103822 KENYA COUNTRY ECONOMIC MEMORANDUM From Economic Growth to Jobs and Shared Prosperity March 2016 Standard Disclaimer: This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank 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. Copyright Statement: The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, http://www.copyright.com All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. Cover Photo: Peter Ndung’u TABLE OF CONTENTS Abbreviations and Acronyms.....................................................................................................................i Acknowledgements...................................................................................................................................ii Foreword...................................................................................................................................................iii Executive Summary...................................................................................................................................V References................................................................................................................................................119 I. Kenya’s Growth Story............................................................................................................................1 II. From Economic Growth to Jobs and Shared Prosperity.........................................................................31 III. Raising Investment through Savings......................................................................................................59 IV. Manufacturing or Services: Where Does the Key to Rapid Growth Lie?.................................................73 V. Non-renewable Resources for Sustainable Development.......................................................................93 LIST OF FIGURES Figure ES.1: Kenya’s uneven growth performance .......................................................................................................vii Figure ES.2: Services have been the main generator of growth. ...................................................................................viii Figure ES.3: Consumption: Main contributor to GDP growth ......................................................................................ix Figure ES.4: Kenya’s GDP per capita has been growing slower than its peers .............................................................ix Figure ES.5: Macroeconomic stability has been maintained since 2012 .....................................................................x Figure ES.6: Kenya has among the highest youth unemployment rates, 2000–14 average .......................................xiii Figure ES.7: Kenya’s minimum wage is highest among its peers .................................................................................xvi Figure ES.8: Jobs were added in low-productivity sectors between 2009 and 2013 ..................................................xvii Figure ES.9: Kenya has witnessed rapid growth of formal business startups...............................................................xvii Figure ES.10: Within-sector differences in productivity are high among Kenyan firms ..............................................xviii Figure ES.11: Innovation Is widespread in Kenya by global standards .........................................................................xix Figure ES.12: Different development trajectories of oil rich countries: Angola, Malaysia, and Republic of Congo ...xx Figure 1.1: Kenya’s peers with similar GDP per capita ..................................................................................................1 Figure 1.2: Kenya’s economy had lower growth and higher volatility than its peers, 2003–14 .................................2 Figure 1.3: Services have been driving growth in Kenya . .............................................................................................2 Figure 1.4: Each of Kenya’s sectors has fared worse compared with peers, 2003–12.................................................4 Figure 1.5: Consumption has been driving growth while net exports have been a drag . ..........................................4 Figure 1.6: Kenya’s goods exports have been low and declining .................................................................................5 Figure 1.7: Several countries in SSA and East Asia have achieved Kenya’s desired pace of growth, 2000–10............6 Figure 1.8: The output gap has turned negative since 2008.........................................................................................7 Figure 1.9: Comparison of projected actual and potential GDP growth rates for 2014–18 .......................................7 Figure 1.10: Fiscal policy moved from being pro-cyclical to countercyclical in 2008 ..................................................8 Figure 1.11: Monetary policy has not been fully in sync with the real economy ........................................................9 Figure 1.12: Kenya has the highest secondary and lowest tertiary enrollment among the peers .............................11 Figure 1.13: Urbanization in Kenya has been low relative to GDP per capita . ............................................................12 Figure 1.15: The financial sector is more developed than in peers .............................................................................13 Figure 1.14: Manufacturing in Kenya Is underdeveloped compared with its non-Africa peers .................................13 Figure 1.16: Kenya has the lowest investment-to-GDP ratio and the highest investment risk, 2005–14 ..................14 Figure 1.17: Kenya’s trade openness remains resilient, although it is lower than that of its peers . ..........................14 Figure 1.18: Development spending is relatively high, 2005–2012 average ...............................................................15 Figure 1.19: BTI indicators of Governance in Kenya and benchmark countries ..........................................................17 Figure 1.20: Governance indicators show Kenya lags behind benchmarking economies ..........................................17 Figure 1.21: Perception of corruption in Kenya and the peer group, 2015 .................................................................19 Figure 1.22: Development in governance indicators, 1995–2013 ...............................................................................20 Figure 1.23: Kenya’s economic management ranks above its peers . ..........................................................................22 Figure 1.24: There is a clear shift in counties’ sector spending priorities ....................................................................25 Figure 1.25: The high inherited wage bill is crowding out development spending in most counties ........................27 Figure 1.26: Kisumu County’s new fees and charges have no coherent basis and are above all previous levels ......28 Figure 2.1: GDP growth has been driven by consumption ...........................................................................................34 Figure 2.2: Poverty has been on the decline in Kenya ..................................................................................................37 Figure 2.3: The majority of Kenya’s working poor live in rural areas ...........................................................................38 Figure 2.4: Estimated trends in the poverty rate under different scenarios between 2005 and 2020 ......................38 Figure 2.5: Formal employment, although desired by many, remains a privilege for a few........................................39 Figure 2.6: Jobs are created largely in informal trade and hospitality services ...........................................................40 Figure 2.7: Productivity growth is fastest in sectors with few workers ........................................................................40 Figure 2.8: Informal firms are mostly young .................................................................................................................42 Figure 2.9: Informal businesses employ few people and pay minimum wages or less ..............................................42 Figure 2.10: Most owners of informal enterprises are young and literate ..................................................................43 Figure 2.11: What is the largest obstacle faced by informal firms? .............................................................................43 Figure 2.12: Own funds constitute the main source of finance ...................................................................................43 Figure 2.13: Why do they choose to operate informally? Why aren’t they registered? .............................................44 Figure 2.14: Kenya is among the countries with restrictive product market regulations ...........................................46 Figure 2.15: Finding skilled workers is becoming a major challenge for employers ...................................................47 Figure 2.16: Minimum wage is highest in Kenya among peer countries .....................................................................50 Figure 2.17: EPZ exports rising but employment steady ..............................................................................................54 Figure 2.18: Relative contribution of apparel declining . ..............................................................................................54 Figure 2.19: Evolution of export growth in selected global SEZs from year of launch ................................................55 Figure 2.20: Monthly downtime caused .......................................................................................................................56 Figure 2.21: Days to clear imports from customs .........................................................................................................56 Figure 3.1: Investment and growth are highly correlated ............................................................................................60 Figure 3.2: Savings are correlated with investment and growth ..................................................................................60 Figure 3.3: Decade average savings rate: Kenya and its peer countries.......................................................................61 Figure 3.4: Kenya’s public savings is relatively low and declining . ...............................................................................62 Figure 3.5: Corporate savings have been increasing, 2005–13.....................................................................................63 Figure 3.6: Credit to households is growing rapidly ......................................................................................................63 Figure 3.7: Lower youth dependency ratio associated with higher savings, average, 1980–2013 ............................66 Figure 3.8: Kenya’s effective youth dependency ratio is much lower, 1975–2010 . ....................................................66 Figure 3.9: Real deposit savings rates have been negative for much of the past decade ...........................................67 Figure 3.10: Financial inclusion in Kenya is high relative to peer countries .................................................................72 Figure 4.1: Declining share of manufacturing in Kenya’s peer group . .........................................................................75 Figure 4.2: Export concentration trend .........................................................................................................................77 Figure 4.3: Kenya has witnessed rapid growth of formal business startups.................................................................78 Figure 4.4: Kenyan manufacturing firms tend to be capital intensive .........................................................................78 Figure 4.5: Capital-to-labor ratio dispersion is high across most sectors, 2010 ..........................................................79 Figure 4.6: Entrant firms are less productive.................................................................................................................79 Figure 4.7: Nairobi attracts new firms but more jobs are created in less urbanized areas .........................................80 Figure 4.8: Services exports are rising much faster than exports of goods..................................................................81 Figure 4.9: The contribution of services in countries’ exports is undervalued ............................................................82 Figure 4.10: The role of services as an input to other sectors’ exports is low in Kenya...............................................83 Figure 4.11: Direct and total value-added exports by sector........................................................................................83 Figure 4.12: Higher productivity service activities have higher average wages but employ few workers .................85 Figure 4.13: Entrant firms are more productive but pay less per worker ....................................................................86 Figure 4.14: Kenya does well in product and process innovation (percent of surveyed firms) ..................................88 Figure 4.15: Kenya is not spending enough on research and development ...............................................................89 Figure 4.16: Kenya’s main source of information for innovation is customer feedback and the internet .................90 Figure 4.17: Kenya has solid managerial capacity but still below the frontier ............................................................90 Figure 5.2: Estimates for oil production and fiscal revenues, 2020–75 .......................................................................94 Figure 5.1: Proven oil reserves by region/country, 2013...............................................................................................94 Figure 5.3: Oil prices have been particularly volatile since 2000 .................................................................................95 Figure 5.4: Different development trajectories: Angola, Democratic Republic of Congo and Malaysia, 1972–2010 .96 Figure 5.5: Public investment management efficiency index and sub-indexes (Selection from a sample of 71) ......100 Figure 5.6: Alternative expenditure scenarios ..............................................................................................................106 Figure 5.7: Investment composition scenarios .............................................................................................................108 Figure 5.8: Predicted output gaps in real time versus actual output gaps, 175 countries, 1990–2011 .....................115 LIST OF TABLES Table 1.1: Inflation, volatility of GDP growth, REER, and TOT movements in Kenya and peer economies . .....22 Table 1.2: Local revenue collection has improved ............................................................................................ 26 Table 2.1: Average annual growth rates to reach MTP-2 formal jobs target ..................................................... 41 Table 2.2: Private returns to tertiary education are high, 2006 (%) ................................................................. 49 Table 3.1: Saving behavior by groups, individuals ages 15+ (%) ....................................................................... 70 Table 3.2: The SACCO industry is growing rapidly............................................................................................. 71 Table 4.1: Top three most important upstream sectors for each downstream sector...................................... 80 Table 4.2. Innovation Activities, 2010–12.......................................................................................................... 89 Table 5.1: Predicted output gaps in real time versus actual output gaps, 175 countries, 1990–2011 .............117 LIST OF BOXES Box 1.1: Corruption and access to water . ........................................................................................................ 16 Box 2.1: From growth to shared prosperity—The different paths of Rwanda and Nigeria .............................32 Box 2.2: Despite challenges, there are still opportunities in the agriculture sector ........................................ 34 Box 2.3: Why agriculture will continue to matter for Kenya’s growth and poverty reduction . .......................35 Box 2.4: The changing context of Kenya’s rural labor market ......................................................................... 36 Box 2.5: Who can Kenya learn from about improving the business environment? ....................................... 45 Box 2.6: Examples of labor regulations that are causing firms to become risk averse in hiring ...................50 Box 2.7: Locking labor market entrants in low-productivity jobs limits their long-term earning potential . 51 Box 2.8: South Africa’s informal economy policy was born in eThekwini/Durban ........................................... 52 Box 3.1: Determinants of Savings...................................................................................................................... 65 Box 3.2. Kenya’s Financial Inclusion................................................................................................................... 72 Box 4.1. How is economic complexity measured and what does it (not) represent? .................................76 Box 4.2: Data source for analysis of manufacturing firms ................................................................................ 77 Box 4.3: Measuring the value added in exports a ....................................................................................... 82 Box 4.4: Tourism is Kenya’s leading service export—can it become even larger? ........................................... 84 Box 4.5: Integrated survey of services .............................................................................................................. 85 Box 5.1: Alternative approaches to scaling up public investments . ................................................................ 102 Box 5.2. Fiscal Rule Examples ......................................................................................................................... 112 ABBREVIATIONS AND ACRONYMS AGOA Africa Growth and Opportunity Act MTP-2 Second Medium-Term Plan AIC Aggressive Infrastructure-based Composition NBER National Bureau of Economic Research ASC Aggressive Skill-based Composition NFRK National Fund of the Republic of Kazakhstan BC Balanced Composition NSSF National Social Security Fund BIH Bird-in-hand OECD Organisation for Economic Co-operation and Development BPO Business Process Outsourcing OSHA Occupational Safety and Health Act BTI Bertelsmann Foundation Transformation Index PIH Permanent Income Hypothesis CPI Consumer Price Index PPP Purchasing Power Parity CPIA Country Policy and Institutional Assessment PWC PriceWaterhouseCoopers DSGE Dynamic Stochastic General Equilibrium R&D Research and Development EAC East African Community RBA Retirement Benefit Authority ECI Economic Complexity Index REER Real Effective Exchange Rate EITI Extractive Industries Transparency Initiative SACCO Savings and Credit Cooperative Organization EPZ Export Processing Zones SASRA SACCO Society Regulatory Authority ES-IM14 2014 Enterprise Survey Innovation Module SAYG Spend-as-you-go FDI Foreign Direct Investment SEZ Special Economic Zone Findex Financial Index SSA Sub-Saharan Africa GDP Gross Domestic Product SWF Sovereign Wealth Fund GNDI Gross National Disposable Income TFP Total Factor Productivity GNI Gross National Income TOT Terms of Trade GPNs Global Production Networks UMICs Upper-middle-income Countries IMF International Monetary Fund UN COMTRADE United Nations Commodity Trade Statistics Database KNBS Kenya National Bureau of Statistics UNPSA United Nations Public Service Awards MDG Millennium Development Goal VAR Vector Autoregression MF Moderate Frontloading WEO World Economic Outlook MFA Multi-Fiber Arrangement WIBA Work Injury Benefits Act MFN Most Favored Nation WTTC World Travel and Tourism Council M-Pesa Mobile Money WWII World War II FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y i ACKNOWLEDGEMENTS T his report is the outcome of the collaborative efforts of many. It was led by Borko Handjiski, Senior Economist, with supervision and direction from Apurva Sanghi, Lead Economist, throughout the preparation. The team comprised World Bank staff from various departments. Chapter 1 was written primarily by Jane Bogoev and Borko Handjiski, with contributions from George Larbi, Angelique Umutesi, John Randa, Jane Kiringai, Patrick Nderitu, and Kathy Whimp. Chapter 2 was prepared by Paul Gubbins, Johan Mistiaen, John Randa, Angelique Umutesi, Tom Farole, and Borko Handjiski. Chapter 3 was written by Toru Nishiuchi and Borko Handjiski. Bill Battaile, Ralph van Doorn, Sebastian Saez, Claire Hollweg, Xavier Cirera, Paulina Mogollon, Georgia Dowdall, and Borko Handjiski worked on Chapter 4. Harun Onder was the primary author of Chapter 5, with inputs from Paul Levine and Giovanni Melina. Angelique Umutesi provided statistical data and analysis on various aspects of the report. The team acknowledges contributions from Keziah Muthembwa. Valuable guidance and comments were received from various colleagues and external peer reviewers. The team also benefited from the guidance and supervision by Diarietou Gaye, World Bank Country Director for Kenya; Pablo Fajnzylber and Albert Zeufack (Practice Managers). Desktop publication was done by Martin Buchara and Lydie Ahodehou. We are also grateful to the following peer reviewers for constructively critical comments and feedback: Thomas O’ Brien; Alan Gelb; Helen Grandvoinnet; Niko Hobdari; Mwangi Kimenyi; Praveen Kumar; Daniel Lederman; Jane Mariara; Armando Morales; Terry Ryan; Aly Khan Satchu; and David Sperling. The report benefited from extensive review from the government of Kenya. ii K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M FOREWORD It is my pleasure to present the Kenya Country Economic Memorandum (CEM) titled: “From Economic Growth to Jobs and Shared Prosperity”. The CEM is a strategic World Bank product that analyzes key aspects of the country’s economic development with the main aim of providing an integrated and long term perspective of the country’s development priorities. This particular edition of the CEM has benefitted from extensive review from various stakeholders, including the government, academia and the private sector. The Kenya CEM has five main messages. First, Kenya has performed well in the past decade in terms of economic growth, and modern services are behind the acceleration of growth. Expansion in these services, such as financial intermediation and mobile communications have stimulated demand for other services such as trade. The CEM discusses how to maximize the potential of services, especially given that most formal, high quality jobs are created in this sector. Second, agriculture, which still contributes to over a quarter of the economy, and manufacturing have stagnated. The CEM discusses the reasons behind this stagnation, noting that agriculture and manufacturing have not been able to create enough jobs for Kenya’s growing working age population. Most of the jobs are created by the informal economy and are concentrated in low productivity segments of trade, hospitality, and jua kali. Improving the ease of doing business is one way towards job creation and higher productivity. However there is still a need for creating job opportunities for the rural poor, for poverty reduction and achieving shared prosperity. Reviving agriculture, in particular, remains the pathway for poverty reduction. Third, accelerating growth to meet Kenya’s development goals requires technological advances and innovation that raise firms’ productivity. Although the likelihood of Kenyan firms to innovate is high compared with firms in several other countries, the CEM finds that there is room to increase innovation. Only a few Kenyan firms have come up with products that are actually new to the domestic market. Moreover, the share of firms spending on research and development (R&D) remains low. And Kenya can leverage its stock of managerial capacity to increase innovation. At the same time, attracting foreign firms can stimulate productivity enhancement as technologies spill over to domestic firms. Fourth, achieving rapid growth will require macroeconomic stability to boost investment and savings. And as the government strives to build Kenya’s energy and transport infrastructure, this needs to be complemented with improvements in the public investment management process and better execution. Fifth, the discovery of oil opens a possibility for raising Kenya’s growth. Kenya’s recent oil discoveries, if used prudently, can contribute to achieving the Vision 2030 goals. Resource extraction can make a direct contribution to economic output and the main transmission channel will be fiscal. So appropriate management of resource revenues can generate resources that could be used to raise public investment, human capital, and productivity in the non-resource sectors of the economy. The World Bank Group is proud of its long-standing relationship with Kenya, and looks forward to continuous collaboration with both National and County Governments and other partners. Working together, Kenya can realize its potential to lift millions of families out of poverty and achieve shared prosperity. Diariétou Gaye Country Director for Kenya World Bank FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y iii iv RAN K E N YA C O U N T R Y E C O N O M I C M E M O Photo: M Authority D UPorts Kenya EXECUTIVE SUMMARY Case Stories Laetitia Mukungu from Bukura in western Kenya was one of the best students in her primary school, but her dream to continue schooling was shattered by lack of finances after her mom lost her job. Driven by her passion for agriculture, Laetitia discovered a way to earn money to cover her schooling expenses. She spent hours researching on the Internet and came up with the idea of raising rabbits. Kenya’s urban centers, in particular high-end hotels and restaurants, demanded more rabbit meat than the market could supply. So Laetitia convinced the local school to lend her K Sh 50,000, which she used to purchase several New Zealand rabbits, and employed 15 women to work in her rabbit center. Today, the business continues to be profitable and provides livelihoods for the women working in the rabbit center. Through her online research and practice, Laetitia discovered the rabbit hatch is a rich source of organic fertilizer and pesticide. Hence, the women in the center started planting maize and sukuma wiki, which supplemented their income. Finally, the business enabled Laetitia to realize her dream of a better education: in 2012 she was accepted at a renowned secondary school, the African Leadership Academy in South Africa. October 2014 was not a good month for the several hundred workers at the Eveready East Africa and Mondelez (formerly Cadbury) production plants; the owners of the two plants had decided to move production outside Kenya. The reason for the closure of both plants was the same: it would be cheaper to produce the products, dry-cell batteries in the case of Eveready and confectionary (chocolate) products for Cadbury, elsewhere—interestingly, to the Arab Republic of Egypt in both instances—and import them to Kenya. While closing production, both companies noted that Kenya is a growing market for their products and they plan to expand sales. In addition, Eveready announced that it would invest in real estate development on the 20 acres of land where the plant is located in Nakuru, and Mondelez said that Kenya would serve as its business hub for East Africa. January 2015 brought a cheer among Nairobi’s residents as the fast-growing taxi company Uber became available on their smartphones. Following its establishment in South Africa and Nigeria, Uber decided on Nairobi, which it sees as “the Green City in the Sun and East Africa’s economic powerhouse.”1 Modern service companies such as Uber need three things to thrive: middle-class city dwellers (urban population) with credit cards (access to financial services) and smartphones (high penetration of mobile Internet). Nairobi clearly has all three. Indeed, other leading multinational companies, such as IBM, Intel, and Google, have moved to Nairobi and are using it as their base for operations in Africa. 1 http://blog.uber.com/nairobilaunch. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y v These three examples pretty much capture the most important messages that the rest of this report conveys from a macro perspective. Laetitia exemplifies the potential of Kenya’s large and growing youth: entrepreneurial spirit, access to the Internet (even in a remote village), and expanding primary education. Businesses such as the rabbit center diversify the economy; moving from raising rabbits to producing fertilizer and pesticides is what economists call expansion of production capabilities (or “economic complexity”). At the same time, Laetitia portrays the struggles young people face: how to finance post-primary education, how to get access to finance, as well as the fact that most young entrepreneurs and workers typically go into services or production of goods that are protected from outside competition, that is, in the non-tradable sector. The second example illustrates how difficult life is for those who have to compete with the rest of the world. Eveready and Cadbury are the most recent examples on a long list of manufacturing firms that have closed production in Kenya and moved elsewhere, while Egypt’s list of newly opened manufacturing plants, in particular in the food industry, is rapidly growing. Kenya's growing economy has been consumption driven, so opportunities to meet the demand for goods and services for the rising middle class are plenty. However, importing those goods and services is more profitable than producing them domestically, given Kenya’s high cost of labor and utilities (electricity, land, and transport). The third example touches on a successful part of Kenya’s economy—rapid development and penetration of mordern services, such as mobile technologies and finance. These sectors of the economy, together with land development and commercial services, have been booming, particularly in Nairobi, which is the regional hub for most services and industrial firms. However, nine in 10 Kenyans do not live in Nairobi and do not work in the modern services sector. Having a small part of the economy pulling up the rest will not be enough to meet Kenyans’ development expectations. Some may think that the recent oil discoveries will fill the gap, but oil should not be taken for granted; the discoveries to date are neither groundbreaking nor guaranteed to become a “blessing” for the majority of Kenyans. Kenya’s Economy in the 21st Century: accelerated in the past decade, a prosperous What Have We Learned? society for all Kenyans has not yet been achieved. The economy remains among the poorest 25 Over the past half-century, Kenya has percent of countries in the world, and poverty established itself as an important regional player is high at around 40 percent of the population. on the continent, and has achieved successes Unlike in some earlier episodes of Kenya’s history, on multiple fronts. The turn of the century at present there is a strong demand and will to marked an economic revival that has been bend the arc of history, and to achieve faster and accompanied by a rise in citizens’ expectations. more inclusive growth. The first step in deciding Following the rebasing of gross domestic product what can be done to accelerate growth and (GDP) in September 2014, the country joined shared prosperity is to understand the upsides the celebrated ranks of the lower-middle- and downsides of past performance. income countries. Although economic growth vi K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Services Have Taken Off… But Not Enough to Following two decades of per capita income Make Kenya a Star Performer stagnation, Kenya’s economy showed signs of revival at the turn of the century. The market Over the long term, Kenya has performed reforms that began in the early 2000s released relatively well compared with others. A temperate the economy’s potential and GDP growth climate, coastal access, and other natural and accelerated steadily from below 1 percent in 2002 geographic advantages have provided a strong economic base for Kenya’s growth. Despite to 7 percent in 2007.2 This is the only episode of structural policy mishaps, episodic political five-year accelerating growth in independent violence, and crime, the country has successfully Kenya’s history (Figure ES.1), and it was also the avoided the outright implosion of many of its first time since 1986 that GDP growth reached 7 neighbors: GDP per capita stagnated during percent. However, since 2007, the economy has 1984–2003, but never crashed. Beginning in the been hit by several shocks. GDP flattened in 2008 2000s, the country demonstrated the capacity for and then picked up to 8.4 percent in 2010, but innovative services (especially in mobile telecom immediately slowed to 5 to 6 percent afterward. and banking), which was facilitated by Kenya’s role as a regional hub. And Kenya has a manufacturing Services, modern and traditional,3 are behind base. But global and regional competition have the acceleration of growth. Between 2006 and threatened the manufacturing base, and, until 2013, 72 percent of the increase in GDP came recently, modernization of the service sector from services. Expansion in modern services, has been hampered by the difficult business such as financial intermediation and mobile environment. Thus, export growth other than in communications—partly owing to innovative the tourism sector has stalled. At the same time, solutions such as M-Pesa (mobile money)— oil and gas prospects have come into view, but stimulated demand for traditional services such existing weaknesses in public expenditure policy as trade. For example, there are more than and management will have to be tackled for this 40,000 M-Pesa retail agents who also sell other new revenue stream to be transformative. products and services. Investment and promotion Figure ES.1: Kenya’s uneven growth performance 25 Post independence-boom Two decades of stagnating Economic revival, disrupted by disrupted by droughts (1970) income per capita (domestic and external) shocks 20 15 10 5 0 -5 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 GDP growth (annual, %) Sources: World Bank and Kenya National Bureau of Statistics. Note: GDP = gross domestic product 2 Use of GDP data throughout the report took into consideration the revised GDP series in September 2014. 3 Modern services comprise communications, finance, professional, scientific, and technical activities, and other services. Traditional services include construction, trade, transport, hospitality, public administration, education, social, recreation, and administrative services. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y vii of tourism have boosted hospitality, real estate, to disappear at the county level, but citizens’ and transport services. Re-orientation of public demand for more accountability is rising, which is resources toward public and social infrastructure expected to boost agricultural development. has promoted educational services as well as construction and transport. As for manufacturing, the puzzle is not why Kenya does not have a manufacturing sector—it does In contrast, agriculture and manufacturing grew have one—but why this sector has not been able slower during 2006–14. Agriculture suffered to expand. Factors highlighted by Rodrik (2015), weather shocks, which caused the sector’s share such as the way globalization and trade have in GDP to decline from 26.5 percent in 2006 to worked to the disadvantage of African countries, 22.0 percent in 2014. Manufacturing stagnated are part of the story. But the economy has also at 11.8 percent of GDP on average during the struggled to develop the deep public-private same period. Some subsectors within agriculture networks of regulation, facilitation, skills, and and manufacturing, such as horticulture and infrastructure, which advanced manufacturing food production, have prospered, but the overall economies need. It is revealing that Kenya does story for the two sectors has been disappointing well in sectors where networks are somewhat (Figure ES.2). The success of the two subsectors easier to establish, as in banking and telecom, is to some extent interlinked: countries with but struggles with the more intensive network successful structural transformation typically are capabilities needed for modern manufacturing. able to increase the value added in agriculture by moving up the value chain (toward improved Figure ES.2: Services have been the main generator of growth quality of produce and further processing). (% of GDP) Despite its relatively weak performance, 100 agriculture continues to be the mainstay of 80 Kenya’s economy, as seven in 10 Kenyans depend on it for their livelihood. Some parts that have 60 Percent seen no direct government intervention, such as horticulture, have been booming, while food 40 crops, such as maize, have underperformed. 20 Multiple strategies and reforms have been 0 2006 2007 2008 2009 2010 2011 2012 2013 2014 designed and adopted, but never fully Agriculture Industry Services implemented. The liberalization of the maize Source: Kenya National Bureau of Statistics. market began in the late 1980s, yet the government has resisted exiting this market and Looking at the expenditure side, consumption remains active through the National Cereals has contributed the most to GDP growth. and Produce Board. Despite the commitment Rising private consumption has been the main to allocate 10 percent of budget revenue to the contributor to growth (Figure ES.3), propelled by sector, only half of that has been spent over the growing middle class, booming informality the past few years. Devolution is expected to in services, increasing credit to households, and bring a positive change to the sector: major income from abroad. Increased investment has agricultural functions have been transferred to also had a positive, although less significant, county governments. Patronage is not expected contribution, in particular fueled by a shift in viii K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M public spending from recurrent to “development capita. Among the peers, Kenya has had the lowest spending.”4 What differentiates Kenya from the per capita GDP growth since 2003 (Figure ES.4). other peer countries, in particular those outside the East Africa region, is the clogged “exports If the Kenyan economy had grown as fast as its engine.” Exports of goods, as a percentage of peers in Sub-Saharan Africa (SSA) over the past GDP, have been falling since 2005, while imports decade, by 2014 the average Kenyan’s income of goods have been increasing. The reason for would have been 15 percent higher. If the these trends has been strong capital goods economy had matched the growth of the Asian imports, associated with investment related peers, then Kenya’s income per capita would to oil exploration, and a decline in agricultural have been 45 percent higher. exports. In contrast, services exports have been Figure ES.4: Kenya’s GDP per capita has been growing expanding, but not by enough to offset the slower than its peers GDP per capita (2000 = 100) widening gap between the exports and imports 200 of goods. 180 Figure ES.3: Consumption: Main contributor to GDP 160 growth Contribution to GDP growth (in percentage points) 140 8 120 6 100 4 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2 Kenya SSA peers Rest of the world 0 Sources: World Bank World Development Indicators; Kenya National Bureau 2009 2010 2011 2012 2013 2014 of Statistics. -2 Note: Sub-Saharan Africa peers: Burkina Faso, Ghana, Senegal, Tanzania, and Uganda. Non-Africa peers: Bangladesh, Cambodia, India, Pakistan, and Vietnam. -4 -6 Consumption Net investment Net exports Comparison with similar economies reveals Source: Kenya National Bureau of Statistics. Note: A percentage point is 1 percent or the unit for the arithmetic difference several distinctions about Kenya’s growth of two percentages. Therefore, adding all the contributions from each expenditure for a given year should give the annual economic growth. GDP model. Chapter 1 shows that Kenya is unique in = gross domestic product. its services-based growth model: in all the other countries, except Senegal, industry5 had a much Taken as a whole, the past decade’s economic larger contribution, and in most of them so did performance can be described as remarkable by agriculture. In Kenya, rising consumption— Kenyan standards, but in a broader perspective it propelled by rising, mostly informal, employment, is not even close to stellar. Instead of comparing credit to households, and income from abroad— Kenya with global or regional (Sub-Saharan Africa) has been the main engine of growth. In contrast low- and middle-income country averages, where to most of the peer countries, Kenya’s net exports a few large economies such as China and Nigeria have made a negative contribution to growth pull the averages, this report benchmarks Kenya and this has come mostly as a result of stagnant against a group of peer countries from Africa and exports of goods. the rest of the world that had similar income per 4 Development spending in Kenya denotes public spending on capital investment. 5 Industry comprises mining, manufacturing, and utilities (electricity, water, and gas). FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y ix Growth Has Been Volatile … political stability. Since 2012, the economy has benefited from macroeconomic stability (Figure Another feature of Kenya’s recent economic ES.5) and a peaceful election cycle in 2013. performance is its volatility, which comes Consequently, volatility decreased while growth primarily from domestic sources. Kenya’s has been maintained above 4.5 percent. economic growth has exhibited higher volatility than that of its peers since 2003. Moreover, … and Not Particularly Inclusive growth volatility increased after 2008. The sources of volatility have been exogenous (through The positive but volatile growth since 2006 trade or global commodity prices) and domestic translated to rapid poverty reduction. Poverty (election cycle), but the latter had a larger impact and inequality in Kenya was last measured in on the economy. Chapter 1 finds that shocks from 2005–06. No survey has been fielded since then major trading partners, that is, the fall in demand to update these estimates.6 In the absence of for Kenyan exports, are instantly transmitted actual data since 2006, current poverty estimates to Kenya’s economy, although the impact is and projections are highly uncertain and depend not as significant. Global food price shocks also exclusively on modeling assumptions, but influence inflation; oil prices have less of an some likely trends are plausible and these are influence. However, the most important finding is presented in chapter 2. In June 2013, the World that much of the volatility has been domestically Bank estimated that poverty fell from 46 percent driven and domestic shocks—such as political in 2006 to below 40 percent by 2012. An updated instability or drought—typically have longer model7 points to weaker poverty outcomes: the effects than exogenous shocks. poverty rate8 is estimated to have fallen to around 42 percent by 2013. Poor and uneven agricultural The silver lining perhaps is that reducing performance has certainly contributed to volatility is primarily a question of domestic poverty. Although the economy has been growing policies. The past four years attest to the continuously, income per capita fell during 2007– importance of domestic macroeconomic and 09, which set back poverty reduction, especially as agriculture shrank during this period. At the Figure ES.5: Macroeconomic stability has been maintained same time, the prices of food and transport—two since 2012 expense categories that affect the poor more— 35 120 spiked during 2009–11. 30 100 Public policies have had mixed effects on Ksh per USD 25 80 Percent 20 60 poverty reduction. Fiscal policy has been pro- 15 40 poor in two ways: the cash transfers program to 10 20 the poor has expanded (yet it reaches less than 5 0 10 percent of the poor), and so has spending 0 on education (although the impact is muted by Oct Oct Oct Oct Oct Oct Apr Apr Apr Apr Apr Apr Jan Jan Jan Jan Jan Jan Jul Jul Jul Jul Jul Jul 2010 2011 2012 2013 2014 2015 high teacher absenteeism). These two reforms Inflation (annual average, %) Interbank rate (%) US$ exchange rate can have a strong effect on poverty reduction, Sources: Kenya National Bureau of Statistics; Central Bank of Kenya. as experience in other countries has shown. 6 Fieldwork for the 2015–16 Kenya Household Budget Survey commenced on September 1, 2015, and the resulting data set, finally after a decade-long wait, will enable updating poverty and inequality measures. 7 The model was updated with revised GDP figures for 2006–13. 8 The share of the population living on less than US$1.25 a day. x K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M For example, Brazil’s success in reducing poverty How to Accelerate Growth to Meet and inequality in the 2000s is attributed largely Kenya’s Development Goals? to expanding social assistance and increasing the wages of low-skilled workers by investing in skill Vision 2030 sets a goal for Kenya to join the ranks development. In contrast, Kenya’s health spending of upper-middle-income countries,10 a group that has remained inadequate at below 2 percent comprises countries with gross national income of GDP over the past decade, corroborated per capita of $4,125 in 2014 (almost four times by high maternal deaths (more than 400 per Kenya’s) (box ES.1).11 This goal is formidable but 100,000 births) and prevalent child malnutrition, achievable—several countries have made such which affects the poor disproportionally.9 The progress over the past few decades—and will first year of devolution is unlikely to have made require GDP growth of about 7 percent per year a significant impact on poverty, but this may until 2030. For Kenya’s authorities, this will not change as counties develop a more proactive role be an easy challenge; the economy grew by more in agricultural development, given agriculture’s than 7 percent in only four of the past 40 years. potential to reduce poverty. Monetary policy Looking forward, attaining those rates, although affects poverty primarily through its effects on difficult, is possible. Moreover, there is a sense inflation, so in this regard the spike in inflation of urgency to this agenda, as the growth targets in 2011 had negative consequences for Kenya’s in the Second Medium-Term Plan (MTP-2) are poor, although since then inflation has been already slipping (2013 growth was 5.7 percent within the Central Bank’s targets. vis-à-vis the MTP-2 target of 6.1, and 2014 growth was also lower than the MTP-2 target of 7.2 percent). Box ES.1: What does it mean to be an upper-middle-income country? Tunisia today is just above the upper-middle-income threshold. Looking at Tunisia’s economic and social indicators, Kenyans have a lot to look forward to. First, Tunisians live on average 14 years longer than Kenyans. Practically no Tunisian lives in extreme poverty, and all Tunisians have access to electricity. Finally, Tunisians are better educated and two-thirds live in urban areas, compared with less than one-third of Kenya’s population. (see Table B.ES.1.1.) Table B.ES.1.1: Standard of living indicators, Kenya and Tunisia Years Kenya Tunisia Access to electricity (% of population) 45 (2015) 99.5 (2010) Life expectancy at birth, total (years) 61 (2013) 75 (2013) Poverty headcount ratio at $1.25 a day (PPP) (% of population) 43 (2005) 1 (2010) School enrollment, tertiary (% gross) a 8.6 (2014) 35 (2012) Urban population (% of total) b 32 (2014) 66 (2013) Sources: World Bank World Development Indicators; Kenya Power and Lighting Company; Kenya Ministry of Education. Note: PPP = purchasing power parity. a. Estimates based on Kenya National Bureau of Statistics and United Nations Educational, Scientific, and Cultural Organization data. b. World Bank estimates. 9 World Bank Kenya Economic Update 8 (June 2013). 10 According to the World Bank’s income classification. 11 The World Bank’s country classification by income is based on gross national income per capita. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y xi Achieving the desired growth targets entails measures. To ensure higher savings, the simultaneous improvements on two fronts: government will pursue prudent macroeconomic increased physical and human capital, and faster policies, to achieve lower economic volatility and productivity growth. In the short to medium improved public investment management. term, adding capital and labor can stimulate growth. Kenya’s investment-to-GDP ratio, at 20 Compared with other fast-growing economies, percent in 2013, is targeted to rise to 31 percent Kenya invests less and the share of investment by 2018 in the MTP-2, and at the same time there financed by foreign savings is higher. The is potential to increase employment, given that economic literature and post-World War II history a portion of Kenya’s labor force is unemployed illustrate that investment determines how fast or underemployed (no precise labor market data an economy can grow. Kenya’s investment, at exist). As the labor force and the share of income around 20 percent of GDP, is lower than the 25 that can be set aside for investment have their percent of GDP benchmark identified by the limits (and face diminishing returns), growth in the Commission on Growth and Development (2008). long run can only be sustained through productivity Kenya’s investment rate, as a share of GDP, has enhancements. This report finds there is scope to also been several percentage points lower than accelerate growth on both fronts. the rate in its peer countries. At the same time, the economy has largely relied on foreign savings Save More to Invest More as a source for new investment since 2007, while national savings have been declining. National Although the economy has relied on foreign savings—measured as a share of gross national funding to increase investment since 2006, disposable income (GNDI)—has not surpassed national savings would have to increase to reach the 15 percent mark over the past decade. In the desired investment levels. Investment rose contrast, Pakistan’s savings is above 20 percent from 15.6 to 19.6 percent of GDP between 2006 of GNDI, and Vietnam’s is more than 25 percent. and 2013, but this increase was financed from Cambodia had a low savings rate in the 1990s, foreign inflows, that is, by rising current account but it more than doubled the rate in the 2000s. deficits. Vision 2030 and the MTP-2 endorse this approach and set ambitious targets for High unemployment and volatile inflation augmenting public and private investment. To this are two of the reasons behind low saving, in end, the MTP-2 aims to increase the investment particular by households. The falling youth rate to 31 percent of GDP by 2018, an ambitious dependency ratio, that is, declining fertility, increase of 11 percentage points from the 2013 should have promoted saving. However, this may level, while raising national savings from 16.4 not have been the case because a large share of to 25.7 percent in the same period to finance youth has been jobless and thus continues to be investment in a sustainable manner. As chapter dependent although the youth are of working 5 discusses, it is only in 2020 at the earliest when age (Figure ES.6). oil could start flowing and contribute to fiscal revenues to be used for public investment. Until Another reason behind the low household then, the government aims to raise national saving is the fact that rates on deposits at savings by implementing a contributory pension financial institutions have been low, and even scheme for public servants and tax incentive negative in real terms in some years. Volatile xii K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M inflation has been one of the main reasons for opportunities for the new labor market entrants the negative deposit rates, as nominal deposit will support savings. Second, keeping inflation rates have not always adjusted fully to changes low and stable (within the Central Bank’s limits) in inflation, in particular when higher inflation would raise the real deposit rate and incentivize has been unexpected, such as in 2011. Other saving. Third, re-orienting public spending from factors that influence the deposit rate include recurrent to capital expenditure will represent the level of competition in the sector, and banks’ an increase in public savings. To this end, if the price-setting behavior.12 Chapter 3 looks into 70-30 rule on recurrent versus development the saving-investment nexus in greater detail, spending is implemented in practice rather than while Spotlight 3 (at the end of chapter 3) shows on paper (2013/14 had a 69-31 budgeted ratio, there is potential to increase saving in rural but the executed ratio was 71-29), the share of areas by improving the access to and design of investment in GDP would increase by up to 2.5 saving schemes. Saving and credit cooperative percentage points. County governments were organizations (SACCOs) have been successful partly responsible for the under-execution in mobilizing savings in Kenya and channeling of development spending: only a third of the savings to investment projects at the local budgeted 2.1 percent of GDP was executed in level. Moreover, connected to mobile saving 2013/14. Spotlight 1 (at the end of chapter 1) schemes, such as M-Shwari, Kenya’s SACCOs have discusses the impact of devolution on investment increasingly attracted savings and contributed and growth, and finds that the lower execution of to the realization of the saving and investment development spending was accompanied by an target of Vision 2030. increase in recurrent spending by the counties. In Figure ES.6: Kenya has among the highest youth promoting private investment, Kenya’s financial unemployment rates, 2000–14 average system is relatively developed, so the onus should Youth (age 15-24) unemployment (in percent) 20.0 be on lowering production (infrastructure) costs and improving the business environment. Finally, 15.0 oil revenue may become a significant source of savings in the long term, although the potential 10.0 (discoveries) is still uncertain and the outcomes will depend on how the oil sector (and revenue) 5.0 is managed (chapter 5 looks into this). 0 Create Jobs for the Growing Number of Youth Source: World Bank World Development Indicators. Kenya’s growing labor force is not being put to productive use, which in turn is hurting growth. Although there is potential to increase savings The share of the working-age population rose through policy measures, it is more important from 47 percent in 1990 to 56 percent in 2014, to channel savings to productive investment. and by 2050 it is expected to be 62 percent.13 This First, the demographic trend of an increase in the opportunity for a demographic dividend, a boost share of the working-age population is expected in GDP growth caused by the increasing share of to continue in the next decade, so ensuring job working-age relative to dependent population, 12 World Bank’s Kenya Economic Update, December 2013. 13 World Bank’s Kenya Economic Update, December 2012. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y xiii will be reaped only if the new potential workers Kenya’s labor market entrants for the next 15 are able to find jobs. Putting Kenya’s human years have already been born, and getting them capital to productive use has proven to be a major employed will require much faster job creation challenge. Between 2009 and 2013, three million than in the past. Kenya’s working-age population youth became of working age, yet the economy is projected to be 39.2 million in 2030, from the was able to add only 2.6 million nonfarm jobs, current 25.5 million in 2015.15 The government’s and the growth in employment (24 percent) strategies recognize this, and to this end the MTP- could not keep up with GDP growth (26 percent). 2 targets 570,000 new formal jobs between 2015 If employment grew at the same pace as GDP, an and 2017, from 110,000 in 2013. However, for this additional 150,000 jobs would have been created to happen, most of the sectors of the economy by 2013, which still would have been insufficient would have to quadruple the job creation rates to absorb all the new entrants. Although official they achieved in 2013 (which were already higher statistics are not available, unemployment and than in previous years) and sustain them until underemployment are rampant, especially 2017. Job creation in the informal sector is also among the youth. projected to increase—from 664,000 new jobs in 2013 to 859,000 new jobs in 2017—and again Although formal jobs are in high-growth and high- this will require proactive policies to promote productivity sectors, the job-creating potential growth in the informal sector, and also higher of these sectors is relatively low, so most job productivity so that income from informal work seekers end up in low-productivity, informal can lift people out of poverty. activities. Kenya’s modern service sectors, such as financial services and communications, but also Further improvements in the business the education sector, mining, and utilities, have environment, quality of skills and education, been adding jobs and raising labor productivity and labor regulations are expected to promote at the same time. However, the job-creating job creation. A conducive business environment potential of these sectors is small. For example, is one of the key pillars for job creation, and while although financial services and communications Kenya has historically fared poorly in this regard, recorded among the fastest employment growth the Government of Kenya has recently made this rates (7 percent per year between 2009 and 2013), a priority. Although many of the positive changes fewer than 10,000 jobs were added per year. The introduced by the government were removed four sectors with the highest productivity growth from the cycle to be captured in Kenya’s latest between 2009 and 2013 accounted for only 7 Doing Business report, the country’s 2016 rank percent of total employment. The majority of job improved an impressive 21 places, from 129 to seekers go to the jua kali,14 in trade, hospitality, 108. The Kenyan government recognized the or manufacturing, and many of them are challenges, and has invested significantly, under underemployed. the coordination of the Ministry of Industry and Enterprise Development, in unlocking business 14 This is a commonly used term for Kenya’s informal sector. 