importers and LICs has been more resilient. above its average in the 1980s and 1990s, which was well below the pre-crisis surge in productivity A. Global, advanced-economy, and B. EMDE productivity growth growth. In commodity importers, average EMDE productivity growth productivity growth in 2013-18 has remained more than twice its 1980s average and one-third above its 1990s average. However, in commodity- exporting EMDEs, the post-crisis commodity price plunge has returned productivity growth to just 0.6 percent, rates which are weak but still above the growth rates of the 1980s. LICs. On average, LIC productivity growth has fallen only modestly to 2.4 percent during 2013- C. Economies with 2013-18 D. EMDE average productivity growth, pre- and post-crisis productivity growth below historical 18, substantially above the negative rates of the averages 1980s and early 1990s. However, productivity growth has again slowed sharply or turned negative in some fragile and conflict-afflicted states (Burundi, Mozambique). Regions. Productivity growth decelerated in all EMDE regions during 2013-18 from their pre- crisis (2003-08) averages (Box 3.1). This slowdown occurred amid heightened debt levels which increase the probability of financial crises E. Productivity growth in EMDE regions F. Cumulative productivity losses relative to 2003-08 trend and crowd out productive investments. The most pronounced slowdown (by 3.8 percentage points to 1.5 percent in 2013-18) occurred in Europe and Central Asia (ECA), where the global financial crisis and subsequent Euro Area debt crisis caused severe economic disruptions. Productivity growth has also fallen steeply in Latin America and the Caribbean (LAC), the Middle East and North Africa (MNA), and Sub-Saharan Source: Penn World Table; The Conference Board; World Bank, World Development Indicators. Africa (SSA), to near zero. Productivity growth Note: Productivity is defined as output per worker. Data are from a balanced sample between 1981- declined substantially in East Asia and Pacific 2018 and includes 29 advanced economies (AEs), and 74 emerging market and developing economies (EMDEs) including 11 low-income countries (LICs), as of 2019 World Bank classifications, (EAP) and more modestly in South Asia (SAR) 52 commodity exporters and 22 commodity importers. GDP-weighted (at constant 2010 prices and exchange rates) aggregates. from pre-crisis levels, but it continued to be A.B. GDP weighted averages (at 2010 prices and exchange rates). C. Share of economies for which average productivity growth during 2013-18 was lower than the robust, remaining above 5 percent in both regions. long-run (1981-2018) average or the pre-crisis (2003-2008) average. E. GDP-weighted productivity growth for 8 EMDEs in East Asia and the Pacific (EAP), 10 EMDEs in Eastern Europe and Central Asia (ECA), 18 EMDES in Latin America and the Caribbean (LAC), 10 Missed opportunities. The steep productivity EMDEs in Middle East and North Africa (MNA), 2 EMDEs in South Asia (SAR), and 26 EMDEs in Sub-Saharan Africa (SSA). growth slowdown since the global financial crisis F. Percent fall in productivity level by 2018 relative to a counterfactual scenario where productivity implies considerable output losses relative to a continued to grow at its 2003-08 average growth rate from 2009 onwards. Click here to download data and charts. counterfactual of productivity growth continuing at its pre-crisis trend. Output per worker in advanced economies would be 5 percent higher today had productivity growth continued at its 198 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 3.1 EMDE regional labor productivity trends and bottlenecks The post-crisis slowdown in productivity growth was particularly severe in East Asia and Pacific, Europe and Central Asia, and Sub-Saharan Africa amid slowing investment growth, financial market disruptions, and a post-crisis commodity price slide. Meanwhile, productivity growth in Latin America and the Caribbean and the Middle East and North Africa—the slowest even before the global financial crisis—has fallen to near-zero as investment collapsed amid political uncertainty, episodes of financial stress in major economies, and falling commodity prices. As a result, the pace of catch-up to advanced-economy productivity levels has slowed in most regions since the global financial crisis and, in some regions, productivity is even falling further behind. In almost all regions, productivity gains from the reallocation of labor from low-productivity to higher-productivity sectors have slowed sharply. To boost productivity, policies are needed to address key obstacles to productivity growth. Some of these obstacles are shared across EMDE regions, including resource-reliant economies, widespread informality, shortcomings in education, and weak governance, and some are region-specific bottlenecks. Introduction East and North Africa (MENA) growing faster than the advanced economy average (Rodrik 2011; Roy, Kessler Although common across all EMDE regions, the post- and Subramanian 2016; Figure 3.1.1). Since the global crisis productivity growth slowdown has differed markedly financial crisis (2013-18), however, productivity growth in severity. Generally, it was more pronounced in more has slowed from pre-crisis (2003-08) rates in all EMDE open EMDE regions that are closely integrated into regions. advanced-economy supply chains. Meanwhile, in regions with a large number of commodity exporters, productivity The slowdown was particularly steep in East Asia and the growth has fallen close to zero. As a result, to varying Pacific (EAP), especially in China, as well as in Europe and degrees, the catch-up to advanced-economy productivity Central Asia (ECA) and Sub-Saharan Africa (SSA). In levels has slowed since the global financial crisis and, in these regions, investment growth has declined sharply some regions, productivity is even falling further behind. from pre-crisis levels amid a policy-guided public Policy priorities to reignite productivity growth differ investment slowdown in China (EAP), financial system across regions. disruptions associated with the Euro Area crisis (ECA), This box draws out differences in regional productivity and the commodity price collapse of 2014-16 (ECA, SSA). trends and policy priorities (summarizing Boxes 2.1-2.6).1 However, in all three regions, there were important Specifically, it addresses the following questions: exceptions to the sharp slowdown. In EAP, the slowdown was concentrated in China while productivity growth • How has the evolution of productivity varied across continued to be robust in other major EAP economies, regions? especially some ASEAN economies (the Philippines and Vietnam), as FDI and investment growth remained robust • What factors were associated with stronger (Box 2.1). In ECA, the slowdown was muted in productivity growth? agricultural economies in Central Asia that shifted their For the purposes of this box, productivity is defined as economic ties towards China and in Central European labor productivity—that is, real GDP per worker (at 2010 economies that continued to integrate into Western prices and exchange rates). European supply chains and benefited from investment financed by European Union structural funds. In SSA, Evolution of productivity productivity growth accelerated in agricultural commodity exporters. Post-crisis labor productivity growth slowdown. An exceptional pre-crisis surge in productivity growth was The slowdown was mildest in South Asia (SAR), in part broad-based across regions, with productivity in more than because the region is the least open EMDE region to 50 percent of economies in each region except The Middle global trade and finance, continued to urbanize rapidly, and, as a predominantly commodity-importing region, benefited from the commodity price slide. In MENA, the Note: This box was prepared by Gene Kindberg-Hanlon with research slowdown was mild since limited links to global financial assistance from Shijie Shi. markets insulated commodity-importing economies from 1 To be as representative of each region as possible, this box uses a global financial stress. broader sample than the main text in Chapter 3, resulting in a shorter time horizon under consideration. This box and the regional boxes cover a sample containing 127 EMDE economies, compared to 74 in the main Post-crisis productivity growth across regions. text. Productivity growth in Latin America and the Caribbean G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 199 BOX 3.1 EMDE regional labor productivity trends and bottlenecks (continued) FIGURE 3.1.1 Evolution of regional labor productivity The post-crisis slowdown in labor productivity growth was particularly severe in EAP, ECA and SSA as these regions struggled with slowing investment growth, financial market disruptions, and weaker commodity prices. In EAP and ECA, the slowdown in productivity growth has reflected both a slower pace of capital deepening and weaker TFP growth. In MENA and SAR, TFP has continued growing or stabilized after earlier contractions (MENA). A. Labor productivity growth in EMDE B. Share of economies growing faster C. Annual rate of productivity regions than the average advanced economy convergence, 2003-08 and 2013-18 D. Regional average productivity differen- E. Contributions to regional productivity F. Contributions to regional productivity tials, GDP-weighted, 2018 growth; EAP, ECA, LAC growth: MNA, SAR, SSA Source: International Monetary Fund; Penn World Table; The Conference Board; World Bank, World Development Indicators. A.B.C.D. Productivity refers to output per worker at 2010 prices and exchange rates. Sample includes 35 advanced economies (AE) and 16 EMDEs in East Asia and the Pacific (EAP), 21 EMDEs in Eastern Europe and Central Asia (ECA), 25 EMDES in Latin America and the Caribbean (LAC), 14 EMDEs in Middle East and North Africa (MNA), 7 EMDEs in South Asia (SAR), and 44 EMDEs in Sub-Saharan Africa (SSA). A. GDP-weighted average labor productivity growth. B. Share of economies with faster productivity growth than the advanced-economy average in each period. C. Rate of convergence calculated as the difference in productivity growth rates with the average advanced economy divided by the log difference in productivity levels with the average advanced economy. Regional rate of convergence is the GDP-weighted average of EMDE members of each region. D. Whiskers show the range within the region as a percent of the advanced economy average while bars show the GDP-weighted average level of productivity relative to advanced economies. Productivity reflects output per worker measured in US dollars at 2010 prices and exchange rates. E.F Aggregates calculated using GDP weights at 2010 prices and exchange rates. The sample includes 92 emerging market and developing economies (EMDEs), including 8 East Asia and Pacific, 21 Europe and Central Asia, 19 Latin America and the Caribbean, 12 Middle East and North Africa, 2 South Asia, and 30 Sub-Saharan Africa economies. Click here to download data and charts. (LAC), MNA, and SSA—even before the crisis, the EAP and SAR, where investment growth is still higher slowest—has fallen to near zero as investment collapsed than in other EMDE regions (EAP, SAR) or the shift amid political uncertainty, episodes of financial stress in towards more productive sectors has accelerated (SAR). In major economies, and falling commodity prices (Box 2.3). these two regions, productivity continues to converge As a result, productivity growth in the majority of EMDEs towards advanced-economy levels at approximately the in LAC, MNA, and SSA now lags that in advanced pre-crisis pace. economies and, on average in these regions, productivity levels are diverging from those in advanced economies. In Regional dispersion of productivity. On average, contrast, productivity growth continues above 5 percent in productivity in EMDEs was just 19 percent of the 200 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 3.1 EMDE regional labor productivity trends and bottlenecks (continued) advanced-economy average in 2018.2 Among EMDE In MNA and SAR, in contrast, TFP continued growing at regions, average labor productivity is highest in the MNA the pre-crisis pace (SAR) or stabilized after earlier (45 percent of the advanced-economy average), LAC and contractions (MNA), even as capital deepening slowed ECA (about 22-30 percent, respectively) and lowest in sharply (SAR) or reversed (MNA). In MNA, the oil price SAR (6 percent) and SSA (11 percent). However, these collapse of 2014-16 weighed heavily on investment in oil regional averages disguise wide dispersion within some exporters and political tensions discouraged investment in regions, especially MNA, ECA, and SSA. In some Gulf commodity importers. However, macroeconomic and Cooperation Council (GCC) countries in MNA, for structural reform efforts helped stem pre-crisis contractions example, productivity is near advanced-economy averages in TFP. In SAR, persistent post-crisis investment whereas in heavily agricultural economies, such as the Arab weakness—in part due to disruptive policy changes and Republic of Egypt and Morocco, it amounted to 10 tapering growth of FDI inflows—was offset by percent of the advanced-economy average (Box 2.4). productivity-enhancing sectoral reallocation, as labor Similarly, close trade integration with Western Europe moved out of agriculture into more productive sectors and, increasingly, China and major reforms since the amid rapid urbanization (Box 2.5). collapse of the Soviet Union have helped raise average productivity levels in ECA to the second-highest among Conversely, in SSA and LAC, TFP contracted. In major EMDE regions (30 percent). However, there is wide LAC economies, continued post-crisis credit extension or heterogeneity, with Poland producing around 38 percent intensifying economic distortions (such as trade of the advanced economy average worker, while some restrictions and price controls) allowed unproductive firms agricultural economies in Central Asia produce just 3 to survive to a greater extent than pre-crisis. In SSA, the percent (Box 2.2). In SSA, LICs produce about 2 percent contraction in TFP was partly offset by accelerating capital of the advanced economy average whereas oil exporters deepening as a number of countries invested heavily in such as Gabon produce 33 percent (Box 2.6). In contrast, public infrastructure, typically financed by debt. closely integrated EAP has a narrower range of productivity levels (2-25 percent of the advanced-economy Regional sources of productivity growth and average). bottlenecks Capital deepening versus total factor productivity A wide range of factors have weighed on productivity growth. Productivity growth can be decomposed into the growth since the global financial crisis, but their relative use of factor inputs (human or physical capital) or the role has differed across regions. In all regions other than effectiveness of their use (total factor productivity, or TFP, SAR, productivity gains from the reallocation away from Figure 3.1.1). In EAP and ECA, the post-crisis slowdown low-productivity (usually agriculture) sectors to higher- in productivity growth has reflected both a slower pace of productivity sectors have slowed (Enache, Ghani, and capital deepening and weaker TFP growth, albeit to O’Connell 2016). In addition, the pre-crisis pace of varying degrees. Two-fifths of the slowdown in EAP improvements in various aspects of the supporting reflected slowing capital deepening, the remainder slowing environment for productivity growth has slowed. TFP growth. In EAP, a policy-guided move towards more Productivity levels in all regions remain less than half of sustainable growth in China and trade weakness weighed those in advanced economies, providing significant scope on investment and capital deepening. In ECA, most (two- for faster productivity growth. However, significant thirds) of the productivity growth slowdown reflected a bottlenecks to productivity convergence remain, many of collapse in investment growth as conflict erupted in parts which differ across regions. of the region, sanctions were imposed on the Russian Federation, political and economic shocks unfolded in Sectoral reallocation Turkey, financial systems transformed after the Euro Area Declining gains from sectoral reallocation. In all regions debt crisis, and the commodity price collapse hit except MNA, switching employment from low- commodity exporters (Arteta and Kasyanenko 2019). productivity sectors to sectors with above-average productivity levels supported productivity growth during 2003-08, especially in EAP, ECA, and SSA (Figure 3.1.2). 2 In this section, GDP-weighted averages of productivity are used to In SSA, it accounted for more than half of growth in the compare productivity levels across economies—in the main text, simple median economy during 2003-2008 (Diao, McMillan, averages are used. and Rodrik 2017). G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 201 BOX 3.1 EMDE regional productivity trends and bottlenecks (continued) FIGURE 3.1.2 Sectoral contributions to regional productivity growth Since the global financial crisis productivity gains from sectoral reallocation have faded across all regions (with the exception of SAR). In SAR and SSA, around half of employment is in the agricultural sector, which only accounts for around 20 percent of output, reflecting low productivity in this sector. The wide dispersion of sectoral productivity levels within regions demonstrates the importance of introducing measures to reduce misallocation and boost productivity in the weakest sectors. A. Within and between sector B. Within and between sector contribu- C. Composition of employment by sector, contributions to regional productivity tions to regional productivity growth: 2015 growth: EAP, ECA, LAC MNA, SAR, SSA D. Composition of value-added by sector, E. Sectoral contribution to aggregate F. Sectoral productivity levels dispersion 2015 productivity growth, 2013-15 within regions, 2015 Source: APO productivity database; Expanded African Sector Database; Groningen Growth Development Center Database; Haver Analytics; ILOSTAT; OECD STAN; United Nations; World KLEMS. Note: Sample includes 46 EMDEs, of which 8 are LICs and 9 East Asia and Pacific, 6 Europe and Central Asia, 6 Latin America and the Caribbean, 3 Middle East and North Africa, 3 South Asia, and 19 Sub - Saharan African economies. A.B. Median contribution for each region. Growth within sector shows the contribution of initial real value added-weighted productivity growth rate of each sector and ‘between sector’ effect shows the contribution arising from changes in sectoral employment shares. E. Median contribution to productivity growth. F. Range of (regional averages of) sector-specific productivity levels relative to advanced-economy average productivity for the same sector in 2015, valued at 2011 purchasing power adjusted exchange rates. The range for MNA excludes sectoral productivity for mining which exceeds 1000 percent of the advanced-economy average. Click here to download data and charts. Since the global financial crisis, however, productivity in SAR, the move of labor out of low-productivity gains from sectoral reallocation have faded across all agriculture into more productive sectors accelerated as regions (with the exception of SAR). In commodity-reliant rapid urbanization continued and strong consumption regions such as LAC, MNA, and SSA, this in part reflected growth fueled employment in higher-productivity trade lower absorption of labor by services and construction services. sectors as real income losses in resource sectors spilled over into weaker demand. In EAP, it reflected slowing labor Looking ahead, further sectoral reallocation continues to reallocation as overcapacity was gradually being unwound. have a high potential to lift productivity growth in SSA In ECA, high-productivity manufacturing, financial, and and SAR, where low-productivity agriculture accounts for mining sectors suffered during the Euro Area debt crisis around 50 percent of employment and 20 percent of and the post-crisis commodity price collapse. Meanwhile, output. Substantial gaps in productivity between sectors 202 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 3.1 EMDE regional productivity trends and bottlenecks (continued) remain, offering the potential for further aggregate accounts for 25-40 percent of official GDP (22 percent of productivity gains from resource reallocation between GDP in MNA); however, reflecting heterogeneity in sectors. productivity levels, informal employment (measured as self -employment) varies widely from 22 percent (MENA) to Bottlenecks to productivity growth 62 percent (SSA) of total employment (World Bank 2019a). Several bottlenecks to higher productivity are shared, to varying degrees, by multiple EMDE regions. These Limited human capital. Higher-skilled and better- include commodity-reliance, widespread informality, poor educated labor forces tend to adopt new technologies, education, and weak governance. Other bottlenecks are including new ICT and manufacturing technologies, more mostly region-specific. readily and more effectively (World Bank 2019c). In EAP and ECA, expected years of schooling for children are now Reliance on commodity exports. In LAC, MNA, and within one year of advanced economies on average, but SSA, commodities account for over 20 percent of exports SAR and SSA lag more than 3 years behind the advanced- on average. In ECA, they account for 30 percent of economy average (Figure 3.1.3). Even where years of exports, largely due to Russia, where around 60 percent of schooling are on par with advanced economies, education exports are (mostly energy) commodities. Economies that can be ineffective where learning outcomes are poor are highly reliant on a narrow range of commodity exports (World Bank 2018a). In learning-adjusted terms, which can also suffer from misallocation and procyclical trends controls for the quality of education in addition to years of for productivity growth (Frankel 2010). Conversely, attainment, SAR and SSA lag substantially (six or more producing across a broad range of sectors can insulate learning-adjusted years) behind advanced economies. economies from external shocks, and can facilitate knowledge transfer to strengthen productivity (Kraay, Region-specific factors. In each region, some challenges to Soloaga, and Tybout 2002; Schor 2004). In EAP, for improving or sustaining productivity growth are notable: example, high pre-crisis productivity growth was spurred by rapid integration into global supply chains and • In EAP, the region faces challenges in sustaining attraction of FDI which enabled a substantial increase in productivity growth as rapid trade integration, which the range and sophistication of production in the region spurred productivity growth in the 2000s, fades. With (Wei and Liu 2006). maturing supply chains and weak global trade, the Weak governance and institutions. In most EMDE priority has shifted towards improving the allocation regions, governance and business climates are less business- and efficiency of investment, including in a wider friendly than in advanced economies. The largest distances range of sectors (World Bank and DRCSC 2019). to the frontier (the most business-friendly climates) are in SSA, SAR, and LAC, but also in pockets of ECA (Central • In ECA, reform momentum has stalled in many Asia and Eastern Europe) and MNA (North Africa). In all economies since the global financial crisis. This regions, a large majority of EMDEs fall below the global follows on the heels of a period of rapid progress in average for tackling corruption. Poor institutions have the 1990s and 2000s in the transition to market-based been associated with weak firm productivity and inefficient economies and, in Central Europe, in the accession to government investment in productivity-augmenting the European Union (Georgiev, Nagy-Mohacsi, and infrastructure (Cirera, Fattal-Jaef, and Maemir 2019). In Plekhanov 2018). Restrictive product market and EAP, poor corporate governance in some sectors services regulations now hinder competition and deter contributes to resource misallocation and weighs on foreign investment. productivity. • In MNA, the government accounts for a large share of Informality. Informality is pervasive in EMDEs, although employment relative to other regions. About one-fifth there are large differences in the productivity of informal of the workforce is employed in the public sector. sectors across regions. Informal firms are less productive This is in part driven by a sizable wage premium for than those in the formal sector and, by competing on more public-sector workers and a bias in the education favorable terms, can deter investment and erode the system toward training for public sector employment. productivity of formal firms (Amin, Ohnsorge, and Okou The non-GCC private sector is anemic, with lower 2019). In all regions except MNA, the informal sector firm turnover than in other EMDE regions. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 203 BOX 3.1 EMDE regional productivity trends and bottlenecks (continued) FIGURE 3.1.3 Potential bottlenecks to productivity growth Several bottlenecks to higher productivity are shared, to varying degrees, by EMDE regions. These include undiversified economies, weak governance, widespread informality, poor learning outcomes, and low trade and financial openness. A. Share of commodities in total exports, B. Government effectiveness, 2013-2018 C. Informal economy, 2016 2013-2018 D. Educational attainment, 2017 E. Trade and financial openness, F. Business climates, 2020 2013-2018 Source: United Nations; World Bank, Doing Business, Human Capital Project, World Development Indicators, Worldwide Governance Indicators. A. Exports of metals, agricultural and energy products in percent of total exports. GDP-weighted average for each region. Average during 2013-2018. B. WGI index defined as capturing perceptions of the quality of public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formation and implementation, and the credibility of the government’s commitment to such policies. Bars show interquartile range. C. Average informal output (DGE-based estimates, percent of official GDP) and employment estimate (self-employment, percent of total employment) in each region. Based on World Bank (2019a). D. Expected years of schooling and learning-adjusted years of schooling from the World Bank’s Human Capital Project. Learning-adjusted years of schooling uses harmonized cross-country test scores to adjust the average years of schooling. E. Unweighted average of trade (exports plus imports) in percent of GDP and net foreign direct investment inflows in percent of GDP. F. Unweighted average distance to frontier measure of the ease of doing business score from the 2020 Doing Business Indicators. A higher value indicates a business climate that is closer to best practices. Bars show range. Click here to download data and charts. • In LAC, productivity could be boosted by policies to • In SSA, low productivity reflects the presence of large improve innovation and competition. Greater trade agricultural sectors, including widespread subsistence integration and more welcoming environments for agriculture. A policy priority is therefore to lift FDI could lift productivity growth through productivity in the agricultural sector. In addition, knowledge and technology transfers. SSA economies tend to be involved in supply chains only at early stages of production, producing primary • In SAR, productivity has been held back by below- products, and have few exporting firms. average international trade integration and FDI, which limits technology and knowledge spillovers, and restricted access to finance from a banking system that is heavily state-dominated. 204 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 average pace ahead of the crisis (2003-2008). differences) drive two-thirds of global inequality Losses relative to the exceptionally high rate of (World Bank 2018c). productivity growth in EMDEs ahead of the crisis are closer to 14 percent, and higher still at 19 • Commodity importers and exporters. Relative percent for EMDE commodity exporters. productivity levels are slightly higher in commodity-importing EMDEs on average Labor productivity (19 percent of advanced-economy produc- tivity) than in commodity-exporting EMDEs convergence (17 percent) and, lower in non-oil exporters (10 percent) than in oil exporters (28 percent) EMDE productivity levels are less than one-fifth of (Chapter 2 boxes). the advanced-economy average, falling to just 2 percent in LICs. In some large EMDEs, such as • LICs. In LICs, productivity is just 2 percent of China and India, productivity is growing the advanced-economy average, having made substantially faster than in advanced economies, negligible progress in narrowing this gap since resulting in productivity catch-up. However, average the 1990s (World Bank 2019b). EMDE productivity growth is just half a percentage point faster than in advanced economies, requiring • Regions. Productivity is lowest on average in more than a century to halve productivity gaps. SSA and SAR (8 and 7 percent of the advanced-economy average respectively). Faster productivity growth occurs in countries Within SSA, which hosts most LICs and with lower initial productivity levels when mostly non-oil commodity exporters, controlling for factors such as the level of human productivity is even lower in many economies, capital and institutional quality (Durlauf, falling to just 2 percent of the advanced Johnson, and Temple 2005; Johnson and economy average in the bottom quartile of the Papageorgiou 2018). At 3.6 percent in 2018, region (Box 3.1). It is highest in MNA (36 productivity growth in EMDEs remained more percent of the advanced-economy average), than four times as high as in the average advanced which hosts several high-income oil exporters, economy (0.8 percent). However, this aggregate and ECA (19 percent of the advanced- growth rate is dominated by China and India, the economy average), parts of which are closely largest EMDEs by output and population, where integrated with EU supply chains and EU productivity growth is above five percent. Many labor markets. Throughout the 2000s, pre- as EMDEs are growing at a substantially slower pace well as post-crisis, the gap with advanced than China and India: on average, EMDE economies has closed fastest in EAP and SAR productivity is growing by just 0.5 percentage but continued to widen in parts of LAC, point faster than in advanced economies. MNA, and SSA. Productivity gaps. Despite some narrowing of the Pace of productivity convergence. Productivity productivity gap in 60 percent of EMDEs since convergence between low and high-productivity the 1990s, output per worker in EMDEs remains economies became broad-based in the late 1990s, less than one-fifth that of the average advanced with little evidence for convergence prior to this economy (Figure 3.4).10 This productivity (Patel, Sandefur, and Subramanian 2018; Figure differential accounts for a considerable proportion 3.4).11 While the presence of convergence during of global income inequality since global per capita the 2000s is reassuring, its pace is disappointing. income differences (reflecting mainly productivity At current productivity growth rates, productivity gaps to advanced-economy average productivity 10 This productivity gap is measured using output per worker in 2010 U.S. dollars at market exchange rates. When measured at 11 The speed of productivity convergence can be formally assessed Purchasing Power Parity (PPP) adjusted U.S. dollars, the gap to using a “β convergence” test, where productivity growth is regressed advanced economies is smaller, with EMDE productivity around one on the initial level of productivity (Barro 1991; Barro and Sala-i- -third of the advanced economy average (World Bank 2018a). Martin 1992). G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 205 are narrowing by 0.3 percent per year on FIGURE 3.4 Distribution of productivity levels and average—requiring more than a century just to convergence progress close half of the gap. But the pace of convergence On average, productivity in EMDEs is less than one-fifth of the advanced- differs across regions. At current rates of economy average, and in LICs it is just 2 percent. EMDE productivity gaps with the advanced-economy average widened during the 1970s-1990s but productivity growth, less than 20 percent of narrowed from 2000 onwards. However, the implied pace of convergence economies in LAC, MNA or SSA—but at least 50 is low—even at the peak of EMDE growth, the productivity gap would have percent of those in EAP and SAR—are on course taken over a century to halve. to halve their productivity gap over the next 40 A. EMDE productivity levels, 2013-18 B. Simple average of productivity years. simple average relative to advanced economies by region, 2013-18 Sources of post-crisis slowdown in labor productivity growth Aggregate labor productivity growth can be decomposed into its sources: into factor inputs and the efficiency of their use, or into sectors. These C. Share of EMDEs with narrowing D. EMDE productivity levels since the decompositions suggest that the post-crisis productivity productivity gap to advanced 1990s, GDP-weighted average growth slowdown in EMDEs, in approximately economies equal measure, reflected weak investment and a slowdown in total factor productivity growth, as well as fading gains from factor reallocation towards more productive sectors. Decomposition into factor inputs Approach. In the first step, productivity growth is decomposed into contributions from individual E. Estimated annual decline in F. Share of economies, by years to factor inputs (capital and human capital) and the productivity gap halve the productivity gap with advanced economies effectiveness of their use (total factor productivity, or TFP, growth), assuming a Cobb-Douglas production function (Annex 3.2). Capital deepening directly increases labor productivity, while human capital improvements (e.g. education and training) enhances the quality of labor input and therefore the resulting output produced. TFP measures the efficiency with which all factors are employed, and is often considered a proxy for the Source: Penn World Table; The Conference Board; World Bank, World Development Indicators. technology behind the production process.12 TFP Note: Productivity defined as output per worker in U.S. dollars (at 2010 prices and exchange rates). Based on 29 advanced economies and 74 EMDEs, which include 22 commodity-importing EMDEs growth can also be affected by non-technology and 52 commodity-exporting EMDEs. A. Blue bars indicate unweighted average output per worker during 2013-18 relative to the advanced- economy average. Whiskers indicate interquartile range relative to the advanced-economy average. B. Unweighted average productivity during 2013-18 relative to average advanced economy by region (2013-18). Includes 29 advanced economies and 74 EMDEs = 8 EMDEs in East Asia and the Pacific (EAP), 10 EMDEs in Eastern Europe and Central Asia (ECA), 18 EMDES in Latin America and the 12 The decomposition above is an accounting framework that does Caribbean (LAC), 10 EMDEs in Middle East and North Africa (MNA), 2 EMDEs in South Asia (SAR), not control for dynamic interactions between TFP and investment and 26 EMDEs in Sub-Saharan Africa (SSA). C. Share of EMDEs with faster productivity growth than the advanced-economy average. growth. However, there is evidence that weak underlying TFP and D. GDP-weighted (at 2010 prices and exchange rates) averages. investment growth reinforce each other, which could have amplified E. Line shows the implied annual rate of decline of the productivity gap based on a regression of labor the post-crisis productivity slowdown. Weaker rates of investment productivity growth on initial productivity. Shaded area indicates 90 percent confidence intervals. reduce TFP growth by reducing the incorporation of new Estimation performed over 10-year rolling windows in the specification technologies into the production process (Adler et al. 2017; Hulten logΔyt c βyt-10 εt where y is output per worker. Coefficient converted to the average annual decline in the productivity gap following Sala-i-Martin (1992). 1992). Conversely, slower technological change reduces the expected F. The proportion of EMDEs in each region that will close half of the productivity gap with the average return on capital and, hence, the incentives to invest. advanced economy in each bracket of years based on average growth during 2013-18 relative to average advanced economy growth and the outstanding productivity gap over the same period. Click here to download data and charts. 206 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 factors, such as changing levels of capital and labor (2003-08) relative to the 1980s and 1990s utilization—therefore estimates may over or and has now recovered modestly.15 understate the true change in the influence of technology on productivity. Efforts to control for • EMDEs. The post-crisis slowdown in EMDE utilization have found that while some of the pre- productivity growth from pre-crisis averages crisis surge in productivity in EMDEs was a reflected, in approximately equal measure, demand-driven phenomenon of increased investment weakness and slowing TFP utilization, a large proportion of the subsequent growth. In commodity-exporters, the con- slowdown was structural, reflecting factors other tribution of capital accumulation faded almost than fading demand after the global financial crisis entirely, after having accounted for about (Dieppe, Kiliç Çelik, and Kindberg-Hanlon, half of productivity growth pre-crisis. This Forthcoming). was compounded by contracting TFP growth, which had accounted for most of the Factors inputs versus the effectiveness of their remainder of pre-crisis productivity growth. use. Globally, the post-crisis (2013-18) slowdown Investment stalled or contracted in com- in labor productivity growth from pre-crisis modity exporters during the commodity (2003-08) averages amounted to half of a prices collapse of 2011-16 (Aslam et al. 2016; percentage point, the majority of which was a World Bank 2017). TFP growth has also result of a slowdown in capital accumulation been weak historically, contributing little to (both public and private; World Bank 2019b). In catch-up growth (De Gregorio 2018). In advanced economies, the slowdown in TFP commodity-importers, especially China, capital growth was a minor source of the post-crisis deepening accounted for much of the decline in labor productivity growth, due to a productivity gains over the past four decades. structural slowdown prior to the crisis.13 In This momentum has slowed since the global EMDEs, however, it accounted for about one-half financial crisis reflecting diminishing growth of the slowdown in labor productivity growth. prospects, heightened uncertainty, and weak FDI inflows. In the early 2000s, TFP was • Advanced economies. Investment weakness boosted by earlier reforms that allowed greater accounted for virtually all of the post-crisis FDI inflows in the 1990s and WTO accession slowdown in productivity growth from pre- in 2001 which unleashed a productivity boom crisis averages in advanced economies (Figure in China and its trading partners, while a 3.5). From 2008, investment growth slowed decade of service-sector oriented reforms sharply in response to weak and highly boosted productivity in India (Bosworth and uncertain growth prospects, heightened policy Collins 2008; He and Zhang 2010; Tuan, uncertainty, and credit constraints in the Ng, and Zhao 2009). aftermath of the global financial crisis.14 Investment contracted by an average of 6 • LICs. In LICs, heavy public infrastructure percent per year between 2008-09. While the investment and business climate improve- investment share of GDP has recovered close ments have supported post-crisis output and to pre-crisis levels, it has been accompanied by productivity growth (World Bank 2019c). strong rates of employment growth, such that This followed on the heels of a decade of the growth of capital per worker has remained heavy investment into mines and oil fields subdued (ECB 2017). TFP growth had already declined in the pre-crisis period 15 Much of the recent discussion of advanced economy TFP growth has focused on the slowdown in the United States, where TFP has weakened further since the crisis following a surge from the mid-1990 to 2000s (Fernald et al. 2017; Cowen 2011; Gordon 13 This finding is in line with previous studies of the United States 2018). In contrast, average TFP growth was low in the pre-crisis and other advanced economies (Adler et al. 2017; Fernald et al. period in major European economies such as Germany and France 2017). (0.1-0.4), and even negative in Italy and Spain, such that the post- 14 See for details Duval, Hong, and Timmer (2017) and Ollivaud, crisis TFP slowdown is much less pronounced for advanced Guillemette, and Turner (2016). economies in aggregate. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 207 amid surging pre-crisis commodity prices. As FIGURE 3.5 Decomposition of productivity growth a result, continued post-crisis strength in Almost three-quarters of the post-crisis slowdown in global productivity productivity growth reflected increased capital growth from pre-crisis averages—and virtually all in advanced accumulation. Modest improvements in economies—reflected a slowdown in capital accumulation. The post-crisis slowdown in EMDE productivity growth from pre-crisis averages reflected, human capital partly offset increasingly in approximately equal measure, investment weakness and slowing TFP negative TFP growth in these economies. A growth. In LICs, strong investment has supported post-crisis output and continued concentration in the agricultural productivity growth. and extractives sectors has led to low technological progress, with additional A. Contributions to productivity growth in advanced economies B. Contributions to productivity growth in EMDEs negative shocks from conflict and from high levels of debt in the 1980s and 1990s also contributing to frequently negative TFP growth (Claessens et al. 1997; IMF 2014). • EMDE regions. Capital accumulation accounted for virtually all of the post-crisis slowdown in productivity growth in MNA, where oil-exporting EMDEs suffered stalled or contracting investment amid the oil price C. EMDE commodity exporter and D. Contributions to productivity collapse of 2014-16 (Stocker et al. 2018). It importer productivity contributions growth in LICs also accounted for most of the slowdown in ECA, whose banking systems were hard-hit by the Euro Area crisis and the subsequent retreat from the region of EU-headquartered banks (Arteta and Kasyanenko 2019). In EAP, a deliberate policy-guided public investment slowdown in China is underway and slower capital accumulation accounted for about two-fifths of the slowdown in post-crisis productivity growth. In SSA, which hosts E. Contributions to regional F. Contributions to regional productivity growth: MNA, SAR, SSA most LICs, and in LAC, the slowdown was productivity growth: EAP, ECA, LAC entirely driven by declining TFP growth. In contrast to other EMDE regions, TFP growth strengthened in MNA, from negative pre- crisis rates amid heavy resource investment, and in SAR, which was little-affected by the disruptions of the global financial crisis. Decomposition into sectors Source: Barro and Lee (2015); International Monetary Fund; Penn World Tables; The Conference Approach. Higher aggregate productivity growth Board; United Nations; Wittgenstein Centre for Demography and Global Human Capital; World Bank, World Development Indicators. in EMDEs in the pre-crisis period was associated Note: Productivity defined as output per worker. Aggregate growth rates calculated using constant with a reallocation of resources towards more 2010 US dollar weights. 52 commodity exporters, 22 EMDE commodity importers, 8 East Asia and Pacific, 10 Europe and Central Asia, 18 Latin America and the Caribbean, 10 Middle East and North productive sectors in addition to productivity Africa, 2 South Asia, and 26 Sub - Saharan Africa economies. GDP weights. The sample includes 29 advanced economies, and 74 emerging market and developing economies including 11 low-income growth within sectors (Diao, McMillan, and countries. Click here to download data and charts. Rodrik 2017). More recently, pre-crisis gains from such reallocation appear to have faded. This is illustrated in a decomposition of economy-wide labor productivity growth into within- and between-sector productivity growth for 80 economies, including 38 EMDEs, of which 7 208 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 are LICs, for nine sectors during 1995-2015 spillovers between sectors may also be diminishing (Box 3.2). (Foerster et al. 2019). However, productivity gaps between sectors in EMDEs remain sizeable. In Wide differentials in sectoral productivity. Labor contrast to other regions, productivity gains from productivity varies widely across sectors, being reallocation continue to be sizable in SAR, lowest by far in agriculture and highest in mining, accounting for one-half of post-crisis productivity financial and business services, and utilities. In growth, as agricultural employment moves into EMDEs, labor productivity in mining and industrial sectors. financial and business services, which are often foreign-owned, is thirty to forty times the level of Challenges for within-sector productivity productivity in the agriculture sector, which is growth. Within-sector productivity gains also often characterized by smallholder farms (Figure decelerated post-crisis, in EMDEs as well as 3.6; Lowder, Skoet, and Raney 2016). In advanced economies. The post-crisis slowdown advanced economies, this differential is may reflect the challenges faced by the most considerably narrower (three times). As a result, productive firms (large, export-oriented ones) agricultural productivity in EMDEs lags far amid post-crisis trade and investment weakness behind that in advanced economies—in the (Box 3.3). In many EMDEs, an additional average EMDE, agricultural productivity is less challenge may arise from the sheer size of the than one-fifth that in the average advanced- informal sector (World Bank 2019a). The labor economy. In contrast, services sectors such as productivity of informal firms is, on average, only transport or financial and business services are one-quarter of the productivity of formal firms. small in EMDEs, accounting for 22 percent of Informal firms are less able than formal firms to value-added in total, but feature productivity that reap the productivity gains from economies of is two-fifths to one-half of advanced-economy scale (size), accumulated experience (age), productivity on average. agglomeration benefits (location), and best managerial practices (Fajnzylber, Maloney, and Fading gains from factor reallocation in EMDEs. Montes-Rojas 2011). Moreover, aggressive In EMDEs, about one-half of the post-crisis competition from informal firms can erode the (2013-15) slowdown in productivity growth from productivity of exposed formal firms by about 24 pre-crisis (2003-08) averages reflected fading gains percent relative to those formal firms that do not from resource reallocation towards more face informal competition (Loayza 2016; World productive sectors. In the 1990s and pre-crisis, Bank 2019a). A more conducive business climate, such resource reallocation had accounted for more and economic development more broadly, can than one-third of average labor productivity alleviate some of the corrosive productivity effects growth, in line with earlier findings (Diao, of informal competition on formal firms. McMillan, and Rodrik 2017). Productivity gains from such a reallocation were particularly large in Fading gains from reallocation away from Sub-Saharan Africa, where they accounted for over agriculture in LICs. In LICs, agriculture accounts half of productivity growth during 2003-2008, for 31 percent of GDP, on average, but amid a large fall in the share of agricultural agricultural productivity is low (Cusolito and employment. Maloney 2018). As a result, a reallocation of employment, especially from agriculture, to Post-crisis, the contribution of reallocation to higher-productivity sectors accounted for almost productivity growth fell to less than one-quarter two-thirds of LIC productivity growth prior to the on average in EMDEs. To some degree as global financial crisis (Box 3.2). Since then, countries reach middle-to high income, sectoral however, this engine of LIC productivity growth reallocation tends to become a less important appears to have stalled. In part, this is due to a driver of productivity growth (de Nicola, collapse in global industrial commodity prices, Kehayova, and Nguyen 2018; Mason and Shetty which have discouraged further growth in 2019). In addition, technology and knowledge employment in the mining and extraction sector, G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 209 BOX 3.2 Sectoral sources of productivity growth Labor reallocation towards higher-productivity sectors has historically accounted for about one-third of aggregate productivity growth in EMDEs. This mechanism has, however, weakened since the global financial crisis. Fading productivity gains from labor reallocation have accounted for about one-half of the post-crisis productivity slowdown in EMDEs. In commodity-exporting EMDE regions, deindustrialization contributed to the slowdown. Introduction Against this backdrop, this box examines the sources of the post-crisis slowdown in productivity growth from a Factor reallocation towards higher-productivity sectors has sectoral angle. Specifically, it addresses the following long been recognized as one of the most powerful drivers questions. of aggregate productivity growth (Baumol 1967).1 It has been identified as an important driver of productivity • What are the main features of sectoral productivity? growth in economies as diverse as Sub-Saharan Africa, China and Vietnam (Cusolito and Maloney 2018; de • What was the role of sectoral reallocation in the post- Vries, de Vries and Timmer 2015; Fuglie et al. 2019). crisis productivity growth slowdown? Especially in East Asia, the move out of agriculture into higher-productivity industry and services has been credited Much of the earlier literature on sectoral productivity has with rapid productivity growth (Helble, Long, and Le focused on three sectors (agriculture, manufacturing, and 2019). services) with only a limited number of cross-country studies including more sectors.2 There is evidence that the In part as a result of several decades of sectoral reallocation findings of reallocation are sensitive to the level of away from agriculture, agriculture now accounts for only aggregation (de Vries et al. 2012; Üngör 2017). To explore 10 percent of EMDE value-added—one-quarter less than these issues, this box draws on a comprehensive dataset for two decades earlier and less than one-third the share of 80 countries and 9 sectors over 1995-2015. industrial production (Figure 3.2.1). LICs are an exception where agriculture still accounts for one-third of value- Features of sectoral productivity added, more than industry, and accounts for over 60 of employment. Wide productivity differentials across sectors. Productivity differs widely across sectors, offering large Meanwhile, services sectors have grown rapidly over the potential for productivity gains by factor reallocation past two decades. They now account for about one-half of across sectors (Figure 3.2.3). In the average EMDE, value-added in EMDEs as well as LICs, compared with productivity in the most productive sector—mining, three-quarters of value-added in advanced economies. which accounts for 4 percent of value-added—is twelve Services sectors have also been the main source of post- times that in the least productive sector—agriculture, crisis productivity growth, accounting for almost two- which accounts for 10 percent of value-added.3 In the thirds of productivity growth in the average EMDE average LIC, the range is even larger: productivity in the (compared with one-fifth accounted for by industry) and most productive sector—financial and business services, more than three-quarters in the average LIC. accounting for 13 percent of value-added—is twenty-two times that in the least productive sector—agriculture, Services describe a highly heterogeneous set of activities. which accounts for almost one-third of value-added Whereas industry mostly consists of manufacturing (64 percent in the average EMDE), services include in almost equal measure trade services, transport services, financial and business services, and government and personal 2 Diao, McMillan, and Rodrik (2017) and McMillan, Rodrik, and services. These service subsectors vary widely in their skill- Verduzco-Gallo (2014) employ 38 and 39 countries; Martins (2019) use 7 sectors and 169 countries, and International Monetary Fund (2018) and capital-intensity as well as their productivity. use 10 sectors and 62 countries. Further disaggregation using micro panel data (such as by Hicks et al. 2017) would help to ensure differences in marginal product are accounted for. 3 The high productivity extractive sectors offer few opportunities for Note: This box was prepared by Alistair Dieppe and Hideaki sectoral reallocation and are intrinsically limited by the size of the re- Matsuoka. source, and market power. It should be noted that refining and pro- 1 Throughout this box, productivity refers to labor productivity, cessing of extractives can sometimes be classified as manufacturing in resource rich countries. defined as value added per employed worker. 210 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 3.2 Sectoral sources of productivity growth (continued) FIGURE 3.2.1 Agriculture, industry and services In part as a result of a several decades of sectoral reallocation away from agriculture, agriculture now accounts for only 10 percent of EMDE value-added—one-quarter less than two decades earlier and less than one-third the share of industrial production. LICs are an exception; agriculture still accounts for one-third of value-added in these economies, more than industry. Meanwhile, services sectors—which include a highly heterogeneous set of activities—have grown rapidly over the past two decades, accounting for about half of post-crisis productivity growth. A. Composition of value-added B. Composition of employment C. Contributions to productivity growth Source: APO productivity database, Expanded African Sector Database, Groningen Growth Development Center Database, Haver Analytics, ILOSTAT, OECD STAN, United Nations, World KLEMS. Note: Based on sample of 80 countries. A.B. Share of agricultural, industry and services in value added. Industry includes mining, manufacturing, utilities, and construction. Services include trade services, transport services, financial and business services, government and personal services. Black horizontal line indicates 50 percent. Click here to download data and charts. (Figure 3.2.2).4 Since the 1990s, the productivity (pre-crisis) or near zero (post-crisis) in mining to the dispersion within the manufacturing and service sectors, highest sectoral growth rates (4.8 percent) in transport has narrowed. Similar differentials, between the most services in EMDEs in 2003-08 (Duernecker, Herrendorf, productive sector (financial and business services) and the and Valentinyi 2017).5 The post-crisis (2013-15) least productive sector (agriculture), in advanced slowdown in manufacturing productivity growth was the economies are considerably narrower. largest among all nine sectors, nearly 2 percentage points below the pre-crisis average (2003-08). Wide sectoral productivity differentials across countries. Productivity in all sectors is lower in EMDEs than in In advanced economies, the post-crisis productivity growth advanced economies, and lower again in LICs. The gap slowdown was broad-based across almost all sectors (except between EMDE and advanced-economy productivity is construction). More than one-half of the post-crisis (2013- particularly wide (almost 80 percent) in agriculture, which 15) slowdown in productivity growth from pre-crisis rates tends to be characterized by smallholder ownership and (2003-08) in the average EMDE originated in the family farms in EMDEs (Lowder, Skoet, and Raney manufacturing sector. The slowdown in agricultural 2016). This reflects in part slow technology adoption in productivity growth had only a limited aggregate effect in the agriculture sector in some of the poorest EMDEs. In EMDEs due to its relatively small share in the economy. mining, which tends to be dominated globally by a few In contrast, EMDE productivity growth picked up after large companies, the productivity gap is considerably narrower (just over 20 percent). 5 Two waves of service sector growth have been identified in the litera- Sectoral productivity growth. Productivity growth in the ture: a first wave in countries with relatively lower income levels and a various subsectors of services varied widely, from negative second wave in countries with higher income levels. The first wave ap- pears to be made up primarily of traditional (personal) services, the sec- ond wave of modern (financial, communication, computer, technical, 4 As agricultural workers often do not work full time in agriculture, the legal, advertising and business) services that are receptive to the applica- sectoral gap is diminished if productivity is measured per hours instead of tion of information technologies and tradable across borders per worker (McCullough 2017). However, even after accounting for (Eichengreen and Gupta 2013). Moreover, there is evidence of the sec- hours and human capital per worker, a large sectoral gap remains for ond wave also occurring in lower income countries after 1990 which are many of countries (Gollin, Lagakos, and Waugh 2014). democracies, and have high trade and financial openness. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 211 BOX 3.2 Sectoral sources of productivity growth (continued) FIGURE 3.2.2 Sectoral labor productivity Productivity differs widely across sectors and subsectors, especially in EMDEs and even more so in LICs. Productivity in all sectors is lower in EMDEs than in advanced economies, and lower again in LICs. The gap to advanced-economy productivity is particularly wide in agriculture, and narrow in mining. Industry was the main source of pre-crisis productivity growth; its slowdown accounted for more than half the post-crisis slowdown in aggregate productivity in EMDEs. A. Sectoral productivity relative to B. Sectoral productivity relative to the C. Sectoral productivity growth in within-group average productivity advanced-economy median advanced economies D. Sectoral productivity growth in EMDEs E. Contributions to productivity growth F. Contributions to productivity growth slowdown between 2003-08 and 2013-15 Source: APO productivity database, Expanded African Sector Database, Groningen Growth Development Center Database, Haver Analytics, ILOSTAT, OECD STAN, United Nations, World KLEMS. Note: Based on samples of 80 countries. Median of the county-specific productivity level, or growth rate. A. Bar charts range from the minimum to the maximum sector productivity gap. B. Sectoral productivities compared at PPP exchange rates. E.F. “Industry” includes mining, manufacturing, utilities, and construction; “Finance” includes business services; “Government” includes personal services. Click here to download data and charts. the global financial crisis in construction, utilities and decompose aggregate labor productivity into within-sector mining. and between-sector components (Wong 2006, Padilla- Pérez and Villarreal 2017). Within-sector productivity Role of sectoral reallocation growth captures changes in aggregate labor productivity Framework. The productivity differentials between sectors growth due to productivity improvements within sectors. offer the potential for productivity gains from labor This may reflect improvements in human capital, reallocation towards higher-productivity sectors, in investments in physical capital, or the reallocation of addition to within-sector productivity gains (Figure resources from the least to the most productive firms 3.2.3).6 This is captured in a shift-share analysis that within each sector. Between-sector productivity growth is driven by the change in employment share and the productivity differential. It reflects both the reallocation of 6 However, Fuglie et al. (2019) point out that different factor shares in resources to sectors with higher productivity levels (static value added would result in a gap of average labor productivity even if the factor allocation is efficient. A gap in average productivity is not sufficient sectoral effect), and the reallocation of employment evidence of misallocation because labor productivity can be equalized at towards sectors with higher productivity growth (dynamic the margin. 212 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 3.2 Sectoral sources of productivity growth (continued) FIGURE 3.2.3 Between- and within-sector sources to productivity growth While productivity growth in advanced economies has predominantly originated within sectors, between-sector gains have accounted for a sizable portion of EMDE productivity growth, and its post-crisis slowdown. In EMDEs, the between-sector productivity gains have involved shifts out of agriculture into higher-productivity sectors that have differed over time. A. Sectoral productivity relative to B. Contributions to productivity growth C. Contributions to within-sector produc- country productivity tivity growth D. Contributions to between-sector E. ECA, LAC, MNA: Composition of F. ECA, LAC, MNA: Contributions to productivity growth employment productivity growth Source: APO productivity database, Expanded African Sector Database, Groningen Growth Development Center Database, Haver Analytics, ILOSTAT, OECD STAN, United Nations, World KLEMS. B-D. Growth within sector shows the contribution of initial real value-added weighted productivity growth rate and structural change effect give the contribution arising from changes in the change in employment share. Median of the county-specific contributions. Based on samples of 80 countries. “Manuf.” includes mining and utilities; “Finance” includes business services; “Government” includes personal services. Click here to download data and charts. sectoral effect). Underlying drivers of such between-sector Decomposition of aggregate productivity growth. While productivity growth include changes in household’s productivity growth in advanced-economies has preferences and changes in relative sectoral productivity, in predominantly originated within sectors, between-sector part as a result of diverging evolutions of labor quality gains have accounted for one-third of EMDE productivity (Lagakos and Waugh 2013).7 growth since the 1990s. In part as a result of narrowing cross-sector productivity differentials and, in some regions, labor movements into lower-productivity sectors, fading sectoral reallocation has accounted for about one-half of 7 Improvements in agricultural productivity can significantly reduce the post-crisis slowdown in EMDE productivity growth. agriculture’s share of employment, contributing to between-sector The between-sector EMDE productivity gains have productivity growth (Gollin, Parente, and Rogerson 2007). The role of involved shifts out of agriculture into higher-productivity agriculture in structural change depends on economic integration within the domestic economy and with global markets (Barrett et al. 2017). sectors that have differed over time. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 213 BOX 3.2 Sectoral sources of productivity growth (continued) • Advanced economies. Productivity growth in advanced productivity agriculture generate aggregate productivity economies, where sectoral productivity differentials gains. In LICs, a somewhat larger portion (almost half) of tend to be narrower than in EMDEs, has been almost the 10 percentage point decline in the share of agricultural entirely driven by within-sector productivity growth employment was absorbed by trade services and only just since the 1990s. Within-sector productivity growth over one-third by industry. The phenomenon of has dwindled to 0.6 percent during 2013-15—less employment shifting out of agriculture into services has than half its 1990s average (Figure 3.2.3). The been dubbed “leapfrogging” in the context of concerns predominant structural change has been the about premature deindustrialization (Rodrik 2016). reallocation of resources from manufacturing to the Looking ahead, productivity gains arising from low-skilled financial and business services sector, two sectors with labor shifting out of agriculture into manufacturing or comparable levels of productivity. services may diminish if robotization and artificial intelligence discourage this movement. • EMDEs. In contrast, between-sector productivity gains in EMDEs boosted productivity growth pre- Deindustrialization. In three regions—Europe and crisis (2003-08) by 1.1 percentage points. Post-crisis, Central Asia (ECA), Latin America and the Caribbean this contribution fell to 0.5 percentage points, (LAC), the Middle East and North Africa (MNA)—the accounting for about one-half of the slowdown in manufacturing sector’s (as well as agriculture’s) share of EMDE productivity growth. Between-sector produc- employment has shrunk since the crisis, continuing a pre- tivity gains have mainly reflected a move out of crisis trend.9 Employment has largely shifted into agriculture and manufacturing into services. In LICs, construction (MNA), finance (ECA, LAC) and trade between-sector gains accounted for almost half of services (ECA, MNA). Since some of these sectors, post-crisis productivity growth, down from almost especially construction and trade services, have lower three-quarters of pre-crisis productivity growth.8 productivity than manufacturing, this has resulted in a Whereas pre-crisis between-sector productivity gains sharply lower contribution (ECA) or even negative in LICs mainly reflected a shift out of agriculture into contribution (LAC, MNA) of between-sector sources of manufacturing, their main post-crisis source was a productivity growth (Rodrik 2016). In LAC, for example, shift out of agriculture into services such as trade trade liberalization in the 1990s led to cheaper services and finance and business services that have manufacturing imports and a contraction in employment benefited from information and computing in the uncompetitive manufacturing sector. Much of this technologies (Eichengreen and Gupta 2013). labor was absorbed in construction and trade services that were buoyed by pre-crisis commodity boom (Gollin, Leapfrogging. Over the two decades until the global Jedwab, and Vollrath 2015). financial crisis, one-third of the EMDE employment that left agriculture moved into industrial sectors Conclusion (predominantly manufacturing and construction) and another one-third into trade services. The share of Large sectoral productivity differentials in EMDEs and agricultural employment in EMDEs declined by 9.4 LICs offer the potential of additional productivity gains percentage points between 1995 and 2008 while the shares when labor moves towards higher-productivity sectors. of industry and trade services rose by 2.5 and 3.0 Such between-sector productivity gains have contributed percentage points, respectively. Although trade services importantly to productivity growth in EMDEs and LICs and construction typically have below-average productivity since the 1990s. However, since the global financial crisis, and manufacturing productivity is near the EMDE these gains appear to have faded. average, the employment shift out of extremely low- 8 This is consistent with Diao, McMillan, and Rodrik (2017) and, for 9 To some degree this could reflect an outsourcing of parts of the Sub-Saharan Africa, McMillan, Rodrik, and Verduzco-Gallo (2014). manufacturing sector to the service sector. 214 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 3.6 Sectoral productivity developments many of these and some other correlates of Productivity varies widely across sectors, with agricultural productivity in productivity growth have slowed or gone into reverse. EMDEs lagging both advanced economies and other sectors in EMDEs. These include investment weakness; a slower pace of Fading gains from resource reallocation towards more productive sectors urbanization; maturing gains from macroeconomic have accounted for about half of the post-crisis slowdown in productivity growth. Within-sector productivity growth has also slowed. stability and global integration; and diminishing improvements or stagnation in educational attain- A. Sectoral productivity relative to B. Sectoral productivity in EMDEs ment, gender equality, and governance. country average relative to advanced-economy levels A large number of variables have been proposed as possible drivers of productivity (Annex 3.3).16 These drivers can be grouped into three categories: the quality and quantity of factors of production and the effectiveness of their use, such as capital, education, and innovation; the supporting economic environment, such as institutions and social conditions; and the degree of market development, such as trade integration and C. Composition of value-added D. Contribution to aggregate financial market development. This section productivity growth presents the correlations of productivity growth with initial conditions for these drivers and, in a second step, discusses the evolution of these drivers. Correlation between productivity growth and its drivers Methodology. The contributions of potential Source: APO productivity database, Expanded African Sector Database, Groningen Growth Development Center Database, Haver Analytics, ILOSTAT, OECD STAN, United Nations, World drivers of productivity growth are estimated in a KLEMS. cross-section regression to identify the main initial Note: Sample includes 80 economies (including 46 EMDEs, of which 8 are LICs). “Manuf.” includes mining and utilities; “Finance” includes business services; “Government” includes personal services. country features associated with subsequently A. Deviation of sectoral productivity level from country-specific average productivity. B. Grey horizontal line indicates 50 percent. higher long-term productivity growth (1960-2018 C. Share of total value added. and 1995-2018) for 59 countries, including 38 D. Growth “within sector” shows the contribution to aggregate productivity growth of each sector holding employment shares fixed. The ‘between sector’ effect shows the contribution arising from EMDEs. Key correlates of productivity growth are changes in sectoral employment shares. Median of the country-specific contributions. Click here to download data and charts. selected from a pool of 29 variables by Bayesian techniques to systematically exclude variables that which have above-average productivity levels in have poor explanatory power for productivity LICs. Despite having high productivity levels, the growth and overlapping variables which reflect the mining and extraction sectors often offer limited same underlying driver (Annex 3.3). scope for expanding employment outside of commodity booms, and therefore few Key initial conditions for higher productivity opportunities for sustainable sectoral reallocation. growth. Productivity in economies with favorable starting conditions in the 1960s grew significantly Long-run drivers of faster than other economies annually. A better educated workforce (proxied by years of schooling) productivity growth and stronger institutions (proxied by improvements in the rule of law), greater During the pre-crisis productivity surge in EMDEs, growth was highest in those economies with more favorable institutional environments, more developed product and factor markets, and higher or higher- 16 See Durlauf, Johnson, and Temple (2005) and Kim and Loayza quality factor inputs. Subsequently, improvements in (2019). G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 215 BOX 3.3 Patterns of total factor productivity: A firm perspective There is substantial variation in firm-level total factor characteristics in EMDEs to address the following productivity (TFP) across industries and across regions. Weak questions: firm productivity in emerging market and developing economies (EMDEs) partly reflects the divergence between a • How does firm-level TFP vary across EMDE sectors few highly productive firms and a large number of firms that and regions? operate far from the productivity frontier. The difference between frontier and laggard firms is, on average, larger in • What firm characteristics account for the dispersion in EMDEs than in advanced economies. Among EMDE firms, TFP? large firms tend to be more productive than small firms. TFP variation across sectors and regions Firms in technology-intensive industries, mainly located in East Asia and Pacific (EAP), Europe and Central Asia Productivity varies across firms, within sectors, and across (ECA), and South Asia (SAR), tend to be more productive regions (Goñi and Maloney 2017). By focusing on TFP, than firms in more traditional sectors. Measures to promote differences due to capital deepening or other factor inputs exports and improve business climates can help close the can be abstracted from. This allows to identify where observed TFP gap. TFP dispersion and gaps are the largest, and where steps are needed to improve productivity. Firm-level TFP data Introduction are obtained from surveys conducted by the World Bank Firm-level productivity in emerging markets and from 2007 to 2017 (Cusolito et al. 2018). The database of developing economies (EMDEs) has been low relative to survey results contains TFP for 15,181 manufacturing advanced economies, and growth has lost momentum over firms in 108 EMDEs, including 20 low-income countries the past decade. This has diminished prospects among (LICs). A cross-sectional analysis of the firm-level TFP many EMDEs to catch up with the advanced economies database is undertaken, which complements longitudinal (Andrews, Criscuolo, and Gal 2016; Cusolito and studies that use micro-level panel data, but with a smaller Maloney 2018). country coverage.2 Two measures of TFP are constructed: output and value-added revenue TFP measures. The Numerous factors have been identified as underlying the latter is obtained by subtracting the value of intermediate low firm-level productivity observed in EMDEs: weak inputs (materials, electricity, etc.) from output before institutions and pervasive informality, slow technology computing TFP. TFP measurement challenges are innovation and adoption, subdued investment and poor discussed in Annex 3.5. quality infrastructure, low human capital and poor firm management practices, protectionist trade policies and TFP across sectors. Differences in firm-level TFP across weak economic integration (Cusolito and Maloney 2018; sectors have been frequently emphasized in the literature.3 World Bank 2019d, 2019e).1 Moreover, outdated On average, firms in technology-intensive industries have technologies, lagging innovation, misallocation of labor to higher TFP than those in other sectors (Figure 3.3.1.A). inefficient sectors, and market rigidities weigh on Technology-intensive industries, denoted by TINT, productivity and contribute to dispersion in total factor include computing and electrical machinery, precision productivity (TFP) across countries (Araujo, equipment, electronics, information, and communication Vostroknutova, and Wacker 2017; Bahar 2018; Syverson sectors (as in Fernald 2015). One explanation for this 2011). In some EMDEs, low participation in global value observation is that firms operating in a technology- chains, or lack of openness to foreign direct investment intensive industry rely more on research and development and migration, has resulted in missed opportunities for a (R&D) and network linkages than physical assets, and as productivity boost through the transfer of innovative processes and managerial capabilities (Goldberg et al. 2010; World Bank 2019d). 2 This analysis does not explore the time series dimension because World Bank’s firm output and input data used to construct TFP This box undertakes a cross-sectional study to analyze estimates were collected at different time in different countries. For firm-level TFP patterns, and maps these to firm example, these firm surveys were conducted in 2007 in South Africa and in 2017 in Ecuador. Moreover, the number of surveyed firms in many countries is small, which does not allow to conduct robust within and Note: This box was prepared by Cedric Okou. cross-country comparisons. 1 Many studies focus on labor productivity, which depends on both 3 See for example, Bartelsman and Doms (2000) and Levchenko and TFP and capital per worker–also known as capital deepening. Zhang (2016). 216 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 3.3 Patterns of total factor productivity: A firm perspective (continued) FIGURE 3.3.1 Firm TFP and distance-to-frontier in EMDEs by industry Firms in technology-intensive industry (TINT) have higher average TFP. These technology-intensive firms are also more tightly clustered around their industry-specific frontier than firms in other sectors. A. TFP estimates, by industry B. Distance-to-frontier and average C. Distance-to-frontier and average output TFP, by industry value-added TFP, by industry Source: World Bank Enterprise Surveys, World Bank. Note: Firm-level TFP is computed using a Cobb-Douglas production function for each industry, assuming that elasticities of output with respect to inputs are the same across countries in a given income group. The distance-to-frontier of TFP is computed within each industry, excluding the top 2.5 percent of firms. For each sector, the location shows the average and the size of the marker (circle) is proportional to one standard deviation of distance to frontier of TFP. Averages and standard deviations are computed using survey weights. Sample includes 15,181 firms in 108 EMDEs, including 20 LICs, for the period 2007-17. Firms operate in 15 industries: APPA = apparel, CHEM = chemicals, FABM = fabricated metals, FOOD = food, FURN = furniture, LEAT = leather, MACH = non-electrical machinery, META = metals, MINE = non-metallic minerals, MOTO = motor vehicles, PAPE = paper, RUBB = rubber, TEXT = textiles, TINT=technology-intensive, WOOD = wood. The technology-intensive industry (TINT) includes firms in computing and electrical machinery, precision equipment, electronics, information, and communication sectors. A. In the manufacture of paper (PAPE) industry, the value-added TFP is positive and much higher than the corresponding (negative) output TFP due to a relatively high elasticity of output with respect to intermediate inputs. B. C. Distance-to-frontier of firm-level TFP (minus) and TFP (log), by industry. The right-hand-side y-axis represent the frontier. Click here to download data and charts. such can reap the benefits of technology to boost in other regions (Figure 3.3.2.A). EAP also has the highest productivity (Chevalier, Lecat, and Oulton 2012). proportion of large size firms and firms exporting more than half of their sales (Figure 3.3.2.C and Figure Distance to TFP frontier across sectors. TFP dispersion 3.3.2.D). Most firms in technology-intensive industries are may signal rigidities in the generation, transfer and located in EAP, Europe and Central Asia (ECA), and acquisition of technology across firms in a sector. To assess South Asia (SAR) (Figure 3.3.2.B; regional boxes in within-sector productivity dispersion, a firm’s distance to Chapter 2). Perceptions of corruption and licensing as an industry-specific TFP frontier is computed.4 Firms in obstacles for firm operation seem to correlate negatively basic manufacturing industries, such as non-electrical with total factor productivity (Figure 3.3.2.E-F). machinery (MACH), textiles (TEXT), leather (LEAT), and basic metals (META), are not only on average less Robustness of TFP dispersion. Substantial TFP dispersion productive than firms in other sectors, but also relatively may signal misallocation of factor inputs or rigidities in the far from their industry-specific frontiers (Figure 3.3.1.B generation, transfer, and acquisition of technology across and 3.3.1.C). By contrast, firms in technology-intensive firms (Hsieh and Klenow 2009). However, commonly industries (TINT) are more tightly clustered around their used dispersion metrics can also reflect mismeasurements, industry-specific frontiers and are more productive.5 quality differences, adjustment costs, markups, and investment risks, among other factors. Recent evidence TFP across regions. Across regions, firms in East Asia and shows that half of the dispersion is unrelated to Pacific (EAP) are, on average, more productive than those misallocation, and driven rather by markups and technology wedges (Cusolito and Maloney 2018). Thus, 4 For a given firm i , the distance to an industry-specific TFP frontier dispersion results should be interpreted with caution. (97.5th quantile) is computed as DTFi = TFP0.975 - TFPi≤0.975. The top 2.5 Nonetheless, the variation in distance to frontier in percent firm-level TFP values are dropped to minimize the impact of technology-intensive industries is less than one-fifth of that extreme values. Results are robust to alternative 1 and 5 percent cutoffs of in basic manufacturing industries (leather, metals, top firm TFP values. 5 This finding is broadly in line with the evidence in Hallward- machinery), suggesting that firms in technology-intensive Driemeier and Nayyar (2017). industries are much closer to their sector-specific frontier. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 217 BOX 3.3 Patterns of total factor productivity: A firm perspective (continued) FIGURE 3.3.2 Firm TFP by regions Firms in EAP are more productive than those located in other EMDE regions. EAP also has the highest share of large-size firms and those exporting more than half of their sales. Most firms in technology-intensive industry (TINT) are located in EAP, ECA, and SAR. Perceptions of corruption and licensing as obstacles for firm operation correlate negatively with total factor productivity (TFP). A. Firm-level TFP, by region B. Percentage of firms in each region, C. Firm size, by region by industry D. Exporting firms, by region E. Perception of corruption, by region F. Perception of licensing obstacles, by region Source: World Bank Enterprise Surveys, World Bank. Note: Firm-level TFP is computed using a Cobb-Douglas production function for each industry, assuming elasticities of output with respect to inputs are the same across countries in a given income group. Unweighted regional averages are computed. Sample includes 15,181 firms in 108 EMDEs, including 20 LICs, for the period 2007-17. EAP = East Asia and Pacific, ECA = Europe and Central Asia, LAC = Latin America and the Caribbean, MNA = Middle East and North Africa, SAR = South Asia, and SSA = Sub-Saharan Africa. A. Solid lines are averages of output TFP (log) for EMDEs (orange) and LICs (red). EMDEs = emerging markets and developing economies, LICs = low-income countries. B. Bars show in each industry the percentage of firms in each region, by industry. Firms operate in 15 industries: APPA = apparel, CHEM = chemicals, FABM = fabricated metals, FOOD = food, FURN = furniture, LEAT = leather, MACH = non-electrical machinery, META = metals, MINE = non-metallic minerals, MOTO = motor vehicles, PAPE = paper, RUBB = rubber, TEXT = textiles, TINT = technology-intensive, WOOD = wood. The technology-intensive industry (TINT) includes firms in computing and electrical machinery, precision equipment, electronics, information, and communication sectors. C. Firm size in terms of number of employees. D. Share of exporting firms. High, medium, and low exports firms export more than 75 percent, between 50 and 75, and up to 25 percent of their sales, respectively. E. Share of firms that perceive corruption as an obstacle for their operations. F. Share of firms that perceive licensing and permits as an obstacle for their operations. Click here to download data and charts. Firm characteristics associated with higher EMDEs: within-firm upgrading and spillovers, regulatory TFP growth environment, and managerial ability. Heterogeneous characteristics related to entering, Within-firm upgrading and technology spillovers. incumbent, and exiting firms can explain the observed Controlling for both size and exports, firms in the patterns of TFP dispersion (Bartelsman and Doms 2000). technology-intensive industry are on average much closer A large and expanding literature points to three broad to the TFP frontier than firms in traditional industries categories of correlates of sectoral TFP dispersion in such as non-electric machinery, food, and non-metallic 218 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 3.3 Patterns of total factor productivity: A firm perspective (continued) FIGURE 3.3.3 Distance-to-frontier of TFP, firm characteristics, and regulations The average firm in the technology-intensive industry (TINT) is significantly closer to the frontier than the average firm in non- electric machinery (MACH), food (FOOD), and non-metallic minerals (MINE) industries, after controlling for firms’ size and exports. As firms grow by number of employees and increase their ratios of exports to total sales, they move closer to the TFP frontier. A conducive business environment can enhance firm-level TFP. Improvements in business freedom and control of corruption are correlated with a reduction in the distance-to-frontier of TFP. A. Distance to TFP frontier differential B. Distance to TFP frontier differential C. Distance to TFP frontier differential between traditional industries and the between firms in lowest and highest between firms in lowest and highest technology-intensive industry quartile of firm size and exports quartile of business environment Source: World Bank Enterprise Surveys, World Bank. Note: Firm-level TFP is computed using a Cobb-Douglas production function for each industry, assuming that elasticities of output with respect to inputs are the same across countries in a given income group. The distance-to-frontier (DTF) of TFP is computed within each sector, excluding the top 2.5 percent of firms. Sample includes 15,181 firms in 108 EMDEs, including 20 LICs, for the period 2007-17. A. Distance-to-frontier of TFP differential between traditional industries, such as manufacturing of non-electric machinery (MACH), food (FOOD), and non-metallic minerals (MINE), and the technology-intensive (TINT) industry, controlling for firm characteristics (firm size and exports). Based on OLS regressions of the DTF of TFP (dependent variable) on industry dummies, controlling for firm characteristics and using the technology-intensive industry (TINT) as the base category as per Annex 3.5. B. Distance to TFP frontier differential between the median firm in the lowest quartile and highest quartile of firms in terms of firm size (number of workers) and exports (share of exports in total sales). Based on OLS regressions of the DTF of TFP (dependent variable) on industry dummies, controlling for firm characteristics and using the technology-intensive industry (TINT) as the base category (Annex 3.5). A positive DTF differential implies that firms in the lowest quartile in terms of size and exports are far from the frontier relative to firms in the highest quartile. The lowest quartile of exports is zero, as more than half of firms have no exports. C. Distance to TFP frontier differential between the median firm in the lowest quartile and highest quartile of firms in terms of business freedom and control of corruption index, controlling for firm characteristics. Based on OLS regressions of the DTF of TFP (dependent variable) on industry dummies and business environment quality, controlling for firm characteristics and using a technology-intensive industry (TINT) as the base category as per equation 3. A positive DTF differential implies that firms in the lowest quartile in terms of business freedom and control of corruption are far from the frontier relative to firms in the highest quartile. Click here to download data and charts. minerals industries (Figure 3.3.3.A). Knowledge, practices and organization structures of “nearby” highly experience, R&D, and information technology can raise productive firms (Dercon et al. 2004; Syverson 2011). TFP through improvements in product quality and Knowledge is also transferred through contacts with other production process upgrading within firms.6 Firms with a firms, courtesy of trade, foreign direct investment and large number of employees are significantly closer to the migration (De Loecker 2007). Firms with a high share of TFP frontier, as larger firms can invest more in R&D and exports are significantly closer to the TFP frontier. A firm bring together a richer set of ideas. On average, the in the top quartile of exports, measured as a share of productivity of a firm in the highest quartile of size is exports in total sales, is about 4 and 6 percent closer to about 12 and 22 percent closer to output and value-added output and value-added TFP frontiers relative to a firm in TFP frontiers relative to a firm in the lowest quartile of the lowest quartile of exports (Figure 3.3.3.B). Enabling size (Figure 3.3.3.B). Moreover, technology in frontier effective innovation policies appears critical to boosting firms can have positive spillovers for productivity in other innovation gains (Cirera and Maloney 2017). firms through agglomeration linkages and cross-border flows of goods, capital and people. Firms can reap Regulatory environment. Institutions reflect political and agglomeration benefits by emulating the best production legal forces that shape social and economic environments. Regulations and policies affect firms’ productivity through incentives to acquire human capital, physical capital, and technology (Bartelsman and Doms 2000). Firm 6 See Brynjolfsson and Hitt (1995) and Goldberg et al. (2010). productivity tends to drop in poorly-regulated markets, G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 219 BOX 3.3 Patterns of total factor productivity: A firm perspective (continued) due to adverse incentives and the lack of creative practices can raise productivity by more than 10 percent destruction (Goldberg et al. 2010). In contrast, (Van Reenen 2011). A policy shift that is more focused on improvements in the business environment are associated enhancing firm managerial capabilities can, therefore, with lower distance to TFP frontier, even after controlling strengthen production synergies and bolster TFP gains for firm characteristics. Conducive regulatory practices— (Cusolito and Maloney 2018). reflected in highest quartile values of business freedom index—may entail up to 9 percent reduction in the Conclusion distance-to-frontier of TFP relative firms in the lowest quartile. Similarly, high quality governance—proxied by The dispersion of firm-level TFP within and across the top quartile estimates of control of corruption index— industries in emerging markets and developing economies is associated with up to 12 percent drop in the distance to (EMDEs) is associated with various firm characteristics. TFP frontier relative to firms in the bottom quartile TFP dispersion correlates negatively with firm size, partly (Figure 3.3.3.C). because large firms can invest more in R&D to innovate. Exports also facilitate the transfer and adoption of new Managerial ability. TFP also reflects how efficiently technologies, and therefore, can help close the gap between productive factors—labor, capital, and intermediate laggards and frontier firms. Moreover, a conducive inputs—are assembled. Through their talents or the business climate characterized by a greater freedom in quality of their practices, managers coordinate the entrepreneurship and less corruption can support TFP integration of factor inputs in the production process. improvements. Undertaking policies to support R&D and Management and organizational styles may vary across innovation, promote exports, combat corruption, increase firms due to competition, location, ownership, and trade the ease of doing business, and enhance firm managerial ties. Intervention-led improvements in management capabilities, appears critical to boosting productivity. innovation (proxied by higher per capita patents), manufacturing and service sectors (Box 3.2). The stronger investment (as a share of GDP), higher level of education and investment also produces levels of urbanization (proxied by population larger impacts on productivity in EMDEs in the density), price stability, and a diverse and long-run estimation, highlighting their sophisticated economic structure (proxied by the importance at lower levels of productivity. Since economic complexity index of Hidalgo and 1995, the relationship between labor productivity Hausmann 2009), are all significantly associated and the economic complexity of tradable goods with higher productivity growth (Figure 3.7).17 has strengthened in EMDEs. Differences between EMDEs and advanced Evolution of the drivers of productivity economies. The estimated impact of improved levels of each driver of productivity growth Pre-crisis improvements. There were substantial depends on the stage of development and gains in many of the underlying drivers of therefore differs between EMDEs and advanced productivity growth in the pre-crisis period, economies. The extent of urbanization has a larger growing faster in EMDEs than advanced impact on productivity growth in EMDEs than in economies (Figure 3.8). The selected drivers can advanced economies, reflecting higher returns to be aggregated to an index based on the size of the reallocation of workers away from rural their estimated impacts on productivity— agricultural production to higher productivity demographics, economic complexity, the number of patents filed, and price stability are all considered to be key determinants of productivity 17 These are largely consistent with existing studies which tend to growth over this period by the econometric have shorter time spans and smaller cross-sections (Durlauf, model. Cumulatively over 1995-2008, produc- Kourtellos, and Tan 2008; Kim and Loayza 2019). tivity in the one-quarter of EMDEs with the most 220 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 3.7 Impact of drivers on productivity growth EMDEs. For educational attainment, growth has Productivity in economies with favorable initial conditions grew by up to 0.8 been three times higher than in advanced percentage point per year faster than other economies. The scale of these economies. Nonetheless, as countries catch up (as effects varies over time and between EMDEs and advanced economies. In measured by average years of schooling), the 1960, the importance of innovation and economic complexity was lower in EMDEs. Demography and economic complexity have become increasingly potential for further growth has slowed.18 important determinants of EMDE productivity growth in recent decades. Other factors that had helped spur EMDE A. Effects of initial level of drivers on B. Effects of initial level of drivers for productivity growth also have deteriorated since productivity growth, 1960-2018 EMDEs on productivity growth, 1960- the crisis. For example, the trend toward 2018 vs. 1995-2018 broadening production to a more diverse range of products at more upstream stages of the value chain slowed partly because the expansion of global value chains stagnated after 2008 (World Bank 2019d). In addition, improvements in inequality and measures of institutional quality have also stagnated or declined in many countries. Finally, gains in price stability, which had significantly improved operating environments for Source: World Bank. firms in the 1990s, slowed (Ha, Kose, and A. B. Estimated marginal contribution to annual long-term productivity growth if the driver improves from the 25th to the 75th percentiles. Sample includes 59 economies, 36 of which are EMDEs. Ohnsorge 2019). Groups which are not significant in both 1960-2018 and 1995-2018 (Finance, Income equality, and health) are excluded from the chart. Variables corresponding to each concept are: Institutions = ICRG rule of law index, Geography=share of non-tropical area, Innovation=patents per capita, Investment=investment to GDP ratio, Income equality=(-1)*Gini coefficient, Urbanization=urban population (% total), Econ. complexity = Economic Complexity Index of Hidalgo and Hausmann Prospects for productivity (2009), Education=years of schooling, Demography=share of working-age population, Gender equality= female average years of education minus male average years. See Annex 3.3 for details. growth B. Marginal contribution of demography for 1995-2018 is 0.83. Click here to download data and charts. The post-crisis weakness in several fundamental drivers of productivity growth is expected to persist or deepen. The weak outlook for the drivers can be favorable initial conditions grew by nearly 15 improved though a concerted reform effort. percent more than productivity in those with the least favorable initial conditions. Among LICs, the Weakening investment. The post-crisis period has differential between the two groups was even been characterized by pronounced investment larger (53 percent). LICs were better able to weakness reflecting adverse terms-of-trade shocks benefit from catch-up growth in the presence of for commodity exporters, slowing foreign direct favorable initial conditions. investment inflows for commodity importers, spillovers from advanced-economy growth Post-crisis slowdown in improvements. The pace weakness, heightened policy uncertainty, and of growth of the drivers most strongly associated private debt burdens (World Bank 2017). The with productivity growth has slowed in EMDEs legacy of weak investment since the crisis and since 2008, consistent with the slowdown in diminishing long-term outlook for investment productivity growth over this period (Figure 3.9). growth raises concerns about future productivity growth (World Bank 2019b). Moreover, subdued Investment growth in EMDEs slowed, reflecting investment growth, especially in R&D-dependent weak activity and spillovers from advanced sectors, can hinder technological progress and economies, weaker growth of commodity demand, TFP growth through weaker capital-embodied and political uncertainty. In addition, earlier technological change (Adler et al. 2017). favorable demographic trends in many EMDEs have waned as the population ages. From 2018 to 2030 the working-age share of the population is 18 While the gap in average years of education with advanced expected to decline by 3 percentage points in economies has declined, substantial gaps in the quality of education advanced economies and 2.5 percentage points in remain (World Bank 2018b). G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 221 Slower growth at the technology frontier. There FIGURE 3.8 Pre-crisis developments in productivity has been a broad-based slowdown in both labor drivers and productivity growth productivity and TFP growth in advanced All drivers of productivity growth in EMDEs, except for innovation, gender economies since the early 2000s with limited signs equality and institutions, improved more than in advanced economies of an impending upturn. To the extent that this during the pre-crisis period, helping to narrow the productivity gap with advanced economies. There was a strong link between drivers and reflects slowing productivity growth in productivity growth— those economies with better initial conditions in the multinationals and the origins of foreign direct 1990s grew at faster rates subsequently. The benefits of improving drivers are larger for LICs. investment—two major channels for knowledge and technology spillovers to EMDEs—this is likely to weigh on EMDE productivity, too A. Share of EMDEs with faster improvements in drivers relative to B. Quartiles of productivity drivers and average EMDE productivity (Wooster and Diebel 2010). However, there are advanced economies, 1995-2008 growth, 1995-2008 mixed views on the prospects of groundbreaking technological progress that could return growth to historical norms, and also spillovers to EMDEs. On the one hand, the impact on productivity growth of new innovations compared to 20th- century innovations seems to be reduced (Fernald 2015; Gordon 2016). On the other hand, recently introduced new digital technologies and those on the horizon such as artificial intelligence and Source: Barro and Lee (2015); International Monetary Fund; Observatory of Economic Complexity; innovations in IT sectors may begin to feed United Nations; World Bank, World Development Indicators. A. Share of EMDE countries whose improvement in drivers are larger than average changes for through to measured productivity (Cusolito and advanced economies. Variables corresponding to each concept are (sample in parentheses): Institutions (74) = WGI Rule of Law Index, Innovation (30) = patents per capita, Investment (72)= Maloney 2018). investment to GDP ratio, Income equality (72) = (-1)*Gini coefficient, Urbanization (74) = Urban population (% total), Econ complexity (56) defined as Economic Complexity Index of Hidalgo and Hausmann (2009), Education (69) = years of schooling, Demography (74) = share of working age Fewer opportunities for technology transfer. population, Gender equality (28) = Ratio of female to male labor market participation. B. Average level of productivity growth and “index of drivers” in each quartile over 1995-2008. “Index Substantial productivity gaps to the frontier are of drivers” created by weighting normalized levels of each potential driver in chart A by its estimated impact on productivity growth (Figure 3.7; Annex 3.3). still present in EMDEs, providing opportunities Click here to download data and charts. for rapid productivity growth. However, routes to technology transfer are narrowing. The expansion of global value chains has come to a halt in the development increasingly challenging in the future post-crisis period after rapid expansion in the pre- (Hallward-Driemeier and Nayyar 2017; Sinha crisis period (World Bank 2019d). Rising 2016). Furthermore, gains from faster productivity implementation of protectionist measures risks growth in the agricultural sector, freeing up further compounding the weakness in global value workers to transition to other sectors, have chains and trade. Moreover, firms in EMDEs may declined. lack the necessary capabilities to adopt new technologies without sustained improvements in Rising debt risk in EMDEs. Amid record-high human capital such as enhancements in EMDE debt, a wide range of adverse shocks could educational quality and management abilities precipitate a financial crisis in EMDEs, which despite the progress in education attainments could do severe damage to productivity (Box 3.4). (Cirera and Maloney 2017). Since 2010, total debt in EMDEs has risen markedly by 54 percentage points, to 168 percent A more challenging environment for structural of GDP in 2018, with private debt growing faster transformation. As highlighted in Box 3.2, the than public debt, reaching 120 percent of GDP in contribution to productivity growth from the 2018 (Chapter 4). Low productivity growth and manufacturing sector has been in decline and rising sovereign debt burdens may even reinforce presents fewer opportunities for EMDE one another (Posen and Zettelmeyer 2019). productivity growth. Secular trends, such as a declining employment share in the manufacturing Climate change. Over the longer-term, climate sector in some economies and risks from change will likely increase the challenges to automation will make manufacturing-led improving productivity in the agricultural sector, 222 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 3.9 Post-crisis slowdown of the drivers of structures more readily (Maestas, Mullen, and productivity growth Powell 2016). The working-age share of the In EMDEs, improvements in a broad range of productivity drivers slowed population rose by 13 percentage points of the after 2008. Investment growth slowed to one-third of its pre-crisis rate in population during 1995-2008 in MENA, the EMDEs. Working-age population shares are expected to contract in the coming years. And the growth of educational attainment has also slowed fastest-growing region, and 8 percentage points in as EMDEs reduce the gap with advanced economies. EAP, the second-fastest growing. In the coming years, EMDE populations are set to age. In EAP A. Share of EMDEs with a post-crisis B. Average investment growth and ECA, the working-age share of the population slowdown in the growth of underlying is expected to decline by 3-4 percentage points of drivers of productivity the population by 2030, while, in LAC, MENA, SAR, and SSA it will stagnate. Policy implications Concerns about prospects for productivity growth in EMDEs call for a renewed emphasis on structural policies that can unlock productivity gains, but undertaking the right structural policies is C. Average annual growth in D. Change in working-age share challenging. Drawing on the findings in this chapter, educational attainment of the population four strands of policy options emerge. The results suggest that a four-pronged policy approach can lift productivity. First, policies can raise labor productivity economy-wide by stimulating private and public investment and improving human capital. Second, policies can foster firm productivity by exposing firms to trade and foreign investment and strengthening human Source: Barro and Lee (2015); International Monetary Fund; Observatory of Economic Complexity; capital, and upgrading workforce skills including Penn World Table; IMF World Economic Outlook; United Nations; World Bank, World Development Indicators; Wittgenstein Centre for Demography and Global Human Capital. that of firm managers. Third, policies can facilitate A Post-crisis slowdown defined as the share of economies where improvements in each underlying driver of productivity during 2008-2017 was less than zero or the pace of improvement during the the reallocation of resources towards more pre-crisis period 1998-2007. Variables corresponding to each concept are (sample in parentheses): Investment (69)=investment to GDP ratio, Demography (74)=share of working-age population, productive sectors and a more diversified set of Innovation (33)=patents per capita, Gender equality(32)= Share of female labor market participation sectors. Finally, to be effective, these policies need rate to male, Urbanization (74)=Urban population (% total), Institutions (74)= WGI Rule of Law Index, Income equality (72)=(-1)*Gini coefficient, Education (72)=years of schooling, ECI (55) defined as to be set in the context of a growth-friendly Economic Complexity Index of Hidalgo and Hausmann (2009). Price stability excluded due to demand-side influences on inflation following the global financial crisis. macroeconomic and institutional environment B. GDP-weighted average annual investment (gross fixed capital formation) growth. C. GDP-weighted change (at 2010 prices and exchange rates) in average years of education. (Cirera and Maloney 2017). D. Changes in the working-age share of the population (aged 15-64). Click here to download data and charts. Within these four broad strands, specific priorities depend on country characteristics. For example, countries with large unmet investment needs may with large falls in crop yields expected as global want to prioritize expanding fiscal resources to temperatures rise (Fuglie et al. 2019). Agriculture achieve more and better public investment. currently accounts for 30 percent of GDP in LICs, Countries with anemic private investment may compared to just 9 percent in non-LIC EMDEs. want to prioritize business climate and In addition, EMDEs in several regions are heavily institutional reforms, reduce support for state- reliant on agriculture: around half of employment owned enterprises, and broadening access to is in the agricultural sector in SAR and SSA. finance to allow private sector investment to flourish. Countries with predominantly low- Less favorable demographics. Younger popu- skilled workers may want to improve health and lations and larger working-age population tend to education for workers and managers alike. adopt new technologies, skills, and organizational Countries with lethargic innovation may want to G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 223 BOX 3.4 Debt, financial crises, and productivity High debt levels increase the probability of financial crises and weigh heavily on productivity growth through a wide range of channels. During debt accumulation episodes associated with financial crises, cumulative productivity gains three years into the episode are 2 percentage points lower than in episodes without crises. Financial crises are accompanied by large and protracted declines in productivity: five years after the financial crisis, productivity is 6.5 percent lower than it would have been without a crisis. Introduction positive effects on productivity and growth (Reinhart and Rogoff 2010; Poirson, Pattillo, and Ricci 2004). However, Productivity growth is vulnerable to a range of adverse debt accumulation can impede productivity by shocks including those associated with financial crises, encouraging a misallocation of resources towards projects especially in the context of rapid debt accumulation that yield short-term returns at the expense of long-term (Chapter 4). Following the global financial crisis and returns or offer low risk at the expense of high returns subsequent global recession of 2007-09, a broad range of (Poirson, Pattillo, and Ricci 2002; Checherita-Westphal countries experienced a rapid accumulation of debt and Rother 2012). These short-term projects can include together with a significant slowdown of productivity those that rely heavily on returns from asset price growth. Debt accumulation raises both long-term and appreciation on expectations of rapid future growth short-term risks to productivity growth. In the long-term, (Claessens and Kose 2017, 2018). it can lead to misallocation of resources towards low productivity projects, worsen investment prospects, weigh Debt overhangs. Rapid debt accumulation can lead to on competitiveness, and curb technological transfers debt overhangs whose debt service crowds out productive embodied in investment.1 In the short-term, debt investment.3 At the firm level, a large outstanding debt accumulation also increases the probability of financial stock can weigh on investment and, hence, the crises that sharply raise borrowing cost, worsen balance productivity growth that technology embedded in this sheets and depress productivity growth, which can last investment can generate. At the government level, debt over an extended period.2 service on high debt may crowd out other productivity- enhancing spending, including for education, health or Against this backdrop, this box discusses the linkages infrastructure. between productivity and financial crises as well as rapid debt accumulation. Specifically, it addresses the following Policy uncertainty. Especially high government debt two questions: increases uncertainty about growth prospects. For investors, large projected government debt service cost • Through which channels does debt affect creates policy uncertainty because they may eventually productivity? compel governments to introduce distortionary taxation • What is the empirical link between financial crises (including on future investment returns), curtail growth- and productivity? enhancing spending, or delay reforms that may support innovation and productivity (IMF 2018). Such Channels of transmission uncertainty lowers incentives to invest in productivity- enhancing technologies (Krugman 1988). Elevated debt levels can affect productivity growth via several channels. These include misallocation of resources, Higher probability of financial crises. Higher debt policy uncertainty and debt overhangs that weigh on increases the probability of financial crises. These tend to productivity-enhancing investment, and a higher proba- be associated with severe short-run productivity losses and bility of financial crises. lasting productivity weaknesses. Financial crises include debt, banking, and currency crises. Misallocation of resources. If used to fund productive investments with high rates of return, debt can have • Sovereign debt crises. Higher government debt may encourage governments to shift towards lower-cost Note: This box was prepared by Alistair Dieppe, Sinem Kilic Celik, and Cedric Okou. 1 Blanchard and Wolfers (2000); Bulow and Rogoff (1989). 