15 The total population is 46 million and 65.4 million in 2015 and 2030, respectively. xiv K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M environment bottlenecks (with the creation with other skills that are valued by employers, of the Ease of Doing Business Delivery unit). such as accessing information, using computers, Momentum has been gained in prioritizing knowing how to interact professionally with reforms, particularly in core bottlenecks, including clients, solving complex problems, and learning company registration, electricity connections, new skills while on the job. The present system property transactions, and access to credit. has several deficiencies: it is not flexible to labor Many aspects of the business environment take a market needs, capacity is limited, and there relatively small amount of time and resources to is limited successful measurement of quality transform. Chapter 2 offers examples of countries and outcomes. Overall, the main priorities for that have made rapid progress on various aspects improving the employability of youth are (i) of reform. Prime examples are the actions better evaluation of existing programs that taken by Kenya, Rwanda, and Uganda to reduce would inform policy; (ii) better coordination of barriers on the Mombasa to Kigali trade corridor, youth policies; (iii) improved access to vocational which eliminated roadblocks and administrative training, particularly for the poor; (iv) better barriers that slowed traffic on that important targeted support to entrepreneurship; and (v) trade route. As a result, transit times fell by improved design of training programs to meet about 50 percent and Kenya and Rwanda each employers’ needs.16 improved by around 50 positions in the World Bank’s logistics performance index. Finally, the 2007 changes to the labor code seem to be dis-incentivizing formal employment, so Following the education reform successes, some aspects of the legislation may need to which yielded notable improvements in access be revisited. Before the major revision of labor to education, the focus of the education legislation took place in 2007, only 4 percent of system has now moved to raising the quality of firms found labor regulations to be a constraint, produced skills. The fastest growing sectors in the which was less than elsewhere in SSA (12 percent economy are increasingly struggling with finding on average). In 2013, one-fifth of firms were suitable employees. In 2007, only 2 percent complaining about the regulations, which in turn of services firms identified skills as a major probably incentivized informal employment. Two constraint. By 2013 more than a third of services particular issues have come up. First, regulations firms were struggling to find qualified workers. are strict in terms of employer-employee To remedy this trend, the quality of education disputes; workdays lost to such disputes rose from needs to be improved, which includes ensuring 15,000 in 2008 to 175,000 in 2011, according to that basic foundational skills are mastered the Kenya National Bureau of Statistics (KNBS). and outcome competencies and programs are Second, the minimum wage may be pushing firms relevant for employment. Key components of toward informality. The ratio of the minimum skill building include acquiring job-relevant wage to value added per worker was found to be technical skills (for example, through technical higher in Kenya compared with its peer countries and vocational education, higher education, pre- (Figure ES.7). employment, and on-the-job training), along T16 Youth Employment Initiatives in Kenya (draft World Bank report). 16 FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y xv Figure ES.7: Kenya’s minimum wage is highest among its entrepreneurs’ decision regarding informality. peers 300 2.0 It should be noted that counties will have a 250 primary responsibility for these policies, but the US dollars per month 1.6 200 outcomes since the start of devolution have been 1.2 discouraging. Spotlight 1 (at the end of chapter 1) Ratio 150 100 0.8 points to the introduction of new taxes, fees, and 50 0.4 charges by counties, which is generating concern 0 0 over the potential impact on local-level business Parkistan costs, especially for small business operators. Kenya Bangladesh Cambodia Uganda Burkina Faso Ghana Tanzania Vietnam India Senegal Minimum wage for a full-time worker (US$/month) Creating job opportunities for the rural poor is Ratio of minimum wage to value added per worker particularly relevant for the poverty reduction Source: World Bank 2015a. agenda. As in many other countries in SSA, Kenya has witnessed low growth elasticity of poverty Since the majority of job entrants will still end reduction, because although most of the poor up in the jua kali, public policy should focus on are in agriculture, growth has been happening promoting productivity growth in the sector. elsewhere (notably in services). To begin, Doing so will complement the growing formal promoting agricultural productivity is paramount sector and help to set a smoother transition from for poverty reduction and growth. Although the informal sector to the formal sector over this report looks at agricultural performance time. The informal sector will absorb most of the and the rural poor, examination of agricultural urban and rural unskilled youth. Chapter 2 reveals productivity is beyond its focus. Nevertheless, that the jua kali is a dynamic sector in terms increasing the value added of agricultural of market entry, but informal establishments produce (through higher crop yields, better typically stay small (one employee) and hence packaging, higher quality, or further processing) do not create additional jobs once established. would boost productivity in the sector and raise The main challenge that informal entrepreneurs farmers’ incomes. Malaysia, for example, is one face is access to finance, followed by access of the world’s largest exporters of papaya, and to utilities, corruption, and crime. Moreover, the government—through its Malaysian Agrifood although informal business owners are aware of Corporation—has played a key role in propelling the benefits of formal operation, most opt to stay the food supply chain business through the informal because of cumbersome registration application of new technology, logistic solutions procedures, as well as to avoid paying taxes. (packaging), and promotion of international Chapter 2 offers examples from other countries food safety standards. Other ways to generate in SSA that have successfully addressed some income for the rural poor include helping youth of the constraints of informal businesses. A key to transit to nonfarm (informal) employment, as lesson here is that public policies should have well as promoting their mobility. On the latter, a dual focus: (i) they should address the largest measures to encourage movement to urban constraints (for example, access to finance or skill centers, or improvements in rural roads so that building), and (ii) they should enhance the quality rural youth can work in urban centers, would of services and governance, as these influence boost employment and poverty reduction. xvi K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Boost Productivity through Policy Reform and section discussed what could be done to create Investment opportunities for the increasing labor force; this section discusses what could be done to improve Whereas adding capital and labor to the the productivity of firms. economy can accelerate growth in the short to medium term, sustained and rapid growth Various data suggest that Kenya is an requires technological advances and innovation entrepreneurial nation and, as figure ES.9 that raise firms’ productivity. From a macro illustrates, Kenya has witnessed rapid growth perspective, productivity can increase by moving of formal business startups. There is higher firm labor from low- to high-productivity sectors churning and more diversified production than (for example, from agriculture to services), or in some of its peers with similar income levels. by increasing within-sector productivity (such Moreover, Kenyan firms are proactive in reaching as firms generating higher value added per foreign (including regional) markets, although employee). Kenya’s economy offers scope for their successes are rare. However, the use of both, but one in itself will not suffice to reach export data as a proxy for production capabilities the Vision 2030 objectives. Shifting jobs to more shows that Kenya’s production capabilities (its productive sectors will only bring productivity up economic complexity) are lower than those of to the level of the best performing sectors, but its peers. Kenya’s capabilities are diversified, will still be low compared with the development but mostly in low complexity goods such as tea targets. And although there is scope for improving or coffee, and they have not been increasing productivity across the board, some sectors, in recent years. The reasons for this likely lie such as agriculture, face limits to productivity in high production costs. Wages in Kenya are growth, so labor would eventually have to move much higher than in peer countries at a similar out. Between 2009 and 2013, inter-sector shifts level of development. Transport, energy, and made a negative contribution to productivity, land costs, which account for half of total costs such that labor was moving to sectors with excluding raw materials and labor, are also below average productivity, such as informal likely to be higher compared with competitor trade and hospitality (Figure ES.8). The previous economies. Moreover, improving workers’ Figure ES.9: Kenya has witnessed rapid growth of formal Figure ES.8: Jobs were added in low-productivity sectors business startups between 2009 and 2013 (2004 = 100) 60 Average employment 450 growth (24.3%) 50 Mining 400 40 Employment (percent) 350 Burkina Faso 30 Communication Education Ghana 20 300 Agriculture Trade and India 10 hospitality Construction Utilities 250 Kenya Finance 0 Pakistan Transport Manufacturing -10 200 Senegal Public Average productivity administration Other -20 Uganda growth (1.7%) 150 -30 0 5 10 15 20 25 30 35 40 100 2004 2005 2006 2007 2008 2009 2010 2011 2012 Productivity (percent) Source: World Bank estimates. Source: World Bank Entrepreneurship Snapshot. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y xvii skills through better quality education will boost Kenya’s service economy is competitive at productivity. Transport and energy costs could the global stage: services exports more than be reduced through investment and regulatory doubled between 2009 and 2013. Services have reform (for example, to reduce waiting time at performed well, in particular the sectors with a ports and borders), which in turn would increase higher share of formality. Several formal services firm productivity and enable firms to grow as sectors, such as telecommunications and finance, they become more competitive. have prospered, and so has tourism over the past decade (2013 and 2014 are exceptions). In addition, the dispersion in firm productivity Looking at services firms through a micro lens, within the same manufacturing subsector is competition seems to work better among formal high, which means firms are not catching up services firms: unlike in manufacturing, entrant with their successful peers. Typically, firms that services firms are typically more productive operate in the same sector learn from each than the established ones. At the same time, the other and the ones with the lowest productivity dispersion of productivity within the same sector improve over time. Chapter 3 finds that this does is high, similar to manufacturing and possibly not seem to be happening in Kenya, so it is worth for the same reasons, but this is something that examining what is preventing low-productivity would warrant further research. firms from catching up to the more productive firms that produce similar products (Figure Interestingly, the successful and internationally ES.10). The analysis shows that entrant firms are competitive services have followed an isolated less productive than established firms, which path of development, which is different may be explained by weaknesses in the business than in many other countries. Gross exports environment (for example, costly procedures of services, as calculated in trade statistics, for starting up, or poor access to finance). Or typically undervalue the contribution of services perhaps established firms are able to draw to a country’s exports, because domestic higher privileges in terms of access to inputs services are a significant component in the (such as electricity) or markets (government production of export goods. This is also the procurements), which would allow them to be case in Kenya. However, chapter 3 shows that more productive than new firms. although services exports are relatively high, their indirect contribution, that is, linkages, to Figure ES.10: Within-sector differences in productivity are other sectors is actually relatively lower than in high among Kenyan firms Produc tivity of firms at the 80 th percentile relative to comparator countries. This suggests a dualistic produc tivity of firms at the 20th percentile 25 economy in Kenya, where services sectors such as telecommunications, finance, and transport 20 have prospered, while most of agriculture and 15 manufacturing have not. Constraints to the 10 business environment that prevail in Kenya seem 5 to have had disproportionate effects on services vis-à-vis manufacturing firms. The finding of 0 business surveys that manufacturing firms are two and a half times more likely than services firms to find electricity as a major constraint supports this Source: World Bank 2015b. Note: Sectors are for Kenya; bars for China and India are averages. xviii K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M argument. The implication is that services will actually come up with things that are new to the continue to expand, irrespective of what happens domestic market. The fact that these innovations with agriculture and manufacturing. have not been accompanied by productivity gains in most cases confirms the point that Irrespective of the sector, innovation is the Kenya still has a long way to go. And when actual critical element of accumulation of production investment in innovation is compared across capabilities—in other words, adding complexity countries, the magnitude of innovation of Kenyan to the economy. At the aggregate level, classical firms becomes less impressive. The share of firms and modern theories of economic growth put spending on research and development (R&D) in innovation at the core of the growth process. At Kenya is 40 percent lower than in Ghana or the the micro or firm level, which is where innovation Arab Republic of Egypt, and less than 50 percent occurs, there is also strong evidence of a robust, that in South Africa. And a relatively lower share positive relationship between innovation and of Kenyan firms acquire machinery, equipment, productivity and growth. Therefore, a critical and software, and the same argument holds for predictor of countries’ potential to grow is how spending on training. innovative their firms are. At the same time, Kenya’s managerial capacity Innovation among Kenyan firms is widespread is an asset that can support higher innovation. by global standards. The results from a 2014 Managerial capacity—a key recipe for enhancing innovation survey (World Bank Enterprise Survey productivity through innovation—is relatively 2014) suggest that the likelihood of Kenyan high in Kenya, considering the country’s level firms to innovate is high compared with firms in of income per capita. Nevertheless, the overall several other countries (Figure ES.11), although quality of management in Kenya remains far the subjective nature of the results makes cross- from the frontier, so there is scope for further country comparisons challenging (Spotlight 4 at growth through improving the quality of tertiary the end of chapter 4). Nevertheless, although education and increasing the linkages between most firms say they have introduced some type academia and businesses. of product or process innovation, only a few have Finally, productivity enhancements can come by Figure ES.11: Innovation Is widespread in Kenya by global standards attracting foreign firms to produce high-value Percent of firms that introduced product or process innova tion in year surveyed goods and services in Kenya. Foreign direct Kenya investment (FDI) in high-productivity sectors Philippines can stimulate productivity enhancements in the Israel Malaysia economy, as technologies and knowledge can Brazil spill over to domestic firms. Although it is not a China close comparator, China has been particularly Uruguay South Africa successful at this: nearly 90 percent of the Colombia foreign-owned firms had adopted the firm’s Russia core technology in local production and more Egypt, Arab Rep. than 60 percent relied on local firms for over 0 10 20 30 40 50 60 50 percent of the production components. At Source: World Bank Enterprise Surveys. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y xix the same time, nearly 50 percent of foreign However, the outcome of resource discoveries is firms train over 80 percent of their staff, and not guaranteed; some countries have lived up to many of these trained employees eventually the potential, while others are examples of how move to domestic firms.17 Kenya has not been natural resources can turn out to be a “curse” able to attract significant foreign investment (in rather than a “blessing” (Figure ES.12). production capacities), irrespective of the data source used for estimating FDI. To this end, the Figure ES.12: Different development trajectories of oil rich countries: Angola, Malaysia, and Republic of Congo revamping of the special economic zones (SEZs) GDP vs. Oil Production (1972 - 2010) program holds potential to attract more FDI and 16 Malaysia, 2010 GDP per person (2005 USD, PPP) in turn enhance productivity. Spotlight 2 (at the 14 end of chapter 2) points to the lessons learned 12 from similar such programs to maximize the 10 Angola, 197 Angola,2010 benefits, in particular related to the spillover 8 Congo, Dem . Rep, 1972 Congo, Dem. Rep. 2010 of productivity enhancements to the domestic 6 economy. These include ensuring that SEZs 4 address the most binding constraints to investors 2 Malaysia, 1972 (such as infrastructure), leveraging competitive 0 0 1 2 3 4 5 6 advantages and agglomeration rather than Oil Production Per Person (tonnes) Sources: BP Statistical Review of World Energy; World Bank World supporting lagging regions, promoting linkages to Development Indicators. the domestic economy, and understanding that developing SEZs takes time (5–10 years). Kenya may become an oil exporter in a few years’ time, and the timing and potential resources can Natural Resources: Kenya’s Hidden boost Kenya’s path toward its Vision 2030 goals. Source of Long-Term Growth? By 2014, commercial oil exploration in northern Kenya brought the discovery of up to one billion Kenya’s recent oil discoveries, if used wisely, barrels of oil resources.18 More exploration can contribute to achieving the Vision 2030 (for oil as well as gas) is ongoing onshore and goals. The discovery of oil, and possibly gas, offshore. These reserves would not turn Kenya opens a possibility for raising Kenya’s growth into a significant oil producer at the global potential. Resource extraction will make a direct stage, but will generate fiscal revenue that can contribution to economic output, but more raise the country’s human and physical capital. importantly, it will generate fiscal resources that The timing of the discoveries relative to Kenya’s could be used to raise public investment, human development cycle is appropriate. Kenya’s level capital, and productivity in the non-resource of development is sufficiently high to be able to sectors of the economy. As chapters 2, 3, and 4 absorb the windfall revenue and establish the point out, for sustained and rapid growth, Kenya needed legal framework and infrastructure for needs further investment in physical and human developing the oil sector. However, as pointed capital to promote productivity growth. Windfall out in chapter 1, the economy (in particular revenues could help bridge Kenya’s infrastructure the exporting segment) is not too advanced to gaps and skills deficiencies and expand the be vulnerable to a Dutch disease type of shock. provision of health or other social services. According to industry estimates, oil could start http://www.tullowoil.com/index.asp?pageid=137&category=&year=Latest&month=&tags=84&newsid=878. 18 xx K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M flowing by 2019, which gives sufficient time problems will be more important than others. to complete the legal framework and build Each will be relevant—to different extents—in the needed human capital (industry skills) and the Kenyan context. For example, managing physical infrastructure (for example, pipeline and the floating exchange rate will become more port upgrades). challenging under volatile inflows of U.S. dollar revenue. Almost all countries with weak Oil extraction will have considerable economic institutions and traditions of patronage politics implications, primarily through fiscal revenue. have experienced challenges in transforming The upstream oil (and gas) industry is not labor revenues derived from natural resources into intensive, nor does it have strong backward or assets for growth and shared prosperity. Further, forward linkages to the rest of the economy. there is evidence that the presence of natural The primary channel through which it will affect resource revenues has a negative impact on the Kenya’s economy is through the fiscal revenue it quality of institutions and governance.21 The will generate. An economic model presented in opportunities arising from increased natural chapter 5 finds that even the reserves that have resources in Kenya are therefore accompanied been discovered so far will bring substantial by significant governance risks that can best be revenue to the budget over a period of several managed with a strong emphasis on transparency, decades, and the development impact of this accountability, and stakeholder involvement. revenue is a factor for policy decisions.19 To respond to some of the questions raised, Resource discovery is not a guarantee of chapter 5 advises on what share of revenues development and whether it becomes a curse or should be saved or spent, how to allocate a blessing will depend primarily on the decisions spending, and what institutional mechanisms to taken related to five policy problems. Although put in place. For saving versus spending, this report the outcomes of the resource discovery are suggests that a permanent income hypothesis22 uncertain, it is certain the policy makers will have approach would best suit the characteristics of to solve five main problems associated with the the Kenyan economy. For where to allocate the resource extraction: (i) pressures on the tradable additional resources, priority should be given to sectors that would come from the inflows of increasing health spending while maintaining foreign exchange (Dutch disease), (ii) oil price the current share of infrastructure investment volatility (complicating public expenditure and in the total envelope for health, education, and investment decisions), (iii) over-borrowing infrastructure. The simulations show that such (borrowers and lenders feel less constrained an approach would bring the highest boost to in anticipation of future revenue flow), (iv) non-resource GDP and would lead to favorable sustainability (the amount of the natural wealth fiscal outcomes. This finding is derived from the to be preserved for future generations), and economic principle of diminishing returns to (v) corruption and mismanagement of revenue (infrastructure) investment, which is especially (the larger the resources, the more voracious true when there are implementation constraints, the incentives for rent-seeking).20 Some of these as is the case of Kenya. 19 Although there could be significant non-oil minerals produced in Kenya in the near term, the chapter limits its focus to oil. 20 See Giugale (2014). 21 See World Bank (2012b). 22 This approach implies constant government consumption (in real terms) of oil resources over time that is equivalent to interest income on the net present value of the country’s oil wealth. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y xxi Among the institutional mechanisms, fiscal Global experiences illustrate that there are many rules, transparency mechanisms, and a solutions to the design and management of an sovereign wealth fund are suggested as most SWF. The Kenyan authorities have been drafting important for achieving positive outcomes. First, an SWF bill since 2014, incorporating a broader establishing a fiscal rule, that is, imposing long- policy framework for managing resources. lasting constraints on discretionary fiscal actions, early on and strictly committing to targets Improving policy coordination in resource would serve well to enhance policy credibility. management is crucial for achievement of the International practice demonstrates that effective expected outcomes from resource revenues. fiscal rules are well defined, transparent, simple, Policy decisions require careful analysis and and to some extent flexible. The last feature deliberation, in particular for countries with is particularly tricky; Spotlight 5 argues that in multiple tiers of government that share countries with weak institutional capacity and responsibilities over the use of public resources. data—and Kenya would fall in this category—too It is unclear if Kenyan policy makers’ current much flexibility for countercyclical policies may legislative efforts are sufficient and aligned actually increase economic volatility, so a more with the best practices for the development rigid rule may be more effective. In the Kenyan impact of natural resources. The Constitution context where expenditure pressures may come of Kenya, especially articles 69 to 72, provides from the counties, a rigid fiscal rule may help the broad foundation of obligations for the central government in maintaining fiscal regulating environmental and natural resource stability. Second, transparency (and oversight) is management. Some progress has been made a critical pillar in the institutional framework, and to enact the necessary laws to operationalize although it cannot ensure the responsible use of these principles. For example, a mining law has resource revenues, without transparency, abuse been submitted to the Parliament and legislation is almost certain.23 One step in this direction is to related to various resources, such as ore, oil, or implement the Extractive Industries Transparency gas, is being drafted. However, much is left to be Initiative (EITI) standards.24 Nineteen African desired. Legislative efforts are typically done in countries, including neighboring Mozambique an isolated manner, whereas some of the policy and Tanzania, have already subscribed to issues, such as how to share resource revenue the EITI standards. In July 2015, the Kenyan among the levels of government, necessitate authorities announced that a focal point for EITI unified solutions. In addition, legislation has been implementation would be established within six proposed (for example, on mining) in the absence months. The authorities also announced that of a clear policy for the sector. Various policy a transparent policy and legislative framework proposals on critical issues, such as revenue would be adopted for the oil and gas sector, sharing, are being presented from different parties including the adoption of a transparent process in the form of legislative proposals and driven by for licensing and publication of contracts. special interests, which make consensus difficult. Last but not least, the sovereign wealth fund And legislative proposals are being drafted (for (SWF) has proven to be a good instrument for example, on the SWF) without in-depth analysis managing resource revenue, and such fund(s) to guide the proposed legislative solutions. may serve a saving or stabilization function. Unless the various stakeholders, in particular 23 Humphreys, Sachs, and Stiglitz (2007). 24 EITI.org. xxii K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M the central government, county governments, have a more developed higher education system and Parliament, start making coordinated and than Kenya. There is a wide consensus that these informed decisions on the management of are among the most important determinants natural resources, the oil discoveries are at risk of growth, and Kenya’s peers have indeed been to become a curse rather than a blessing for the growing faster over the past decade. Kenyan people. Finally, inadequate attention has been paid to managing the expectations The main policy implication of the above is that and needs of the local communities, and as oil achieving rapid growth and shared prosperity happened to be discovered in Kenya’s poorest will require continued action on multiple fronts. and conflict-prone region, addressing the Improving on the key determinants of growth economic and social needs of the people in those necessitates not only enactment of legislation, areas is critical for avoiding unnecessary conflict. but also its enforcement; more public investment Kenya can learn from the mistakes of other oil- and better execution of capital projects; greater producing countries in Africa and elsewhere to political and economic stability; and improved avoid falling into the same trap. governance. Progress on multiple fronts is happening in Kenya. What remains is to move up Conclusions and Next Steps a gear as well as to expand the focus on those areas that have received less than the needed To sum, several aspects of Kenya’s growth attention. This report argues for a range of short-, model over the past decade have been positive medium-, and long-term actions that would and can be built on, while in other areas there put the economy on the trajectory envisaged in are challenges to be overcome. What most Vision 2030. Kenyans know is that the economy has become much more dynamic and innovative in the past Short-Term vs. Long-Term Reform Priorities decade. Looking at the data, however, shows that more work is needed to make such economic For transformational growth, Kenya requires trends transformational: agriculture remains complementary policy reforms and investments. the mainstay of the economy (more than two- First, macroeconomic distortions that contribute thirds of Kenyans continue to live in rural to pro-cyclical policy and reduce saving should be areas); manufacturing has been unsatisfactory; addressed: lower inflation and less volatility in and the modern sectors, such as finance and public spending. Second, the government should communications, account for only a marginal consider investments that can unlock Kenya’s share of employment. comparative advantage; energy and transport infrastructure are among the main bottlenecks, The fact that in a global context Kenya’s growth and there is notable progress in these areas. performance has been modest at best comes as Third, the government will also need to consider no surprise when Kenya is benchmarked against pressing social sector needs, especially in the most important determinants of growth. education and health, in the context of the Countries at a similar level of development challenges and opportunities that devolution typically have greater macro-stability and higher brings. Fourth, breaking elite capture in the urbanization, are more open, invest more, spend agriculture sector—which is primarily a question more on health, have better governance, and of political will—and removing policy distortions FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y xxiii can give a boost to the currently undercapitalized affecting the informal sector. Fiscal policy farms and rural enterprises. Finally, for private should underpin macroeconomic stability by sector development, reforms should center maintaining a countercyclical role, given that on those areas that inhibit formalization and economic volatility has been high in recent years. enterprise growth. The government has recently Whereas the government has effectively started prioritized reform improvements to promote to direct more spending toward stemming the private investment and has made significant largest constraints to economic development, strides in improving the business environment, such as electricity generation, transport, and and should continue to further these reforms, as health, measures should be put in place to chapter 2 points out. ensure sustainability over the medium term without compromising fiscal space. The focus of In the short term, fiscal expenditure reforms hold education reform should move beyond increasing promise to yield immediate gains. As discussed primary and secondary enrollment to making in chapter 1, Kenya’s overall macroeconomic vocational training more in line with market management ranks better than that of its peers. needs and improving the quality of tertiary However, there is room for improvement, education. Kenya’s urbanization can be used to particularly on the fiscal front, where there drive economic growth and poverty reduction is a need to maintain and execute the share with better public infrastructure and services of the budget allocated to investment. Three in cities. Urbanization can also drive policies elements pose risk in this respect: (i) the rising that foster the specialization and agglomeration wage bill, (ii) falling execution of capital projects, economies that firms need to create more and (iii) weak coordination among central and formal sector wage jobs, while also promoting county governments. If the wage bill spirals out measures that create opportunities for informal of control, the share of public investment in entrepreneurs (chapter 2 shows examples from the budget will be permanently reduced. One other countries), given that the majority of those important issue that emerges is the need to moving to urban areas will not immediately be increase revenues collected from taxes at the able to join the formal economy. Devolution has county and national levels. However, this needs placed the responsibility for many public services to be complemented with improvements in the and some aspects of the business environment with public investment management process. Even county governments. Counties are following the if funds continue to be raised and budgeted for positive lead of the national government in looking investment projects, the economy will not benefit to improve their local business environments, unless projects are adequately implemented. but devolution work will be an essential piece of For capital projects, improvements in the the growth agenda. Finally, removing the policy public investment management framework and obstacles that choke agricultural development improved coordination between counties and would make a dent in poverty, given that most of the central government would help to raise the the poor are in rural areas effectiveness of public investment. Three elements of the long-term agenda stand The list of actions broadens over the medium out: increasing innovation, making the most term to other aspects of fiscal policy, as well of Kenya’s newly discovered natural resources, as education, urbanization, and agriculture and improving governance. Innovation activity is and areas of the business environment widespread in various parts of Kenya’s economy. xxiv K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Innovation typically takes the form of introducing which is a long-standing concern that remains new products or production processes, but the challenging, and where deeper and faster improvements tend to be marginal as firms in progress is still much needed. Kenya spend relatively little on R&D compared with firms in similar countries. In the long Next Steps for Further Exploration term, the government can stimulate more R&D This report takes a bird’s-eye view of Kenya’s (including through public funding) as well as economy, and a logical next step would be to enhance the quality of education to produce zoom in on some of the key bottlenecks and the needed skills. Kenya is expected to become come up with “how-to” ideas. The chapters an oil exporting country in 2020 at the earliest. in this report provide in some instances However, to make oil a success story, several recommendations or examples from other things would have to be in place: legal framework, countries. But the main focus of the analysis is accompanying infrastructure (pipelines), and institutions to manage the sector. The work on on identifying the problems. Hence, some of the each of these building blocks would need to start areas would require further work to get from today; as oil starts to flow, the emphasis would understanding the problem to designing solutions. shift to how best to put the resource revenue First, services range from construction to public to use. Great strides have been made in recent administration, and addressing the constraints to years to improve public sector governance: The growth requires delving into the specific issues in Constitution 2010, devolution, strengthening each subsector. Therefore, decoupling services of oversight institutions and improvement in and proposing solutions for particular subsectors core revenue and public financial management would be a natural continuation of the work in procedures are noteworthy achievements this report. Second, understanding better the which demonstrates significant commitment linkages between services and manufacturing and reform capability. However, for Kenya may help unleash the potential of the latter to further accelerate growth and poverty sector. Finally, understanding better the impact reduction, efforts are needed to follow through of subsectors of the economy—including the on these reforms and avoid reversing on the rapidly developing oil sector—is essential for progress made. A significant issue is corruption promoting shared prosperity. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y xxv xxvi RAN K E N YA C O U N T R Y E C O N O M I C M E M O Photo: M D U Muthembwa/World Keziah Bank CHAPTER 1 KENYA’S GROWTH STORY Kenya’s Recent Growth Performance Although the improvement in economic performance in the past decade has been Following two decades of stagnation in per remarkable, benchmarking Kenya’s economy capita income and high volatility of economic against similar peers from across the world activity, Kenya’s economy moved to a path of sheds light on Kenya’s relative success. For this accelerating growth after 2002. Gross domestic purpose, the report has identified a handful of product (GDP) growth increased steadily from countries from the continent and elsewhere at below 1 percent in 2002 to 7 percent in 2007. This a similar level of development as Kenya was a was the only episode of five-year accelerating decade ago.25 The countries in Sub-Saharan Africa growth in independent Kenya’s history, and it was (SSA) include Burkina Faso, Ghana, Senegal, the also the first time since 1986 that GDP growth Tanzania, and Uganda. The non-Africa peers reached 7 percent. Since 2007, the economy has are Bangladesh, Cambodia, India, Pakistan, and been hit by several shocks, starting with the post- Vietnam. In addition, Kenya’s performance is election violence in January 2008, which brought benchmarked against the high-growth countries GDP growth to a halt, followed by a slow recovery of the 1980s and 1990s that had a similar income in 2009. Economic growth has started to rebound level as Kenya in the 2000s (these include China, since 2010 and has stabilized since, although at Indonesia, and Thailand). Kenya’s average GDP rates lower than before 2008. per capita is higher than its SSA peers (Figure 1.1). However, growth volatility remained high in Figure 1.1: Kenya’s peers with similar GDP per capita Average GDP per capita, 2003 - 2014 (constant 2005 US$) the recovery phase. The standard deviation of 1200 GDP growth was the same (1.8) in the 1990s and 1000 post-2012. In recent years, political turmoil and 800 violence after the political elections in 2007 and 600 the global economic crisis magnified volatility. 400 This fluctuation in growth was caused by various 200 factors, such as political shocks (elections years have been associated with lower growth since 0 India Vietnam Senegal China Pakistan Kenya Ghana Cambodia Bangladesh Burkina Faso Tanzania Uganda Thailand Indonesia the 1990s), exogenous shocks (drought, oil prices, and global crisis), and macroeconomic policy shocks (relatively high inflation). Source: Calculations based on World Bank World Development Indicators. Note: The data for Thailand are for 1971–80. GDP = gross domestic product. Peer countries had +30 and -50 percent of Kenya’s GDP per capita in 2005. 25 FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 1 Compared with the average for the selected Figure 1.3: Services have been driving growth in Kenya peer economies, Kenya’s GDP growth was 4 lower. The average growth rate for Kenya was 1.7 3 percentage points lower than that for the (non- weighted) SSA peers and 3.1 percentage points Percentage points 2 lower than the high-growth economies during 1 the 1980s and 1990s (Figure 1.2). Furthermore, Kenya’s GDP volatility—which has been identified 0 2007 2008 2009 2010 2011 2012 2013 2014 as one of the main concerns for policy makers—is -1 the highest compared with the country’s peers. -2 Agriculture Industry Services The first step to understanding the reasons Source: Calculations based on World Bank World Development Indicators data. behind Kenya’s relatively weaker economic performance is to identify the drivers of economic Among the services sectors, communications, growth in the past decade. The approach taken trade, and financial services have been the star to this end is to examine the sector growth trends, performers. Their share in total gross value added the demand side of economic activity, as well as increased from 15.3 to 19.2 percent between changes in production factors. 2006 and 2013. Wholesale and retail trade has flourished, boosted by job market entrants Services have been the main engine of Kenya’s who find this to be the easiest way to generate economy over the past decade. Expansion of the services sectors accounted for almost two-thirds income outside farming. Road transport achieved of the increase in output between 2006 and 2014 the fastest growth—supported by the increase (Figure 1.3). Industry contributed more than 20 in the number of vehicles and rising regional percent of the increase, and the remaining 15 trade—and its share in total transport value percent came from agriculture. Services also added rose from 47 to 65 percent. Air transport proved to be most resilient following the 2008 output doubled because of increased tourist election crisis; while agricultural output fell and arrivals and expansion of operations by Kenya industrial output growth slowed in 2008 and Airways (the number of passengers rose from 2.0 2009, growth in services accelerated in 2009. million to 3.6 million between 2005 and 2013). Figure 1.2: Kenya’s economy had lower growth and higher volatility than its peers, 2003–14 Average rate of GDP growth Coefficient of variation 10 0.5 8 0.4 6 0.3 4 0.2 2 0.1 0 0 High growth Rest of the world SSA peers Kenya Kenya High growth SSA peers Rest of the world economies peers economies peers Source: Calculations based on World Bank World Development Indicators data. Note: SSA peers: Burkina Faso, Ghana, Senegal, Tanzania, and Uganda; non-Africa peers: Bangladesh, Cambodia, India, Pakistan, and Vietnam; high-growth economies: China, Indonesia, and Thailand. GDP = gross domestic product; SSA = Sub-Saharan Africa. 2 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M In contrast, railway traffic has been declining, in performance of the sector can be attributed cargo and passengers. Rapid mobile penetration partly to the efficient air transport that is used (from 19 percent in 2005 to 75 percent in 2013) for this product, while inefficiencies at the port and development of mobile payment services of Mombasa have been increasing the costs of boosted output in the communications sector. sea exports. Tobacco is another promising sector; The financial sector has also grown rapidly and output almost tripled in the eight years after Kenyan banks have begun to widen their presence 2005. Among the traditional sectors, although in the regional market. tea production has continued to grow, coffee production has slowed. As coffee prices have Within industry, mining and energy achieved fallen in real terms since 2005, production has above average growth, while manufacturing moved away from coffee (total production area was below average. Manufacturing, which fell by a third). Tea production area increased by accounts for the bulk of industry, has had a mixed 40 percent by 2013. However, the average yield performance. The food industry has proven to for tea fell in the same period, while for coffee be successful, while the other sectors have had the average yield grew 37 percent. slow or negative growth. The share of labor costs in total value added rose from 34 percent Compared with the peer group, Kenya stands in 2005 to 38 percent in 2013, which implies out in two ways: the contribution of services that wages are eating into the competitiveness to overall growth is highest, and the relative of the manufacturing sector or that low-skill performance of each sector is lagging behind labor intensive industries have been growing that of the peers. Services generated almost faster. Data from the financial statements of half of the GDP growth in the peer groups and manufacturing firms listed at the stock exchange two-thirds of Kenya’s growth. Correspondingly, confirm this trend: the share of labor cost in sales the contribution of industry in Kenya was lower rose from 8.6 to 10.6 percent between 2009 and than practically in all other countries, with the 2013. Mining has expanded rapidly since 2006, exception of Senegal. Although growth was because of the booming demand for and rising skewed toward services in Kenya, the sector grew prices of soda ash, gold, and fluorspar, although slower than in most of the peer group countries the price trends reversed in 2013. Finally, the (Figure 1.4). The growth of agriculture and energy sector has witnessed a marginal increase industry was even weaker in comparison with the in its share in the economy thanks to rising peer countries. The discrepancy is particularly electricity production in recent years. Electricity noticeable for industry: some countries from the generation increased by 31 percent between peer group had growth rates for industry that 2010 and 2014. Whereas geothermal grew were two or three times higher than Kenya’s. faster, hydro generation remained the main The gap is even wider when comparing the source, accounting for over 40 percent during performance of Kenya’s manufacturing with that the same period. of the high-growth economies identified in the Growth Commission Report 2008. In the latter, The agriculture sector is still endowed with rapid development of the manufacturing sector opportunities. Horticultural production has was found to be the key driver of rapid and boomed and the volume of flower exports rose sustainable economic growth. 32 percent between 2005 and 2013. The stellar FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 3 Figure 1.4: Each of Kenya’s sectors has fared worse share had fallen to 6 percent. For example, Kenya’s compared with peers, 2003–12 (GDP growth, %) market share in EAC’s market for chemicals and 9 paper has been stagnating or declining over the 7 past decade, while Chinese and Indian exporters have been expanding and have surpassed Kenyan 5 exporters in market share. At the same time, the structure of exports, comprising animal, mineral, 3 wood, and footwear products, has remained 1 largely unchanged. As a whole, net exports have been a drag on growth throughout most of the High growth SSA peers Rest of the world Kenya -1 economies peers past decade. Agriculture Industry Services Not Defined Figure 1.5: Consumption has been driving growth while Source: Calculations based on World Bank World Development Indicators net exports have been a drag data and Kenya National Bureau of Statistics. Note: High-growth economies data refer to the 1980s. GDP = gross domestic 12 product; SSA = Sub-Saharan Africa. 10 8 The finding of services being the driver of Growth rate (percentage points) 6 output growth is concurrent with the fact that 4 consumption has been the main contributor 2 to growth on the expenditure side. Rising 0 consumption (average annual growth of 5.1 2 4 percent), propelled by rising formal employment 6 (average annual growth of 2.8 percent) and credit 8 to the private sector (average annual increase of 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Consumption Net Exports Investment Not defined 20 percent) fueled growth throughout the past Source: Calculations based on World Bank World Development Indicators. decade (Figure 1.5). Investment has also made a positive contribution to growth year after year The clogged “exports engine” is what since 2003—a stark difference compared with differentiates Kenya from the peer countries, the preceding two decades. Interestingly, fiscal in particular those outside the Africa region. policy has been behind the increase in public Kenya’s goods exports have been relatively low investment, while private investment fell from within the peer group. In 2012, exports of goods 15 to 13 percent of GDP in 2012. The rising were 12 percent of GDP, while the successful consumption and investment generated rising East Asian countries have been producing and demand for imports of goods, which has not been exporting several times more (Figure 1.6). accompanied by a matching increase in exports. The weakness of the export sector has been On the contrary, exports as a percent of GDP exacerbated in recent years. Kenya was one of have declined since 2006. Contrary to popular the few countries in the group that recorded a expectations, exports to the fast-growing regional decline in the export-to-GDP ratio between 2005 East African Community (EAC) market—which and 2012. Several factors are suspected to be takes up a fifth of Kenya’s exports—have been the culprits for this trend: high cost of transport particularly disappointing. Kenyan exporters have (partly caused by inefficiencies in getting goods been losing market share: in 2006, 11 percent of to and from Mombasa port), appreciating real EAC’s imports came from Kenya and by 2013 the exchange rate, and weak manufacturing sector. 4 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Figure 1.6: Kenya’s goods exports have been low and with a GNI per capita (Atlas method) of US$1,160, declining following the rebasing of GDP. 40 Becoming a UMIC by 2030 is a formidable task. 30 The income thresholds move up practically each year—in line with inflation in the high-income Percent 20 economies—and in 2030 the threshold for UMICs is projected to be around US$5,600. To 10 reach this level, Kenya’s GNI per capita would need to increase fivefold over the next 15 years. 0 High growth economies Rest of the world peers SSA peers Kenya Looking in the rearview mirror, that is, at Kenya’s Export of goods % of GDP in 2012 Export of services % of GDP in 2012 past performance, the Vision 2030 goal seems Source: Calculations based on the International Monetary Fund World Economic Outlook database. farfetched. To begin, Kenya’s GNI per capita was Note: The data for Thailand are for 1980. GDP = gross domestic product; SSA = Sub-Saharan Africa. just below the lower-middle-income threshold in 1988 when the income classification was Services exports have fared much better. Unlike introduced, and then kept slipping for more most of the peer economies, Kenya has a than a decade. Things have improved in the past strong export-oriented services economy: only decade; however, even if Kenya’s GNI per capita Thailand and Uganda in the 1980s had a higher were to continue to grow at the historic rate services exports-to-GDP share. Travel services26 of the past 10 years, by 2030 its GNI per capita (tourism) are the largest services export, would still be far from the UMIC threshold. followed by transport. Services exports grew faster than GDP between 2005 and 2013, and Nevertheless, the Vision 2030 goal is achievable, transport services accounted for almost half of and similar successes have been noted the increase in exports. throughout the world. Achieving UMIC status by 2030 implies that Kenya’s GNI per capita Vision 2030 Goal would need to grow 10 percent annually for the Kenya has set a goal to become an upper-middle- next 15 years. Such rapid and sustained growth income country by 2030. The World Bank places has been witnessed in several countries around countries in four income groups: low-income the world, including in SSA. The lower-middle- countries, lower-middle-income countries, income countries’ GNI per capita grew at 10.6 upper-middle-income countries (UMICs), and percent per annum in the 2000–10 decade. This high-income countries. For this purpose, income high growth came mostly from the East Asia is defined as gross national income (GNI) per and Pacific region—which recorded an average capita based on the World Bank’s Atlas method. annual growth rate of 14 percent—but several The GNI per capita (Atlas method) thresholds countries in SSA have also grown at such a rapid for 2013 were: US$1,045 or less for low income, pace. Angola and Equatorial Guinea were among US$1,046 to US$4,125 for lower-middle income, the fastest growing countries in the world in the US$4,126 to US$12,745 for upper-middle income, 2000s: their GNI per capita rose by over 22 percent and US$12,746 or more for high income. In 2013, per annum during the decade (Figure 1.7). This Kenya became a lower-middle-income country growth was driven by oil exports, similar to the 26 Inflows from foreign tourists (spending on travel, accommodations, and food) is recorded as exports of travel services the balance of payments account. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 5 Figure 1.7: Several countries in SSA and East Asia have achieved Kenya’s desired pace of growth, 2000–10 (annual average GNI per capita, %) 25 20 15 10 5 0 Burundi Gambia Liberia Kiribati Togo CAR Madagascar Swaziland Tanzania Cote d'Ivoire Cameroon Kenya Mozambique Uganda Lesotho Comoros South Africa Congo, DR Mauritius Niger Senegal Malawi Rwanda Benin Botswana Mauritania Sierra Leone Cape Verde Philippines Thailand Namibia Burkina Faso Gabon Cambodia Mali Vietnam Ethiopia Guinea-Bissau Chad Zambia Ghana Lao PDR Congo, Rep. Sudan Nigeria Angola Eq. Guinea Source: World Bank. Note: GNI = gross national income; SSA = Sub-Saharan Africa. case of the next two fastest growing countries economy under the baseline scenario. The first in SSA: Nigeria and Sudan. Nevertheless, non- step in understanding how likely it is to achieve resource rich countries such as Ethiopia, Ghana, the required growth rate of 6.8 percent for a Guinea-Bissau, and Zambia also recorded GNI per sustained period of time is to assess how the capita growth rates of 11 percent or higher. economy has performed relative to its potential and what is its medium-term potential. The The required 10 percent GNI per capita growth potential growth rate refers to the rate of growth to achieve Vision 2030 translates into a real GDP of GDP when all available production factors annual growth rate of 6.8 percent. Under the (capital, labor, and technology) are fully utilized baseline scenario,27 Kenya’s annual GDP growth without producing inflationary pressures. Details rate should average 6.8 percent for the country on the methodological approach to estimating the to become a UMIC by 2030. Under pessimistic potential growth rate are provided in appendix C. scenarios that assume higher fertility or exchange rate depreciation, the required GDP growth rate The output gap estimates for the past two would need to be higher. There have been only decades indicate that the economy has been a few occasions in Kenya’s history when the mostly below its potential. For the historic economy grew by 6.8 percent. Looking forward, performance, three filter methods have been sustained annual growth of 6.8 percent or more applied to ensure the robustness of the results. is possible, but it would require bending the arc The results illustrate that despite higher growth of history. in the 2000s, the economy has been below potential for most of the past 15 years. The Estimating Kenya’s Potential Growth negative output gap was largest in 2003, and The previous section estimated the required then began to narrow, moving to a positive value growth rate to achieve the Vision 2030 goal; this in 2006. The positive output gap peaked in 2007, section assesses the potential growth for the but then turned negative after 2008 (Figure 1.8). The assumptions are population growth of 2.2 percent per year, a stable $US-K Sh exchange rate, and parity between GDP and GNI. 27 6 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Figure 1.8: The output gap has turned negative since 2008 Figure 1.9: Comparison of projected actual and potential (percent) GDP growth rates for 2014–18 4 8 7 3 6 2 5 1 Percent 4 0 3 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 -1 2 -2 1 -3 0 2014 2015 2016 2017 2018 -4 Poten tial GDP growth rate IMF projections Hodric-Prescott filter Source: Calculations based on data from World Bank World Development Source: Calculations based on data from World Bank World Development Indicators and International Monetary Fund World Economic Outlook. Indicators and International Monetary Fund. Note: GDP = gross domestic product; IMF = International Monetary Fund. Kenya’s potential growth rate is below what is to rely mostly on expansionary fiscal policy to sought in Vision 2030 and the Second Medium- give a short-term boost to growth (beyond Term Plan (MTP-2). The potential GDP growth potential). It seems that Kenya’s economy is at rate up to 2020 is just above 6 percent. Under a low risk of overheating, that is, there is no risk this scenario, Kenya would not become a UMIC of domestic demand-driven inflationary pressures by 2030. Assuming the potential growth rate over the medium term. This situation implies that remains at 6 percent until 2030, GDP per capita there is room for policy interventions aimed at in 2030 will be $2,303, which is lower than that accelerating GDP growth. To understand the policy for a UMIC. options for monetary and fiscal decision makers, it is useful to begin by examining the cyclicality of Unlike the past decade, Kenya’s GDP growth fiscal and monetary policy in the past. starting from 2015 is projected to be above potential. The latest International Monetary Fiscal policy has been increasingly correlated Fund (IMF) projections indicate that growth will with the business cycle since the beginning of average 7 percent over 2015–18, which is higher the century. The synchronization between fiscal than the potential growth rate of 6.2 percent policy and the business cycle is assessed using (Figure 1.9). These projections rest on the two measures of fiscal outcomes: the primary assumption that the external environment will be budget balance and the cyclically adjusted28 positive and macroeconomic policies, in addition primary budget balance as an indicator for the to structural reforms, will be growth oriented. discretionary fiscal policy stance. The former fiscal At the same time, the economy will face several policy indicator does not point to a relationship downside risks, including on the policy front. between fiscal policy and the output gap prior to 2008. The cyclically adjusted primary budget Although in the long term the authorities aim balance is found to be negatively correlated with to raise potential output through investment the output gap, which suggests that discretionary in infrastructure and education, they are likely fiscal policy was on average pro-cyclical prior The cyclical adjustment of the budget balance has been done according to the aggregated approach that is conducted by the IMF. 28 According to this approach, the elasticity of revenues and expenditures with respect to output gap fluctuations are taken to be 1 and 0, respectively. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 7 to 2008. Since 2008, fiscal policy outcomes However, this policy direction has to take have improved as both fiscal indicators signal a into consideration the fiscal weaknesses that countercyclical fiscal policy stance (Figure 1.10). undermine the effectiveness of countercyclical The correlation coefficients range from 0.3 to fiscal policy. More precisely, the decision on 0.8, which is a relatively strong (statistically further fiscal loosening should incorporate the significant) positive relationship. In the absence following: (i) contingent liabilities arising from of this fiscal boost, the economic recovery after the devolution process; (ii) pressures to increase 2008 would have been weaker and possibly recurrent spending (wage bill), which is difficult more volatile. to reverse; and (iii) the deteriorating efficiency of investment expenditure. Ambiguity in some Figure 1.10: Fiscal policy moved from being pro-cyclical to countercyclical in 2008 aspects of the devolution process has created 4 potential liabilities that may end up in the central 3 government’s budget. For example, lack of 2 1 clarity in the transfer of staff and assets from the 0 central to the county level, as well as potential Percent -1 borrowing by counties, may pose a burden to the -2 central budget. The wage bill, which has swelled -3 at the central and county levels, is another risk -4 -5 to fiscal sustainability, since these entitlements 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 are difficult to curb. Finally, concerns about the General government primary balance % of GDP Cyclically adjusted primary budget balance % of GDP declining efficiency of public investment stem from falling execution rates, cutting of operations Output gap Source: Calculations based on World Bank World Development Indicators and International Monetary Fund World Economic Outlook database. Note: GDP = gross domestic product. and maintenance spending, and inadequate prioritization of spending. All these weaknesses Kenya’s macroeconomic position allows for would need to be addressed for countercyclical continued countercyclical fiscal policy to smooth fiscal policy to boost output growth in a the negative output gap in the medium term. sustainable manner. Revenue collection has not been stellar over the past few years, in particular in the administration Monetary policy has been focused mostly on of value-added tax, and the IMF projects that neutralizing price shocks in the economy. This the revenue-to-GDP ratio will remain at the focus is somewhat expected, as the monetary same level in the medium term. At the same policy regime is a type of inflation targeting, so time, the MTP-2 envisages increased spending the Central Bank of Kenya reacts primarily to on infrastructure and social sectors, which will inflationary pressures, which in the past few years result in continued budget deficits. The 2014 have often come via exogenous shocks (drought IMF-World Bank debt sustainability analysis or oil prices) rather than domestic demand concludes that Kenya’s public debt is sustainable pressures. For that reason, there was not much and the country faces a low risk of debt distress space for monetary policy to react to business (IMF 2014). Hence, a moderate increase in the cycle fluctuations (Figure 1.11). For the medium budget deficit would be possible. term, a substantial shift in the policy stance is not expected. The data suggest that, in absence of exogenous supply-side shocks, key policy rates would stay unchanged. 8 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Figure 1.11: Monetary policy has not been fully in sync Shocks in economic activity in major export with the real economy partners have an immediate impact on Kenya’s 20 economy, but with low magnitude. A shock 15 of one standard deviation in the foreign GDP 10 per capita growth has a 0.7 percentage point 5 effect on Kenya’s GDP per capita growth. The Percent 0 impact is transmitted in the current period -5 and persists for almost a year. The variance decomposition results indicate that variations in -10 foreign economic activity account for around 15 -15 2006 2007 2008 2009 2010 2011 2012 2013 2014 percent of the variation in domestic economic Nominal central bank rate (end of period) Real central bank rate (end of period) Output gap HP filter activity. This suggests that foreign GDP shocks Source: Calculations based on data from World Bank World Development Indicators and the Central Bank of Kenya. are quickly transmitted to Kenya, but their size Note: HP = Hodrick-Prescott of transmission is less than complete and they can explain only a limited part of the overall Managing Economic Volatility fluctuations in domestic output. In addition to growing below the desired, or potential, level, Kenya’s economy over the past In the transmission of foreign shocks to decade has also been coping with high GDP inflation, again the results indicate a positive volatility. Although growth has been positive and statistically significant relationship. A one throughout the decade, the economy recorded standard deviation shock in foreign effective significant swings, with annual GDP growth inflation leads to an increase in domestic inflation ranging from less than 1 to 7 percent. Recognizing of 6 to 9 percentage points (appendix A). Foreign the causes of economic volatility—the most inflation shocks are transmitted in the current important ones and how long the pass-through period and persist for a longer time horizon (even effect lasts—is a necessary step toward achieving up to four years). The variance decomposition higher and sustained growth, so that policy method implies that fluctuations in foreign makers can prepare for and react to anticipated inflation are the dominant factor dictating price shocks. Broadly defined, the causes of economic movements in Kenya, as they account for up to 71 fluctuations can be external or domestic. percent of the variability in inflation. The transmission of exogenous shocks to the Digging deeper, the results show that shocks in domestic economy can be assessed via two foreign food prices are likely to affect Kenya’s channels. The first channel is when economic inflation to a much greater extent compared with activity in key trading and/or investor partner shocks in oil prices. To examine the impact of the countries influences domestic economic activity. transmission of food and oil process on Kenya’s The second channel is via the pass-through inflation, the same method is applied using effects of foreign inflation, in particular, foreign quarterly data for 2001–13, because variations effective inflation, global commodity prices, and in food and oil prices are greater and usually oil prices, on domestic inflation. The empirical last shorter (a few months), so the pass-through approach and results can be found in appendix A. effect can be captured more precisely with lower FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 9 frequency data. The impulse response functions illustrates that exogenous foreign shocks are indicate that there is a significant reaction of not fully transmitted to Kenya’s economy and Kenya’s inflation to shocks in food prices, but not domestic shocks explain much of the variability to oil prices, at the 5 percent level of significance. in GDP growth. These findings imply that A shock of one standard deviation in food prices reducing volatility is primarily a question of results in an increase in Kenya’s inflation from domestic policies. The next section describes the 2.4 percentage points within two quarters delay determinants of growth in Kenya, which in turn and up to 10 percentage points after 10 quarters should guide decision makers toward the policy delay. The variance decomposition exercise priorities that deserve greater focus to reduce indicates that fluctuations in world food prices volatility and accelerate growth. explain up to two-thirds of the variation in Kenya’s inflation. The transmission of shocks in Factors behind Kenya’s Economic oil prices to domestic inflation in Kenya was not Performance found to be statistically significant, even when One of the main questions for Kenya’s policy using monthly frequency data. makers is how to accelerate economic growth. Vision 2030 and the MTP-2 call for faster and The same empirical approach illustrates that sustained growth, an outcome that is different much of the economic volatility that Kenya from historic economic performance, which has has experienced over the past decade has been below potential and volatile. To understand been domestically driven. First, changes in the the determinants of growth, this chapter adopts investment-to-GDP ratio are shown to have a twofold approach. The starting point is the a positive effect on GDP per capita growth of vast economic literature on the determinants of between 0.7 and 2.5 percentage points. The growth. Then, Kenya’s performance in each area shocks are transmitted with a delay of one is benchmarked to its group of peer countries, year and persist up to six years. A shock of which also includes some of the East Asian one standard deviation in government final Tigers during their initial boom periods. Chapter consumption positively affects GDP per capita 4 complements the analysis on the growth growth between 0.6 and 1.5 percentage points. determinants using a complementary approach: The transmission of the shocks occurs with a it applies the product space methodology delay of one year and persists up to three years. (Hidalgo et al. 2007) to examine the complexity Government final consumption explains between of the Kenyan economy. 15.5 and 20.9 percent of the volatility in GDP per capita growth. Finally, a shock of one standard The economic literature groups the determinants deviation in Kenya’s inflation has a negative of growth broadly into three categories: impact on GDP per capita growth of 1.1 to 1.6 structural policies and institutions, stabilization percentage points. The shock from inflation is policies, and external conditions. Structural transmitted within the first year and persists up policies and institutions typically include the to two years. Inflation explains nearly one-fifth country’s human capital, business environment of the variance in GDP growth. and financial sector, size of government, trade openness, and quality of public institutions The main conclusion from this analysis is that and governance. Stabilization policies capture economic stability is primarily a function of macroeconomic conditions, including inflation, domestic policies. In sum, the empirical analysis output volatility, and the real exchange rate. 10 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M External conditions refer to the exogenous factors enrollment rates (primary, secondary, and tertiary) that influence growth, such as the external and the average years of schooling. Although environment in major trading partners. the quantity of education is important, so is quality; however, many low- and middle-income The analysis of growth determinants for Kenya countries, including Kenya, lack internationally follows a benchmarking approach. Under this comparable data on the quality of their education approach,29 the most important factors of growth systems.31 In primary and secondary school are analyzed and compared with Kenya’s peer enrollment rates, Kenya outperforms most of countries. The peer group comprises economies the countries in its peer group (Figure 1.12). with current similarities to Kenya (in GDP per Consequently, Kenya’s population’s average capita and population), as well as some of years of schooling, at 6.5 years, is among the the high-growth economies during the 1980s highest, with only Ghana having a more educated identified by the Growth Report (Commission on population. However, marginalized groups Growth and Development 2008).30 The rationale (those living in arid lands, pockets within urban for including some of the East Asian Tigers in settlements, and some of the coastal areas) the list is because in the 1980s or 1990s these have not witnessed the same improvements in countries were growing rapidly and had initial enrollments. In contrast, tertiary enrollment is conditions that were similar to today’s Kenyan an area where Kenya is a clear outlier, having economy. the lowest enrollment rate in the group in 2009.32 Since then, the authorities have taken Structural Determinants strong action to boost university education and Human capital is perceived as one of the enrollment increased by 10 percent in 2011. This most significant structural determinants for trend of a growing number of university students sustainable long-term growth. For international is expected to continue, but the transition from comparison purposes, the quality of human secondary to university education, which was 6.5 capital is typically captured through gross school percent in 2010, remains low. Figure 1.12: Kenya has the highest secondary and lowest tertiary enrollment among the peers Secondary school enrollment (gross) Tertiary school enrollment (gross) 60 14 12 50 10 40 8 30 6 20 4 10 2 0 0 Kenya Rest of the world "High-growth" SSA peers Peers from the rest "High-growth" SSA peers Kenya peers economies of the world economies Source: World Bank World Development Indicators. Note: The data for secondary school enrollment for Vietnam are for 1998 only. The data for Kenya for gross tertiary school enrollment are for 2009 only. SSA = Sub-Saharan Africa. 29 See Johnson, Ostry, and Subramanian (2006, 2007). 30 Those economies are Indonesia, Malaysia, Morocco, and Thailand during the 1980s. 31 Information on the quality of education is available from several sources: Kenya National Examinations Council has annual student learning achievement data; Southern and Eastern Africa Consortium for Monitoring Educational Quality collects regional student learning data; and UWEZO collects annual national student assessment data. 32 According to the World Development Indicators, the gross tertiary school enrollment rate increased from 3 in 2005 to 4 in 2009. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 11 Urbanization is another important characteristic another third lives in cities with 100,000–500,000 of successful low- and middle-income population. Looking at 2030, Kenya’s urbanization economies, and this is an area where Kenya rate is projected to increase rapidly, mainly has a lot of catching up to do. The urban share because of the increase in the number of cities of the population has been confirmed to be a with population of more than 250,000. relevant indicator for a country’s development. Urbanization positively influences economic In a fast-urbanizing, non-resource rich, low- growth through greater technological progress or middle-income country, it is typically occurring in urban areas mainly through the manufacturing sector that generates manufacturing production and some services, integral migration to urban areas.33 Increasing which in turn raises labor productivity. Another employment in manufacturing creates benefit of urbanization is the agglomeration “production cities” that in turn generate effect, that is, when the know-how, knowledge, demand for urban goods and services. However, and technology found in urban areas are applied urbanization in Kenya has been driven largely in agricultural production, as well as more by the services economy, in particular informal efficient commuting between urban and rural services such as trade. areas. According to the United Nations’ definition of urban population, Kenya’s level of urbanization Post-World War II (WWII) history clearly is the lowest among its peer group, and lower illustrates that countries relied on manufacturing relative to its GDP per capita (Figure 1.13). or abundant natural resources to achieve rapid However, those estimates do not include the and sustained growth. The Growth Report peri-urban population, which, if included, would (Commission on Growth and Development 2008) put Kenya’s urbanization rate at 30–35 percent, found that expanding the manufacturing sector which would put Kenya ahead of half of its peers, was one of the key ingredients behind the success but still below the expected value relative to its of most of the dozen or so economies that GDP per capita. More than a third of the urban managed to grow at 7 percent per annum for an population lives in Nairobi and Mombasa, while entire generation. In practically all cases, growth Figure 1.13: Urbanization in Kenya has been low relative to GDP per capita Source: World Bank World Development Indicators. Note: GDP = gross domestic product; PPP = purchasing power parity. Gollin, Lagakos, and Waugh (2013). 33 12 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M in manufacturing production was accompanied Figure 1.14: Manufacturing in Kenya Is underdeveloped by rising manufacturing exports, that is, compared with its non-Africa peers 80 countries were producing for global markets and in particular for high-income economies. The 60 manufacturing sector is perceived to be one of the key drivers of technology development, know- 40 how, and, consequently, productivity growth. 20 Kenya has yet to move to a path of expanding manufacturing production and exports. The 0 Rest of the "High-growth" Kenya SSA peers share of manufacturing value added to GDP world peers economies and the share of manufacturing exports to total Share of manufacturing exports to merchandise exports Share of manufacturing to GDP merchandise exports in Kenya are higher than in Source: World Bank World Development Indicators. Note: Values are for 2005–14 for Kenya, SSA, and the rest of the world peers, the SSA peers, but lower than in the high-growth and for the 1970s or 1980s for the high-growth economies. GDP = gross domestic product; SSA = Sub-Saharan Africa. economies and peer countries in the rest of the world (Figure 1.14). The share of this sector measured by the ratio of credit to the private in Kenya’s GDP has remained unchanged over sector to GDP, is impressive for the country’s the past decade. This suggests that the Kenyan level of development (Figure 1.15). This suggests manufacturing sector should play catch-up with that the financial sector has been a driving force its peers in the rest of the world and the high- behind Kenya’s economic performance. growth economies. Capital markets are relatively well developed Kenya has reaped the benefits of expanding in Kenya. Stock market capitalization, as a share financial services. Broad empirical evidence of GDP, is the best measure of the development identifies the financial sector as a catalyst for of capital markets. Kenya’s stock market economic growth. A more developed financial capitalization surged to over 50 percent of GDP system increases financial inclusion and thus in 2014, higher than in most peer countries. helps the economy to mobilize savings and Some 60 companies are listed, which is more allocate them more easily and more efficiently to than the listings in Tanzania’s and Uganda’s investment needs. Kenya’s financial sector depth, bourses combined. Figure 1.15: The financial sector is more developed than in peers 200 40 160 30 120 80 20 40 Kenya 10 0 0 4 5 6 7 8 9 10 11 12 Kenya Rest of the "High -growth" SSA peers world peers economies Sources: World Bank World Development Indicators and Finstats data set. Note: The graphs show the private credit-to-GDP ratio and log GDP per capita (scatter) and the stock market capitalization-to-GDP ratio (bar chart), 2005–13. The red spot in the scatter graph is Kenya; the black spots are Kenya’s peers; and the blue spots are the countries in the rest of the world. GDP = gross domestic product; SSA = Sub-Saharan Africa. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 13 Sustained growth requires high levels of Burkina Faso, where the public sector had higher investment and this is an area where Kenya levels of investment. aims to achieve more. High-growth economies in the post-WWII period have had high levels Trade openness, that is, trade integration of investment, typically 25 percent of GDP or with the rest of the world, is a well-known higher (Figure 1.16). Kenya’s investment rate was determinant of growth on which Kenya performs below 25 percent of GDP during 2005–14. Thus, well. Greater international trade stimulates Kenya has the lowest investment rate among the economic growth, as it allows domestic producers peer group, with the exceptions of Cambodia to expand their production, or sales, to foreign and Pakistan. Investment in Kenya during the markets. Openness to trade also facilitates the reference period came largely from the private transfer of technology and know-how, which sector. Among the peer group, the private sector boost productivity. Another contribution of played a dominant role in investment, except for trade openness is that it increases competition Figure 1.16: Kenya has the lowest investment-to-GDP ratio in the domestic market and makes local and the highest investment risk, 2005–14 production more efficient.34 Kenya’ policies Gross fixed capital formation to GDP 30 to promote trade regionally and beyond have 25 paid off. Trade openness, measured by the sum of exports and imports to GDP, has remained 20 above 50 percent in the past decade (Figure 1.17). 15 However, Kenya’s trade openness is relatively low 10 within the peer group (it is higher compared with 5 the high-growth economies because global supply chains were in an infant stage in the 1980s and 0 High growth SSA peers Rest of the world Kenya 1990s) (Figure 1.17). At the same time, Kenya has economies peers Gross private fixed capital formation to GDP subscribed to an open trade policy—its applied Sources: World Bank World Development Indicators; Political Risk Services Most Favored Nation tariff is among the lowest group (www.prsgroup.com). Note: The data for Cambodia, Ghana, India, Pakistan, Tanzania, and Thailand in the group (only Uganda has a marginally lower are averages for 2005–13. GDP = gross domestic product; SSA = Sub-Saharan Africa. tariff rate). Figure 1.17: Kenya’s trade openness remains resilient, although it is lower than that of its peers Exports and imports (% of GDP) 100 70 80 60 60 40 50 20 40 0 Rest of the SSA peers Kenya "High -growth" 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 world peers economies 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Trade openness to GDP Simple aaverage MFN applied rate Source: World Bank World Development Indicators. Note: The data for the tariff rate are through 2012. The data refer to the 1980s for Thailand and Indonesia, and the 1990s for China. GDP = gross domestic product; MFN = most favored nation; SSA = Sub-Saharan Africa. A more comprehensive overview of the role of international trade in growth is provided in the Growth Report (Commission on Growth 34 and Development 2008) and Calderón, Fajnzylber, and Loayza (2005). 14 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Government spending can stimulate growth than Kenya (6.5 percent of GDP), but Vietnamese if the spending is geared toward physical and 15-year-old students are among the best in the human capital enhancement, but it can also be world in mathematics, reading, and science. a drag on growth if it is excessive and focused on Vietnam was among the top 15 performers on non-productive items. If government spending the 2012 Programme for International Student is directed toward productive sectors, like Assessment test—ahead of Australia, the United infrastructure, education, and health, then greater Kingdom, and the United States—and although government spending should facilitate economic Kenya has not participated in this international growth. As identified by the Growth Report, the benchmarking exercise, the evidence points success of the high-growth economies is partially to much worse performance. Public health owed to the high public spending in some of the expenditures are relatively low in Kenya and the aforementioned areas. Excessive non-growth- effectiveness of saving is likely lower as measured enhancing spending, such as civil service wages, by outcomes. Kenya’s life expectancy is 60 years, which spur domestic demand in the short term, compared with 65.4 years for the peer group. may have negative consequences on economic Figure 1.18: Development spending is relatively high, growth. For example, it may lead to crowding out 2005–2012 average the private sector, increasing indebtedness, and 30 raising uncertainty. Kenya’s public expenditures are by no means excessive. However, what the 20 money is spent on and how the money is spent Percent of GDP are the more important questions. The allocation and spending of resources has shifted under the 10 new system of devolved government, so the counties will have increasing responsibility in the public spending–growth relationship. Spotlight 1 0 SSA peers Kenya Rest of the "High -growth" (at the end of this chapter) discusses in greater Public expenditures on education to GDP world peers economies Public investment to GDP Public expenditures on health to GDP detail the effects of devolution on growth. Other publuic expenditure to GDP Sources: World Bank World Development Indicators; International Monetary Fund World Economic Outlook. Note: The data for the high-growth countries are for 2005–12 because data Public spending on education in Kenya is for the 1980s and 1990s are not available. GDP = gross domestic product; SSA = Sub-Saharan Africa. the highest, but not necessarily efficient or sufficient. Kenya’s education budget, at over Good governance can facilitate inclusive 6 percent of GDP, is larger than in any of the growth, and is one of the key impediments peer countries (Figure 1.18). However, not all to fully unleashing Kenya’s growth potential. the spending is growth enhancing. For example, Strong, effective, transparent, and accountable a 2012 Public Expenditure Tracking/Social governance institutions create the enabling Development Indicators Education Survey found environment for broad economic growth. that public teachers—whose salaries account for Ineffective governance, and corrupt institutions 70 percent of total expenditure in the sector—are reduce the prospects for sustainable growth not teaching 45 percent of the scheduled time. In and poverty reduction and take away resources contrast, some of the benchmark countries have that are meant for delivering public services to advanced education systems that deliver better businesses (such as infrastructure, business results while spending less. For example, Vietnam services, regulation) and citizens (such as health, spends less on education (6.2 percent of GDP) drinking water, and education).35 (see box 1.1.) World Bank (2012b, 10). 34 FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 15 Box 1.1: Corruption and access to water Corruption is estimated to raise the price of connecting a household to a water network in low- and middle- income countries by as much as 30 to 45 percent. Poor people living in slums not connected to the water grid frequently pay far more for water than connected customers do. Globally it is estimated that 20 to 70 percent of lost resources in the water sector could be saved if transparency was widespread and corruption was eliminated. In Kenya, corruption in the water sector is characterized by bribery, unaccounted for water fees, and procurement processes that are not transparent. According to survey work by TI Kenya, 87 percent of respondents in Nairobi had witnessed the payment of bribes to connect to the city’s water network. With wide-scale corruption in the water sector, achieving the global Millennium Development Goal (MDG) target of improved access to water could cost an estimated US$48 billion more than planned. Further, the survey results based on data for 51 countries reveal that a population’s access to safe drinking water is negatively correlated with the level of bribery practiced in the country irrespective of the level of national per capita income and the money invested by the government in public infrastructure for water and other services. Overall, the study revealed that increased transparency, accountability, and integrity could lead to better education and health outcomes and increased access to water, which are important MDGs. Source: Transparency International 2010, 2014. Over the past decade, Kenyans have made systems. Above all, the elections were peaceful progress in a range of governance reforms, and were not accompanied by macro fiscal accelerated by the 2010 Constitution. Progress mismanagement. The reforms bear testimony has been made in areas of economic governance, to significant support and capacity for including revenue administration at the national governance reforms. level; the passing of a public financial management law that, inter alia, regulates the use of budget Despite the progress, Kenya still faces significant and control; and the establishment of the Office governance challenges. Various indicators are of the Controller of the Budget and a Supreme available to assess governance in Kenya vis-à- Audit Institution. A Treasury Single Account has vis in the rest of the peer group. Although these been introduced and work on program-based governance indicators have their limitations, budgeting is ongoing. Reforms of the judiciary using several sources of information confirms have been initiated and work is in progress to the robustness of the assessment that Kenya has put in place a credible, public, and transparent several governance challenges and weaknesses. process for scrutinizing the appointments of One source of information on the relative quality all senior public officers. The defunct Anti- of governance in Kenya is the Bertelsmann Corruption Commission has been replaced with Foundation Transformation Index (BTI). The a new Ethics and Anti-Corruption Commission overall BTI comprises many components that with some operational independence, although measure various themes beyond governance, it has recently been embroiled in controversies. such as investment/trade restrictions, etc. For Devolution has taken off since the election of the purpose of this analysis, a restricted: index March 2013, bringing governance and service based on relevant governance subcomponents of delivery closer to citizens, but with significant the index is used.36 The following components of challenges in the capacity of institutions and the BTI index are included (Figure 1.19): The aggregate value of the governance index is calculated as a simple average of the selected components that capture governance. 36 This approach is consistent with the methodology for calculating the value of the overall BTI. 16 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M 1. To what extent does the government The BTI suggests that Kenya has the potential for successfully contain corruption? improvement compared with the benchmarking 2. How effective is the government in countries. This is especially the case in the implementing its own policies? area of monopoly on the use of force through its whole territory (reflecting incidents of 3. To what extent are private companies terrorism), property rights, and anti-corruption permitted and protected? Are privatization processes conducted in a manner consistent policy. Corruption, in particular, has been a with market principles? major challenge for Kenya’s performance on international measures. In recent times, the 4. To what extent do government authorities current president has begun to take some action, ensure well-defined rights of private property including suspension of some members of the and regulate the acquisition, benefits, use, cabinet and senior civil servants who are being and sale of property? investigated for allegations of corruption. 5. To what extent do safeguards exist to prevent the development of economic monopolies Despite reform initiatives in Kenya, the World and cartels, and to what extent are they Bank’s Worldwide Governance Indicators show enforced? similar results and indicate that there has been 6. To what level have the fundamentals of little measurable progress in the past decade. market-based competition developed? Figure 1.20 shows the governance indicators for 7. To what extent are public officeholders Kenya compared with all countries in SSA, high- who abuse their positions prosecuted or growth peer countries, and other peer countries penalized? used for comparison in this report. The indicators 8. To what extent does the state’s monopoly on include control of corruption, rule of law, political the use of force cover the entire territory of stability and absence of violence/terrorism, the country? voice and accountability, and government effectiveness. The further away a measurement Figure 1.19: BTI indicators of Governance in Kenya and Figure 1.20: Governance indicators show Kenya lags benchmark countries behind benchmarking economies 9.0 Control of Corruption 8.0 7.0 Voice and Government 6.0 Accountability Effectiveness 5.0 4.0 3.0 2.0 Political Stability and Rule of Law Absence of 1.0 Violence/Terrorism on n ty l y ed of n pti tio te se er op as n on se y o ce rru y ta iva ri op ts on y t-b tio uti u ol for en Pr erp Pr righ -m olic ke peti ec e ab op of - co olic em t ti r s n nti p pl en An p a M com o c Pr offi o e M e us A Im Regulatory Quality th SSA peers Rest of the wor ld Kenya High gr owth A ver age benchmar king gr oup SSA Average Kenya Rest of the world High growth - Source: http://www.bti-project.org. Source: World Bank Governance Indicators, http://data.worldbank.org/data- Note: SSA peers: Burkina Faso, Ghana, Senegal, Tanzania, and Uganda; catalog/worldwide-governance-indicators. Data extracted in February 2016. high-growth: China, Indonesia, and Thailand; rest of the world: Bangladesh, Note: SSA peers: Burkina Faso, Ghana, Senegal, Tanzania, and Uganda; Cambodia, India, Pakistan, and Vietnam. BTI = Bertelsmann Foundation high growth: China, Indonesia, and Thailand; rest of the world: Bangladesh, Transformation Index; SSA = Sub-Saharan Africa. Cambodia, India, Pakistan, and Vietnam. Values are expressed in units of the standard distribution, ranging from -2.5 to +2.5. The World Bank Governance Indicators are aggregate indicators that combine the views of many enterprise, citizen, and expert survey respondents. The indicators are based on 32 individual data sources. SSA = Sub-Saharan Africa. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 17 point is from the center of the figure, the higher Competitiveness Report for 2015–16 (World or better the score on the indicators. The figure Economic Forum 2015), while at a 99th rank out shows, for example, that Kenya is aligned with the of 140 countries placing Kenya above its SSA average for SSA for voice and accountability, but is peers identifies corruption as the top problematic below the average for SSA for the rule of law. factor for doing business in Kenya. Kenya fares well on regulatory quality and The rule of law governance indicator captures voice and accountability although below the perceptions of the extent to which agents have level of high growth countries. Regulatory confidence in and abide by the rules of society. quality captures perceptions of the ability of the The particular focus of the indicator is on the government to formulate and implement sound quality of contract enforcement, property rights, policies and regulations that permit and promote police, and courts, as well as the likelihood of private sector development. Kenya scores crime and violence. The implementation of relatively well on this indicator. reforms under Kenya’s 2010 Constitution is beginning to bring improvements across the Kenya scores high on the voice and accountability board, and especially related to property rights indicator. This indicator captures perceptions of and rule of law in general. For example, the the extent to which citizens are able to participate government launched a land titling program in in selecting their government, as well as freedom 2013 with a target to issue three million land titles of expression, freedom of association, and a free in four years. There are ongoing reforms in the media. These aspects of governance have been police and judiciary, and judicial independence strengthened by Kenya’s 2010 Constitution, has significantly improved to strengthen the rule which is very strong on citizen participation and of law. consultation as a requirement, for example, in planning and budget formulation. For example, Kenya has had a history of political violence there have been instances where governance with detrimental effects on the economy. Kenya, policies and legislation have been challenged together with India, Indonesia, and Uganda, successfully in court for lack of consultation with scores low on the sub-index of political stability and participation of citizens. and violence. The indicator reflects the violent and fierce competition that has historically The control of corruption indicator captures characterized Kenya’s elections and created perceptions of the extent to which public uncertainties before and after elections. This in power is exercised for private gain. This refers turn has resulted in volatile economic activity to petty and grand forms of corruption, as well around the election cycle, with investors adopting as “capture” of the state by elites and private a “wait and see” attitude in the period preceding interests. This indicator suggests that corruption and immediately after elections. The elections of is perceived to be present to a great extent March 2013 marked a positive turn, in the sense in Kenya. According to this indicator, Kenya, that they were relatively peaceful and with no together with Bangladesh and Cambodia, is significant adverse effect on economic activity. perceived by its citizens to be among the most Nevertheless, in recent years, insecurity has been corrupt societies of the benchmark countries. a concern, with terrorist attacks in parts of the This perception is corroborated by the Global country and negative impacts on tourism, which 18 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M is a key sector of the economy. To illustrate, IDA country average of 3.2. Kenyas lowest score the number of international visitors dropped was on the public sector management cluster from 1.7 million in 2012 to 1.5 million in 2013.37 with 3.4. Kenya ranked 139th on Transparency Recently (August 2015), several countries have International’s Corruption Perception Index in eased their travel restrictions, and charter flights 2015 (Figure 1.21). from Europe have begun to resume. Figure 1.21: Perception of corruption in Kenya and the peer group, 2015 Government effectiveness is another area Cambodia for improvement especially compared to Bangladesh high growth countries. This indicator captures Kenya perceptions of the quality of public services, Uganda Pakistan quality of the civil service and degree of its Tanzania independence from political pressures, quality Vietnam of policy formulation and implementation, and India credibility of the government’s commitment Burkina Faso Senegal to such policies. Kenya is grouped at the lower Ghana end, together with Bangladesh, Cambodia, and 0 20 40 60 80 100 120 140 160 Pakistan. However, and as in other countries, Source: Transparency International. Note: CPI = Corruption Perception Index there are nuances to this overall performance. For example, the Huduma Kenya program, which is coordinated by the Ministry for Devolution and Weak enforcement, red tape, and corruption Planning through the Huduma Kenya Secretariat, are some of the main culprits for the prevailing was the first place winner in Category 1 for informality and low growth and investment in the Improving Delivery of Public Services in the 2015 formal sector. Corruption and weak enforcement United Nations Public Service Awards (UNPSA). of regulations increase the cost of investment and In 2007, the UNPSA went to the Performance doing business. As chapter 2 illustrates, Kenya is Contracts Steering Committee Secretariat and among the most regulated economies when it the Kenya Open Data Initiative was one of three comes to doing business, which, combined with international finalists at the Open Data Institute’s the high incidence of bribery (one in four formal 2015 awards for its publisher award, celebrating firms faces at least one bribery request per year), high publishing standards and the use of reduces the return on investment. challenging data. There is hope that the new system arising Other governance indicators support the from the 2010 Constitution will overcome impression that Kenya has its priorities right some of the governance weaknesses, but legal when putting governance and public sector solutions on their own cannot achieve the modernization high on the reform agenda. full impact. The 2010 Constitution provides an Governance is one of the areas for improvement in opportunity to strengthen and build strong and the World Bank’s Country Policy and Institutional effective governance institutions that support Assessment (CPIA) index. Kenya’s overall CPIA development effectiveness and sustained was 3.8 in 2014, which is above the south Saharan growth. Devolution in particular brings about Republic of Kenya, Economic Survey 2014. Kenya National Bureau of Statistics, page 204. 37 FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 19 an opportunity for increased participation and and Uchmi grew larger, they started to demand stronger accountability. However, the political more efficient transport, as the poor transport culture and patronage-based politics are likely situation was hurting their profits and growth. to continue to impede investment and growth. Consequently, EAC governments undertook Vested interests remain strong and impunity action to improve quality, reduce transport times, continues to pose a challenge. Corrupt practices and strengthen regulatory compliance. are institutionalized and much remains to be done in the fight against Significant reforms have been undertaken and the outcomes are likely to register in the data Another driver of change could be the in the coming years if reform initiatives are private sector, which, as it grows, can put followed through. However, retrospectively, stronger demands on the state for efficiency there has been only modest change in the improvements. For example, road transport in indicators over time, as illustrated in figure 1.22. Kenya and the EAC region as a whole for years Regulatory quality has been on a slight downward had been characterized by poor compliance with trend in recent years, while rule of law and maybe regulations and slow transit time caused by low- government effectiveness has been on a slight quality roads, many roadblocks, and weighbridges. upward trend. Other indicators have been flat at As large retailers such as Nakumatt, Tuskys, a relatively low level. Figure 1.22: Development in governance indicators, 1995–2013 Voice and accountability Political stability and absence of violence/terrorism 100 100 50 50 0 0 1996 1998 2000 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Government e ectiveness Regulatory quality 100 100 50 50 0 0 Rule of law Control of corruption 100 100 50 50 0 0 Source: World Bank Governance Indicators, http://data.worldbank.org/data-catalog/worldwide-governance-indicators. Note: The inner, thicker blue line shows the selected country’s percentile rank on each of the six aggregate governance indicators. 20 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Stabilization Policies and External Kenya’s real exchange rate has been appreciating Conditions over the past few years, which is not a good signal for its exporters unless it is driven by Macroeconomic stability has long been productivity growth. In parallel to the inflation- considered a key precondition for long- targeting monetary policy setting, the nominal term growth. Some of the most commonly exchange rate has been flexible (more than other used indicators in the literature to measure currencies in the region) and without significant macroeconomic stability include inflation, the volatility. This partly explains the continued real exchange rate, and the terms of trade inflows of short-term foreign capital. In addition (TOT). Each of the three indicators can be to recorded inflows, Kenya’s economy is influenced by exogenous factors, but primarily attracting unreported foreign exchange inflows, all three are driven by the country’s monetary which in turn has increased the value of the and fiscal policy. shilling. The appreciation of the real effective exchange rate (REER),39 which has also been a Although Kenya has experienced episodes of high characteristic for Ghana, is opposite from the inflation, effective monetary policy has played experience of the high-growth economies in an important role in maintaining price stability. the 1980s, which recorded depreciation of the This outcome has been a result of exogenous REER. The depreciation of the REER enabled shocks and domestic policy. On the former, the those countries to keep a fast pace of export transmission of changes in global food prices has growth. Nevertheless, according to World Bank been the main cause of inflation in Kenya, and estimates, Kenya’s real exchange rate seems to there is limited action that can be taken to this be close to equilibrium. end. On the latter, inflation is typically high when monetary policy lacks credibility or when fiscal In addition to the appreciating REER, the terms policy puts pressure on domestic demand. Both of trade have been worsening. The TOT indicator factors seem to have influenced price changes in is complementary to the REER and depicts the Kenya over the past few years. Inflation began average price of imports (stated in domestic to accelerate in late 2010, primarily because currency) relative to the price of exports.40 Thus, of rapidly growing credit to the private sector unlike the REER, which involves tradable and non- that boosted domestic demand, but fiscal tradable goods and services, the TOT involves policy contributed to the boost in demand only traded goods and services. The TOT indicator and the monetary authorities contributed, as has been increasing in Kenya during recent they were hesitant to respond to the building years, which implies a worsening of Kenya’s TOT inflationary pressures. (lower demand and/or worsening of the price competitiveness of Kenya’s exports) (Table 1.1). 39 The weighted average of a country’s currency relative to an index or basket of other major currencies adjusted for the effects of inflation. 40 A negative value of the TOT capacity indicates greater capacity to export than to import and vice versa. An increased rate of growth when the TOT indicator is negative implies improvement of the TOT conditions of the country, whereas increase of the growth of the TOT indicator when it is positive indicates worsening of the country’s TOT and a possible deterioration of the current account balance. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 21 Table 1.1: Inflation, volatility of GDP growth, REER, and TOT movements in Kenya and peer economies Inflation (%) REER trend Terms of trade Uganda 9.9 Depreciating Worsening Tanzania 9.4 n/a Worsening Burkina Faso 3.1 n/a Improving Ghana 12.5 Appreciating   Senegal 2.3 n/a Improving Kenya 11.5 Appreciating Worsening Bangladesh 7.8 n/a Improving Cambodia 6.6 n/a Worsening Pakistan 11.3 Depreciating Improving Vietnam 10.8 n/a Worsening India 8.6 n/a Worsening Indonesia 9.6 n/a Improving Thailand 5.8 Depreciating Worsening China 7.8 Depreciating Improving Sources: World Bank World Development Indicators; respective central bank websites of the countries. Note: GDP = gross domestic product; n/a = not available; REER = real effective exchange rate; TOT = terms of trade. However, Kenya’s overall macroeconomic The Priorities: What Are the Binding management ranks better than that of its peers. Horizontal Barriers to Growth? Using the World Bank CPIA, Kenya has maintained its position in economic management, with an The benchmarking of Kenya against its peer average score 4.2 points, which is higher than SSA countries on the most relevant determinants of peers and the rest of the world (Figure 1.23). This growth shows mixed results. The positive news situation is the result of government measures is that progress has been recorded in recent years toward macroeconomic, fiscal, and debt policies. on many of the growth determinants in which Kenya performs below average. For a few of the Figure 1.23: Kenya’s economic management ranks above its peers determinants, the outcomes have been stagnant; Economic Management Cluster (CPIA) score, 2005 -2014 hence, a strong impetus of reform is needed to 4.7 bridge the gap with the peer countries. On a 4.5 Kenya positive note, Kenya is quite advanced, relative to the same group, in several areas that are 4.3 important for growth and shared prosperity. 4.1 SSA peers Kenya ranks below its peers on governance and 3.9 Rest of the World has had a flat profile of performance in the period 3.7 observed. However, the significant reforms that 3.5 have been initiated, if they are sustained, could 3.3 result in improvements. 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Source: World Bank World Development Indicators. Note: CPIA = Country Policy and Institutional Assessment; SSA = Sub-Saharan Africa. 22 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Not all growth determinants are of equal None of these growth determinants on its own importance, and their roles are not the same at is likely to alter a country’s growth trajectory. all levels of development. Improving on some A good example is the following: in 2013, the of the discussed growth determinants could average years of schooling of Kenya’s labor force unleash growth in the short term, but to achieve was 6.75, and GDP per capita was around US$730 the desired sustained growth, improvements (in 2005 U.S. dollars). The United Kingdom’s labor across the board will be required. For example, force had the same quantity of education in reforms that stimulate investment may bring 1964 and its GDP per capita was 25 times higher, immediate benefits in growth acceleration, but and France reached that level of education may only define progress up to a certain point. in 1985 when its GDP per capita was 40 times After all, investment (as a share of GDP) can that of Kenya.41 Growth occurs when various only go so far and sustained growth requires aspects of the economic environment stimulate technological advances and innovation that the creation of productive knowledge, that is, raise productivity. At the same time, productivity the knowledge and ability to produce different jumps typically come with a lag, so filling some of types of goods and services. This process is the performance gaps will have a medium-term complex and gradual, so emphasizing one rather than immediate impact. growth determinant will not necessarily deliver equivalent outcomes if other aspects remain a More importantly, the above list of growth bottleneck to knowledge creation. determinants is not all encompassing. On the contrary, in addition to making progress across The remainder of this report examines several the horizontal areas that have been discussed, of the growth determinants. Chapter 2 focuses growth can occur on the mezzo (sector) or micro on jobs and touches on access to finance, (firm) scale by developing the capability to governance, and urbanization. Chapter 2 also produce a particular good or service. The Atlas of assesses the link between growth and poverty, Economic Complexity, by Hausmann et al. (2011), as poverty reduction is a complementary goal illustrates that a country’s economic complexity, in the government’s strategy. The issue of which reflects the knowledge embedded in its how to increase investment is discussed in productive structure, drives income per capita Chapter 3, which looks at the role of saving growth. Changes in economic complexity as a driver of investment. Chapter 4 looks occur at the sector and firm levels, whereby into the performance and growth potential of the potential for improvement is defined by manufacturing, but also services, as Kenya has the country’s starting point, that is, its current proven to be more successful in unleashing the capabilities and complexity. potential of several service industries. Last but not least, chapter 5 looks a bit further into the future and explores how the discovery of oil changes Kenya’s growth potential. Barro and Lee (2001). 41 FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 23 What Does Devolution Mean for Growth? Devolution is the centerpiece of Kenya’s governments, of which K Sh 230.65 2010 Constitution, involving large-scale billion was expected to be transferred political, fiscal, and administrative to them. The National Assembly and decentralization, with fiscal equalization Senate have recently agreed that as a major objective. Underpinning the counties will be paid a total of K Sh 308 devolution agenda was the need to: (i) billion in 2015/16. In the medium term, address deeply entrenched disparities counties’ equitable share allocation of in development between regions; (ii) nationally-raised revenue is expected to improve equity in access to social and remain stable at about 4 percent of GDP. economic services at the county level; and, (iii) work progressively toward Experiences from other countries SPOTLIGHT 1 equalizing opportunities for all Kenyans. illustrate that fiscal decentralization can catalyze economic growth but there Significant service delivery functions are also downside risks. The benefits of have been devolved from the central decentralized government include the government to the counties. Under following: (i) public policies tailored to the 2010 Constitution, counties are local needs through closer proximity responsible for policy implementation to the people; (ii) better governance and service delivery in primary and and accountability structures, since secondary health care, water supply, they are closer to the people; (iii) more rural electrification, urban service cost-effective approaches to delivery delivery, trade licensing, transport of services, through peer competition; (county roads), and agriculture. In the and, (iv) where there is subnational tax first year of devolution (fiscal year autonomy, increased accountability with 2013/14), the equivalent of 3.9 percent a positive relationship to growth. At the of GDP was transferred to county same time, antagonists of devolution governments, against an original budget argue that devolution can undermine of 4.3 percent of GDP. The shortfall growth potentially through: (i) increased was comprised of donor financed bureaucratic burden; (ii) separation of conditional grants and the allocation spending and taxing responsibilities, to the equalization fund, both of which which can undermine efficiency and were budgeted as transfers, but not lead to arrears; and (iii) newly created actually paid. For 2014/15, a total of K Sh subnational governments that may face 247.21 billion was budgeted to county capacity constraints. 