3 Debt overhang can occur in the presence of high levels of debt, as 2 See Aguiar and Gopinath (2006), Arteta and Hale (2008), Blanchard, potential investors hold back new investments because they face Cerutti, and Summers (2015), Cerra and Saxena (2008, 2017), Furceri heightened uncertainty about tax rates on future investment returns, and Mourougane (2012a), Jordà, Schularick, and Taylor (2013), and given the government’s large projected revenue needs to service the Reinhart and Rogoff (2009, 2010). outstanding debt. 224 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 3.4 Debt, financial crises, and productivity (continued) FIGURE 3.4.1 Productivity in debt accumulation episodes and financial crises About 40 percent of all episodes of debt accumulation are associated with financial crises. During those episodes, productivity gains are significantly lower than during other episodes. Specifically, a financial (banking, currency and debt) crisis is accompanied on average by a 6.5 percent cumulative decline in the level of labor productivity after 5 years, and the negative effect is protracted, exceeding 7 percent at an 8 year-horizon. A. Total debt accumulation episodes B. Cumulative productivity gains during C. Impact of financial crises on EMDE around crises episodes of rapid debt accumulation productivity and output levels Source: World Bank. A. Share of total (government and private) debt accumulation episodes that were associated with financial (banking, currency, debt) crises. B. * and ** indicates 10 and 5 percent significance level for the difference between productivity growth during the median total debt accumulation associated with crises and the median total debt accumulation episode not associated with crises. C. Bars show the average loss in labor productivity and output levels in EMDEs, expressed in percent, at impact, 1, 2, … and 8 years after a financial crisis (Laeven and Valencia 2018). Financial crises include banking, currency and debt crises. Whiskers represent 90 percent confidence intervals. The estimation is based on local projection method (Jordà 2005), which includes control variables (country fixed effects, lagged shocks, forward bias correction terms, and lagged TFP growth) and bias correction (Teulings and Zubanov 2014) for forward values of the crisis dummy between time t and t+h-1. Click here to download data and charts. but higher-risk debt issuance such as at shorter unemployment erodes human capital.5 Because of maturities or in foreign currency (Kalemli-Özcan, their shorter duration, currency crises are typically less Laeven, and Moreno 2018). This heightens the harmful to productivity. However, combined banking probability that financial market stress precipitates a and currency crises can be particularly damaging for sovereign debt crisis that sharply raises investor risk economic activity and productivity. premia and borrowing cost.4 These tend to coincide with severe economic disruption just as sovereign debt Empirical link between financial crises and distress prevents governments from supporting productivity activity with counter-cyclical fiscal policy (Reinhart and Rogoff 2010). This depresses public and private Productivity gains during rapid debt accumulation investment and restricts other productivity-enhancing episodes. Long-term productivity gains during rapid debt public spending. accumulation episodes have been considerably lower when these debt accumulation episodes were associated with • Banking and currency crises. Other types of financial financial crises. As in Chapter 4, rapid debt accumulation crises, including systemic banking crises and currency episodes are defined as an expansion from trough to peak crises, can also do lasting damage to productivity of total debt-to-GDP ratios by more than one standard (Cerra and Saxena 2017; Oulton and Sebastiá-Barriel deviation, with troughs and peaks identified using the 2017). The disruptions in financial intermediation Harding and Pagan (2002) algorithm. This yields 190 during banking crises curb the funding of episodes, of which almost half were associated with productivity-enhancing technologies and typically financial crises—identified as in (Laeven and Valencia trigger recessions (De Ridder 2017). In the 2018) —at some point during the episode. subsequent protracted weakness, elevated long-term 4 Aguiar and Gopinath (2006); Arellano (2008); Sandri (2015). 5 See Blanchard and Wolfers (2000) and Furceri and Mourougane (2012b). G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 225 BOX 3.4 Debt, financial crises, and productivity (continued) In a debt accumulation episode accompanied by a crisis, percent at the end of five years (Figure 3.4.1). The effect median productivity three years into the episode was 3 persists into the eighth year. This is consistent with earlier percent higher than at the beginning of the episode. This studies that document protracted effects of financial crises is statistically significantly less than during a debt on productivity growth (Obstfeld 1996; Morris and Shin accumulation episode that was not associated with a crisis 1998; Barro 2001).6 (5 percent). The difference may reflect the severe short- term damage to productivity driven by financial crises. Conclusion Two years later (five years into the episode), productivity differences between the two types of episodes were no Financial crises weigh heavily on productivity growth longer statistically significant. through a wide range of channels. During debt accumulation episodes associated with financial crises, Impact of financial crises on productivity. The cumulative productivity gains three years into the episode productivity losses associated with financial crises are are 2 percentage points lower than in episodes without estimated in a local projections model of productivity crises. Financial crises are accompanied by large and levels in financial crises episodes. These episodes are protracted productivity losses—following an initial drop of identified as in (Laeven and Valencia 2018). There are 299 2.2 percent, productivity falls by a cumulative 6.5 percent financial crisis episodes for which labor productivity five years after the onset of the crisis. In this context, the estimates are available. 72 percent of these episodes rapid post-crisis build-up of debt in EMDEs increases occurred in 71 middle- or high-income EMDEs and 10 vulnerability to financial crises and represents an important percent in 13 low-income countries. downside risk to productivity growth (Chapter 4). Financial crises are accompanied by large and lasting productivity losses. Immediately after the onset of a debt 6 The damage to output and productivity does not differ statistically crisis, labor productivity declines on average by significantly over the first eight years following the crisis. about 2.2 percent and then falls by a cumulative 6.5 expose their private sectors to foreign knowledge slowdown in labor productivity growth. Elsewhere and technologies through greater trade and foreign (SSA, SAR), sizable infrastructure deficits restrict direct investment (Boxes 2.1-2.6). firms’ ability to improve productivity. Better physical capital and infrastructure—transport, Policy interactions can lead to unintended power, telecommunications—can reinforce a consequences. For instance, trade liberalization country’s competitiveness and boost its produc- reforms can increase the exposure of private sector tivity (Calderón, Moral-Benito, and Servén 2015). firms to foreign knowledge and frontier A key challenge is to prioritize investments to technologies, and boost productivity. However, reconcile large development needs with funding trade liberalization can also be associated with constraints and to improve public investment greater informality in the short-run if labor management. Low– and middle-income countries markets are not flexible, thus counteracting will need to spend between 4.5 to 8.2 percent of policies that aim at facilitating the reallocation of GDP on new infrastructure annually to 2030 in resources towards more productive sectors (Bosch, order to meet infrastructure-related Sustainable Goni, and Maloney 2007; World Bank 2019a). Development Goals (Rozenberg and Fay 2019).19 Therefore, these potential interactions should be Where fiscal space exists, governments should accounted for when designing a policy mix for a fund infrastructure spending in areas likely to country. generate high-returns. SSA is estimated to have the Improving factors of production 19 SDG targets for universal access to safely managed water, Meet infrastructure investment needs. In several sanitation, and hygiene services, improved irrigation infrastructure to regions (ECA, MNA, SAR), weaker rates of capital improve food supplies, universal access to electricity and improved deepening accounted for most of the post-crisis transport infrastructure. 226 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 3.10 EMDE infrastructure and education gaps highest infrastructure deficit required to meet the Infrastructure needs to meet the Sustainable Development Goals are SDGs (Figure 3.10). Poor infrastructure, such as highest in SSA. While education gaps, measured as years of schooling, power supply problems, have been found to lower are closing in many regions, they remain large in SAR and SSA. The gaps manufacturing TFP in Bangladesh and reduce to advanced economy levels are even larger after adjusting for educational quality. export diversification in lower-income EMDEs (Osakwe and Kilolo 2018). A range of A. Infrastructure gaps B. Years of education and learning- infrastructure investments in the road and adjusted years of education (2017) telecommunications networks in South Africa were found to have positive effects on manufacturing TFP (Bogetic and Fedderke 2009). Remove private sector investment constraints. Removing business environment constraints, labor and product market inefficiencies, and improving corporate governance should be prioritized (World Bank 2019a). In addition, credit constraints can Source: Rozenberg and Fay (2019); World Bank, Human Capital Project. also hold back investment, with many EMDEs A. Investment and maintenance needs based on the Sustainable Development Goals as set out in Rozenberg and Fay (2019) including both new investment and maintenance of existing capital stock. lacking developed capital markets and financial Infrastructure investment includes investment in electricity, transport, water supply and sanitation, flood protection, and irrigation. Preferred is defined as the infrastructure “pathway [that] limits products for much of the population (Sahay et al. stranded assets, has a relatively high per capita consumption due to electric mobility, and invests 2015). Weak access to finance is a key constraint mostly in renewable energy and storage.” B. GDP-weighted expected years schooling and learning-adjusted years of schooling from the World to small and medium firms in SAR–especially for Bank’s Human Capital Project. Leaning-adjusted years of schooling use harmonized cross-country test scores to adjust average years of schooling. women-owned businesses—and holds back firm- Click here to download data and charts. level productivity gains in India (Box 2.5). Efforts are needed to encourage the use of fintech FIGURE 3.11 Developments in Fintech and Govtech products in regions where access to traditional banking products and sources of finance is low, Economies with the largest “unbanked” populations have also seen the biggest increases in fintech innovations to payment systems and other while addressing associated risks of these financial services. The rise of fintech has been largest in SSA. These technologies, such as financial crime and systems are critical to improving access to finance to make productivity- cybersecurity risks (Figure 3.11; IMF and World enhancing investments. EMDE government transparency still lags advanced economies. New ICT can facilitate the rapid dissemination of Bank 2019). Investing an additional 4.5 percent of information within and outside of government to monitor performance and GDP annually in infrastructure in EMDEs would service shortfalls. lift long-run productivity growth by 0.3 percentage point (Figure 3.12). A. Access to banking services and B. Information openness: national mobile money accounts government data availability Raise human capital. Better-educated and healthier workers hold better-paying jobs, have more stable careers, and are more productive. Moreover, a better educated and healthier workforce is more capable of advanced technology adoption (Bils and Klenow 2000). Educational gaps with advanced economies are largest in SAR and SSA, where expected years of schooling is 3 and 5 years lower than in advanced economies, Source: GSM Association (GSMA), Open Knowledge Foundation, World Bank. A. Mobile money accounts based on a sample of 16 EMDEs, excluding China, in East Asia and the respectively. This gap increases to 6 and 7 years Pacific (EAP), 7 EMDEs in Eastern Europe and Central Asia (ECA), 18 EMDES in Latin America and the Caribbean (LAC), 9 EMDEs in Middle East and North Africa (MNA), 7 EMDEs in South Asia when adjusting for quality, suggesting that (SAR), and 40 EMDEs in Sub-Saharan Africa (SSA). Bank accounts, defined as depositors at commercial banks, based on a sample of 22 EMDEs, educational reforms should be a priority in these excluding China, in East Asia and the Pacific (EAP), 24 EMDEs in Eastern Europe and Central Asia regions (Figure 3.10). In addition, tailored (ECA), 32 EMDES in Latin America and the Caribbean (LAC), 19 EMDEs in Middle East and North Africa (MNA), 8 EMDEs in South Asia (SAR), and 48 EMDEs in Sub-Saharan Africa (SSA). interventions at early ages are important. These B. Global Open Data Index is a proxy for the availability of open national government data at large. GDP weighted average. 2016/7 data. It based on a sample of 27 Advanced economies, 14 EMDEs in can include measures to expand school attendance, Eastern Europe and Central Asia (ECA), 6 EMDEs in East Asia and the Pacific (EAP), 25 EMDES in Latin America and the Caribbean (LAC), 2 EMDEs in Middle East and North Africa (MNA), 6 EMDEs provide student grants, support nutrition in South Asia (SAR), and 12 EMDEs in Sub-Saharan Africa (SSA). Click here to download data and charts. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 227 programs for early childhood development, FIGURE 3.12 Productivity growth: reform scenario upgrade teachers’ training, foster teacher A reform package that combines filling investment needs, boosting human accountability and incentivize performance, which capital, and improving the adoption of new technologies could lift can boost educational outcomes. Conditional cash productivity by just over half of a percentage point over 10 years. Replicating the success of China and Vietnam in shifting out of transfer programs can have persistent effects agriculture towards manufacturing and trade services could provide a on educational attainment and the quality of significant boost to productivity growth in low-income economies. employment (Kugler and Rojas 2018). Transitioning to lower fertility rates can reduce A. EMDE productivity reform scenario B. Sectoral reallocation scenario dependency rates and free up resources to invest in education and health—Botswana and Ethiopia have experienced rapid declines in fertility rates in recent decades, alongside large falls in poverty rates (World Bank 2019f). By increasing educational attainment at the same rate as its fastest 10-year cumulative increase ending between 2000-2008, EMDEs could raise long-run productivity growth by about 0.1 percentage point Source: World Bank (Figure 3.12). Note: GDP-weighted average. EMDEs = emerging markets and developing economies. A. The reform scenario assumes: (1) Fill investment needs: the investment share of GDP increases by 4.5 percentage points as in the Rozenberg and Fay (2019) “preferred” infrastructure scenario. The Another key component of human capital is increase is phased in over 10 years (2) Boost human capital: average years of education increases in each EMDE at its fastest cumulative 10-year pace ending during 2000-08; (3) Reinvigorate health. Although life expectancy at birth in technology adoption: economic complexity (Hidalgo & Hausmann 2009) increases at the same pace as its fastest 10-year rate of increase ending during 2000-08. EMDEs has increased to 70 years on average as of B. The sectoral reallocation scenario assumes the sectoral reallocation reform replicates the 2017, this is still about 10 years below average successful transformation of China and Vietnam during 2003-2008. The share of employment in the agriculture sector falls by 15 percent and is reallocated to the manufacturing and trade services advanced-economy levels (81 years). Improve- sectors over a 5 year period. Click here to download data and charts. ments in access to clean water, the provision of adequate sanitation, health care, training, and performance-based payments to health service participation in global value chains to boost providers can yield substantial rewards on the knowledge diffusion on management practices well-being of the population and lift productivity (Bloom et. al. 2013; Cirera and Maloney 2017). (World Bank 2012, 2018b). However, private firms may be reluctant to undertake costly investments in R&D to open Boosting firm productivity foreign markets if competitors can free-ride. Policies that ensure property rights and promote Foster firm capabilities. The structural slowdown public-private partnerships to create technology in TFP growth in EMDEs suggests a need to extension centers in sectoral clusters can increase reinvigorate technology adoption and innovation. firm participation in global value chains, and lift Interventions to ease international and domestic productivity (Cirera and Maloney 2017).20 knowledge diffusion and boost firm absorptive capacities will buttress innovative activities (De Firm-level analysis suggests that to benefit from Visscher, Eberhardt, and Everaert 2018). On-the- technology spillovers EMDEs need to foster trade job training and targeted educational reforms can and financial integration (Box 3.3). Reducing update skills to complement current and newly trade restrictions, alongside increasing levels of introduced technologies, many of which require human capital, increase export diversification and higher cognitive skills and tertiary education levels reduce reliance on commodity exports (Giri, compared to previous technologies. Firm Quayyum, and Yin 2019). Efforts to improve management capabilities have been shown to be trade openness can include regional trade key in generating high-quality R&D and agreements, such as the African Continental Free technology adoption. In India, firms provided with training on management practices saw 20 Technology extension centers generate and transfer new foreign productivity rise by 17 percent—a key factor for and domestic technologies, tailored to a country’s specific needs, to improving management quality has been local users. 228 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 Trade Area which includes economies in MENA Furthermore, middle-income countries tend to be and SSA. In India, reforms in the 1990s to boost highly diversified across a broad range of both foreign (and domestic) competition in the service manufacturing and service sectors, although at sector also had large positive spillovers to high levels of development diversification tends to manufacturing productivity (Arnold et al. 2016). slow and there is a re-specialization (Imbs and Bangladeshi garment exporters increased Wacziarg 2003). productivity after gaining tariff-free access to EU markets in 2001, which also boosted productivity Sectoral diversification is of vital importance to in domestically-focused firms (World Bank economies with a high reliance on commodity 2019d). In China, firms’ participation in foreign extraction, who have usually experienced the supply chains and FDI complemented lowest levels of productivity growth globally domestically-led research and development, (Bahar and Santos 2018).21 Commodity exporting spurring homegrown innovation (Hu, Jefferson, economies in LAC, MENA, and SSA have had and Jinchang 2005). Enhancing technology highly procyclical investment and low average adoption in EMDEs—returning economic TFP growth during the past three decades. The complexity growth to its fastest pace during the benefits of diversification include greater EMDE growth and trade surge during 2000- macroeconomic stability as well as higher average 2008—could increase productivity growth by 0.2 rates of productivity growth. Economies that have percentage point annually (Figure 3.12). successfully reduced their reliance on oil exports, such as Malaysia, Mexico, and Indonesia, initially Address informality. The informal sector is expanded to complementary industries, such as associated with lower average productivity levels natural-resource processing and manufacturing, or and accounts for around 70 percent of expanded to labor-intensive manufacturing, before employment in EMDEs, with particularly high expanding to more complex manufacturing or concentrations in SSA and SAR (World Bank services sectors. In addition, these economies 2019a). In Paraguay, informal firms have been established free trade zones, used tax incentives, found to be not only less productive than formal and established industrial clusters to promote FDI firms, but to have negative spillovers on formal (Cherif and Hasanov 2016). firms’ productivity (Vargas 2015). Reducing the scope for rent-seeking bureaucratic processes that Seek opportunities in services, boost lagging obstruct formalization, improving the fairness of sectors. Many high value-added service sectors regulation, and enhancing the even-handedness of provide opportunities for rapid productivity catch- regulatory and tax enforcement have been up growth (Box 3.2; Hallward-Driemeier and associated with a more efficient reallocation of Nayyar 2017). High-productivity service sectors input factors from less productive informal such as finance, ICT, accounting and legal services activities to more productive formal ones (Amin are likely to become increasingly tradable due to and Islam 2015; Amin, Ohnsorge, and Okou technological advances, but require an enhanced 2019). Beyond formalization, pro-productivity education, including at the tertiary level due to and skill-upgrading interventions could be more their skill-intensive nature. In LICs, notwith- focused on informal small-scale firms and standing rapid pre-crisis productivity gains, unskilled workers (Nguimkeu and Okou 2019). productivity levels in the agricultural sector remain less than 10 percent of the average advanced Encouraging sectoral reallocation economy. SSA hosts the largest number of LICs and may stand to benefit most from reallocation Support sectoral reallocation and diversification. away from agriculture. Yet, LICs in SSA have so Sectoral reallocation is an important engine of productivity growth (Box 3.2). The largest gains 21 EMDE commodity-exporters have historically experienced a in productivity occur at low levels of income as “crowding-out” effect on other faster growth industries during workers shift away from the agricultural sector, periods of high commodity prices, which has hindered them from with lower benefits in middle-income EMDEs. closing the productivity gap with advanced economies. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 229 far shifted away from agriculture towards Creating a growth-friendly environment industrial sectors at a slower pace than LICs in Asia (Box 2.6). Strengthen institutions and government efficiency. Over the long term, institutional Agricultural productivity can be improved through quality is one of the most important determinants targeted measures to increase infrastructure in of productivity growth (Figure 3.7). Productivity these regions, ensure secure land tenures, and gains have been shown to stem from fair promote access to finance. Productivity led growth competition, even-handed contract enforcement, in agriculture could free-up input factors. In simplified and transparent legal processes, and Vietnam, successful reforms included contained political risk (Acemoglu et al. 2019). strengthening of land property rights and relaxed Governments can promote productivity growth by restrictions on external and internal trade of lowering transaction costs, increasing trust in agriculture goods. This could facilitate the institutions and facilitating long-term contracts reallocation of resources from agriculture to more (Leipziger and Thomas 1993). Major governance productive sectors such as manufacturing and reform spurts are associated with faster TFP and services, and boost overall productivity (Fuglie et investment growth (Figure 3.13).22 Other al. 2019). If EMDEs replicated the successful measures to improve the business environment, such as product market and trade reforms or 2003-08 sectoral reallocation of China and cutting red tape, may boost productivity by more Vietnam from the agriculture sector to in the presence of good governance (IMF 2019). manufacturing and trade services, this would lift New information and communications productivity growth by 0.3 percentage points. technologies (“Govtech”) can provide one channel Given sizeable differences in sectoral productivity, through which governments can facilitate the LICs would particularly benefit, with a boost of rapid dissemination of information within and over 1.5 percentage points (Figure 3.12). outside of government to monitor performance Address market failures. Government efforts to and service shortfalls and improve transparency promote specific sectors should first identify (Figure 3.11; World Bank 2018d). market failures that have prevented sectoral Safeguard macroeconomic stability. As reallocation. In addition, the complexity and scale highlighted in Box 3.4, episodes of rapid debt of interventions to foster new industries need to be accumulation and other triggers for financial crises balanced against government and institutional have historically had scarring effects on capacity to manage risks such as political capture productivity. Total EMDE debt has risen by 54 by special interests (Maloney and Nayyar 2018). percentage points since 2010 and currently stands In addition, distortions that prevent the efficient at 168 percent of GDP, exposing many EMDEs to allocation of resources to productive sectors and the risk of financial instability (Chapter 4). Even firms should be removed. Productivity in firms in excluding China, where corporate debt has soared India and China may be 30-60 percent lower due post-crisis, total EMDE debt has risen to a near- to misallocation of capital and labor across sectors record 107 percent of GDP in 2018. Private sector which may be driven by market distortions (Hsieh debt vulnerabilities can be contained with and Klenow 2009). Where firm entry is costly— macroprudential policies and supervisory whether due to high levels of regulation or monitoring of risks. Where sovereign debt regulations that favor state-owned firms— vulnerabilities exist, including those from regulations can be streamlined, access to finance contingent private-sector liabilities, establishing expanded, implicit subsidies reduced, and fiscal rules can increase confidence in the corporate governance standard improved. In sustainability of debt, lengthening the maturity of regions with high energy subsidies (LAC, MNA), lowering these subsidies can also reduce the misallocation of resources into low-productivity 22 These spurts are defined as those that improve at least one of and inefficient energy-intensive sectors. four Worldwide Governance Indicators (government effectiveness, control of corruption, rule of law, and regulatory quality) by at least 2 standard deviations over two years. 230 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 3.13 Effect of governance reform spurts existing debt can ease near-term financing hurdles, Governance reform spurts have been associated with increased potential and improving the quality of spending towards TFP and investment growth. Setbacks, where perceptions of the quality of high-return infrastructure investment can yield governance decline sharply, are associated with slowing investment and growth improvements. TFP growth. A. Average change in potential TFP B. Average change in investment Improve gender equality. Improvements in growth around World Governance growth around World Governance gender equality, in particular by narrowing Indicators reforms Indicators reforms differentials in education and labor force participation, can drive sustained improvements in productivity growth by enhancing the human capital available for production. Women currently comprise only about one-fifth of the labor force in MNA and one-quarter of the labor force in SAR. In SSA, where female employment rates are high, female entrepreneurs tend to have lower profits and access to capital. Gender inequality can be Source: World Bank. addressed by ensuring equal legal rights, targeted Note: TFP growth refers to potential TFP growth, as estimated in World Bank 2018e A.B. Simple averages of potential TFP (A) and investment (B) growth during reform spurts and training programs, relieving capital and financing setbacks (minus simple average potential TFP and investment growth outside such episodes) for all constraints for women, and addressing social countries (“Global”) or for EMDEs only (“EMDE”) using World Governance Indicators. Based on an event study of 305 statistically significant reform events—defined as two-standard-error changes in norms that constrain women’s economic one of four World Governance Indicators—for 136 EMDEs and 36 advanced economies. Data are from 1996-2018. opportunities. Policies to empower women and Click here to download data and charts. boost their productivity include building skills beyond those taught in traditional training programs, such as a greater focus on developing an entrepreneurial mindset—this approach has been found to lift sales and profits in Togo (World Bank 2019f). In the analysis of the underlying drivers of productivity, economies with the lowest gap between female and male educational attainment grew by an average of 0.2 percentage point faster each year than those with the highest differential when controlling for other characteristics of the economy (Figure 3.7). G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 231 ANNEX 3.1 Challenges of EMDE economies (Fajnzylber, Maloney, and Montes-Rojas 2011).1 Productivity Measurement • Intensity of labor input. The number of people There are two primary ways of measuring involved in the production process does not productivity: labor productivity and total factor take into account various work-arrangements productivity (TFP). The former is defined by the that vary the intensity of labor input (Katz total output produced by a unit of labor, the latter and Krueger 2016; Brandolini and Viviano measures the efficiency with which factor inputs 2018). The intensity of labor input is, for are combined. TFP can also be interpreted as the example, better captured by hours worked but technology embedded in the production process, these data are not available for many but may also incorporate wider factors such as countries. organizational characteristics. This annex reviews the different techniques and challenges of these • Quality of labor input. The effectiveness of different productivity measures and explains how labor input may be affected by the level of they are tackled in this study. education, training, and health of workers. These aspects of human capital can be Labor productivity. One of the common addressed by estimating the years of schooling approaches is measuring labor productivity as for education and the number of expected output per worker by taking the number of years of life for health. However, the quality employees as the unit of labor input. Its advantage of formal education and health, and the is in its wide availability across countries. Its amount of on-the-job training is difficult to disadvantage rests in the failure to account for the measure consistently in a panel setting. quality and intensity of labor input. Total factor productivity. One of the most • Comprehensiveness. Having high ratios of commonly used measures of technological informality in EMDEs makes it challenging to enhancement is total factor productivity growth. appropriately measure productivity. While The standard growth accounting approach is one both output and employment might be of the most common methodologies in the mismeasured due to non-registration, many literature to estimate TFP. It is appealing due to national statistics offices estimate the size of its simple nature and its ease of interpretation. the informal sector and adjust their GDP Being estimated as residual, it depends on the estimates accordingly (SNA 1993, 2008; assumed functional form and any measurement UNECE 2008; Charmes 2012). The error for factor inputs. In the context of the difficulty in estimating the scale of informal United States, this has triggered a debate about output and lack of consistency in approach the extent to which TFP growth adequately allows scope for productivity misme- reflects new technologies. asurement. Labor input is intended to capture all of those involved in the production • Functional form. TFP is defined as “a shift in process. Thus, total employment figures the production function”, in contrast to include self-employment, which accounts for biased technological change. Its calculation a large proportion of informal employment in assumes the existence of a well-behaved and EMDEs (World Bank 2019a). However, stable production function which also some self-employment does not involve the accurately describes the technology in use informal sector, while the scale of additional (Baqaee and Farhi 2018). One of the employment in the informal sector is also subject to uncertainty—therefore, difficulties in both the measurement of informal output 1 The direction of the bias depends on how national statistics and employment contribute to uncertainty offices adjust their employment and official GDP to cover the around the productivity level, particularly in informal sector, which may vary across countries (UNECE 2008). 232 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 commonly used functional forms is Cobb- measured productivity (Byrne, Fernald, and Douglas with constant returns to scale and Reinsdorf 2016). Others find evidence of sizable unitary elasticities of substitution between mismeasurement and attribute part of the United capital and labor. If the assumption of States productivity slowdown to measurement constant returns to scale is not valid, TFP biases, particularly due to the increasing share of estimations may be biased by market power in the services sector in output (Brynjolfsson and final goods (Dribe et al. 2017). McAfee 2014; Feldstein 2017). Overall, while there is some evidence for mismeasurement, it is • Capital measurement. Physical capital is unlikely that a significant part of the slowdown difficult to value accurately. Its value depends can be explained by it alone (Cerra and Saxena on the longevity of assets (short-lived assets 2017; Syverson 2016). such as computers versus long-lived assets such as roads) and the nature of capital (intangible ANNEX 3.2 Data and Growth capital such as research and development or marketing expenditures). A common way of Accounting Approach measuring the capital stock is to apply the perpetual inventory methodology to the flow Data. The data on capital services and human of expenditure on assets and their depreciation capital are taken from the Penn World Table 9.1, rates. Since data for the initial capital stock is while data on other macroeconomic aggregates usually not available, assumptions are made such as GDP are primarily drawn from the World on capital to output ratio of the initial year Bank’s World Development Indicators (WDI) but this ratio can be highly country-specific database, complemented by the ILO and (Feenstra, Inklaar, and Timmer 2015). Conference Board estimates of employment. This results in annual labor productivity, TFP and • Factor utilization. Since TFP is measured as a capital services data for 103 economies, of which residual, it estimates not only technological 73 are EMDEs (including 11 low-income change but also any mismeasurement of economies) and 29 are advanced economies, for capital and labor input (Basu, Fernald, and 1981-2018. All aggregates are GDP-weighted Kimball 2006). The capital stock measures averages at 2010 prices and exchange rates. These the total physical capital available for economies account for 96 percent of global GDP. production without necessarily considering how much of the existing capital is actually Growth accounting. Following Caselli (2005), used in the production process. Similarly, productivity is decomposed into contributions labor input, even if it is finely measured as from several factor inputs: total working hours, does not include labor Labor productivity Yt/Lt At Kt /Lt 1-α Htα effort. This may lead to an overly cyclical measure of productivity. Following Solow (1957), a Cobb-Douglas production function with constant returns to scale New technologies and output measurement. is assumed. By taking log differences, labor There have been concerns that quality productivity growth can be decomposed into the improvements in information technology have not following factor inputs. been accurately captured because price deflators for information and communications technology ΔLPt 1 - α Δkt αΔht Δat understate the true price declines in these assets (Hatzius et al. 2016). Mismeasurement of new IT Where kt K log L t and ht log Ht , and at is technologies could, therefore, explain some of the t the log of TFP, calculated here as a residual of slowdown in measured productivity growth. Some labor productivity growth after subtracting the studies find evidence of mismeasurement in both change in capital deepening and human capital the pre and post-crisis period, such that indices, weighted by their respective shares in the mismeasurement explains little of the slowdown in production function ( 1 - α and α ). G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 233 Capital services (Kt). Data on capital services are • supporting environments, such as institutions, from the Penn World Table 9.1 (PWT) (Feenstra, policies and social conditions; and Inklaar, and Timmer 2015). In contrast to previous versions of PWT, this edition utilizes • market development, such as trade integration capital services as a measure of capital inputs and financial deepening. instead of capital stocks (Inklaar, Woltjer, and Gallardo 2019). This annex reviews the theoretical and empirical literature that establishes linkages between each of Human capital (Ht). The human capital index the most commonly identified drivers and from the Penn World Table 9.1 is used productivity growth and assesses differences across throughout the sample. This measure uses average EMDE regions as well as over time. years of schooling of the working-age population in combination with an estimate of the global Inputs of production returns to education. Innovation. Technical innovations create better Labor share estimates. The output-labor elasticity ways to produce goods, deliver services, and (α), proxied by the labor income share, is also improve within-sector productivity of firms. derived from the PWT 9.1 database. It is Despite large productivity gaps in EMDEs relative estimated using the labor compensation to output to advanced economies, most EMDEs invest ratio, including adjustments to take account of much less in formal research and development mixed-income and wages from self-employment. (R&D) than advanced economies (Goñi and Labor shares are allowed to vary across countries Maloney 2017). The number of patents per in this chapter’s decompositions. This analysis capita—one indicator of the pace of innovation— uses constant labor shares over time, defined as the is particularly low in Latin America and the long-term average of labor share data from PWT Caribbean (LAC), South Asia (SAR), and Sub- 9.1, although it varies across countries. Saharan Africa (SSA; Annex Figure 3.3.1). Nonetheless, gradual improvements in process or product quality have been reported across all income levels (Goñi and Maloney 2017). New patents tend to be more productivity-enhancing in ANNEX 3.3 Drivers of countries with ample supply of highly educated productivity1 and skilled labor force, while gradual improvements in productivity can be achieved Productivity improvements are key for spurring even with low human capital levels (World Bank sustained economic growth and social progress in 2018e). the presence of limited quantity and quality of factor inputs—labor inputs, physical capital, and Physical capital. Labor productivity can be natural resources (Easterly and Levine 2001; boosted by capital accumulation, underpinned by Caselli 2005). Drawing from growth theories, the investment and matched with adequate absorptive empirical literature has identified many potential capacity (Eberhardt and Presbitero 2015). In drivers of productivity growth.2 These can be particular, investments in infrastructure, including classified into three broad categories: inputs of transport, water and sanitation, power, and production, such as innovation, physical capital telecommunications can complement and labor; technological progress and lift productivity.3 Infrastructure needs in EMDEs remain large. Achieving infrastructure-related SDGs in low- and middle-income countries will require an average 1 This annex was prepared by Alistair Dieppe, Atsushi Kawamoto, Yoki Okawa, and Cedric Okou. 2 As some concepts overlap there could be alternative classifications 3 See, for example, Aschauer (1989); Servén (2015); and Martins which focus on other concepts such as competition, geography, and social fragmentation. (2019). 234 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 ANNEX FIGURE 3.3.1 Productivity drivers in 2017, by • Education. As a labor force becomes better region trained and more highly skilled, it has a All EMDE regions fall short of advanced-economy conditions in important greater propensity to contribute to productivity drivers, including innovation, human capital, institutions, technological advancements and to benefit macroeconomic stability, and trade openness. There is considerable from positive innovation. Countries with variation across regions: SSA and LAC tend to rank low in many of these dimensions whereas EAP and ECA tend to rank highly. better-educated working-age populations tend to have higher productivity (Barro and Lee A. Patents per capita B. Years of schooling 2015). This could reflect workforces in EMDEs moving jobs from sectors requiring limited skills, such as agriculture, to sectors requiring greater skill levels, such as manufacturing and services (Box 3.2). Despite significant catch-up over the past five decades, the gap in average years of schooling between EMDEs (8.6 years) and advanced economies (12.3 years) remains sizeable. There is a substantial dispersion among EMDE regions. C. Rule of Law D. Inflation For instance, Europe and Central Asia (ECA) has the highest years of schooling among EMDEs, just one year short of the advanced- economy average. By contrast, SSA and SAR has low years of schooling, less than half of the advanced-economy average (Annex Figure 3.3.1). • Health. Healthy workers can work more efficiently and learn faster; they are also more E. Trade openness F. Economic complexity committed to improving their skills and are better equipped to innovate (World Bank 2018e). Better health complements education in reinforcing the supply of good-quality labor, in turn raising human capital, attracting investment, and improving productivity. • Demographic trends. Workforce aging is often negatively associated with productivity growth (Aiyar, Ebeke and Shao 2016; Aksoy et al. Source: Observatory of Economic Complexity; United Nations; World Bank, Worldwide Governance 2019).4 New technologies can disrupt the Indicators. Note: Data for 2017. Unweighted averages. Trade openness is the sum of exports and imports in value of existing human capital, as senior and percent of GDP. Samples include 22-33 countries in advanced economies, 6-23 countries in EAP, 4-24 countries in ECA, 17-31 countries in LAC, 7-16 countries in MNA, 3-8 countries in SAR, and unskilled workers may need retraining. The 10-48 countries in SSA, strength of this mechanism may depend on Click here to download data and charts. the economic structure of the country, as productivity benefits more from experience in yearly investment of 4 to 8 percent of GDP some occupations and from innovation in during 2015-30 (Rozenberg and Fay 2019; others. This effect is particularly pronounced Vorisek and Yu, forthcoming). in advanced economies, where the working- age share of the population shrank Labor. The productivity of labor can be improved in several ways. A better-educated or healthier 4 The Solow model suggests a decline of the working age work-force can adjust more easily to productivity- population could increase the capital per worker and positively affect enhancing changes. labor productivity. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 235 by 3 percentage points between 2008 to 2018. economic growth (Gramacy, Malone, and Horst In the decades ahead, EMDEs are projected to 2014). Price stability in EMDEs, proxied by follow the same path. Between 2018 and inflation, has substantially improved over time, 2030, the working-age population share is and currently stands at about 4 percent (except in expected to decline by 3 percentage points SSA), down from 18 percent in 1990 (Annex for advanced economies and 2.5 percentage Figure 3.3.1). Nevertheless, in many EMDEs, points for EMDEs. In East Asia and Pacific monetary and fiscal policy frameworks still lag (EAP) and ECA, the working-age population behind best practices (Koh and Yu 2019). share has already begun to decline, whereas SSA continues to benefit from rising working- Income equality. Income inequality has been age population shares. Realizing the potential explored as a potential underlying driver of low of a youthful population requires investing in productivity growth. However, the literature is education and accelerating job creation. agnostic about the impact of inequality on productivity and economic growth (Herzer and Supporting environment Vollmer 2012; Alvaredo et al. 2018). The elusive empirical link may be due to the u-shaped Institutions. Institutions are the entities that relationship between income equality and the shape human interactions within a society (North stage of development: the adverse effects of 1990). Institutions come in many forms—rule of income inequality tends to be high for low-income law, barriers to firm creation and operation, and and high-income countries, but not high in system of government, to name a few. Better middle-income countries (Banerjee and Duflo quality institutions are associated with fairer 2003). Income inequality has fallen in some competition and higher productivity (Easterly and EMDE regions, such as LAC. Yet, it remains Levine 2003; Levchenko 2007). Increased much higher in EMDEs than in advanced competition is found to support innovation and economies. As of 2017, inequality measured by raise productivity through improvements in the Gini index, was 41 for EMDEs, compared to management and product quality (Van Reenen 33 for advanced economies.5 2011). Acemoglu et al. (2019) find that the transition to democracy raises productivity by 20 Gender equality. Large gaps between women and percent in the subsequent 25 years, but the results men in measures of education, health, and access vary across studies and some have not uncovered to economic opportunities can lower productivity. an effect (Ruiz Pozuelo, Slipowitz, and Vuletin Better income-earning opportunities for women 2016). Productivity improvements depend on a can increase human and physical capital country’s distance to the technology frontier investment through higher household income and (Prati, Onorato, and Papageorgiou 2013). There higher returns for building women’s human remains a large gap between the quality of capital (Klasen and Santos Silva 2018). It may also institutions, proxied by the government lower fertility and, hence, help provide each child effectiveness index, between all EMDE regions with better education and health care. An and advanced economies, and the gap has increasing share of women in the labor force, with remained almost unchanged over the past twenty fair pay and equal job opportunity, can also be years (Annex Figure 3.3.1). beneficial for productivity growth, as it brings a richer collection of perspectives to the decision- Price stability. Price stability in part reflects the making and production process (Gallen 2018). By absence of major distortions and uncertainty in contrast, the exclusion of all women from the macroeconomic environment (Rodrik, managerial positions can reduce income per capita Subramanian, and Trebbi 2004b). Price by 12 percent (Cuberes and Teignier 2012, 2014). instability, which can be reflected by high The gap between EMDEs and advanced inflation or a large difference between the black market and official exchange rates, may hinder 5 The Gini index is a measure of the distribution of income across investment, lead to sizeable capital outflows, and income percentiles, presented on a scale of 0 to 100, where 100 is the are negatively correlate with productivity and most unequal. 236 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 economies for the latter of these indicators has measured as a composite indicator that compares declined during the last five decades. each country’s sectoral export shares with the sector’s share in world trade. The economic Market development complexity is higher if the country exports more “complex” goods such as X-ray appliances, which Trade. Trade can significantly improve can be exported from only a few other economies productivity growth (World Bank 2019d) (Hausmann et al. 2014). Greater export although some studies find only a weak complexity has been associated with higher labor relationship between trade and productivity productivity through its association with the (Rodrik, Subramanian, and Trebbi 2004). diversification and sophistication of a country’s Imports of machinery or high-technology goods economic structure (Hausmann and Hidalgo can directly improve productivity at the firm, 2010). EMDEs largely lag behind advanced sector, and country level. Lower tariffs can economies in terms of economic complexity increase imports, facilitate knowledge transfers, (Annex Figure A3.3.1) and strengthen firm-level productivity (Kraay, Soloaga, and Tybout 2002). Exporting firms tend Urbanization. Urbanization can facilitate to have higher productivity than non-exporting agglomeration benefits such as knowledge ones. The high productivity of exporting firms can spillovers, and improved skills matching within be explained by self-selection in some cases the labor force. Densely populated areas bring (Clerides, Lach, and Tybout 1998). However, people and firms closer together, making it easier evidence from Kenya and the Republic of Korea to share ideas, exchange information, invent new suggests that exports can increase productivity technologies, design new projects, engage in new after controlling for self-selection (Graner and partnerships, and start new businesses (Abel, Dey, Isaksson 2009). Learning-by-exporting effects on and Gabe 2012). These agglomeration benefits productivity depend on the income level of can in turn lift productivity. importers or exporters. The learning effect is large when the exporter and importer have similar Finance. Well-developed financial markets can productivity levels or importer’s human capital is improve the efficiency of capital allocation, high (Graner and Isaksson 2009; Keller 2004; facilitate technology spillovers and help firms take Blalock and Gertler 2004; Aw, Chung, and advantage of productivity-enhancing investments Roberts 1998). ECA and EAP are the EMDE (Fisman and Love 2003; Levine 1997). Financial regions that are most open to trade whereas SAR is development and integration are associated with the least open (Annex Figure 3.3.1). productivity growth (Aghion, Howitt, and Mayer- Foulkes 2005). Financial markets allow firms to Foreign direct investment. Investment from diversify investment risk, increase liquidity, and abroad can bring advanced technology, improved stimulate entrepreneurship and productivity. organizational structure, and good management practices from frontier technology economies, Estimating impacts of drivers on productivity boosting productivity in host economies where it growth is lagging (Griffith, Redding, and Simpson 2003). Methodology. A cross-section analysis is Cross-border capital flows have a positive effect on undertaken where the dependent variable is the productivity, especially those with a high level of long-run growth of productivity during 1960- development and high-quality institutions. 2018 and separately over 1995-2018. In addition However, this positive relationship is weaker for to the initial level of log productivity (y0), other EMDEs (Keller and Yeaple 2009). In developing regressors (X0)—discussed in the literature and countries, the cost of subsidies offered to firms to measured at the beginning of the period—are attract foreign investments can exceed the positive included: effect of FDI on productivity (Haskel, Pereira, and Slaughter 2007). yT,j - yo,j βyo,j Xo,jγ εj, Economic complexity. Economic complexity is where εj is a disturbance term, and j denotes a G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 237 specific economy.6 The set of coefficients γ ANNEX FIGURE 3.3.2 Productivity changes in capture how each covariate (X0,j) drive productivity drivers, by region productivity dynamics over the long-run. The Productivity drivers—here captured in a composite index—have improved wide range of potential drivers associated with considerably in EMDEs since the 1980s. However, in several regions, productivity growth leads to a large range of including EAP, ECA, LAC and SAR, the pace of improvement appears to have stalled since the global financial crisis. potential model specifications (Fernández, Ley, and Steel 2001; Durlauf, Kourtellos, and Tan A. East Asia and Pacific B. Europe and Central Asia 2008, Durlauf, Johnson, and Temple 2005). In order to reduce the bias stemming from an ad-hoc selection and omission of variables, a Bayesian Model Averaging (BMA) approach is applied, which considers different subsets of potential variables and evaluates their inclusion probabilistically. Nonetheless, the estimation results can be unstable in the presence of strong collinearity, as many variables can essentially represent the same concepts (Ghosh and Ghattas C. Latin America and the Caribbean D. Middle East and North Africa 2015). Therefore, based on existing literature and growth theories, variables that represent common concepts are grouped together. The posterior distributions of the coefficients obtained from the BMA procedure are then aggregated to the group level.7 Impacts. The estimation is undertaken for 59 countries, including 36 EMDEs.8 It shows that F. Sub-Saharan Africa better educated workforce, stronger institutions, E. South Asia greater innovation, stronger investment, higher levels of urbanization, price stability and a diverse and sophisticated economic structure are all significantly associated with higher productivity growth (Figure 3.7). Furthermore, the estimated impact depends on the stage of development and has changed over the more recent period. The estimated coefficients can be interpreted as the hypothetical coefficient of each theoretical driver Source: World Bank Note: For each country, index is a weighted average—weighted by the normalized coefficients shown of productivity growth. Using these coefficients an in Figure 3.7—of the normalized value of each driver of productivity. Drivers include the ICRG rule of law index, patents per capita, share of non-tropical area, investment in percent of GDP, ratio of female average years of education to male average years, share of population in urban areas, Economic Complexity Index, years of schooling, and share of working-age population. Regional and EMDE indices are GDP-weighted averages. Samples include 7 economies in EAP, 8 economies in ECA, 18 economies in LAC, 6 economies in MNA, 4 economies in SAR, and 11 economies in SSA. 6 Most candidate variables can be viewed as outcomes of Click here to download data and charts. productivity, in addition to drivers of productivity, which constrains the interpretation of causal claims from the regressions. To counter the reverse causality issue, the variables used in the analysis are levels aggregate index of drivers of productivity growth in 1960 (or 1995), based on the assumption that serial correlation in is formed. It shows it grew rapidly on average in 1960 and average growth for the next 58 years is small. 7 Parametric estimations cannot exclude the possibility of omitted EMDEs in the pre-crisis period supporting variable bias. Panel estimation focusing on more recent periods can productivity growth (Annex Figure 3.3.2). reduce this issue by the inclusion of country fixed effects and a wider However, since the global financial crisis, range of potential variables, but usually rely on the constant country effects assumption and can suffer from serial correlation and other improvements in the drivers have begun to level types of biases. off as the pace of improvement has slowed, 8 Variables related to theories in the existing literature are chosen particularly in several EMDE regions (EAP, ECA, where data exists before 1970 for a large sample of economies. Observations which are not available in the particular year was LAC, and SAR) amid a productivity growth substituted by the observations in the closest year available. slowdown. 238 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 ANNEX TABLE 3.3.1 Variables included in the regressions and sources Group Variable Source Financial development Ratio of domestic credit to GDP World Development Indicators (WDI) Investment Ratio of gross fixed capital formation to GDP WDI Education Years of schooling Barro & Lee, UN Human capital UNDP Years of tertiary schooling Barro & Lee, UN Years of primary and secondary schooling Barro & Lee, UN Economic Complexity Economic Complexity Index plus Economic observatory (Exports + Imports)/GDP WDI Innovation Patents per capita WDI Patents per capita * years of tertiary schooling WDI Equality 100 - Gini coefficient UNU wider database Institutions Political Rights Index Freedom House Civil Rights Index Freedom House Rule of Law Index International Country Risk Guide, PRS Ratio of government consumption to GDP WDI and various other sources Urban Share of population in urban areas WDI Population density WDI Health Survival rate after 5 years per 1000 births = 1000-Infant mortality rate WDI Life expectancy at birth WDI Demography Share of population aged 15-64 WDI Share of population aged below 15 WDI Gender Ratio of years of schooling of female to male Barro & Lee, UN Ratio of years of primary schooling of female to male Barro & Lee, UN Ratio of labor participation rate of female to male WDI Geography Dummy for landlocked countries WDI Share of land which is in tropical regions WDI EMDE energy exporter dummy World Bank Stability (-1) * CPI Inflation Rate WDI Black market exchange rate relative to the official rate WDI Note: Sources and list of variables included in the Bayesian selection model. Variables selected with the highest probability of inclusion for each category are in bold. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 239 ANNEX 3.4 Data and Shift-share analysis. Following (Wong 2006) and (Padilla-Pérez and Villarreal 2017), this chapter methodology for sectoral employs a shift-share-analysis which decomposes productivity aggregate labor productivity into the growth within a sector and shifts between sectors: Data. The database consists of sectoral and aggregate labor productivity statistics for 80 countries, and nine sectors covering the period up to 2015. Compared with the literature using nine- sector data, it employs a large and diverse sample Intra-Sectoral Static Sectoral Dynamic Sectoral of countries. The database combine data from the Effect Effect Effect Shift OECD STAN database, World KLEMS (EU, LAC and Russia), the Groningen Growth Structural Change Effect Development Center (GGDC) database (de Vries, de Vries and Timmer 2015), and the Expanded where y is aggregate labor productivity, yj is labor Africa Sector Database (EASD, Mensah and productivity of sector j, Yj is initial value added of Szirmai 2018) for value added data and sector j, sj is employment share of sector j. employment. The APO Productivity Database, Structural changes are driven by the change in UN data, ILOSTAT and National sources are employment share. They are further decomposed used for supplementary purposes. Following into those which are due to the reallocation of McMillan, Rodrik, and Verduzco-Gallo (2014), sources to sectors which higher productivity levels local currency value added is converted to U.S. (static sectoral effect), and those due to dollars using 2011 PPP exchange rate obtained reallocation toward sectors with higher from Penn World Table for the international productivity growth (dynamic sectoral effect). comparison of productivity levels.1 1 Van Biesebroeck (2009) builds an expenditure-based sector- specific PPP in OECD countries, using detailed price data. ANNEX TABLE 3.4.1 Sectoral classifications Sector name Description 1. Agriculture Agriculture, forestry and fishing 2. Mining Mining and quarrying 3. Manufacturing Manufacturing 4. Utilities Electricity, gas, steam and air conditioning supply 5. Construction Construction 6. Trade services Wholesale and retail trade; repair of motor vehicles and motorcycles; Accommodation and food service activities 7. Transport services Transportation and storage; Information and communication Financial and insurance activities; Real estate activities; Professional, scientific and technical activities; 8. Financial and Business services Administrative and support service activities Public administration and defence; compulsory social security; Education; Human helath and social work activities; 9. Government and Personal Arts, entertainment and recreation; Other service activities; Activities of households as employers; undifferentiated services goods-and services-producing activities of households for own use; Activities of extraterritorial organizations and bodies. 240 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 ANNEX 3.5 Methodology for ANNEX 3.6 Local projection Box 3.3 methodology for Box 3.4 Measurement challenges. Revenue-based TFP The computation of crises impacts follows the (TFPR) measures conflate physical productivity local projection (LP) method (Jordà 2005). The and price effects (Foster et al. 2008; Andrews, dependent variable is the cumulative change in Criscuolo, and Gal 2016). These price effects can output or productivity levels between horizons t-1 substantially distort TFPR estimates in non- and t 1, measured as the natural logarithms competitive markets or when output prices and (yt,j ). The baseline model is given by inputs choice are correlated. For instance, a high- productivity firm with market power can lower output prices to increase its market share. In this case, TPFR estimates can be low even though the firm is highly productive. Producer prices, if available, can be used to deflate firm-level sales and obtain physical TFP (TFPQ) estimates Where h = 0,1,2,…, 8 is the horizon, α h ,j and (Cusolito and Maloney 2018; Van Beveren 2012). τ h ,t are country j and time fixed effects, and u h Moreover, specifying a single production function t,j is an error term. The coefficient of interest β h for a firm using multiple production technologies captures the dynamic multiplier effect (impulse is restrictive and can bias TFP estimates (Bernard, response) of the dependent variable with respect to Redding, and Schott 2010; Goldberg et al. 2010). the event dummy variable Et,j. The number of lags Disaggregated product-level data, if available, can for each variable is denoted by p and set to 1 for be used to construct product-level TFP and help the estimation. The specification controls for (i) account for the richness in production mix. country and time specific trends, (ii) lagged event dates, (iii) future values of the event dummy Methodology. e tted speci cation is between time t and t h-1 to correct for possible forward bias (Teulings and Zubanov 2014), and g DTFi = 40+ ∑ g ρg I ( g ∈G \{ref }) +∑ γ X + v j j ij i (iv) past changes Δyt-s,j. Additional controls for country-specific interactions and non-linear effects g may also be included. where DTFi is the distance-to-frontier of TFP for firm i in industry g, 40 stands for the constant term, ref = TINT is the reference industry, and coefficients ρg are interpreted relatively to the reference group. Xij is firm i’s jth characteristic such as GDP per capita (in 2009 U.S. dollars per worker), size (number of employees), exports (as a proportion of total sales), and business climate (control of corruption, business freedom). The error term is denoted by vi. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 241 Amin, M., F. Ohnsorge, and C. Okou. 2019. “Casting References a Shadow Productivity of Formal Firms and Informali- Abel, J. R., I. Dey, and T. M. Gabe. 2012. ty.” Policy Research Working Paper 8945. World “Productivity and the Density of Human Capital.” Bank, Washington, DC. Journal of Regional Science 52 (4): Andrews, D., C. Criscuolo, and P. Gal. 2016. “The 562–86. Best versus the Rest: The Global Productivity Slow- Acemoglu, D., S. Naidu, P. Restrepo, and J. A. Robin- down, Divergence across Firms and the Role of Public son. 2019. “Democracy Does Cause Growth.” Journal Policy.” OECD Productivity Working Paper 05, Or- of Political Economy 127 (1): ganization for Economic Co-operation and Develop- 47–100. ment, Paris, France. ADB (Asian Development Bank). 2017. Asian Develop- Araujo, J. T., E. Vostroknutova, and K. Wacker. 2017. ment Outlook: Transcending the Middle-Income Chal- “Productivity Growth in Latin America and the Carib- lenge. Manila: Asian Development Bank. bean: Exploring the Macro-Micro Linkages.” Discus- sion Paper 19, World Bank, Washington, DC. Adler, G., R. Duval, D. Furceri, S. Kilic Celik, K. Ko- loskova, and M. Poplawski-Ribeiro. 2017. “Gone with Arellano, C. 2008. “Default Risk and Income Fluctua- the Headwinds: Global Productivity.” IMF Staff Dis- tions in Emerging Economies.” American Economic cussion Note 17/04, International Monetary Fund, Review 98 (3): 690–712. Washington, DC. Arnold, J. M., B. Javorcik, M. Lipscomb, and A. Mat- Ağca, Ş., and O. Celasun. 2012. “Sovereign Debt and too. 2016. “Services Reform and Manufacturing Per- Corporate Borrowing Costs in Emerging Markets.” formance: Evidence from India.” Economic Journal 126 Journal of International Economics 88 (1): 198–208. (590): 1–39. Aghion, P., P. Bacchetta, R. Rancière, and K. Rogoff. Arteta, C., and G. Hale. 2008. “Sovereign Debt Crises 2009. “Exchange Rate Volatility and Productivity and Credit to the Private Sector.” Journal of Interna- Growth: The Role of Financial Development.” Journal tional Economics 74 (1): 53–69. of Monetary Economics 56 (4): 494–513. Arteta, C., and S. Kasyanenko. 2019. “Financial Mar- Aghion, P., P. Howitt, and D. Mayer-Foulkes. 2005. ket Developments.” In A Decade after the Global Reces- “The Effect of Financial Development on Conver- sion: Lessons and Challenges for Emerging and Developing gence: Theory and Evidence.” Quarterly Journal of Eco- Economies, edited by M. A. Kose and F. Ohnsorge. nomics 120 (1): 173–222. Washington, DC: World Bank. Aguiar, M., and G. Gopinath. 2006. “The Elephant Aschauer, D. A. 1989. “Is Public Expenditure Produc- Curve of Global Inequality and Growth.” Journal of tive?” Journal of Monetary Economics 23 (2): 177–200. International Economics 69: 64–83. Aslam, A., S. Beidas-Strom, R. Bems, O. Celasun, S. Aiyar, S., C. Ebeke, and X. Shao. 2016. “The Impact of Kiliç Çelik , and Z. Koczan. 2016. “Trading on Their Workforce Aging on European Productivity.” IMF Terms? Commodity Exporters in the Aftermath of the Working Paper 16/238. International Monetary Fund, Commodity Boom.” IMF Working Paper 16/27. Inter- Washington, DC. national Monetary Fund, Washington, DC. Aksoy, Y., H. S. Basso, R. P. Smith, and T. Grasl. Atkin, D., A. K. Khandelwal, and A. Osman. 2017. 2019. “Demographic Structure and Macro-economic “Exporting and Firm Performance: Evidence from a Trends.” American Economic Journal: Macroeconomics Randomized Experiment.” The Quarterly Journal of 11 (1): 193–222. Economics 132 (2): 551–615. Alvaredo, F., L. Chancel, T. Piketty, E. Saez, and G. Zucman. 2018. “The Elephant Curve of Global Ine- Bahar, D. 2018. “The Middle Productivity Trap: Dy- quality and Growth.” AEA Papers and Proceedings 108 namics of Productivity Dispersion.” Economics Letters (May):103–08. 167: 60–66. Amin, M., and A. Islam. 2015. “Are Large Informal Bahar, D., and M. A. Santos. 2018. “One More Re- Firms More Productive than the Small Informal Firms? source Curse: Dutch Disease and Export Concentra- Evidence from Firm-Level Surveys in Africa.” World tion.” Journal of Development Economics 132 (1): 102– Development 74 (October): 374–85. 14. 242 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 Ball, L. 2009. “Hysteresis in Unemployment: Old and Blanchard, O., and J. Wolfers. 2000. “The Role of New Evidence.” NBER Working Paper 14818, Na- Shocks and Institutions in the Rise of European Unem- tional Bureau of Economic Research, Cambridge, MA. ployment: The Aggregate Evidence.” Economic Journal 110: C1–33. Banerjee, A. V., and E. Duflo. 2003. “Inequality and Growth: What Can the Data Say?” Journal of Economic Bloom, N., B. Eifert, A. Mahajan, D. McKenzie, and J. Growth 8 (3): 267–99. Roberts. 2013. “Does Management Matter? Evidence from India.” Quarterly Journal of Economics 128 (1): 1– Baqaee, D., and E. Farhi. 2018. “The Microeconomic 51. Foundations of Aggregate Production Functions.” NBER Working Paper 2593, National Bureau of Eco- Bogetic, Z., and J. W. Fedderke. 2009. “Infrastructure nomic Research, Cambridge, MA. and Growth in South Africa: Direct and Indirect Productivity Impacts of 19 Infrastructure Measures.” Barrett, C. B., L. Christiaensen, M. Sheahan, and A. World Development 37 (9): 1522–39. Shimeles. 2017. “On the Structural Transformation of Rural Africa.” Journal of African Economies 26 (1): 11– Bosch, M., E. Goni, and W. Maloney. 2007. “The 35. Determinants of Rising Informality in Brazil: Evidence from Gross Worker Flows.” Policy Research Working Barro, R. J. 1991. “Economic Growth in a Cross Sec- Paper 4375, World Bank, Washington, DC. tion of Countries.” The Quarterly Journal of Economics 106 (2): 407–43. Bosworth, B., and S. M. Collins. 2008. “Accounting for Growth: Comparing China and India.” Journal of ———. 2000. “Inequality and Growth in a Panel of Economic Perspectives 22 (1): 45–66. Countries.” Journal of Economic Growth 5 (1): 5–32. Brynjolfsson, E., and L. Hitt. 1995. “Information Technology as a Factor of Production: The Role of ———. 2001. “Economic Growth in East Asia Before Differences among Firms.” Economics of Innovation and and After the Financial Crisis.” NBER Working Paper New Technology 3 (3–4): 183–99. 8330, National Bureau of Economic Research, Cam- bridge, MA. Brynjolfsson, E., and A. McAfee. 2014. The Second Machine Age: Work, Progress, and Prosperity in a Time of Barro, R. J., and J.-W. Lee. 2015. Education Matters: Brilliant Technologies. New York: W W Norton & Co. Global Schooling Gains from the 19th to the 21st Centu- ry. New York: Oxford University Press. Bulow, B. J., and K. Rogoff. 1989. “Sovereign Debt: Is to Forgive to Forget?” American Economic Review 79 Barro, R. J., and X. Sala-i-Martin. 1992. (1): 43–50. “Convergence.” Journal of Political Economy 100 (2): 223–51. Bustos, P., B. Caprettini, J. Ponticelli, D. Atkin, F. Buera, V. Carvalho, G. Gancia, et al. 2016. Bartelsman, E. J., and M. Doms. 2000. “Agricultural Productivity and Structural Transfor- “Understanding Productivity: Lessons from Longitudi- mation: Evidence from Brazil.” American Economic nal Microdata.” Journal of Economic Literature 38 (3): Review 106 (6): 1320–65. 569–94. Byrne, D. M., J. G. Fernald, and M. B. Reinsdorf. Basu, S., J. G. Fernald, and M. S. Kimball. 2006. “Are 2016. “Does the United States Have a Productivity Technology Improvements Contractionary?” American Slowdown or a Measurement Problem.” Finance and Economic Review 96 (5): 1418–48. Economics Discussion Series 2016 (17): 1–74. Baumol, W. J. 1967. “Macroeconomics of Un- Calderón, C., E. Moral-Benito, and L. Servén. 2015. balanced Growth: The Anatomy of Urban Crisis.” “Is Infrastructure Capital Productive? A Dynamic Het- American Economic Review 57 (3): 415–26. erogeneous Approach.” Journal of Applied Econometrics Bils, M. J., and P. Klenow. 2000. “Does Schooling 30 (2): 177–98. Cause Growth?” American Economic Review 90 (5): Caselli, F. 2005. “Accounting for Cross-Country In- 1160–83. come Differences.” In Handbook of Economic Growth, Blanchard, O., E. Cerutti, and L. Summers. 2015. edited by Philippe Aghion and Steven N. Durlauf, “Inflation and Activity - Two Explorations and Their 1:679–741. Amsterdam: Elsevier. Monetary Policy Implications.” NBER Working Paper Cerra, V., and S. C. Saxena. 2008. “Growth Dynamics: 21726, National Bureau of Economic Research, Cam- The Myth of Economic Recovery.” American Economic bridge, MA. Review 98 (1): 439–57. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 243 Cerra, V., and S. C. Saxena. 2017. “Booms, Crises, and Cusolito, A. P., D. C. Francis, N. Karalashvili, and J. Recoveries: A New Paradigm of the Business Cycle and R. Meza. 2018. “Firm Level Productivity Estimates.” Its Policy Implications.” IMF Working Paper 17/250, Methodological Note. World Bank, Washington, DC. International Monetary Fund, Washington, DC. Cusolito, A. P., and W. F. Maloney. 2018. Productivity Charmes, J. 2012. “The Informal Economy World- Revisited - Shifting Paradigms in Analysis and Policy. wide: Trends and Characteristics.” The Journal of Ap- World Bank, Washington, DC. plied Economic Research 6 (2): 103–32. Dabla-Norris, E., S. Guo, V. Haksar, M. Kim, K. Kochhar, K. Wiseman, and A. Zdzienicka. 2015. “The Checherita-Westphal, C., and P. Rother. 2012. “The New Normal: A Sector-Level Perspective on Productiv- Impact of High Government Debt on Economic ity Trends in Advanced Economies.” IMF Staff Discus- Growth and Its Channels: An Empirical Investigation sion Note 15/03, International Monetary Fund, Wash- for the Euro Area.” European Economic Review 56 (7): ington, DC. 1392–1405. De Gregorio, J. 2018. “Productivity in Emerging- Cherif, R., and F. Hasanov. 2016. “Breaking the Oil Market Economies: Slowdown or Stagnation?” PIIE Spell: The Gulf Falcons’ Path to Diversification.” IMF Working Paper 18-12, Peterson Institute for Interna- Working Paper 14/177, International Monetary Fund, tional Economics, Washington, DC. Washington, DC. De Loecker, J. 2007. “Do Exports Generate Higher Chevalier, P. A., R. Lecat, and N. Oulton. 2012. Productivity? Evidence from Slovenia.” Journal of Inter- “Convergence of Firm-Level Productivity, Globalisa- national Economics 73 (1): 69–98. tion and Information Technology: Evidence from France.” Economics Letters 116 (2): 244–46. de Nicola, F., V. Kehayova, and H. Nguyen. 2018. “On the Allocation of Resources in Developing East Cirera, X., R. Fattal-Jaef, and H. Maemir. 2019. Asia and Pacific.” Policy Research Working Paper “Taxing the Good? Distortions, Misallocation, and 8634, World Bank, Washington, DC. Productivity in Sub-Saharan Africa.” The World Bank Economic Review 00 (0): 1–26. De Ridder, M. 2017. “Investment in Productivity and the Long-Run Effect of Financial Crises on Output.” Cirera, X., and W. F. Maloney. 2017. The Innovation CESifo Working Paper 6243, Center for Economic Paradox: Developing-Country Capabilities and the Unre- Studies and Ifo Institute, Cambridge, U.K. alized Promise of Technological Catch-Up. Washington, DC: World Bank. De Visscher, S., M. Eberhardt, and G. Everaert. 2018. “Estimating and Testing the Multicountry Endogenous Claessens, S., E. Detragiache, R. Kanbur, and P. Wick- Growth Model.” Working Paper, 1–41. ham. 1997. “HIPCs’ Debt Review of the Issues World Bank and International Monetary Fund.” Journal of de Vries, G., K. de Vries, and M. Timmer. 2015. African Economies 6 (2): 231–54. “Structural Transformation in Africa: Static Gains, Dynamic Losses.” Journal of Development Studies 51 Claessens, S., and M. A. Kose. 2017. “Asset Prices and (6): 674–88. Macroeconomic Outcomes: A Survey.” BIS Papers 676, Bank for International Settlements, Basel. de Vries, G. J., A. A. Erumban, M. P. Timmer, I. Vos- koboynikov, and H. X. Wu. 2012. “Deconstructing the ———. 2018. “Frontiers of Macrofinancial Linkages.” BRICs: Structural Transformation and Aggregate BIS Papers 95, Bank for International Settlements, Productivity Growth.” Journal of Comparative Econom- Basel. ics 40 (2): 211–27. Clerides, S. K., S. Lach, and J. R. Tybout. 1998. “Is Dercon, S., M. Fafchamps, C. Pattillo, R. Oostendorp, Learning by Exporting Important? Micro-Dynamic J. Willem Gunning, P. Collier, A. Zeufack, et al. 2004. Evidence from Colombia, Mexico, and Morocco.” The “Do African Manufacturing Firms Learn from Export- Quarterly Journal of Economics 113 (3): 903–47. ing?” Journal of Development Studies 40 (3): 115–41. Cuberes, D., and M. Teignier. 2012. “Gender Gaps in Diao, X., M. McMillan, and D. Rodrik. 2017. “The the Labor Market and Aggregate Productivity.” Shef- Recent Growth Boom in Developing Economies: A field Economic Research Paper Series Sheffield, U.K. Structural Change Perspective.” NBER Working Paper 23132, National Bureau of Economic Research, Cam- ———. 2014. “Gender Inequality and Economic bridge, MA. Growth: A Critical Review.” Journal of International Development 26 (2): 260-276. Dieppe, A., S. Kiliç Çelik, and G. Kindberg-Hanlon. 244 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 Forthcoming. “Global Productivity Trends: Looking View.” Policy Research Working Paper 7743, World Through the Turbulence.” In Productivity in Emerging Bank, Washington, DC. and Developing Economies. Washington, DC: World Bank. Fajnzylber, P., W. F. Maloney, and G. V. Montes- Rojas. 2011. “Does Formality Improve Micro-Firm Dribe, M., M. Breschi, A. Gagnon, D. Gauvreau, H. Performance? Evidence from the Brazilian SIMPLES A. Hanson, T. N. Maloney, S. Mazzoni, et al. 2017. Program.” Journal of Development Economics 94 (2): “Socio-Economic Status and Fertility Decline: Insights 262–76. from Historical Transitions in Europe and North America.” Population Studies 71 (1): 3–21. Feenstra, R. C., R. Inklaar, and M. P. Timmer. 2015. “The Next Generation of the Penn World Table.” Duernecker, G., B. Herrendorf, and A. Valentinyi. American Economic Review 105 (10): 2017. “Structural Change within the Service Sector and 3150–82. the Future of Baumol’s Disease.” CEPR Discussion Paper 12467, Center for Economic Policy Research, Feldstein, M. 2017. “Underestimating the Real Growth London, U.K. of GDP, Personal Income, and Productivity.” Journal of Economic Perspectives 31(2): 145–64. Durlauf, S. N., P. A. Johnson, and J. R. W. Temple. 2005. “Growth Econometrics.” In Handbook of Eco- Fernald, J. 2012. “A Quarterly, Utilization-Adjusted nomic Growth, Volume 1A, edited by P. Aghion and S. Series on Total Factor Productivity.” Fed Working N. Durlauf, 555–677. Amsterdam: North Holland. Paper Series 2012–19. Federal Reserve Bank of San Francisco, San Francisco, CA. Durlauf, S. N., A. Kourtellos, and C. M. Tan. 2008. “Are Any Growth Theories Robust?” The Economic ———. 2015. “Productivity and Potential Output Journal 118 (527): 329–46. Before, During, and After the Great Recession.” Work- ing Paper Series 01-51, Federal Reserve Bank of San Duval, R., G. H. Hong, and Y. Timmer. 2017. Francisco. “Financial Frictions and the Great Productivity Slow- down.” IMF Working Papers 17/129, International Fernald, J., R. E. Hall, J. H. Stock, and M. W. Wat- Monetary Fund, Washington, DC. son. 2017. “The Disappointing Recovery Of Output After 2009.” NBER Working Paper 23543, National Easterly, W., and S. Fischer. 1994. “The Soviet Eco- Bureau of Economic Research, Cambridge, MA. nomic Decline Revisited.” Policy Research Working Paper 1284, World Bank, Washington, DC. Fernández, C., E. Ley, and M. F. J. Steel. 2001. “Model Uncertainty in Cross-Country Growth Regres- Easterly, W., and R. Levine. 2001. “What Have We sions.” Journal of Applied Econometrics 16 (5): 563–76. Learned From a Decade of Empirical Research on Growth? It’s Not Factor Accumulation: Stylized Facts Feyrer, J. 2008. “Aggregate Evidence on the Link be- and Growth Models.” World Bank Economic Review 15 tween Age Structure and Productivity.” Population and (2): 177-219. Development Review 34 (2008): ———. 2003. “Tropics, Germs, and Crops: How En- 78–99. dowments Influence Economic Development.” Journal Fisman, R., and I. Love. 2003. “Financial Develop- of Monetary Economics 50 (1): 3–39. ment and the Composition of Industrial Growth.” Eberhardt, M., and A. F. Presbitero. 2015. “Public NBER Working Paper 9583, National Bureau of Eco- Debt and Growth: Heterogeneity and Non-Linearity.” nomic Research, Cambridge, MA. Journal of International Economics 97 (1): 45–58. Foerster, A., A. Hornstein, P.-D. Sarte, and M. Wat- ECB (European Central Bank). 2017. “The Slowdown son. 2019. “Aggregate Implications of Changing Sec- in Euro Area Productivity in a Global Context.” ECB toral Trends.” Federal Reserve Bank of San Francisco. Economic Bulletin 3/2017. European Central Bank, Foster-McGregor, N., A. Isaksson, and F. Kaulich. Frankfurt. 2016. “Importing, Productivity and Absorptive Capaci- Eichengreen, B., and P. Gupta. 2013. “The Two ty in Sub-Saharan African Manufacturing and Services Waves of Service-Sector Growth.” Oxford Economic Firms.” Open Economies Review 27 (1): 87–117. Papers 65 (1): 96–123. Frankel, J. 2010. “The Natural Resource Curse: A Sur- Enache, M., E. Ghani, and S. O’Connell. 2016. vey.” NBER Working Paper 15836, National Bureau “Structural Transformation in Africa: A Historical of Economic Research, Cambridge, MA. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 245 Fuglie, K., M. Gautam, A. Goyal, and W. Maloney. and C. Villegas-Sanchez. 2017. “Capital Allocation and 2019a. Harvesting Prosperity: Technology and Productivi- Productivity in South Europe*.” The Quarterly Journal ty Growth in Agriculture. World Bank, Washington, of Economics 132 (4): 1915–67. DC. Gordon, R. J. 2016. The Rise and Fall of American Furceri, D., and A. Mourougane. 2012a. “The Effect of Growth: The U.S. Standard of Living since the Civil Financial Crises on Potential Output: New Empirical War. New Jersey: Princeton University Press. Evidence from OECD Countries.” Journal of Macroeco- nomics 34 (3): ———. 2018. “Why Has Economic Growth Slowed 822–32. When Innovation Appears to Be Accelerating?” NBER Working Paper 24554, National Bureau of Economic ———. 2012b. “How Do Institutions Affect Structur- Research, Cambridge, MA. al Unemployment in Times of Crises?” Panoeconomicus 59 (4): 393–419. Graner, M., and A. Isaksson. 2009. “Firm Efficiency and the Destination of Exports: Evidence from Kenyan Gallen, Y. 2018. “Motherhood and the Gender Plant-Level Data.” The Developing Economies 47 (3): Productivity Gap.” HCEO Working Paper 91, Human 279–306. Capital and Economic Opportunity Global Working Group, Chicago, IL. Griffith, R., S. Redding, and H. Simpson. 2003. “Productivity Convergence and Foreign Ownership At Galor, O., and D. N. Weil. 1996. “The Gender Gap , the Establishment Level.” Fertility , and Growth.” American Economic Review 86 (3): 374–87. Ha, J., M. A. Kose, and F. Ohnsorge, eds. 2019. Infla- tion in Emerging and Developing Economies: Evolution, Georgiev, Y., P. Nagy-Mohacsi, and A. Plekhanov. Drivers, and Policies. Washington, DC: World Bank. 2018. “Structural Reform and Productivity Growth in Emerging Europe and Central Asia.” ADB Economics Hall, R. E. 2014. “Quantifying the Lasting Harm to Working Paper Series. Asian Development Bank, Ma- the U.S. Economy from the Financial Crisis.” NBER nila, Philippines. Working Paper 20183, National Bureau of Economic Research, Cambridge, MA. Giri, R., S. N. Quayyum, and R. J. Yin. 2019. “Understanding Export Diversification: Key Drivers Hall, R. E., and C. I. Jones. 1999. “Why Do Some and Policy Implications.” IMF Working Paper. Inter- Countries Produce So Much More Output Per Worker national Monetary Fund, Washington, DC. than Others?” The Quarterly Journal of Economics 114 (1): 83–116. Goldberg, P. K., A. K. Khandelwal, N. Pavcnik, and P. Topalova. 2010. “Multi-Product Firms and Product Hallward-Driemeier, M., and G. Nayyar. 2017. Trou- Turnover in the Developing World: Evidence from ble in the Making? The Future of Manufacturing-Led India.” Review of Economic Statistics 92 (4): 1042–49. Development. Washington, DC: World Bank. Gollin, D., R. Jedwab, and D. Vollrath. 2015. Haltiwanger, J., S. Scarpetta, and H. Schweiger. 2014. “Urbanization with and without Industriali-zation.” “Cross Country Differences in Job Reallocation: The Journal of Economic Growth 21 (1): Role of Industry, Firm Size and Regulations.” Labour 35–70. Economics 26 (January): 11–25. Gollin, D., D. Lagakos, and M. E. Waugh. 2014. “The Harding, D., and A. Pagan. 2002. “Dissecting the Cy- Agricultural Productivity Gap.” Quarterly Journal of cle: A Methodological Investigation.” Journal of Mone- Economics 129 (2): 939–93. tary Economics 49 (2): 365–81. Gollin, D., S. L. Parente, and R. Rogerson. 2007. “The Haskel, J. E., S. C. Pereira, and M. J. Slaughter. 2007. Food Problem and the Evolution of International In- “Does Inward Foreign Direct Investment Boost the come Levels.” Journal of Monetary Economics 54 (4): Productivity of Domestic Firms?” Review of Economics 1230–55. and Statistics 89 (3): 482–96. Goñi, E., and W. F. Maloney. 2017. “Why Don’t Poor Hatzius, J., Z. Pandl, A. Phillips, D. Mericle, E. Countries Do R&D? Varying Rates of Factor Returns Pashtan, D. Struyven, K. Reichgott, and A. Thakkar. across the Development Process.” European Economic 2016. “Productivity Paradox v2.0 Revisited.” US Eco- Review 94 (May): 126–47. nomics Analyst. Goldman Sachs. Gopinath, G., Ş. Kalemli-Özcan, L. Karabarbounis, Hausmann, R., and C. Hidalgo. 2010. “Country Di- 246 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 versification, Product Ubiquity, and Economic Diver- Structural Reforms Play?” In World Economic Outlook: gence.” HKS Working Paper 10-045, Kennedy School Global Manufacturing Downturn, Rising Trade Barriers. of Government, Harvard University, Cambridge, MA. Washington, DC: International Monetary Fund. Hausmann, R., C. A. Hidalgo, S. Bustos, M. Coscia, S. IMF and World Bank. 2019. “Fintech: The Experience Chung, J. Jimenez, A. Simoes, et al. 2014. The Atlas of so Far - Executive Summary.” International Monetary Economic Complexity: Mapping Paths to Prosperity. Map- Fund, Washington, DC. ping Paths to Prosperity. Harvard Center for Interna- Inklaar, R., P. Woltjer, and D. Gallardo. 2019. “The tional Development, Cambridge, MA. Composition of Capital and Cross-Country Productivi- He, D., and W. Zhang. 2010. “How Dependent Is the ty Comparisons.” The Fifth World KLEMS Confer- Chinese Economy on Exports and in What Sense Has ence 36. University of Groningen, Netherlands. Its Growth Been Export-Led?” Journal of Asian Eco- Johnson, P., and C. Papageorgiou. 2018. “What Re- nomics 21 (1): 87–104. mains of Cross-Country Convergence?” MPRA Paper Helble, M., T. Long, and T. Le. 2019. “Sectoral and 89355, Munich Personal RePEc Archive, Munich, Skill Contributions to Labor Productivity in Asia.” Germany. ADBI Working Papers, March. Jordà, Ò. 2005. “Estimation and Inference of Impulse Responses by Local Projections.” American Economic Hicks, J. H., M. Kleemans, N. Y. Li, E. Miguel, D. Review 95 (1): 161–82. Albouy, J. Alvarez, L. Beaman, et al. 2017. “Reevaluating Agricultural Productivity Gaps with Jordà, Ò., M. Schularick, and A. M. Taylor. 2013. Longitudinal Microdata.” 23253. NBER Working “When Credit Bites Back.” Journal of Money, Credit Paper Series. and Banking 45 (2). Hidalgo, C., and R. Hausmann. 2009. “The Building Jorgenson, D. W., M. S. Ho, and K. J. Stiroh. 2008. Blocks of Economic Complexity.” Proceedings of the “A Retrospective Look at the U.S. Productivity Growth National Academy of Sciences of the United States of Resurgence.” Journal of Economic Perspetives 22 (1): 3– America 106 (26): 10570-10575. 24. Hsieh, C.-T., and P. J. Klenow. 2009. “Misallocation Kalemli-Özcan, S., L. Laeven, and D. Moreno. 2018. and Manufactring TFP in China and India.” Quarterly “Debt Overhang, Rollover Risk, and Corporate Invest- Journal of Economics 124 (4): 1403–48. ment: Evidence from the European Crisis.” NBER Working Paper 24555, National Bureau of Economic Hu, A. G. Z., G. H. Jefferson, and Q. Jinchang. 2005. Research, Cambridge, MA. “R&D and Technology Transfer: Firm-Level Evidence from Chinese Industry.” Review of Economics and Sta- Keller, W. 2004. “International Technology Diffu- tistics 87 (4): 780–86. sion.” Journal of Economic Literature 42 (3): 752–782. Imbs, J., and R. Wacziarg. 2003. “Stages of Diversifica- Keller, W., and S. R. Yeaple. 2009. “Multinational tion.” American Economic Review 93 (1): 63–86. Enterprises, International Trade, and Productivity Growth: Firm-Level Evidence from the United States.” IMF (International Monetary Fund). 2018. Review of Economics and Statistics 91 (4): 821–31. “Macroeconomic Developments and Prospects in Low- Income Developing Countries - 2018.” IMF Policy Kim, Y. E., and N. V. Loayza. 2019. “Productivity Paper, International Monetary Fund, Washington, Growth: Patterns and Determinants across the World.” DC. Policy Research Working Paper 8852, World Bank, Washington, DC. ———. 2014. “Sustaining Long-Run Growth and Macroeconomic Stability in Low-Income Countries - Koh, W. C., and S. Yu. 2019. “Macroeconomic and the Role of Structural Transformation and Diversifica- Financial Sector Policies.” In A Decade after the Global tion - Background Notes.” IMF Policy paper. Interna- Recession: Lessons and Challenges for Emerging and De- tional Monetary Fund, Washington, DC. veloping Economies, edited by A. Kose and F. Ohnsorge. Washington, DC: World Bank. ———. 2018. “World Economic Outlook: Cyclical Kraay, A., I. Soloaga, and J. Tybout. 2002. “Product Upswing, Structural Change.” International Monetary Quality, Productive Efficiency, and International Tech- Fund, Washington, DC. nology Diffusion: Evidence from Plant-Level Panel ———. 2019. “Reigniting Growth in Low-Income Data.” Policy Research Working Paper 2759, World and Emerging Market Economies: What Role Can Bank, Washington, DC. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 247 Krugman, P. 1988. “Financing vs. Forgiving a Debt McCullough, E. B. 2017. “Labor Productivity and Overhang.” Journal of Development Economics 29 (3): Employment Gaps in Sub-Saharan Africa.” Food Policy 253–68. 67 (February): 133–52. Kugler, A., and I. Rojas. 2018. “Do CCTs Improve Mcmillan, M., D. Rodrik, and C. Sepulveda. 2017. Employment and Earnings in the Very Long-Term? “Structural Change, Fundamentals and Growth: A Evidence from Mexico.” NBER Working Paper 24248, Framework and Case Studies.” NBER Working Paper National Bureau of Economic Research, Cambridge, 23378, National Bureau of Economic Research, Cam- MA. bridge, MA. Laeven, L., and F. Valencia. 2018. “Systemic Banking Mcmillan, M., D. Rodrik, and Í. Verduzco-Gallo. Crises Revisited.” IMF Working Paper 18/206, Inter- 2014. “Globalization, Structural Change, and Produc- national Monetary Fund, Washington, DC. tivity Growth, with an Update on Africa.” World De- velopment 63 (November): 11–32. Lagakos, D., and M. E. Waugh. 2013. “Selection, Agri- culture, and Cross-Country Productivity Differences.” Melo, P. C., D. J. Graham, and R. Brage-Ardao. 2013. American Economic Review 103 (2): 948–80. “The Productivity of Transport Infrastructure Invest- ment: A Meta-Analysis of Empirical Evidence.” Region- Leipziger, D. M., and V. Thomas. 1993. “The Lessons al Science and Urban Economics 43 (5): 695–706. of East Asia: An Overview of Country Experience.” World Bank, Washington, DC. Mensah, E. B., and A. Szirmai. 2018. “Africa Sector Database (ASD): Expansion and Update.” UNU- Levchenko, A. A. 2007. “Institutional Quality and MERIT Working Paper Series 2018-020,UNU- International Trade.” The Review of Economic Studies MERIT, Maastricht, Netherland. 74 (3): 791–819. Moretti, E. 2004. “Workers’ Education, Spillovers, and Levchenko, A., and J. Zhang. 2016. “The Evolution of Productivity: Evidence from Plant-Level Production Comparative Advantage: Measurement and Welfare Functions.” American Economic Review 94 (3): 656–90. Implications.” Journal of Monetary Economics 78 (April): 96–111. Morris, S., and H. S. Shin. 1998. “Unique Equilibrium in a Model of Self-Fulfilling Currency Attacks.” Ameri- Levine, R. 1997. “Financial Development and Eco- can Economic Review 88 (3): 587–97. nomic Growth: Views and Agenda.” Journal of Econom- ic Literature 35 (2): 688–726. Nakamura, L. 2001. “What Is the U.S. Gross Invest- ment in Intangibles? (At Least) One Trillion Dollars a Loayza, N. V. 2016. “Informality in the Process of Year!” Working Papers 01-15, Federal Reserve Bank of Development and Growth.” World Economy 39 (12): Philadelphia. 1856–1916. Nguimkeu, P., and C. Okou. 2019. “Informality.” In Lowder, S. K., J. Skoet, and T. Raney. 2016. “The The Future of Work In Africa: Harnessing the Potential of Number, Size, and Distribution of Farms, Smallholder Digital Technologies for All, edited by J. Choi, M. Dutz, Farms, and Family Farms Worldwide.” World Develop- and Z. Usman, 107–39. Washington, DC: World ment 87 (November): 16–29. Bank. Maestas, N., K. Mullen, and D. Powell. 2016. “The North, D. C. 1990. Institutions, Institutional Change Effect of Population Aging on Economic Growth, the and Economic Performance. Cambridge: Cambridge Labor Force and Productivity.” Working Paper 1063. University Press. RAND Labor & Population, Santa Monica, CA. Obstfeld, M. 1996. “Models of Currency Crises with Maloney, W. F., and G. Nayyar. 2018. “Industrial Self-Fulfilling Features.” European Economic Review 40 Policy, Information, and Government Capacity.” The (95): 1037–47. World Bank Research Observer 33 (2): 189–217. OECD (Organisation for Economic Co-operation and Martins, P. M. G. 2019. “Structural Change: Pace, Development). 2015. The Future of Productivity. Paris: Patterns and Determinants.” Review of Development Organisation for Economic Co-operation and Develop- Economics 23 (1): 1–32. ment. Mason, A. D., and S. Shetty. 2019. A Resurgent East Ollivaud, P., Y. Guillemette, and D. Turner. 2016. Asia: Navigating a Changing World. Washington, DC: “The Links Between Weak Investment and the Slow- World Bank. down in OECD Productivity and Potential Output 248 CHAPTER 3 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 Growth.” Working Papers No. 1304, OECD Econom- ment.” Journal of Economic Growth 9 (2): 131–65. ics Department 1304. Organisation for Economic Co- Roy, S., M. Kessler, and A. Subramanian. 2016. operation and Development, Paris. “Glimpsing the End of Economic History? Uncondi- Osakwe, P. N., and J.-M. Kilolo. 2018. “What Drives tional Convergence and the Missing Middle Income Export Diversification? New Evidence from a Panel of Trap.” Center for Global Development Working Paper Developing Countries.” UNCTAD Research Paper 3, 438, Center for Global Development, Washington, United Nations Conference on Trade and Develop- DC. ment, New York. Rozenberg, J., and M. Fay. 2019. Beyond the Gap: How Oulton, N., and M. Sebastiá-Barriel. 2017. “Effects of Countries Can Afford the Infrastructure They Need While Financial Crises on Productivity, Capital and Employ- Protecting the Planet. Washington, DC: World Bank. ment.” Review of Income and Wealth 63 (February): S90 Ruiz Pozuelo, J., A. Slipowitz, and G. Vuletin. 2016. –112. “Democracy Does Not Cause Growth: The Im- Patel, D., J. Sandefur, and A. Subramanian. 2018. portance of Endogeneity Arguments.” IDB Working “Everything You Know about Cross-Country Conver- Paper 694, Inter-American Development Bank, Wash- gence Is (Now) Wrong.” Center for Global Develop- ington, D.C. ment. Accessed October 15. https://www.cgdev.org/ Sahay, R., M. Cihak, P. N’Diaye, R. B. Barajas, D. blog/everything-you-know-about-cross-country- Ayala, Y. Gao, A. Kyobe, et al. 2015. “Rethinking Fi- convergence-now-wrong. nancial Deepening: Stability and Growth in Emerging Pereira, A. M., and J. M. Andraz. 2013. “On the Eco- Markets.” Staff Discussion Note 15/08, International nomic Effects of Public Infrastructure Investment: A Monetary Fund, Washington, DC. Survey of the International Evidence.” College of Wil- Sandri, D. 2015. “Dealing with Systemic Sovereign liam and Mary Department of Economics Working Debt Crises: Fiscal Consolidation, Bail-Ins or Official Paper 108, College of William and Mary, Williams- Transfers?” IMF Working Paper 17/223, International burg, VA. Monetary Fund, Washington, DC. Poirson, H., C. Pattillo, and L. Ricci. 2002. “External Schnitzer, M. 2002. “Debt v. Foreign Direct Invest- Debt and Growth.” IMF Working Papers 02/69, Inter- ment: The Impact of Sovereign Risk on the Structure national Monetary Fund, Washington, DC. of International Capital Flows.” Economica 69: 41–67. ———. 2004. “What Are the Channels Through Sinha, R. 2016. “Sectoral Productivity Gaps and Ag- Which External Debt Affects Growth?” IMF Working gregate Productivity.” Policy Research Working Paper Paper 04/15, International Monetary Fund, Washing- 7737. World Bank, Washington, DC. ton, DC. SNA (System of National Accounts). 2008. System of Posen, A., and J. Zettelmeyer. 2019. Facing Up to Low National Accounts. New York: United National Statisti- Productivity Growth. Washington, DC: Peterson Insti- cal Commission. https://unstats.un.org/unsd/ tute for International Economics. nationalaccount/sna2008.asp. Reinhart, C. M., and K. S. Rogoff. 2009. “The After- Solow, R. M. 1957. “Technical Change and the Aggre- math of Financial Crises.” American Economic Review gate Production Function.” The Review of Economics 99 (2): 466–472. and Statistics 39 (3): 312–20. ———. 2010. “Growth in a Time of Debt.” American Stocker, M., J. Baffes, Y. M. Some, D. Vorisek, and C. Economic Review: Papers & Proceedings 100 (May): 573 M. Wheeler. 2018. “The 2014–16 Oil Price Collapse –78. in Retrospect: Sources and Implications.” Policy Re- Rodrik, D. 2011. “The Future of Economic Conver- search Working Paper. 8419 World Bank, Washing- gence.” NBER Working Paper 17400, National Bureau ton, DC. of Economic Research, Cambridge, MA. Syverson, C. 2011. “What Determines Productivity?” ———. 2016. “Premature Deindustrialization.” Jour- Journal of Economic Literature 49 (2): 326–365. nal of Economic Growth 21 (1): 1–33. ———. 2016. “Challenges to Mismeasurement Expla- Rodrik, D., A. Subramanian, and F. Trebbi. 2004. nations for the U.S. Productivity Slowdown.” NBER “Institutions Rule: The Primacy of Institutions Over Working Paper 21974, National Bureau of Economic Geography and Integration in Economic Develop- Research, Cambridge, MA. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 3 249 Teulings, C. N., and N. Zubanov. 2014. “Is Economic World Bank. 2012. World Development Report: Gender Recovery a Myth? Robust Estimation of Impulse Re- Equality and Development. Washington, DC: World sponses.” Journal of Applied Econometrics 29 (3): 497– Bank. 514. ———. 2017. Global Economic Prospects: Weak Invest- Tuan, C., L. F. Y. Ng, and B. Zhao. 2009. “China’s ment in Uncertain Times. January. Washington, DC: Post-Economic Reform Growth: The Role of FDI and World Bank. Productivity Progress.” Journal of Asian Economics 20 (3): 280–93. ———. 2018a. Africa’s Pulse: An Analysis of Issues Shaping Africa’s Economic Future. October. Washing- UNECE (United Nations Economic Commission for ton, DC: World Bank. Europe). 2008. Non-Observed Economy in National Accounts. Geneva: United Nations Economic Commis- ———. 2018b. World Development Report 2018: sion for Europe. Learning to Realize Education’s Promise. Washington, DC: World Bank. Üngör, M. 2013. “De-Agriculturalization as a Result of Productivity Growth in Agriculture.” Economics Letters ———. 2018c. Global Economic Prospects: Broad-Based 119 (2): 141–45. Upturn, but for How Long? January. Washington, DC: World Bank. ———. 2017. “Productivity Growth and Labor Real- location: Latin America versus East Asia.” Review of ———. 2018d. Improving Public Sector Performance: Economic Dynamics 24 (March): 25–42. Through Innovation and Inter-Agency Coordination. Washington, DC: World Bank. van Ark, B., M. O’Mahony, and M. P. Timmer. 2008. “The Productivity Gap between Europe and the Unit- ———. 2018e. The Human Capital Project. Washing- ed States: Trends and Causes.” Journal of Economic ton, DC: World Bank. Perspectives 22 (1): 25–44. ———. 2019a. Global Economic Prospects: Darkening van Biesebroeck, J. 2009. "Disaggregate Productivity Skies. January. Washington, DC: World Bank. Comparisons: Sectoral Convergence in OECD Coun- ———. 2019b. Global Economic Prospects: Heightened tries." Journal of Productivity Analysis 32 (2): 63-79. Tensions, Subdued Investment. June. Washington, DC: Van Reenen, J. 2011. “Does Competition Raise World Bank. Productivity through Improving Management Quali- ———. 2019c. “Growth in Low-Income Countries: ty?” International Journal of Industrial Organization 29 Evolution, Prospects, and Policies.” In Global Economic (3): 306–16. Prospects, Heightened Tensions, Subdued Investment. Vargas, M. 2015. “Informality in Paraguay; Macro- Washington, DC: World Bank. Micro Evidence and Policy Implications.” IMF Work- ———. 2019d. World Development Report 2020: Trad- ing Papers 15/245, International Monetary Fund, ing for Development in the Age of Global Value Chains. Washington, DC. Washington, DC: World Bank. Vorisek, D., and S. Yu. forthcoming. “Understanding ———. 2019e. World Development Report: The Chang- the Cost of Achieving the Sustainable Development ing Nature of Work. Washington, DC: World Bank. Goals.” Policy Research Working Paper, World Bank, Washington, DC. ———. 2019f. Africa’s Pulse: An Analysis of Issues Shap- ing Africa’s Economic Future. October. Washington, Wei, Y., and X. Liu. 2006. “Productivity Spillovers DC: World Bank. from R and D, Exports and FDI in China’s Manufac- turing Sector.” Journal of International Business Studies World Bank and DRCSC (Development Research 37 (4): 544–57. Center of the State Council, the People's Republic of China). 2019. Innovative China: New Drivers of Wooster, R. B., and D. S. Diebel. 2010. “Productivity Growth. Washington, DC: World Bank. Spillovers from Foreign Direct Investment in Develop- ing Countries: A Meta-Regression Analysis.” Review of Development Economics 14 (3): 640–55. CHAPTER 4 THE FOURTH WAVE Rapid Debt Buildup G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 253 The global economy has experienced four waves of debt accumulation over the past fifty years. The first three ended with financial crises in many emerging market and developing economies. During the current wave, which started in 2010, the increase in debt in these economies has already been larger, faster, and more broad- based than in any of the previous three waves. Current low interest rates —which markets expect to be sustained into the medium term—appear to mitigate some of the risks associated with high debt. However, emerging market and developing economies are also confronted by weak growth prospects, mounting vulnerabilities, and elevated global risks. A menu of policy options is available to reduce the likelihood of the current debt wave ending in crises and, if crises were to take place, to alleviate their impact. Introduction triggered a lively debate about the benefits and risks of further government debt accumulation to Waves of rapid debt accumulation have been a finance increased spending. It is generally agreed recurrent feature of the global economy over the that public borrowing can be beneficial, past fifty years, in both advanced economies and particularly in EMDEs with large development emerging market and developing economies challenges, if it is used to finance growth- (EMDEs). Since the 2008-09 global financial enhancing investments, such as infrastructure, crisis, another wave has been building, with global health care, and education. Debt accumulation debt reaching an all-time high of about 230 can also be appropriate temporarily as part of percent of global GDP in 2018 (Figure 4.1). counter-cyclical fiscal policy, to boost demand and activity in economic downturns. Total EMDE debt reached almost 170 percent of GDP in 2018 ($55 trillion), an increase of 54 However, high debt carries significant risks for percentage points of GDP since 2010. Although EMDEs, as it makes them more vulnerable to China accounted for the bulk of this increase—in external shocks. The rollover of existing debt can part due to its sheer size—the debt-buildup was become increasingly difficult during periods of broad-based: In about 80 percent of EMDEs total financial stress, potentially leading to a crisis. High debt was higher in 2018 than in 2010. Following government debt levels can also limit the size and a steep fall during 2000-10, debt has also risen in effectiveness of fiscal stimulus during downturns, low-income countries (LICs), reaching 67 percent and can dampen longer term growth by weighing of GDP (around $270 billion) in 2018, up from on productivity-enhancing private investment. 48 percent of GDP (around $140 billion) in 2010. EMDEs have been navigating dangerous waters as the current debt wave has coincided with a decade In contrast, in advanced economies, total (public of repeated growth disappointments, and they are and private) debt has remained steady near the now confronted by weaker growth prospects in a record levels reached in the early aftermath of the fragile global economy (Kose and Ohnsorge global financial crisis, at 264 percent of GDP in 2019). In addition to their rapid debt buildup 2018 ($130 trillion). While government debt has during the current wave, these economies have risen to a high of 104 percent of GDP ($50 accumulated other vulnerabilities, such as growing trillion), private sector debt has fallen slightly fiscal and current account deficits, and a amid deleveraging in some sectors. compositional shift toward short-term external debt, which could amplify the impact of shocks. The current environment of low interest rates, combined with subpar global growth, has Thus, despite current exceptionally low real interest rates, including at long maturities, the latest wave of debt accumulation could follow the Note: This chapter was prepared by a team led by M. Ayhan Kose, Peter Nagle, Franziska Ohnsorge, and Naotaka Sugawara, with historical pattern and eventually culminate in contributions from Jongrim Ha, Alain Kabundi, Sergiy Kasyanenko, financial crises in EMDEs. A sudden global shock, Wee Chian Koh, Franz Ulrich Ruch, Lei (Sandy) Ye, and Shu Yu. It is based on Kose et al. 2019. Vanessa Banoni, Julia Norfleet, Jankeesh such as a sharp rise in interest rates or a spike in Sandhu, Shijie Shi, and Jinxin Wu provided research assistance. risk premia, could lead to financial stress in more 254 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 4.1 Evolution of debt employs a wide range of approaches, including Global debt has trended up since 1970, reaching around 230 percent of event studies, econometric models, country case GDP in 2018. Debt has risen particularly rapidly in EMDEs, reaching a studies, and a detailed review of historical peak of about 170 percent of GDP in 2018. Much of the increase since episodes. Specifically, it examines the following 2010 has occurred in the private sector, particularly in China. Debt in low- income countries has started to rise after a prolonged period of decline questions. following debt relief measures in the late 1990s and 2000s. Advanced- economy debt has been broadly flat since the global financial crisis, with • How have previous waves of debt in EMDEs increased government debt more than offsetting a mild deleveraging in the private sector. evolved? • How does the current wave of debt A. Global debt B. Debt in EMDEs accumulation compare to earlier waves? • What are the macroeconomic implications of rapid debt accumulation? • What are the lessons and policy implications for EMDEs? Contributions. An extensive literature has studied various aspects of debt accumulation, especially in C. Debt in LICs D. Debt in advanced economies the context of government and private debt crises. This chapter adds to this literature in five dimensions. First, the chapter provides the first in- depth analysis of the similarities and differences among four distinct waves of broad-based debt accumulation in EMDEs since 1970.1 Each wave contains episodes that have been widely examined in the literature but they have rarely been put into a common framework. Examining debt buildups Source: International Monetary Fund; World Bank. as waves allows a richer analysis by considering the Note: Averages computed with current U.S. dollar GDP as weight and shown as a 3-year moving average. Vertical lines in gray are for years 1970, 1990, 2002, and 2010. interaction of global drivers with country-specific B. Dashed lines refer to EMDEs excluding China. conditions. Earlier work has taken on a longer Click here to download data and charts. historical perspective and focused mainly on debt developments in advanced economies, typically vulnerable economies. Indeed, these risks were based on case studies. Second, in contrast to earlier illustrated by the recent experiences of Argentina studies, the chapter puts the ongoing (fourth) and Turkey, which witnessed sudden episodes of wave of broad-based debt accumulation in sharply rising borrowing costs and severe growth EMDEs into historical perspective.2 Third, the slowdowns in 2018. Among LICs, the rapid increase in debt and the 1 Previous studies have examined the impact of mounting government debt in advanced economies (BIS 2015; Cecchetti, shift from concessional toward financial market Mohanty, and Zampolli 2011; Erhardt and Presbitero 2015; and non-Paris Club creditors have raised concerns Eichengreen et al. 2019; Mbaye, Moreno-Badia, and Chae 2018a; about debt transparency and debt collateralization. OECD 2017; Panizza and Presbitero 2014; Reinhart, Reinhart, and Rogoff 2012). For EMDEs, previous studies have often analyzed Elevated debt in major EMDEs, including China, certain periods of debt distress, or crises in individual countries. For could amplify the impact of adverse events and example, contagion from the Asian crisis has been examined by Baig and Goldfajn (1999); Chiodo and Owyang (2002); Claessens and trigger a growth slowdown, posing risks to global Forbes (2013); Glick and Rose (1999); Kaminsky and Reinhart and EMDE growth. (2000, 2001); Kawai, Newfarmer, and Schmukler (2005); Moreno, Pasadilla, and Remolona (1998); and Sachs, Cooper, and Bosworth Against this backdrop, this chapter compares the (1998). 2 The recent debt accumulation, without the historical context, current wave of debt buildup to previous episodes have been discussed in IMF (2019a, 2016a) and World Bank (2015, and considers the policy implications. The chapter 2016a, 2017a). G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 255 chapter undertakes the first comprehensive to lower external vulnerabilities and strengthen empirical analysis of a large number of individual policy frameworks. These similarities notwith- episodes of rapid government and private debt standing, the financial instruments used for accumulation in 100 EMDEs since 1970. The borrowing have shifted over time as new separate analysis of individual episodes offers key instruments or financial actors emerged. The insights into the macroeconomic consequences, at nature of EMDE borrowers in international the country level, of debt accumulation. Earlier financial markets has also changed, with the work has examined developments in government private sector accounting for a growing share of and private debt markets separately, or focused on borrowing through the first three waves. a smaller group of (mostly advanced) economies or regions.3 Fourth, the chapter identifies the most Another global wave of debt underway. The debt frequent triggers of crises and the country-level buildup in EMDEs in the fourth wave, which vulnerabilities that contribute to or exacerbate started in 2010, has already been larger, faster and crises. Fifth, armed with insights from an extensive broader-based than in any of the previous waves. analysis of the global and national waves of debt The annual increase in EMDE debt since 2010 accumulation and the empirical linkages between has been larger, by some margin, than during the elevated debt and financial crises, as well as the first three waves. Whereas previous waves were earlier literature, the study distills lessons and largely regional in nature, the fourth wave was presents a rich menu of policy options that can global, with total debt rising in more than 70 help EMDEs boost resilience to future crises. percent of EMDEs in all regions and rising by at least 20 percentage points of GDP in more than The chapter documents the following findings. one-third of EMDEs. In the fourth wave, most national episodes of debt accumulation combined Three previous waves. Prior to the current wave, government and private debt accumulation, in EMDEs experienced three waves of broad-based contrast to the previous three waves which had a and rapid debt buildup. The first (1970-89) was greater focus a single sector. focused in Latin America and the Caribbean (LAC) and Sub-Saharan Africa (SSA), the second Debt buildups often associated with crises. Since (1990-2001) in East Asia and Pacific (EAP) and 1970, there have been about 520 national episodes some other EMDEs in Europe and Central Asia of rapid debt accumulation in 100 EMDEs. (ECA) and LAC, and the third (2002-09) was Around half of these episodes were accompanied chiefly in ECA. The fourth wave (2010 onwards), by a financial crisis, with sizeable economic costs. in contrast, has covered all EMDE regions. Crises during rapid government debt buildups featured larger output losses than crises during Similarities and differences among previous rapid private debt buildups. waves. All debt waves began during prolonged periods of very low real interest rates, and were Debt accumulation as shock amplifier. While often facilitated by changes in financial markets financial crises during rapid debt accumulation that contributed to rapid borrowing. The three episodes were often triggered by external shocks, earlier waves all ended with widespread financial such as sudden increases in global interest rates, crises and coincided with global recessions (1982, domestic vulnerabilities often increased the 1991, and 2009) or downturns (1998, 2001). likelihood of crises and amplified their adverse Crises were usually followed by reforms designed impact. Most countries where crises erupted suffered from unsustainable combinations of inadequate fiscal, monetary, or regulatory 3 Government debt crises have been discussed in Kindleberger and Aliber (2011); Reinhart, Reinhart, and Rogoff (2012); Reinhart frameworks. Crises were more likely, or the and Rogoff (2010, 2011); and World Bank (2019a). Credit booms economic distress they caused was more severe, in have been examined in Dell’Arricia et al. (2014, 2016); Elekdag and countries with higher external debt—especially Wu (2013); Jordà, Schularick, and Taylor (2011); Mendoza and Terrones (2008, 2012); Ohnsorge and Yu (2016); and Tornell and short-term—and lower levels of international Westermann (2005). reserves. 256 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 Policy implications. While there is no magic and is followed by two consecutive years of bullet of a policy prescription to ensure that the decline. current debt wave proceeds smoothly, the experience of past waves of debt points to the • The dating of the end of waves is consistent critical role of policy choices in determining the with the approximate timing of policies to outcomes of these episodes. Sound debt resolve the financial crises that they management and debt transparency can help engendered. In 1989, for example, Mexico reduce borrowing costs, enhance debt issued the first Brady bonds, marking the sustainability, and dampen fiscal risks. Strong beginning of resolution of the Latin American monetary, exchange rate, and fiscal policy debt crisis. In 1998-2001, a series of IMF frameworks can safeguard EMDEs’ resilience in a programs led to debt resolution after the East fragile global economic environment. Robust Asian and Russian financial crises. In 2009, regulatory and supervisory regimes, which are also governments implemented a large-scale, well coordinated between home and host internationally coordinated policy stimulus to supervisors of foreign banks, can help contain combat the adverse effects of the global financial market risks and encourage prudent financial crisis. lending to the private sector. Good corporate governance can help ensure that debt is used for Features of the first three waves the most productive purposes. This identification yields three historical waves of global debt accumulation and one ongoing. The Evolution of past waves first wave runs from 1970 to 1989, the second of debt from 1990 to 2001, the third from 2002-09, and the fourth since 2010. The buildup of EMDE debt since 1970 has not been linear. At different points in time, different First wave countries, and regions, have undergone periods of rapid debt accumulation (Figure 4.2). These have The first wave spanned the 1970s-80s, with often been followed by crises, and periods of borrowing primarily accounted for by deleveraging. This section examines “waves” of governments in LAC and low-income countries in broad-based debt accumulation in EMDEs, and SSA (Kose et al. 2019). The combination of low considers their similarities and differences. It interest rates and a rapidly growing syndicated identifies four waves of debt since 1970, of which loan market encouraged EMDE governments to the fourth is still ongoing. borrow heavily (Gadanecz 2004). Identification of the four waves LAC. The debt buildup was greatest in LAC, which accounted for over half of all debt flows to The dating of the four waves meets some basic EMDEs in 1973-81 (Bertola and Ocampo 2012; criteria. Devlin 1990). As part of a strategy of import substitution industrialization, countries relied on • The first wave begins in 1970.4 Data external debt to finance infrastructure and limitations prevent more detailed analysis of investment in heavy industries (Baer 1972; Bruton the period prior to 1970. 1998; Diaz-Alejandro, Krugman, and Sachs 1984). Many LAC economies borrowed from • The end of a wave is broadly defined as the international banks via new syndicated loan year in which the total debt-to-GDP ratio in markets, which provided a way to recycle dollar- the affected region or country group peaks denominated oil revenues from oil-exporters to importers (Altunbaş, Gadanecz, and Kara 2006). 4 1970 is also used as the starting year by Laeven and Valencia Vulnerabilities mounted, as widening current (2018) in their database of nancial crises. account and fiscal deficits were financed by G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 257 external debt, and inflation rose, while pegged FIGURE 4.2 Debt in EMDEs exchange rate regimes were backed by low levels of The region and sector of debt accumulation has varied substantially over reserves. The late 1970s and early 1980s saw a the four EMDE waves (1970-1989, 1990-2001, 2002-09, and since 2010). series of global shocks, including an oil price spike and U.S. monetary policy tightening that A. Total debt B. External debt accompanied a global recession. The crisis began in 1982 with Mexico announcing that it would not be able to service its debt, and spread rapidly to other LAC and SSA countries. The U.S. administration’s Brady plan eventually provided comprehensive debt relief in 1989 (Cline 1995; Unal, Demirgüç-Kunt, and Leung 1993). The debt crisis resulted in a “lost decade” in LAC, with GDP per capita not recovering its pre-crisis level until 1993, after having grown by 50 percent C. Government debt D. Private debt during 1970-1980 (Loayza, Fajnzylber, and Calderón 2005). SSA. Many low-income countries (LICs), especially in SSA, borrowed heavily in the 1970s and 1980s from official creditors (Daseking and Powell 1999). Debt was typically used to finance domestic-focused industry (Greene 1989). Amid rising global interest rates and deteriorating terms of trade, several countries suffered debt crises in E. Government debt in EMDE regions, F. Private debt in EMDE regions, excluding China excluding China the 1980s (Dornbusch, Branson, and Cline 1985). In response, the World Bank and IMF provided financial support for adjustment programs, while the Paris Club creditors agreed to “flow rescheduling,” under which debt principal and interest payments were delayed. While these policies helped with liquidity issues, they led to a steady increase in debt (Dicks 1991). While growth in LICs was robust in the 1970s, it Source: International Monetary Fund; World Bank. was persistently weak in the subsequent two A. B. Light blue and yellow lines exclude China. C.D. Averages computed with current U.S. dollar GDP as weight and shown as a 3-year moving decades with income per capita falling during average. Dashed lines for EAP refer to EAP excluding China. Lines for ECA start in 1995 due to smaller sample size prior to that year. Vertical lines in gray are for years 1970, 1990, 2002, and 2010. 1980-99 amid rapid population growth. E.F. GDP-weighted averages. EAP = East Asia and Pacific, ECA = Europe and Central Asia; Eventually, the World Bank and IMF, along with LAC = Latin America and Caribbean; MNA = Middle East and North Africa; SAR = South Asia; SSA = Sub-Saharan Africa. other multilaterals and bilateral creditors, Click here to download data and charts. announced the “Heavily Indebted Poor Countries” (HIPC) initiative in 1996, which was followed by the Multilateral Debt Relief Initiative governments in ECA to borrow heavily; it ended (MDRI) in 2005 (IMF 2006; World Bank and with crises in these regions in 1997-2001. IMF 2017a). EAP. The EAP region registered one of the fastest Second wave increases in private debt in the 1990s. Poor bank regulation and supervision, together with implicit The second wave ran from 1990 until the early government guarantees for banks and corporates, 2000s as financial and capital market liberalization encouraged risk taking by the domestic financial enabled banks and corporates in EAP and sector and allowed already highly leveraged 258 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 corporates to borrow heavily (Kose et al. 2019). Third wave Countries also suffered from poor corporate governance, a prominent presence of state-owned The third wave was a runup in private sector enterprises (e.g., Thailand), weak business climates borrowing in ECA from U.S. and EU- (e.g., Indonesia), and heavy investment in non- headquartered “mega-banks” after regulatory tradeable sectors such as commercial real estate easing and amid initially accommodative (e.g., Thailand; Krugman 2000). Rising private monetary policy in advanced economies (Cetorelli debt, particularly short-term debt, left several EAP and Goldberg 2011). While the buildup of debt in countries (Indonesia, Malaysia, Philippines, and the third wave primarily occurred in advanced Thailand) vulnerable to a reversal in capital flows. economies, the emerging mega-banks fueled a steep increase in direct cross-border lending on the In early 1997, capital inflows to Thailand began to interbank market, lending through subsidiaries, taper off amid investor concerns about external and investment in EMDE debt markets debt sustainability. Despite government (Balakrishnan et al. 2011). intervention in early 1997, Thailand was forced to abandon its currency peg in July 1997. Financial This wave ended when the global financial crisis markets quickly turned on countries with similar disrupted bank financing in 2008-09 and tipped vulnerabilities, and Indonesia, Korea, Malaysia, several ECA economies into deep recessions and the Philippines experienced large capital (Aslund 2010). The crisis in ECA was short-lived, outflows which resulted in substantial pressure on in part due to IMF and EU support (Berglof et al. their currencies (Corsetti, Pesenti, and Roubini 2009). In contrast to the ECA region (and 1998; Kawai, Newfarmer, and Schmukler 2005). advanced economies), most EMDEs proved resilient to the global financial crisis, in part Corporates were unable to service their debt, because they had limited exposures to the actual resulting in large loan losses for banks and global shocks at the time (Kose and Prasad 2010). triggering banking crises. Governments created Many EMDEs also improved debt management, “bad banks” to absorb non-performing loans of supporting a reduction in currency, interest and commercial lenders, recapitalized banks, and maturity risks (Anderson, Silva, and Velandia- improved corporate debt restructuring regimes Rubiano 2010). (Mishkin 1999). Prior to the crisis, the sharp rise in borrowing among EAP countries was Similarities and differences between waves accompanied by rapid GDP growth but, during the crisis, GDP and investment growth The first three waves of broad-based debt plummeted. accumulation featured several similarities (Box 4.1). At the beginning of each wave, the initial LAC and ECA. The late 1990s saw crises occur in debt buildup was associated with low or falling some other major EMDEs, notably Russia, global interest rates and major changes in financial Argentina, and Turkey. These countries markets, often in response to deregulation. The experienced sovereign debt crises when a broad- first three waves eventually witnessed severe and based loss of investor confidence triggered capital widespread financial crises in EMDEs with severe outflows and forced governments to abandon macroeconomic consequences, usually triggered by currency pegs. A notable exception was Brazil, external shocks and amplified by domestic which suffered a currency crisis in 1999, but vulnerabilities. Financial crises were typically avoided a banking and sovereign debt crisis. The followed by reforms in affected countries to lower authorities dampened exchange rate depreciation, external vulnerabilities and strengthen policy but at considerable fiscal cost. The earlier “Tequila frameworks. crisis” in 1995 also falls into the second wave, when Mexico accepted assistance from the IMF There were also noticeable differences between the and others to stem a currency crisis but avoided a three waves. The sectors and regions that were the full sovereign debt crisis (Laeven and Valencia most active borrowers, and the financial 2018; Kose et al. 2019). instruments involved changed over the course of G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 259 the three waves: borrowing shifted from the FIGURE 4.3 The fourth wave: Debt accumulation government sector to the private sector, while the Since the global financial crisis, another wave of debt accumulation has type of debt moved from syndicated loan markets been underway. The fourth wave has been especially rapid in EMDEs, and in the first wave, to government bond markets and has seen government debt increasing in tandem with mounting private sector debt. The share of debt accounted for by bonds has continued to international private sector borrowing in the rise, and large EMDEs have seen a sharp increase in domestically issued second wave, to cross-border and foreign-owned bonds. bank lending in the third wave. In all three waves, A. Total debt B. Government debt financial crises resulted in substantial economic damage, but their severity varied between waves and across regions. The waves also varied in terms of the speed of resolution, with sovereign debt crises typically taking longer to resolve, and having much larger negative macroeconomic impacts than private debt crises. The current wave of debt in historical context C. EMDE external debt, by borrower D. Change in EMDE bond issuance, and type of instrument 2010-16, by sector and domicile Since 2010, another wave of debt accumulation has been building. The buildup has been global, but especially fast in EMDEs (Box 4.2, Figure 4.3). As a result, total debt in EMDEs has risen to almost 170 percent of GDP, on average, in 2018—a record high—from 114 percent of GDP in 2010 (Kose et al. 2019). China, where corporate debt has soared post-crisis, accounted for the bulk of this buildup—partly due to its sheer Source: International Monetary Fund; World Bank. size—but the buildup was broad-based. Excluding C. “Public-official” includes “private other” which is chiefly accounted for by export guarantee agencies. China, total EMDE debt has risen to a near-record D. Chart shows the change in debt securities (in percentage points of GDP) between 2010 and 2016 (last observation). Other EMDEs includes 8 countries. Data for India are unavailable. 107 percent of GDP in 2018. The debt-to-GDP Click here to download data and charts. ratio has risen in all EMDE regions with the exception of SAR, where it has been broadly flat, and in almost 80 percent of EMDEs, with more nancial markets, and very low interest rates (as a than one-third seeing increases of at least 20 result of accommodative monetary policy percentage points of GDP. following the global nancial crisis). Financial The current, fourth, wave of debt accumulation systems in EMDEs have deepened and become bears many similarities to the previous waves. But more complex (Didier and Schmukler 2014). there are also important differences. Among these Both corporate and sovereign borrowers have is its sheer magnitude: it is the largest, fastest and increasingly accessed capital markets, in some regions following the retrenchment of large most broad-based wave of debt accumulation yet. international banks. Over the past decade, more Similarities with the previous three waves than 20 EMDEs have accessed international capital markets for the rst time. In SSA, e fourth wave shares a number of features with Eurobond issuance has grown, with several earlier waves: a changing global nancial countries tapping the Eurobond market for the landscape, mounting vulnerabilities, and concerns rst time. about ine cient use of borrowed funds. Domestic debt has also become increasingly Financial landscape. As in the previous three important, with a rising share of local currency- waves, the current wave has seen changes in denominated bonds (Essl et al. 2019; Kose and 260 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 4.1 Similarities and differences between the previous three waves In each of the first three waves of broad-based debt accumulation, the initial runup in debt was facilitated by changes in financial markets, and low real interest rates in major advanced economies. These waves witnessed severe financial crises in EMDEs, usually triggered by external shocks and amplified by domestic vulnerabilities. They typically led to policy reforms in affected countries to lower external vulnerabilities and strengthen monetary and fiscal policy frameworks. The three waves differed in the composition of borrowers; the financial instruments involved; the speed of crisis resolution; and their macroeconomic impact. Introduction low or falling global interest rates, and major changes in financial markets, often in response to deregulation. These Since 1970, there have been four waves of EMDE debt enabled previously credit constrained borrowers to access accumulation, of which the fourth one is still underway international financial markets and accumulate debt. (see Kose et al. 2019 for a detailed discussion of each of Shortcomings in domestic policy frameworks often these waves). The first wave spanned the 1970-80s, with a contributed to rapid debt buildups, and exacerbated the rapid accumulation of debt by governments in LAC and severity of crises. SSA which led to a series of defaults in the early 1980s, and ended with debt relief and restructuring occurring in Low or falling global interest rates. The beginning of each the late 1980s-90s (LAC), and 1990s-2000s (SSA). The of the three waves was associated with low, or falling, second wave ran from 1990 until the early 2000s as global real interest rates, which encouraged borrowing financial and capital market liberalization enabled banks (Figure 4.1.1). In the first wave, the U.S. real policy rate and corporates in East Asia and the Pacific (EAP) and averaged around 0.6 percent over 1970-79, with several governments in Europe and Central Asia (ECA) to borrow years of negative real interest rates. During the second heavily; it ended with a series of crises in these regions in wave, the U.S. real policy rate declined from a high of 5 1997-2001. The third wave was a runup in private sector percent in 1989 to a low of 0.5 percent in 1993, as the borrowing in ECA from U.S. and EU-headquartered Federal Reserve cut policy rates in response to the global “mega-banks” after regulatory easing; this wave ended recession in 1991. Similarly, the U.S. real policy rate fell when the global financial crisis disrupted bank financing in into negative territory at the beginning of the third wave 2008-09 and tipped especially ECA countries into deep following the 2001 recession in the United States. (albeit short-lived) recessions. New financial instruments. The emergence of the This box synthesizes the main features of the three waves syndicated loan market in the 1970s set the stage for the that have by now concluded. In particular, it addresses the first wave. The introduction of Brady bonds in the 1990s following questions in detail. spurred the development of sovereign bond markets that underpinned sovereign borrowing in the second wave, • What were their similarities? while capital account liberalization in many EMDEs in the 1990s, especially in EAP, facilitated private sector • What were their differences? borrowing. The third wave in the 2000s largely consisted of cross-border flows via international banks in advanced Similarities economies after deregulation in the United States and the The first three waves of broad-based debt accumulation EU. featured several similarities. All of the waves had common Economic upturns. The beginnings of the first and second drivers, including changes in financial markets and low waves coincided with recoveries from global recessions interest rates. The waves also typically ended in crises with (1975, 1991, 2009) and the beginning of the third wave substantial macroeconomic impacts, which led to policy with the recovery from the global slowdown of 2001 (Kose changes. In part as a result of these policy changes, and Terrones 2015). countries weathered subsequent crises better. During the waves: Borrower country policies Beginning of the waves: Low global interest rates, changes in financial landscape Borrower country policies often encouraged rapid debt accumulation, or exacerbated the risks associated with it. The initial debt buildup in each wave was associated with Fixed exchange rate regimes and weak prudential frameworks encouraged risk taking; weak fiscal frameworks Note: is box was prepared by Peter Nagle. encouraged unfunded government spending; and G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 261 BOX 4.1 Similarities and differences between the previous three waves (continued) FIGURE 4.1.1 Comparison of previous waves The start of each wave generally coincided with a period of low, or falling, interest rates. The end of waves was also associated with a sharp slowdown in capital inflows, which restarted as new waves got underway. Debt episodes that ended in banking crises typically resulted in large increases in government debt. A. U.S. interest rates B. Capital flows to EMDEs C. Government debt during past banking crises Source: Bloomberg; International Monetary Fund; World Bank. A. Start of a wave defined as the first three years of the wave. Crisis defined as the year before, and year of, widespread crises. For the first wave, these are 1970-72, and 1981-82. For the second wave, these are 1990-92, and 1996-97. For the third wave, these are 2002-04, and 2008-09. For the final wave, the start is 2010-12, and the “latest” is the final two years of the sample, 2017-18. Real interest rates are calculated as the difference of nominal interest rates and the GDP deflator. B. Net capital inflows to EMDEs, in percent of GDP. The start of each wave is the first year, the peak is the peak capital inflow before the start of crises in the wave, and the trough is the lowest point after the crisis year. For the first wave, these dates are 1970, 1978, and 1988 respectively. For the second wave, they are 1990, 1995, and 2000. For the third wave, they are 2002, 2007, and 2009. The fourth wave begins in 2010 and the latest data are for 2018. C. “Before” and “after” denote, respectively, one year before and after the onset of banking crisis, as shown by numbers below the corresponding country names, taken from Laeven and Valencia (2018). Indonesia refers to central government debt only. Click here to download data and charts. government spending priorities or weak prudential Weak fiscal frameworks. In episodes of rapid government supervision directed funding to inefficient uses. debt accumulation, in LAC and SSA in the first wave and in ECA in the second wave, many countries ran persistent Fixed exchange rate regimes. During the first and second fiscal deficits financed with external debt. waves, especially, fixed or managed exchange rates in LAC, EAP and ECA encouraged capital inflows by leading Inefficient use of debt. While debt flows were often used lenders and borrowers to underestimate exchange rate to finance productive investment, in some cases debt was risks. With interest rates on foreign currency loans below used for domestic-facing investments, such as import those for domestic currency loans and the fixed exchange substitution industrialization that eroded competitiveness rate interpreted as an implicit guarantee of foreign in LAC in the first wave or construction and property exchange claims, borrowers readily took on foreign booms that did not raise export revenues in EAP and ECA currency-debt and domestic banks offered dollarized or in the second and third waves. Weak corporate euro-ized accounts on a large scale to local clients governance, including inadequate oversight of projects and (Impavido, Rudolph, and Ruggerone 2013; Magud, investment decisions as well as declining profitability, also Reinhart, and Rogoff 2011). led to inefficient investment in several EAP countries (Capulong et al. 2000). Weak prudential frameworks. Structural changes in financial markets were typically not accompanied by End of waves: Financial crises appropriate reforms to prudential or supervisory frameworks, allowing excessive risk-taking. In the second Rapid debt accumulation initially supported growth but wave, for example, rapid liberalization of capital markets was often associated with nancial crises. encouraged EAP banks to borrow heavily from Triggers. Financial crises have often been triggered by international markets (Furman et al. 1998). In the third shocks that raised investor risk aversion, risk premiums wave, the risks posed by growing cross-border lending and and borrowing costs, followed by a sudden stop of capital macro-financial linkages were underappreciated by ows, or by growth slowdowns that eroded debt financial supervisors (Briault et al. 2018; Claessens and sustainability (Frankel and Rose 1996; Easterly 2002; Kose 2018). Kaminsky and Reinhart 2000; Summers 2001). In the rst 262 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 4.1 Similarities and differences between the previous three waves (continued) wave, around the global recession of 1982, deteriorating implemented policies that helped build resilience to future risk sentiment restricted access to new borrowing in LAC nancial stress. In the rst and second waves, LAC and and SSA. In the second wave, capital ows to EMDEs EAP governments took measures to increase reserves and stalled or reversed in the global slowdown of 1998, amid a limit future buildups of external debt. Many moved loss of investor con dence following the East Asian and towards in ation targeting and exible exchange rates. In Russian crises (Kaminsky 2008; Kaminsky and Reinhart the second and third waves, EAP and ECA governments 2001). In the third wave, banking system liquidity dried eventually strengthened bank supervision, corporate up in the 2008 global nancial crisis, interrupting cross- bankruptcy laws and scal frameworks. However, progress border lending in ECA. Domestic political events also has varied across countries, with some remaining more contributed to some crises, for example in Turkey and vulnerable to shocks than others. Argentina in the third wave (Ozatay and Sak 2002). Differences across the waves Types of financial crises. Many crises began with sharp depreciations and capital out ows, which were The three waves differed in the most active borrowing occasionally the precursor to sovereign debt crises. Large sector and their regional focus; the financial instruments depreciations increased debt service on dollar-denominated involved; the speed of resolution of crises; and their debt and led to surges in in ation. Sudden stops or macroeconomic impact. reversals in capital ows complicated debt rollover. In all Borrowing sector and region three waves, countries that slid into crises had sizable vulnerabilities, such as large external, short-term foreign In the first wave, borrowing was primarily accounted for currency-denominated or variable-rate debt; low reserves; by the public sector in LAC and SSA (Figure 4.1.2).2 In pegged exchange rates; and weak monetary, scal, and these two regions, governments ran persistent fiscal deficits prudential frameworks. which were used to fund current expenditure in some countries, as well as investment. In the second wave, both Macroeconomic impact. Debt buildup in the rst three the private sector (EAP) and the public sector (ECA, LAC) waves was often followed by crises or stagnation, especially played a role. In the third wave—with fewer countries when the debt buildup was predominantly driven by with large debt runups than in the previous two waves— sovereign debt. Currency depreciations were often large, the private sector in ECA was the primary source of especially during the rst and second wave, and triggered borrowing. Sovereign debt levels in most EMDEs were sharp spikes in in ation and deteriorating debt-to-GDP either muted or falling in the third wave. Governments in ratios when debt was denominated in dollars. at said, EAP (second wave) and ECA (third wave) typically had there were considerable di erences in the severity of sound fiscal positions in the run-up to crises. As a result of macroeconomic outcomes between the waves, as discussed these shifts, the share of the public sector in external below. borrowing fell from a high of 95 percent in 1989 to 53 Fiscal impact. Financial crises were often scally costly. In percent in 2018. the rst wave, defaulting governments in LAC lost capital Financial instruments and debt resolution market access for many years. In the second and third waves, governments had to support ailing banks in Financial instruments. The source of credit in each wave recognition of implicit guarantees for nancial systems.1 also evolved. In the first wave, sovereigns borrowed from 90 percent of banking crises have required bank the official sector through bilateral lending and restructuring, and roughly 60 percent have led to the multilateral loans, as well as from commercial banks via nationalization of one or more banks. the syndicated loan market (lending from commercial banks accounted for around one-third of total external Policy responses. In all waves, the countries su ering crises 2 The first and third waves were global in the sense of total EMDE 1 For a global sample, the average cost of government intervention in debt rising whereas the second wave had a narrower regional focus. the financial sector during crises in 1990-2014 amounted to 9.7 percent During the first wave, EMDE government debt rose sharply; similarly, of GDP, with a maximum of 55 percent of GDP (IMF 2016a). The during the third wave, EMDE private debt rose sharply, driving up average cost of government intervention in public sector enterprises EMDE total debt (Figure 4.1). In contrast, during the second wave, during 1990–2014 amounted to about 3 percent of GDP and the average EMDE government debt declined while EMDE private debt, resulting in cost of the realization of contingent liabilities from public-private a limited overall increase in total EMDE debt over the course of the partnerships was 1.2 percent of GDP (Bova et al. 2016). second wave. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 263 BOX 4.1 Similarities and differences between the previous three waves (continued) FIGURE 4.1.2 Changes in debt by sector and region Whereas earlier waves were concentrated in a few regions, the debt buildup in the fourth wave has been broad-based. Like the third wave, private and government sectors accounted almost equally for external borrowing. A. Change in government EMDE debt, B. Change in private EMDE debt, C. Composition of external debt in by region by region EMDEs Source: World Bank. A.B. EAP = East Asia and Pacific; ECA = Europe and Central Asia; LAC = Latin America and Caribbean; MNA = Middle East and North Africa; SAR = South Asia; SSA = Sub-Saharan Africa. C. Long-term external debt only. Click here to download data and charts. public debt in EMDEs by 1980-81). The introduction of Turkey and Argentina, which required IMF Brady bonds in the 1990s spurred the development of assistance. Restructuring after Argentina’s 2001 debt sovereign bond markets, and in the 2000s, local bond default was not completed until many years later.4 markets deepened, allowing governments to obtain long- term finance, including from foreign investors. In the ECA • Faster private debt resolution. In the second wave, region, the private sector accessed cross-border lending by private sector debt in EAP was resolved quite quickly, European banks, whose subsidiaries and branches were with speedy support from the public sector through based in ECA countries but headquartered in advanced bank recapitalization and other support schemes, economies. As a result, there has been a shift from often with IMF assistance. Non-financial corporate international debt to domestic debt, and a move toward debt resolution, particularly among larger debt securities, including local currency bonds. conglomerates, was much slower than for the financial sector, and non-performing loans remained elevated Debt resolution: speed, scope, and mechanisms. The for several years after the crisis (Kawai 2002). In the speed of resolution largely depended on whether the third wave, globally accommodative policies; IMF debtors were in the public or private sector. The difficulty assistance; the European Bank Coordination of debt restructuring led to gradual progress in debt (“Vienna”) Initiative in 2009; and other banking resolution and restructuring mechanisms. system support together helped stem currency and banking crises. • Slow government debt restructuring. In the first wave, the resolution of widespread sovereign debt defaults in • New resolution mechanisms. At the start of the first LAC and SSA was slow, given Paris Club countries’ wave, there was little consideration for borrowers’ concerns about advanced economy bank solvency and ability to service their debt. Over time, creditors the lack of a well-defined restructuring mechanism moved toward acceptance of some debt reduction. (Callaghy 2002).3 In the second wave, debt resolution This paved the way for the conversion of syndicated was again prolonged for sovereign debt crises in 4 Argentina arranged a rst restructuring of its debt in 2005, which was 3Borensztein and Panizza (2009) find that the reputational and accepted by about three-quarters of bond holders (Hornbeck 2013). A economic cost of sovereign debt defaults is significant although short- second restructuring was agreed in 2010, which two-thirds of the lived, in part because crises precede defaults and defaults tend to happen remaining bondholders accepted. 7 percent of bondholders were at the trough of the recession. “holdout” creditors, who eventually reached a settlement in 2016. 264 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 4.1 Similarities and differences between the previous three waves (continued) FIGURE 4.1.3 GDP per capita in EMDEs during the four waves In the first wave of debt, countries in LAC and SSA saw prolonged stagnation in per capita growth after debt crises erupted. In the second wave, rapid growth in EAP was interrupted by the Asian financial crisis in 1998 but growth soon recovered. In the third wave, growth in ECA was robust throughout the period but fell in the final year when the crisis hit. A. First wave B. Second wave C. Third wave Source: World Bank. Note: Data are per capita GDP level (at 2010 prices and exchange rates) in each region at the pre-crisis peak and the end of the wave in each region, indexed to the start of the wave. For LAC and SSA in the first wave, the peak was in 1980; in EAP and ECA in the second wave it was in 1997; and in ECA in the third wave it was in 2008. The orange diamonds in Figures A-C show the average for all EMDEs excluding the highlighted regions in each chart, for the corresponding years. EAP = East Asia and Pacific, ECA = Europe and Central Asia; LAC = Latin America and the Caribbean; SSA = Sub-Saharan Africa. Click here to download data and charts. loans to Brady bonds, and later the HIPC and MDRI second wave also generated severe output losses. In debt relief initiatives for official debt in low-income contrast, in the second wave, EAP countries with countries. Collective action clauses (CACs) were later predominantly private debt buildups experienced only a introduced to facilitate sovereign debt restructuring temporary slowdown from the East Asia crisis. In the third with multiple bondholders (Eichengreen, Kletzer, and wave, ECA countries with predominantly private debt Mody 2003). For private debt, the Insolvency and buildups saw large but short-lived declines in output. Creditor Rights Standard developed best practices for national insolvency systems (Leroy and Grandolini Currency depreciations. Depreciations were substantially 2016). There has been a substantial improvement in larger and more common in the rst and second waves, insolvency protections over the course of the three when exchange rates were mostly xed or crawling pegs, waves (World Bank 2019a). and often had to be abandoned in the face of speculative attacks. By the third wave, more countries had exible Macroeconomic impact exchange rates, reducing the likelihood of substantial overvaluations to begin with. During the rst three waves, nancial crises did substantial economic damage, but the severity varied between the Inflation. Inflation following crises rose more in the first waves, and across regions. wave, and to a lesser extent, in the second. In part, this was due to larger depreciations in these waves. It also reflected Output cost. In the rst wave, LAC su ered a lost decade subsequent improvements in monetary frameworks—a of stagnant per-capita incomes following the 1982 crisis move toward inflation-targeting and independent central (Figure 4.1.3). Per capita incomes in SSA fared even banks that helped anchor inflation expectations (Ha, Kose, worse, with GDP per capita declining for many years. and Ohnsorge 2019). Sovereign debt crises in Turkey and Russia during the G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 265 Ohnsorge 2019; Turner 2002).5 Especially in the FIGURE 4.4 The fourth wave: Vulnerabilities and use of largest EMDEs, domestic bond issuance has risen borrowed funds rapidly. Foreign portfolio investors are also The fourth wave has seen growing vulnerabilities in EMDEs, with a rise in becoming more active in local bond markets, both domestic and external debt as countries have run persistent current accounting for a growing share of local currency- account and fiscal deficits. The composition of debt has shifted, with a greater share held by non-residents and a rise in non-concessional debt. denominated sovereign bonds. Public investment has fallen sharply in EMDEs, suggesting that rising debt is being used for current spending, rather than growth-enhancing e current wave has also seen a signi cant investment, despite a fall in interest payments. increase in nonbank nancial intermediation in EMDEs. ese nonbank nancial institutions A. EMDEs with current account B. EMDEs with fiscal deficits deficits have expanded rapidly in a number of EMDEs, particularly large economies. Vulnerabilities. Over the course of the fourth wave, vulnerabilities have once again grown (Ruch 2019). Since 2010, EMDE total external debt has risen to 26 percent of GDP on average in 2018, re ecting sizable and persistent current account de cits. In 2018, 55 percent of EMDEs had weaker current account balances than in 2010; 76 percent ran current account de cits (compared C. Average maturity and D. Non-resident share of government non-concessional debt in EMDEs debt, foreign currency share of with 69 percent in 2010); and 44 percent had corporate debt current account de cits in excess of 5 percent of GDP (Figure 4.4). e number of countries with scal de cits has also risen. In addition, both government and private debt have shifted toward riskier forms in many EMDEs, with a rise in the share of debt that is held by non-residents (for governments), is denominated in foreign currency (for corporates) and is on non-concessional terms. A greater share E. Public expenditures in EMDEs F. Cumulative change in house prices, of corporate debt than before the global nancial selected country groups crisis is held by rms with riskier nancial pro les, as supportive nancing conditions have allowed rms to issue more debt with weaker credit quality (Beltran and Collins 2018; Feyen et al. 2017; IMF 2015a). EMDE nancial markets are now more tightly integrated into the global nancial system, which could in some circumstances facilitate the contagion of global nancial shocks both to foreign currency and, to a lesser extent, local Source: Bank for International Settlements; International Monetary Fund; OECD; World Bank. currency debt markets. C. Median of 65 EMDEs for maturity and 122 EMDEs for non-concessional debt D. Non-resident share of government debt is average for 45 EMDEs, with a smaller sample size for earlier years. Foreign currency share of corporate debt of average for 21 EMDEs. F. Chart shows the cumulative percentage increase in house prices over the course of a wave, prior to the crisis. The range covers 1990-97 for EAP, 2001-2008 for ECA, and 2010-18 for EMDEs. EAP contains three countries, ECA contains 5, and EMDEs contains 31 countries. Orange diamonds denoted the median, and blue bars the interquartile range of country groups. Click here to download data and charts. 5 However, such a switch may bring other risks, as countries switching from external to domestic debt could be trading a currency mismatch for a maturity mismatch (Panizza 2008; Broner, Lorenzoni and Schmukler 2013). Nominal interest rates on domestic debt tend to be higher than on external debt (IMF 2015a). 266 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 4.2 The fourth wave Since 2010, another wave of debt accumulation has been building and total debt in EMDEs has reached almost 170 percent of GDP, on average, in 2018—a record high—from 114 percent of GDP in 2010. This increase was accompanied by shifts toward borrowing from non-traditional creditors and financial institutions, as well as capital markets. As with previous waves, the fourth wave has seen mounting vulnerabilities for EMDEs. The fourth wave of debt buildup among EMDEs began in period government debt-to-GDP has risen in three- 2010. It was broad-based across EMDE regions and quarters of EMDEs and by at least 10 percentage borrowing sectors. The debt buildup has been points in almost 60 percent of them. Government accompanied by a decade of anemic growth in EMDEs debt saw a marked increase among commodity- (Kose and Ohnsorge 2019). Changes in advanced- exporting countries in the aftermath of the economy financial sectors also propelled shifts in creditors commodity price plunge in 2014 (particularly oil to EMDE governments and corporates. This box examines prices), as fiscal deficits surged amid declining revenue the fourth wave by addressing the following questions. and large fiscal stimulus (World Bank 2018c). • How did debt evolve in the fourth wave? • Private debt. The private sector has also rapidly accumulated debt since the global financial crisis, • Which factors have contributed to debt accumulation particularly in China. About two-fifths of EMDEs during the fourth wave? witnessed private sector credit booms in at least one year during 2011-18 (Ohnsorge and Yu 2016; World Evolution of debt Bank 2016a).2 The rise in debt in China has been focused in a few sectors, notably the real estate, Broad-based public and private debt buildup. Since 2010, mining, and construction sectors, and among state- another wave of debt accumulation has been underway. owned enterprises. The buildup has been especially fast in EMDEs, with government debt increasing in tandem with mounting Shifts to riskier debt. Both government and private debt private sector debt. As a result, total debt in EMDEs has have shifted toward riskier funding sources in many risen to almost 170 percent of GDP, on average, in EMDEs, making these countries more vulnerable to a 2018—a record high—from 114 percent of GDP in 2010 deterioration in global investor sentiment (Figure 4.2.1). (Kose et al. 2019). The debt-to-GDP ratio has risen in all EMDE regions with the exception of SAR, where it has • Government debt. The increase in government debt been broadly flat, and in 80 percent of EMDEs, with more has been accompanied by a growing share of than one-third seeing an increase of at least 20 percentage non-resident investors (to 43 percent in 2018) and an points of GDP.1 Excluding China, where corporate debt increasing reliance on non-concessional terms. has soared post-crisis, total EMDE debt has risen to a Sovereign ratings have also been downgraded for near-record 107 percent of GDP in 2018. The pace of many EMDEs since 2010. This also increases the increase in EMDE debt excluding China has slowed since fragility of EMDE banks where there is some evidence 2016, with a modest decrease in private sector debt that exposures to sovereigns have increased (Feyen offsetting a small increase in government debt. However, and Zuccardi 2019). this masks substantial variation between regions, with large increases in debt-to-GDP ratios in SSA and LAC and • Private debt. On average, across EMDEs with declines in MNA and ECA. available data, foreign currency-denominated corporate debt has risen from 19 percent of GDP in • Government debt. Since 2010, EMDE government 2010 to 26 percent of GDP in 2018, although its debt has risen, on average, by 12 percentage points of share of total corporate debt remained around 40 GDP to 50 percent of GDP at end-2018. Over this percent over this period (IIF 2019b). By end-2018, one third of these EMDEs had foreign currency denominated corporate debt above 20 percent of Note: This box was prepared by Peter Nagle. GDP. In addition, a greater share of corporate debt 1 Total debt has risen particularly rapidly in Argentina, Cambodia, Chile, and China. Turkey stands out as having the third fastest increase in private sector debt after Cambodia and China. Among low-income 2 About half of all credit booms are followed by at least a mild countries, Mozambique, The Gambia, and Togo and have seen the largest increases in debt. deleveraging within three years (Ohnsorge and Shu 2016). G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 267 BOX 4.2 The fourth wave (continued) FIGURE 4.2.1 The fourth wave: Debt developments Low-income countries have seen a sharp increase in borrowing from non-Paris club bilateral sovereign lending and non- concessional lending. As EU- and U.S.-headquartered banks have downsized their EMDE operations, cross-border bank lending to EMDEs shifted to EMDE-headquartered banks. EMDE corporate and sovereign borrowers have increasingly turned to capital markets to raise new debt. A. Share of non-concessional debt in B. Creditor composition of LIC external C. Pan-regional banks LICS public debt D. Global assets of 10 largest G-SIBs by E. Debt securities outstanding F. Claims on the official sector bank domicile Source: Bank for International Settlements; Claessens and van Horen 2014; International Monetary Fund; World Bank. A. Dashed blue lines denote the interquartile range, while solid blue line is the median. Includes 30 low-income countries and excludes Somalia, South Sudan, and Syria due to data restrictions. B. GDP-weighted average across 32 low-income countries. Bilateral includes public and publicly guaranteed (PPG) loans from governments and their agencies (including central banks), loans from autonomous bodies, and direct loans from official export credit agencies. Multilateral includes PPG loans and credits from the World Bank, regional development banks, and other multilateral and intergovernmental agencies. It excludes loans from funds administered by an international organization on behalf of a single donor government. Private include PPG bonds that are either publicly issued or privately placed; PPG debt from commercial bank loans from private banks and other private financial institutions; as well as export and supplier credits. C. GFC = global financial crisis. Based on annual bank statements; before the GFC = 2008 or 2009 depending on data availability; after GFC = 2018 or latest data available. D. Based on the Financial Stability Board 2018 list of global systemically important banks (G-SIBs). E. Sample includes Argentina, Brazil, Colombia, India, Indonesia, Malaysia, Mexico, Philippines, Russia, South Africa, Thailand and Turkey. F. BIS estimates of the claims by foreign banks on official sector: sample includes Argentina, Brazil, Chile, Colombia, Hungary, India, Indonesia, Israel, Malaysia, Mexico, Poland, Russia, Thailand, Turkey, Republic of Korea, and South Africa Click here to download data and charts. than before the global financial crisis has been owed creditors, notably China, as well as commercial creditors by firms with riskier financial profiles, as supportive over the past decade (World Bank 2018b; World Bank financing conditions have allowed firms to issue more and IMF 2018a). In 2016, non-Paris Club debt accounted debt with weaker credit quality (Beltran and Collins for more than a fth of the median LIC’s external debt, 2018; Feyen et al. 2017). and about 13 percent of their public debt, raising concerns about debt transparency as well as debt collateralization LIC government debt. In LICs, debt has also shifted (Essl et al. 2019). toward non-concessional, non-Paris Club bilateral 268 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 4.2 The fourth wave (continued) Estimates of current public debt levels in LICs also su er European banks withdrew.3 LAC was an exception, with a from limited debt transparency, including issues related to growing role of domestic banks, rather than of banks based contingent liabilities, state-owned enterprise debt and PPP in other countries in the region, as domestic banks transactions, and the assets held by LIC governments. acquired assets from exiting foreign lenders. e regional These data limitations are especially acute for debt owed to expansion of EMDE banks has yet to reach the scale of commercial and non-Paris Club creditors. Poor data pre-crisis cross-border activity of lenders from the coverage can give rise to unexpected sudden increases in advanced economies. debt, for example when the debt of loss-making SOEs migrates onto the books of the central government. For Finally, the domestic institutional investor base has example, in Mozambique and the Republic of Congo, the continued to grow in EMDEs, o ering the prospect of a revelation of unreported debt led to large upward revisions potentially stabilizing pool of domestic savings. Assets of to official debt figures, which resulted in debt distress pension funds and insurance companies had risen to 46 (IMF 2018a). Only a third of the 59 countries eligible for percent of GDP by end-2016, on average, in EMDEs. International Development Association borrowing report Such assets remain equivalent to only about half of the private sector external debt statistics (World Bank and assets of the bank and non-bank nancial system (World IMF 2018b). Bank 2019c).4 Changes in the composition of creditors. Since the global Contributing factors to debt accumulation financial crisis, borrowing by EMDEs has shifted toward Evolving financial instruments. e latest wave has been capital markets and regional banks, and away from global associated with a growing importance of domestic debt, banks. Bond issuance has allowed firms to access finance while external debt grew more slowly than in the most when bank credit supply tightened or at different terms a ected regions during previous waves. e fourth wave from bank loans (Cortina, Didier, and Schmukler 2016). has seen rising demand for EMDE bonds from The role of regional EMDE banks has also grown as large international investors such as asset managers (Shin 2014). international banks have retrenched from EMDEs in the Domestic bond issuance has risen sharply, particularly in aftermath of the global financial crisis (BIS 2018; Feyen large EMDEs, while exceptionally long-term (50- and and Gonzalez de Mazo 2013). As large international banks 100-year) international bonds have been issued by some retrenched, cross-border bank lending to EMDEs shifted EMDEs, including Mexico in 2010, and Argentina in to EMDE-headquartered banks, which greatly expanded 2017. Over the past decade, more than 20 EMDEs their regional presence, most notably in SSA (Cerutti and accessed international capital markets for the rst time. Zhou 2017, 2018; IMF 2015b; World Bank 2018c). New frontier market bond indices, such as J.P. Morgan’s NEXGEM launched in 2011 or MSCI’s Frontier Market Chinese banks accounted for two-thirds of EMDE-to- Index launched in 2007, have facilitated international EMDE lending between 2013 and 2017 and for most of capital market access and broadened the investor base for the doubling in cross-border claims on SSA economies in countries which thus far only had intermittent capital the same period, to over 10 percent of GDP on average market access. (Cerutti, Koch, and Pradham 2018; Dollar 2016). Other EMDE banks have also increased their presence in EMDEs within their respective regions. A notable exception has been the Middle East and North Africa region, where declining current account surpluses resulting 3 For example, example, Russia’s largest lender, Sberbank, acquired from weaker oil revenues have reduced the region’s ability Volksbanken’s VBI Eastern European operations in 2012. to recirculate savings from high-income oil exporters to 4 Data on assets of pension funds and insurance companies are only lower-income EMDEs with persistent current account available for 22 EMDEs. Foreign institutional investors’ role in EMDE financial markets has also grown but in some sectors remains small. For deficits (World Bank 2019b). example, in just under 1000 infrastructure projects since 2011, the share of institutional investors has more than tripled but still accounts for only In SSA, banks headquartered in Togo, Nigeria and South 0.7 percent of the average project value (World Bank 2018a). Some Africa have expanded rapidly to other EMDEs in the institutional investors in EMDEs have been shown to behave procyclically, leaving EMDE financial markets during times of stress region (Arizala et al. 2018). In ECA, Russian banks rather than acting as stabilizing investors with deep pockets (Raddatz and initially expanded post-crisis within the region, as Western Schmukler 2012). G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 269 BOX 4.2 The fourth wave (continued) The share of corporate debt financed by debt securities on currency bonds issued by Poland, and Hungary—currently average rose from 16 percent to 25 percent of total lending trade at negative yields.7 Spreads on emerging market debt between end-2007 and end-2018. This included issuance both for corporate and sovereign bonds reached all-time on both international and domestic debt markets. The lows in 2017, boosting borrowing. Average spreads on volume of international debt securities issued by EMDEs corporate bond issuance have fallen for all EMDEs, increased more than three times between 2007 and 2018. including LICs. Spreads have also fallen for lower rated Domestic debt issuance excluding China increased from corporate bonds. 33 percent of GDP in 2007 to 47 percent of GDP in 2018. An additional reason for rapid debt accumulation has been a sharp slowdown in growth over the course of the fourth EMDE sovereign borrowers are also relying more heavily wave that eroded EMDE scal positions and resulted in on capital markets. From 2007 to 2017, debt securities additional borrowing to maintain current spending levels. issued by EMDE governments increased by 4.4 percentage Government debt levels in commodity exporters surged points of GDP on average, to 22 percent of GDP. In SSA following the collapse in commodity prices, particularly Eurobond issuance has grown, with several countries after the oil price plunge in 2014, driving much of the tapping this market for the rst time. Sovereign debt increase in EMDE debt (excluding China) in the second issuance has grown particularly rapidly in domestic bond half of the current wave (World Bank 2018a). markets, especially in EAP (G20 IFAWG 2018). In some EMDEs, the share of nonresident investors in local Growing non-bank financial intermediation. e current currency sovereign bond holdings exceeds 30 percent, wave has also seen a signi cant increase in shadow banking which makes these economies more vulnerable to sudden activities in EMDEs. Shadow banking refers to non-bank shifts in investor con dence (G20 IFAWG 2018). nancial intermediation that takes place outside of the regulated nancial system and may provide credit to riskier New nancing vehicles such as infrastructure bonds and borrowers who often lack access to bank credit. Shadow green nance bonds have stimulated lending to speci c banking systems, which were small before the global EMDE sectors where banks used to be the primary source recession, have expanded rapidly in a number of EMDEs, of funding (FSB 2018a; McKinsey Global Institute particularly in large economies such as China and India 2018).5 However, infrastructure nancing, in general, has (IMF 2014). In these two countries, assets of non-bank declined in EMDEs following the sharp reduction in nancial institutions now represent over a third of total cross-border lending and stricter post-crisis regulations in nancial system assets. In China alone, this share has more the nancial sector (G20 2013; Kose and Ohnsorge than doubled over the last decade, and the size and 2019).6 complexity of its non-bank nancial sector is becoming comparable to those of advanced economies (Ehlers et al. Very low interest rates, weak growth. Interest rates have 2018). been at very low levels throughout the fourth wave as a result of unconventional monetary policy among central A decade of lighter regulation of non-banks than banks, banks, including negative policy rates and quantitative combined with rapid growth, has increased maturity easing. is has encouraged an aggressive search for yield, mismatches and credit risks in non-banks (IMF 2018c). large capital ows to EMDEs, and a sharp fall in bond Financial stress in non-banks may quickly propagate to the spreads. Around one quarter of sovereign and corporate rest of the nancial system, owing to its bonds in advanced economies—and some foreign- interconnectedness with banks (FSB 2017, 2018b, 2019; Pozsar et al. 2013). is has been illustrated by a recent shift toward stricter regulations and supervision of non- 5 In advanced economies, financial instruments that were widely used banks in China and a default of one of the largest non- before the crisis have regained popularity. Especially in the United States, bank lenders in India, which have already created tighter leveraged loan issuances—the majority of which are now covenant-lite nancial conditions for the private sector in those with lesser protections for creditors, and which are predominantly held in Collateralized Loan Obligations (CLOs) and loan funds—have risen economies (IMF 2019b). again above elevated pre-crisis levels. Concerns have been raised whether CLO prices are fully aligned with risks (Domanski 2018; FSB 2019). 6 Grants and concessional loans are the primary source of 7 In the two EMDEs with negative yielding sovereign bond issuances, infrastructure finance in LICs, with bank lending providing a government, household and corporate debt have risen only marginally (at complementary source of funding only in a small number of countries most 7 percentage points of GDP) over the past decade. (Gurara et al. 2017). 270 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 4.5 Comparison of features of fourth wave and private debt has nanced residential construction, earlier waves: Debt which does not yield export earnings. The fourth wave has seen the largest and fastest increase in debt-to-GDP ratios among EMDEs. It has also been the most broad-based increase in Differences from the previous three waves debt across regions and borrowing sectors. e fourth wave has featured the largest, fastest A. Change in total debt B. Annual average change in total and most broad-based debt accumulation in debt EMDEs yet. In contrast to earlier waves, government debt has risen in tandem with mounting private sector debt. Compared to the rst and third waves—when advanced-economy debt accumulation outpaced EMDE debt accumulation—the fourth wave has been accompanied by near-stable advanced-economy debt-to-GDP ratios. However, some other developments have been more reassuring. During the latest wave, there have been reforms that have C. Share of economies with increase D. Share of economies with increase in government debt, by region in private debt, by region made the international nancial system more resilient and enlarged the global nancial safety net. Many EMDEs have improved their macroeconomic and prudential policy frameworks over the past two decades. Largest, fastest, and most broad-based wave yet. Including or excluding China, the annual increase in EMDE debt since 2010 (almost 7 percentage points of GDP, on average) has been larger, by Source: . International Monetary Fund; World Bank. Note: First wave: 1970-89; second wave: 1990-2001; third wave: 2002-09; fourth wave: 2010 some margin, than during the rst three waves onwards. A. Change in total debt-to-GDP ratio over the source of each wave. (Figure 4.5). In contrast to previous waves, which B. Average annual change calculated as total increase in debt-to-GDP ratio over the duration of the were largely regional in nature, the fourth wave wave, divided by the number of years in a wave. C.D. Sample includes 142 EMDEs. Data show the share of economies where the debt-to-GDP ratio was global. Total debt has risen in more than 70 increased over the duration of the wave. Regions are excluded if country-level data are available for less than one-third of the full region. percent of EMDEs in all regions—previous waves Click here to download data and charts. saw higher rates of increase within speci c regions, but not across all regions simultaneously. More than one-third of EMDEs have seen an increase in debt of at least 20 percentage points of GDP. Use of debt. In the current wave of debt, there Finally, the majority of debt accumulation have been signs that government debt is being episodes have featured combined government and used for “less e cient” spending rather than on private debt buildups—in contrast to the previous productive investment in physical or human three waves when the majority of debt capital that could boost potential growth in accumulation episodes were either predominantly EMDEs. Public investment in EMDEs fell from government or predominantly private episodes. an average of 2.1 percent of GDP in 2002-09, to 0.9 percent in 2010-18 (IMF 2019c). Among Stronger policy frameworks. Many EMDEs learnt commodity exporters, declining tax revenues the lessons from crises in the previous waves and following the commodity price plunge in 2014-16 adopted reforms designed to improve resilience. widened scal de cits and raised debt despite ese include greater exchange rate exibility, and lower investment (World Bank 2018a). more robust monetary policy frameworks and Meanwhile, house prices have risen sharply in central bank transparency—since 1999, the some EMDEs, suggesting that some of the rise in number of EMDEs who have adopted in ation G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 271 targeting has increased from 3 to 24 (Figure 4.6). FIGURE 4.6 Comparison of fourth wave and earlier EMDEs have also made reforms to scal waves: Policies and institutions frameworks, with the number of countries with Many EMDEs learned lessons from crises in the previous waves and scal rules rising from 12 in 1999 to 62 in 2018, adopted policies to improve resilience. These include more robust monetary and exchange rate policy frameworks, fiscal rules, and substantial improvements in debt macroprudential tools, higher foreign exchange reserves relative to management policies and tools (World Bank external debt, and improved bankruptcy processes. 2013). Foreign exchange reserves to debt have risen markedly across EMDE regions, although A. EMDEs with inflation targeting B. EMDEs with fiscal rules they have fallen from the highs of 2009-10. More central banks EMDEs are using macroprudential tools, particularly placing stricter limits on foreign exchange positions. Bankruptcy rights have also been strengthened, but there is still considerable room for improvement (Kose and Ohnsorge 2019). Financial regulatory reforms. Financial sector reforms implemented since the global nancial crisis are also increasing resilience (BIS 2018). e C. Foreign reserves in EMDE regions D. EMDEs with flexible exchange rates G20 global nancial regulatory reform agenda has implemented major nancial reforms since the global nancial crisis, including the international adoption of the Basel III capital and liquidity standards (FSB 2018c). Global nancial safety nets have been signi cantly expanded, with resources available in country- speci c, regional and multilateral nancial safety nets tripling between 2007 and 2016, including E. Macroprudential policy in EMDEs F. Bankruptcy rights protection in through the creation of regional nancing EMDEs arrangements (RFAs), expanded IMF resources, and increased international reserve holdings (IMF 2018c). 6 Stable debt in advanced economies. In contrast to the rst and third waves—when advanced- economy debt accumulation outpaced EMDE debt accumulation—the fourth wave of EMDE debt accumulation was accompanied by near- Source: Cerutti, Claessens, and Laeven (2017); Dincer and Eichengreen (2014); Ha, Kose, and stable advanced-economy debt-to-GDP ratios. Ohnsorge (2019); Huidrom et al. (2019); International Monetary Fund; Kose et al. (2017); World Bank. Advanced economies have also seen pronounced A. Inflation targeting as classified in the International Monetary Fund’s Annual Report of Exchange Arrangements and Exchange Restrictions. private-sector deleveraging which reduced the B. An economy is considered to be implementing a fiscal rule if it has one or more fiscal rules on expenditure, revenue, budget balance, or debt. share of private debt in total debt during the D. An economy is considered to have a flexible exchange rate if it is classified as “Floating” or “Free Floating” in the International Monetary Fund’s Annual Report of Exchange Arrangements and fourth wave. Exchange Restrictions. E. Sample includes 123 EMDEs. Unweighted average of the Macroprudential Policy Index of Cerutti, Claessens, and Laeven (2017). The Macroprudential Policy Index measures the number of tools used by authorities and is based on a simple sum of up to 12 including, but not limited to, countercyclical capital buffers and loan-to-value ratios. F. Distance to frontier score for strength of insolvency resolution. A higher index indicates reforms 6 The global financial safety net consists of 1) self-insurance that improve the business climate. EAP, ECA, LAC, MNA, SAR, and SSA include 22, 22, 32, 19, 8, against external shocks using foreign reserves or fiscal space at and 46 economies, respectively. Advanced economies include 36 economies. Based on World Bank Doing Business reports for 2010, and 2019. national level, 2) bilateral are swap lines among countries, 3) regional Click here to download data and charts. financing arrangements, and 4) the global financial backstop provided by the IMF (Brueggemann et al. 2018). 272 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 4.3 Debt and crises This box conducts an econometric exercise to illustrate the extent to which debt accumulation can increase the likelihood of a crisis. A substantial rise in either government or private debt is associated with a significantly higher probability of a crisis occurring in the following year. A combined increase in government and private debt had a particularly strong association with the probability of a currency crisis in the next year. A high share of short-term debt, or large debt servicing costs, similarly raised the likelihood of a crisis. Countries that experienced crises typically had major institutional shortcomings, including debt and fiscal mismanagement, inadequate banking regulation, poor corporate governance, and political uncertainty. The event study suggests that episodes of debt and several control variables in a panel logit model with accumulation that were accompanied by crises often random effects for a sample of 139 EMDEs over 1970- featured larger debt buildups than episodes without crises. 2018 (Annex 4.2). All explanatory variables are lagged This box quantifies the effect of debt accumulation on the because the focus is on pre-conditions that make crises likelihood of financial crises using an econometric analysis. more likely. In addition, the use of lagged variables Specifically, it answers the following questions. attenuates potential endogeneity bias caused by contemporaneous interactions between economic • What factors have been found to correlate with fundamentals and crises. Regressions are estimated financial crises? separately for sovereign debt crises, currency crises and banking crises since these are likely to be associated with • What factors are associated with an increased different sectoral vulnerabilities. likelihood of crises? The correlates of crises are drawn from a rich empirical • What were the common features of crisis episodes? literature on the determinants of financial crises, or of the Empirical literature vulnerabilities that worsen the impact of crises. This literature has identified the following correlates of higher The econometric exercise here builds on an extensive crisis probabilities: literature on early warning models.1 The first generation of early warning models, in the 1980s and 1990s, aimed at • Factors that increase rollover risk. These are particularly predicting currency crises and largely focused on relevant during periods of financial stress; the include macroeconomic and financial imbalances. Measures of high short-term external debt and high or rapidly balance sheet health became more prominent in such growing total, government, or private debt; current models after the Asian financial crisis, especially in account deficits; predicting banking crises. A combination of government • Factors that restrict policy room to respond. These solvency and liquidity indicators have also been used in include low international reserves; large fiscal and studies of sovereign debt crises. current account deficits; and weak institutions. Debt accumulation and financial crises: • Factors that suggest overvaluation of assets. These An econometric analysis indicate potential for large asset price corrections; the Econometric specification. In the baseline regression include exchange rate misalignment and credit or asset specification, the probability of a financial crisis is price booms. estimated as a function of the pace of debt accumulation Of these potential correlates, the regression model identi es several that are statistically signi cant and robust correlates of the probability of nancial crises.2 ese Note: This box was prepared by Wee Chian Koh and Peter Nagle, include external vulnerabilities (higher short-term debt, with contributions from Jongrim Ha, Alain Kabundi, Sergiy Kasyanenko, Wee Chian Koh, Franz Ulrich Ruch, Lei (Sandy) Ye, and Shu Yu. 1 See Berg, Borensztein, and Patillo (2005); Chamon and Crowe 2 Annex 4.1 lists the variables used in the baseline model and presents (2012); Frankel and Saravelos (2012); Kaminsky, Lizondo, and Reinhart (1998) for extensive reviews of the literature on early warning models. a number of robustness tests; for example, for alternative model For models involving currency crises, see Eichengreen, Rose, and specifications (random effects probit model) and twin crises. Twin crises Wyplosz (1995); Frankel and Rose (1996); and Kaminsky and Reinhart are defined as the simultaneous occurrence of any two types of financial (2000). For models involving banking crises, see Borio and Lowe (2002); crises (sovereign debt, banking, or currency). Such episodes are usually Demirgüç-Kunt and Detragiache (1998); and Rose and Spiegel (2012). associated with much larger changes in typical leading indicators. The For models involving debt crises, see Dawood, Horsewood, and Strobel correlates in the baseline model indeed have higher statistical significance (2017) and Manasse, Roubini, and Schimmelpfenning (2003). in predicting twin crises than individual crises. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 273 BOX 4.3 Debt and crises (continued) higher debt service, lower international reserves), adverse averaged -1 percent. Contractions of this magnitude shocks (higher U.S. interest rates, lower domestic output increased the probability of entering a sovereign debt crisis growth), and faster debt accumulation—especially if true in the subsequent year to 1.9 percent from 1.2 percent of both government and private debt.3 ese ndings are outside crisis episodes. A 2-percentage point increase in broadly consistent with the existing literature on leading U.S. real interest rates—half the cumulative increase indicators of nancial crises, particularly with regard to the during a typical tightening phase of U.S. monetary important role of the composition of debt and pace of debt policy—increased the probability of entering a currency accumulation.4 In addition, the regressions here suggest crisis by one-half to 6.0 percent from 4.1 percent. that combined private and government debt buildups signi cantly increase the probability of a currency crisis. External vulnerabilities. A larger share of short-term debt in external debt, greater debt service cost and lower reserve Debt accumulation. An increase in debt, either cover were associated with signi cantly higher probabilities government or private, was associated with signi cantly of nancial crises. higher probabilities of crises in the following year. For example, an increase of 30 percentage points of GDP in • Short-term debt. Compared to the probability of a government debt over the previous year (equivalent to the sovereign debt crisis of 1.2 percent associated with a median buildup during a government debt accumulation share of short-term debt of 10 percent of external debt episode) increased the probability of entering a debt crisis (the average during non-crisis episodes), a 30 percent to 2.0 percent (from 1.4 percent) and that of entering a share of short-term debt in external debt (Mexico’s currency crisis to 6.6 percent (from 4.1 percent). For share before it plunged into a twin currency and debt private debt, a 15 percentage points of GDP increase in crisis in 1982) raised the probability of entering a debt (equivalent to the median increase during a private sovereign debt crisis in the following year to 2.0 debt accumulation episode) doubled the probability of percent. entering a banking crisis, to 4.8 percent, or a currency crisis, to 7.5 percent, in the following year—probabilities • Debt service. A 50 percent ratio of debt service to that are considerably larger than those for a similarly-sized exports—Mexico’s average debt service burden in the buildup in government debt. early 1980s—was associated with probabilities of entering a sovereign debt crisis of 2.8 percent and a Combined government and private debt increase. banking crisis of5.5 percent. is was more than Simultaneous increases in both government and private double the probabilities associated with a 15 percent debt ampli ed the probability of a currency crisis. us, a debt service-to-export ratio in the average non-crisis 15 percentage points of GDP increase together with a 30 episode. percentage point of GDP increase in government debt resulted in a 24 percent probability of entering a currency • Reserve coverage. e probability of a debt or banking crisis the next year—more than six times the probability crisis exceeded 3 percent, and that of a currency crisis had debt remained stable (3.9 percent) and about one- 5 percent, for a reserve coverage of 1 month of third more than similarly-sized government or private debt imports (which was the case in Mexico in the early buildup separately. 1980s) compared to probabilities of 0.6-2.0 percent for banking and debt crises, and 3.8 percent for Adverse shocks. Compared to average growth outside currency crises, when the reserve coverage amounted crises (4 percent), growth in EMDE crisis episodes to 4 months of imports (the average for non-crisis episodes). 3 The same variables remain statistically significant in a regression that combines sovereign debt and banking crises, but the change in Other vulnerabilities. Other vulnerabilities identi ed government debt becomes insignificant. This may reflect the fact that tended to be more speci c to certain types of crises or banking crises have been more than twice as common as sovereign debt borrowing sectors. crises since 1970. Since almost all crises in the sample are associated with debt accumulation episodes, dummy variables indicating the presence of a private or government or combined (private and government) debt • Wholesale funding. Higher wholesale funding, proxied accumulation episode are not statistically significant. by the ratio of credit to deposits, was associated with a 4 Relevant empirical regularities are discussed in, for example, greater probability of a banking crisis but appears to Manasse, Roubini, and Schimmelpfenning (2003) on sovereign debt crises; Kauko (2014) on banking crises; and Kaminsky, Lizondo, and have been largely unrelated to the probabilities of Reinhart (1998) on currency crises. sovereign debt and currency crises. 274 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 4.3 Debt and crises (continued) • Real exchange rate overvaluation. Real exchange rate Macroeconomic policies overvaluation was associated with a higher probability of a currency crisis but tended to be largely unrelated Inefficient use of debt. In addition to financing import to banking and sovereign debt crises (Dornbusch et al. substitution policies, public debt was used in some 1995). countries in the first wave to finance current government spending and policies that led to overly expansionary • Concessional debt and FDI ows. A higher share of macroeconomic policies (Argentina, Brazil, Chile, Peru). concessional debt, which consists of loans extended on In other countries, rapid private-sector borrowing resulted more generous terms than commercial ones, was in debt-fueled domestic demand booms, including associated with a lower probability of a sovereign debt property booms (Thailand, Ukraine) or inefficient crisis but tended to be largely unrelated to banking manufacturing investment (Korea). and currency crises. Larger FDI in ows, a more stable form of nance than portfolio in ows, were associated Inadequate fiscal management. Many countries had severe with a lower probability of a currency crisis. fiscal weaknesses. This included weak revenue collection (Argentina, Brazil, Indonesia, Russia), widespread tax Crisis probabilities: Small or large? In isolation, some of evasion (Argentina, Russia), public wage and pension these probabilities may appear small. is is expected since indexing (Argentina, Brazil, Mexico, Uruguay), monetary they are associated with individual indicators. However, financing of fiscal deficits (Argentina, Brazil), and the probabilities could cumulate rapidly when multiple substantial use of energy and food subsidies (Egypt, indicators deteriorate at the same time as has frequently Venezuela). happened prior to nancial crises. Indeed, as documented in the previous chapters, in a typical nancial crisis, an Risky composition of debt. Many of the crisis countries adverse shock is often compounded by elevated debt and borrowed in foreign currency. They struggled to meet debt multiple other vulnerabilities. service obligations and faced steep jumps in debt ratios following currency depreciations (Indonesia, Mexico, Lessons from financial crisis episodes Thailand). In Uruguay, for example, almost all public debt The preceding section quantified how shocks and was denominated in U.S. dollars in the mid-1990s. Several vulnerabilities have affected the likelihood of crises. In countries relied on short-term borrowing and faced addition, beyond measures that can be easily quantified, rollover difficulties when investor sentiment deteriorated countries with financial crises during or after a debt (Indonesia, Korea, Philippines, Russia in the late 1990s). accumulation episode shared some structural and In Europe and Central Asia (ECA) in the 2000s, countries institutional weaknesses that made their economies more borrowed cross-border from nonresident lenders and faced prone to crises once an adverse shock hit. These structural a credit crunch once liquidity conditions tightened for and institutional weaknesses are explored in this section in global banks that were the source of this lending a set of selected country case studies of financial crises. (Hungary, Kazakhstan in the late 2000s). These case studies look into 43 crisis episodes in 34 Balance sheet mismatches. A substantial number of EMDEs that have witnessed rapid government or private currency and banking crises, and the majority of debt accumulation episodes since 1970 (Annex 4.3). Most concurrent currency and banking crises, were associated of these cases (65 percent) involved overlapping private with balance sheet mismatches (Indonesia, Malaysia, and government debt accumulation episodes. Almost all Mexico, and Russia in the late 1990s). Sovereign debt cases (90 percent) involved twin crises, and 40 percent crises less frequently involved balance sheet mismatches, involved triplet crises.5 except when banking supervision was weak (Indonesia, Turkey in the 1990s). 5 The main references for these country case studies are provided in Structural and institutional features Kose et al.(2019). For a discussion of some of these macroeconomic, structural and institutional shortcomings see Balassa (1982); Kaufmann Poorly designed growth strategies. Many of the case (1989); and Sachs (1985, 1989) on growth strategies and uses of debt; Roubini and Wachtel (1999) on current account sustainability; studies of crises in the 1970s and early 1980s showed Daumont, Le Gall, and Leroux (2004) and Kawai, Newfarmer, and heavy state intervention through state-led industrialization, Schmukler (2005) on inadequate banking regulation; Brownbridge and state-owned companies, and state-owned banks (Balassa Kirkpatrick (2000) on balance sheet mismatch; and Capulong et. al. 1982). Industrial policy in countries such as Argentina, (2000) for poor corporate governance. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 275 BOX 4.3 Debt and crises (continued) Brazil, and Venezuela focused on import substitution commodity prices for commodity exporting industrialization, typically financed by external borrowing. economies (LAC and SSA in the 1980s, Russia in the 1990s), and contagion from both global crises (2007- Lack of economic diversification. A number of the crisis 09 global financial crisis) and regional crises (East countries had undiversified economies, which increased Asian financial and Russian crises in the 1990s), their vulnerability to terms of trade shocks. Several which generated sudden withdrawals of capital countries in Latin America and the Caribbean (LAC) and inflows. Sub-Saharan Africa (SSA), in particular, were heavily dependent on both oil and non-oil commodity exports • Natural disasters and domestic shocks. Natural disasters (Bolivia, Niger, Nigeria, Paraguay, Uruguay in the 1970s such as droughts were a major contributing factor to and 1980s). When commodity prices fell in the 1980s, the crises in some countries, typically smaller, less profitability of (often state-owned) corporates in the diversified economies (e.g. Bangladesh in the 1970s, resource sector, fiscal revenues, and export proceeds Nepal in the 1980s, Zimbabwe in the 2000s). collapsed, which triggered financial crises. • Other domestic shocks. In a small number of countries, Inadequate banking regulation. Poor banking regulation crises were triggered, or exacerbated, by other was a common feature in many case studies. Several SSA domestic shocks. Typically, these were episodes of countries experienced banking crises in the 1980s political turmoil (Turkey, Zimbabwe). primarily because of the failure of banks that were typically Crisis resolution. Many, though not all, crises were state-owned and subject to little oversight (Cameroon, resolved by policy programs of adjustment and structural Kenya, Niger, and Tanzania). In EAP, financial reform supported by financing from the IMF, World deregulation contributed to insufficient regulation and Bank, and other multilateral bodies and partner countries. oversight of the financial sector in the second wave (Indonesia, Korea, Malaysia, Philippines, and Thailand). • IMF support. The vast majority of countries in these This resulted in growing weaknesses, including balance case studies adopted IMF-supported policy programs sheet mismatches, and excessive risk taking by corporates to overcome their crises. The countries that did not (see below). In several countries in ECA during the 2000s, use IMF support typically had stronger fundamentals, cross-border lending was inadequately regulated by including lower public debt and larger international domestic regulators (Hungary and Kazakhstan). reserves (e.g. Colombia, Kazakhstan, Malaysia). Poor corporate governance. Among case studies of the • Debt restructuring. Among the case studies of 1980s and 1990s, poor corporate governance was a sovereign debt crises, many ended with default and common shortcoming, notably in some East Asian restructuring of debt (e.g. Argentina, Mexico, countries (Indonesia, Korea, and Thailand). Along with Nigeria). These cases were more common in the poor bank regulation, this led to inefficient corporate 1980s, 1990s, and early 2000s. Debt restructuring investment, as banks lent to firms without rigorously was often prolonged and occurred well after the initial evaluating their creditworthiness. sovereign debt crisis. Political uncertainty. Many sovereign debt crises were • Reforms. IMF support was conditional on the associated with severe political uncertainty (Indonesia, implementation of macroeconomic and structural Philippines, Turkey, Venezuela). reforms. For many EMDEs in LAC in the 1980s and in EAP in the late 1990s, crises were the trigger for Triggers and resolution of crises policy changes to allow greater exchange rate flexibility and strengthen monetary policy regimes. Triggers. The case studies suggest that crises were usually triggered by external shocks, although in a small number Conclusion of countries domestic factors also played a role. Crises are typically sparked by an adverse shock, such as an • External macroeconomic shocks. The most common increase in global interest rates or a growth slowdown, trigger of crises was an external shock to the real whose impact is amplified and propagated via country economy. These included a sudden rise in global vulnerabilities such as high levels of debt, especially short- interest rates (LAC in the 1980s), a slowdown in term debt, and low international reserves. In line with the global growth (ECA in the 2000s), a fall in literature, the econometric exercise conducted here 276 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 BOX 4.3 Debt and crises (continued) documents that a rapid rise in government or private debt robust bank supervision to the forefront of policy increases the probability of crises. A combined runup in discussions (Brownbridge and Kirkpatrick 2000; IMF government and private debt—as has been the case during 1999). The launch of the Financial System Assessment the fourth global wave—increases the probability of a Program in 1999 started systematic assessments of currency crisis. financial sectors (IMF 2000). The 2007-09 global financial crisis shifted an earlier consensus on the use of In several cases, crises revealed shortcomings that were capital controls. Before 2008, capital controls were largely mainly recognized ex post but had rarely been flagged considered ineffective and detrimental (Forbes 2004, before these crises. Following these crises, research 2007). After the global financial crisis, the literature shifted (described in academic studies and policy reports) shifted to a guarded endorsement of capital controls is its focus to these issues. For example, the Asian financial appropriately designed and implemented in the “right” crisis propelled the challenges of balance sheet mismatches circumstances (Forbes, Fratzscher, and Straub 2015; IMF and weak corporate governance as well as the need for 2012, 2015b). Rapid debt accumulation episodes of rapid government debt accumulation in 99 EMDEs since 1970, among a sample of 100 episodes EMDEs with available data for 1970-2018. It also yields 263 episodes of rapid private debt Spurts in debt buildups are common in EMDEs. accumulation in 100 EMDEs, out of a sample of When they coincide in many EMDEs, they form 100 EMDEs with available data for 1970-2018. the global waves of debt discussed above. is section examines the implications of national rapid Frequency of episodes. Debt accumulation debt accumulation episodes at the country level. It episodes have been common (Figure 4.7). EMDEs uses an event study approach that compares rapid in SAR, SSA, and LAC—the regions with the debt accumulation episodes that coincided with a largest number of episodes per country—had, on nancial crisis (which might be a currency, average, about 3 government and 3 private debt banking, or sovereign debt crisis) with those that accumulation episodes since 1970. Most episodes escaped a crisis. Box 4.3 analyses the factors which occurred in SSA (34 percent of all government increase the likelihood of a nancial crisis and 33 percent of all private debt accumulation occurring, including quantifying the impact of a episodes), in part re ecting the large number of rise in debt. countries in the region. Features of national rapid debt Duration. e average duration—the time accumulation episodes between trough and peak debt-to-GDP ratios— for both private and public episodes varied widely Definition of episodes. An episode of rapid debt but amounted to 7 years for the median accumulation is de ned as a period during which government episodes and 8 years for the median the government debt-to-GDP ratio or the private private episode. Most episodes had run their sector debt-to-GDP ratio rises by more than one course in less than a decade. However, 21 percent standard deviation from a trough to its next peak. of government episodes and 29 percent of private is approach closely follows the dating of turning debt episodes lasted for more than a decade. e points of business cycles but the key results are long duration of some of these episodes suggests robust to using a de nition more closely aligned that the debt buildup in part re ected nancial with the literature on credit booms (Claessens, development. Kose, and Terrones 2012; Mendoza and Terrones 2012; Annex 4.1). is approach results in 256 Amplitude. Although again with wide G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 277 heterogeneity among the episodes, the government FIGURE 4.7 Episodes of rapid debt accumulation in debt buildup in the median government debt EMDEs accumulation episode (30 percentage points of Episodes of rapid debt accumulation have been common among EMDEs, GDP from trough to peak) was double the private in both the government and private sectors. In the average year between debt buildup in the median private debt 1970 and 2018, three-quarters of EMDEs were in either a government or a private debt accumulation episode or both. Since the early 2000s, the accumulation episode (15 percentage points of number of combined government and private debt accumulation episodes GDP). Variation in the amplitude of debt has increased. During 1970-2018, the median debt accumulation episode lasted 7-8 years. During rapid debt accumulation episodes, government accumulation episodes across countries was debt typically rose (trough to peak) by 30 percentage points of GDP, and particularly wide for government debt private debt by 15 percentage points of GDP. accumulation episodes. In one-quarter of such episodes, the government debt buildup amounted A. EMDEs in rapid debt accumulation B. EMDEs in rapid debt accumulation episodes episodes to more than 50 percentage points of GDP. Debt accumulation on such a scale was rare for the private sector: in three-quarters of private debt accumulation episodes, private debt rose by less than 30 percentage points of GDP. Combined government and private debt accumulation episodes. About 70 percent of government and private debt accumulation episodes overlap. ese overlapping, combined C. Rapid government debt D. Rapid private debt accumulation government and private episodes, are statistically accumulation episodes by region episodes by region signi cantly shorter and more pronounced than solely-private or solely-government debt accumulation episodes (Annex Table 4.1.1). Episodes coinciding with crises. Financial crises— de ned as in Laeven and Valencia (2018)—can occur at any point during a debt accumulation episode, and more than one type of crisis can occur during an episode. Since 1970, based on all E. Duration of rapid debt F. Change in debt during rapid episodes that have concluded, more than half of accumulation episodes accumulation episodes government debt accumulation episodes and 40 percent of private debt accumulation episodes have been associated with crises (Figure 4.8). Crises were particularly common during the rst and second waves. Most crises occurred well before the end of the debt accumulation episode (Annex 4.1). Crises were equally common in longer-lasting (such as those lasting a decade or more) and shorter episodes (lasting less than a Source: International Monetary Fund; World Bank. decade). A.