24 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M The effects of devolution on economic County spending in 2013/14 reflected growth in Kenya will manifest a worrisome trend on expenditure through multiple channels. The first priorities; the 2014/15 county budget is the macroeconomic effect of public allocation reflects a clear shift in sector spending, including how county spending commitments (Figure 1.24). resources are shared between recurrent Counties’ general public services expenditure and investment. The public accounted for 84 percent of the total financial management law (2012) county spending in 2013/14, a share requires governments (national and that reduced significantly to 36 percent county) to spend at least 30 percent of of the total county budget in 2014/15. their revenue on investments. Second, Counties’ allocation to the health sector counties will define the quality and scope (which is a fully devolved function) of growth-enhancing services, such as increased to 19 percent of the total county health care or urban service delivery. budget in 2014/15. Expenditure allocation Finally, the responsibility of improving to counties’ economic affairs sectors, governance is being shared between which include agriculture, transport, the central and county authorities. This and other economic affairs subsectors, SPOTLIGHT 1 spotlight focuses on the first channel. accounted for 26 percent of the total county budget in 2014/15, reflecting a County governments have taken over shift toward productive spending. the delivery of devolved services, starting with an expenditure layout of Figure 1.24: There is a clear shift in counties’ sector spending priorities 5.4 percent of GDP or 20 percent of total expenditure in 2013/14. At the outset, Counties sector spending, percent of county budget the intention was to increase productive General public services 35.9 83.8 Economic affairs 6.9 spending through devolution. The Health 5.3 25.7 19.4 2013/14 fiscal data reveal important Education 0.8 8.3 emerging trends in county expenditure: Housing and community ammenities 2.2 6.0 Environmental protection (i) overall expenditure execution is low Recreation, culture and religion 3.0 1.2 (the overall budget execution rate was Social protection 0.5 63 percent of approved expenditure); 0 20 40 60 80 100 (ii) administrative expenditures have Percent 2013/14 2014/15* built up rather quickly (78 percent of Source: Kenya National Bureau of Statistics 2015. total spending was on recurrent costs); (iii) underspending was concentrated on the development budget, where The majority of counties face revenue only a handful of counties allocated and expenditure gaps. On the revenue at least one-third of their budget side, many counties budgeted for for development projects; and (iv) “hidden deficits” through inflated the counties experienced an overall estimates of own source revenues; but revenue gap, mostly on own revenues, several have been unable to match the with a collection gap of 57 percent. revenues of defunct local authorities. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 25 More importantly, from the prism of in 2014/15 (33 percent shortfall) accelerating growth, counties were (Table 1.2). Pressure to inflate revenue not able to execute their expenditure estimates appears to be coming from two plans fully, and development spending interlocking factors: the requirement suffered in particular. Only a third of to budget 30 percent for development the budgeted 2.1 percent of GDP on spending—which is beyond the fiscal development spending was executed in capacity of counties that inherited large 2013/14 but this figure improved in the wage bills—and the political pressure second year of devolution to 2/3 of the on county governments to appease budgeted amount. The reasons for low multiple interests in the budget process, execution included: increasing cost of to get the county budget passed. If the administration and wages; significantly causes behind these trends are not undershooting local revenue targets; addressed promptly, fiscal policy will and late receipt of one-sixth of the become a drag on economic growth. equitable share, which arrived too late to be spent in both fiscal years. County Data on revenue collection and budget governments ended 2014/15 with a execution performance for 2014/15 SPOTLIGHT 1 surplus of K Sh 17.9 billion compared to are more encouraging. By the end of K Sh 54.8 billion in 2013/14. 2014/15, the counties had spent a total of K Sh 90 billion on development, Local revenue collection is improving compared with only K Sh 36.6 billion for and revenue forecasts are also becoming the same period the year before. The more realistic in the second year of county government budget execution devolution. Actual amount collected rate for the year was 79.1 percent, was K Sh 23.6 billion in 2013/14 (66 which was not far behind the national percent shortfall) and K Sh 33.9 billion budget execution rate of 84.6 percent. Table 1.2: Local revenue collection has improved 2013/14 2014/15 2015/16 Revised Actual Revised Actual Budgeted Forecast Budget realized Budget realized revenue increase in (K Sh (K Sh (K Sh (K Sh (K Sh revenue millions) millions) millions) millions) millions) (K Sh millions) Nairobi 15,448 10,026 13,323 11,500 17,528 4,205 Mombasa 7,345 1,716 5,122 2,493 5,182 60 Kisumu 3,417 622 1,500 971 1,869 369 Nakuru 2,555 1,817 2,756 2,200 29,112 26,356 Uasin Gishu 1,754 564 890 801 1,037 147 Machakos 2,542 1,175 2,850 1,357 2,372 -478 Kakamega 3,500 325 904 5,169 1,000 96 Nyeri 479 432 1,344 681 1,488 144 Source: Office of the Controller of the Budget. 26 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M A particular expenditure challenge To address the fiscal burdens arising affecting many counties is the high from inherited staff, the Commission and/or growing wage bill. Personnel for Revenue Allocation has proposed costs accounted for almost half of the a revised revenue-sharing formula counties’ budgets in 2013/14 (Figure that will take into consideration the 1.25). For 16 counties, the proportion wage bill in allocating resources; but exceeded 50 percent; in Taita Taveta the recommendation has not yet been County, the figure was 73 percent. For accepted by the Parliament, which is some of the counties, in particular the authorized to decide the formula. In the urban ones, this is a legacy issue: bloated long term, rationalization of county staff workforces were inherited from defunct will greatly help ease this problem. At local administrations. Poor design of present, county governments lack the the process for transferring staff from legal power to remove staff who were central-level to county-level institutions transferred to counties as national public also contributed to overstaffing, which servants. A rationalization program in turn has inflated the wage bill. Last is being conducted by the national but not least, counties introduced government. Counties need clear SPOTLIGHT 1 additional allowances for county staff, guidance on their authority to reduce which has further increased wage costs. staff numbers, and a clear process and County wage bills still grew in 2014/15. technical support for doing so. Counties spent K Sh 77.4 billion on wages in 2013/14. At the end of 2014/15, they Another risk to growth, which stems from had spent K Sh 103 billion. These figures the challenges that have been described, reflect the reality that counties did not is that some counties are trying to bridge necessarily inherit the skills they need the fiscal gap through uncontrolled to carry out their new mandates. introduction of new fees and charges. Having limited opportunity, or will, to Figure 1.25: The high inherited wage bill is adjust expenditure, many counties are crowding out development spending in most counties introducing, or increasing, county-level fees and charges. These charges include Total county expenditure, 2013/14 (Ksh billion) 100 parking fees, business permits, health 80 inspection and transport licenses, rents, 60 and payments for billboard adverts. 40 In many cases, the new charges were 20 45.8 significantly higher than previous local 30.5 0 21.5 2.2 authority levels, a situation that has Personnel O&M Development Debt repayment and pending bills generated concern over the potential Percent of total expenditure impact on local-level business costs, especially for small business operators Source: Office of the Controller of the Budget. (Figure 1.26). In general, such drastic increases in local taxes could also have FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 27 Figure 1.26: Kisumu County’s new fees and charges have no coherent basis and are above all previous levels Selected old versus new fees and charges by Kisumu (K Sh. average Old New 30,000 25,000 20,000 15,000 10,000 5,000 - Annual public Annual public Monthly rent for Inspecti on for Approval of Approval of Helath inspecti on Applicati on for Occupati onal health licence health licence (non- shops located registrati on structural plans by commercial fee for export medical services certifi cate from (chemicals, food food category) within residenti al (hopitals, schools town engineer building plans by permit certifi cate certifi cate to export town engineer for and general estates and other (commercial blocks) town enginner foodstuff comercial property hygiene category) businesses) (other zones; new (per complete unit) charge) Source: Kisumu County Finance Act (2013). detrimental effects on county revenues cadaster, which is likely to meet political in the medium term, particularly if they resistance. In Kiambu, Mombasa, drive away business and investment. Nairobi, and Nakuru, past attempts at SPOTLIGHT 1 Some counties are also charging taxes for updating the valuation rolls have been goods that are transported through their stalled by political interference. jurisdiction, in effect applying domestic trade taxes. These charges have dubious Last but not least, subnational fiscal legal validity, since counties have been borrowing, if not well managed, is likely empowered to impose only two taxes, to become another source of concern for property rates and entertainment macroeconomic stability and growth. As tax. The domestic trade taxes are also counties are pressured to deliver on public economically inefficient and prejudice services, they may embark on borrowing producers whose goods have the domestically and internationally to spend furthest to travel to market. more. At present, there is a restriction on county borrowing, during the three- At the same time, county governments year transition period. The restriction are not fully exploiting the main revenue is to allow more time for finalization stream available to them—property of the regulatory framework, which rates. Former local authorities raised is currently being formulated. In any about 20–25 percent of their revenue case, the Constitution disallows county from property rates, although valuation borrowing without sovereign guarantee. rolls for some of the largest urban centers Nevertheless, based on the experience are woefully out of date. For example, of countries such as Brazil, India, Mexico, Nairobi’s property roll is estimated to and South Africa, the main concern in cover less than one-quarter of the total Kenya is that with inadequate monitoring properties, and values are now almost of debt issuance and weak enforcement 35 years out of date. Realizing the full of borrowing regulations, counties potential of property rates will require could generate unsustainable levels of complete reconstruction of the fiscal contingent liability for the central budget. 28 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Therefore, specific elements of the In addition to these issues, inadequate subnational borrowing framework coordination between national- and will need further attention. The county-level actors is hampering definition of debt is unclear. Decisions successful implementation of public are needed on whether debt includes service provision. Inadequate contingent liabilities, including multi- coordination has led to inconsistencies year obligations and public-private and, in some cases, duplication and partnerships, which many counties conflict. For instance, a pre-devolution are entering into. The establishment arrangement to transfer functions of borrowing limits needs more work. and funding in phases, based on the Proposed debt stock limits of 20 percent level of preparedness of each county, of recurrent revenues are out of step with was abandoned shortly after the international norms (usually around 200 2013 elections. Many complexities percent). The current debt of Nairobi, experienced during the first year of estimated at K Sh 42 billion, is more than devolution, including disruption of 10 times the stock limit of around K Sh crucial services, could be attributed 4 billion. The process for issuing debt to this approach. In addition, some SPOTLIGHT 1 does not yet include a comprehensive functions, notably responsibility for assessment of county creditworthiness. secondary roads and drug distribution, Finally, a comprehensive framework are yet to be clearly assigned to either of ex post rules should include triggers level of government, a situation that for insolvency, options for intervention continues to cause duplication, including and financial restructuring, and a clear in public spending. More broadly, pathway for counties to exit from devolution has remodeled power-cum- interventions. A further area of potential public resource relations between key concern is the proposal by some counties institutions (and vis-à-vis civil to form a regional bank. Although the objective of financing economic society) in ways that may not have development in neglected areas is been fully anticipated. To minimize an understandable one, if imprudent conflicts and achieve a more successful investments are underwritten by county implementation of devolution, it will governments, they could jeopardize be important to get intergovernmental county fiscal solvency. This is the relations to work effectively, while sequence of events that generated the also expanding the political space for subnational debt crises in Brazil two engagement, including with citizens. decades ago. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 29 30 RAN K E N YA C O U N T R Y E C O N O M I C M E M O Photo: DUM Gerando Pesantez/World Bank CHAPTER 2 FROM ECONOMIC GROWTH TO JOBS AND SHARED PROSPERITY Introduction Moving forward, the focus should be on generating job opportunities, as the poor rely The ultimate objective of the Kenyan mostly on labor income. As the poor in Kenya government and its people is to achieve not depend primarily on labor income, the key is to only growth, but also shared prosperity. Vision provide them with job opportunities. However, 2030 and the Second Medium-Term Plan (MTP- job creation has yet to catch up with demographic 2) set targets for gross domestic product (GDP) trends, as half of the increase in output is being growth, but at the same time aim for poverty generated by sectors with low labor elasticity, reduction and job creation. The two goals are such as finance or communications. Creating interdependent, but higher growth does not formal jobs has been a major struggle; only always imply lower poverty or more jobs. The 75,000 formal jobs per year are being created. The African continent presents some of starkest remaining 90 percent of labor market entrants examples of this inconvenient truth. Equatorial end up being part of the informal economy, which Guinea is the richest country on the continent, is characterized by low productivity, that is, low with GDP per capita above US$20,000—almost earnings, and stunted growth potential. 20 times higher than Kenya’s—and its economy grew at double digits over the past decade, yet The labor market entrants of 2030 have already more than 60 percent of its population lives on been born, and a large share of them will most less than US$1.25 (in purchasing power parity) likely be starting their employment in the per day.42 For Kenya, inclusive growth is to be informal sector. Despite having formalization as spread across sectors: in services, which have a priority, the informal sector will remain part of high poverty elasticity, but also in agriculture, Kenya’s reality even as the country moves toward where most of Kenya’s poor are. being classified as an upper-middle-income country. Going forward, focus should be placed Kenya’s economic model has not been on improving the productivity of the jua kali.43 particularly inclusive; hence, poverty remains The reasons for the low productivity and growth high. New estimates—in the absence of actual potential are multiple. Jua kali entrepreneurs face poverty data since 2006—point to lower difficulties with access to finance and access to poverty reduction since 2006 than previous utilities (including land), among others, which in studies have found. This finding is partly turn stunts their growth. This chapter concludes explained by below average and volatile that public policies should center on enabling growth in agriculture, and also a result of the jua kali to prosper in addition to removing above average price increases for food and constraints for formal businesses. The chapter transportation, which represent a significant also provides examples from other countries share of the poor’s consumption basket. that have successfully promoted the informal economy (box 2.1). 42 http://www.africaneconomicoutlook.org/en/countries/central-africa/equatorial-guinea/. 43 A commonly used term for Kenya’s informal sector. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 31 Box 2.1: From growth to shared prosperity—The different paths of Rwanda and Nigeria Poverty rates have been falling across Africa since the beginning of the 2000s. As gross domestic product (GDP) growth accelerated to around 5 percent per year, the percentage of people living on less than US$1.25 per day in Sub-Saharan Africa declined at an average rate of 0.8 percent per year. However, beneath the regional averages, there is considerable variation in the elasticity of poverty to growth. Rwanda and Nigeria illustrate this point. Rwanda is one of Africa’s economic success stories of the 2000s. GDP grew at an average of 8.2 percent per year between 2005 and 2010, and GDP per capita grew at 5.2 percent. This translated into growth of mean consumption per person among the rural poor of 2.1 percent per year, which brought poverty down from 62 to 48 percent.a Rwanda’s growth story was pro-poor in the absolute sense and the relative sense, as consumption growth of the poor rose faster than consumption of the non-poor (0.4 percent). The key to Rwanda’s success was increasing agricultural production, which doubled at the household level, as did the share of households selling surplus harvests on the market. Behind this success was a cohesive agricultural strategy focused on increasing investments in agricultural inputs, land consolidation, and infrastructure. If Kenya had the consumption trends in rural Rwanda, Kenya’s rural poverty rate would have declined to 38 percent. Like in Rwanda, Nigeria’s growth has been robust, but poverty reduction much less so. Annual non-oil GDP growth rates have averaged 8 percent per year since 2003 and GDP per capita has grown at 3.5 percent per year. However, unlike Rwanda, the results from household surveys conducted in 2004 and 2010 suggest that overall poverty declined by only 2 percentage points (to 46 percent) and rural poverty went down from 57 to 53 percent. Although there are lingering questions about the quality of the consumption data for 2010, the evidence points to growing inequality in the country. A decomposition of the change in poverty between 2004 and 2010 indicated that poverty would have been 4 percentage points lower had inequality not increased.b Nigeria’s experience also highlights how regional variations in economic performance affected the national trajectory of inequality. While states in the coastal regions and the federal capital enjoyed inclusive growth, poverty reduction in the rest of the country was set back because growth was accompanied by increasing inequality or stagnation. In Kenya, there are stark spatial divides, primarily between the high-potential agro- ecological zones in central, western, and coastal Kenya, and the arid and semi-arid pastoral regions of northern and southern Kenya. Uneven geographic development will likely work to accentuate national inequality. a. World Bank 2012a. b. World Bank 2013b What Do We Know about Kenya’s Poor? pro-poor, inclusive, or shared,44 the income of the bottom quintiles should increase. As the poor From Growth to Poverty Reduction rely primarily on labor income, and mostly in There is no guarantee that a growing economy rural areas, inclusive growth must capture those will reduce poverty. Where and how economic segments of the economy where the poor are growth is concentrated, which sectors lead, currently active, or generate new opportunities and whether institutions harness growth into for the poor in the nonfarm, urban sectors of improved public services all play a role in the the economy. extent of poverty reduction. For growth to be Given the magnitude of poverty in Kenya, for the purposes of this discussion references to “inclusive” or “shared” (referring to the 44 bottom 40 percent) growth connote reducing poverty. 32 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Understanding Poverty in Kenya disproportionately hurt the poor. Agriculture suffered from several shocks, including low Going back several decades, the story on poverty rainfall, extreme temperatures, and reduced alleviation in Kenya, as in many other African demand in key export markets (such as North countries, is a disappointing one. In 1981, the Africa and Europe) for cash crops.49 Consequently, poverty rate was estimated at 48 percent,45 and agricultural output shrank by 5 percent in 2008 a generation later it was only 1 percentage point and by another 2.3 percent in 2009; it then lower. This is not surprising, since real GDP per bounced back in 2010 (10 percent growth) and capita measured in constant 2005 U.S. dollars has been growing at 2–5 percent per year since. actually fell by 2.5 percent, from US$537 to US$525, between 1980 and 2005.46 In addition, inflation has been high and volatile, and market distortions in key food items (sugar In 2005, 16.7 million Kenyans (47 percent of the and maize) have contributed to rising prices. population) were unable to afford the cost of a Food prices spiked in 2011 in response to a rising basic needs bundle of food and nonfood goods global food crisis; food inflation increased from deemed necessary to avoid living in poverty.47 10 to 26 percent between 2009 and 2011. In This bundle of goods was valued at K Sh 1,562 addition, various policy measures are in place per month per person for those living in rural that raise the prices of maize and sugar, which areas and K Sh 2,912 for the urban population. are key consumption items for poor households. In 2005, 47 percent of the population lived under Because of high import tariffs, nontariff barriers, the international poverty line of US$1.25 per state intervention, and anticompetitive conduct day (K Sh 1,246 per person per month).48 At that by firms, the wholesale price of sugar in Kenya time, 85 percent of poor households lived in rural is almost three times the world price, and the areas, were headed by members with less than price of maize is 20 percent higher (Box 2.2). five years of formal education (33 percent never Transportation inflation, another expense attended school), and worked primarily in family category that disproportionately affects the poor, farming (41 percent). Moreover, 95 percent lived doubled between 2009 and 2011. without electricity in the home, 49 percent lived without a decent source of drinking water in the Moreover, economic growth has been mostly home, and 26 percent lived without some kind of consumption driven, which is disproportionately waste infrastructure in the home. good for the poor in times of accelerating growth and bad in times of economic slowdown. As Economic performance since 2005 has been noted in chapter 1, Kenya’s economy has been solid, but not spectacular, as chapter 1 describes. riding on a consumption-driven growth model Seen through the prism of poverty, the good since 2005 (Figure 2.1). In years when growth news is that GDP growth has been positive and was high or accelerating, private consumption sustained. However, as output growth slowed, per capita—the average amount a person average income per capita fell between 2007 consumes in a year—grew faster than GDP per and 2009. Then the economy, and especially the capita. However, in years of economic slowdown, agriculture sector, exhibited high volatility that such as 2008 and 2011, private consumption 45 Rural Household Budget Survey, 1981/82, obtained from Gamba, Mghenyi. Rural Poverty Dynamics, Agricultural Productivity and Access to Resources, Tegemeo. 46 Data from World Bank World Development Indicators 2013. 47 Kenya Integrated Household Budget Survey 2005/06. 48 World Bank 2009. 49 World Bank 2011. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 33 Box 2.2: Despite challenges, there are still opportunities in the agriculture sector Agriculture is the main source of employment in rural Kenya and has significant potential to reduce poverty. Kenya has seen several successes in agriculture, including in tea, fertilizer, and horticulture. Overall, the sector continues to have a large potential for growth and contribution to poverty reduction and shared prosperity. The agriculture sector challenges are well Figure B2.2.1: Wholesale prices of maize in Nairobi, represented by the maize market, where Kampala, and international commodity markets competition is limited and consumer prices are high, with significant negative impact on welfare and poverty (Figure B2.2.1). State intervention, tariff barriers, nontariff trade barriers, and anticompetitive conduct by firms are among the explanations behind the high prices (Argent and Begazo 2015). The mechanisms in place to maintain high prices reflect a complex political economy. Maize is produced by smallholders for private consumption in large Source: Argent and Begazo 2015 parts of Kenya, but a considerable proportion of maize traded for consumption is produced by large Kalenjin farmers in the Northern Rift Valley. Using household surveys, the Tegemeo Institute shows that although maize production is widespread, 70 percent of maize farmers produce for their own consumption only. The same surveys show that 50 percent of revenues from traded maize was earned by 2 percent of the farmers. The government, through the National Cereals and Produce Board, is engaged in setting the price for maize through price announcements in the immediate post-harvest period each year and through maize purchasing schemes. The National Cereals and Produce Board has largely been managed by officials with ties to the Northern Rift Valley. Figure 2.1: GDP growth has been driven by consumption per capita fell by more than GDP per capita. The 8 increased investment model that chapters 1 and 3 argue for would of course have the opposite 6 effect in the short term, although in the medium 4 term more investment would raise the economy’s 2 potential and its growth rate, which in turn would Percent 0 bring higher consumption. 2006 2007 2008 2009 2010 2011 2012 2013 2014 -2 The volatility of the agricultural sector is perhaps -4 the most influential dimension of Kenya’s growth -6 in terms of its effect on poverty. Although the GDP per capita growth Private consuption per capita growth net impact of food prices on poverty depends Source: World Bank. on whether the poor are net buyers or sellers of grain,50 without corresponding increases in wages 50 Food makes up 70 percent of the poor’s monthly consumption budget. Analyses from Tegemeo suggest that the poorest deciles in rural settings purchase food for more than three months of the year, compared with only 1 month for the top decile. 34 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Box 2.3: Why agriculture will continue to matter for Kenya’s growth and poverty reduction Agriculture continues to be a key pillar of Kenya’s economy. It accounts for about a quarter of gross domestic product (GDP) and two-thirds of exports. And agriculture accounts for more than 70 percent of employment in rural areas. Moreover, adding the food industry and indirect (spillover) effects, the total contribution of the agriculture sector goes up to half of total output. The sector’s performance over the past decade can be characterized as weak and erratic at best. Apart from a few exceptional segments, productivity and output growth in agriculture have been quite weak. The causes of the underperformance are well known. The World Bank Agriculture Sector Review (forthcoming) finds the following to be particularly high ranking: (i) institutional and regulatory weaknesses (legal framework that supports elite capture), (ii) land fragmentation and administration (average farm size of just above one hectare and significant share of unregistered land), (iii) lack of downstream facilities (such as drying and storage), and (iv) poor rural infrastructure (less than 2 percent of agricultural land is irrigated). From a macro perspective, sorting out the challenges in the sector would bring positive spillover effects. First, increases in agricultural output and productivity would directly raise exports and GDP, as well as the incomes of the bottom 40 percent of the population. Second, because of its excessive reliance on climate conditions, the sector has been one of the main contributors to economic volatility over the past five years. Hence, productivity measures that stabilize food security, such as greater stability in grain policy (tariffs and quotas) or subsidies for seeds or fertilizer, would also reduce volatility in the nonfarm sector. Third, improvements in agriculture would facilitate a transition of labor to the more productive nonfarm sectors, following a decade in which the share of urban population has been stable. Such urbanization would in turn reduce rural population density and land scarcity, which in turn would abate the political risks that may stem from land scarcity. Finally, raising the incomes of the rural population would make a significant dent in poverty, as most of the poor live in rural areas. and incomes, overall increases in the cost of work.51 Research based on a panel survey data living threaten to pull households consuming set fielded in maize-growing areas suggests that close to the poverty line into poverty (Box 2.3). the within-sample poverty rate declined from However, strong growth in private consumption 42.3 percent in 2000 to 37.6 percent in 2007.52 (private consumption per capita has increased by The households that escaped poverty were more 2.2 percent per year since 2006) and growth in likely to have better educated members, more services suggest that there are increased earnings land under cultivation, and more non-land assets and employment opportunities for poor families (that is, more diversified income). These findings that are diversifying into nonfarm livelihoods or imply that diversifying income beyond farming seeking opportunities in cities through migration. is an effective poverty reduction strategy, and education helps rural Kenyans to obtain skills to perform wage work or become self-employed Poverty in Rural Kenya (Box 2.4). Since most of the rural poor live Rural poverty has been on a decline, primarily relatively close to the largest urban centers, as a result of rural workers doing nonfarm promoting internal mobility—through better 51 A panel survey was conducted by Tegemeo Institute (Egerton University) in 1997, 2000, 2004, and 2007. Data were collected from 1,275 households across eight agro-regional zones in Kenya (excluding pastoral areas). 52 Suri et al. (2008). FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 35 Box 2.4: The changing context of Kenya’s rural labor market The broader context of jobs in Kenya is one in which an increasing number of people are moving away Figure B2.4.1: Wage work and self-employment are increasing rapidly from small-scale farming. Figure B2.4.1 provides (population, millions, by primary source of income) estimates from census data of the population of 40 people primarily dependent on family farming, 35 nonfarm self-employment, and wage work. A 30 person is classified as being “primarily dependent” 25 on one of these job classes if that person resides 20 in a household where more than 50 percent of the 15 working-age members work in that job class. The 10 starkest changes are the reduction in the share of 5 the population dependent on family farming, the 0 increase in the share dependent on nonfarm self- 1989 1999 2009 employment, and the stagnation of the share of Family farming Non-farm self employment Wage work Other employment in the wage sector. Source: Kenya National Bureau of Statistics Population Censuses. Within households, working-age members are increasingly incorporating nonfarm self-employment into the mix of income sources. Income data from the Tegemeo panel survey also suggest that rural households have increasingly been diversifying away from crop income. In 2000, 50 percent of household income was derived from crops and 16 percent from nonfarm business; by 2007, crop income comprised 38 percent of household income and nonfarm business comprised 21 percent.a The census data also suggest that the bottleneck of low job creation in the wage sector may in part explain the increase in nonfarm self-employment. In a setting of stagnant wage job growth and increasing demand for nonfarm jobs, one short-run solution for individuals with some schooling is to try their hand at starting an informal business. In the foreseeable future, it is inevitable that informal self-employment will continue to grow, given the rapid growth of the workforce and increasing access to education in a setting of low wage job growth. a. Suri et al. 2008. transport links, public goods, access to credit, in (in)equality, which is something that can be and land tenure—holds promise for reducing measured only through a household survey. In rural poverty. the absence of such, four scenarios are presented in the report (Figure 2.2). Under the first scenario, Overall, Poverty Has Been Declining which assumes that consumption per capita grew at the same rate as the GDP per capita growth rate Poverty in Kenya has been on a steady but slow and (in)equality did not change, the poverty rate decline. In the absence of official measurement of fell to less than 39 percent in 2013. If inequality, poverty since 2006, this report estimates poverty trends up to 2013 building on earlier estimates measured through the Gini coefficient, fell by in the 2013 Kenya Economic Update (World 1 percent per year between 2005 and 2013, Bank 2013c) and using the revised GDP data. poverty fell to 35 percent, while a 1 percent The magnitude of poverty reduction depends per year increase in inequality brings poverty on the distribution of income, that is, change to about 42 percent by 2013. The last scenario 36 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M uses sector growth rates and population shares The Way out of Poverty to estimate household consumption. According The past economic performance, if maintained to that scenario, poverty fell to about 40 in the future, is unlikely to make a big dent in percent by 2013. poverty. Between 2006 and 2013, average annual Figure 2.2: Poverty has been on the decline in Kenya output growth in agriculture, industry, and services was 2.5, 3.9, and 5.9 percent per year, respectively. If each sector continues to expand at the same pace, and assuming the employment transition out of agriculture to services continues at its historic pace, by 2020 the poverty headcount will be in the vicinity of 35 percent. One intervention that can reduce poverty further, if targeted well, is the government’s expansion of social protection programs, primarily cash transfer programs for the poor. Enrolling and sustaining the almost one million households in Source: World Bank estimates. Note: The graph shows estimated trends in the poverty rate under different cash transfer programme with transfers equal models between 2005 and 2013. to cover their average poverty gap, and with a 75 percent targeting accuracy, would reduce Another way to understand poverty trends is by poverty by an additional 16 percentage points.54 asking people how they feel about it. In a 2011 survey among adults across the country, when Moving forward, the path of poverty reduction asked to rate the government’s performance will be mostly defined by what happens in rural on a range of matters, more than 80 percent Kenya. The majority of Kenya’s poor live in rural responded “very bad” or “fairly bad” in the areas: 90 percent of Kenyans in the bottom 40 areas of reducing income gaps, improving the percent of the income distribution live in rural living standards of the poor, and creating jobs. areas (Figure 2.3). Hence, to make a big dent in In addition, despite growth, there has been a poverty, the incomes of rural poor would have growing tide of discontent. Between 2003 and to rise. As the poor in Kenya depend primarily 2011, the proportion of Kenyans describing their on labor income, the key is to provide them living conditions as very bad or fairly bad doubled, with job opportunities, which is the focus of the from 36 to 72 percent.53 remainder of the chapter. 53 Afrobarometer 2011. 54 Kenya has four cash transfer programs implemented under the Ministry of Labor and Social Services: cash transfers for vulnerable children, older persons, persons with severe disability, and poor urban households, and one program under the Ministry of Devolution and Planning: the Hunger Safety Net Program, which operates in four arid counties in northern Kenya. The programs give K Sh 4,000 to enrolled households on a bi-monthly basis. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 37 Figure 2.3: The majority of Kenya’s working poor live in rural areas Source: Kenya Integrated Household Budget Survey 2005/06. If the economy is able to shift growth into higher to 33 percent. Rather than providing credible gear, as laid out in the MTP-2, with agriculture, predictions of the trajectory of poverty, these services, and industry growing 6.8, 9.4, and scenarios suggest that raising the productivity 8.6 percent per year, respectively, by 2020 of agriculture carries more potential than other poverty would drop to 20 percent. Although in sectors for poverty reduction, given the current the long run services hold the largest potential patterns of population change in each sector. for economic development, in the medium term improvements in agriculture are key to Although future outcomes in the sector are not reducing poverty. If agricultural growth were certain, there is no doubt that productivity must boosted to 4.6 percent (from an average of 2.3 increase to achieve a substantial acceleration in percent) while services and industry continued at agricultural output growth. Kenya has suffered their historic averages, poverty would fall to 27 through a long period of stagnant yields in percent by 2020 (Figure 2.4). Boosting the rate of agriculture. For example, cereal yields have services growth to 10.6 percent while holding the remained practically unchanged for more than a other sectors constant would see poverty decline decade, while other countries have achieved rapid to 30 percent. And boosting industry’s rate of productivity growth. In Rwanda, for example, growth to 9 percent would reduce poverty only crop yields have been rising 8 percent per year for more than a decade and have surpassed Kenya’s. Figure 2.4: Estimated trends in the poverty rate under different scenarios between 2005 and 2020 Uganda and Vietnam have also seen crop yields grow at 2–3 percent annually since 2000. A small- scale experiment on maize farms in western Kenya illustrated that yields may double simply through applying improved crop management practices, and adding fertilizer would give a further boost. Poverty reduction outcomes would be greater by boosting agricultural productivity and liberalizing the sugar and maize markets. The removal of market restrictions in the sugar and Source: World Bank estimates. maize markets is expected to yield poverty 38 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M reduction gains: 20 percent reductions in sugar Job creation has not been able to keep up with and maize retail prices are estimated to reduce population growth. While the working-age poverty by 1.5 and 1.8 percent, respectively. population increased by three million between This could be achieved through lowering import 2009 and 2013, 2.6 million jobs—or 80 percent tariffs for sugar and maize, applying more of the working-age population—were created stringent competition policies, and reducing the in this period. The economy expanded by 26 role of the National Cereals and Produce Board percent in the five-year period, while total in determining prices (Argent and Begazo 2015). employment—formal and informal, excluding subsistence farming—rose by 24 percent, Last but not least, it is difficult to design effective which points to the contribution of productivity poverty reduction policy in the absence of increases to GDP growth. Micro data from listed precise poverty data. The historic estimates and companies at the Nairobi Stock Exchange confirm the forward-looking outlook on poverty that are this. Sales (of those companies that publish data presented here are based on assumptions about on employment) rose by 56 percent between what may have been happening with the income 2009 and 2013, while employment increased of the poor since 2006. However, this approach by only 22 percent. At the same time, almost 90 cannot replace having comprehensive data on percent of the jobs were created by the informal the income trends and characteristics of the poor. economy, where four in five Kenyan workers are Thus, for effective poverty reduction policies, it is employed (Figure 2.5). In contrast, only 75,000 absolutely critical to conduct household budget new formal wage jobs were added each year. surveys at regular time intervals. To this end, Figure 2.5: Formal employment, although desired by the Kenya National Bureau of Statistics (KNBS) is many, remains a privilege for a few (Kenya’s demographic trends, millions) expected to complete a new household budget 50 survey (Kenya Integrated Household Budget 45 40 Survey-II) in 2016. 35 30 Has Growth Created Jobs and Where? 25 20 15 Although good data on the labor market are 10 not available—in the absence of labor market 5 surveys—-official statistics confirm that Kenya is - 2005 2006 2007 2008 2009 2010 2011 2012 2013 suffering from high rates of underemployment Informal employment Wage employment and youth unemployment. Kenya has not had a Othera Dependent population Source: Kenya National Bureau of Statistics Economic Surveys. labor force survey in the past decade, although a. Other includes small-scale farmers, pastoralists, as well as the unemployed/ inactive population. KNBS compiles annual (formal and informal) employment by sector. Estimation from the Although the informal economy cushions 2009 population census points to pervasive Kenya’s unemployment rates and is Kenya’s underemployment—hidden unemployment— largest employer, it does not contribute directly which is transposed into a high share of labor in to government budgets. Businesses in the subsistence agriculture (Kenya Economic Update informal sector are able to use public services Edition 7). Although this phenomenon is not and contribute to them through value-added unique to Kenya, this section looks at job creation taxes, but they face challenges accessing private and the constraints to employment. services, with access to finance being the most FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 39 important. Even if these businesses were to Figure 2.6: Jobs are created largely in informal trade and hospitality services formalize, it is unlikely this situation would change, Employment by Sectors because the businesses would be too small to 7,000 boost government budgets significantly, and 6,000 number of workers (in 1000's) they would still have difficulties accessing formal 5,000 4,000 finance, given high collateral requirements. 3,000 2009 2013 2,000 Most of the new labor market entrants find 1,000 jobs in trade and hospitality, predominantly as 0 Mining Education administration Trade, hotels, Communications Construction Other Agriculture Manufacturing Electricity, water, Transport Financial and food… services informal workers. In the informal sector, almost Public and gas two-thirds of new jobs between 2009 and 2013 were in trade and hospitality. Manufacturing contributed to 18 percent of job creation in the Source: Kenya National Bureau of Statistics. Note: Agricultural employment refers to formal employment only (informal and subsistence farming data do not exist). jua kali, while formal manufacturing jobs rose by 7 percent during the four-year period. The Productivity is increasing in the sectors with construction sector also recorded strong growth, formal employment. Modern services, such as with formal employment growing much faster, financial services and education, and industry although informal employment continues to (excluding manufacturing) have seen increasing account for two-thirds of the total. More than productivity. At the same time, the four sectors half of informal jobs are created in rural areas— with the highest productivity growth account which is not surprising, given that three-quarters for only 7 percent of total employment (Figure of Kenyans live in rural areas—however, it is 2.7), so their contribution to the overall labor interesting that job creation in urban areas did productivity of the economy is minimal. The not grow faster than in rural areas until 2013. This good news is that informal trade and hospitality, trend is behind the relatively slower urbanization which attract the majority of labor entrants, are in Kenya over the past decade compared with becoming more sophisticated: productivity in peers in Sub-Saharan Africa (SSA), but 2013 the sector grew by 6 percent between 2009 and seems to have marked a break in the trend, as 90 2013. Data on (informal) agricultural employment percent of the 630,000 additional informal jobs are not available, so the change in agricultural were in urban areas. productivity cannot be measured. Figure 2.7: Productivity growth is fastest in sectors with In contrast, the number of formal workers is few workers small and predominantly in the public sector. Output per Worker by Sectors (in Ksh) Kenya’s formal job market is tiny: only 2.3 million 4,000,000 Kenyans were formally employed in 2013, of 3,500,000 which about 700,000 were in the public sector. 3,000,000 2,500,000 Although data are not available, it seems that 2,000,000 formal jobs are relatively stable, as practically 1,500,000 2009 2013 every sector of the economy has been adding jobs 1,000,000 year after year since 2009 (Figure 2.6). The public 500,000 sector’s role in job creation has been limited: 0 administra… water, and… hotels, and… Mining Communicat Manufacturi Other Financial Education Transport Constructio Electricity, services Public Trade, ions 75,000 workers—most of them teachers—were ng n added in this period. Source: Calculations based on Kenya National Bureau of Statistics data. 40 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M To conclude, Kenya’s jobs story does not Table 2.1: Average annual growth rates to reach MTP-2 correspond with the strategic goals of formal jobs target accelerated and shared growth. The good 2012–13 2013–17 actual estimates news is the economy added more than 2.6 Agriculture 2.7 10.7 million nonfarm jobs between 2009 and 2013. Manufacturing 3.3 13.3 The bad news is the structure of job creation Construction 12.3 49.3 is unfavorable. First, workers are going into the Trade and hospitality 6.7 26.9 lowest productivity sector, trade and hospitality. Transport 1.6 6.4 Second, labor productivity is growing fastest Mining 4.4 4.4 in sectors that employ few workers. Third, Utilities 8.0 8.0 manufacturing—one of the priority sectors Other 5.5 10.9 for the government—is showing declining Communications 8.2 8.2 productivity. Had manufacturing recorded the Financial services 8.8 8.8 same productivity growth as trade and hospitality Education 4.2 4.2 (6.4 percent between 2009 and 2013), which, Public administration 5.1 5.1 similar to manufacturing, comprises largely Sources: Kenya National Bureau of Statistics; staff estimates. informal employment, GDP in 2013 would have Note: Green = high growth potential; yellow = moderate growth potential; red = historic growth trend. MTP-2 = Second Medium-Term Plan. been 1.6 percent higher. areas have been forced to create their own Furthermore, the ambitions of the MTP-2 are opportunities and employment. Consequently, much higher than the performance over the the number of (informal) jobs in wholesale and past five years. Job creation is targeted to double, retail trade, hotels, and restaurants has exploded. from 723,000 new jobs in 2013 to 1.4 million Other than the fact that the jua kali is a fallback jobs in 2017. The structure of the labor targets option for most new labor market entrants, little is even more ambitious: new formal wage jobs was known about the characteristics of this are to go up from 110,000 in 2013 to 570,000 sector and the constraints that the millions of in 2017. Under this scenario, the proportion of informal entrepreneurs are facing. An informality modern sector employment would increase from survey, conducted in 2013 and financed by the 12 percent in 2012 to 40 percent by the end of World Bank, offers insights on these issues that the plan period. However, to achieve the 570,000 are highly relevant for targeting public policies in new formal jobs target in 2017, job creation the sector. growth would need to quadruple compared with the 2012–13 growth rates in the sectors that are The age structure of the surveyed informal not constrained by resource or fiscal limitations business enterprises corroborates the story of (Table 2.1). the structural change that began to take place in the early 2000s. The data on the year of What Do We Know about the Jua Kali? establishment show a clear break in the number The jua kali has been part of Kenya’s economic of “births” between the 1990s and 2000s (Figure reality for more than three decades, but 2.8), which is likely to capture the change in the information on the sector has been scarce until economic environment after the end of the Moi now. In the absence of robust demand for wage regime (the rest of the difference being explained labor in the formal sector, those moving to urban by exit of unsuccessful firms). One of the key changes was the relaxation by commercial banks FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 41 and micro finance institutions of lending to the years. In addition to being small, informal firms general public. The spurt of lending generated have low productivity, which is illustrated in the lots of informal manufacturing business, mostly wages of their employees. Most informal firms jua kali activities to support the boom in the pay minimum wages or less, which confirms the informal economy. findings from the macro-data on output and employment that were discussed in the previous Figure 2.8: Informal firms are mostly young section. Three-quarters of the people employed by informal businesses earn between K Sh 1,000 dk 11 and K Sh 9,000 (US$12 to US$104) per month. 1970 1 1973 2 1981 1 What year did this business or activity originally start? 1982 1 1985 3 The minimum wage in Kenya is about K Sh 7,000 1987 2 1988 2 1989 2 1990 6 1991 1 1992 2 1993 1994 1995 1996 2 3 5 7 (US$80). Only 10 percent of firms pay their 1997 4 1998 1999 2000 2001 2 8 9 16 workers more than K Sh 10,000 per month. 2002 16 2003 26 2004 17 2005 25 2006 23 2007 31 Jua kali entrepreneurs are mostly young with 2008 42 2009 41 2010 67 2011 55 2012 75 some education. Almost three-quarters of the 2013 25 0 20 40 60 80 frequency surveyed owners of firms are younger than 40 years of age (Figure 2.10). Adding the fact that Source: Kenya Informality Survey 2013. Note: dk = don’t know. 20 percent of the firms are more than a decade old, with some in operation since the 1980s, Most of the informal businesses remain small, it becomes clear that Kenya’s youth has been with no more than one employee. Although behind the creation and growth of the jua kali. more than half of the surveyed firms were At the same time, the majority of the owners in established prior to 2010, three-quarters had the informal sector have some form of education, no more than one person employed, that is, the with three-quarters having undergone vocational owner (Figure 2.9). Moreover, three-quarters of training or secondary school. This fact reiterates the firms did not hire additional employees or the point that informal entrepreneurship has acquire machinery or space over the past three been a “must” for Kenya’s youth. Figure 2.9: Informal businesses employ few people and pay minimum wages or less 400 Average monthly salary for a worker 80 338 73.55 300 60 Frequency 200 Percent 40 118 100 20 40 13.51 10.32 12 17 4 2 1 1 2.627 0 0 1 2 3 4 5 6 8 10 12 15 0 0 Below 1000 1000-9000 10000-25000 don't know How many people who work in this business or activity were paid Source: Kenya Informality Survey 2013. Note: The number of employees includes the owner. 42 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Figure 2.10: Most owners of informal enterprises are young and literate b. Age a. Educa tion What is the highest level of education of the largest owner? Age of the largest owner 40 No Education 10 39.02% University training (complete 13 or not) 31.33% 30 dk 17 Percent 20 Primary School (complete or not) 147 18.01% Vocational Training 172 10 Secondary School (complete or not) 174 7.13% 2.81% 1.69% 0 0 50 100 150 200 18 and don't know 20-29 30-39 40-49 50-59 60-99 frequency Source: Kenya Informality Survey 2013. Note: dk = don’t know. Since informal firms are not growing and they informal firms, the other obstacles are also operate at low productivity, the main question quite severe. For example, corruption, although for policy makers is about the main obstacles not the largest, is a major obstacle for a third of that these firms face. Lack of access to finance the surveyed firms. And more than a quarter of is the main cause of stagnation for 43 percent of firm owners report crime to be a major obstacle. the informal businesses (Figure 2.11). Only 10 Seven percent of those surveyed stated they had percent of informal firms have received funding suffered a loss from crime in the past month. from banks or microfinance institutions. The rest have to rely on own resources, and few are able In addition to identifying the most important to get suppliers credit (Figure 2.12). Other key constraints to doing business, the informality obstacles include getting electricity (7.5 percent survey also reveals the reasons for operating of firms), corruption (6.8 percent), access to land informally. The two main reasons for staying (6.7 percent), and crime (5 percent). Although informal are registration procedures and taxes there is consensus on the largest obstacle for (Figure 2.13). The procedures for starting a formal Figure 2.11: What is the largest obstacle faced by informal Figure 2.12: Own funds constitute the main source of firms? finance Own Funds Constitute the Main Source of Finance Limited access to tecnology 3 Most used sources of finance Inadequately educated workforce (including the owner) 7 don't know 10 Don't know 52 Problems with the water supply 12 Own funds 372 Crime 24 Credit from suppliers 36 Limited access to land 36 Moneylenders 2 Corruption 36 Microfinance institutions 29 Problems with the electricity supply 40 Banks 17 not applicable 135 Friends or relatives 22 Limited access to finance or loans 230 Other 3 0 50 100 150 200 250 frequency 0 100 200 300 400 frequency Source: Kenya Informality Survey 2013. Source: Kenya Informality Survey 2013. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 43 Figure 2.13: Why do they choose to operate informally? Why aren’t they registered? a. Registration procedures b. Taxes Because of time, fees and papaer work required to register Because of taxes to be paid if registered dk 33 dk 39 No 219 No 213 Yes 52.7 % 281 Yes 52.7% 281 0 100 200 300 frequency 0 100 200 300 frequency Source: Kenya Informality Survey 2013. Note: dk = don’t know. business are indeed complex and onerous in in the informal sector has a ceiling, and moving Kenya. Formal registration of course brings many toward upper-middle-income status necessities benefits, such as better access to services and productivity increases that only the formal institutions (such as courts), lower burden from sector can achieve. These three constraints are inspections and other government officials, as particularly relevant for job creation. well as limited liability of the owner vis-à-vis the firm. Informal entrepreneurs are aware of these Business Environment benefits: half of them responded that they felt A sound business environment is a foundation formal registration would bring them benefits. for enterprise growth and employment creation. Thus, easing the regulations for starting and Kenya has historically remained behind its peers operating a business should increase formality, in most aspects of the business environment and in turn productivity and output. (Box 2.5) until it gained a noteworthy momentum in its business environment reforms last year. What Is Constraining Job Creation in According to the Doing Business 2016 Report, Kenya? Kenya’s increased by 21 places, moving from Several factors constrain the labor market in 129th to 108th place globally. Kenya introduced a Kenya, on the demand and supply sides. This total of 4 reforms making it easier to do business in chapter centers on: (i) the business environment; the areas of starting a business, getting electricity, (ii) the education system, that is, skills supply, registering property and getting credit. These as school enrollment has rapidly expanded efforts helped Kenya be recognized as the third and employers are concerned with challenges most improved economy globally in the period related to quality and mismatch; and (iii) labor from June 2, 2014 to June 1 2015, and positioned regulations. The latter relates to a small segment it as the second most business-friendly economy of the labor market—formal employment only— in the East African Community (EAC). but in the long run it is relevant, as productivity 44 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Box 2.5: Who can Kenya learn from about improving the business environment? While Kenya fares well compared to its regional, and even global competitors in the area of Getting Credit, it shows room for improvement in the 9 other areas measures by Doing Business indicators, where its ranking remains in triple-digits. Areas such as starting a business, dealing with construction permits and resolving insolvency, in particular, could benefit from additional streamlining to reduce procedure count and cost needed to complete these processes, as well as strengthening the underlying legal framework. While several efforts to pass various bills aiming to improve these areas are ongoing, implementation and communication remains to be an important aspect of reforms that should not be neglected moving forward. Starting a business. Senegal has made solid progress in facilitating the creation of formal businesses by Table B2.5.1: Kenya Remains behind in the Doing shortening the time to register a business to six days Business Indicators (involving only four procedures). In Kenya, it takes Selected peer Kenya countries 26 days and 11 procedures to start a company. Starting a business Senegal (90) 143 Dealing with construction permits: The region’s Getting electricity Ghana (151) 71 best performer in this area is Mozambique, where Registering property Vietnam (33) 136 it takes only 10 steps and 111 days to obtain Paying taxes Rwanda (27) 102 a construction permit, while the cost for the Trading across borders Tanzania (137) 153 procedure is at 3.7% of the warehouse value. In Enforcing contracts Vietnam (47) 137 Kenya, the procedure count is above the regional Resolving insolvency Uganda (98) 134 average (15 vs 14.5) and it takes 146 days to obtain Source: World Bank 2016 (www.doingbusiness.org). a construction permit, while the cost is also above the regional average with 6.9% of warehouse value. Registering property. Rwanda, is the region’s best performer in this area and it takes 3 steps, 32 days and costs only 0.1% of property value to transfer a property. In Kenya, the process is twice as long with 61 days, and the procedure count is 3 times bigger, at 9 steps. The cost is at 4.2% of property value. Getting electricity: Mauritius is the region’s best performer in this area, where it takes only 4 steps, 81 days and costs 260% of income per capita to get a new electricity connection. Mauritius also scores 6 out 8 points on the new reliability of supply and transparency of tariff index. Ranked 127th on this indicator, Kenya takes the same amount of steps to get a new electricity connection, but shows significant room for improvement in streamlining the time and cost needed for new electricity connection, and improvement under the new reliability of supply index, where it scores zero. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 45 Resolving insolvency. Mauritius is the region’s full interoperability between operators and best performer in this area, where the recovery the existence of exclusive contracts between rate is 67.4 cents on the dollar and the strength operators and cash merchants reduce of insolvency framework index is 9.5 out of 16 the possibility of new market entrants. In points. In Kenya, the recovery rate is only 27.9 professional services, limitations on advertising cents on the dollar and strength of insolvency and on partnerships hinder new market entry index score is 5 out of 16 points. Kenya’s product and investment. In agribusiness, unclear market regulations are also unnecessarily and unnecessarily burdensome licensing restrictive. Regulatory characteristics of the requirements for processors, as well as rules business environment determine the incentives that require consent from incumbents to grant and ability of firms to participate in markets and a new license, discourage entry. Likewise, the compete. Compared with other countries, in Kenya lack of a market-oriented process for spectrum the regulatory framework presents a high degree assignment may be hurting growth in the of restrictiveness to firm entry and expansion, as telecommunications sector. Finally, lack of clarity shown by the Barriers to Entrepreneurship indicator on whether the competition law applies to state calculated using the Product Market Regulation corporations raises concerns about potential methodology designed by the Organisation for market distortions. Economic Co-operation and Development (OECD). The level of restrictiveness in Kenya is higher Multiple factors are behind the weak business compared with the OECD average and many low- environment: lack of implementation capacity and middle-income economies, such as South and the challenges of multiple agency Africa (Figure 2.14).55 coordination top the list. In the past few years, several initiatives to streamline business Figure 2.14: Kenya is among the countries with restrictive product market regulations regulations have failed to be implemented, as other legislation took priority over business law review and amendments. Lack of effective 4.0 coordination among the various institutions in Product market regulation (scale 0-6) 3.5 charge of business regulations has undermined 3.0 2.5 2.0 the capacity to drive these complex reforms, 1.5 1.0 and the recent devolution of national functions 0.5 0.0 has added further challenges to this process, as capacity constraints are generally exacerbated at Romania India Chile Costa Rica Turkey Colombia Kenya Jamaica Bulgaria Peru Mexico Israel Brazil China Honduras El Salvador Nicaragua OECD avg Dominican R. Russia Federation Croatia South Africa the county level. Reform in some aspects of the business environment has also been hampered Sources: OECD Product Market Regulations database; OECD-World Bank by vested interests. For example, lawyers may Group Product Market Regulations database. Note: A score of 6 indicates the most restrictive regulatory framework. benefit from the complex processes associated with company registration, public officials may In addition to the above horizontal measures benefit from the lack of automation in property that constrain business activity, there are and land transactions, and restrictions on market many examples of sector-specific obstacles entry and competition may be a result of political to doing business. For example, in the market lobbying by incumbent firms. for mobile payment systems, the absence of 55 Data are not available for the peer countries (except India) used throughout the report. 46 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M More recently, the Government of Kenya Quality of Skills and Education has made significant efforts to improve the Kenya’s education system is failing to meet business environment, under the coordination market needs, as it does not prepare the of the Ministry of Industry and Enterprise labor market entrants with appropriate skills. Development (with the creation of the Ease of Although the quantity of graduates is rising Doing Business Delivery unit). Momentum has rapidly, businesses are increasingly complaining been gained in prioritizing reforms, particularly in about shortages of skills in the labor market core bottlenecks, including company registration, (Figure 2.15). This is particularly true for services electricity connections, property transactions, firms, which represent the fastest growing and access to credit. Counties are following suit segment of the economy. In 2013, more firms to improve their local business environments. in Kenya were identifying skills as a major One of the first movers is Mombasa county, constraint than in the rest of SSA. In large part, which has embraced the use of technology to interact with businesses through automated the mismatch of skills seems to be caused by the systems for construction and business permitting. quality of the education system. Kenya’s basic The results of these efforts should improve the education system continues to overemphasize operating environment for companies and are teaching facts and imparting knowledge, rather expected to improve Kenya’s poor performance than the development of analytical and problem- on international benchmarks. solving skills (Murthi and Sondergaard 2012). The system is also weak on creating job-relevant Acting on the weakest links is expected to technical skills (for example, through technical boost investment and market competition. and vocational education, higher education, pre- There is growing evidence that streamlining employment, and on-the-job training), along business regulations stimulates economic with other skills valued by employers, such as activity. For example, reducing the burden for accessing information, using computers, solving starting a business is expected to result in the complex problems, and learning new skills while establishment of more firms, and this has been on the job. confirmed by several case studies.56 According Figure 2.15: Finding skilled workers is becoming a major to the Doing Business 2013 report (World Bank challenge for employers 2013a), improved regulations are associated Percent of firms identifying inadequtely educated workforce as a with higher inflows of foreign direct investment. major constraint Haidar (2012)57 goes further and examines the link between “doing business” reforms and Services economic growth, and finds a positive significant Manufacturing relationship between the two. In addition, several studies show that streamlining product market All firms regulations promotes innovation, employment, 0 5 10 15 20 25 30 35 40 and productivity growth. Nevertheless, improving 2007 2013 the business environment is necessary but not sufficient to achieve the desired jobs and poverty Source: enterprisesurveys.org. reduction targets; human capital and streamlined labor regulations are also critical. 56 http://www.doingbusiness.org/reports/case-studies/topic/starting-a-business. 57 Haidar 2012. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 47 The current 8-4-4 education system, which was platform established in 2010, precisely with the introduced in the 1980s, has failed to keep up objective to facilitate linkages between firms and with the fast-changing labor market and needs educational institutions.58 reform. The system needs reform to focus on promoting the acquisition of strong generic Tertiary education among youth has expanded (cognitive and noncognitive) skills and enhancing rapidly. The coverage of tertiary education has the alignment of academic curricula with market expanded more rapidly in the past two decades, demands, particularly at the secondary and mainly as a result of upgrading of colleges to tertiary levels. The expansion of tertiary education universities as well as through what is referred should be managed through the strengthening of to as “parallel programs” where students quality assurance frameworks and the provision pay tuition for part-time or distance learning of better information on labor market prospects programs. Expansion has occurred in all public in various fields. In addition, emphasis should be universities and in private, for-profit providers of placed on the production of skills for innovation tertiary education. However, Kenya has not joined at the tertiary level. The skills demanded by firms the trend of attracting world class universities today are rapidly shifting from routine, manual, to establish an in-country presence (campus); and cognitive skills toward more non-routine, neighboring Rwanda, for example, saw the opening higher-order skills. of a campus by Carnegie Mellon University.59 Technical and vocational education needs Increasing the pool of university graduates is adequately designed expansion. In an effort to good for productivity growth. A study from address the large youth unemployment, there 2006 showed that returns to tertiary education are pressures to expand technical and vocational are high in Kenya’s urban centers (Kimenya et education. However, very little is known about the al. 2006). Returns for women are significantly current system, its quality, and the employment greater than for men in rural and urban settings trajectories of its graduates. Addressing skill (Table 2.2). Increasingly, however, research mismatches among young workers calls for well- demonstrates that rates of returns vary with designed apprenticeship programs that ease the the quality of the skills imparted. Recent studies transition from school to work by developing show that rather than years of schooling, it is behavioral skills. International experience the quality of cognitive skills that determines underscores the importance of the governance individual earnings. Labor markets clearly offer of technical and vocational training, and close returns to those who have skills that are relevant, partnerships with private sector employers. rather than years in school per se. For example, the recent development of oil exploration and extraction, which is new to The rapid expansion of tertiary education Kenya, will necessitate that job entrants have carries significant risks, including for the quality specific skills. Hence, education policy should be of learning. The expansion of post-secondary geared toward creating those skills. There have educational opportunities has been driven by been positive steps toward strengthening the link demographic pressures as well as pressures between businesses and academia. One example arising from earlier reforms undertaken in for this is the “Linking Industry with Academia” primary and secondary education. However, it 58 http://liwatrust.org/index.php/about-us/linkages. 59 http://www.cmu.edu/rwanda/. 48 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Table 2.2: Private returns to tertiary education are high, 2006 (%) Completed Completed TVET University Category primary secondary National 7.7 23.4 23.6 25.1 Urban 9.3 34.4 26.2 34.8 Rural 7.8 21 22.4 14.2 All males 4.4 21.2 12.8 23.3 Urban males 6.1 25.6 17.9 30.7 Rural males 4.2 20.2 12.4 12.6 All females 13.2 36.3 43.5 62.5 Urban females 6.2 44.9 28 66 Rural females 16 30.3 51.5 18.6 Source: Kimenya, Mwabu, and Manda 2006. Note: TVET = technical vocational education and training. seems evident that most tertiary institutions have 4 percent of firms found labor regulations to emphasized revenue generation, while placing be a major constraint to doing business, far less weak or nonexistent mechanisms to maintain or than the SSA average of 12 percent. However, improve quality. At the same time, the expansion by 2013 the share of firms rose to 20 percent, of tertiary education carries a risk of inequitable while in the rest of the continent the share access if the system fails to equalize opportunities of firms complaining about labor regulations for key constituents, such as women, the rural remained unchanged at 12 percent.61 The main population, and those with low income. grievances concern the strict medical surveillance requirements, health and safety audits (Box 2.6), Labor Regulations as well as the high minimum wage. The 2007 changes to the labor code seem to have disincentivized formal employment. Labor The strict regulations partly explain the rise in legislation was drastically revised in 2007, without disputes between employers and employees. wide stakeholder participation, and firms have Industrial disputes between employers and not been happy with the reform. The reform was employees have been on a rise in recent years. done primarily as an appeal to the trade unions According to KNBS data, the number of workdays ahead of the December 2007 elections; the lost as a result of such disputes skyrocketed from Central Organization of Trade Unions claims to 15,000 in 2008 to 175,000 in 2011. Disputes in the have 1.7 million members.60 Many of the changes agriculture and forestry sectors have been most that were introduced have been disputed by prevalent, accounting for half the workdays lost. employers and their business associations, The transport, manufacturing, and construction who continue to voice their concerns about the sectors follow, while in other sectors of the strict regulations. Businesses’ perceptions are economy such incidents are rare. also noted in enterprise surveys. In 2007, only 60 http://cotu-kenya.org/about/ 61 www.enterprisesurveys.org FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 49 Box 2.6: Examples of labor regulations that are causing firms to become risk averse in hiring The Work Injury Benefits Act (WIBA) and the Occupational Safety and Health Act (OSHA) provisions seem to have become a source of costly litigation between employers and employees. Although these laws seek to protect employees’ rights and were enacted to root out oppressive practices at the workplace, they have resulted in higher costs for formal hiring. For example, the definition of a dependent in the WIBA is too wide and it can result in unnecessary litigation instead of the law limiting the dependents to the immediate family only. The costs of annual safety and health audits and risk assessments have been loaded onto employers. OSHA Provision No. 15 of 2007 introduced compulsory annual safety and health audits, risk assessment, and the requirement for a health and safety statement by all employers. The costs of these undertakings are loaded onto the employers instead of the agencies undertaking the audits. The cost of compliance with this requirement will drive out small investors who are unable to conform because of lack of capacity to conduct the audits and assessments. Compensation levels for injury at work are also perceived to be high by employers. According to the Kenya Association of Manufacturers, an employee earning K Sh 50,000 a month would be eligible for compensation of K Sh 4.8 million, and one earning K Sh 1.2 million a month could get K Sh 115 million in case of permanent disablement. Such high amounts make businesses risk averse when it comes to employment. The prescribed minimum wage may also be Figure 2.16: Minimum wage is highest in Kenya among peer countries pushing firms toward informality. Kenya’s 140 1.0 average prescribed minimum wage in 2012 120 was K Sh 5,704 for agricultural workers and K 100 0.8 US dollars Sh 10,646 and K Sh 13,471 (depending on the 80 0.6 Ratio 60 urban area) for those working in manufacturing 40 0.4 and services, respectively. Consequently, and to 20 0.2 no surprise, KNBS data also show that less than 0 0.0 4 percent of formal wage workers in trade and hospitality earn less than K Sh 20,000, and at Minimum wage applicable to the worker assumed in the case study (US$/month) Minimum wage for a 19-year old worker or an apprentice (US$/month) the same time there are more than six million Ratio of minimum wage to value added per worker informal workers in the sector. The same goes Source: World Bank 2014b. for manufacturing and construction, where less than 1 percent of formal workers earn less than Understanding Kenya’s Segmented Labor K Sh 15,000, while there are two million and Market 300,000 informal workers, respectively, in the To meet the MTP-2 goals on job creation, Kenya’s two sectors. The minimum wage is relatively government would have to focus in parallel high in an international comparison. According on three segments of the working population. to the Doing Business 2014 report, Kenya had Job creation is a multifaceted policy challenge, the highest minimum wage for a young market and one way to approach it is by grouping new entrant among a group of peer countries (Figure labor market entrants in three groups: (i) urban 2.16). Moreover, the ratio of the minimum wage educated youth, (ii) urban low-skilled youth, and to worker productivity (measured as value added (iii) rural (low-skilled) youth. Different stimuli will per worker) was by far the highest. be required for each group. 50 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M The urban educated youth demand “modern” investment and productivity growth, as well as jobs. To make this cohort fit for modern jobs, the reforms to make labor regulations more friendly to education system, in particular tertiary education, hiring low-wage workers. And active labor market should focus on quality as well as on producing policies, in particular those that involve the private skills that the market needs. In addition, firms sector, could boost demand for such jobs. in modern sectors, including financial and business services, or high-skill manufacturing But in the short to medium term, low-skilled sectors, should be allowed to operate in an urban youth are more likely to end up in the environment that is conducive to investment and jua kali. The informal sector will continue to be innovation. Moving to such an environment will the employer of last resort, as it has been for the require policy reforms to remove some of the past few decades. Hence, public policies should regulatory burden, as well as investment in the be oriented toward boosting productivity in the infrastructure that these sectors need. Empirical sector rather than limiting its presence. Countries evidence from neighboring Tanzania (box 2.7) in SSA have taken divergent approaches on this illustrates the need to place qualified market topic. South Africa is a good example (box 2.8). entrants in adequate jobs at the start of their World Bank (2014a) considers informality in low- career; otherwise they risk going for, and staying and middle-income countries and points to the in, low paid traps. important fact that there is heterogeneity among informal firms, so different policies need to be The low-skilled urban youth will aim for low- devised for different types of firms. Nevertheless, skill formal wage jobs. Kenya has the potential to generate formal jobs in low value-added a broad lesson is that public policies should have a manufacturing, construction, or tourism- dual focus. First, they should aim to improve firm related services such as hotels and restaurants. productivity by boosting skills, improving access Unleashing this potential requires business to finance, or financing business services. Second, environment reforms and infrastructure they should be aimed at enhancing the quality of improvements that promote firm creation, services, governance, and the institutional system Box 2.7: Locking labor market entrants in low-productivity jobs limits their long-term earning potential A recent study by Falco et al. (2014) uses the Ghana and Tanzania Urban Panel Surveys to examine the determinants of earnings, earnings growth, and low-pay/high-pay transitions in the high growth period 2004–08. The findings highlight the relative importance of job characteristics—over workers’ endowments— in determining earnings and earnings growth. The findings also point toward path dependence in pay trajectories. This conclusion is reinforced by the finding that being in low-paid employment has a scarring effect: it undermines future earnings prospects. In other words, low pay is a persistent condition and new entrants in the labor market risk being trapped in low-paying occupations, and some groups of workers— women and youth—are particularly at risk of falling into low-pay traps. Falling into low pay undermines individuals’ prospects for obtaining high-paying jobs in the future. Source: Falco et al. 2014. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 51 in general, as these features strongly influence Box 2.8: South Africa’s informal economy policy was the decision of firms regarding informality. Other born in eThekwini/Durban African countries are doing more in this regard, The broader Durban metropolitan area, eThekwini, according to MasterCard’s African Cities Growth has been at the vanguard of establishing a friendly Index 2014, which ranks Nairobi at position 19 policy environment for the informal economy. Its of the 49 large African cities surveyed. The index pro-informal economy interventions have been concludes that the capital has low potential for emulated by other local governments in the country growth in the next five years, based on capital and serve as the catalyst for a broader national formation, political stability, GDP per capita, reflection on local government and informality governance, and household consumption.62 in South Africa. What began with a department dedicated to street trader management in 1990s One way to boost productivity in the informal has grown into a full-fledged policy that aims to sector is through offering market information. support the informal sector and allocate resources Even in rural areas, information technologies and to infrastructure development for these firms. mobile penetration are becoming an effective Source: http://wiego.org/informal-economy/durbanethekwini- south-africa-informal-economy-policy. platform for disseminating market information, such as price trends or market opportunities. These technologies are already in place. For throughout Africa. Kenya benefits from a well- example, farmers can check produce prices with organized informal sector, so jua kali associations their phones. Expansion of such solutions—either could be actively involved in designing and by the public or by promoting private sector implementing apprenticeship programs. involvement—would raise productivity and boost informal employment. County authorities can also learn from Kenyan examples of how to support informal Another way to promote the jua kali is through businesses. The Muthurwa market in Nairobi, developing targeted skill-building programs, a US$9 million project, created the then largest including apprenticeships. Informal workers market in East and Central Africa, with capacity typically depend on their income for survival, and for 8,000 traders. The market was constructed so cannot attend training during the day. They with the aim to boost the market efficiency also have limited access to new technologies and productivity of thousands of traders who, and pedagogical sources of training (World Bank because of lack of adequate space, were selling 2013a). Hence, the report suggests that training on the streets of Nairobi. The project was programs should be tailored to informal workers: successful in moving traders from the street to be flexible and affordable, provide access to the market; however, just a few years down the relevant technology, and offer the broad range of road, Muthurwa market was in a dilapidated and skills needed (World Bank 2013a). Evidence from unhygienic condition as a direct result of the lack past programs in Kenya shows that vouchers can of management and support infrastructure.63 This stimulate private sector provision of such skill- illustrates the point that putting up infrastructure building programs. In addition, apprenticeships is necessary but not sufficient. A contrasting have been proven to be an effective tool for case is an example from Bamako, Mali, where promoting skill building and youth employment the local authorities delegated the management http://www.businessdailyafrica.com/Nairobi-ranked-low-on-inclusive-growth-in-Africa/-/539546/2342528/-/hvqspez/-/index.html. 62 http://www.delog.org/cms/upload/pdf-africa/managing_informality_local_governments_practices_towards_the_informal_economy.pdf. 63 52 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M of markets to the informal traders. The results production, and exports. Policy reforms could were improved market conditions (hygiene, change the incentives for farmers to diversify sanitation, and access to water) and increased and shift toward more productive crops. Risk tax collection. This example shows that the buy- management interventions, such as insurance, in and involvement of stakeholders in project could help pastoralists cope with droughts and management are essential, and that partnerships generate income even in times of bad weather. between local authorities and informal traders can bring win-win outcomes. The second priority is to expand the possibilities for nonfarm income in rural areas, Finally, a comprehensive job creation strategy including through promoting mobility (better must recognize that more than half of Kenyans infrastructure) so that those living in rural areas will continue to live in rural areas. Even if could work in urban centers. Third, as agricultural urbanization picks up at a faster pace, the number productivity increases and services and of Kenyans living in rural areas will increase by manufacturing are allowed to grow, the transition about five million by 2020 and the vast majority from rural to urban areas should be encouraged by would be involved in small farming. Hence, the policies that promote urbanization. Such policies first step is to take measures to enhance the could include improved public services, clear productivity of small farmers, and this comprises property rights, and better urban infrastructure. a mix of public investment, institutional reforms, Some reforms along these lines are already being and regulatory changes. Investment in irrigation implemented. For example, there has been an offers huge returns: even moderately successful expansion of public investment in rural roads. investment in smallholder agricultural water However, these reforms would have to be further development could triple per capita farm expanded and accentuated to achieve the desired incomes. And investment in infrastructure can pace of change. facilitate downstream industries, such as food FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 53 The Jobs Potential of Special Economic Zones Kenya’s Experience with EPZs in 2005—and the new competition from East Asia—investment in the EPZs The export processing zones (EPZ) slowed sharply, despite some success program of two decades ago ended up in diversifying production in other being heavily focused on the apparel sectors, notably agro-processing. Total industry, but expansion of production employment has remained around slowed after 2005. With the EPZ in 40,000, and between 2009 and 2013, place, Kenya was well-positioned to the number of EPZ firms declined from take advantage of the opportunities 83 to 81 (figures 2.17 and 2.18). available through the U.S. Africa Growth and Opportunities Act (AGOA), which Since 2011, the EPZs have been boosted SPOTLIGHT 2 came into effect in 2000. The EPZs grew by AGOA-linked apparel exports, owing rapidly and by 2004 more than 40 zones to rapidly rising wages in China. By were established, employing close to 2013, the zones hosted more than 40,000 workers and contributing 10 80 enterprises, employed more than percent of national exports. Production 35,000 Kenyans, and exported more at the EPZs was highly concentrated in than US$500 million (of which 8 percent clothing exports to the United States to the East African Community (EAC) under AGOA, accounting for around 80 market). Nevertheless, on a global scale, percent of EPZ exports and 90 percent of Kenya’s EPZ results are not impressive. employment. Following the expiration Countries like Costa Rica and Vietnam, of the Multi-Fiber Arrangement (MFA) as well as China, ramped up investment Figure 2.17: EPZ exports rising but employment Figure 2.18: Relative contribution of apparel steady declining Source: Data from EPZA Annual Performance Reports (various Source: Data from EPZA Annual Performance Reports (various years). years). Note: EPZ = export processing zone. Note: EPZ = export processing zone. 54 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M and exports much more quickly (Figure Nevertheless, like other African zone 2.19). The MTP-2 aims to change this programs, Kenya’s EPZs also suffered trend, so learning from own successes from some critical weaknesses in design and failures, as well of those of other and delivery, including the following: countries, should be a first step in revamping the zones program. Overemphasis on single-factory units. Zone authorities are generally stretched Figure 2.19: Evolution of export growth in selected global SEZs from year of launch to carry out their mandate within the main zones, and servicing single-factory units (with infrastructure, support services, etc.) is even more difficult. Failure to address infrastructure and other barriers to competitiveness outside the EPZ gates. Kenya’s EPZs offered a better investment climate than what was available to firms in SPOTLIGHT 2 the domestic economy. The quality of Source: Farole 2011. infrastructure and services was also Note: SEZs = special economic zones. superior to other African EPZs, yet Lessons Learned from the EPZ it still lagged considerably behind what Experience Asian and Latin American countries Kenya’s unsatisfactory experience with offered. One of the reasons for this was EPZs mirrors that of most African zone that although the EPZ infrastructure programs. Part of the story is simply and regulatory environment was one of bad timing. The rapid growth of effective, little was done to address economic zones worldwide and their gaps beyond the EPZ gates, including success in stimulating export-led growth issues such as electricity outages (Figure owes in part to an unprecedented era 2.20), electricity costs, customs (Figure of globalization of trade and investment 2.21), transport logistics, and the low that took place during the 1980s and productivity of the labor force. 1990s, with the rise of global production networks (GPNs). But African countries, Rigid model that restricts potential for most of which launched programs diversification and local integration. The only well into the 1990s and 2000s, EPZ model adopted in Kenya was relevant face a much more difficult competitive only for export-oriented, assembly- environment. This situation resulted not related activities relying on imported only from the expiration of the MFA, but inputs. Companies in the EPZs were also from the entrenchment of “factory disadvantaged when it came to serving Asia,” the consolidation of GPNs, and the the large EAC market, and services post-2008 slowdown in global demand. companies were excluded altogether. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 55 Figure 2.20: Monthly downtime caused Figure 2.21: Days to clear imports from customs 50 50 45 40 34 35 30 26 25 20 16 15 9 10 5 6 4 5 2 2 0 Samir Bangladesh Athi River Single Factory Unit Tanzania Honduras Lesotho Vietnam Ghana Non_EPZ exporters Source: Farole 2011. Source: Farole 2011. From EPZs to SEZs—Getting It Right are seldom the most important criteria This Time Around for decision. For example, in Bangladesh, electricity infrastructure was a big Kenya is in the process of transforming constraint. This constraint was resolved SPOTLIGHT 2 its economic zones program; however, in the zones by establishing a new law SEZs as a model are not a panacea for allowing investors to establish power economic transformation, and simply plants and resell production within the adopting a new regime is no guarantee zones. The largest constraints for foreign of success. What can Kenya do to help companies in Kenya include regulations ensure the likelihood of greater success and tax policy, followed by logistics and with the new regime? In addition to infrastructure. Obtaining work permits learning from own experience and the for expat staff, typically managers, has experiences of other EPZ programs in Africa, Kenya’s government can draw been especially difficult according to on the experiences described in this a survey done by the International section. The new SEZ bill is expected to Finance Corporation of foreign investors address some of the issues, although in Kenya. Having restrictive policies or the key to success lies in successful practices on employing foreign workers implementation. seems unnecessary, as two-thirds of the managers in SEZs are Kenyans, a level SEZs should address the most binding similar to other zones across the world. constraints to investors. Too often zones are developed with little that Use SEZs to leverage competitive differentiates them from the national advantages and facilitate agglomeration environment. It is critical to begin with rather than to support the development understanding “what are the most of lagging regions. International important constraints to investment experience has shown clearly that the in the country?” and addressing those location of an SEZ in a country—in constraints in the zones. In this context, particular, its proximity to major trade it is worth noting that although investors gateways (ports and airports) and the will always ask for fiscal incentives, these country’s largest metropolitan areas— 56 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M is critical to the success of the SEZ. This that contribute to delivering on an SEZ is finding is particularly important for of critical importance. A one-stop shop zones that depend on manufacturers is an effective tool to ensure timely and who require access to imported inputs, efficient approvals for initial setup and business services, large pools of labor, ongoing operations within the SEZ. and transport networks. But it also holds true for knowledge-based zones Actively promote linkages to the (for example, information technology domestic economy. The SEZ regime parks), which require proximity to should actively pursue policies that specialized labor and high-quality promote diffusion of knowledge, backbone services. In Kenya, this technology, and backward linkages means focusing on Nairobi (especially with domestic firms. The difference is for services-oriented investments), obvious: in the Republic of Korea’s main Mombasa (exports for world markets), SEZ in 1971, domestic firms supplied and Kisumu (exports for EAC). only 3 percent of inputs to foreign firms in the zone; four years later they supplied Infrastructure remains a critical factor a quarter; and a few years down the for success. In almost all low- and road, their share rose to almost half of SPOTLIGHT 2 middle-income countries, the provision the value of inputs. In the Dominican of quality infrastructure will be the Republic, by contrast, the share of single most important way in which domestic value added started at a similar a zone program can offer a “special” level as in Korea, but never moved up. environment to investors. This may The Korean government encouraged include the provision of land and factory backward linkages with domestic shells with flexible lease terms (reducing risk for investors), but most critically firms. For example, local firms supplying it means delivering an environment the zones had preferential access to where the supply of utilities (water, raw materials and technical assistance telecommunications, and most was provided to subcontractors. In importantly electricity) are dependable contrast, the Dominican Republic and available at a reasonable cost. did not introduce any incentives to However, although the infrastructure of a domestic firms and actually made zone is important, it is equally important their operations difficult by requiring to develop the connective infrastructure difficult-to-obtain export licenses to between SEZs, cities, and ports. sell to the firms in the SEZs. Institutions matter—build capacity SEZs take time. Finally, it is important and ensure coordination. The to remember that reaping benefits implementation of SEZ programs is of SEZs requires time and consistent as important as the design and legal effort. Experience from countries with framework of this instrument. Given successful SEZ programs, such as China the reality that few governments or Malaysia, shows that it takes at 5 to ultimately establish a powerful and 10 years to build momentum in SEZs, autonomous SEZ authority, effective in particular with regards to backward coordination across the many agencies linkages with the domestic economy. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 57 58 RAN K E N YA C O U N T R Y E C O N O M I C M E M O Photo: M D UBank World CHAPTER 3 RAISING INVESTMENT THROUGH SAVINGS Introduction some years negative, real deposit rate, which is likely an outcome of volatile inflation. In As chapter 1 illustrates, stimulating investment addition to the macro determinants of private is the most effective way for policy makers to savings, Spotlight 3 (at the end of this chapter) accelerate economic growth in the medium looks at the behavioral aspects of saving. Finally, term, and Kenya’s investment rate is lower higher savings would need to be accompanied than the rates of its peers. The current level by better macroeconomic management, that is, of investment would not yield the Vision 2030 lower economic volatility and improved public and Second Medium-Term Plan (MTP-2) growth investment management; otherwise, savings targets. The MTP-2 sets an ambitious target— may end up in low-return investments. investment rate to reach 31 percent of gross domestic product (GDP) by 2018—and this target Going forward, higher savings can be reinforced is accompanied by an anticipated jump in savings. through economic policy. First, the demographic trend of the past decade is expected to continue; So far, the increase in investment has resulted hence, more job opportunities for youth would in a rising current account deficit that is promote higher savings. Second, if inflation is unsustainable in the long term. The increase kept under control, deposit rates are envisaged in investment over the past decade has been to turn positive, which in turn would incentivize financed primarily from foreign savings, which saving. Third, recent changes in fiscal policy, such brings into question the sustainability of this as the re-orientation of spending from recurrent model, although non-debt-creating inflows have to capital expenditure, are projected to continue, so far accounted for the bulk. Domestic savings, which implies an increase in public savings. private and public, started to increase after 2011, Finally, growth-accelerating economic policies are likely driven by the rise in corporate savings. expected to boost income and promote saving. Public savings are relatively low, although on an upward trend. Encouraging saving, in particular household saving, is even more relevant in the context Several factors, including youth unemployment, of recent oil discoveries. First, oil production negative real deposit rate, and low public and exports are not expected to occur before savings, are the most relevant for the low 2020-2022, and the government’s strategy is to savings. Demographic trends over the past boost savings and investment by 2018. Second, decade, such as falling youth dependency ratios, even under a conservative spending scenario should have promoted saving. However, this may (of oil rents), Kenya’s budget is likely to rely on not have been the case, given that the effective borrowing in the early years of oil production. youth ratio (which takes into account youth Thus, encouraging household savings could employment) has fallen only slightly. Another be seen as a way to mitigate pro-cyclical fiscal reason for low savings has been the low, and in responses to anticipated oil revenue. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 59 Investment and Long-Term Growth The economic literature demonstrates that low and declining savings hampers long-run Chapter 1 concluded that Kenya needs higher investment and growth. Across the globe, saving rates of investment to accelerate growth. The rates exhibit high correlation with investment Growth Report (Growth Commission 2009) and economic growth. Although there has been concluded that an investment rate of 25 percent controversy on causality between savings and of GDP or higher was common among the high- growth, the causality that runs from savings to growth countries of post-World War II. Previous growth plays a critical role through the capital literature has confirmed that investment accumulation process.64 In theory, it does not determines how fast economies can grow (Figure matter how investment is financed. In practice, 3.1). However, economic growth cannot rely however, a close connection between the two solely on capital accumulation in the long term is observed, especially in the long run (Figure because of diminishing returns. Sustainable 3.2). Investment is not necessarily financed by growth needs human capital enhancements and national savings if a country has access to external technological improvements. sources. Nonetheless, the country cannot rely Figure 3.1: Investment and growth are highly correlated on external financing over a long-term horizon, given that large current account deficits cannot 10 be sustainable. 8 GDP per capita growth 6 Kenya’s savings65 have not only declined, but are Kenya 4 also the lowest among selected peer countries.66 2 In the 1980s, Kenya’s average saving rate was 0 higher than the saving rates in several peer -2 countries (Figure 3.3). Since then, Kenya’s rate -4 has declined, while its peers have boosted their 5 10 15 20 25 30 35 Investment rate (% of GDP) saving. Over the past decade, Ghana, Senegal, and Uganda, which had among the lowest Source: World Bank World Development Indicators. Note: GDP = gross domestic product. Figure 3.2: Savings are correlated with investment and growth 40 10 35 8 Investment rate (% of GDP) 30 GDP per capita growth 6 25 4 20 Kenya 2 15 0 10 Kenya 5 -2 0 -4 0 10 20 30 40 0 10 20 30 40 Savings rate (% of GDP) Savings rate (% of GDP) Source: World Bank World Development Indicators. Note: GDP = gross domestic product; GNDI = gross national disposable income. 64 Aghion, Comin, and Howitt (2006) developed a theory that domestic saving affects economic growth in low- and middle-income countries that are far from the technological frontier. 65 Gross domestic savings are measured relative to GDP and are defined as the difference between GDP and final consumption. Gross national savings are measured against gross national disposable Income (GNDI), which equals GDP plus net income and transfers from abroad. In the case of Kenya, where transfers from abroad, which include development aid, are significant, gross national savings is the more appropriate indicator. 66 See chapter 1 on the selection criteria for the peer countries. 60 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M saving rates in the 1980s, have surpassed Kenya around 16 percent of gross national disposable in savings. Neighboring Tanzania has already income (GNDI). Since then, savings have declined passed the 20 percent mark (23 percent in 2012), to 12 percent of GNDI, while investment has although its gross national income per capita is been on a gradual upward trend, reaching 20 lower than Kenya’s. In addition to the increase percent of GNDI. The consequent widening of in income and higher GDP growth, population the current account deficit has been financed growth and longer life expectancy contributed mostly by non-debt-creating foreign inflows, to higher saving because of the increase in the including foreign direct investment, portfolio working-age population. The same trend of rising equity investments, and real estate investments, savings can be observed across most of Sub- although it is difficult to pinpoint the financing Saharan Africa (SSA). Low-income SSA’s average sources because of weaknesses in the balance saving rate went from 11 percent in the 1990s, of payments statistics.68 Nevertheless, in 2014 to 12 percent during the 2000s, and by 2013 it the government moved toward external sources reached 18 percent.67 Similarly, saving rates for to finance the fiscal deficit. In the second half of lower-middle-income SSA rose from 13 percent 2014, €2.75 billion in Eurobonds were issued, and in the 1990s, to 17 percent in the 2000s, and a US$3.6 billion loan was signed with the Export- 20 percent in 2013. For example, Tanzania and Import Bank of China for the financing of the Uganda have achieved remarkable investment standard gauge railway project. Consequently, rates, and high saving rates are a large contributor external debt is projected to continue to increase to this success. over the medium term. Figure 3.3: Decade average savings rate: Kenya and its peer countries (Percent) Nevertheless, the high current account deficits 35 are not sustainable in the long term, and even 30 less so if investment is to reach the targets 25 set in the MTP-2. The total external debt is 20 projected to increase from 27 percent of GDP in 15 2010 to 37 percent of GDP in 2018, which puts 10 a question mark on the long-term sustainability 5 of the savings-investment gap. The rising stock 0 of short-term foreign flows, including portfolio equity investments, exposes the economy to 1980s 1990s 2000s shifts in investor confidence or preferences. The Source: World Bank World Development Indicators. monetary tightening in the United States in early 2014, for example, illustrated the risk of reversal The widening gap between savings and of flows to emerging markets. In addition, these investment has so far been financed largely projections are based on the assumption of by non-debt-creating foreign inflows. Despite maintaining investment at around 23 percent of the low and declining savings, investment has GDP over the medium term, which falls short of been on the rise over the past decade. In 2003, the MTP-2 target. saving and investment rates were equal at 67 Low-income countries are defined as non-fragile countries with average per capita gross national income of less than US$1,045 in 2013. Those countries included Benin, Burkina Faso, Ethiopia, Ghana, Kenya, Madagascar, Malawi, Mali, Mozambique, Niger, Rwanda, Senegal, Tanzania, and Uganda. 68 In the September 2014 Article IV staff report, the International Monetary Fund estimates annual foreign direct investment (FDI) inflows at above 4 percent of GDP between 2009 and 2011 (based on a foreign investor survey), while in official balance of payment statistics FDI is below 1 percent of GDP. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 61 In the long term, oil rents could become a savings is very different from most of the peer major source for financing investment, but it economies, which have public savings rates will be years before a significant amount of oil that are several times higher than Kenya’s. For revenue starts flowing to the budget. As chapter instance, neighboring Tanzania’s public savings 5 discusses, oil production will start in 2020-2022 was around 8 percent in the 2000s and accounted at the earliest (and the sector is facing headwinds for 50 percent of national savings. in meeting this target date). Then, the first few years after oil production begins are likely to see Public savings, defined as total revenue minus increased public borrowing. Hence, domestically total expenditure and public investment, financed investment up to 2020 and probably has been fairly stable in Kenya. Changes in beyond will rely on increasing domestic savings. recurrent expenditures have mirrored trends in revenue collection, while the increase in public Financing Investment from Domestic investment, that is, development spending, Resources since 2006 has largely been financed by public borrowing. Consequently, public debt rose from The first step to understanding the domestic 40 percent of GDP in 2010 to 44 percent in 2013. sources of financing investment is to understand who saves in Kenya. Decomposing national Corporate savings is estimated to have been savings into private and public savings reveals on an upward, although volatile, trend since two characteristics of Kenya’s savings. First, public 2005. Official data on corporate savings are not savings have been low and mostly stable over the available. The data used here were estimated past three decades (Figure 3.4, panel a). Second, based on the retained earnings of 56 corporations private savings constitutes a large portion of listed at the Nairobi Stock Exchange for which national savings and has been the driving force financial data are available. Corporate savings of its trend. doubled from 1 percent of GNDI in 2005 to 2.4 percent in 2010, but then fell to 2.1 percent in In the past decade, Kenya’s public savings rates 2013 (Figure 3.5). The financial sector generates fluctuated around 1 percent of GNDI, and have the bulk of savings—a quarter of listed companies been less than 10 percent of national savings is banks or insurance companies—although it is (Figure 3.4, panel b). This trend of low public also the most volatile. Figure 3.4: Kenya’s public savings is relatively low and declining 25 10 8 20 6 % of GNDI 15 Percent 4 10 2 5 0 ya a a l a 0 -2 tan n di di ga an da as o am ni a kis Ke In bo ne Gh an aF tn za 1983 1986 1989 1992 1995 1998 2001 2004 2007 2010 Pa r am Se Ug in Vi e Ta n C rk Public Savings Private Savings Bu Source: International Monetary Fund World Economic Outlook. Note: In panel b, national savings as a percentage of GNDI is the average for the 2000s. GNDI = gross national disposable income. 62 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Figure 3.5: Corporate savings have been increasing, 2005–13 One way for Kenyan households to save is (K Sh, thousands) through formal contributory pension systems. Retirement saving through various pension schemes is relatively well developed by SSA standards. Kenya has three pension schemes: the pay-as-you-go Civil Service Pension Scheme, the fully-funded National Social Security Fund (NSSF), and various occupational pension funds. The Kenyan pensions system manages assets of about 17 percent of GDP and is dominated by the NSSF. The number of employees contributing to the NSSF rose from less than four million in 2008 Source: Calculations based on the financial data of corporations listed on the to five million in 2013. The pension schemes Nairobi Stock Exchange. Note: Data for some companies are missing in certain years. are regulated by the independent Retirement Benefits Authority (RBA), which is mandated to Data on household savings are not available, but ensure prudent management of pension assets. the analysis implies that households have been saving less over the past eight years at least. Similar to households in other countries in SSA, The last household budget survey was conducted one reason behind low savings is the continued in 2005/06; hence, there are no primary data increase in the value of immovable property. to show trends in saving since then. However, International experience shows that it is common the fact that national savings have been falling for individuals to save in nonmonetary assets since 2005, while corporate and public savings in an environment of low returns to saving in have increased marginally, implies a reduction financial assets, or low real rates on deposits. And in household saving. One of the explanations if property values are continuously growing— for this could be the increased borrowing by as has been the case in Kenya—owners expect households as access to credit has become such trends to continue and in turn generate easier. Commercial banks’ credit to households future income, thus lowering their need to save. rose fivefold between 2005 and 2013, to over 5 Although data on the housing stock and prices percent of GDP (Figure 3.6). are sparse, two trends back the hypothesis that Figure 3.6: Credit to households is growing rapidly Kenyans save in this way. First, housing prices (K Sh, billions) in the two main economic centers have soared. 350 According to the Wealth Report 2012 by Knight 300 Frank and Citi Private Bank (2012), real estate 250 prices in Nairobi and Mombasa increased by 25 200 and 20 percent, respectively, in 2012, placing 150 the two cities in the first and second positions, respectively, of 71 cities surveyed globally. This 100 happened at a time when construction of new 50 residential buildings exploded. The number of 0 new constructed residential buildings, according 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Source: Central Bank of Kenya. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 63 to the Kenya National Bureau of Statistics (KNBS) Kenya’s oil discoveries could make such a scenario Economic Survey 2013, quadrupled between feasible, it is yet unclear when the oil production 2008 and 2012. In addition to these factors, will reach its peak—probably in about a decade— Spotlight 3 (at the end of this chapter) looks at (chapter 5 discusses the macroeconomic effects the behavioral aspects of saving. of oil production). Consequently, long-term sustained growth requires a mix of more (and How Much Savings Is Needed to Achieve sustainable) investment and higher domestic the Desired Investment Growth? savings over the medium term. What level of savings is needed to achieve the What Determines Savings? development objectives of Vision 2030 and the MTP-2? Over the long run, growth requires The discussion so far argues that investment investment, and to be sustainable investment it should be accompanied by domestic savings. To should be accompanied by savings. The following put forward policy options on how to stimulate simulations employ a model that investigates the savings, this section reviews the literature on link between savings, investment, and growth. determinants of savings from the macro and micro The analysis answers the following questions: perspectives. The determinants of savings fall (i) what is the potential growth with the current into two categories: (i) nonpolicy determinants, saving and investment rates, and (ii) what saving such as GDP growth and demographic change; rates correspond to the desired Vision 2030 and (ii) policy determinants, including financial growth rates? sector development, macroeconomic stability, and income volatility (see box 3.1). As discussed in chapter 2, economic growth relies on the accumulation of factors of At a first glance, Kenya’s stubbornly low savings in production and productivity improvements. a period with a rapidly falling youth dependency Higher total factor productivity (TFP) implies ratio seems peculiar. The negative correlation that the economy can produce larger output between savings and the youth dependency ratio with a given level of physical and human capital. has been quite strong. In East Asia, for example, However, sustaining high TFP growth is not easy: a 10 percentage point decline in the youth only 5 percent of countries in the world achieved dependency ratio has been associated with a 3 an average 2 percent TFP growth during 1965–95. percentage point increase in the saving rate. In The average TFP growth rates for SSA and low- Kenya, the youth dependency ratio declined by income countries in the 2000s were 1.5 and 1.4, 25 percentage points between the 1980s and the respectively. Kenya’s TFP growth was negative up 2000s, while the saving rate increased by only 0.3 to 2003, and then increased to almost 2 percent percentage points (Figure 3.7). before it turned negative again after 2008. This inconsistency is a result of the high youth Kenya’s ambitious growth target of around 7 unemployment rate in Kenya and in SSA in percent per year requires much higher savings general. What ultimately matters for savings and investment. To achieve the growth target, is the effective youth dependency ratio, which savings and investment would need to more than takes into consideration whether the job market double. Only China and a few other resource-rich entrants are actually employed. Kenya suffers countries have achieved such results. Although from high youth unemployment: new labor 64 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Box 3.1: Determinants of Savings Although a strong positive relationship has been found between savings and income growth, the causality between the two runs both ways. There are two major arguments in the controversy. The first argument is causality between savings and growth; the second argument is how income growth affects savings. Loayza, Schmidt-Hebbel, and Serven (2000) find that a 1 percentage point increase in income growth increases the saving rate by roughly the same amount. Moreover, the same authors show that this relationship is stronger in low- and middle-income economies. In contrast, Rodrik (2000) concludes that a permanent increase in the saving rate induces a temporal increase in output growth, whereas a permanent increase in income growth is followed by a permanent increase in the saving rate. This argument on the causality has profound policy implications. If saving causes growth, savings-enhancing policies are likely to induce growth. If the direction is opposite, such policies may fail to promote permanent growth. These policies may promote growth in the short run by fueling investment through savings. In the long run, however, they may fail to realize permanent growth because returns to capital diminish and savings itself does not affect total factor productivity, that is, the long-term determinant of economic growth. Demographic Changes Demographics are another strong determinant of savings. The life-cycle theory predicts that savings follow a hump-shaped pattern, that is, young workers dis-save against their future income, middle-aged workers save for their retirement, and the elderly dis-save upon their retirement (Modigliani 1970). Therefore, demographic changes have significant impacts on household saving patterns. The microeconomic and macroeconomic literature confirms that rises in youth and old dependency ratios tend to lower the savings rate. Macroeconomic Stability Macroeconomic stability has been found to be among the most important policy determinants of savings. The stability of prices is the main parameter of macroeconomic stability that influences savings. Balassa (1986) claims that maintaining low and stable inflation encourages saving. Kenya’s decade-average inflation rate declined from 17.4 percent in the 1990s to 10.9 percent in the 2000s. This macroeconomic stabilization could be accompanied by an increase in the saving rate. Nevertheless, price volatility has risen since 2010, which in turn discourages saving. Financial Development Financial development leads to more efficient domestic resource mobilization, but has two opposite effects on savings. As the banking sector grows, individuals have more opportunities to save, although at the same time they are able to borrow more, which in turn leads to dis-saving. Kenya’s financial sector has developed rapidly over the past years, with the share of population with access to finance increasing from 69 to 75 percent (of which access to formal prudential banking rose from 22 to 33 percent) (IMF 2014). It seems that the increased access has supported saving, as the number of customers and deposits rose faster—157 and 92 percent, respectfully—than credit to households (84 percent growth) between 2009 and 2013. Urbanization Urbanization has been found to influence savings, as it depicts an increase in income stability. As people move from rural to urban areas, their income becomes less volatile and uncertain, which in turn has been found to lead to lower savings. Rural households are expected to save a larger portion of their income for the precautionary motive (for example, in case of a poor harvest), while nonfarming households tend to save less because their income is more predictable. Kenya has witnessed a trend of shift in population from its rural to urban areas (see chapter 1), which possibly contributed negatively to saving rates. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 65 Fiscal Policy Empirical studies have shown that an increase in public savings can lead to an increase in household savings in the short and long runs. The permanent income hypothesis predicts that a change in the timing of taxation does not affect household consumption behavior. Given a sequence of public expenditures, it does not matter when the government raises taxes to finance these expenditures. Therefore, the theory predicts that change in public savings fully crowds out private savings, leaving national savings unchanged (Ricardian equivalence). However, most empirical studies show that public savings only partially crowd out private savings. Because the private sector needs some time to adjust its saving behavior, the impact of public savings on household savings is more pronounced in the short run. This situation suggests that public savings can be an effective policy instrument to raise savings. Figure 3.7: Lower youth dependency ratio associated with Figure 3.8: Kenya’s effective youth dependency ratio is higher savings, average, 1980–2013 much lower, 1975–2010 50 1.4 45 % of total working age popula tion 40 1.2 Savings rate (% of GNDI) 35 30 1.0 25 20 0.8 Kenya 15 10 0.6 5 0 0.4 0 20 40 60 80 100 120 1975 1980 1985 1990 1995 2000 2005 2010 Youth dependency ratio (% of working age population) Youth dependency ratio Eff ective youth dependency Source: World Bank World Development Indicators. Source: World Bank staff calculations based on United Nations World Note: GNDI = gross national disposable income. Population Prospects and Omolo 2010. market entrants, those around age 20 years, face In countries with a low real interest rate, an unemployment rate of around 35 percent.69 increasing the real interest rate typically raises Comparison with the peer group shows that savings. Savers, be it households or firms, are Kenya’s youth unemployment rate (for those attracted to save by the real returns they expect ages 15–24) is the highest, at 17 percent. to get on their deposits. Negative real deposit Consequently, even as the young-age population interest rates in general discourage household, bulge entered the labor market, the effective but also corporate, saving in the banking system. dependency on income earners did not decline Although Kenyan banks offer relatively high much (Figure 3.8). One positive contribution in nominal rates on deposits, the real deposit this process may be the fact that an increasing interest rates in Kenya have been negative share of youth is in school rather than searching for most of the past decade (Figure 3.9). The for jobs. Although this makes them dependents, variability of the real rate comes as a result of the at least parents are spending on education, which wide fluctuation of inflation. Nevertheless, even is considered consumption, although it is in fact when inflation has been low and stable (as it has investment in human capital. been since 2012), the real deposit rate has been 69 UNDP (2013). 66 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M low. The high interest spread, which averaged to increase. The consequent decline in the youth more than 10 percent over the past decade, has dependency ratio should lead to more saving, but meant that nominal deposit rates are low. this will only happen if those entering the labor market are actually able to find a job. Otherwise, Figure 3.9: Real deposit savings rates have been negative for much of the past decade Kenya’s youth dependency ratio will continue to 20 be high; that is, the number of savers will not 15 increase. For example, the youth dependency 10 ratio is projected to fall to 0.6 by 2030, which is Percent 5 0 higher than the one Vietnam observed in 1996. -5 If the youth unemployment rate remains high, -10 Kenya’s effective youth dependency rate will go -15 -20 down only to 0.7. As the youth dependency ratio -25 is a critical determinant of savings, reducing the -30 effective youth dependency ratio through jobs 1995 1997 1999 2001 2003 2005 2007 2009 2011 Real deposit interest rate Nominal deposit interest rate holds tremendous potential for enhancing saving Source: World Bank World Development Indicators. rates. Chapter 2 discusses policies that could promote job creation, in particular for youth. What Can Be Done to Expand Resources for Investment? Maintaining macroeconomic stability and low inflation could also reap high benefits in terms As the analysis so far illustrates, saving behaviors of savings. Inflation has been volatile and high are determined by various policy factors. From throughout the past decade, during which the the perspective of a policy maker in Kenya, it rate on deposits would be positive, which should would be useful to know in which way the wind incentivize individuals to save more. However, may blow, that is, to be aware of the trends in any policy shocks, for example, inflation spikes each determinant of savings and what could be that in turn lower real deposit rates, may lead to done to promote saving. loss of confidence and reduce savings. To begin, economic policies that would stabilize growth will also promote savings. In contrast to To turn real deposit rates positive, policy the volatile period since 2008, Kenya’s economic makers should focus on reducing the interest growth is expected to be high and sustained over rate spread, in addition to maintaining low and the medium term. Hence, following years of stable inflation. Two possible causes of the high uncertainty and shocks, the stable growth outlook interest spread seem to stand out: the lack of should boost the confidence of households and competitiveness in the banking sector and the firms to save. high cost of financial intermediation. Kenya’s financial market is highly segmented between Demographic trends also support a move large banks and small cooperatives. Large banks away from consumption and toward saving. have the market power to maintain a wide spread Although below the peak of the 1970s, fertility at the expense of borrowers and depositors. in Kenya is still high, at around 3. The expected Measures to expand the number of players or drop in fertility over the next decade will bring a products in the sector (for example, M-Shwari) “demographic dividend” for Kenya, whereby the and limit the market power of the largest bank share of the working-age population is expected FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 67 could result in a lower interest rate spread and Changes in pension policy can also affect saving higher real deposit rates. The first step to this behavior. Although evidence from other countries end would be to develop a strategy for promoting reveals that the expansion of pension schemes competition in the banking sector. does not necessarily lead to more savings, there is strong evidence that pay-as-you-go pension The high interest rate spread points to structural systems reduce national saving rates while fully- deficiencies in the business environment for funded contributions increase national savings. banking, which also need to be addressed The main challenges that Kenya’s pension system to reduce the interest rate spread. The main faces are contingent (unfunded) liabilities and priority in this regard is to reduce information inadequate corporate governance. Moving from asymmetries and risks by improving the credit a Pay As You Go to a funded public pension information systems, regulatory framework scheme recognizes government liabilities and (collateral, creditor, and insolvency laws), land will create a significant pool of assets. The RBA and company registries and titling, and process is working to increase savings amongst informal of taking and realizing collateral. sector workers and on improving the governance of the NSSF.” Increasing public savings, which is a direct outcome of fiscal policy, can raise overall savings Finally, it matters less where savings are in the short and long runs. Policy makers can generated compared with how they are stimulate savings through lowering recurrent channeled. In some countries, such as the East expenditure or raising tax revenue. Kenya’s public Asian Tigers, households were the primary savings is low (at 1 percent) compared with the generators of savings during high-growth peer economies and the continent in general episodes, while in others, such as resource rich (the SSA average is 4 percent). In this context, countries, the state (public budget) played this raising public savings could be an effective tool to role. What ultimately matters for economic shift resources from consumption to investment, growth is that savings is invested in productive especially through reduction in recurrent assets. For public saving, the main priority of the expenditure. Lowering the overall public sector government should be increasing infrastructure wage bill would not only strengthen fiscal investment (chapter 1 raises the risks of shifting sustainability, but also promote savings. It should resources toward recurrent spending). In the case be noted, however, that the fiscal classification of households, the onus is on how to mobilize of expenditures is not fully consistent with their savings in savings and credit cooperative the definitions of savings and investment. For organizations (SACCOs) to good small business example, some recurrent expenditures, for investment opportunities, including in the jua example, teacher salaries, are in fact investment in kali (chapter 2 looks into the performance and human capital. Hence, redistribution of spending challenges of Kenya’s informal entrepreneurs). should take into consideration the social returns For corporations and enterprises, the main to such (recurrent) spending, relative to capital objective should be to create the right conditions spending on infrastructure. that would allow them to invest in expanding their business (chapter 4 discusses enterprise developments in manufacturing and services). 68 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Insights from Behavioral Economics on Savings Behavioral Economics to perpetuate the status quo, so that if people have not had bank accounts, Standard economics says that a primary they cannot save even when they have school graduate will continue his or her access to bank accounts (Samuelson education if the benefits from further and Zeckhauseer 1988). education exceed the costs of education. It also says that a poacher will stop killing How Can Saving Behaviors Change? elephants if he or she can earn more from safari tourism than from smuggling Saving is crucial not only for economic ivory. In reality, the secondary school growth for a country, but also for poverty enrollment rate in Kenya is still low and alleviation at the household level. SPOTLIGHT 3 smuggling ivory remains a big issue in Saving allows households to accumulate many countries in SSA. People do not capital to invest in education, pay for always make rational decisions. health care, and provide seed funding for entrepreneurship. Failure to save Behavioral economics seeks to explain under poverty could lead to dropping why people make irrational decisions out of school, having limited access to from the perspectives of psychology medical services, and missing out on and economics. This spotlight presents business opportunities. insights from behavioral economics on saving behavior, that is, why people do Low savings among the poor are often a not save even if the long-term benefits result of the way people think rather than from savings exceed the short-term lack of income. The World Development benefits from consumption. Report 2015 shows two examples from Kenya that illustrate this. An experiment Behavioral economics encompasses with poor households that were in need theories such as (i) hyperbolic of preventive health products, such discounting, which suggests that as insecticide-treated mosquito nets, people tend to discount the future and showed that providing people with a overweigh the present when making lockable metal box to save money or a decisions (Thaler 1981; Laibson 1997); dedicated savings account for health (ii) loss aversion, which implies that emergencies can increase savings; in this people react differently to losses and case, investment in these products rose gains of equal size, such that people by 66–75 percent. A second example is feel losses more keenly than gains from rural Kenya, where poor farmers of equal magnitude (Tversky and typically use less than the optimal Kahneman 1992); and (iii) the tendency amount of fertilizer because of lack of FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 69 funds at the time of applying fertilizer. such as rotating saving and credit However, in an experiment, an option associations or even keeping money to pre-purchase fertilizer at the end of in the house, are the preferred way the harvest season (when farmers have to save. Among the formal channels, money) was introduced, which in turn M-Pesa (mobile money) is the prevalent resulted in higher fertilizer use than if instrument for saving in financial the cost of fertilizer was subsidized by 50 assets—for a quarter of the surveyed percent at the time of applying fertilizer. poor, it was their preferred instrument— followed by saving in a bank. In addition, lack of access to banking has been a challenge, especially for However, this challenge does not mean the rural poor. Households without that the high cost of banking services access to banks often save at home causes poor households not to save. or save in nonmonetary assets, like Women in rural Kenya, having faced cows. In either case, interest cannot limited access to saving instruments, be accrued. Financial institutions are have the desire to save. Dupas and reluctant to provide services to the Robinson (2013a) provided access to SPOTLIGHT 3 rural poor because of the high cost. non-interest-bearing bank accounts The number of individuals in the world to female market vendors in rural without access to banks is estimated Kenya. Despite high withdrawal fees, a at two billion to three billion (Karlan substantial share of the women used and Morduch 2009). In Kenya, three- the bank accounts, saved, and increased quarters of the rural labor force saved their productive investment and private in the past year, the majority of whom expenditures. Once the main barriers saved in a saving club, such as a SACCO to saving are removed, then savings are (Table 3.1). However, most households, expected to increase. and especially poor ones, save primarily in nonfinancial assets such as a house, Formal and informal commitments to land, or livestock. Even for those who save, such as at a financial institution save in financial instruments, only 9 or through a rotating savings and credit percent of households’ assets are in association, can help households to save. formal financial instruments, according Dupas and Robinson (2013b) performed to the Kenya Financial Diaries survey another experiment in rural Kenya by in 2014 (Zollman 2014). For many poor providing access to four saving schemes households, informal saving channels, with different levels of commitment. The Table 3.1: Saving behavior by groups, individuals ages 15+ (%) Income, bottom Income, top In the past year, did you… Total Rural 40% 60% Save any money 76 75 70 80 Save at a financial institution 30 28 13 42 Save using a savings club 40 40 38 42 Source: Global Financial Inclusion (Findex) database 2014. 70 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M schemes ranged from a simple lockbox to policy reforms aimed at strengthening a health pot with a binding commitment capital adequacy, broadening the rules to save and the opportunity to withdraw for investment in land and buildings, in case of emergency. The higher the improving credit information sharing, level of commitment, the more the rural allowing prudentially supervised Kenyans saved. SACCOs to access the payment system if they meet the operational requirements, SACCOs are another important segment and expanding the liquidity management of the financial sector that helps channel system. Some of these reforms are savings to investment, particularly in expected to result in consolidation of the rural areas. More than 60 percent of SACCO industry, which in turn should lead membership in deposit-taking SACCOs to greater efficiency and lower lending comes from farmer- and community- rates (unlike the lending rates in the based SACCOs, although their assets banking sector, SACCOs have not lowered are only 16 percent of the total. Hence, their lending rates since the 2012 hike in SACCOs play an important role in response to the rising inflation). mobilizing savings in rural Kenya and SPOTLIGHT 3 channeling it to investment projects at Recent technological advances the local (community or county) level. have enabled saving schemes with Although membership and assets (and dramatically low transaction costs. asset quality) have been growing rapidly M-Pesa, the largest and most rapidly (Table 3.2), many policy initiatives growing mobile money platform in low- are being discussed by the SACCO and middle-income countries, provides Society Regulatory Authority (SASRA) saving schemes through M-Kesho that would enhance the efficiency and M-Shwari in Kenya. M-Shwari’s of the financial intermediation of customer savings have reached more SACCOs, by expanding and improving than K Sh 24 billion in just over two years the quality of investments. In its 2013 since the launch of the mobile phone– supervision report,70 SASRA outlined based bank account. M-Shwari attracts Table 3.2: The SACCO industry is growing rapidly Indicator 2010 2011 2012 2013 Number of active SACCOs a 1,821 1,954 1,989 1,995 Of which deposit taking 215 Membership 3,294,829 Total assets (K Sh million) 261,144 248,765 293,827 335,437 Member deposits (K Sh million) 157,540 180,003 213,080 240,805 Loans and advances (K Sh million) 157,926 186,149 221,554 251,879 Total capital (K Sh million) 20,115 21,324 25,297 Turnover (K Sh million) 27,721 31,464 37,286 43,271 Source: SACCO Society Regulatory Authority. Note: SACCO = savings and credit cooperative organization. a. SACCOs that filed their audited financial statements with the commissioner for cooperative development as a legal requirement. http://www.sasra.go.ke/index.php/resources/publications?download=64:sacco-supervision-annual-report-2013. 70 FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 71 K Sh 200 million in deposits daily from Box 3.2. Kenya’s Financial Inclusion more than six million users.71 Moreover, Kenya’s financial inclusion is relatively mobile platforms are now connected high, which in turn encourages formal to SACCOs, which increasingly are saving. The banking sector is well attracting savings, especially from the developed in geographical coverage and youth, and contributing to realization products. The use of banking accounts of the savings and investment target is widespread—more than in any peer of Vision 2030. The World Council of country—with almost a quarter of Credit Unions nominated 18 Kenyan the population saving at a financial SACCOs as 20 leading SACCOs in SSA. institution (figure 3.10). Penetration of Increasing use of mobile technology has retail banking has accelerated in recent placed Kenya’s SACCOs as the highest years: the number of deposit accounts growth achievers and the leaders of the in commercial banks increased from 4.7 million in 2007 to 21.1 million in industry in Africa (box 3.2). September 2013. In addition, Kenya’s mobile revolution has led to the This research shows that as long as establishment of mobile saving accounts access to and design of saving schemes SPOTLIGHT 3 (M-Shwari), which have recorded a steep are improved, people can change their increase since their inception. saving behavior. The key finding from the research is that the poor attempt to make productive use of their resources Figure 3.10: Financial inclusion in Kenya is high relative to peer countries in the environment in which they 60 operate. Making the environment such 50 that incentives are offered to remind people to save for investment in items 40 with high payoffs, be it health, education, 30 or improving agricultural yields, can lead to greater savings. Lowering the 20 transaction cost for saving and having a 10 diverse portfolio of saving opportunities is another way to stimulate saving, in 0 particular for the poorest segments of Source: Global Findex database. the population. Account at a financial institution (% age 15+) Saved at a financial institution (% age 15+) Zollman (2014). 71 72 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M CHAPTER 4 MANUFACTURING OR SERVICES: WHERE DOES THE KEY TO RAPID GROWTH LIE? Introduction sustainable growth of firms. Many firms attempt to export, which is good, but most of them Vision 2030’s target of rapid and sustained fail. Large firms do not seem to be expanding, growth is envisaged to rest on the development especially in producing higher complexity of agriculture, manufacturing, and services. goods. This situation is confirmed by export However, as chapter 1 points out, Kenya’s growth data, industrial firm surveys, and data from the model since 2007—when Vision 2030 was put export processing zones. The high dispersion in forward—has been led predominantly by a few productivity within the same sector and the entry booming services, fueled largely by growing of lower productivity firms point to constraints to private consumption and rising exports. Growth of firm creation and growth. High costs of production, agriculture and manufacturing has been sluggish, in particular energy, are a major obstacle, which with sporadic success stories. Moving forward, is why more firms are being established in sectors the key question is to understand why the with low energy intensity. Other constraints, such manufacturing sector has underperformed and as the weak business environment, are discussed whether there is potential to accelerate further in chapter 5. the growth of services. Although it is one of the six priority sectors in Vision 2030, agriculture is Services have followed an independent not the focus of this chapter. Chapter 2 looks at development path and hold the potential to the importance and potential of agriculture from achieve Kenya’s objectives for growth and the perspective of poverty reduction. job creation. Services have been expanding independently of manufacturing, unlike in peer The high share of services is not what is peculiar countries where services have typically been about Kenya or Sub-Saharan Africa (SSA) in general; it is the low share of manufacturing that pulled by manufacturing. In addition to the differentiates Kenya from other fast-growing, boost from rising domestic demand, services low- and middle-income economies. The growth exports have also been booming and will soon story of East Asia and a few other successful take over goods exports. New entrants in the economies has been that (export-oriented) services sector are more productive than existing industrialization was the main engine of growth. ones, which adds to within-sector productivity, Kenya and SSA in general have not followed the although, as with manufacturing, dispersion is same path; instead, services have been the main high. Innovation seems to be high, although the driver of growth, while manufacturing has been type of innovation that Kenyan firms engage stagnant or declining. in is not as productivity-enhancing as in peer countries. In Kenya, innovation comprises mainly Manufacturing is being held back. The paradox marginal improvements, while investment in in the manufacturing sector is that high research and development, for example, is lower entrepreneurial dynamism does not lead to than in the peers. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 73 Evolving Roles of Industrialization and The disappointing outcomes in Africa’s Services as Poles of Growth manufacturing sector are explained by high labor and external costs, which are also found Episodes of growth similar to what Vision 2030 in the case of Kenya, although they may in some sets for Kenya have been rare in the economic cases be offset by high productivity. Wages in history of the past six decades (post World War the manufacturing sector in Africa are relatively II). The Growth Report (Growth Commission high compared with wages in other countries at 2009) found only 13 episodes where countries a similar level of development. For example, the managed to sustain an average 7 percent median labor cost in manufacturing in Kenya is growth for 25 years or more, and only one of almost four times as high (although productivity them (Botswana) is from the African continent. is even more so) as in Bangladesh, which has With few exceptions, including Botswana, which a similar income per capita (Gelb, Meyer, and transformed natural resources into rapid growth, Ramachandran 2013). Moreover, labor costs the main engine of growth in the majority of cases seem to be increasing: their share in the value was (export-oriented) industrialization. This has added of manufacturing firms increased between been the story behind the growth in East Asia. 2009 and 2013, and so did their share in total sales for listed manufacturing firms at the Nairobi African countries have not been able to replicate Stock Exchange. Part of this is explained by the the East Asian successes, and the primary fact that Africa’s, and Kenya’s, manufacturing reason lies in Africa’s difficulties in developing sector is characterized by dualism: a few formal manufacturing. Many countries in SSA, including (high productivity) firms coexist with many Kenya, have witnessed an economic revival in informal (low productivity) firms. The formal this century that has been driven largely by the productive firms face a steeper labor cost services economy. However, the huge importance curve. The other reason, which is also very of services is not exceptional to Africa. The much relevant for Kenya, is that external costs, average share of services in low- and middle- related to electricity, logistics, transport, and income economies is only slightly lower than that corruption, are higher than elsewhere (Eifert, in Kenya. What is exceptional for Kenya, and the Gelb, and Ramachandran 2008). As shown in a rest of SSA, is the low share of manufacturing recent study by Iacovone, Ramachandran, and and high share of agriculture (and extractives Schmidt (2014), infrastructure gaps and weak the in some SSA countries) in the economy. Africa’s business environment raise external costs and manufacturing sector remains underdeveloped make it difficult for firms to grow. and very few countries have managed to diversify into export-oriented manufacturing Nevertheless, recent insights on the drivers that goes beyond the processing of raw materials of long-term growth show that ultimately (Gelb, Meyer, and Ramachandran 2013). Like productivity drives economic development, in the peers in SSA, the agricultural sector in irrespective of which sector is behind the Kenya remains critical to the economy, and productivity growth. Productivity growth is the manufacturing is closely linked to it. Indeed, food engine of sustainable development and can production accounted for 32 percent of Kenya’s generally be driven by within-sector productivity total manufacturing output in 2013. gains or structural movement of labor and other resources across sectors. There is ample evidence that shows that market competition boosts firm- 74 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M level (within-sector) productivity. Kenya’s product at which the share of industry (in total gross markets could be made more competitive. domestic product (GDP) or employment) peaks The Organisation for Economic Co-operation is happening earlier in the development process. and Development’s (OECD’s) Product Market For example, in 1988, for the world as a whole, the Regulations indicators and the Bertelsmann peak share of manufacturing was 30 percent and Foundation Transformation’s market-based attained at a per capita GDP level of US$21,700. competition sub-index indicate that competition By 2010, the peak share of manufacturing had rules in Kenya are weaker compared with those fallen to 21 percent and at a per capita GDP level in middle-income countries such as Brazil, China, of US$12,200. This change implies that services India, and South Africa. Productivity gains can are increasingly becoming a lead engine of also occur through shifts of labor from lower growth. Interestingly, Kenya and its peer group to higher productivity sectors, even if sector seem to be showing a de-industrialization trend productivity remains flat. at even lower levels of development: the share of industry rose in the 1990s, but has fallen since In Africa, and to a lesser extent in Kenya, (Figure 4.1). labor has been moving to lower productivity sectors. MacMillan and Rodrik (2012) examined Figure 4.1: Declining share of manufacturing in Kenya’s peer group productivity trends in selected African economies since 1990 and found productivity improvements at the sector level, yet at the aggregate level these Kenya were offset by a large movement of labor toward lower productivity activities (growth-reducing structural change). Kenya’s story is a bit more positive, as the economy has achieved within- sector productivity gains and a shift of labor to more productive sectors since 1990, although in recent years the bulk of labor market entrants has gone into the informal trade and hospitality Source: World Bank World Development Indicators. sector (chapter 2 analyzes this in greater detail). Note: GDP = gross domestic product; PPP = purchasing power parity. In contrast, East Asian economies witnessed within-sector productivity growth and “growth- Nevertheless, experiences from elsewhere can enhancing” structural change during their high- only be of limited help to policy makers devising growth episodes. a country’s development strategy; policy reforms and growth poles are highly contextual. For many low- and middle-income countries, Kenya’s context is such that manufacturing and including Kenya, services have been a primary services hold the potential to contribute to driver behind productivity increase and growth faster and sustained growth, as individual firms over the past three decades. The past three or subsectors, be it East African Breweries or decades unmistakably point to a change in the Safaricom, have proven. Even in agriculture, development path, a process defined by Rodrik examples such as cut flower exports demonstrate (2013) as an early relative decline of industry, Kenya’s potential. The remainder of this chapter or premature de-industrialization. Basically, examines the performance of manufacturing and the economic data show that the point of time services from macro and micro lenses. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 75 Economic Complexity of Kenya’s Kenya’s Manufacturing Sector Is Diversifying Manufacturing Sector Looking at Kenya’s export structure beginning Economic complexity and income per capita in the early 1990s, manufacturing showed display high correlation; hence, countries positive signs of diversification toward higher attempt to increase their complexity. Economic complexity (value) goods. Kenya has traditionally complexity reflects the amount of knowledge been an agricultural commodity exporter, and that is embedded in the productive capacity of the share of agriculture in total exports has an economy (Hausmann, Hwang, and Rodrik increased since 2009 because of the rapid growth 2007). Economic complexity is a measure of the of tea and horticultural exports. Nevertheless, production capacity of a country’s manufacturing the non-agriculture part of exports showed sector: which products it can or cannot signs of diversification and movement toward produce (box 4.1). Rich countries, excluding more complex products (including beer, plastic resource-endowed ones, display high economic packaging, and pharmaceutical products) until complexity, while poor countries generally have the mid-2000s. lower economic complexity. The manufacturing sector in Kenya is more More important, economic complexity can diversified than that in other countries with be a driver of future economic performance. similar income per capita. Kenya’s concentration Countries that increase their complexity tend to among manufacturing exports (measured by the grow faster in subsequent periods (Hausmann, Herfindahl Index) is relatively low (Figure 4.2), Hwang, and Rodrik 2007). Countries that have which implies a diversified manufacturing base.72 relatively higher complexity, given their level A World Bank (2014)73 report found that Kenya’s of income, tend to grow faster than those manufacturing exports are highly diversified countries that are “too rich” given their level of at the product level, relative to other African economic complexity. countries and other export powerhouses, such as Box 4.1. How is economic complexity measured and what does it (not) represent? The economic complexity index (ECI) is measured using disaggregated export data (from UN COMTRADE), which are used as a proxy for the industrial structure of the entire economy. All products are mapped in terms of their linkages based on how similar they are in their complexity. The most complex products with the largest number of connections are located in the center or core of the network. There are many products in the periphery that are only weakly connected to other products. The ECI is not a measure of export diversification or trade openness. The aim of the ECI is not to prescribe specific sectors that a country should develop, but to serve as a big picture examination of the potential of the economy. Data in the Atlas of Economic Complexity, which captures more than 120 countries, can be found at http://atlas.cid.harvard.edu/ Sources: Lall 2000; Hidalgo and Hausmann 2009. 72 Strictly speaking, the Herfindahl Hirschman Index is meant to measure firm concentration in a market and thus inform the assessment of competition in the marketplace. Nevertheless, the index is used extensively to measure the degree of concentration in other spheres: product exports and imports, trade partners, etc. In this particular case, it measures Kenya’s concentration in manufacturing exports. 73 Manufacturing export competitiveness in Kenya report by Farole and Mukim (2013). 76 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Vietnam. Diversification comes with benefits for complexity since 2005. This finding is supported long-term growth, as it is associated with greater by World Bank (2014), which found that although macroeconomic stability, lower vulnerability Kenya’s (export-oriented) manufacturing sector is to shocks, and lower terms of trade volatility dynamic, (export) product survival is low. (Lederman and Maloney 2012). It seems that although Kenya has a few firms Figure 4.2: Export concentration trend that produce complex products, the majority of production is of low complexity. Kenya’s top four exports by total value are among the least complex goods traded globally. Nevertheless, several manufacturing sectors, for example apparel or iron and steel, in which Kenya is strong, offer potential for expanding production to products that are similar in complexity to what is already produced. Hence, the well performing firms in this sector have room to grow and expand their product portfolio. Source: PRMED calculations using UN-Comtrade data. The Manufacturing Sector from a Micro But Overall Economic Complexity Is Neither Lens High nor Increasing Manufacturing firms in Kenya are relatively old, Kenya’s economic complexity, as measured by but entrepreneurship seems to have picked up its exports structure, is relatively low and not over the past decade. According to the Kenya increasing. Kenya’s economic complexity is higher National Bureau of Statistics’ industrial firms than that of its peers in SSA as well as Bangladesh census (box 4.2), two-thirds of firms have been and Cambodia. But Kenya’s export complexity is operating for more than 10 years. One-fifth of lower than that of its better performing peers, the firms are up to three years old, which allows such as India, Pakistan, and Vietnam. The three peers from East Asia (China, Indonesia, and comparing established versus entrant firms. Thailand) had much higher complexity in 1995 Although no other census of this type has been (first year of available data) compared with conducted in the past three decades, the data Kenya’s in 2012. More importantly, Kenya’s point to intensified firm establishment after 2002: manufacturing production capabilities have 50 percent more firms, of those still in existence stagnated. Kenya is one of the few countries in 2010, were established between 2002 and that have recorded a decrease in economic 2005 than between 1998 and 2001. Box 4.2: Data source for analysis of manufacturing firms The analysis in this section draws on the 2010 Census of Industrial Production conducted by the Kenya National Bureau of Statistics (KNBS). The data set covers 2,252 firms and 109 International Standard Industrial Classification four-digit product categories in manufacturing, mining, electricity, and water supply. These firms accounted for 324,841 jobs (including proprietors) in 2009, which implies that the data capture formal and informal workers in the sector, given that total formal wage employment in manufacturing was 270,000 in 2009 according to KNBS. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 77 Another data source shows that Kenya saw The majority of Kenya’s manufacturing firms are a surge in (formal) entrepreneurial activity located in low value-added sectors. Most of the following the measures to open up the economy large firms, in terms of number of workers, are in in the early 2000s. Between 2004 and 2008, sectors with low value added. The largest share of the number of registered new (limited liability) employment in manufacturing is in food products firms rose threefold, from less than 7,000 new (41 percent of total employment), followed firms in 2004 to 18,000 in 2008 (Figure 4.3). by textiles (8 percent), wearing apparel (5.7 Only Rwanda, among the peer group countries, percent), wood and wood/cork products except witnessed a similar trend, which accelerated furniture (3.7 percent), and leather and related even further after 2008: by 2012 the number of products (2.5 percent). In contrast, the highest newly registered firms was five times higher than value-added sectors, such as manufacture of coke in 2008. Data are not available for Kenya post and refined petroleum products, employ very 2008, although a similar trend is not expected, few workers (0.01 percent of total employment). given the relatively weaker and more volatile Sectors such as beverages and tobacco and repair performance of the economy post 2008. and installation of machinery and equipment also have a relatively high level of value added per Figure 4.3: Kenya has witnessed rapid growth of formal business startups worker, yet low employment numbers. (2004 = 100) 450 400 Kenya’s manufacturing firms tend to be relatively Burkina Faso capital intensive. In most industrial sectors, 350 Ghana compensation per worker, that is labor cost, is well 300 India below the value added per worker, which implies 250 Kenya Pakistan high capital intensity. In food products—a low 200 Senegal value-added sector—compensation per worker 150 Uganda is 40 percent of value added per worker (Figure 100 4.4). This signals that formal manufacturing firms in Kenya tend to be capital intensive, a trend that 2004 2005 2006 2007 2008 2009 2010 2011 2012 is supported by the evidence of high costs of Source: World Bank Entrepreneurship Snapshot. labor (relatively high salaries). Figure 4.4: Kenyan manufacturing firms tend to be capital intensive 1.00 Size=# employees 1.60 Millions 0.90 Ksh Millions 1.40 0.80 Repaid & Install Coke & Ref Petrol Mach & Eqm 1.20 0.70 Compensation employee Electr, gas, steam, Compensation employee 0.60 1.00 A/C Paper Non-metal mineral 0.50 Beverages 0.80 Repaid & Install Tobacco Mach & Eqm 0.40 Size=# employees 0.60 0.30 0.40 0.20 Basic metals 0.10 0.20 Food prod Waste collec tion 0.00 0.00 -1 0 1 2 3 4 -10 10 30 50 70 Ksh Millions VA per employee VA per employee Millions Sources: Kenya National Bureau of Statistics; World Bank. 78 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M The typical catching-up effect in firm among firms in these sectors. The dispersion is productivity within a sector seems to be limited highest in sectors with relatively fewer workers, and productivity varies between and within such as coke production. sectors in Kenya. Within sectors there is a wide dispersion of firms in terms of productivity, The manufacturing sector shows relatively high whereas typically firms within a sector would be churning, with 20 percent of all firms being learning from each other and converge toward entrants, although the newcomers do not seem the productivity level of the more successful to be raising overall productivity. Firms that ones. Firm-level performance within the large were less than three years old had on average a sectors, such as food products, show great lower value added, output, and compensation dispersion, and large firms are not necessarily per worker than the established firms (Figure on the productivity frontier. This pattern is 4.6). This is true even if comparing entrants with consistent with the findings by Gelb, Meyer, and established firms at the regional level (Nairobi, Ramachandran (2014) of slow convergence of Central, Eastern, and Rift Valley). However, new productivity in SSA.74 They find that SSA’s formal entrants added more new jobs compared with manufacturing sector appears to be dominated old firms between 2009 and 2010. by a limited number of larger firms with higher Figure 4.6: Entrant firms are less productive labor productivity that coexists with a long tail of (value added, output, and compensation per worker, for entrants ages 0-3 years and established firms) lower-productivity firms. 3,500,000 3,000,000 Capital intensity varies significantly across firms 2,500,000 within the same sector. This holds true across sectors, which is expected, but also among firms 2,000,000 K Sh within the same sector (Figure 4.5). Looking at 1,500,000 the largest sectors, food production is relatively 1,000,000 capital intensive, while garments are on the low 500,000 side. At the same time, these two sectors have 0 relatively low dispersion in the capital-to-labor Entrant Established ratio, which may imply productivity convergence Value added per worker Output per worker Compensation per worker Sources: Kenya National Bureau of Statistics; World Bank. Figure 4.5: Capital-to-labor ratio dispersion is high across most sectors, 2010 10,000,000,000 Minimum-to-maximum range 1,000,000,000 Average capital-to-labor ratio, by sector Overall average 100,000,000 10,000,000 K Sh per worker 1,000,000 100,000 10,000 1,000 100 10 1 Source: Census of Industrial Production. See Gelb, Meyer, and Ramachandran (2014). They find that, “to a large extent, SSA’s formal manufacturing sector appears to be dominated 74 by a limited number of larger firms with higher labor productivity that coexists with a long tail of lower-productivity firms.” FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 79 Higher agglomeration attracts new firms. As infrastructure development, this trend may be shown in figure 4.7, Nairobi has the highest explained in part by differences in the regulations concentration of established firms and related for starting a business. According to the 2012 employment (about 40 percent of both), and subnational Doing Business report, Nairobi had continues to attract significantly more new the fifth most business-friendly regulations for entrants (about 45 percent of new firms) and starting a business. new jobs (just over 30 percent). Other regions with high concentrations of established firms In terms of the inputs (other than raw materials) are the more urban central region and Rift that Kenyan manufacturing firms use, transport, Valley, where the new entrants have provided real estate, and energy matter the most. a greater proportion of new jobs (21 and 17 Overall, energy is the second largest cost for percent, respectively). Unfortunately, historically manufacturing firms (excluding raw material and marginalized regions have low concentrations labor costs). There is some variation depending of established firms and do not appear to be on the sector, but for food products, which is the attracting significant numbers of new firms or largest subsector, energy is also one of the largest employment. In addition to differing levels of (non-raw materials) costs (Table 4.1). The energy Figure 4.7: Nairobi attracts new firms but more jobs are created in less urbanized areas Entrants (0-3 years) Established Firms (>3 Years) 50 45 45 40 40 35 35 30 30 Percent Percent 25 25 20 20 15 15 10 10 5 5 0 0 Central Coast Eastern Nairobi North Nyanza Rift Valley Western Central Coast Eastern Nairobi North Nyanza Rift Valley Western Eastern Eastern % of Firms % of Jobs % of Firms % of Jobs Sources: Kenya National Bureau of Statistics; World Bank. Table 4.1: Top three most important upstream sectors for each downstream sector Top-3 upstream sectors excluding materials Downstream sector (percent of total expenditure by downstream sector) Manufacturing of coke and refined Transportation (12 Repairs and maintenance (7 Mining and quarrying petroleum products (53 percent) percent) percent) Transportation and storage (18 Manufacturing Electricity (15 percent) Real estate (15 percent) percent) Manufacturing of coke and refined Electricity Electricity (22 percent) Real estate (6 percent) petroleum products (48 percent) Water, sewerage, and Transportation (16 percent) Electricity (14 percent) Financial services (14 percent) waste management Source: Census of Industrial Production. 80 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M dependency of manufacturing contrasts with accounted for more than half of the increase in that of many of the services sectors, in particular total exports. If the trends of the past four years the successful banking and telecommunications continue, services exports will overtake goods sectors. For example, a Kenyan spends less than 1 exports by 2015. dollar a year on charging a smartphone, and less than 20 dollars a year on using a laptop. It comes Services have a direct contribution to exports as as no surprise that four in 10 manufacturing firms well as an indirect contribution, serving as inputs in 2013 indicated energy as a major constraint to to production in other sectors. A new World Bank business (a slight improvement over 2007 when trade database (box 4.3) dissects the direct and more than half of firms found it to be a major indirect roles of services for the tradable part of constraint).75 Kenya’s industrial electricity tariff, the economy, which in turn allows measurement at more than US$0.30 per kilowatt hour, is higher of the value chain linkages between services and than the tariff in many SSA countries and the non- the rest of the economy. This includes forward Africa peers. Appendix table A1.1 shows for each linkages, the contribution of a particular sector manufacturing sector the decomposition of costs as an input to others sectors’ exports. (Box 4.3 across utilities and services sectors (not including presents the definitions and methodology.) materials).76 Gross exports of services, as calculated in Services as a Driver of Growth trade statistics, typically undervalue the total contribution (total value added) of services to a Services are the largest and most dynamic part country’s exports, and this is the case in Kenya. of Kenya’s economy. As chapter 1 points out, Services are embedded as inputs in exports domestic services have been behind the growth of manufactured and agriculture goods, while in domestic demand over the past decade. In the production of services does not necessarily addition, services exports have increased rapidly involve a significant input from the latter two. over the past decade. Exports of services have Hence once considering services used as inputs, been growing faster than goods exports (Figure the ratio between total value added exports to 4.8), and since 2005 services exports have gross exports is higher than one, and this is the Figure 4.8: Services exports are rising much faster than case for Kenya and all other peer countries except exports of goods (K Sh, millions) the Arab Republic of Egypt (Figure 4.9). This 600,000 characteristic—that services exports in value- 500,000 added terms once considering forward linkages tend to be greater than gross exports—does not 400,000 necessarily extend to the manufacturing sector; 300,000 its ratio is well below 1. 200,000 100,000 0 00 001 002 003 004 005 006 007 008 009 010 011 012 013 20 2 2 2 2 2 2 2 2 2 2 2 2 2 Goods Services Source: Central Bank of Kenya. 75 Kenya enterprise surveys 2007 and 2013 (www.enterprisesurveys.org). 76 Including materials would reduce the shares of all upstream sectors, as materials account for 36 percent of expenditure in mining and quarrying, 85 percent in manufacturing, 66 in electricity, and 60 percent in water supply. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 81 Box 4.3: Measuring the value added in exports a Exports (of goods and services) can be measured as the following: Gross exports. The transaction value of a sector’s exports, that is, what is published as exports in the balance of payments. Gross exports capture the value added embodied in the production of the export, as well as all (domestic and imported) intermediate inputs. Gross measures of trade statistics are registered in the balance of payments at the transaction value, that is, the price for the goods or services. For example, a business process outsourcing (BPO) company from India sells its services in Europe and India’s gross exports of BPO services capture the invoice price of those services. Direct value added of exports. This is a sector’s domestic value added embodied in its own exports, measured as gross exports excluding inputs. This measure captures the value-added contribution of a sector in the sector’s own exports. For example, an Indian BPO firm uses telecommunications services, from local providers and foreign owners of satellites, which are intermediate inputs. Total value added of exports. This is a measure of the total value added of a sector’s exports from a countrywide perspective. It captures the direct value added of a sector’s exports and the value of the sector’s inputs to other sectors’ exports (forward linkages). Forward linkages. These are the value added of a sector that is exported indirectly through exports of other sectors that contain inputs from the sector. For example, the BPO firm may be providing services to a domestic manufacturing firm that exports its products. The share of the BPO firm’s input to the manufacturing good exports counts as a forward linkage. a. The analysis in this section draws on a new World Bank Value Added Database (Saez et al. 2014) which contains data on both direct exports of services and indirect exports of services. The database uses input-output data from the Global Trade Analysis Project. Figure 4.9: The contribution of services in countries’ Interestingly, services exports in Kenya exports is undervalued have primarily a direct rather than indirect 6 contribution to exports. The share of services 5 in Kenya’s exports is 25 percent when trade is Total value added/gross exports measured in terms of gross value. This large 4 share of services exports is somewhat expected, 3 as tourism is major part of the economy (box 4.4). 