-D. Share of EMDEs in the sample that are in rapid debt accumulation episodes. Click here to download data and charts. Macroeconomic outcomes during national rapid debt accumulation episodes e one-half of debt accumulation episodes that were associated with nancial crises had considerably weaker macroeconomic outcomes than those that subsided without crises. e 278 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 4.8 Crises during rapid debt accumulation with crises were around 10 percent lower than in episodes in EMDEs episodes without a crisis; investment was 22 About half of all episodes of government and private debt accumulation percent lower; and consumption was 6 percent during 1970-2018 were associated with financial crises. Different types of lower. Several external indicators—international crises often occurred at the same time. The number of crises has fallen since the first two waves of debt. reserves, external debt—deteriorated signi cantly more in episodes associated with crisis than in non A. Government debt accumulation B. Private debt accumulation -crisis episodes as governments drew down reserves episodes associated with crises episodes associated with crises in an e ort to stem depreciation. Private debt accumulation episodes. After eight years, private debt accumulation episodes associated with crises featured weaker output and per capita income (by about 6 percent); consumption (by 8 percent); and investment (by 15 percent). Private debt accumulation episodes with crises also saw signi cantly more pronounced deteriorations in external positions—international C. Crises in EMDEs D. Crises during debt waves reserves, external debt—than non-crisis episodes. Similarities. Regardless of the borrowing sector, rapid debt accumulation episodes with crises featured considerably worse macroeconomic outcomes and vulnerabilities than those not associated with crises. Both types of debt accumulation episodes associated with crises saw larger falls in reserves and greater increases in Source: International Monetary Fund; World Bank. external debt than non-crisis episodes. Fiscal and Note: Episodes associated with crises are those which experienced financial crises (banking, currency, and debt crises, as in Laeven and Valencia 2018) during or within two years after the end of current account de cits widened in both types of episodes. For definition of episodes and sample, see Annex 4.1. Click here to download data and charts. episodes but more in government debt accumulation episodes than in private debt accumulation episodes. macroeconomic implications have tended to be Differences. Government debt accumulation worse when rapid debt growth stemmed from episodes associated with crises tended to be more both the government and the private sector.7 costly than private debt accumulation episodes Government debt accumulation episodes. associated with crises, with much larger shortfalls Government debt accumulation episodes that in output growth, especially in the early years after involved crises were typically associated with a crisis. Conversely, government debt greater debt buildups, weaker economic outcomes, accumulation episodes associated with crises and higher vulnerabilities than non-crisis episodes featured much larger drops in investment than (Figure 4.9). In the episodes associated with similar private debt accumulation episodes, nancial crises, the government debt buildup was possibly re ecting greater disruptions to nancing about 14 percentage points of GDP larger after conditions in crises during government debt eight years than in non-crisis episodes. After eight accumulation episodes. years, output and output per capita in episodes What comes next? 7 Combined government and private debt accumulation episodes were accompanied by signi cantly weaker investment and e current wave, not yet a decade old, has already consumption growth than solely-private episodes. Excluding episodes included the euro area debt crisis and several associated with crises, combined episodes also featured slower overall growth than solely private debt accumulation episodes. EMDE currency crises. Although EMDEs have G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 279 gone through periods of volatility during the FIGURE 4.9 Macroeconomic developments during debt current wave of debt, they have not yet accumulation episodes experienced widespread nancial crises. e key Rapid debt accumulation episodes associated with financial crises show question is whether the current wave of debt slower output, investment and consumption growth. Private debt accumulation will at some point end in nancial accumulation episodes associated with crises also had lower international reserves and higher external debt than episodes without any crisis events. crises in many EMDEs, as all its predecessors eventually did, or whether such crises will be A. Government episodes: Government B. Private episodes: Private debt avoided perhaps because EMDEs have learned and debt applied their lessons from the past. Prolonged period of low interest rates. e current environment of low interest rates and persistently low in ation in advanced economies alleviates some risks associated with the latest wave of debt. Policy interest rates in many advanced economies are near historical lows after major central banks recently reverted to an easing stance after winding down tightening cycles in 2018 C. Government episodes: Output, D. Private episodes: Output, (Figure 4.10). Monetary policy in advanced investment and consumption investment and consumption economies is likely to be accommodative for the foreseeable future as growth prospects and in ation expectations remain subdued. Interest payments on government debt in EMDEs have fallen from an average of 2.6 percent of GDP in 2000-07, to 1.6 percent of GDP in 2010-18, despite the increase in debt over that period. At current nominal GDP growth and long-term interest rates, debt appears to be on stable or E. Government episodes: International F. Private episodes: International falling trajectories in almost half of EMDEs. reserves and external debt reserves and external debt An easing of U.S. nancial conditions, a bellwether for global nancial conditions, has typically accompanied an increase in capital ows to EMDEs (Feyen et al. 2015). But increased borrowing can also raise vulnerability to a future rebound in interest rates. Historically, rising global interest rates have been a key trigger for nancial crises (Bulow et al. 1992; Bulow and Rogo 1989; Source: Bruegel; International Monetary Fund; Laeven and Valencia (2018); World Bank. Reinhart and Rogo 2010, 2011). Hence, low or Note: Median for episodes with data available for at least 8 years from the beginning of the episode. Year “t” refers to the beginning of rapid government debt accumulation episodes. Episodes falling global interest rates provide no sure associated with crises are those that experienced financial crises (banking, currency, and debt crises, as in Laeven and Valencia (2018)) during or within two years after the end of episodes. “*”, “**”, and protection against nancial crises for EMDEs. “***” denote that medians between episodes associated with crises and those with no crises are Half of all crises during episodes of rapid debt statistically different at 10 percent, 5 percent, and 1 percent levels, respectively, based on Wilcoxon rank-sum tests. accumulation occurred in years when U.S. long- A.B. Government (A) or private (B) debt in percent of GDP two and eight years after the beginning of the government debt accumulation episode (t). term (10-year) interest rates were falling and one- C.D. Cumulative percent increase from t, based on real growth rates for output (GDP), output (GDP) per capita, investment, and consumption. eighth of episodes occurred in years when U.S. E.F. Series shown as percent of GDP. long-term real interest rates were below 1 percent Click here to download data and charts. (as they have been since 2016). Weak growth prospects. In addition to interest rates and scal positions, growth is another major determinant of debt sustainability. An important 280 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 FIGURE 4.10 Fourth wave: Opportunities and risks reason for rapid debt accumulation has been the The current environment of very low interest rates has alleviated immediate sharp growth slowdown over the course of the risks associated with the latest accumulation of debt since long-term fourth wave. EMDE growth slowed after 2020 to interest rates are below growth in about half of EMDEs. However, while a trough of 4.1 percent in 2016 before a modest debt levels in advanced economies are on a sustainable path, debt levels in almost half of EMDEs are on a rising path. Although current levels of recovery took hold (Kose and Ohnsorge 2019). EMDE government or private debt are, on average, still below or near Current trends in fundamental drivers of growth those in the median rapid debt accumulation episode, increases in government or private debt since 2010 have already exceeded those of the suggest that it is likely to slow further over the typical historical episode in about one-quarter of EMDEs. next decade, to a pace about 0.5 percentage point lower than in 2013-17 (Ruch 2019; World Bank A. Long-term interest rates B. Share of economies with interest rates below growth 2018a). For commodity-exporting EMDEs— almost two-thirds of EMDEs—prospects will be further dimmed by the expected slowdown in commodity demand growth as major commodity- consuming EMDEs slow and mature (World Bank 2018b). e past decade has been marked by repeated growth disappointments. If these persist into the next decade, they could lead to growing concerns about debt sustainability, even in a world of low interest rates. C. Sustainability gaps D. Countries with negative sustainability gaps Vulnerability to external shocks. e previous three waves highlight the risks associated with a sharp buildup of debt. Financial crises typically occurred when external shocks hit EMDEs with domestic vulnerabilities. Many EMDEs have improved their monetary and scal policy frameworks over the past two decades. However, elevated debt levels in the current wave of debt accumulation have been accompanied by rising scal, corporate and external vulnerabilities. ese E. Current levels of government debt vs. previous rapid debt accumulation F. Current levels of private debt vs. previous rapid debt accumulation include lower international reserves and larger episodes episodes shares of EMDEs with current account and scal de cits. ere has been a signi cant change in the composition of debt in EMDEs. is shift could generate new vulnerabilities. Increasing issuance of foreign-currency-denominated corporate debt in EMDEs has contributed to rising currency exposures and heightened the risks of nancial Source: Bloomberg; Kose et al. (2017); World Bank. distress in the corporate sector and the banking A. Average long-term nominal government bond yields (with maturity of 10 years) computed with current U.S. dollar GDP as a weight, based on up to 36 advanced economies and 84 EMDEs. system in the event of U.S. dollar appreciation.8 In B. Share of countries where long-term nominal interest rates (represented by 10-year local currency some EMDEs, the share of nonresident-held government bond yields) are below nominal GDP growth for 1990-2018 in up to 34 advanced economies and 83 EMDEs. bonds in local currency bond markets has grown C.D. A sustainability gap is defined as the difference between the actual primary balance and the debt-stabilizing balance. Averages computed with current U.S. dollar GDP as weights, based on at to more than 30 percent. In LICs, debt has been most 34 advanced economies and 83 EMDEs. D. Share of economies in which sustainability gaps are negative (for example, debt is on a rising increasingly nanced by non-concessional sources. trajectory, or fiscal positions are debt-increasing). Sample includes 34 advanced economies and 83 EMDEs. E.F. Median levels of debt during debt accumulation episodes, as defined in Annex 4.1. t=0 indicates the peak of debt accumulation episodes that were completed before 2018. For current debt accumulation, t=0 indicates 2018. 8 This appreciation could be triggered, for example, by reversals Click here to download data and charts. of capital flows to EMDEs on heightened global risk aversion. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 281 Shocks could have several sources: deterioration of corporate nancial perfor- mance, and many corporates are facing • Although it seems unlikely in the foreseeable deteriorating pro tability (Molnar and Lu future, a return to monetary policy 2019; World Bank 2018b, 2019a). In view of normalization in advanced economies could the size of China’s economy, adverse spillovers raise borrowing cost. It would be likely to to other EMDEs would be likely (Ahmed et trigger U.S. dollar appreciation and a turn in al. 2019; World Bank 2016b). investor sentiment that would, especially, a ected those EMDEs with large foreign • LICs have accumulated debt rapidly and participation in local bond markets (Cerutti, increasingly from non-concessional and less Claessens, and Ratnovski 2017; Ruch 2019). transparent sources (Essl et al. 2019). is increases their vulnerability to nancing • A decade of tightening banking regulation has shocks and to the revelation of previously been accompanied by the emergence of credit undisclosed debt obligations (Bova et al. risk and maturity mismatches in the non-bank 2016; Horn, Reinhart, and Trebesch 2019; nancial system in advanced economies (Kose Lee and Bachmair 2019). and Ohnsorge 2019). Financial stress in non- bank nancial institutions could quickly • For some EMDEs, risks related to climate propagate to the rest of the nancial system, change are substantial. e experience of owing to the interconnectedness between several economies in LAC shows that debt nonbanks and banks. Growing linkages crises can be triggered by natural disasters between non-bank nancial systems in (Rasmussen 2004). To the extent that natural advanced economies and EMDEs have disasters are becoming more frequent and increased both the likelihood and the persistent as a result of climate change, they potential magnitude of spillovers from distress pose a growing risk to debt sustainability in in advanced-economy nonbanks to EMDE vulnerable EMDEs. Furthermore, the move to bond markets and broader nancial systems. a low-carbon economy could have a material e ect for energy-exporting EMDEs. A shift • Many commodity-exporting EMDEs rely away from the use of carbon-intensive fuels heavily on revenues from the resource sector could leave the assets of fossil fuel companies, to nd government expenditures and service including state-owned companies, stranded by sovereign debt (Correa and Sapriza 2014). As rules to curb climate change (Carney 2015). a result, commodity price shocks have is could have critical implications for debt periodically disrupted government nances sustainability both at the rm and the country and been a source of nancial instability in level. EMDEs, culminating in some cases in sovereign debt default or other nancial crises. Vulnerability to domestic shocks. Elevated debt increases an economy’s vulnerability to domestic • e large corporate debt buildup in China has nancing and political shocks even in an been primarily to domestic creditors. Its environment of benign global nancing counterpart in the nancial system could conditions. Domestic nancing shocks can trigger eventually reveal non-performing loans and sharp increases in borrowing cost. ese may result in a growth slowdown in China. include the sudden emergence of contingent Concerns remain that the rapid pace of government liabilities, including in state-owned investment growth may have contributed to enterprises or public-private partnerships. Policy overcapacity in some industries (Yu and Shen surprises or sudden bouts of policy uncertainty can 2019; Wang, Wan and Song 2018; also fuel investor concerns about debt repayment, Maliszewski et al. 2016). Although it has causing a jump in borrowing cost. recently declined, high corporate leverage in China, particularly that of state-owned Broader costs of debt accumulation. In addition enterprises, has been associated with a to restricting economies’ ability to weather shocks, 282 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 high debt may also act as a constraint on growth purposes to boost potential growth and exports, as of its own accord through three effects (Kose et al. painfully learned from the experience of the first 2019). First, high debt constrains governments’ wave. Crises were common in countries that ability to respond to downturns. For example, borrowed heavily to finance state-led fiscal stimulus during the 2008-09 global financial industrialization or real estate markets (e.g. crisis was considerably smaller in countries with Argentina and Brazil in the first wave, Thailand in high government debt than in those with low the second wave). government debt (World Bank 2015). Maintain a resilient debt composition. A debt As well as limiting the use of fiscal policy, high composition tilted toward foreign currency- government debt tends to render fiscal policy less denominated, short-term, or nonresident-held effective (Huidrom et al. 2019). Second, high debt debt makes countries more vulnerable to shifts in service costs may crowd out growth-enhancing market sentiment, currency depreciation, or spikes public investment or social safety nets (Obstfeld in global interest rates and risk premia. Crises have 2013; Reinhart and Rogoff 2010; Romer and been more likely when the share of short-term Romer 2018). Third, high debt could also create debt was higher. The first and second waves uncertainty about macroeconomic and policy showed how a high share of foreign currency- prospects (IMF 2018a; Kumar and Woo 2010). denominated debt meant that currency This can crowd out productivity-enhancing depreciations led to an increase in both debt private investment and weigh on output growth. servicing costs and debt ratios. Seven lessons Regulation and supervision matter. Inadequate regulatory and supervisory regimes, including gaps The analysis of waves of global and national debt in coordination between home and host accumulation episodes yields several important supervisors, can encourage excessively risky lessons for EMDEs. Box 4.3 complements the lending and debt buildup. This was the case in the lessons learned by considering 43 episodes of debt Asian financial crisis during the second wave and accumulation followed by financial crises in 34 in ECA countries during the third wave. EMDEs, and examining the similarities and Conversely, a robust regulatory system, that is also differences between these case studies. well coordinated between home and host supervisors of foreign banks, can temper the Accumulate debt with care. Borrowing, when well incentive to take excessive risks resulting from the spent and sustainable, could support growth. public safety net for the financial system (moral Waves of broad-based debt accumulation have hazard; Briault et al. 2018). typically coincided with global upturns amid accommodative monetary policy and financial Beware of external shocks (especially when there market development. However, about half of are domestic vulnerabilities). Crises typically rapid debt accumulation episodes at the country occurred when external shocks hit countries that level were associated with financial crises. Episodes had substantial domestic vulnerabilities, including of rapid government debt accumulation were a reliance on external and short-term debt in more likely than episodes of rapid private debt conjunction with a fixed exchange rate and low accumulation to be associated with crisis, and were levels of international reserves (Bordo, Meissner, costlier than rapid buildups of private debt. and Stuckler 2010; Mishkin 1999). In contrast, countries with higher international reserves were Use debt efficiently. The present combination of significantly more resilient to these types of shocks weak global growth and low interest rates makes (Gourinchas and Obstfeld 2012). In addition to government debt accumulation an appealing external shocks, domestic political shocks option for EMDEs to boost growth-friendly contributed to crises by increasing policy spending (World Bank 2019d). However, it is uncertainty and weakening investor sentiment. critical that the debt be used for productive G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 283 Private debt can rapidly turn into government balance between the benefits and costs of debt. Large private sector losses, including losses additional debt. These include sound debt threatening bank solvency, and the materialization management, high debt transparency, and of contingent liabilities, including those of state- thorough monitoring of contingent liabilities. owned enterprises, can lead governments to While these policies mostly apply to borrowers, provide substantial financial support (Mbaye, creditors also need to implement measures to Moreno-Badia, and Chae 2018b). This occurred mitigate risks associated with excessive debt in the EAP region in the second wave, and in ECA accumulation. in the third wave, with governments providing substantial support to banks. While the provision Sound debt management can help reduce of government support can save the banking borrowing costs, enhance debt sustainability, and system from collapse, it can also lead to a steep dampen fiscal risks.9 Debt managers are jump in public debt which, in turn, can heighten increasingly adopting pro-active policies to build the fragility of banks with large sovereign buffers and make the composition of debt more exposures (Bova et al. 2016; Claessens et al. 2014; resilient, but further progress is needed (World Feyen and Zuccardi 2019; World Bank 2015). Bank 2013). Prudent debt management favors Fiscal space can shrink rapidly as a result even debt contracted on terms that preserve though fiscal deficits may have been moderate. macroeconomic and financial resilience— preferably at longer maturities, at fixed (and Develop effective mechanisms to recognize losses favorable) interest rates, and in local currency. A and restructure debt. Having mechanisms in place debt composition that is less vulnerable to market to promptly recognize and restructure debt can disruptions reduces the likelihood that a decline in improve the prospects for recovery from crisis, market sentiment, sharp depreciations, or interest particularly public debt crises (Kroszner 2003) or rate spikes erode debt sustainability. A well- banking crises (Rutledge et al. 2012). The developed and liquid domestic bond market can protracted resolution after the Latin American reduce the need for foreign currency-denominated crises of the 1970s and the SSA debt distress in the lending and help ensure stability in government 1980s and 1990s were associated with a period of financing (Arvai and Heenan 2008; World Bank very low, or even negative, per capita income and IMF 2001). growth. Growth only rebounded after the Brady Transparency about balance sheets is a pre- plan and the HIPC and MDRI debt initiatives requisite for sound debt management. History resolved debt distress and reduced debt overhangs. shows that public debt spikes can reflect the revelation of previously undisclosed liabilities such Policy implications as those revealed in Mozambique during the fourth wave (Jaramillo, Mulas-Granados, and Policy frameworks have improved in many Jalles 2017; Weber 2012). Greater fiscal EMDEs since the first two waves of debt. These transparency is associated with lower borrowing improvements played a critical role in mitigating costs, improvements in government effectiveness the adverse impact of the global financial crisis on and lower government debt (Kemoe and Zhan these economies at the end of the third wave of 2018; Montes, Bastos, and de Oliveira 2019). debt accumulation. However, there is still Improvements in data collection practices for LIC considerable scope for further improvement. debt would help policymakers undertake better- Specific policy priorities ultimately depend on informed borrowing decisions, and have been country circumstances but there are four broad associated with lower borrowing costs (Cady and strands of policies that can help contain the risks Pellechio 2006; World Bank and IMF 2018c). associated with the recent debt accumulation. Policies for managing debt 9 Recognizing the need for better debt management, the World Bank and IMF have developed guidelines, best practices, and Governments need to put in place mechanisms frameworks to assist countries in implementing debt management and institutions that help them strike the proper strategies (World Bank and IMF 2014). 284 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 Principles and guidelines for debt transparency overstated. During episodes of financial stress, have been created, both by international financial when EMDE currencies tend to depreciate institutions, including the IMF’s fiscal sharply, strong monetary policy frameworks will transparency code, and by the private sector (IIF be helpful not least because the exchange rate pass- 2019a; IMF 2019d). through to inflation tends to be smaller in countries with more credible, transparent and Monitoring and mitigation of contingent independent central banks; inflation-targeting liabilities are integral for sound public debt monetary policy regimes; and better-anchored management. Recent survey evidence suggests that inflation expectations (Kose et al. 2019). With less a majority of public debt managers are monitoring pass-through from depreciation to inflation, risks of contingent liabilities; only a minority, central banks in EMDEs will have more scope to however, use risk mitigation tools, such as reserve support activity. Flexible exchange rates can accounts (40 percent of respondents) or risk provide an effective mechanism for exposure limits on contingent liabilities (30 macroeconomic adjustment and can help avoid percent of respondents; Lee and Bachmair 2019). currency overvaluations and the buildup of large currency mismatches on balance sheets—a Creditors, including international financial common precursor of crises. A flexible exchange institutions, play an important role in mitigating rate regime requires, however, that monetary the risks associated with debt accumulation. For policy pursue a credible policy of inflation control example, while country authorities have the to provide an effective nominal anchor to the primary responsibility to transparently report their economy. Such a policy framework needs to be debt data, international financial institutions complemented by strong macroeconomic and support transparency and sustainable lending institutional arrangements. practices through several measures. The IMF and the World Bank collect and disseminate debt Fiscal rules can help avoid fiscal slippages, ensure statistics that are used by a wide range of that revenue windfalls during times of strong stakeholders; produce published analyses of public growth are prudently managed, and manage and debt data via debt sustainability analyses (DSAs); contain risks from contingent liabilities (Cebotari support countries’ efforts to produce medium- 2008; Currie and Velandia 2002; Romer and term debt management strategies (MTDSs); Romer 2019; Ulgenturk 2017). Strong fiscal publish information on countries’ borrowing frameworks have also been associated with lower capacity; and directly liaise with multilateral, inflation and inflation volatility, supporting the bilateral, and private creditors. All of these efforts central bank in delivering its mandate (Ha, Kose, provide important support to borrowers and and Ohnsorge 2019). EMDEs have made lenders in their decision making. important strides in the adoption and design of fiscal rules (Schaechter et al. 2012).10 However, Macroeconomic policies fiscal rules may only be effective once a certain degree of broader government effectiveness is Notwithstanding substantial improvements since achieved and sound budgetary institutions are in the 1990s, macroeconomic policy frameworks can place.11 be strengthened further in many EMDEs (Kose and Ohnsorge 2019). Monetary policy frameworks and exchange rate regimes can be 10 Schaechter et al. (2012) create an overall fiscal rule index that captures both the number and characteristics of fiscal rules in strengthened to increase central bank credibility. operation in advanced economies and EMDEs and show how Fiscal frameworks can ensure that borrowing EMDEs have played catch-up to advanced economies since 2000. remains within sustainable limits and borrowed Ardanaz et al. (2019) find that well-designed fiscal rules can help safeguard public investment during downturns. funds are used well. 11 Calderón and Nguyen (2016) estimate that fiscal and monetary policy procyclicality is greater in countries with weak institutions. Macroeconomic and exchange rate policy Bergman and Hutchison (2015, 2018) show that fiscal rules are effective only when government effectiveness exceeds a minimum frameworks. The benefits of stability-oriented and threshold. World Bank (2015) discusses the circumstances and resilient monetary policy frameworks cannot be features that can make fiscal rules more effective. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 285 Alternatives to debt accumulation are available to al. 2008). At the same time, however, excessively expand fiscal resources for priority spending. rapid growth in financial markets can generate Public spending can be reallocated to uses that are financial stability risks. A careful balance between more likely to boost future growth, including measures to promote financial market deepening education and health spending as well as climate- and supervision and regulation is critical. smart infrastructure investment to strengthen economic resilience. Government revenue bases Strengthening institutions can be broadened by removing special exemptions and strengthening tax administration (Gaspar, Well-enforced frameworks for sound corporate Ralyea, and Ture 2019; IMF 2019c; World Bank governance can help ensure that funds borrowed 2017b). Government can also take action to foster by private corporates are well used. Sound private sector-led growth. Reform agendas to bankruptcy frameworks can help prevent debt improve business climates and institutions have overhangs from weighing on investment for resulted in significant gains in investment and prolonged periods. productivity EMDEs (World Bank 2018a). In turn, increased private sector growth expands the The promotion of good corporate governance can revenue base and, ultimately, strengthens mitigate risks arising from the corporate sector. government revenues. Stronger corporate governance can tilt firms’ financing towards equity rather than debt (Mande, Financial sector policies Park, and Son 2011); increase hedging of foreign currency positions to protect against external Robust financial sector regulation and supervision shocks (Lel 2012); and encourage more efficient can help prevent risks from building up. Financial firm operation (Henry 2010). Other measures can market deepening can help mobilize domestic also help contain risks from corporate credit savings that may provide more stable sources of growth, such as increased stress testing of listed financing than capital inflows. corporates’ balance sheets. Improved financial system regulation and Effective bankruptcy and insolvency regimes can supervision, by acting on systemic exposures and help in the resolution of private debt crises and ensuring adequate capital buffers, can help prevent have benefits outside of crises (Leroy and risks from building up. Robust prudential Grandolini 2016). Several EMDEs have recently regulation and supervision can help pre-empt the reformed bankruptcy procedures, but in general, buildup of systemic financial weaknesses. EMDE bankruptcy protection laws lag Macroprudential policies can help moderate international best practices.12 Strengthening lending to households and corporates. The use of bankruptcy protection can boost investment and living wills for banks and robust bank bankruptcy facilitate responsible corporate risk-taking, helping regimes can also help with the orderly winding to relieve the costs of debt overhang (World Bank down of insolvent institutions, including through 2014b). Well-functioning legal, regulatory, and the bail-in of creditors. Credibility and institutional frameworks are crucial for predictability of bank resolution can help prevent commercial banks and companies to resolve non- spillovers from the failure of one financial performing loans and facilitate business exit and institution to others by reassuring creditors about reorganization (Menezes 2014). A robust the continued functioning of the financial system insolvency regime can improve financial inclusion as a whole (Hoshi 2011). and increase access to credit, by reducing the cost of lending. Financial market deepening can help expand the pool of stable long-term domestic savings available for domestic investment. This requires an enabling environment of robust institutions, protection of creditor rights, sound regulatory quality and 12 These include the introduction of a new bankruptcy law in Egypt and strengthening of secured creditors’ rights in India. macroeconomic stability (Laeven 2014; Sahay et 286 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 ANNEX FIGURE 4.1.1 Country examples of debt Second, an expansion phase is labeled as a rapid accumulation episodes accumulation episode if an increase in debt-to- GDP ratio (from trough to peak) exceeds the A. Turkey: Government debt B. Mexico: Government debt maximum ten-year moving standard deviation (over the period t-9 to t) of the debt-to-GDP ratio during the phase (Figure A4.1.1). In scaling debt by GDP, this approach implicitly focuses on the concept of debt burden, which captures the ability of borrowers to repay their debt.2 An increase in the debt burden, as measured by the debt-to-GDP ratio used here, could reflect an output collapse, an exchange rate depreciation, C. Philippines: Private debt D. Malaysia: Private debt or outright borrowing. Regardless of these underlying reasons, an increase in the debt burden makes it more challenging to service debt and makes the debt burden more likely to become a source of financial or economic stress. In practice, output contractions are the source of increases in debt-to-GDP ratios only in a minority of episodes identified here (one-third of Source: International Monetary Fund; World Bank. government debt episodes and two-fifths of Note: Blue line indicates debt outside debt accumulation episodes. A period of debt accumulation is private debt episodes). Currency crises are indeed identified with the algorithm in Harding and Pagan (2002). When a change in debt-to-GDP ratios over an accumulation period is above the maximum of 10-year moving standard deviation of the ratios associated with larger debt buildups during the during the period, it is considered as a rapid debt accumulation (shown as an orange area). When it is below the threshold, it is treated as a non-rapid accumulation (shown as a light blue area). If a crisis debt accumulation episodes identified here, but (i.e., banking, currency, or debt crisis) occurs during a rapid debt accumulation period or within two years since the end of the period, it is regarded as an episode of rapid debt accumulation associated these currency crises typically happen well before with a crisis (shown as a red line). An ongoing episode (e.g., the third orange area in Panel C) is also classified as either rapid or non-rapid accumulation, based on the same methodology. (two years before) debt peaks and the increase in Click here to download data and charts. debt during the year of the currency crisis only accounts for one-tenth (private debt episodes) to one-quarter (government debt episodes) of the ANNEX 4.1 Event study total debt buildup during these debt accumulation episodes associated with currency crises. methodology Phases at the beginning and end of data series are Identifying episodes of rapid debt accumulation. also classified as either rapid or non-rapid The identification of episodes of rapid accumulation, if they are on the expansion accumulation of government and private debt trajectory. While they are identified in the same proceeds in two steps. First, the Harding and way as in the other cases, the beginning and end of Pagan’s (2002) algorithm is used to identify the episodes are set when data availability of cyclical turning points in the debt-to-GDP ratios. government and private debt begins and ends. In particular, a debt cycle (from one peak debt-to- GDP ratio to the next peak debt-to-GDP ratio) is An episode of rapid debt accumulation is assumed to last at least five years with a minimum associated with a financial crisis if a crisis— two-year duration of the contraction phase (from banking, currency, or debt crisis—occurs during peak to trough) and the expansion (or the period of rapid debt accumulation or at least accumulation) phase (from trough to peak).1 within two years since the end of the episode. The 2 Debt buildup results from both demand and supply factors. This dating method is documented in Kose, Nagle, Ohnsorge, 1 Regardless of which of these predominates, a high debt-to-GDP ratio and Sugawara (2019). presents a vulnerability in the event of adverse shocks. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 287 ANNEX TABLE 4.1.1 Comparison of combined government and private debt accumulation episodes with solely government or private debt accumulation episodes. Rapid accumulation with crises Rapid accumulation without crises Government Private Both Private Both Government debt debt debt (combined) debt (combined) Duration (years) 7 8 3 7 8 4 Amplitude (percentage points) 42.6 13.1 35.3 21.6 14.8 26.0 Growth (percent) 2.2 3.7 2.7 4.1 4.6 4.2 Per capita growth (percent) 0.1 1.9 0.9 2.0 2.5 2.0 Investment growth (percent) 1.9 5.7 2.2 6.3 7.2 6.1 Private consumption growth 2.5 4.0 2.9 4.1 4.8 4.2 (percent) Reserves (percent of GDP) 7.2 7.2 6.6 12.9 13.2 12.9 Short-term external debt 4.4 4.8 4.3 3.9 3.7 3.8 (percent of GDP) Note: Amplitude for "Both (combined)" is measured as an average of amplitudes of government debt and private debt during a combined part. Bold numbers indicate statistically significant difference from combined episodes. information on crisis years is obtained from also used. The Hausman test suggests that the Laeven and Valencia (2018). The year coverage for random effects model is appropriate for debt and currency crises is extended to 2018, by following banking crises but not for currency crises. the methodology in Laeven and Valencia (2018) However, even for currency crises, the coefficient using data on end-of-year exchange rates vis-à-vis estimates and their statistical significance remain U.S. dollars from the IMF. This association only similar in fixed effects and random effects models. describes the timing or coincidence between rapid To exploit the time and cross-sectional accumulation of debt and financial crisis, and dimensions, a panel dataset of 139 EMDEs with therefore does not imply any causal link between annual data over the period 1970–2018 is the two. constructed. The details of the methodology are described in Kose et al. (2019). Sample. The sample includes data for 100 EMDEs for 1970-2018, while the identification of Selection of explanatory variables. The variables debt accumulation uses data prior to 1970 (see are chosen from a close examination of the Kose et al. 2019 for details). Small states, as empirical findings from the early warning crisis defined by the World Bank, are excluded. This literature (see Chamon and Crowe 2012; Frankel results in 256 episodes of rapid government debt and Saravelos 2012; and Kaminsky, Lizondo, and accumulation and 263 episodes of rapid private Reinhart 1998 for an extensive review). A large debt accumulation in a sample of 100 EMDEs number of variables is included (with various data with available data for 1970-2018. transformations, such as levels, growth rate, percentage point change, deviation from trend) that can be grouped into several categories: ANNEX 4.2 Regression • Debt profile: public and private debt (percent methodology of GDP); short-term debt and concessional debt (in percent of total debt); debt service on Discrete choice modelling. The most common external debt (in percent of exports). estimation methods used in the empirical literature on predicting crises are logit and probit • Capital account: international reserves (in models. The baseline specification used in this months of imports), currency mismatch study is a panel logit model with random effects, (foreign liabilities to foreign assets), net FDI but for robustness purposes, a random effects inflows (in percent of GDP). probit model and a fixed effects logit model are 288 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 • Current account: exchange rate overvaluation ANNEX 4.3 Case studies (percent deviation from Hodrick Prescott- filtered trend). The in-depth literature review of Box 4.3 covered • Foreign environment: U.S. interest rate 43 crisis case studies for 30 EMDEs since 1970. (deflated by GDP deflator, in percent). While non-exhaustive, the case studies were chosen to: (i) be representative of debt • Domestic environment: GDP growth (in accumulation episodes over the past fifty years; (ii) percent). include the large EMDEs in major regional debt crises episodes; (iii) represent crises in low-income • Banking sector: funding ratio (banking system countries; and (iv) a sufficiently comprehensive credit to deposits); literature to base an assessment on. To attenuate potential endogeneity bias caused by contemporaneous interaction between economic In the case of the in-depth literature review, the fundamentals and crises, lagged values of the search covered all publicly available country explanatory variables are used, except for U.S. reports and flagship publications of international interest rate. Robustness checks using alternative financial institutions (Asian Development Bank, model specifications as well as results for African Development Bank, European Bank for probabilities of twin and triplet crises are provided Reconstruction and Development, Inter-American in Kose et al. (2019). Development Bank, International Monetary Fund, World Bank) and academic publications Probability of crises. The probability of crises published during 1970-2018. Publications were occurring are evaluated at specific points of found on the institutions’ websites and, especially interest for illustration (while keeping all other before 1997, in the EconLit database. The main variables at their average values). For details, see sources are detailed in Kose et al. (2019). Kose et al. (2019). G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 289 “Assessing Early Warning Systems: How Have They References Worked in Practice?” IMF Staff Papers 52 (3): 462-502. Ahmed, S., R. Correa, D. A. Dias, N. Gornemann, J. Berglöf, E., Y. Korniyenko, A. Plekhanov, and J. Hoek, A. Jain, E. Liu, and A. Wong. 2019. “Global Zettelmeyer. 2009. “Understanding the Crisis in Spillovers of a China Hard Landing." International Emerging Europe.” EBRD Working Paper 109, Finance Discussion Papers 1260, Federal Reserve Board European Bank for Reconstruction and Development, of Governors, Washington, DC. London. Altunbaş, Y., B. Gadanecz, and A. Kara. 2006. “The Bergman, U. M., and M. Hutchison. 2015. “Economic Evolution of Syndicated Loan Markets.” The Service Stabilization in the Post-Crisis World: Are Fiscal Rules Industries Journal 26 (6): 689-707. the Answer?” Journal of International Money and Anderson, P. R. D., A. C. Silva; A. Velandia-Rubiano. Finance 52 (April): 82-101. 2010. “Public Debt Management in Emerging Market ———. 2018. “Fiscal Procyclicality in Developing Economies: Has This Time Been Different?” Policy Economies: The Role of Fiscal Rules, Institutions and Research Working Paper 5399, World Bank, Economic Conditions.” Unpublished paper available at Washington, DC. https://www.researchgate.net/ Ardanaz, M., E. Cavallo, A., Izquierdo, and J. Puig. publication/325820923_Fiscal_Procyclicality_in_Dev 2019. “Growth-Friendly Fiscal Rules? Protecting Public eloping_Economics_The_Role_of_Fiscal_Rules_Institu Investment from Budget Cuts Through Fiscal Rule tions_and_Economic_Conditions. Design.” IDB Discussion Paper 698, Inter-American Bertola, L., and J. A. Ocampo. 2012. The Economic Development Bank, Washington, DC. Development of Latin America Since Independence. Arizala, F., M. Bellon, M. MacDonald, M. Mlachila, Oxford: Oxford University Press. and M. Yenice. 2018. “Regional Spillovers in Sub- BIS (Bank for International Settlements). 2015. Saharan Africa: Exploring Different Channels.” “Debt.” BIS Papers 80, Bank for International Spillover Notes 18/01, International Monetary Fund, Settlements, Basel. Washington, DC. ———. 2018. “Structural Changes in Banking After Árvai, Z., and G. Heenan. 2008. “A Framework for the Crisis.” CGFS Papers 60, Committee on the Global Developing Secondary Markets for Government Financial System, Bank for International Settlements, Securities.” IMF Working Paper 8/174, International Basel. Monetary Fund, Washington, DC. Bordo, M., C. Meissner, and D. Stuckler. 2010. Aslund, A. 2010. The Last Shall Be the First: The East “Foreign Currency Debt, Financial Crises and European Financial Crisis. Washington, DC: Peterson Economic Growth: A Long-run View.” Journal of Institute for International Economics. International Money and Finance 29 (4): 642-665. Baig, T., and I. Goldfajn. 1999. “Financial Market Borio, C. and P. Lowe. 2002. “Assessing the Risk of Contagion in the Asian Crisis.” IMF Staff Papers 46 Banking Crises.“ BIS Quarterly Review 7 (1): 43-54. (2): 167-195. Bova, E., M. Ruiz-Arranz, F. Toscani, and H. E. Ture. Balakrishnan, R., S. Danninger, S. Elekdag, and I. 2016. “The Fiscal Costs of Contingent Liabilities; A Tytell. 2011. “The Transmission of Financial Stress New Dataset.” IMF Working Paper 16/14, from Advanced to Emerging Economies.” Emerging International Monetary Fund, Washington, DC. Markets Finance and Trade 47 (2): 40-68. Briault, C., E. Feyen, I. Del Mazo, B. Yee, J. , Balassa, B. 1982. Development Strategies in Semi- Rademacher, and I. Skamnelos. 2018. “Cross-border Industrial Economies. Washington, DC: World Bank. Spillover Effects of the G20 Financial Regulatory Reforms: Beltran, D.O., and C. G. Collins. 2018. “How Results from a Pilot Survey. Washington, DC: World Culnerable are EME Corporates?” IFDP Notes K.7, Bank. Board of Governors of the Federal Reserve System, Broner, F., G. Lorenzoni, and S. Schmukler. 2013. Washington, DC. “Why Do Emerging Economies Borrow Short Berg, A., E. Borensztein, and C. Pattillo. 2005. Term?” Journal of the European Economic Association 11 (1): 67–100. 290 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 Brownbridge, M., and C. Kirkpatrick. 2000. “Financial “The Real Effects of Debt.” BIS Working Paper 352, Regulation in Developing Countries.” The Journal of Bank for International Settlements, Basel. Development Studies 31 (1): 1-24. Cerutti, E., S. Claessens, and L. Laeven. 2017. “The Brueggemann, A., M. Bussière, L. Chitu, J. Estefanía, Use and Effectiveness of Macroprudential Policies: S. Gallego Herrero, I. Garrido, C. Janssens, et al. 2018. New Evidence.” Journal of Financial Stability 28 “Strengthening the Global Financial Safety Net.” (February): 203–224. Occasional Paper 207, European Central Bank, Frankfurt. Cerutti, E., S. Claessens, and L. Ratnovski. 2017. “Global Liquidity and Cross-border Bank Flows.” Brunnermeier, M., and L. Pedersen. 2009. “Market Economic Policy 32 (89): 81–125. Liquidity and Funding Liquidity.” The Review of Financial Studies, 22(6): 2201-38. Cerutti, E., and H. Zhou. 2017. “The Global Banking Network in the Aftermath of the Crisis: Is There Bulow, J., and K. Rogoff. 1989. “A Constant Evidence of De-globalization?” IMF Working Paper Recontracting Model of Sovereign Debt.” Journal of 17/232, International Monetary Fund, Washington, Political Economy 97 (1): 155-178. DC. Bulow, J., K. Rogoff, A. Bevilaqua, S. Collins, and M. ———. 2018. “The Global Banking Network: What is Bruno. 1992. “Official Creditor Seniority and Burden- Behind the Increasing Regionalization Trend?” IMF Sharing in the Former Soviet Bloc.” Brookings Papers on Working Paper 18/46, International Monetary Fund, Economic Activity, 1992(1): 195-234. Washington, DC. Cady, J., and A. Pellechio. 2006. “Sovereign Borrowing Cetorelli, N., and L. S. Goldberg. 2011. “Global Banks Cost and the IMF's Data Standards Initiatives.” IMF and International Shock Transmission: Evidence from Working Paper 06/78, International Monetary Fund, the Crisis.” IMF Economic Review 59 (1): 41-76. Washington, DC. Chamon M., and C. Crowe. 2012. “Predictive Calderón, C., R. Duncan, and K. Schmidt-Hebbel. Indicators of Crises.” In Handbook in Financial 2016. “Do Good Institutions Promote Countercyclical Globalization: The Evidence and Impact of Financial Macroeconomic Policies?” Oxford Bulletin of Economics Globalization, edited by G. Caprio, 499–505. London: and Statistics 78 (5): 650-670. Elsevier. Calderón, C., and H. Nguyen. 2016. “The Cyclical Claessens, S., and K. Forbes, eds. 2013. International Nature of Fiscal Policy in Sub-Saharan Africa.” Journal Financial Contagion. Berlin: Springer. of African Economies, 25(4): 548-579. Claessens, S., and M. A. Kose. 2018. “Frontiers of Callaghy, T. 2002. “Innovation in the Sovereign Debt Macrofinancial Linkages.” BIS Papers 95, Switzerland, Regime: From the Paris Club to Enhanced HIPC and Bank for International Settlements, Basel. Beyond.” Operations Evaluation Department, World Bank, Washington, DC. Claessens, S., M. A. Kose, L. Laeven, and F. Valencia. 2014. Financial Crises: Causes, Consequences, and Policy Capulong, M.V., D. Edwards, D. Webb, and J. Responses. Washington, DC: International Monetary Zhuang. 2000. An Analytic Framework of Corporate Fund. Governance and Finance. Corporate Governance and Finance in East Asia: A Study of Indonesia, Republic of Claessens, S., M. A. Kose, and M. Terrones. 2012. Korea, Malaysia, Philippines, and Thailand. Manila: “How do Business and Financial Cycles Interact?” Asian Development Bank. Journal of International Economics 87 (1): 178-90. Carney, M. 2015. “Breaking the Tragedy on the Claessens, S., C. Pazarbasioglu, L. Laeven, M. Dobbler, Horizon—Climate Change and Financial Stability.” F. Valencia, O. Nedelescu, and K. Seal. 2014. “Crisis Speech at Lloyd’s of London, September 29, London . Management and Resolution: Early Lessons from the 2007-09 Financial Crisis.” In Financial Crises: Causes, Cebotari, A. 2008. “Contingent Liabilities: Issues and Consequences, and Policy Responses, edited by S. Practice.” IMF Working Paper 245, International Claessens, M. A. Kose, L. Laeven, and F. Valencia. Monetary Fund, Washington, DC. Washington, DC: International Monetary Fund. Cecchetti, S., Mohanty, M., and Zampolli, F. 2011. Claessens, S., and N. van Horen. 2014. “Foreign G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 291 Banks: Trends and Impact.” Journal of Money, Credit Devlin, R. 1990. Debt and Crisis in Latin America: The and Banking 46 (1): 295-326. Supply Side of the Story. Princeton, NJ: Princeton University Press. Cline, W. 1995. International Debt Reexamined. Washington, DC: Institute for International Dicks, M. 1991. “The LDC Debt Crisis.” Bank of Economics. England Quarterly Bulletin (Q4): 498-508. Correa, R., and H. Sapriza. 2014. “Sovereign Debt Didier, T., C. Hevia, and S. Schmukler. 2012. “How Crises.” International Finance Discussion Paper 1104, Resilient and Countercyclical were Emerging May, Board of Governors of the Federal Reserve Economies to the Global Crisis?” Journal of System, Washington, DC. International Money and Finance 31 (8): 2052–2077. Corsetti, G., P. Pesenti, and N. Roubini. 1998. “What Didier, T., M. A. Kose, F. Ohnsorge, and L. Ye. 2015. Caused the Asian Financial Crisis? Part I: A “Slowdown in Emerging Markets: Rough Patch or Macroeconomic Overview.” NBER Working Paper Prolonged Weakness?” Policy Research Note 4, World 6833, National Bureau of Economic Research, Bank, Washington, DC. Cambridge, MA. Dincer, N., and B. Eichengreen. 2014. "Central Bank Cortina-Lorente, J. J., T. Didier, and S. L. Schmukler. Transparency and Independence: Updates and New 2016. “How Long is the Maturity of Corporate Measures." International Journal of Central Banking 10 Borrowing? Evidence from Bond and Loan Issuances (1): 189-259. Across Markets.” Policy Research Working Paper 7815, World Bank, Washington, DC. Dollar, D. 2016. China’s Engagement with Africa: From Natural Resources to Human Resources. Washington, Currie, E., and A. Velandia. 2002. “Risk Management DC: Brookings Institution. of Contingent Liabilities Within a Sovereign Asset- Liability Framework.” World Bank, Washington, DC. Dornbusch, R., W. Branson, and W. Cline. 1985. “Policy and Performance Links between LDC Debtors Daseking C., and R. Powell. 1999. “From Toronto and Industrial Nations.” Brookings Papers on Economic Terms to the HIPC Initiative: A Brief History of Debt Activity 1985 (2): 303–368. Relief for Low-Income Countries.” IMF Working Paper 99/142, International Monetary Fund, Dornbusch, R., I. Goldfajn, R. Valdés, S. Edwards, and Washington, DC. M. Bruno. 1995. “Currency Crises and Collapses.” Brookings Papers on Economic Activity 1995 Daumont, R., F. Le Gall, and F. Leroux. 2004. (2): 219-293. “Banking in Sub-Saharan Africa: What Went Wrong?” IMF Working Paper 04/55, International Monetary Easterly, W. 2002. “How Did the Heavily Indebted Fund, Washington, DC. Poor Countries Become Heavily Indebted? Reviewing Two Decades of Debt Relief.” World Development 30 Dawood, M., N. Horsewood, and F. Strobel. 2017. (10): 1677-1696. “Predicting Sovereign Debt Crises: An Early Warning System Approach.” Journal of Financial Stability 28: Eberhardt, M., and A. F. Presbitero. 2015. “Public 16-28. Debt and Growth: Heterogeneity and Non-linearity.” Journal of International Economics 97 (1): 45-58. Dell’Ariccia, G, D. Igan, L. Laeven, and H. Tong. 2014. “Policies for Macrofinancial Stability: Dealing Ehlers, T., S. Kong, and F. Zhu. 2018. “Mapping with Credit Booms and Busts.” In Financial Crises: Shadow Banking in China: Structure and Dynamics.” Causes, Consequences, and Policy Responses, edited by S. Working Paper 701, Bank for International Claessens, M.A. Kose, L. Laeven, and F. Valencia. Settlements, Basel. Washington, DC: International Monetary Fund. Eichengreen, B., A. El-Ganainy, R. Esteves, and K. J. ———. 2016. “Credit Booms and Macrofinancial Mitchener. 2019. “Public Debt through the Ages.” Stability.” Economic Policy 31 (86): 299–355. NBER Working Paper 25494, National Bureau of Economic Research, Cambridge, MA. Demirgüç-Kunt, A., and E. Detragiache. 1998. “The Determinants of Banking Crises in Developing and Eichengreen, B., K. Kletzer, and A. Mody. 2003. Developed Countries.” Staff Papers 45 (1): 81-109. “Crisis Resolution: Next Steps.” In Brookings Trade 292 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 Forum 2003, edited by S. M. Collins and D. Rodrik. Crashes in Emerging Markets: An Empirical Washington, DC: Brookings Institution. Treatment.” Journal of International Economics 41 (3- 4): 351-366. Eichengreen, B., A. K. Rose, and C. Wyplosz. 1995. “Exchange Market Mayhem: The Antecedents and Frankel, J. A., and G. Saravelos. 2012. “Can Leading Aftermath of Speculative Attacks.” Economic Policy 10 Indicators Assess Country Vulnerability? Evidence from (21): 249-312. the 2008-09 Global Financial Crisis.” Journal of International Economics 87 (2): 216-231. Elekdag, S., and Y. Wu. 2013. “Rapid Credit Growth in Emerging Markets: Boon or Boom-Bust?” Emerging FSB (Financial Stability Board). 2017. Assessment of Markets Finance and Trade 49 (5): 45-62. Shadow Banking Activities, Risks and the Adequacy of Post-Crisis Policy Tools to Address Financial Stability Essl, S., S. Kilic Celik, P. Kirby, and A. Proite. 2019. Concerns. Basel: Financial Stability Board. “Debt in Low-Income Countries: Evolution, Implications, Remedies.” Policy Research Working ———. 2018a. Evaluation of the Effects of Financial Paper 8794, World Bank, Washington, DC. Regulatory Reforms on Infrastructure Finance. Basel: Financial Stability Board. Feyen, E., N. Fiess, I. Z. Huertas, and L. Lambert. 2017. “Which Emerging Markets and Developing ———. 2018b. Global Shadow Banking Monitoring Economies Face Corporate Balance Sheet Report 2017. Basel: Financial Stability Board. Vulnerabilities? A Novel Monitoring Framework.” Policy Research Working Paper 8198, World Bank, ———. 2018c. Implementation and Effects of the G20 Washington, DC. Financial Regulatory Reforms. Basel: Financial Stability Board. Feyen, E., S. Ghosh, K. Kibuuka, and S. Farazi. 2015. “Global Liquidity and External Bond Issuance in ———. 2019. “Implementation of G20/FSB financial Emerging Markets and Developing Economies.” Policy reforms in other areas: Summary of Key Findings Based Research Working Paper 7373, World Bank, on the 2018 FSB Implementation Monitoring Washington, DC. Network (IMN) Survey.” Financial Stability Board, Basel, Switzerland. Feyen, E., and I. Gonzalez de Mazo. 2013. “European Bank Deleveraging and Global Credit Conditions: Furman, J., J. Stiglitz, B. Bosworth, and S. Radelet. Implications of a Multi-Year Process on Long-Term 1998. “Economic Crises: Evidence and Insights from Finance and Beyond.” Policy Research Working Paper East Asia.” Brookings Papers on Economic Activity 1998 6388, World Bank, Washington, DC. (2): 1-135. Feyen, E., and I. Zuccardi. 2019. “The Sovereign-Bank G-20 (Group of 20). 2013. “Long-Term Investment Nexus in EMDEs: What Is It, Is It Rising, and What Financing for Growth and Development: Umbrella Are the Policy Implications?” Policy Research Working Paper.” Paper 8950, World Bank, Washington, DC. G20 IFAWG. 2018. “Recent Development of Local Forbes, K. J. 2004. “Capital Controls: Mud in the Currency Bond Markets in Emerging Economies.” Wheels of Market Discipline.” NBER Working Paper Staff note, International Financial Architecture 10284, National Bureau of Economic Research, Working Group. Cambridge, MA. Gadanecz, B. 2004. “The Syndicated Loan Market: ———. 2007. “The Microeconomic Evidence on Structure, Development and Implications.” BIS Capital Controls: No Free Lunch.” In Capital Controls Quarterly Review, December, Bank for International and Capital Flows in Emerging Economies: Policies, Settlements, Basel. Practices and Consequences, edited by S. Edwards, 171- Gaspar, V., J. Ralyea, and E. Ture. 2019. “High Debt 202. Chicago: University of Chicago Press. Hampers Countries’ Response to a Fast-Changing Forbes, K., M. Fratzscher, and R. Straub. 2015. Global Economy.” IMFBlog; Insights & Analysis on “Capital-Flow Management Measures: What Are They Economics & Finance (blog), April 10. Good For?” Journal of International Economics 96 Glick, R., and A. K. Rose. 1999. “Contagion and (July): S76-S97. Trade: Why are Currency Crises Regional?” Journal of Frankel, J. A., and A. K. Rose. 1996. “Currency International Money and Finance 18 (4): 603-617. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 293 Gourinchas, P.-O., and M. Obstfeld. 2012. “Stories of ———. 2015b. Pan-African Banks: Opportunities and the Twentieth Century for the Twenty-first,” American Challenges for Cross-Border Oversight. Washington, DC: Economic Journal: Macroeconomics 4 (1): 226-65. International Monetary Fund. Greene, J. 1989. “The External Debt Problem of Sub- ———. 2016. Analyzing and Managing Fiscal Risks— Saharan Africa.” IMF Staff Papers 36 (4): 836-874. Best Practices. Washington, DC: International Monetary Fund. Ha, J., M. A. Kose, and F. Ohnsorge. 2019. Inflation in Emerging and Developing Economies: Evolution, Drivers ———. 2018a. Fiscal Monitor: Capitalizing on Good and Policies. Washington, DC: World Bank. Times. April Washington, DC: International Monetary Fund. Henry, D. 2010. Agency Costs, Ownership Structure and Corporate Governance Compliance: A Private ———. 2018b. Macroeconomic Developments and Contracting Perspective. Pacific-Basin Finance Prospects in Low-Income Developing Countries. Journal 18 (1): 24-46. Washington, DC: International Monetary Fund. Horn, S., C. M. Reinhart, and C. Trebesch. 2019. ———. 2018c. Global Financial Stability Report. A “China’s Overseas Lending.” NBER Working Paper Decade After the Global Financial Crisis: Are We Safer? 26050, National Bureau of Economic Research, Washington, DC: International Monetary Fund. Cambridge, MA. ———. 2019a. Global Financial Stability Report. Lower Hoshi, T. 2011. “Financial Regulation: Lessons from for Longer. October. Washington, DC: International the Recent Financial Crises.” Journal of Economic Monetary Fund. Literature 49 (1): 120-28. ———. 2019b. Global Financial Stability Report. Huidrom, R., M. A. Kose, J. J. Lim, and F. Ohnsorge. Vulnerabilities in a Maturing Credit Cycle. Washington, 2019. “Why Do Fiscal Multipliers Depend on Fiscal DC: International Monetary Fund. Positions?” Journal of Monetary Economics. Advance online publication. https://doi.org/10.1016/. ———. 2019c. Fiscal Monitor: Curbing Corruption. April. Washington, DC: International Monetary Fund. IMF (International Monetary Fund). 1999. Report of the Managing Director to the Interim Committee on ———. 2019d. “Fiscal Transparency Initiative: Progress in Strengthening the Architecture of the Integration of Natural Resource Management Issues.” International Financial System. Washington, DC: IMF Policy Paper, International Monetary Fund, International Monetary Fund. Washington DC. ———. 2000. Financial Sector Assessment Program Impavido, G., H. Rudolph, and L. Ruggerone. 2013. (FSAP) A Review: Lessons from the Pilot and Issues Going “Bank Funding in Central, Eastern and South Eastern Forward. Washington, DC: International Monetary Europe Post Lehman: A 'New Normal'?” IMF Fund. Working Paper 13/148, International Monetary Fund, Washington, DC. ———. 2006. “The Multilateral Debt Relief Initiative.” Factsheet. International Monetary Fund, Jaramillo, L., C. Mulas‐Granados, and J. T. Jalles. Washington, DC. 2017. “Debt Spikes, Blind Spots, and Financial Stress.” International Journal of Finance & Economics 22 (4): ———. 2012. The Liberalization and Management of 421-437. Capital Flows—An Institutional View. Washington, DC: International Monetary Fund. Jordà, Ò., M. Schularick, and A. M. Taylor. 2011. “Financial Crises, Credit Booms, and External ———. 2014. Global Financial Stability Report: Risk Imbalances: 140 Years of Lessons.” IMF Economic Taking, Liquidity, and Shadow Banking. International Review 59 (2): 340-378. Monetary Fund, Washington, DC. Kaminsky, G. L. 2008. “Crises and Sudden Stops: ———. 2015a. Global Financial Stability Report:, Evidence from International Bond and Syndicated- Vulnerabilities, Legacies, and Policy Challenges - Risks Loan Markets.” Monetary and Economic Studies 26: Rotating to Emerging Markets. International Monetary 107-129. Fund, Washington, DC. 294 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 Kaminsky, G. L., S. Lizondo, and C. M. Reinhart. Kose, M. A., and E. Prasad. 2010. Emerging Markets: 1998. “Leading Indicators of Currency Crises.” IMF Resilience and Growth Amid Global Turmoil. Staff Papers 45 (1): 1-48. Washington, DC: Brookings Institution Press. Kaminsky, G. L., and C. M. Reinhart. 1999. “The Kose, M. A., and M. E. Terrones. 2015. Collapse and Twin Crises: The Causes of Banking and Balance-of- Revival: Understanding Global Recessions and Recoveries. Payments Problems.” American Economic Review 89 Washington, DC: International Monetary Fund. (3): 473-500. Kroszner, R. S. 2003. “Sovereign Debt Restructuring.” Kaminsky, G. L., and C. M. Reinhart. 2000. “On American Economic Review 93 (2): 75-79. Crises, Contagion, and Confusion.” Journal of International Economics 51 (1): 145-168. Krugman, P. 2000. Currency Crises. Chicago: University of Chicago Press. ———. 2001. “Bank Lending and Contagion: Evidence from the Asian Crisis.” In Regional and Global Kumar, M., and J. Woo. 2010. “Public Debt and Capital Flows: Macroeconomic Causes and Consequences, Growth.” IMF Working Paper 174, International edited by T. Ito and A. O. Krueger, 73-99. Chicago: Monetary Fund, Washington, DC. University of Chicago Press. Laeven, L. 2014. “The Development of Local Capital Kaufman, R. R. 1989. “The Politics of Economic Markets: Rationale and Challenges.” IMF Working Adjustment Policy in Argentina, Brazil, and Mexico: Paper 14/234, International Monetary Fund, Experiences in the 1980s and Challenges for the Washington, DC. Future.” Policy Sciences 22 (3): 395-413. Laeven, L., and F. Valencia. 2018. “Systemic Banking Kawai, M. 2002. “Exchange Rate Arrangements in East Crises Revisited.” IMF Working Paper 18/206, Asia: Lessons from the 1997-98 Currency Crisis.” International Monetary Fund, Washington, DC. Monetary and Economic Studies 20 (S1): 167-204. Lee, A., and F. Bachmair. 2019. “A Look Inside the Kawai, M., R. Newfarmer, and S. Schmukler. 2005. Mind of Debt Managers: A Survey on Contingent “Crisis and Contagion in East Asia: Nine Lessons.” Liabilities Risk Management.” World Bank Treasury Eastern Economic Journal 31 (2): 185-207. Public Debt Management, World Bank, Washington, DC. Kemoe, L., and Z. Zhan. 2018. “Fiscal Transparency, Borrowing Costs, and Foreign Holdings of Sovereign Lel, U. 2012. “Currency Hedging and Corporate Debt.” IMF Working Paper 18/189, International Governance: A Cross-country Analysis.” Journal of Monetary Fund, Washington, DC. Corporate Finance 18 (2): 221-237. Kindleberger, C. P., and R. Z. Aliber. 2011. Manias, Leroy, A., and G. Grandolini. 2016. Principles for Panics and Crashes: A History of Financial Crises. Effective Insolvency and Creditor and Debtor Regimes. London: Palgrave Macmillan. Washington, DC: World Bank. Kose, M. A., S. Kurlat, F. Ohnsorge, and N. Sugawara. Loayza, N., P. Fajnzylber, and C. Calderón. 2005. 2017. “A Cross-Country Database of Fiscal Space.” "Economic Growth in Latin America and the Policy Research Working Paper 8157, World Bank, Caribbean: Stylized Facts, Explanations, and Washington, DC. Forecasts." Working Papers of the Central Bank of Kose, M. A., H. Matsuoka, U. Panizza, and D. Chile 265, Central Bank of Chile. Vorisek. 2019. “Inflation Expectations: Review and Magud, N. E., C. M. Reinhart, and K. S. Rogoff. Evidence.” Policy Research Working Paper 8785, 2011. “Capital Controls: Myth and Reality—A World Bank, Washington, DC. Portfolio Balance Approach.” NBER Working Paper Kose, M. A., P. Nagle, F. Ohnsorge, and N. Sugawara. 16805, National Bureau of Economic Research, 2019. Global Waves of Debt: Causes and Consequences. Cambridge, MA. Washington, DC: World Bank. Maliszewski, W., S. Arslanalp, J. Caparusso, J. Garrido, Kose, M. A., and F. Ohnsorge, eds. 2019. A Decade S. Guo, J. Kang, W. Lam, et al. 2016. “Resolving Since the Global Recession: Lessons and Challenges for China’s Corporate Debt Problem.” IMF Working Emerging and Developing Economies. Washington, DC: Paper 16/203, International Monetary Fund, World Bank. Washington, DC. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 295 Manasse, P., N. Roubini, and A. Schimmelpfennig. Obstfeld, M., and Rogoff, K. 1986. “Ruling Out 2003. “Predicting Sovereign Debt Crises.” IMF Divergent Speculative Bubbles.” Journal of Monetary Working Paper 221, International Monetary Fund, Economics 17 (3): 349-362. Washington, DC. OECD (Organisation for Economic Co-operation and Mande, V., Y. K. Park, and M. Son. 2012. Equity or Development). 2017. OECD Economic Outlook. Debt Financing: Does Good Corporate Governance November. Paris: Organisation for Economic Co- Matter? Corporate Governance: An International operation and Development. Review 20 (2): 195-211. Ohnsorge, F., and S. Yu. 2016. “Recent Credit Surge Mbaye, S., M. Moreno-Badia, and K. Chae. 2018a. in Historical Context.” Policy Research Working Paper “Bailing Out the People? When Private Debt Becomes 7704, World Bank, Washington, DC. Public.” IMF Working Paper 18/141, International Monetary Fund, Washington, DC. Ozatay, F., and G. Sak. 2002. “Banking Sector Fragility and Turkey's 2000-01 Financial Crisis.” In ———. 2018b. “Global Debt Database: Methodology Brookings Trade Forum 2002, edited by S. M. Collins and Sources.” IMF Working Paper 18/111, and D. Rodrik. Washington, DC: Brookings International Monetary Fund, Washington, DC. Institution. Mendoza, E. G., and M. E. Terrones. 2008. “An Panizza, U. 2008. “Domestic and External Public Debt Anatomy of Credit Booms: Evidence from Macro in Developing Countries.” United Nations Conference Aggregates and Micro Data.” NBER Working Paper on Trade and Development Discussion Paper 188, 14049, National Bureau of Economic Research, New York. Cambridge, MA. Panizza, U., and A. F. Presbitero. 2014. “Public Debt ———. 2012. “An Anatomy of Credit Booms and and Economic Growth: Is There a Causal Effect?” their Demise” NBER Working Paper 18379, National Journal of Macroeconomics 41 (September): 21–41. Bureau of Economic Research, Cambridge, MA. Pozsar, Z., T. Adrian, A. Ashcraft, and H. Boesky. Menezes, A. 2014. “Debt Resolution and Business Exit: 2010. “Shadow Banking.” FRBNY Economic Policy Insolvency Reform for Credit, Entrepreneurship, and Review 458: 3-9. Growth.” Viewpoint; Public Policy for the Private Sector series, No. 343, World Bank, Washington, DC. Rasmussen, M. 2004. Macroeconomic implications of natural disasters in the Caribbean (No. 4-224). Mishkin, F. 1999. “Global Financial Instability: International Monetary Fund. Framework, Events, Issues.” Journal of Economic Perspectives 13 (4): 3-20. Reinhart, C., V. Reinhart, and K. Rogoff. 2012. “Public Debt Overhangs: Advanced-Economy Episodes Molnar, M., and J. Lu. 2019. “State-owned Firms Since 1800.” Journal of Economic Perspectives 26 (3): Behind China’s Corporate Debt.” OECD Economics 69-86. Department Working Paper 1536, Organisation for Economic Co-operation and Development, Paris. Reinhart, C. M., and K. S. Rogoff. 2010. “Growth in a Time of Debt.” American Economic Review 100 (2): Montes, G. C., J. C. A. Bastos, and A. J. de Oliveira. 573–578. 2019. “Fiscal Transparency, Government Effectiveness and Government Spending Efficiency: Some ———. 2011. “From Financial Crash to Debt Crisis.” International Evidence Based on Panel Data American Economic Review 101 (5): 1676-1706. Approach.” Economic Modelling 79 (June): 211-225. Romer, C. D., and D. H. Romer. 2018. “Phillips Moreno, R., G. Pasadilla, and E. Remolona. 1998. Lecture—Why Some Times are Different: “Asia’s Financial Crisis: Lessons and Policy Responses.” Macroeconomic Policy and the Aftermath of Financial Pacific Basin Working Paper Series 98-02, Federal Crises.” Economica 85 (337): 1–40. Reserve Bank of San Francisco. ———. 2019. “Fiscal Space and the Aftermath of Obstfeld, M. 2013. “On Keeping Your Powder Dry: Financial Crises: How it Matters and Why.” NBER Fiscal Foundations of Financial and Price Stability.” Working Paper 25768, National Bureau of Economic Monetary and Economic Studies 31 (November): 25–37. Research, Cambridge, MA. 296 CHAPTER 4 G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 Rose, A. K., and M. M. Spiegel. 2012. “Dollar 1995.” Brookings Papers on Economic Activity 1996 (1): Illiquidity and Central Bank Swap Arrangements 147–215. During the Global Financial Crisis.” Journal of International Economics 88 (2): 326-340. Tornell, A., and F. Westermann. 2005. Boom-Bust Cycles and Financial Liberalization. Cambridge, MA: Roubini, N., and P. Wachtel. 1999. “Current-Account MIT Press Books. Sustainability in Transition Economies.” In Balance of Payments, Exchange Rates, and Competitiveness in Turner, P. 2002. “Bond Markets in Emerging Transition Economies, edited by M. I. Blejer and M. Economies: An Overview of Policy Issues.” BIS Papers Škreb. Boston: Kluwer Academic Publishers. 11, Bank for International Settlements, Basel. Ruch, F. U. 2019. “Prospects, Risks and Ülgentürk, L. 2017. "The Role of Public Debt Vulnerabilities.” In A Decade After the Global Recession: Managers in Contingent Liability Management." Lessons and Challenges for Emerging and Developing OECD Working Paper on Sovereign Borrowing and Economies, edited by M. A. Kose and F. Ohnsorge. Public Debt Management 8, Organisation for Washington, DC: World Bank. Economic Co-operation and Development, Paris. Rutledge, V., M. Moore, M. Dobler, W. Bossu, N. Unal, H., A. Demirgüç-Kunt, and K. Leung. 1993. Jassaud, and J. Zhou. 2012. “From Bail-Out to Bail- “The Brady Plan, 1989 Mexican Debt-Reduction In: Mandatory Debt Restructuring of Systemic Agreement, and Bank Stock Returns in United States Financial Institutions.” Staff Discussion Note 2012/03, and Japan.” Journal of Money, Credit and Banking 25 International Monetary Fund, Washington, DC. (3): 410-429. Sachs, J. 1985. “External Debt and Macroeconomic Performance in Latin America and East Asia.” Brookings Wang, D., K. Wan, and X. Song. 2018. “Quota Papers on Economic Activity 1985 (2): 523-573. Allocation of Coal Overcapacity Reduction Among Provinces in China.” Energy Policy 116: 170-181. ———. 1989. “New Approaches to the Latin American Debt Crisis.” In Essays in International Weber, A. 2012. “Stock-Flow Adjustments and Fiscal Finance 174, Princeton University Press, Princeton, NJ. Transparency: A Cross-Country Comparison.” IMF Working Paper 12/39, International Monetary Fund, Sachs, J., R. Cooper, and B. Bosworth. 1998. “The Washington, DC. East Asian Financial Crisis: Diagnosis, Remedies, Prospects.” Brookings Papers on Economic Activity 1998 World Bank. 2013. Europe and Central Asia: Sovereign (1): 1-90. Debt Management in Crisis: A Toolkit for Policymakers. Washington, DC: World Bank. Sahay, R., M. Čihák, P. N’Diaye, A. Barajas, R. Bi, D. Ayala, Y. Gao, et al. 2008. “Rethinking Financial ———. 2015. Global Economic Prospects Report: Deepening: Stability and Growth in Emerging Having Fiscal Space and Using It. January. Washington, Markets.” Staff Discussion Note 5/08, International DC: World Bank. Monetary Fund, Washington, DC. Schaechter, A., T. Kinda, N. T. Budina, and A. Weber. ———. 2016a. Global Economic Prospects: Divergences and Risks. June. Washington, DC: World Bank. 2012. “Fiscal Rules in Response to the Crisis-Toward the 'Next-Generation' Rules: A New Dataset." IMF ———. 2016b. Global Economic Prospects: Spillovers Working Paper 12/187, International Monetary Fund, Amid Weak Growth. January. Washington, DC: World Washington, DC. Bank. Shin, H. 2014. “The Second Phase of Global Liquidity and Its Impact on Emerging Economies.” In Volatile ———. 2017a. “Heavily Indebted Poor Countries Capital Flows in Korea: Current Policies and Future (HIPC) Initiative and Multilateral Debt Relief Responses, edited by K. Chung, S. Kim, H. Park, C. Initiative (MDRI) - Statistical Update.” World Bank, Choi, and H. S. Shin, 247-257. London: Palgrave Washington, DC. Macmillan. ———. 2017b. Tax Revenue Mobilization: Lessons from Tornell, A., A. Velasco, and J. Sachs. 1996. “Financial World Bank Group Support for Tax Reform. Crises in Emerging Markets: The Lessons from Washington, DC: World Bank. G LO BAL EC O NO MIC P ROS P EC TS | J AN U ARY 2020 CHAPTER 4 297 ———. 2018a. Global Economic Prospects: Broad-Based World Bank and IMF (International Monetary Fund). Upturn, but for How Long? January. Washington, DC: 2001. Developing Government Bond Markets: A World Bank. Handbook. Washington, DC: World Bank. ———. 2018b. Global Economic Prospects: The ———. 2014. “Revised Guidelines for Public Debt Turning of the Tide. June. Washington, DC: World Management.” World Bank, Washington, DC. Bank. ———. 2018a. Debt Vulnerabilities in Emerging and ———. 2019a. Doing Business 2019: Training for Low-Income Economies. October 2018 Meeting of the Reform. Washington, DC: World Bank. Development Committee. Washington DC: World Bank. ———. 2019b. MENA Economic Update: Reforms and External Imbalances: The Labor-Productivity Connection ———. 2018b. “G-20 Note: Improving Public Debt in the Middle East and North Africa. Washington, DC: Recording, Monitoring and Reporting Capacity in Low World Bank. and Lower Middle-Income Countries: Proposed Reforms.” World Bank, Washington, DC. ———. 2019c. Global Financial Development Report 2019/2020: Bank Regulation and Supervision a Decade Yu, B. and C. Shen. 2019. “Environmental Regulation after the Global Financial Crisis. Washington, DC: and Industrial Capacity Utilization: An Empirical World Bank. Study of China.” Journal of Cleaner Production. Advance online publication. https://www.science ———. 2019d. Global Economic Prospects: Heightened direct.com/science/article/pii/S0959652619338569. Tensions, Subdued Investment. June. Washington, DC: World Bank. STATISTICAL APPENDIX G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 S T A TI S T I C A L A P P E N D IX 301 Real GDP growth Annual estimates and forecasts1 Quarterly estimates2 (Percent change) (Percent change, year-on-year) 2017 2018 2019e 2020f 2021f 2022f 18Q2 18Q3 18Q4 19Q1 19Q2 19Q3e World 3.2 3.0 2.4 2.5 2.6 2.7 3.1 2.8 2.5 2.4 2.3 .. Advanced economies 2.4 2.2 1.6 1.4 1.5 1.5 2.4 2.0 1.7 1.8 1.6 1.7 United States 2.4 2.9 2.3 1.8 1.7 1.7 3.2 3.1 2.5 2.7 2.3 2.1 Euro Area 2.5 1.9 1.1 1.0 1.3 1.3 2.2 1.6 1.2 1.4 1.2 1.2 Japan 1.9 0.8 1.1 0.7 0.6 0.4 1.0 -0.3 -0.3 0.8 0.8 1.9 Emerging market and developing economies 4.5 4.3 3.5 4.1 4.3 4.4 4.5 4.3 4.0 3.6 3.6 .. East Asia and Pacific 6.5 6.3 5.8 5.7 5.6 5.6 6.4 6.2 6.1 6.1 5.9 5.7 Cambodia 7.0 7.5 7.0 6.8 6.8 6.8 .. .. .. .. .. .. China 6.8 6.6 6.1 5.9 5.8 5.7 6.7 6.5 6.4 6.4 6.2 6.0 Fiji 5.2 4.2 1.0 1.7 2.9 3.0 .. .. .. .. .. .. Indonesia 5.1 5.2 5.0 5.1 5.2 5.2 5.3 5.2 5.2 5.1 5.1 5.0 Lao PDR 6.9 6.3 5.2 5.8 5.7 5.6 .. .. .. .. .. .. Malaysia 5.7 4.7 4.6 4.5 4.5 4.5 4.5 4.4 4.7 4.5 4.9 4.4 Mongolia 5.3 7.2 5.7 5.5 5.2 5.5 5.3 4.5 9.1 8.6 6.2 4.2 Myanmar 6.8 6.5 6.6 6.7 6.8 6.8 .. .. .. .. .. .. Papua New Guinea 3.5 -0.8 5.6 2.9 2.9 3.0 .. .. .. .. .. .. Philippines 6.7 6.2 5.8 6.1 6.2 6.2 6.2 6.0 6.3 5.6 5.5 6.2 Solomon Islands 3.0 3.5 2.9 2.8 2.8 2.7 .. .. .. .. .. .. Thailand 4.0 4.1 2.5 2.7 2.8 2.9 4.7 3.2 3.6 2.8 2.3 2.4 Timor-Leste -3.5 -1.1 4.2 4.6 4.9 5.0 .. .. .. .. .. .. Vietnam 6.8 7.1 6.8 6.5 6.5 6.4 6.7 6.8 7.3 6.8 6.7 7.3 Europe and Central Asia 4.1 3.2 2.0 2.6 2.9 2.9 3.9 3.0 1.8 1.1 1.3 .. Albania 3.8 4.1 2.9 3.4 3.6 3.5 4.3 4.7 3.3 2.4 2.3 .. Armenia 7.5 5.2 6.9 5.1 5.2 5.2 .. .. .. .. .. .. Azerbaijan -0.3 1.4 2.5 2.3 2.1 2.1 .. .. .. .. .. .. Belarus 2.5 3.0 1.0 0.9 0.5 0.5 3.9 2.2 1.3 1.3 0.5 .. Bosnia and Herzegovina 3.2 3.6 3.1 3.4 3.9 3.9 3.9 3.1 3.9 2.8 2.6 .. Bulgaria 3.5 3.1 3.6 3.0 3.1 3.1 3.0 3.3 3.0 4.5 3.8 3.1 Croatia 3.1 2.7 2.9 2.6 2.4 2.4 3.2 3.0 2.2 4.1 2.4 2.9 Georgia 4.8 4.8 5.2 4.3 4.5 4.5 5.4 3.6 5.1 5.0 4.6 5.8 Hungary 4.3 5.1 4.9 3.0 2.6 2.6 5.0 5.3 5.3 5.3 4.9 5.0 Kazakhstan 4.1 4.1 4.0 3.7 3.9 3.7 4.3 3.9 4.1 3.8 4.4 .. Kosovo 4.2 3.8 4.0 4.2 4.1 4.0 .. .. .. .. .. .. Kyrgyz Republic 4.7 3.5 4.2 4.0 4.0 4.2 .. .. .. .. .. .. Moldova 4.7 4.0 3.6 3.6 3.8 3.8 5.3 3.3 3.8 4.4 5.8 4.3 Montenegro5 4.7 5.1 3.0 3.1 2.8 3.2 .. .. .. .. .. .. North Macedonia 0.2 2.9 3.1 3.2 3.3 3.1 1.7 2.4 6.2 3.9 3.4 3.6 Poland 4.9 5.1 4.3 3.6 3.3 3.1 5.4 5.5 4.4 4.6 4.1 4.1 Romania 7.1 4.0 3.9 3.4 3.1 3.1 3.8 4.0 4.2 5.0 4.4 3.0 Russia 1.6 2.3 1.2 1.6 1.8 1.8 2.2 2.2 2.7 0.5 0.9 1.7 Serbia 2.0 4.4 3.3 3.9 4.0 4.0 5.0 4.2 3.5 2.7 2.9 4.8 Tajikistan 7.1 7.3 6.2 5.5 5.0 5.0 .. .. .. .. .. .. Turkey 7.5 2.8 0.0 3.0 4.0 4.0 5.6 2.3 -2.8 -2.3 -1.6 0.9 Turkmenistan 6.5 6.2 5.0 5.2 5.5 5.5 .. .. .. .. .. .. Ukraine 2.5 3.3 3.6 3.7 4.2 4.2 3.8 2.8 3.5 2.5 4.6 4.1 Uzbekistan 4.5 5.1 5.5 5.7 6.0 6.0 .. .. .. .. .. .. 302 S T A TI S T I C A L A P P E N D IX G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 Real GDP growth (continued) Annual estimates and forecasts1 Quarterly estimates2 (Percent change) (Percent change, year-on-year) 2017 2018 2019e 2020f 2021f 2022f 18Q2 18Q3 18Q4 19Q1 19Q2 19Q3e Latin America and the Caribbean 1.9 1.7 0.8 1.8 2.4 2.6 1.8 1.6 1.1 0.6 0.7 .. Argentina 2.7 -2.5 -3.1 -1.3 1.4 2.3 -3.8 -3.6 -6.1 -5.8 0.0 -1.7 Belize 1.9 2.1 2.7 2.1 1.8 1.8 .. .. .. .. .. .. Bolivia 4.2 4.2 2.2 3.0 3.2 3.4 4.8 4.0 3.3 3.4 2.8 .. Brazil 1.3 1.3 1.1 2.0 2.5 2.4 1.1 1.5 1.2 0.6 1.1 1.2 Chile 1.3 4.0 1.3 2.5 3.0 3.0 5.3 2.6 3.6 1.5 1.9 3.3 Colombia 1.4 2.6 3.3 3.6 3.9 3.9 2.9 2.6 2.7 3.2 3.0 3.3 Costa Rica 3.4 2.6 2.0 2.5 3.0 3.2 3.8 2.6 1.3 2.2 0.6 .. Dominican Republic 4.7 7.0 5.3 5.0 5.0 5.0 7.3 7.6 6.3 5.6 3.7 .. Ecuador 2.4 1.4 -0.3 0.2 0.8 1.2 1.4 1.5 0.8 0.6 0.3 .. El Salvador 2.3 2.5 2.4 2.5 2.5 2.5 2.9 2.2 2.2 2.5 1.8 .. Grenada 4.4 4.2 3.5 2.9 2.9 3.2 .. .. .. .. .. .. Guatemala 2.8 3.1 3.4 3.0 3.2 3.2 3.6 3.6 3.5 3.1 3.5 .. Guyana 2.1 4.1 4.5 86.7 10.5 14.6 .. .. .. .. .. .. Haiti3 1.2 1.5 -0.9 -1.4 -0.5 1.4 .. .. .. .. .. .. Honduras 4.8 3.7 3.3 3.5 3.5 3.5 4.0 3.3 4.5 2.8 1.7 2.6 Jamaica2 1.0 1.9 1.0 1.1 1.2 2.0 2.2 1.9 2.0 1.8 1.3 .. Mexico 2.1 2.1 0.0 1.2 1.8 2.3 3.0 2.5 1.4 1.2 -0.9 -0.3 Nicaragua 4.7 -3.8 -5.0 -0.5 0.6 1.0 -5.2 -4.4 -7.7 .. .. .. Panama 5.6 3.7 3.5 4.2 4.6 4.8 3.0 3.3 4.2 3.1 2.9 2.7 Paraguay 5.0 3.7 0.7 3.1 3.9 3.8 6.9 1.6 1.0 -2.1 -3.0 .. Peru 2.5 4.0 2.6 3.2 3.5 3.6 5.4 2.5 4.8 2.4 1.2 3.0 St. Lucia 2.6 0.9 1.8 3.2 3.0 2.4 .. .. .. .. .. .. St. Vincent and the Grenadines 1.0 2.2 2.3 2.3 2.3 2.3 .. .. .. .. .. .. Suriname 1.8 2.6 2.2 2.5 2.1 2.1 .. .. .. .. .. .. Uruguay 2.6 1.6 0.5 2.5 3.5 3.2 2.2 1.8 0.6 -0.5 0.0 0.9 Middle East and North Africa 1.1 0.8 0.1 2.4 2.7 2.8 2.4 2.8 3.5 2.4 1.7 .. Algeria 1.3 1.4 1.3 1.9 2.2 2.2 .. .. .. .. .. .. Bahrain 3.8 2.2 2.0 2.1 2.4 2.4 2.1 1.9 5.3 2.6 0.8 .. Djibouti 5.1 5.5 7.2 7.5 8.0 8.4 .. .. .. .. .. .. Egypt3 4.2 5.3 5.6 5.8 6.0 6.0 5.4 5.3 5.5 5.6 5.7 5.6 Iran 3.8 -4.9 -8.7 0.0 1.0 1.0 2.5 .. .. .. .. .. Iraq -2.5 -0.6 4.8 5.1 2.7 2.5 .. .. .. .. .. .. Jordan 2.1 1.9 2.0 2.2 2.4 2.5 .. .. .. .. .. .. Kuwait -3.5 1.2 0.4 2.2 2.0 2.0 0.6 2.7 2.0 0.9 0.4 .. Lebanon 0.6 0.2 -0.2 0.3 0.4 0.5 .. .. .. .. .. .. Morocco 4.2 3.0 2.7 3.5 3.6 3.8 .. .. .. .. .. .. Oman 0.3 1.8 0.0 3.7 4.3 4.3 .. .. .. .. .. .. Qatar 1.6 1.5 0.5 1.5 3.2 3.2 1.9 1.5 0.5 0.8 -1.4 .. Saudi Arabia -0.7 2.4 0.4 1.9 2.2 2.4 1.6 2.4 4.3 1.7 0.5 .. Tunisia 1.8 2.5 1.6 2.2 2.6 2.6 .. .. .. .. .. .. United Arab Emirates 0.5 1.7 1.8 2.6 3.0 3.0 .. .. .. .. .. .. West Bank and Gaza 3.1 0.9 0.5 2.5 2.6 2.7 .. .. .. .. .. .. G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 S T A TI S T I C A L A P P E N D IX 303 Real GDP growth (continued) Annual estimates and forecasts1 Quarterly estimates2 (Percent change) (Percent change, year-on-year) 2017 2018 2019e 2020f 2021f 2022f 18Q2 18Q3 18Q4 19Q1 19Q2 19Q3e South Asia 6.7 7.1 4.9 5.5 5.9 6.0 7.8 6.9 6.4 5.8 4.9 4.5 Afghanistan 2.7 1.8 2.5 3.0 3.5 3.5 .. .. .. .. .. .. Bangladesh3,4 7.3 7.9 8.1 7.2 7.3 7.3 .. .. .. .. .. .. Bhutan3,4 6.3 3.8 3.9 5.6 7.6 6.2 .. .. .. .. .. .. India 3,4 7.2 6.8 5.0 5.8 6.1 6.1 8.0 7.0 6.6 5.8 5.0 4.5 Maldives 6.9 6.7 5.2 5.5 5.6 5.6 .. .. .. .. .. .. Nepal3,4 8.2 6.7 7.1 6.4 6.5 6.6 .. .. .. .. .. .. Pakistan3,4 5.2 5.5 3.3 2.4 3.0 3.9 .. .. .. .. .. .. Sri Lanka 3.4 3.2 2.7 3.3 3.7 3.7 3.9 3.5 1.8 3.7 1.5 2.7 Sub-Saharan Africa 2.7 2.6 2.4 2.9 3.1 3.3 2.0 2.7 2.8 2.2 2.5 .. Angola -0.1 -1.2 -0.7 1.5 2.4 3.0 .. .. .. .. .. .. Benin 5.8 6.7 6.4 6.7 6.7 6.7 .. .. .. .. .. .. Botswana 2.9 4.5 4.0 4.1 4.2 4.2 5.3 4.1 4.2 4.3 3.1 3.1 Burkina Faso 6.3 6.8 6.0 6.0 6.0 6.0 .. .. .. .. .. .. Burundi 0.5 1.6 1.8 2.0 2.1 2.2 .. .. .. .. .. .. Cabo Verde 3.7 5.1 5.0 5.0 5.0 5.0 .. .. .. .. .. .. Cameroon 3.5 4.1 4.0 4.2 4.3 4.5 .. .. .. .. .. .. Chad -3.0 2.6 3.0 5.5 4.8 4.8 .. .. .. .. .. .. Comoros 3.8 3.4 1.7 4.8 3.7 3.6 .. .. .. .. .. .. Congo, Dem. Rep. 3.7 5.8 4.3 3.9 3.4 3.6 .. .. .. .. .. .. Congo, Rep. -1.8 1.6 2.2 4.6 1.9 2.4 .. .. .. .. .. .. Côte d'Ivoire 7.7 7.4 7.3 7.0 7.1 7.1 .. .. .. .. .. .. Equatorial Guinea -4.7 -6.1 -4.3 -2.3 1.0 -4.8 .. .. .. .. .. .. Eswatini 2.0 2.4 1.3 2.6 2.5 2.5 .. .. .. .. .. .. Ethiopia 3 10.0 7.9 9.0 6.3 6.4 7.1 .. .. .. .. .. .. Gabon 0.5 0.8 2.9 3.0 3.2 3.3 .. .. .. .. .. .. Gambia, The 4.8 6.6 6.0 6.3 5.8 5.5 .. .. .. .. .. .. Ghana 8.1 6.3 7.0 6.8 5.2 4.6 5.4 7.4 6.8 6.7 5.7 5.6 Guinea 10.0 5.8 5.9 6.0 6.0 6.0 .. .. .. .. .. .. Guinea-Bissau 5.9 3.8 4.6 4.9 5.0 5.0 .. .. .. .. .. .. Kenya 4.9 6.3 5.8 6.0 5.8 5.8 6.4 6.4 5.9 5.6 5.6 .. Lesotho -0.4 1.5 2.6 0.7 2.1 2.8 0.2 -3.1 3.0 -1.4 1.2 .. Liberia 2.5 1.2 -1.4 1.4 3.4 4.2 .. .. .. .. .. .. Madagascar 4.3 5.1 4.7 5.3 4.4 5.0 .. .. .. .. .. .. Malawi 4.0 3.5 4.4 4.8 5.2 5.3 .. .. .. .. .. .. Mali 5.3 4.7 5.0 5.0 4.9 4.9 .. .. .. .. .. .. Mauritania 3.0 3.6 6.4 5.7 5.8 8.7 .. .. .. .. .. .. Mauritius 3.8 3.8 3.9 3.9 4.0 4.0 .. .. .. .. .. .. Mozambique 3.7 3.4 2.0 3.7 4.2 4.4 .. .. .. .. .. .. Namibia -0.9 -0.1 -0.5 0.9 1.7 1.9 .. .. .. .. .. .. Niger 4.9 6.5 6.3 6.0 5.6 11.9 .. .. .. .. .. .. Nigeria 0.8 1.9 2.0 2.1 2.1 2.1 1.5 1.8 2.4 2.1 2.1 2.1 Rwanda 6.1 8.6 8.5 8.1 8.0 8.0 .. .. .. .. .. .. Senegal 7.1 6.8 6.3 6.8 7.0 7.0 .. .. .. .. .. .. Seychelles 4.3 4.1 3.5 3.3 3.3 3.4 .. .. .. .. .. .. Sierra Leone 3.8 3.5 4.8 4.9 4.9 5.0 .. .. .. .. .. .. 304 S T A TI S T I C A L A P P E N D IX G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 Real GDP growth (continued) Annual estimates and forecasts1 Quarterly estimates2 (Percent change) (Percent change, year-on-year) 2017 2018 2019e 2020f 2021f 2022f 18Q2 18Q3 18Q4 19Q1 19Q2 19Q3e Sub-Saharan Africa (continued) South Africa 1.4 0.8 0.4 0.9 1.3 1.5 0.1 1.3 1.1 0.0 0.9 0.1 Sudan 4.3 -2.3 -2.6 -1.4 -0.6 0.2 .. .. .. .. .. .. Tanzania 6.8 5.4 5.6 5.8 6.1 6.2 6.1 7.1 7.1 6.5 7.2 .. Togo 4.4 4.9 5.3 5.5 5.5 5.5 .. .. .. .. .. .. Uganda3 3.9 5.9 6.1 6.5 5.9 6.0 5.0 6.5 6.8 5.6 5.4 .. Zambia 4.1 3.1 1.8 2.6 2.6 4.0 4.7 6.0 2.5 2.3 2.2 .. Zimbabwe 4.7 3.5 -7.5 2.7 2.5 2.8 .. .. .. .. .. .. World Bank and Haver Analytics. Note: e = estimate; f = forecast. 1. Aggregate growth rates calculated using GDP weights at 2010 prices and market exchange rates. 2. Quarterly estimates are based on non-seasonally-adjusted real GDP, except for advanced economies, as well as Ecuador and Poland. Data for Bosnia and Herzegovina are from the production approach. Quarterly data for Jamaica are gross value added. Regional averages are calculated based on data from following countries. East Asia and Pacific: China, Indonesia, Malaysia, Mongolia, Philippines, Thailand, and Vietnam. Europe and Central Asia: Albania, Belarus, Bosnia and Herzegovina, Bulgaria, Croatia, Georgia, Hungary, Kazakhstan, North Macedonia, Poland, Romania, Russia, Serbia, Turkey, and Ukraine. Latin America and the Caribbean: Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, Jamaica, Mexico, Nicaragua, Panama, Paraguay, Peru, and Uruguay. Middle East and North Africa: Bahrain, Egypt, Iran, Kuwait, Qatar, and Saudi Arabia. South Asia: India and Sri Lanka. Sub-Saharan Africa: Botswana, Ghana, Kenya, Nigeria, South Africa, Tanzania, Uganda, and Zambia. 3. Annual GDP is on fiscal year basis, as per reporting practice in the country. 4. GDP data for Pakistan are based on factor cost. For Bangladesh, Bhutan, Nepal, and Pakistan, the column labeled 2019 refers to FY2018/19. For India, the column labeled 2018 refers to FY2018/19. 5. Quarterly data are preliminary. Click here to download data. G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 S T A TI S T I C A L A P P E N D IX 305 Data and Forecast Conventions The macroeconomic forecasts presented in this ments Statistics, and IMF International Financial report are prepared by staff of the Prospects Statistics. Group of the Development Economics Vice- Presidency, in coordination with staff from the Aggregations. Aggregate growth for the world and Macroeconomics, Trade, and Investment Global all sub-groups of countries (such as regions and Practice and from regional and country offices, income groups) is calculated as GDP-weighted and with input from regional Chief Economist average (at 2010 prices) of country-specific growth offices. They are the result of an iterative process rates. Income groups are defined as in the World that incorporates data, macroeconometric models, Bank’s classification of country groups. and judgment. Forecast Process. The process starts with initial Data. Data used to prepare country forecasts assumptions about advanced-economy growth and come from a variety of sources. National Income commodity price forecasts. These are used as Accounts (NIA), Balance of Payments (BOP), and conditioning assumptions for the first set of fiscal data are from Haver Analytics; the World growth forecasts for EMDEs, which are produced Development Indicators by the World Bank; the using macroeconometric models, accounting World Economic Outlook, Balance of Payments frameworks to ensure national account identities Statistics, and International Financial Statistics by and global consistency, estimates of spillovers the International Monetary Fund. Population from major economies, and high-frequency data and forecasts are from the United Nations indicators. These forecasts are then evaluated to World Population Prospects. Country- and ensure consistency of treatment across similar lending-group classifications are from the World EMDEs. This is followed by extensive discussions Bank. DECPG databases include commodity with World Bank country teams, who conduct prices, data on previous forecast vintages, and in- continuous macroeconomic monitoring and house country classifications. Other internal dialogue with country authorities and finalize databases include high-frequency indicators such growth forecasts for EMDEs. The Prospects as industrial production, consumer price indexes, Group prepares advanced-economy and com- house prices, exchange rates, exports, imports, and modity price forecasts. Throughout the forecasting stock market indexes, based on data from process, staff use macro-econometric models that Bloomberg, Haver Analytics, OECD Analytical allow the combination of judgement and House Prices Indicators, IMF Balance of Pay- consistency with model-based insights. 306 S E L E CTE D TO P I C S G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 Global Economic Prospects: Selected Topics, 2015-20 Growth and Business Cycles Informality Growing in the shadow: Challenges of informality January 2019, Chapter 3 Linkages between formal and informal sectors January 2019, Box 3.1 Regional dimensions of informality: An overview January 2019, Box 3.2 Casting a shadow: Productivity in formal and informal firms January 2019, Box 3.3 Under the magnifying glass: How do policies affect informality? January 2019, Box 3.4 East Asia and Pacific January 2019, Box 2.1.1 Europe and Central Asia January 2019, Box 2.2.1 Latin America and the Caribbean January 2019, Box 2.3.1 Middle East and North Africa January 2019, Box 2.4.1 South Asia January 2019, Box 2.5.1 Sub-Saharan Africa January 2019, Box 2.6.1 Inflation Low for how much longer? Inflation in low-income countries January 2020, Special Focus 2 Currency depreciation, inflation, and central bank independence June 2019, Special Focus 1.2 The great disinflation January 2019, Box 1.1 Growth prospects Growth in low-income countries: Evolution, prospects, and policies June 2019, Special Focus 2.1 Long-term growth prospects: Downgraded no more? June 2018, Box 1.1 Global output gap Is the global economy turning the corner? January 2018, Box 1.1 Potential growth Building solid foundations: How to promote potential growth January 2018, Chapter 3 What is potential growth? January 2018, Box 3.1 Understanding the recent productivity slowdown: Facts and explanations January 2018, Box 3.2 Moving together? Investment and potential output January 2018, Box 3.3 The long shadow of contractions over potential output January 2018, Box 3.4 Productivity and investment growth during reforms January 2018, Box 3.5 East Asia and Pacific January 2018, Box 2.1.1 Europe and Central Asia January 2018, Box 2.2.1 Latin America and the Caribbean January 2018, Box 2.3.1 Middle East and North Africa January 2018, Box 2.4.1 South Asia January 2018, Box 2.5.1 Sub-Saharan Africa January 2018, Box 2.6.1 Productivity Fading promise: How to rekindle productivity growth January 2020, Chapter 3 EMDE regional productivity trends and bottlenecks January 2020, Box 3.1 Sectoral sources of productivity growth January 2020, Box 3.2 Patterns of total factor productivity: a firm perspective January 2020, Box 3.3 Debt, financial crises, and productivity January 2020, Box 3.4 Labor productivity in East Asia and Pacific: Trends and drivers January 2020, Box 2.1.1 Labor productivity in Europe and Central Asia: Trends and drivers January 2020, Box 2.2.1 Labor productivity in Latin America and the Caribbean: Trends and drivers January 2020, Box 2.3.1 Labor productivity in Middle East and North Africa: Trends and drivers January 2020, Box 2.4.1 Labor productivity in South Asia: Trends and drivers January 2020, Box 2.5.1 Labor productivity in Sub-Saharan Africa: Trends and drivers January 2020, Box 2.6.1 G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 S E L E CTE D TO P I C S 307 Global Economic Prospects: Selected Topics, 2015-20 Growth and Business Cycles Investment slowdown Investment: Weak prospects, strong needs June 2019, Special Focus 1.1 Weak investment in uncertain times: Causes, implications and policy responses January 2017, Chapter 3 Investment-less credit booms January 2017, Box 3.1 Implications of rising uncertainty for investment in EMDEs January 2017, Box 3.2 Investment slowdown in China January 2017, Box 3.3 Interactions between public and private investment January 2017, Box 3.4 Investment slowdown (continued) East Asia and Pacific January 2017, Box 2.1.1 Europe and Central Asia January 2017, Box 2.2.1 Latin America and the Caribbean January 2017, Box 2.3.1 Middle East and North Africa January 2017, Box 2.4.1 South Asia January 2017, Box 2.5.1 Sub-Saharan Africa January 2017, Box 2.6.1 Forecast uncertainty Quantifying uncertainties in global growth forecasts June 2016, Special Focus 2 Cross-border spillovers Who catches a cold when emerging markets sneeze? January 2016, Chapter 3 Sources of the growth slowdown in BRICS January 2016, Box 3.1 Understanding cross-border growth spillovers January 2016, Box 3.2 Within-region spillovers January 2016, Box 3.3 East Asia and Pacific January 2016, Box 2.1.1 Europe and Central Asia January 2016, Box 2.2.1 Latin America and the Caribbean January 2016, Box 2.3.1 Middle East and North Africa January 2016, Box 2.4.1 South Asia January 2016, Box 2.5.1 Sub-Saharan Africa January 2016, Box 2.6.1 Fiscal space Having space and using it: Fiscal policy challenges and developing economies January 2015, Chapter 3 Fiscal policy in low-income countries January 2015, Box 3.1 What affects the size of fiscal multipliers? January 2015, Box 3.2 Chile’s fiscal rule—an example of success January 2015, Box 3.3 Narrow fiscal space and the risk of a debt crisis January 2015, Box 3.4 Revenue mobilization in South Asia: Policy challenges and recommendations January 2015, Box 2.3 Other topics Education demographics and global inequality January 2018, Special Focus 2 Recent developments in emerging and developing country labor markets June 2015, Box 1.3 Linkages between China and Sub-Saharan Africa June 2015, Box 2.1 What does weak growth mean for poverty in the future? January 2015, Box 1.1 What does a slowdown in China mean for Latin America and the Caribbean? January 2015, Box 2.2 Commodity Markets The role of major emerging markets in global commodity demand June 2018, Special Focus 1 The role of the EM7 in commodity production June 2018, SF1, Box SF1.1 Commodity consumption: Implications of government policies June 2018, SF1, Box SF1.2 With the benefit of hindsight: The impact of the 2014–16 oil price collapse January 2018, Special Focus 1 From commodity discovery to production: Vulnerabilities and policies in LICs January 2016, Special Focus After the commodities boom: What next for low-income countries? June 2015, Special Focus 2 Low oil prices in perspective June 2015, Box 1.2 Understanding the plunge in oil prices: Sources and implications January 2015, Chapter 4 What do we know about the impact of oil prices on output and inflation? A brief survey January 2015, Box 4.1 308 S E L E CTE D TO P I C S G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 Global Economic Prospects: Selected Topics, 2015-20 Globalization of Trade and Financial Flows Poverty impact of food price shocks and policies January 2019, Chapter 4 Arm’s-length trade: A source of post-crisis trade weakness June 2017, Special Focus 2 The U.S. economy and the world January 2017, Special Focus Potential macroeconomic implications of the Trans-Pacific Partnership Agreement January 2016, Chapter 4 Regulatory convergence in mega-regional trade agreements January 2016, Box 4.1.1 China’s integration in global supply chains: Review and implications January 2015, Box 2.1 Can remittances help promote consumption stability? January 2015, Chapter 4 What lies behind the global trade slowdown? January 2015, Chapter 4 Monetary and Exchange Rate Policies The fourth wave: Rapid debt buildup January 2020, Chapter 4 Price controls: Good intentions, bad outcomes January 2020, Special Focus 1 Low for how much longer? Inflation in low-income countries January 2020, Special Focus 2 Currency depreciation, inflation, and central bank independence June 2019, Special Focus 1.2 The great disinflation January 2019, Box 1.1 Corporate debt: Financial stability and investment implications June 2018, Special Focus 2 Recent credit surge in historical context June 2016, Special Focus 1 Peg and control? The links between exchange rate regimes and capital account policies January 2016, Chapter 4 Negative interest rates in Europe: A glance at their causes and implications June 2015, Box 1.1 Hoping for the best, preparing for the worst: Risks around U.S. rate liftoff and policy options June 2015, Special Focus 1 Countercyclical monetary policy in emerging markets: Review and evidence January 2015, Box 1.2 Fiscal Policies The fourth wave: Rapid debt buildup January 2020, Chapter 4 Debt: No free lunch June 2019, Box 1.1 Debt in low-income countries: Evolution, implications, and remedies January 2019, Chapter 4 Debt dynamics in emerging market and developing economies: Time to act? June 2017, Special Focus 1 Having fiscal space and using it: Fiscal challenges in developing economies January 2015, Chapter 3 Revenue mobilization in South Asia: Policy challenges and recommendations January 2015, Box 2.3 Fiscal policy in low-income countries January 2015, Box 3.1 What affects the size of fiscal multipliers? January 2015, Box 3.2 Chile’s fiscal rule—an example of success January 2015, Box 3.3 Narrow fiscal space and the risk of a debt crisis January 2015, Box 3.4 G L O B A L E CO N O MI C P R OS P E C TS | J A N U A R Y 20 2 0 S E L E CTE D TO P I C S 309 Prospects Group: Selected Other Publications on the Global Economy, 2015-20 Commodity Markets Outlook Column1 Food price shocks: Channels and implications April 2019, Special Focus The implications of tariffs for commodity markets October 2018, Box The changing of the guard: Shifts in industrial commodity demand October 2018, Special Focus Oil exporters: Policies and challenges April 2018, Special Focus Investment weakness in commodity exporters January 2017, Special Focus OPEC in historical context: Commodity agreements and market fundamentals October 2016, Special Focus Energy and food prices: Moving in tandem? July 2016, Special Focus Resource development in an era of cheap commodities April 2016, Special Focus Weak growth in emerging market economies: What does it imply for commodity markets? January 2016, Special Focus Understanding El Niño: What does it mean for commodity markets? October 2015, Special Focus How important are China and India in global commodity consumption? July 2015, Special Focus Anatomy of the last four oil price crashes April 2015, Special Focus Putting the recent plunge in oil prices in perspective January 2015, Special Focus Inflation in Emerging and Developing Economies Inflation: Concepts, evolution, and correlates Chapter 1 Understanding global inflation synchronization Chapter 2 Sources of inflation: Global and domestic drivers Chapter 3 Inflation expectations: Review and evidence Chapter 4 Inflation and exchange rate pass-through Chapter 5 Inflation in low-income countries Chapter 6 Poverty impact of food price shocks and policies Chapter 7 A Decade After the Global Recession: Lessons and Challenges for Emerging and Developing Economies A Decade After the Global Recession: Lessons and Challenges Chapter 1 What Happens During Global Recessions? Chapter 2 Macroeconomic Developments Chapter 3 Financial Market Developments Chapter 4 Macroeconomic and Financial Sector Policies Chapter 5 Prospects, Risks, and Vulnerabilities Chapter 6 Policy Challenges Chapter 7 The Role of the World Bank Group Chapter 8 Global Waves of Debt: Causes and Consequences Debt: Evolution, Causes, and Consequences Chapter 1 Benefits and Costs of Debt: The Dose Makes the Poison Chapter 2 Global Waves of Debt: What Goes up Must Come Down? Chapter 3 The Fourth Wave: Ripple or Tsunami? Chapter 4 Debt and Financial Crises: From Euphoria to Distress Chapter 5 Policies: Turning Mistakes into Experience Chapter 6 High-Frequency Monitoring Column1 Global Monthly newsletter ECO-AUDIT Environmental Benefits Statement e World Bank Group is committed to reducing its environmental footprint. In support of this commitment, we leverage electronic publishing options and print-on-demand technology, which is located in regional hubs worldwide. Together, these initiatives enable print runs to be lowered and shipping distances decreased, resulting in reduced paper consumption, chemical use, greenhouse gas emissions, and waste. We follow the recommended standards for paper use set by the Green Press Initiative. e majority of our books are printed on Forest Stewardship Council (FSC)-certi ed paper, with nearly all containing 50-100 percent recycled content. e recycled ber in our book paper is either unbleached or bleached using totally chlorine-free (TCF), processed chlorine-free (PCF), or enhanced elemental chlorine-free (EECF) processes. More information about the Bank’s environmental philosophy can be found at http://www.worldbank.org/corporateresponsibility. G lobal growth is projected to be slightly faster in 2020 than the post-crisis low registered last year. While growth could be stronger if reduced trade tensions lead to a sustained reduction in uncertainty, the balance of risks to the outlook is to the downside. Growth in emerging market and developing economies is also expected to remain subdued, continuing a decade of disappointing outcomes. A steep and widespread productivity growth slowdown has been underway in these economies since the global financial crisis, despite the largest, fastest, and most broad-based accumulation of debt since the 1970s. In addition, many emerging market and developing economies, including low-income countries, face the challenge of phasing out price controls that impose heavy fiscal cost and dampen investment. These circumstances add urgency to the need to implement measures to rebuild macroeconomic policy space and to undertake reforms to rekindle productivity growth. These efforts need to be supplemented by policies to promote inclusive and sustainable long-term growth and accelerate poverty alleviation. Global Economic Prospects is a World Bank Group Flagship Report that examines global economic developments and prospects, with a special focus on emerging market and developing economies, on a semiannual basis (in January and June). The January edition includes in-depth analyses of topical policy challenges faced by these economies, while the June edition contains shorter analytical pieces. SKU 211468