2 When measured in terms of direct value added, the share of services exports in total exports 1 reached 34 percent, which is higher than in the 0 comparator countries except Egypt. However, the Kenya SSA peers Non-African High-growth Services Manufacturing Agriculture forward linkages of services are relatively small: Source: Calculations using data from the World Bank Value-Added database. they add an additional 2 percentage points to the share in services value added in total exports of goods and services (Figure 4.10). This contrasts 82 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M with Kenya’s peer countries, where the difference to other exporting goods and services, that is, between direct to total value added services they have significant forward linkages with the exports—an indication of forward linkages—is rest of the economy. Interestingly, the transport much larger. sector, while important on its own, has limited forward linkages, which is probably explained by Figure 4.10: The role of services as an input to other the high volume of transit goods (which count as sectors’ exports is low in Kenya (% of total exports of goods and services) gross exports). 8 Figure 4.11: Direct and total value-added exports by sector 7 (% of direct and total value added of all exports) Total value added / gross exports 6 430 5 380 4 330 3 280 2 230 1 180 0 130 Kenya SSA peers Non-African High-growth 80 Services Manufacturing Agriculture 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Source: Calculations using data from the World Bank Value-Added database. India Cambodia Vietnam Ghana Kenya Tanzania The services sector in Kenya creates fewer Source: Calculations using data from the World Bank Value-Added database. forward linkages than is predicted by the country’s income per capita level. Kenya Kenya’s high services exports are driven by the outperforms other countries of similar income direct value added provided by transport and per capita in all three measures of services export communications services. The total value-added shares (gross, direct, and total). However, forward contribution of transport and communications linkages have a small role to play, given that services to Kenya’s exports is higher than that Kenya’s relative position drops when considering of all comparator countries as well as other the total value added of services exports. This countries with similar income per capita. This finding illustrates that some of the services sectors confirms the finding that these two sectors in Kenya, such as telecommunications, financial comprise firms that have established themselves services, and transport, have prospered owing to in the regional market. Similar conclusions can an intrinsic development path rather than being be made for financial services where Kenya pulled by other sectors through forward linkages. outperforms the peer countries. For the rest of The fact that manufacturing exports have been the services economy, the export contribution is stagnant (relative to GDP) confirms this finding. below or on par with other countries of similar income per capita. Distribution services, which Among those services that have forward in general are important for trade, is one sector linkages to the rest of the economy, information in particular where Kenya underperforms, driven and communications technology and business by the sector’s low direct value-added exports services, distribution and trade, and financial despite its stronger linkages. services stand out. These three sectors are inputs FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 83 Box 4.4: Tourism is Kenya’s leading service export—can it become even larger? There are high expectations of the tourism sector, Figure B4.4.1: Growth of tourism receipts has been solid given the great success of the previous decade. (2000 = 100) International tourism took off in Kenya in 2003, and 430 the growth has been remarkable, also compared 380 to other emerging tourism destinations. Total 330 receipts from foreign tourists jumped fourfold 280 between 2000 and 2012, which was faster than 230 the average for Sub-Saharan Africa (SSA), but not 180 as fast as in Cambodia, Tanzania, or Vietnam (figure 130 B4.4.1). In absolute terms, Kenya’s US$2 billion of 80 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 international tourism receipts were only a fraction India Cambodia of what the more established destinations earn. Vietnam Ghana Kenya Tanzania Egypt’s receipts were five times larger, Morocco’s four times larger, and Vietnam’s three times larger. Source: wttc.org. Tourism is one of the seven core sectors of Kenya’s Figure B4.4.2: Kenya is expected to invest less in the development model. The Second Medium-Term tourism sector relative to its peers Average annual growth 2014-20 (in percent) Plan (MTP-2) aspires to double the number of 9.0 foreign tourists, to three million, by 2018. The 8.0 impact on growth and employment would be 7.0 astounding. Based on data from the World Travel 6.0 and Tourism Council (WTTC), the total (direct 5.0 and indirect) contribution of tourism reached 12 4.0 percent of Kenya’s GDP in 2012.a Moreover, it 3.0 2.0 created 230,000 jobs and an additional 360,000 1.0 jobs indirectly. If foreign visitors increase from the 0.0 1.7 million in 2012 to three million, the sector would Ghana Kenya Tanzania Cambodia Vietnam Uganda India add 10 percent to GDP and half a million more jobs. Capital investment in tourism Tourism receipts Source: wttc.org. Although hopes for the future are running high, from an outsider’s point of view the outlook is not that bright. WTTC projects Kenya’s tourism receipts to rise slower than in the rest of SSA, and far from the needed rate to reach the three million visitors target. Even some of the more mature tourist destinations, such as Indonesia or Morocco, are expected to be more successful in attracting more visitors. The deteriorating security situation has already had a high toll on international visits, and will continue to be a drag on the sector. Disease epidemics, as the 2014 experience with Ebola has shown, can also have detrimental effects. Apart from these exogenous factors, another explanation for the relatively weaker outlook is low investment in the sector (figure B4.4.2). Over the past 12 years, investment in tourism development grew at an average of 7 percent annually in Kenya, while other countries were investing more rapidly. Public policies related to tourism are a key pillar to successful development of the sector. First and foremost, infrastructure capacity has to be consistent with the strategy for tourism development. The airport in Nairobi has a design capacity of 2.5 million and has operated for years at overcapacity (reaching almost six million passengers per year). However, the government has been investing in the airport: a new terminal opened in 2014, and further expansion is ongoing. Other countries are taking similar steps: Vietnam’s main airport, in Hanoi, will 84 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M open a new terminal this year and capacity will rise to 16 million, and Dar el Salaam’s capacity is currently being expanded to six million passengers annually. Second, road infrastructure to key tourist destinations (coast and parks) also needs improvements. The utilization of national parks, Kenya’s premier attraction, is unbalanced, with the top parks (Mara, Amboseli, and Nakuru) being overcrowded, while others, such as Tsavo or Meru, are underutilized. It seems that private reserves are more successful in attracting visitors compared with some of the public national parks. At the same time, the capacity to accommodate foreign visitors, which is a job of the private sector, seems to be there. Bed occupancy ratios have not crossed above 40 percent over the past few years, which is low by industry standards. Services from a Micro Lens Figure 4.12: Higher productivity service activities have higher average wages but employ few workers Although Kenya has become known worldwide 9.0 Average annual growth 2014-20 (in percent) for some of its high-value services, be it M-Pesa 8.0 or high-end luxury safari travel, most of the 7.0 country’s formal services workers are in low 6.0 value-added firms. The largest services firms in 5.0 terms of employment are in wholesale and retail 4.0 3.0 trade, hospitality, public administration, and 2.0 transport. 1.0 0.0 Unlike in manufacturing, services firms exhibit Ghana Kenya Tanzania Cambodia Vietnam Capital investment in tourism Uganda Tourism receipts India a stronger correlation between productivity and Source: Kenya National Bureau of Statistics. wages. Productivity, that is, output per worker, differs greatly across sectors, but more productive Nevertheless, labor costs seem to be rising faster sectors tend to pay higher wages. Finance, ICT than sales, hence lowering the competitiveness and real estate are the star performers in terms of services firms. Looking at the services firms of value added and compensation per worker listed on the Nairobi Stock Exchange (sample of (Figure 4.12). 11 companies with more than K Sh 300 billion in Box 4.5: Integrated survey of services For insight into the characteristics of services firms, this section draws on the 2011 Integrated Survey of Services. The Kenya National Bureau of Statistics conducted this survey of 3,191 formal services firms in 2011, covering firm behavior in 2009 and 2010. Although not intended for national accounts purposes, the results are indicative of firm performance with two important caveats. First, there are data limitations given the intended purpose of the survey, which was to derive input-output structure for supply and use tables rather than to calculate value added and the fact that only 2,300 firms have complete and consistent observations. Second, the timeframe of the survey is not ideal, as it occurred immediately after the global economic crisis, which resulted in a slowdown of growth in Kenya. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 85 annual sales in 2013), labor costs are rising faster firms with high productivity have greater market than sales, hence eroding their competitiveness. share in terms of value added. At the same time, their number of employees grew by only 10 percent between 2009 and 2013 Figure 4.13: Entrant firms are more productive but pay less per worker compared with 60 percent increase in sales, a sign of increasing labor productivity; however, the 40 fact that the labor cost-to-sales ratio is increasing presents a worrying trend. 20 The same as in manufacturing, there is significant 0 dispersion of the productivity of firms within Direct Total Direct Total Direct Total Direct Total the same service activity. Dispersion is especially high in wholesale and retail trade, financial/ Kenya SSA peers Non -African High -growth insurance, real estate, and public administration. Transport Finance Communications The reasons for the high dispersion may be low Distribution & trade Other commercial Source: Kenya National Bureau of Statistics. technology diffusion, poor investment climate (inadequate access to finance for less productive firms), or political economy factors (existence of To sum, services have been the star performer privileged firms). of the past decade. Macro and firm level data illustrate that services have outperformed Entrant firms seem to be leading the increase manufacturing. Moreover, it seems that there in within-sector productivity. Comparison of is more dynamism in the services sector— entrant firms against more established firms entry of more productive firms—which signals shows important differences in the characteristics. that services will continue to grow faster than Entrants are more productive than established manufacturing. Since Kenya’s development firms (Figure 4.13), but generally offer lower goals include job creation in addition to growth, wages (and have lower labor costs). Entrants chapter 5 discusses the relationship between created fewer jobs between 2009 and 2010 than growth and employment, as well as the barriers established firms. Apart from labor, entrants face to job creation. higher costs than established firms. Established 86 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Enhancing Economic Complexity and Productivity through Innovation Knowledge and Innovation of countries’ potential to grow is their investments in innovation activities. A critical element for economic development in the medium and long Innovation Is Widespread in Kenya, terms is the accumulation of knowledge. but Innovation Is Small and Incre- One key input to facilitate knowledge mental accumulation is innovation. At the aggregate level, theories of economic Analysis of the 2014 Enterprise Survey growth have put innovation at the core Innovation Module (ES-IM14)77 suggests of the growth process since the seminal that the rate of innovation in Kenya is work of Solow (1957). The importance large compared with some emerging SPOTLIGHT 4 of the accumulation of knowledge was countries that have implemented reinforced with the emergence of “new national innovation surveys. Around 55 growth theory” (Aghion and Howitt percent of the surveyed firms introduced 1992; Romer 1986). a product or process innovation— defined as a substantial change in At the firm level, which is where products or processes—during 2010– innovation occurs, a large empirical 12 and 68 percent of firms introducing literature documents the importance some marketing innovation. However, a of innovation for moving up the much smaller share of firms introduced development ladder. Hall (2011) shows organizational changes (30 percent), or a robust, positive relationship between logistics and distribution changes (23 innovation and productivity; Harrison percent). The dispersion of innovation et al. (2008) illustrate the impact of in Kenya is high compared with other innovation on employment; and Hall countries that have implemented similar et al. (2011) study the link between firms. On product and process innovation, innovation inputs, such as research and Kenyan firms seem to be more innovative development (R&D), and productivity. than firms in Brazil, China, Malaysia, and Innovation fosters economic other higher-income economies (Figure development, since it facilitates 4.14). However, there are some caveats in technology adoption, improves the comparison of firm innovation across productivity, and as a result increases countries, because of the subjective competitiveness, employment, and nature of measuring innovation wages. Therefore, a critical predictor through surveys. The ES-IM14 sample is a stratified sample by sector and location. It contains 549 firms, 51 percent of which are in 77 manufacturing, and the remaining in services, mainly in wholesale and retail trade, 34 percent. The size composition includes 17 percent large firms, 33 percent medium firms, 43 percent small firms, and 7 percent micro firms. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 87 Figure 4.14: Kenya does well in product and 17 percent of firms’ sales from new process innovation (percent of surveyed firms) products exceed 20 percent of sales in Percent of firms that introduced product or process innovation a given year. in year surveyed Kenya Philippines The innovative efforts lack enough Israel Malaysia knowledge accumulation to have a Brazil China positive impact on productivity growth. Uruguay South Africa The pattern of small incremental Colombia Russia innovation can also be observed by Egypt 0 10 20 30 40 50 60 analyzing the links between these innovations and productivity. The positive impact of innovation on Source: Elaborated from World Bank Enterprise Survey data 2014. productivity growth is a well-established Although these rates of innovation empirical fact for OECD countries. provide a subjective indication of An econometric estimation on the whether substantial changes have innovation premium on productivity78 for the different types of innovation SPOTLIGHT 4 been implemented, it is important to understand how substantial the suggests that there are no statistically changes are. One indicator of more significant productivity differences radical innovation is whether the between innovators and non-innovators, new products introduced are new to except for organizational innovations the national market. Although this where innovators appear to be more is also subjective and relies on the productive. interviewee’s knowledge of the local market, it provides some weighting for Small Investments in Innovation Activities the importance of innovation. In Kenya, only 12 percent of firms introduced Innovation outcomes depend on the such, more radical, innovations. size of the innovation investments and activities that the firms carry It seems the high innovation rates out and that determine the extent of are likely to reflect small incremental knowledge accumulation. Table 4.2 innovations that are distant from the provides a summary of innovation technological frontier. Another metric activities. Around 26 percent of firms to understand the importance of the in the sample carry out R&D, mostly innovation on the firm is to analyze intramural (within the firm) R&D, with the impact of innovation on sales and 36 percent of the innovation related to performance more generally. For the worker training and 47 percent related median innovative firm, 20 percent to the purchase of new equipment of sales are new products and only for innovation. However, the intensity Labor productivity measures (value added per worker) and TFP measures based on a Cobb-Douglas function were 78 regressed on a set of sector dummies, firm size dummies, and a dummy for the type of innovation. All coefficients for all specifications, except organization innovation, are not statistically different from zero. 88 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Table 4.2. Innovation Activities, 2010–12 license, patents Apply patent extramuralb intramurala Equipment trademark Purchase Training Apply R&D R&D R&D Measure Share of firms 26% 25 5.9 36.3 47 3.3 % 5.2 % 9.3 Share of expenditure 0.5 % 0.7 3.8 0.4 % in total sales Source: Elaborated from World Bank Enterprise Survey data 2014. a. R&D activities within the firm. b. R&D activities contracted outside the firm. of these investments is very small, In global terms, Kenya’s innovation especially the investments in R&D and activity—measured through actual training. R&D represents only half a investment in innovation rather than percent of sales (more than 80 percent self-reporting—is less impressive. The of all R&D expenditure in the sample is share of firms spending on intramural carried out by only one firm), training R&D in Kenya is 40 percent lower than SPOTLIGHT 4 in Egypt or Ghana, and less than half of represents 0.7 percent, and new the share in South Africa (Figure 4.15). equipment represents 3.78 percent. In addition, a relatively lower share of Kenyan firms acquire machinery, In addition, there is little purchase equipment, and software, and the same of licenses, patents, and trademarks, conclusion holds for spending on training. with only 3.3 percent of firms buying this form of technology transfer and In general, the data suggest a lack of representing only 0.4 percent of sales. investment in accumulating knowledge Applications for patents and trademarks via innovative effort to converge to the are small, although in line with countries technological frontier. This finding is also of similar income per capita given the clear from the sources of information for lack of innovation capabilities. innovation in figure 4.16, where in most Figure 4.15: Kenya is not spending enough on research and development Innovati on ac ti vity by source (share of fi rms innovati ng) Innovati on ac ti vity by type (share of fi rms innovating) Sources: World Bank Enterprise Survey 2014 data; UNESCO 2014. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 89 Figure 4.16: Kenya’s main source of information for innovation is customer feedback and the internet Customer feedback Internet Products or services In-house R&D and pers Business associations Knowledge from parent Suppliers Consultancy firms Professional journals Don't know Government ministries Recent hires from other Universities and research institutions 0 5 10 15 20 25 30 35 40 45 Percent Source: Elaborated from World Bank Enterprise Survey data 2014. cases information comes from clients or capabilities for improving productivity. products and services, and very little is Kenya’s managerial capacity is high, in-house. Internal sources of innovation relative to its level of development, within-firm or within-group is the largest although there is still catching up to SPOTLIGHT 4 source of information and innovation for do. Figure 4.17 shows the management countries such as Ghana, South Africa, scores of medium and large firms in and most emerging markets. Only about Kenya benchmarked with other countries 15 percent of Kenyan firms rely on in- using the index developed by Bloom house research and development. This and Van Reene (2010) to measure the situation suggests the need for further quality of management. Although the development of internal capabilities. score for Kenya is higher than for other African countries, management quality Lack of Complementary Factors Is is still far from the “managerial” frontier Critical represented by management practices in OECD countries such as Japan or the Managerial and technical capacity United States, and more importantly is often a bottleneck to investing in innovation; Kenya’s endowment in Figure 4.17: Kenya has solid managerial this sense is relatively high. Innovation capacity but still below the frontier (index and income per capita) requires complementary technical and managerial capacity to absorb, implement, and manage product or process improvements. Most of the Management quality index productivity differences observed in Kenya are between firms that implement organizational innovations and those that do not implement such innovations. This is consistent with recent evidence emphasizing the importance of Log of GDP PPP per capita managerial and organizational Source: Graph supplied by Renata Lemos. 90 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M below the managerial capabilities in most internal capabilities and, to some extent, middle-income and emerging markets. lack of management quality. Investments in improving internal capabilities, Overall, although innovation is especially around management and widespread among Kenyan firms, the organization, are critical to facilitate size and depth of these innovations technological innovations that can appear to be limited and significantly substantially increase productivity and constrained by lack of investments in employment in the country. SPOTLIGHT 4 FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 91 92 RAN K E N YA C O U N T R Y E C O N O M I C M E M O Photo: M D U Muthembwa/World Keziah Bank CHAPTER 5 NON-RENEWABLE RESOURCES FOR SUSTAINABLE DEVELOPMENT Introduction revenues can help finance these gaps. This chapter presents scenarios on the fiscal and Kenya has high hopes to become an oil and economic impacts of resource discoveries. gas exporter in a few years’ time. A series of Fiscal revenue projections are based on rule-of- commercial oil explorations in northern Kenya thumb calculations using the production and tax have boosted prospects for Kenya’s upstream oil profiles from a set of benchmark countries to industry.79 Discovered reserves estimated at 600 estimate annual production and revenue.82 Fiscal million barrels were announced in February 2014, analysis compares the consequences of different and follow-up explorations and appraisals have spending patterns by using a multisector dynamic further de-risked the discovered resources.80 In stochastic general equilibrium (DSGE) model addition, several companies have acquired blocks tailored to the specifics of the Kenyan economy. and are drilling (or planning to drill) onshore and offshore. It will take several years before Kenya’s Resource discovery is not a guarantee for oil and gas reserves have been assessed.81 The development, and whether it becomes a curse current slump in oil prices does not accelerate or a blessing depends primarily on the decisions this process; nevertheless, the authorities are taken related to three policy questions. The first already considering the policy and development decision to be made is on how much should be implications of this discovery. spent or saved, and the suggestion here is that If used wisely, oil and possibly gas revenues a permanent income–based approach would can contribute to Kenya’s transformation, best suit the Kenyan context. The consequent and in particular to bridge the saving and decision to be made is where to allocate the investment gap. Natural resources have been resources. This chapter argues for increasing the discovered at a more or less appropriate stage share of health expenditure in total development of Kenya’s development cycle: not too early as spending while maintaining the share of spending in many other Sub-Saharan African economies on education. As to how to implement these, an and not too late. The previous chapters observe effective fiscal rule, accompanied by transparent that Kenya’s savings is low, and at the same decision making and a sovereign wealth fund time the economy is facing infrastructure gaps with saving and stabilization objectives, would that hamper competitiveness, especially of maximize the impact of resource discoveries on manufacturing firms. Thus, the natural resource Kenya’s development. 79 While Kenya also has significant non-hydrocarbon minerals, the analysis in the chapter focuses exclusively on recent developments in hydrocarbon sector and their economic implications. 80 See the press release by Tullow Oil Inc., at http://www.tullowoil.com/index.asp?pageid=137&category=&year=Latest&month=&tags=84&n ewsid=878. 81 Detailed information on field development, production plans, and fiscal revenue estimates are not yet available. 82 These values are projected using production and capital cost aggregates by PWC (2015) and the time profile of the Jubilee field in Ghana, U.S. Energy Information Administration oil price projections, and the average of average effective tax rate profiles from comparable low- and middle-income countries as reported by IMF (2012). FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 93 Making the right decisions is only one part of the broader time frame, which reflects the time problem; managing expectations and resisting required to develop the fields and optimize the the political economy pressures are equally costs of production. Hence, based on the current challenging. Benefits from natural resource exploration results, oil production will not be wealth may not be realized immediately. Thus, substantial and is unlikely to provide a global the initial euphoria and excitement about the market niche for Kenya to specialize. new found wealth could easily transform into Figure 5.1: Proven oil reserves by region/country, 2013 resentment, suspicion, and public anger over (thousand million barrels) a short time period. This is more likely when expectations are not managed, transparency Libya, 48.0 Nigeria, 37.2 is not established, and information is scarce. South Middle It is therefore extremely important for the East, 807.7 Sudan, 3.5 government to establish a communication Africa, 130.3 Kenya, 0.6 strategy to manage the expectations by spreading South and C. America, 328.4 Algeria, 12.2 the information in a timely and reliable manner. North Asia America, 220.2 Pacific, 41.5 From Oil Reserves to Oil Revenues Europe and Egypt, 4.3 Angola, 12.7 Eurasia 40.8 In global terms, Kenya’s discovered resources are relatively small. The country’s 600 million Source: Data from BP Statistical Review of World Energy 2013. barrel stock puts Kenya in the 47th position worldwide in terms of oil reserves, just ahead Nevertheless, oil and gas production is expected of Uzbekistan. This quantity constitutes a small to have a non-negligible impact, especially on fraction of the reserves in resource rich African fiscal revenues. Kenya’s possible recoverable countries like Algeria, Angola, Libya, and Nigeria, reserves could reach about 1.4 billion barrels of in absolute and per capita amounts (Figure 5.1). oil and 1.7 billion barrels oil equivalent of natural By comparison, Saudi Arabia produced about gas (PWC 2015). The most recent estimates show 11.5 million barrels of oil per day in 2012. At that that oil production will start in 2020-2022, and speed of production, Kenya’s reserves would be reach a plateau of about 77 million barrels a year depleted in only 52 days. In practice, however, soon after that (Figure 5.2). Starting in 2032, the production in Kenya will be spread over a production will decrease gradually, reflecting Figure 5.2: Estimates for oil production and fiscal revenues, 2020–75 Oil and Gas Production Oil Prices and Fiscal Revenues 180 10 120 9 160 8 100 140 7 120 6 80 MBBoe/year 5 Billion USD 100 4 60 USD 80 3 60 2 40 40 1 0 20 20 -1 0 0 2015 2018 2021 2024 2027 2030 2033 2036 2039 2042 2045 2048 2051 2054 2057 2060 2063 2066 2069 2066 2069 2060 2063 2054 2057 2051 2042 2045 2048 2036 2039 2030 2033 2024 2027 2015 2018 2021 Oil, MBBL/year Gas, MBBoe/year Fiscal Revenue Oil Prices (RHS) Sources: Authorities; USEIA price projections; World Bank staff calculations. 94 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M the maturing of existing fields. In comparison, could rise even faster than they have in the past the production of natural gas is estimated to couple years, and the infrastructure projects that start in 2025 and peak at 95 million barrels of are needed to establish access to international oil equivalent per year in 2033. Calculating the markets for oil could generate additional jobs for fiscal revenues associated with these production local communities. Moreover, the Government of profiles requires a detailed approach with Kenya has ambitious desires to transform Kenya information on cost profiles and the production into an oil and natural gas hub in East Africa by agreements between the Government of Kenya taking advantage of the scale of the oil and natural and the producing companies. In the absence of gas sector and the country’s strategic position. such information, rough estimates, using World Bank oil price projections and general industry Any forecasts of future oil revenue must rules of thumb, show that Kenya’s fiscal revenues be looked at with caution, as oil prices are from oil production are projected to peak at extremely volatile and unpredictable. Oil prices, about US$8.9 billion in 2033. This is roughly and in turn oil production and revenue, tend to equivalent to 16 percent of Kenya’s 2013 gross fluctuate significantly in short- and long-term domestic product (GDP). time horizons. Figure 5.3 shows the evolution of oil prices between 1987 and 2015 and the Estimated fiscal revenues can help finance associated volatilities. Although in the long term Kenya’s infrastructure deficit. The previous fiscal revenues from oil production in Kenya chapters explain Kenya’s saving gap, which in turn are expected to have a hump-shape, oil price explains the relatively low levels of investment, fluctuations will make fiscal revenue volatile as well as the weaknesses in terms of physical as well. In 2014, oil prices unexpectedly fell by and human capital. The oil sector, if properly more than half. This sharp drop is a reminder that managed, can generate the needed revenue projections may change significantly, but also to bridge the infrastructure and skill gaps. As that government policies should be isolated from an example, the annual fiscal revenue at peak such fluctuations. production under a baseline scenario would be Figure 5.3: Oil prices have been particularly volatile since 2000 sufficient to cover the total cost of the standard gauge railway from Mombasa to Nairobi that is Weekly Brent Spot Price 25 160 currently being constructed. 20 Level Growth 140 15 120 10 100 A well-established oil and gas sector has Percent USD 5 80 other potential benefits as well. In addition to 0 60 -5 40 generating fiscal revenues, the hydrocarbon -10 20 -15 0 sector can also catalyze other economic activities in an indirect manner, which may be difficult to quantify at this early stage of development. For instance, the foreign direct investment inflows Source: Data from U.S. Energy Information Administration. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 95 From Oil Revenues to Economic Resource revenues may actually harm the Development: What Will Determine If It national economy if spending escalates too Is a Blessing or a Curse? rapidly, because it is volatile, unpredictable, and finite. When oil revenues are injected Oil discovery is by no means a guarantee for into the economy directly, or spending from the economic development of a country. On the natural resources reflects the same intrinsic contrary, the recent history outside Kenya points characteristics of the natural resource revenues, to many examples where resource discovery—be the economy may suffer short- and long-term it oil, gas, diamonds, or other minerals—has led consequences. A rapid increase in spending to economic demise and conflict, or has simply would lead to higher demand for goods and not lived up to its potential. The examples of services, which typically translates into a hike Angola, the Democratic Republic of Congo, and in prices of non-tradable items. In parallel, the Malaysia are perfect illustrations of this sad truth, domestic currency could appreciate in response as all three countries started from a similar level to high foreign currency inflows in the domestic of development and oil production back in 1972 economy. As a result, the competitiveness of (Figure 5.4). The Democratic Republic of Congo domestic producers diminishes, which also and Angola increased oil production by 15 and implies long-term losses in potential output, a four times, respectively, between 1972 and 2010, phenomenon known as Dutch disease. Volatile while their GDP per capita was practically the and unpredictable spending magnifies this impact same as it was four decades earlier. In contrast, by diminishing the risk-adjusted returns. Malaysia quadrupled its GDP per capita during the same period, while its oil output increased Another key determinant for making resources a only slightly compared with the oil boom in the blessing is the policy and institutional framework other two countries. These examples show that for managing resources. A comprehensive and oil is neither necessary nor sufficient for rapid clear legal framework, institutional structure, and sustained economic development. transparency mechanisms, and a sovereign Figure 5.4: Different development trajectories: Angola, wealth fund are suggested as most important for Democratic Republic of Congo and Malaysia, 1972–2010 achieving positive outcomes. Managing the oil GDP vs. Oil Production (1972-2010) 16 sector is a complex task that requires a proper GDP per person (2005 USD, PPP) 14 Malaysia, 2010 legal framework that stimulates investment in 12 the sector, adequate institutional setup, and 10 Angola, 19 Angola, 2010 administrative capacity related to monitoring oil 8 Congo, 197 production, collection, and use of oil revenue. Congo, 2010 6 Transparency (and oversight) is a critical pillar 4 in the institutional framework, and although it 2 0 Malaysia, 1972 cannot ensure the responsible use of resource 0 1 2 3 Oil Production Per Person (tonnes) 4 5 6 revenues, without transparency, abuse is almost certain.83 One step in this direction is to implement Sources: BP Statistical Review of World Energy; World Bank Word Development Indicators. the Extractive Industries Transparency Initiative (EITI) standards.84 Nineteen African countries, 83 Humphreys, Sachs, and Stiglitz (2007). 84 EITI.org. 96 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M including neighboring Mozambique and Tanzania, revenue among the levels of government, have already subscribed to the EITI standards. In necessitate unified solutions. In addition, July 2015, the Kenyan authorities announced that legislation has been proposed (for example, on a focal point for the EITI implementations would mining) in the absence of a clear policy for the be established within six months. The authorities sector. Various policy proposals on critical issues, planned to adopt a transparent policy and such as the one on revenue sharing, are being legislative framework for the oil and gas sector, presented from different parties in the form including the adoption of transparent processes of legislative proposals and driven by special of licensing and publication of contracts. Last interests, which makes consensus difficult. but not least, having a sovereign wealth fund Legislative proposals are being drafted (for (SWF) has proven to be a good instrument for example, on the SWF) without in-depth analysis managing resource revenue, and such fund(s) to guide the proposed legislative solutions. may serve a saving or stabilization function. Unless the various stakeholders, in particular Global experiences illustrate that there are many the central government, county governments, solutions to the design and management of an and Parliament, start making coordinated and SWF. The Kenyan authorities have been drafting informed decisions on the management of an SWF bill since 2014, incorporating a broader natural resources, the oil discoveries are at risk policy framework for managing resources. to become a curse rather than a blessing for the Kenyan people. Finally, inadequate attention is Improving policy coordination in resource being paid to managing the expectations and management is crucial for the achievement needs of the local communities. Because oil of the expected outcomes from resource happened to be discovered in Kenya’s poorest revenues. Policy decisions require careful and conflict-prone region, addressing the analysis and deliberation, in particular for economic and social needs of the people in those countries with multiple tiers of government areas is critical for avoiding unnecessary conflict. that share responsibilities over the use of Kenya can learn from the mistakes of other oil public resources. It is unclear if Kenyan policy producing countries in Africa and elsewhere to makers’ current legislative efforts are sufficient avoid falling into the same trap. and aligned with the best practices for the development impact of natural resources. The In the case of Kenya, the socioeconomic Constitution of Kenya, especially articles 69 to characteristics of Turkana County, where oil 72, provides the broad foundation of obligations reserves were found, present new opportunities. that regulate environmental and natural resource Turkana is one of the poorest parts of Kenya; it management. Some progress has been made to has the highest poverty rate, with 94 percent enact the necessary laws to operationalize these of the population living under the poverty line principles. For example, a mining law has been (Kenya Open Data 2009). The territory of Turkana submitted to the Parliament and the drafting of is arid with a difficult climate and terrain, and legislation related to various resources, such as most of the population is nomad pastoralists ore, oil, and gas, is being done. However, much is whose livelihood depends on moving around left to be desired. Legislative efforts are typically in search of good pastures and water for their done in an isolated manner, whereas some of livestock. Three-quarters of Turkana’s population the policy issues, such as how to share resource are thought to depend on humanitarian and FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 97 food aid. Thus, the oil and gas sector provides an financial or otherwise, would be made available. opportunity to revamp the Turkana economy and A similar approach could play an important role in achieve significant development outcomes. preventing any potential conflict among various parties in Kenya. Traditionally, violent competition for scarce resources, such as water and arable land, has So, is oil a curse or a blessing? The answer lies been common among groups in Turkana and in the policies chosen. Whether Kenya’s oil across borders. The theft of livestock, known discovery will contribute to building a Kenyan as raiding, is a traditional activity among the economic miracle depends primarily on the policy groups and in recent years it has become more decisions that will be taken on the management violent and destructive. The proliferation of of the oil sector and oil revenue. Fiscal policy small arms in the area has been identified as a will play a particularly important role in ensuring source of increased violence among pastoralist effective and efficient use of resources as well as communities. The boundaries between Turkana in minimizing the downside risks to the economy. and West Pokot counties have been particularly volatile and the recent discovery of oil in the Translating the finite and volatile oil revenues border areas contested by these two counties has into development outcomes will depend on added to the controversy. how the Kenyan government responds to three questions related to the management of its oil The devolution process that begun in 2013 revenue. As oil starts to flow from the ground and increased Turkana county’s fiscal resources oil revenue begins to pour in, the Government of as well as its participation in national politics. Kenya will have to decide on the following: A previously sidelined county has now taken 1. How much should the government spend national notoriety, and the discovery of oil (and of the oil revenues and how much should it water) resources contributed to that process. In save? this context, managing expectations around the oil industry and ensuring an equitable distribution 2. How should policy makers allocate the of the oil revenues are seen as critical factors in additional spending that is financed by ensuring conflict-free oil development in Turkana. resource revenues? 3. What institutional mechanisms to be used A national communication strategy that helps to implement the answers to the first two to manage expectations, ensure transparency, questions? and publicize new opportunities is warranted. It is extremely important to create clarity in the In principle, the answers to these questions public’s mind on how natural resources are being managed and how benefits are being allocated should incorporate (i) the particular across the constituencies. In the case of Tanzania, characteristics of the oil sector in the country, Zeufack and Woodroffe (2013) suggested a such as its size and the dynamic characteristics communication strategy that informs the public of the resource envelope; (ii) the economic and about how the extractive industry in question social returns associated with public expenditure; works; what skills, goods, and services will be and (iii) the absorptive capacity constraints of the needed in developing the industry; and what implementing agencies. support for relevant entrepreneurial activity, 98 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M To Save or to Spend? On the other hand, low- and middle-income countries also face tighter constraints against The decision of whether to save or spend the effective implementation of infrastructure natural resource revenues is not only a moral spending. The returns to domestic investments problem—how much should current versus are likely to decrease fast in countries with weak future generations spend—but also an economic governance and low absorption capacity. Many and administrative efficiency issue. As in many projects that can be financed with resource economic decision making problems, optimal revenues would require domestically produced levels of savings and spending are determined by inputs, both goods and labor. But the countries wishes and constraints. In the case of revenues generally are not capable of responding to from nonrenewable natural resources like oil, the rapid build-up of demand, creating supply relying only on wishes would lead to a quick bottlenecks in the economy. Therefore, an depletion of the natural assets without creating immediate boost in infrastructure spending is not alternative income-generating processes, and necessarily optimal. The optimal level of spending create unintended negative consequences for from resource revenues is country-specific and the economy. Similarly, considering only on the is determined primarily by the infrastructure constraints would cause inefficient allocation of gap in the economy and the efficiency of public resources and lost opportunities. Therefore, it is investment expenditures. important to emphasize that saving is not only a moral obligation, but a condition determined by Despite the recent effort to scale up investments, country-specific factors that helps to maximize Kenya’s gap in physical infrastructure remains the welfare of the constituents. wide and is one of the key bottlenecks to the ambitious growth plans outlined in the On the one hand, the need for spending is more Vision 2030. These gaps are particularly pronounced in low- and middle-income countries large in transportation, energy, and water with insufficient provision of public services. infrastructures. Kenya produces less than a tenth There is no doubt that public investments are of the electricity produced in middle-income essential for stimulating growth in low-income countries on a per capita basis. Similarly, access countries. For example, Eden and Kraay (2014) to improved water sources in Kenya is lower than find that an extra dollar of public investment in any of the peer countries apart from Ethiopia. raises output by 1.5 dollars in a sample of 39 The road density in Kenya is far lower than in low-income countries. These results imply that any of the richer countries in the peer group. returns on domestic physical assets could be The Government of Kenya has recently initiated higher than investing in financial assets abroad. aggressive infrastructure projects to address In addition, productive investments raise the some of these gaps. The Lamu Port and Lamu- economy’s growth potential, and hence benefit Southern Sudan-Ethiopia Transport Corridor and future generations as well. The issues raised by Standard Gauge Railway projects, when finished, these arguments constitute the “wishes” part of could be important contributions in this regard. the decision-making problem; however, several One aspect of infrastructure where Kenya has counterarguments put limits on the effective done well is telecommunications. Mobile phone implementation of these wishes. penetration is high and, although lower than in FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 99 the richer peer countries, Kenyan citizens use and information and communications technology mobile phones for payments and even saving, projects was 82 percent, and for environmental which is not the case in most other countries. and water projects about 80 percent. In comparison, budget execution in sectors with a The quality of the country’s human capital is large current spending component was higher. mixed: improvements in education in recent For education, for example, it reached about 95 years have closed the gap with peer countries, percent. Another concern about scaling up public but health outcomes remain weak. The secondary investments too rapidly is the efficiency of project school enrollment ratio in Kenya is lower than the selection and management. Dabla-Norris et al. middle-income country average and the ratio in (2010) provide an index of public investment the Arab Republic of Egypt, but the margins are management efficiency for a sample of 71 low- relatively small. However, the gaps in health and middle-income countries (Figure 5.5). Kenya conditions are significant, as indicated by the near performs poorly on most aspects of public 10 year difference between the life expectancies investment management, project management in Kenya and middle-income countries. Overall, being the notable exception. addressing these gaps by undertaking further public investments would be required to boost Consequently, an increase in public investments growth in the non-resource sectors. should be accompanied by capacity enhancements. Ideally, a spending plan should Vision 2030 and the Second Medium-Term Plan be in place even before oil revenue starts to flow. (MTP-2) embrace an aggressive infrastructure Such a plan would take into consideration time- investment strategy, but not all investment varying revenue projections, public investment contributes to accumulation of public capital. gap, and implementation constraints. The next Absorptive capacity constraints, which are already section presents simulation exercises that visible at the current levels of public investment, compare alternative policy scenarios for savings limit the economic impact of public investments. and investment, and discusses the macro-fiscal Between 2009/10 and 2011/12, the average outcomes in each case. budget execution rate for energy, infrastructure, Figure 5.5: Public investment management efficiency index and sub-indexes (Selection from a sample of 71) 1.49 Overall Score 1.65 1.33 Evaluation 1.33 2.27 Managing Selection 1.2 Appraisal 1.33 1.17 Congo, Rep. Cote d'Ivoire Ethiopia Cambodia Mauritania Belarus Namibia Jamaica Indonesia Moldova Kenya Tunisia Botswana Bolivia Djibouti Bangladesh Mozambique Azerbaijan Guinea Rwanda South Africa Armenia Nigeria Mongolia Senegal Uganda Pakistan Ukraine Lesotho Gambia Gabon Zambia Benin Sudan Kazakhstan Burundi Swaziland Jordan Turkey Ghana Burkina Faso El Salvador Tanzania Hai ti Mali Chad Malawi Egypt Source: Dabla-Norris et al. 2010. 100 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Alternative Approaches to Scale Up Public i. Spend-as-you-go (SAYG). This approach Investments presents the least strategic policy stance. In this case, the entire flow of resource It is important to compare the alternative revenues is spent on additional public saving and spending approaches by fully investment projects as the revenues accounting for their implications. This section become available. shows how alternative saving and spending decisions would affect the Kenyan economy ii. Permanent income hypothesis (PIH). This through different channels in the long term. A approach denotes a sustainable spending DSGE framework is tailored and calibrated to path that allocates the spending evenly over reflect the Kenyan economy in 2015.85 Public the time. Public investments are scaled up by investment expenditures are defined broadly to an amount equal to the permanent income include all productive public expenditures. These annuity implied by the present value of include infrastructure expenditures as well as resource wealth. social spending with direct impact on human iii. Bird-in-hand (BIH). This approach represents capital. In the baseline, when public investment the most stringent spending stream. All expenditures are scaled up using the resource revenues from the oil sector are accumulated revenues, all these components are scaled up in a sovereign wealth fund that invests proportionately. The results show the effects on abroad. Only the interest earnings from this public capital accumulation, non-resource GDP fund are used to scale up public investments. growth, sector composition of the economy, iv. Moderate frontloading (MF). This approach public debt stock, and savings in the wealth fund. presents a case where public investments are scaled up rapidly in the beginning of the The simulations compare four alternative simulations. In principle, neither the flow nor expenditure paths by holding fixed the the stock of the additional spending stream composition of investments. To cover a range is linked to the resource wealth under this of possible fiscal policy options with different approach. However, in the long term, the degrees of strategic decision making, the amount that is used for additional public simulations for Kenya compare four alternative investments converges to the permanent approaches (see box 5.1 on the formal definition income annuity to ensure sustainability. and spending simulations for each approach). In all cases, the composition of investments is fixed at the levels given by the latest available data; Formal definitions of these approaches, and however, the scale of investments is adjusted the corresponding spending simulations, are accordingly. These approaches are the following: discussed in box 5.1. The simulations described here do not provide forecasts for macroeconomic variables in the future; but show the marginal impacts of 85 different spending trajectories on these forecasts. For more information on the model specifics, see appendix B and Levine, Melina, and Onder (2015). FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 101 Box 5.1: Alternative approaches to scaling up public investments 102 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M 10.0 Spend As You Go (SAYG) 10.0 Permanent income rule (PIH) 9.0 9.0 8.0 8.0 7.0 Borrowing Oil Revenues 7.0 Oil Revenues 6.0 6.0 PIH Annuity Billion USD 5.0 Billion USD 5.0 4.0 Savings 4.0 3.0 Spending 3.0 2.0 2.0 Use of interest 1.0 1.0 earnings 0.0 0.0 -1.02014 2018 2022 2026 2030 2034 2038 2042 2046 2050 2054 2058 2062 2066 2070 -1.0 2015 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 10.0 Moderate Frontloading (MF) 10.0 Bird in Hand (BIH) 8.0 9.0 Oil Revenues 8.0 6.0 Savings Oil Revenues Billion USD 7.0 BIH Annuity 4.0 6.0 Billion USD 5.0 Borrowing Use of interest Savings 2.0 earnings and 4.0 other resources 3.0 Use of 0.0 interest 2.0 2015 2017 2019 2021 2023 2025 2027 2029 2031 2033 2035 2037 2039 2041 2043 2045 2047 2049 2051 2053 2055 2057 2059 2061 2063 2065 2067 2069 earnings 1.0 -2.0 0.0 -1.02015 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 The SAYG approach does not lead to any savings; therefore, transfers to the budget from the resource boom diminish over time, following the resource revenue depletion. Under PIH, the government transfers about $2.7 billion to the fiscal budget annually (black line in panel b). In the short term, this is financed by borrowing from abroad (the first yellow shaded area), as resource revenues are relatively low at this stage. In the medium term, the resource revenues pick up and reach a peak of about $9 billion. The difference between revenues and transfers is saved in a sovereign wealth fund (green shaded area). Finally, as the revenues gradually die out, interest earnings on the welfare fund assets are used to supplement the transfers to the budget (second yellow shaded area). Under BIH, transfers to the fiscal budget are scaled up over time as resource revenues are saved in the sovereign wealth fund and interest earnings on wealth fund assets increase (black curve line in panel c). Until the early 2040s, resource revenues exceed the transfers; therefore, reserves continue to build up. Later in the projection horizon, accumulation comes to a halt and the BIH annuity reaches a plateau. Finally, the “big push” under the MF approach leads to investments that are financed by borrowings in the short term (first yellow-shaded area). In the medium term, resource revenues exceed spending; however, the difference is smaller than with PIH or BIH. Moreover, the spending converges to the PIH annuity in the long- term; however, spending remains above the PIH. Therefore, stabilization fund savings would be a lot smaller than the levels with PIH or BIH. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 103 All the approaches assume an increase in public Non-resource GDP responds to higher investments; the difference between them lies in public expenditure levels; yet, this impact is the timing and scale of the increases. Figure 5.6 not sustainable under the SAYG approach. shows the evolution of investments under each Although the speed limit reduces public capital approach using the baseline oil price projections. accumulation under the SAYG approach, the large The SAYG approach mimics the dynamics of oil scaling-up of investments still has a significant revenues illustrated by the inverted-U shape in medium-term impact on non-resource GDP. At figure 5.6, panel b; it thus leads to an aggressive its peak, non-resource GDP is about 13 percent scaling-up of public investment expenditures greater than its initial equilibrium value. This is toward the middle of the projection horizon. partially because higher expenditures not only In about two decades, this approach reaches a increase the infrastructure investments, but also maximum, more than doubling public investment build up more physical and human capital, which expenditures compared with the initial level. In do not suffer from absorptive capacity constraints. comparison, the MF and PIH approaches bring However, this impact is not sustainable, because about a permanent and relatively moderate the resource revenues are depleted toward the rise at the outset. The MF approach increases end of the projection horizon. In comparison, public investment expenditures to a maximum steady spending under the PIH and MF of 100 percent relative to the initial level before approaches brings the non-resource GDP close it gradually approaches about 50 percent; the to or even higher than the SAYG value in the steady increase implied by the PIH. The BIH long term, with the MF exceeding it by more approach gradually scales up public investments, than 5 percentage points. The BIH approach, by reaching SAYG only in the mid-2040s when the contrast, keeps the non-oil GDP close to its initial expenditures under the latter approach are levels for a long time before the interest earnings reduced rapidly. become large enough to have a significant impact on non-resource GDP, which occurs only around However, absorptive capacity constraints impose three decades after the revenue starts to flow. a “speed limit” on scaling up public investments in an efficient manner. The simulations show that The sector composition of the GDP shifts higher spending does not automatically translate significantly under the different expenditure into a proportionate increase in public capital. In policies. All the approaches, apart from SAYG, lead the short term, a rapid scaling-up under the MF to a gradual and relatively balanced expansion of approach leads to significant losses in average non-resource GDP over the projection horizon. infrastructure efficiency. However, the loss is In contrast, the SAYG approach leads to more significantly larger under the SAYG approach. As prominent Dutch disease–like symptoms. Under a result, although the SAYG spends significantly the MF and PIH approaches, the tradable and more, the two approaches lead to similar levels non-tradable sectors grow at a relatively stable of public capital accumulation (about 60 percent rate. For the BIH approach, the growth rates in greater than the initial equilibrium) by the mid- both sectors are relatively back-loaded, but they 2050s. This shows that the additional spending are balanced across sectors. In contrast, the under the SAYG is wasted. SAYG approach leads to a rapid expansion of the 104 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M non-tradable sector early in the projections (up fund balances exceed four times the current to 12 percentage points higher than the initial GDP by 2075. Similarly, the savings under the PIH equilibrium), which is sustained for a prolonged approach exceeds 3.5 times the GDP. A slightly period of time. The expansion of the tradable smaller accumulation compared with the BIH sector comes in the second half of the projection approach reflects the borrowings to finance the horizon, and is relatively short-lived. These initial escalation of public investments. differences in the sector compositions can be traced back to Dutch disease symptoms in the Overall, the baseline analysis suggests that the economy. The rapid escalation of expenditures permanent income hypothesis approach best under the SAYG approach leads to a significant suits the characteristics of Kenya’s economy. The and sustained appreciation of the domestic most relevant criteria for Kenya in deciding on the currency in the first half of the projection horizon. optimal approach are the impacts on the non- This leads to an erosion of competitiveness in the resource economy, efficiency of spending, and tradable sector. Currency appreciations under the sustainability of fiscal outcomes. The simulations PIH and MF approaches, however, are relatively show that spending resource revenues as they short-lived and limited to the early years. become available (as in the SAYG approach) are wasteful and incapable of delivering a better The PIH and BIH approaches lead to better result than other approaches in promoting and more sustainable fiscal outcomes. The non-resource growth and sustainability in fiscal simulations evaluate the fiscal implications by balances. Moreover, this approach is most likely comparing asset and liability accumulations under to trigger Dutch disease symptoms in the medium each approach. The lowest debt-to-GDP ratio is term. In contrast, saving all the revenues (as in generated by the SAYG approach, because there the BIH approach) is too stringent. Although this is no debt issuance for financing investments in approach helps to build large quantities of fiscal this case. The debt-to-GDP ratio thus decreases buffers, it falls short of boosting the non-resource from about 45 percent in the beginning of the economy with much needed investments projections to 42 percent by 2075. However, in infrastructure, education, and health. In there is no accumulation of savings either. In comparison, the PIH and MF approaches facilitate contrast, the MF approach raises public debt the non-resource growth; however, the PIH approach most because of the initial “big push.” The debt- performs much better in fiscal outcomes. As to-GDP ratio increases from about 45 percent to the degree of front-loading increases beyond a close to 60 percent over the projection horizon. certain threshold, absorptive capacity constraints At the same time, the oil revenue savings reach limit the effectiveness of the MF approach just about 90 percent of the current GDP. However, like in the case of the SAYG approach. Therefore, the true winners for fiscal sustainability are the any attempt to frontload the investments should more conservative approaches. Under the BIH consider the PIH approach as a benchmark, and approach, the debt-to-GDP ratio decreases by not deviate too far from it, to minimize waste. about 2 percentage points, and stabilization FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 105 Figure 5.6: Alternative expenditure scenarios a. Public Investment b. Average Infrastructure Efficiency c. Public Capital (% Deviation from Initial Steady state) (%) (% Deviation from Initial Steady state) 200 45 70 180 60 160 40 140 50 120 40 35 100 30 80 30 20 60 40 10 20 25 0 0 -20 20 -10 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 d. Non-Oil Output e. Tradable Non -Oil Output f. Non-Tradable Output (% Deviation from Initial Steady state) (% Deviation from Initial Steady state) (% Deviation from Initial Steady state) 18 25 14 16 12 20 14 10 12 15 10 8 8 10 6 6 4 5 4 2 2 0 0 0 -2 -5 -2 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 g. Real Exchange Rate h. Public Debt i. Sovereign Wealth Fund Savings (% Deviation from Initial Steady state) (% of GDP) (% of GDP) 2 70 600 1 65 500 0 60 400 -1 55 300 -2 50 200 -3 45 100 -4 40 -5 35 0 -6 30 -100 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 Spend-As-You-Go Bird-in-Hand Permanent Income Moderate Frontloading Source: Levine, Melina, and Onder 2015. Note: SS = initial steady state. Infrastructure efficiency is assigned by a parameter in production function that is based on the ranking in figure 5.5. Lower oil prices weaken the outcomes of the MF a smaller degree in comparison with the baseline approach, that is, they favor less frontloading. oil price scenario. The difference between long- A low oil price scenario for the simulations leads term impacts on the non-resource GDP between to a stronger separation between alternative the SAYG and PIH approaches remains same; approaches. A reduction in long-term oil price however, it changes sign in favor of the BIH depresses spending under all the approaches approach with adverse oil prices. except the MF approach. This is mainly because the MF approach is interpreted as commitment to How to Allocate Investment? a plan that is unlinked from variations in resource The simulations in this section compare revenues. The balanced outcomes under the PIH alternative compositions of public investments and BIH approaches are also reinforced against on the basis of their long-term implications. the SAYG approach with adverse oil prices. Overall, Actual investment paths for education, health, the absorptive capacity constraints continue to and infrastructure are determined by two forces restrict the impact of the SAYG approach, albeit to 106 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M that are defined by policy decisions. The first one For all types of investments, the scale effect is the scale effect, which describes the changes in dominates the composition effects. In the long the level of total public investments. The public term, education and health investments are investment expenditure scenarios described in higher than their initial levels under all three the previous section determine the magnitude of scenarios. Figure 5.7 shows that by the end of this effect. The second force is the composition projection horizon, investments in education effect, which describes the structure of spending. increase from about 2 percent of 2020 GDP to about 3 percent, and health investments increase What is the optimal composition of spending? from about 0.6 percent of 2020 GDP to about To answer this question, the simulations in this 0.8 percent, even when the AIC is chosen. The section will assume the scale of expenditures investments in each category never fall below the as given by the PIH approach. Then, three initial levels in these simulations, mainly because alternatives to the allocation of spending on the additional investments generated by the PIH infrastructure, education, and health will be approach are large enough to compensate any compared on the basis of their long-term growth potential losses if an unfavorable composition and fiscal implications: approach is chosen. This is particularly clear in the case of infrastructure investments: they * Aggressive infrastructure-based composition increase from about 6 percent of GDP in 2020 (AIC). This approach keeps the shares of to about 6.5 percent in the same period if ASC all components in total public investments is chosen. Proportionately this increase is small fixed at their current levels. These shares because physical capital depreciates faster are approximately the following: 70 percent than education and health. Gross investments infrastructure investments, 24 percent in this case are just large enough to offset the education, and 6 percent health. Total public depreciation under the ASC. investments are set as implied by the PIH approach; thus, the size and composition The simulations show that the limited scaling- of the investments are kept constant up under the PIH approach saves the AIC from throughout the projection horizon. being punished heavily by the absorptive * Aggressive skill-based composition (ASC). The capacity constraint. A rapid scaling-up of public share of education in public investments is investments does not necessarily mean that public gradually increased from the initial level capital is scaled up quickly. Efficiency constraints to about 40 percent at the expense of in public investment projects bind most when investments on infrastructure, whereas the infrastructure investments are scaled up rapidly share of health is kept constant. as under the AIC approach. As a result, higher * Balanced composition (BC). The share of spending in this approach does not necessarily health in public investments is gradually translate into faster capital accumulation if public increased from the initial level to investments are scaled up more rapidly than what about 11 percent at the expense of the PIH approach suggests. However, when the infrastructure, whereas the share of PIH is chosen, this is not a problem. Thus, figure education is kept constant. 5.7 shows that the gap between the public capital FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 107 Figure 5.7: Investment composition scenarios a. Infrastructure Investment b. Education Investment c. Health Investment (% of Initial GDP) (% of Initial GDP) (% of Initial GDP) 9.5 6 1.6 9 1.4 5 8.5 1.2 8 4 1 7.5 3 0.8 7 0.6 6.5 2 0.4 6 1 5.5 0.2 5 0 0 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 d. Average Infrastructure Efficiency e. Public Capital f. Non-Oil Output (%) (% Deviation from Initial Steady state) (% Deviation from Initial Steady state) 45 60 14 40 50 12 35 10 30 40 25 8 30 20 6 15 20 4 10 10 2 5 0 0 0 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 g. Real Exchange Rate h. Public Debt i. Sovereign Wealth Fund Savings (% Deviation from Initial Steady state) (% of GDP) (% of GDP) 1.5 60 400 1 350 55 0.5 300 0 50 250 -0.5 45 200 -1 150 -1.5 40 -2 100 35 -2.5 50 -3 30 0 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 Balanced composition Aggressive infrastructure composition Aggressive skill-based composition Source: Levine, Melina, and Onder 2015. levels among the three approaches widens over the expense of the other one would eventually time. By the end of the projection horizon, the decrease the total output. The BC approach leads gap between the AIC and BC is relatively small, to a more balanced combination of physical and whereas public capital shrinks toward its initial human capital, which is more conducive to long- level under the ASC. term growth. BC public investments translate into higher The three composition approaches bring about growth in the non-resource sector. Public capital similar fiscal sustainability outcomes; however, stock under the BC approach is smaller than under the fiscal buffers are lower in the ASC case. Total the AIC in the second half of the projections. public debt as a share of GDP remains similar However, the human capital stock is greater. As a in all the composition scenarios. In all cases, result, non-resource GDP under the BC approach the public debt-to-GDP ratio climbs from about grows more than it would under the AIC 45 percent in 2020 to about 55 percent in the approach. This is mainly because infrastructure medium term, and then stabilizes around 50 and human capital complement each other’s percent in the long term. Accumulation of savings productivity. With diminishing returns to each in the stabilization fund exhibits significant factor, this implies that increasing one factor at differences. By the end of the projection horizon, 108 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M the BC and AIC approaches lead to savings that investment composition and scaling-up have are about 350 percent of GDP. The savings under abstracted from a critical issue that is important in the ASC approach are about 290 percent of GDP. the actual implementation of these decisions. In As the GDP under the ASC approach is lower than the economic model, which defines the behavior the ones under other approaches, the gaps in of agents with mathematical approximations, savings-to-GDP ratios imply greater differences in Kenyan citizens are assumed to have perfect savings in nominal terms. information about the policy makers’ decisions in the short and long term. However, in practice Overall, a balanced investment composition this is rarely the case. Information is usually is expected to deliver the best long-term not available to everyone, and the credibility of development results in Kenya. The simulations the information may be far from perfect. As a in this section show that a BC investment result, investors and households may act in an approach brings the highest boost to non- inefficient manner to avoid private costs. These resource GDP and leads to favorable fiscal actions may establish a self-fulfilling prophecy outcomes. This outcome is derived from the where investors and households do not believe economic principle of diminishing returns to in the announced policies, and the announced investment, which is especially true when there policies fail to be implemented because investors are implementation constraints. Therefore, even and households do not believe in them and do if public investments are scaled up rapidly, in not act accordingly. Therefore, the next step in the absence of accompanying improvements using the natural resources for development is to in public investment efficiency and matching establish a credible institutional mechanism for buildup of private and human capital, resources making and communicating policy choices on the are likely to be wasted. use of resource revenue. While aiming for a balanced investment How to Implement Spending and composition over time, it should be Allocation Decisions? acknowledged that the optimal composition Decisions for effective scale and composition of might change over time. Demography is one of the investments should be supported by establishing main factors that will influence the composition the right implementation framework to manage of investment. As Kenya’s population pyramid expectations. The recent oil discoveries have changes over the next few decades, investment already led Kenyan citizens to expect immediate decisions will need to take into account the need improvements in their livelihoods. If managed to provide education to the large cohorts. Then, well, the resource revenues will translate into as these educated youths enter the labor market, better livelihoods. However, the magnitude and the economy will benefit from infrastructure speed of this translation may not match the investment that would improve the economy’s expectations. This implies that making the right competitiveness and enable firms to grow and decisions is necessary but not sufficient to achieve create jobs. the desired outcomes in Kenya. These decisions also need to be communicated in a convincing In practice, choosing the right decisions for manner. Policy makers are advised to develop the the expenditure scale and composition are right institutional and communications framework necessary but not sufficient for the optimal for effective management of expectations in use of resources. The simulation exercises for government institutions and the public. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 109 Experience has shown that weak institutions agenda, and enhance the credibility of public and powerful interest groups can lead to plans. Policy makers are advised to act swiftly adverse development outcomes from resource and boldly in establishing clearly defined rules discoveries. In the absence of strong institutions, and mechanisms that ensure compliance. Box 5.2 interest groups such as rival ethnicities or discusses several examples of fiscal rules from subnational governments may lead to graft. other resource rich countries. Once the natural resource revenues start flowing into the economy, the competition between International practice has shown that effective these groups leads to a reduction in the quality fiscal rules have several common characteristics. of investment in renewable capital such as A useful framework that summarizes the physical and human capital. This “race to the characteristics of fiscal rules is provided by bottom” would mainly be driven by the motive of Kopits and Symanksy (1998). The fiscal rules are appropriating the windfall revenues before other the following: power groups do. This mechanism, known as the * Clarity. The target instrument, coverage of “voracity effect,” turns the resource discovery the rule, and institutional responsibilities into an impoverishing process. The solution to should be well defined. In the case of this problem lies in the policy makers’ ability using oil revenues to finance development to commit to a policy framework that intends projects in Kenya, the amount of oil revenue to work for all Kenyan citizens, and reduce the transfers to the fiscal budget would provide concerns regarding the possibility of being left a good target instrument. The chosen fiscal out from the common pool. rule should clearly assign the trajectory of transfers independent of short-term price The only way out from this bad equilibrium is to movements. Moreover, a legal framework strengthen public institutions. The voracity effect that defines the flow of funds between the deepens when policy makers cannot commit government agencies, responsibilities, and to a publicly known strategy that provides the decision making criteria should be in place. required transparency and predictability. In this case, the fear of being left out from the benefits * Transparency. Operations and actions to of a common pool would become self-fulfilling. ensure compliance should be transparent. As a result, re-establishing the policy credibility The public should be able to understand would become much more difficult compared the rule and decision making criteria and be with the periods earlier in resource boom. informed on the flow of funds. * Adequacy. The fiscal rule should be sufficient Establishing fiscal rules early on and strictly to achieve the designated targets. If the committing to targets would serve well to fiscal rule aims to provide accumulation of enhance policy credibility. Fiscal rules impose savings and macroeconomic stabilization, long-lasting constraints on discretionary fiscal then the transfers to the fiscal budget actions. These are publicly announced numerical should be disconnected from the short- targets, which aim to correct the distorted term price fluctuations. PIH and BIH rules incentives, limit the voracity effect and other provide this aspect with the condition of no pressures to overspend, show the commitment frequent revaluation. of the policy makers to an economic and political 110 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M * Consistency. The fiscal rule should be broadly A first step in the right direction for transparency consistent with the designated targets and be is to implement the EITI standard. Guided by in line with other economic policies. Chosen the EITI principles established in 2003, the EITI mechanisms should avoid conflicting results standard promotes transparency in resource rich between the stabilization and sustainability countries. To this effect, several requirements targets of the fiscal rule. One way of doing must be adhered by the member countries. this is to anchor the reference oil prices to a These are (i) effective oversight by the multi- slow-moving process, like long-term moving stakeholder group; (ii) timely publication of EITI averages (structural prices). reports; (iii) EITI reports that include contextual information about the extractive industries; (iv) * Simplicity. The rule should be simple, that is, production of comprehensive EITI reports that easy to be understood and implemented. include full government disclosure of extractive Fixed transfer rules, such as BIH and PIH, industry revenues, and disclosure of all material are relatively simple to communicate. payments to the government by oil, gas, and Low-income countries should stay clear of mining companies; (v) credible assurance complex rules that require a high degree process in applying international standards; (vi) of statistical and implementation capacity EITI reports that are comprehensible, actively (such as the structural balance rule in Chile; promoted, publicly accessible, and contribute see box 5.2). to public debate; and (vii) multi-stakeholder * Flexibility. The fiscal rule can carry group to take steps to act on lessons learned contingency options to facilitate long-term and review the outcomes and impact of EITI compliance without causing a breach. In implementation. The Government of Kenya is exceptional circumstances, such as the recommended to become a member of the EITI, global crisis of 2008, the fiscal rule should and go beyond these requirements to enhance enable a certain degree of flexibility to cope institutional quality. with extraordinary hardships. However, this flexibility should be limited to rare events, The other important issue in managing time- and the procedures that trigger such escape varying resource flows is accumulating the clauses and terminate them should be clearly savings in a sovereign wealth fund. Resource defined in advance. rich countries typically use specialized funds to * Enforceability. There should be clear offset the difference between oil receipts and mechanisms to enforce compliance. The transfers to the budget as assigned by the fiscal designers of the fiscal rule and supporting rule. In the case of a surplus, which typically institutional framework should be particularly occurs with high oil prices, the excess revenues careful in creating enforcement mechanisms. are transferred to a sovereign wealth fund to Conflict of interest among the agencies, such be invested in financial assets abroad. If the as the implementing agencies, by design, revenues fall short of the designated transfer should be understood and the political and amount, the funds flow in the opposite direction, economic pressures should be minimized. from the sovereign wealth fund to the fiscal budget. Therefore, the sovereign wealth fund helps to accumulate national savings and isolate the economy from fluctuations in resource revenues by complementing the fiscal rule. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 111 Box 5.2. Fiscal Rule Examples The following experiences in other countries can serve as benchmarks for planning the institutional framework in Kenya. Chile’s Structural Balance Rule In 2006, Chile’s Fiscal Responsibility Law institutionalized one of the most praised fiscal rules in effect today. The framework characterizes a complex set of mechanisms, including transfer rules and automatic stabilizers that enable long-term sustainability and short-term macroeconomic stabilization. The structural balance rule limits government spending by structural revenues. In the case of natural resource revenues falling short of the target levels (or exceeding them), the differences are compensated by transfers between the two sovereign wealth funds and the fiscal budget. Each year, a minimum of 2 percent of the previous year’s gross domestic product (GDP) is transferred to the Pension Reserve Fund, which accumulates national savings to cover long-term pension liabilities. This transfer can be increased by another 0.5 percent if the natural resource revenues exceed the structural target substantially. Any remaining surplus is transferred to the Economic and Social Stabilization Fund. In the case of a fiscal deficit, which may be driven by low growth performance or low resource prices, withdrawals from this fund cover the gap. In general, both funds invest only in financial assets abroad to avoid overheating the economy (World Bank 2014a). The design and commitment of the Chilean government have turned this practice into a success story. However, implementation of such a complex rule requires strong institutional capacity. To estimate the structural revenues the authorities would need to project the long-term prices of natural resources, interest earnings on financial assets, and the trajectory of potential GDP going forward. In many low-income countries, a simpler rule could prove to be more transparent and implementable. Norway’s Bird-in-Hand Starting in 2001, Norway’s fiscal rule has limited the structural non-oil deficit by an amount equivalent to the real return on resource revenue savings. Accordingly, the revenues net of transfers to the budget are transferred to the Global Government Pension Fund, which serves stabilization and saving purposes by investing in financial assets abroad. As such, the fiscal rule in Norway enables spending to increase gradually without depleting the savings, which are crucial for fiscal sustainability in the long term. For mature producers that are close to the end of their extraction cycle, the bird-in-hand rule enables a sustainable expenditure path. However, this requires substantial savings to be accumulated before the depletion. In Norway, these savings already exceed 100 percent of GDP. In comparison, in low-income countries with relatively young natural resource industries, the bird-in-hand rule may constrain spending on much-needed public goods in the earlier phases of the extraction cycle. Kazakhstan’s Modified Permanent Income Rule Kazakhstan’s sovereign wealth fund, the National Fund of the Republic of Kazakhstan (NFRK), serves a dual purpose of stabilization of the macro-economy and accumulation of national savings. It was established in 2000, and has been modified several times. The more recent modifications were implemented in 2007 and 2012. Since 2007, the NFRK receives all fiscal revenues from gas, oil, and four metals (chrome, zinc, lead, and copper), and makes disbursements to the budget, as enacted by law. 112 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M In 2012, the annual transfer of funds to the budget was changed from the previously fixed amount of US$8 billion to the flexible amount of US$8 billion plus or minus 15 percent (US$6.8 billion to US$9.2 billion), depending on the cyclical position of the economy. In 2012, it was decided that the next three budgets, subject to revision, will receive a transfer of US$9.2 billion, keeping the transfers pegged at the upper end of the range that has been specified. Timor-Leste’s Sustainable Income Rule Timor-Leste’s Petroleum Fund was established in 2005 with an objective of prudently managing the petroleum resources and mitigating the risks to the budget and economy from fluctuations in oil revenues. For this purpose, the Petroleum Fund collects all oil revenues, which constitute more than 70 percent of the economy (McKechnie 2013). Transfers from the Petroleum Fund to the fiscal budget are determined by a modified permanent income rule called Estimated Sustainable Income. Accordingly, each year the Government of Timor-Leste calculates the country’s petroleum wealth using conservative long-term oil price projections, existing Petroleum Fund savings, and the discounted sum of future productions. A maximum of 3 percent of this petroleum wealth is then transferred to the fiscal budget annually. Exceptions that involve transfers exceeding this ceiling are subject to parliamentary approval. Although this rule has been applied successfully in Timor-Leste, it is sensitive to projection biases. Long- term averaging of the prices helps smooth the expenditures by preventing sharp movements. Similarly, limiting the expenditures to a fraction of total oil wealth provides a sustainable approach. However, frequent updates in long-term oil price projections based on current market data could also make public expenditures pro-cyclical. The Government of Timor-Leste has been careful by taking a conservative approach so far. In addition, transparency of the fund, clear definition of responsibilities, and effective management by the Central Bank have led independent evaluators to consider this a success story. The institutional structure of sovereign wealth revenues. Therefore, these funds invest in liquid funds can be adjusted to the country’s needs, assets. In comparison, the National Fund of the and for most low- and middle-income countries, Republic of Kazakhstan combines the saving a single fund would be sufficient. Some countries and stabilization purposes under the same have designed multi-layered organizational roof; however, it also diversifies the investment structures for sovereign wealth funds (box 5.2). portfolio into higher risk or more liquid assets as Typically, these layers separate the saving and needed. Recent studies have shown that a single stabilization functions, which are reflected in fund with a diversified role tends to perform their investment portfolio as well. In Chile and better than a multi-layered structure (IMF 2012). the Russian Federation, the savings funds are mainly responsible for long-term investments However, sovereign wealth funds alone do of savings that correspond to the long-term not guarantee savings and stabilization unless social security commitments of the government. they are complemented by a firm fiscal rule. Stabilization funds, by contrast, serve as a buffer Just as opening a bank account does not ensure zone between long-term savings and short-term savings for households, establishing a sovereign needs by offsetting the fluctuations in resource wealth fund itself would not deliver savings and FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 113 stability without fiscal discipline. International As with the fiscal rule, proper institutional evidence has been mixed about the success of arrangements are crucial to operate the sovereign wealth funds. However, it has been sovereign wealth fund efficiently. In low- and clearly shown that the saving and stabilization middle-income countries, management of the functions of the sovereign wealth funds have sovereign wealth fund is generally provided by been more successful in countries with sound the central bank under clearly defined roles macroeconomic management and a strong and responsibilities. This is mainly because the commitment to fiscal discipline (Fasano 2000). In financial management capacities of the central the absence of a legal and regulatory framework bank are the highest among all government that clearly defines the conditions under which agencies. However, proper coordination with revenues can flow in and out of the fund, the fiscal and monetary policies and selection and fund would typically become subject to pressures implementation of the most appropriate fiscal to underwrite unjustified projects. rules will also require an improvement in the capacity of other government units. 114 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Taking into Consideration the Institutional Constraints: The Case of Stabilizing a Resource Rich Economy Volatility Although countercyclical policies are desirable in principle, Onder and Volatility of resource revenues, if Ley (2013) emphasize an important transmitted to the economy, can limitation against the implementation destabilize the economy and lead to of such policies: the policy makers permanent damage. Van der Ploeg may not reliably observe the cyclical and Poelhekke (2009) estimate that if position of the economy in real time.87 resource rich economies in Africa could This limitation is more prominent in reduce their macroeconomic volatility countries with low institutional capacity to the levels observed in the East Asian and high informality. Tigers, the economies of Africa would SPOTLIGHT 5 gain 3 percentage points in annual Estimating the Output Gap Is Easy; growth rates on average.86 Kenya is one Getting It Right Is Not So Easy of the African economies that have seen increasing volatility over the past decade. Accurately assessing the cyclical position of a non-resource economy in real Regulating the injection of resource time is a very challenging task. This revenues into the economy can improve can be shown by contrasting real-time macroeconomic stability. In practice, predictions with the final data for the this can be done by committing to output gap. Figure 5.8 displays the result fiscal rules that disconnect public of such an exercise by using International expenditures from movements in Figure 5.8: Predicted output gaps in real time resource revenues. In countries with versus actual output gaps, 175 countries, 1990–2011 institutional capacity constraints, these rules are recommended to be simple, transparent, and easily communicable. In several cases, policy makers have assumed a more pro-active position to stabilize the economy. Kazakhstan, for instance, introduced 15 percent flexibility in the permanent income annuities to enable countercyclical actions using discretionary policies. Source: Elaborated from World Bank Enterprise Survey data 2014. Van der Ploeg and Poelhekke (2009). 86 Onder and Ley (2013). 87 FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 115 Monetary Fund World Economic Outlook may actually increase the volatility (IMF-WEO) data for a cross-section of of an economy instead of decreasing countries, which cover more than two it. To elaborate on this point, Onder decades. For each year, the predictions and Ley (2013) run a Monte Carlo correspond to the projections in the simulation using a two-sector model previous year’s fall IMF-WEO. These with uncertainty concerning the precise predictions are subsequently revised cyclical position of the economy. The and the final numbers correspond to model is calibrated for a resource rich the most recent vintage in the data set economy. According to this framework, (IMF 2011). shocks cause the output to deviate from its trend, which, in turn, triggers the If the predictions were reasonably natural stabilizing forces in the economy. good, scatter plots would lie along the The fiscal authority may consider the diagonal, as predicted values would adjustment speed too low in this case, be equal to actual values. However, as and wish to use discretionary fiscal the figure shows, the dispersion is very policies to accelerate the recovery. The large. The correlation between the final efficiency of such policies (the size of SPOTLIGHT 5 and predicted gaps is less than 0.4, the fiscal multiplier) depends on the and deteriorates further as income per size of the actual output gap. However, capita decreases. In more than one- the actual position of the economy third of the cases, the predicted output along the cycle is unknown in real time, gap has the opposite sign of the final and the policy maker uses its estimate output gap estimates. In other words, to calculate the appropriate level of when the economy was estimated to countercyclical fiscal intervention. be underperforming and resources to be idle, in fact, the economy was Table 5.1 shows the volatility in the overheating (and vice versa). These non-resource sector under three policy are the points in the northwest and options: a no policy case (fixed rule), southeast regions in the figure. a moderately limited policy case, and a loosely limited policy case.88 As the The next question is about the extent degree of limitations loosens, the to which these measurement errors government’s degree of flexibility to affect the success of countercyclical counteract the cycle increases. This fiscal policies. exercise is also repeated for two different long-term allocation rules (PIH and BIH) With Biased Assessments, Policies under two different fiscal multiplier Can Affect the Cycle, but Not Necessarily As Intended assumptions (baseline and high). In all cases, average volatility increases as If output gap estimates are significantly greater levels of countercyclical policies biased, then countercyclical policies are allowed. The measure of volatility reported here is the standard deviation of variation around long-term potential GDP growth. 88 116 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Table 5.1: Predicted output gaps in real time versus actual output gaps, 175 countries, 1990– 2011 Approach Fixed rule Moderate limits Loose limits Permanent Income Hypothesis 0.27 0.36 0.47 Bird-in-Hand Rule 0.27 0.36 0.48 High Fiscal Multiplier Scenario Permanent Income Hypothesis 0.28 0.29 0.29 Bird-in-Hand Rule 0.27 0.31 0.32 Source: Automatic Stabilizers errors. In these cases, countercyclical policies that are less prone to discretion This analysis shows that institutional errors are found to be more effective. capacity is an important factor in Anchoring the policies by using rigid implementing countercyclical policies. fiscal rules and enhancing the automatic SPOTLIGHT 5 Low-income countries, particularly those that are at the early stages of the stabilizers, such as a progressive tax resource cycle, may face more problems system and social benefits, can go a long in implementing these policies, as way in providing policy credibility and they exhibit larger measurement smoothing the cycles at the same time. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 117 118 RAN K E N YA C O U N T R Y E C O N O M I C M E M O Photo: M D UBank World REFERENCES Aghion, P., D. Comin, and P. Howitt. 2006. “When Does Domestic Saving Matter for Economic Growth?” NBER Working Paper No. 12275, National Bureau of Economic Research, Cambridge, MA. Aghion, P., and P.Howitt. 1992. “A Model of Growth through Creative Destruction.” Econometrica, 60 (2): 323–25. Akaike, H. 1974. “A New Look at the Statistical Model Identification.” I.E.E.E. Transactions on Automatic Control19: 716–723. ———. 1976. “Canonical Correlation Analysis of Time Series and the Use of an Information Criterion.” In System Identification: Advances and Case Studies, edited by R. K. Mehra and D. G. Lainotis, 52–107. New York:Academic Press. Argent, Jonathan, and Tania Begazo. 2015. “Competition in Kenyan Markets and Its Impact on Income and Poverty: A Case Study on Sugar and Maize.” Policy Research Working Paper 7179, World Bank, Washington, DC. Balassa, B. 1986. “Policy Responses to External Shocks in Developing Countries.” American Economic Review 76 (2): 75–78. Barro, R., and J.W. Lee. 2001. “International Data on Educational Attainment: Updates and Implications.” Oxford Economic Papers 53(3): 541–63. Berg, A., R. Portillo, S. S. Yang, and L.-F. Zanna. 2013. “Public Investment in Resource-Abundant Developing Countries.” IMF Economic Review 61 (1): 92–129. Bloom, Nicholas, and John Van Reenen. 2010. “Why Do Management Practices Differ across Firms and Countries?” Journal of Economic Perspectives 24 (1): 203–24. Buffie, E. F., A. Berg, C. Pattillo, R. Portillo, and L.-F. Zanna. 2012. “Public Investment, Growth, and Debt Sustainability: Putting Together the Pieces.” IMF Working Paper 12/144, International Monetary Fund, Washington, DC. Calderon, Fainzylber, and Loayza 2005. “Economic Growth in Latin America and the Caribbean: Stylized Facts, Explanations, and Forecasts.” Washington, DC: World Bank. Commission on Growth and Development. 2008. “The Growth Report: Strategies for Sustained Growth and Inclusive Development.” Washington, DC: World Bank. Cushman, David O., and Tao Zha. 1997. “Identifying Monetary Policy in a Small Open Economy under Flexible Exchange Rates.” Journal of Monetary Economics 39: 433–48. Dabla-Norris, Era, Richard Allen, Luis-Felipe Zanna, Tej Prakash,Eteri Kvintradze, Victor Lledo, Irene Yackovlev, and Sophia Gollwitzer.2010.“Budget Institutions and Fiscal Performance in Low-Income Countries.” IMF Working Paper WP/10/80, International Monetary Fund, Washington, DC. Dupas, P., and J. Robinson. 2013a. “Savings Constraints and Microenterprise Development: Evidence from a Field Experiment in Kenya.” American Economic Journal: Applied Economics 5 (1): 163–92. ———. 2013b. “Why Don’t the Poor Save More? Evidence from Health Savings Experiments.” American Economic Review 103 (4): 1138–71. Eden, M., and A. Kraay. 2014. “‘Crowding In’ and the Returns to Government Investment in Low-Income Countries.” Policy Research Working Paper 6781, World Bank, Washington, DC. Eifert, B., A. Gelb, and V. Ramachandran. 2008.“The Cost ofDoing Business in Africa: Evidence from Enterprise Survey Data.” World Development 36 (9): 1531–46. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 119 Falco, Paolo, Andrew Kerr, Pierella Paci, and Bob Rijkers. 2014. “Working toward Better Pay: Earnings Dynamics in Ghana and Tanzania.” Washington, DC: World Bank. Farole, Thomas. 2011. “Special Economic Zones in Africa: Comparing Performance and Learning from Global Experience.” Washington, DC: World Bank. Farole, T., and M. Mukim. 2013. “Manufacturing Export Competitiveness in Kenya. Policy Note: NLTA on Revitalizing and Diversifying Kenya’s Manufacturing Sector.” Report ACS6849, World Bank, Washington, DC. Fasano, U. 2000. “Review of the Experience with Oil Stabilization and Savings Funds in Selected Countries.” IMF Working Paper Series No. 112, International Monetary Fund, Washington, DC. Gelb, A., C. Meyer, and V. Ramachandran. 2013. “Does Poor Mean Cheap? A Comparative Look at Africa’s Industrial Labor Costs.” Working Paper 325, Center for Global Development, Washington, DC. ———. 2014. “Development as Diffusion: Manufacturing Productivity and Sub-Saharan Africa’s Missing Middle.” Working Paper 3517, Center for Global Development, Washington, DC. ———. 2014.“Development as Diffusion: Manufacturing Productivity and Sub-Saharan Africa’s Missing Middle.” WIDER Working Paper 2014/042. World Institute for Development Economics Research, Helsinki, Finland. Ghosh, S. R., and A. Kraay. 2000. “Measuring Growth in Total Factor Productivity.” Premnotes No. 42, World Bank, Washington, DC. Giugale, M. and N. Nguyen. 2014. “Money to the People: Estimates of the Potential Scale of Direct Dividend Payments in Africa.” Center for Global Development Policy Paper 043. Gollin, D., D. Lagakos, and M. E. Waugh. 2013. “The Agricultural Productivity Gap.” Quarterly Journal of Economics 129(2): 939–93. Gupta, S., A. Kangur, C. Papageorgiou, and A. Wane. 2011. “Efficiency-Adjusted Public Capital and Growth.” IMF Working Paper WP/11/217, International Monetary Fund, Washington, DC. Haidar, Jamal Ibrahim. 2012. “The Impact of Business Regulatory Reforms on Economic Growth.” Journal of the Japanese and International Economies 26 (3): 285–307. Hall, B.H. 2011. “Innovation and Productivity.” Nordic Economic Policy Review. 2: 167–204. Hannan, E. J., and B. G. Quinn. 1979. “The Determination of the Order of an Autoregression,” Journal of the Royal Statistical SocietyB(41): 190–95. Harrison, R., J. Jaumandreu, J. Mairesse, and B. Peters. 2008. “Does Innovation Stimulate Employment? A Firm- level Analysis Using Comparable Microdata from Four European Countries.” NBER Working Paper No. 14216. National Bureau of Economic Research, Cambridge, MA. Hausmann, R., C. A. Hidalgo, S. Bustos, M. Coscia, S. Chung, J. Jimenez, A. Simoes, and M. Yildirim. 2011. “The Atlas of Economic Complexity.” Cambridge, MA: Puritan Press. Hausmann, R., J. Hwang, and D. Rodrik. 2007. “What You Export Matters.” Journal of Economic Growth 12(1): 1–25. Hidalgo, C., and R. Hausmann. 2009. “The Building Blocks of Economic Complexity.” Proceedings of the National Academy of Sciences 106 (26): 10570–75. Hidalgo, C., B. Klinger, A. Barabasi, and R. Hausmann. 2007. “The Product Space Conditions the Development of Nations.” Science Magazine 317(5837). 120 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Humphreys, Macartan, Jeffrey D. Sachs, and Joseph E. Stiglitz. 2007. “Escaping the Resource Curse.” New York: Columbia University Press. Iacovone, Leonardo, Vijaya Ramachandran, and Martin Schmidt. 2014. “Stunted Growth: Why Don’t African Firms Create More Jobs?” Working Paper No. 353, Center for Global Development, Washington, DC. IMF (International Monetary Fund). 2011. “World Economic Outlook.” April. IMF, Washington., DC. ———. 2012. “Macroeconomic Policy Frameworks for Resource-Rich Developing Countries.” IMF, Washington, DC. ———. 2014. “Kenya: Joint World Bank/IMF Debt Sustainability Analysis.” In Kenya: 2014 Article Iv Consultation staff report, press release, and Statement By The Executive Director For Kenya. IMF Country Report No. 14/302, IMF, Washington, DC. Johnson, Simon, Jonathan D. Ostry, and Arvind Subramanian. 2006. “Levers for Growth: Policy Lessons from Earlier Bouts of Growth in Developing Countries.” Finance and Development43 (1): 28–32. ———. 2007. “The Prospects for Sustained Growth in Africa: Benchmarking the Constraints.” NBER Working Paper No. 13120, National Bureau of Economic Research,Cambridge, MA. Karlan, D., and J. Morduch. 2009. “Access to Finance.” InHandbook of Development Economics, edited by D. Rodrik and M. Rosenzweig, 4703–84. Amsterdam: Elsevier. Kenya National Bureau of Statistics. 2015. “Economic Survey.” Kenya National Bureau of Statistics,Nairobi. Kenya Open Data. 2009. https://opendata.go.ke/. Kimenyi, M. S., Manda, D. K., and G. Mwabu. 2006. “Human Capital Externality and Returns to Education in Kenya.” Eastern Economic Journal 32: 493–514. Knight Frank, and Citi Private Bank. 2012. “The Wealth Report: A Global Perspective on Prime Property and Wealth.” London: Knight Frank. Kopits, G., and S. Symansky. 1998. “Fiscal Policy Rules.” IMF Occasional Papers No. 162, International Monetary Fund, Washington, DC. Kuepie, M., C. J. Nordman, and F. Roubaud. 2009. “Education and Earnings in Urban West Africa.” Journal of Comparative Economics 37: 491–515. Laibson, D. 1997. “Golden Eggs and Hyperbolic Discounting.” Quarterly Journal of Economics 112 (2): 443–77. Lall, Sanjaya. 2000. “The Technological Structure and Performance of Developing Country Manufactured Exports, 1985-98.” Oxford Development Studies 28 (3): 337–69. Lederman, Daniel, and William F. Maloney. 2012. “Does What You Export Matter? In Search of Empirical Guidance for Industrial Policies.” Latin American Development Series. Washington, DC: World Bank. Levine, P., G. Melina, and H. Onder. 2014. “Fiscal Rules in a DSGE Model of Kenya.” Unpublished. (H) Levine, P., G. Melina, and H. Onder. 2015. “Fiscal Rules in a DSGE Model of Kenya.” Unpublished background paper to Kenya Country Economic Memorandum, World Bank, Washington, DC. Loayza, Norman, Klaus Schmidt-Hebbel, and Luis Serven. 2000. “What Drives Private Saving Across the World?” Review of Economics and Statistics 82: 2: 165–81. McKechnie, A. 2013. “Managing Natural Resource Revenues: The Timor-Leste Petroleum Fund.” London: Overseas Development Institute Research Report. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 121 MacMillan, Margaret, and Dani Rodrik. 2012. “Globalization, Structural Change, and Productivity Growth.” IFPRI Discussion Paper 1160, International Food Policy Research Institute, Washington, DC. Melina, G., S. C. Yang, and L. F. Zanna. 2014. “Debt Sustainability, Public Investment and Natural Resources in Developing Countries: The DIGNAR Model.” IMF Working Paper 14/50, International Monetary Fund, Washington, DC. Melina, G. and Xiong, Y. 2013. “Natural Gas, Public Investment and Debt Sustainability in Mozambique.” IMF Working Paper Series WP/13/261. Modigliani, Franco. 1970. “The Life-Cycle Hypothesis and Intercountry Differences in the Saving Ratio.” InInduction, Growth, and Trade: Essays in Honour of Sir Roy Harrod, edited by W. A. Eltis, M. Fitzgerald Scott, and J. N. Wolfe, 197–225. Oxford: Oxford University Press. Murthi, M. and L. Sondergaard. 2012. “Skills, Not Just Diplomas: The Path for Education Reforms in ECA.“ Washington, DC: World Bank. Omolo, O. J. 2010. ”The Dynamics and Trends of Employment in Kenya.” IEA Research Paper Series, No. 1/2010, Institute of Economic Affairs, Nairobi. Onder, H., and E. Ley. 2013. “The National Fund of Republic of Kazakhstan and Countercyclical Fiscal Policy.” Background paper for Kazakhstan Oil Rules: Kazakhstan’s Policy Options in a Downturn. Report No. 81686-KZ, World Bank, Washington, DC. PricewaterhouseCooper. 2015. “Towards a Petroleum Sector Master Plan for Kenya.” PWC, Washington, DC. Raddatz, C. 2007. “Are External Shocks Responsible for the Instability of Output in Low-Income Countries?” Journal of Development Economics84(1):155–87. Rodrik, D.2015. “Premature Deindustrialization.” John F. Kennedy School of Government, Harvard University, Cambridge, MA. Romer, J.P. 1986. “Increasing Returns and Long-Run Growth.” Journal of Political Economy 94 (5): 1002–37. Sáez, S., D. Taglioni, E. van der Marel, C. H. Hollweg, and V. Zavacka. 2014. “Valuing Services in Trade: A Toolkit for Competitiveness Diagnostics.” Washington, DC: World Bank. Samuelson, W., and R. J. Zeckhauser. 1988. “Status Quo Bias in Decision Making.” Journal of Risk and Uncertainty 1: 7–59. Schwarz, G. 1978. “Estimating the Dimension of a Model.” Annals of Statistics 6: 461–64. Solow, M.R. 1957. “Technical Change and the Aggregate Production Function.” Review of Economics and Statistics 39 (3): 312–20. Suri,T., D. Tschirley, C. Irungu, R. Gitau, and D. Kariuki. 2008. “Rural Incomes, Inequality and Poverty Dynamics in Kenya.” WPS 20/2008, Tegemeo Institute of Agricultural Policy and Development, Nairobi. Thaler, R. 1981. “Some Empirical Evidence on Dynamic Inconsistency.” Economic Letters 8 (3):201–07. Transparency International.2010. “Corruption Perceptions Index.” Transparency International, Berlin. ———.2014. “Corruption Perceptions Index.” Transparency International, Berlin. Tversky, A., and D. Kahneman. 1992. “Advances in Prospect Theory: Cumulative Representation of Uncertainty.” Journal of Risk and Uncertainty5: 297–323. 122 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M UNDP (United Nations Development Programme). 2013. “Kenya’s Youth Employment Challenge.” New York: UNDP. Van der Ploeg, F., and S. Poelhekke. 2009. “Volatility and the Natural Resource Curse.” Oxford Economic Papers 61: 727–60. World Bank. 2009. “Kenya Poverty and Inequality Assessment: Executive Summary and Synthesis Report.” World Bank, Washington, DC. ———. 2011. “Navigating the Storm, Delivering the Promise.” Kenya Economic Update No. 5. World Bank, Washington, DC. ———. 2012a. “Rwanda Economic Update: Managing Uncertainty for Growth and Poverty Reduction” World Bank, Washington, DC. ———. 2012b. “Strengthening Governance: Tackling Corruption.” The World Bank Group’s Updated Strategy and Implementation. World Bank, Washington, DC. ———. 2013a. Doing Business 2013: “Smarter Regulations for Small and Medium-Size Enterprises.” Washington, DC: World Bank. ———. 2013b. “Where Has All the Growth Gone? A Poverty Update for Nigeria.” World Bank, Washington, DC. ———. 2013c. Kenya Economic Update: Poverty. ———. 2014a. Doing Business 2014: “Understanding Regulations for Small and Medium-Size Enterprises.” Washington, DC: World Bank. ———. 2014b. “Anchoring High Growth: Can Manufacturing Contribute More?.” Kenya Economic Update Edition 11. World Bank, Washington, DC. ———. 2015a. Doing Business 2015: “Measuring Regulatory Quality and Efficiency.” Washington, DC: World Bank. ———. 2015b. World Development Report 2015: “Mind, Society, and Behavior,” Vol. 1. Washington, DC: World Bank. ———. 2015b. Food prices. World Economic Forum. 2013. “The Global Competitiveness Report 2013-14.” World Economic Forum, Geneva. Zeufack, A. and Woodroffe 2013. “A New Approach to Engaging On Extractives: Lessons from Tanzania. Unpublished.” World Bank. Washington, DC. Zollman, J. 2014. “Kenya Financial Diaries: Shilingi Kwa Shilingi, the Financial Lives of the Poor.” Financial Sector Deepening, Nairobi. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 123 Appendix A: Examination of exogenous and endogenous shocks to the Kenyan economy The applied method for examining the size and lag effects of exogenous and endogenous shocks on the economy is based on the work of Raddatz (2007). It begins by identifying the economies that unilaterally influence Kenya’s economy (no significant reverse impact). The list includes Kenya’s major exogenous export partners, the European Union, India, Pakistan, and the United States. Foreign effective demand for Kenya (GDP growth rates of the aforementioned economies) is constructed with normalized weights that sum to one, computed according to the average export share of Kenya’s exports of goods and services in 2009–12. Then, foreign effective inflation is constructed (consumer price index--based inflation) using the same approach, this time using the main exogenous partners from which Kenya imports goods and services, China, the European Union, India, Japan, South Africa, and the United States. The assessment of the transmission of the shocks is based on impulse response functions and variance decomposition methods for which bivariate vector autoregression (VAR) models are employed. VAR is one of the most frequently used methods in the empirical literature for examining the size and the time-lag of the reaction of one variable to a shock in another. Because the core aim of the analysis is to assess the transmission of the exogenous shocks, by following the approach of Cushman and Zha (1997) and Raddatz (2007), the so-called block exogeneity assumption is imposed in the VARs.89 The selection of the VAR models is done according to the common approach in the literature.90 The same methodological approach can be utilized to assess how domestic (endogenous) shocks affect the economy. To this end, the impacts of shocks from the following factors were examined: investment (gross and fixed investments, and net foreign direct investment inflows), fiscal policy outcomes (total government expenditure, final government consumption, budget deficit, and public debt expressed as ratios of GDP), domestic inflation, population growth, and human capital variables (average years of schooling and gross secondary school enrollment). The analysis is again based on bivariate VAR models, but without the imposition of the block exogeneity assumption due to endogeneity issues. 89 The imposition of the block-exogeneity assumption in the VARs means that the two-side effects of the variables that are included in the model are restricted. For example, in the usual VAR framework the two-way causation is assessed between the variables used (X and Y), whereas the block exogeneity assumption allows only the one-way impact of the variables to be assessed, that is, from the exogenous (X) to the endogenous (Y) variable, but not the other way around, as is allowed under the usual VAR framework. 90 First, the lag length of the variables is chosen by the information selection criteria (Akaike (1974, 1976), Schwarts (1978), and Hannan and Quinn (1979)), and then the residual diagnostic tests are conducted for serial correlation, normal distribution, and autoregressive conditional heteroskedasticity. After the model has been specified, the impulse response functions are estimated with bootstrap confidence intervals of 100 repetitions. 124 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M Table A.1. Summary of the Results of the Impulse Response Functions and Variance Decomposition of the Impact of Shocks on Kenya’s GDP per Capita Growth and Inflation GDP per capita growth of Kenya   Impulse response functions Variance decomposition   Size of the impact Time horizon Explaining the variance of GDP (in percentage points) per capita growth Exogenous variables:       Foreign effective GDP per capita 0.6 0 to 1 years Between 13% and 15% growth Endogenous variables:       Government final consumption 0.6 to 1.5 1 to 3 years Between 15.5% and 20.9% (1st diff) Gross investment (1st diff) 0.7 to 2.5 1 to 6 years Between 35.3% and 36.2% Inflation -1.1 to -1.6 1 to 2 years Between 23% and 24% Inflation in Kenya Impulse response functions Variance decomposition   Size of the impact Time horizon Explaining the variance of GDP (in percentage points) per capita growth Foreign effective inflation (CPI- 6.1 to 9.2 0 to 4 years Between 70% and 71% based inflation)a World food price inflationb 2.4 to 10 2 to 10 quarters Between 31% and 67% Sources: Calculations based on data from World Bank World Development Indicators, IMF World Economic Outlook, and Kenya National Bureau of Statistics. Note: CPI = consumer price index; GDP = gross domestic product; IMF = International Monetary Fund. a. Foreign effective inflation is calculated as a geometric average of the CPI-based inflation from the exogenous importing partners of Kenya. b. World food price inflation is based on the inflation of food price index published by the IMF Primary Commodity Prices database. 89 The imposition of the block-exogeneity assumption in the VARs means that the two-side effects of the variables that are included in the model are restricted. For example, in the usual VAR framework the two-way causation is assessed between the variables used (X and Y), whereas the block exogeneity assumption allows only the one-way impact of the variables to be assessed, that is, from the exogenous (X) to the endogenous (Y) variable, but not the other way around, as is allowed under the usual VAR framework. 90 First, the lag length of the variables is chosen by the information selection criteria (Akaike (1974, 1976), Schwarts (1978), and Hannan and Quinn (1979)), and then the residual diagnostic tests are conducted for serial correlation, normal distribution, and autoregressive conditional heteroskedasticity. After the model has been specified, the impulse response functions are estimated with bootstrap confidence intervals of 100 repetitions. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 125 Appendix B: Simulation model for Kenya’s oil revenue The simulations in chapter 5 were performed by using a dynamic stochastic general equilibrium model developed by Levine, Melina, and Onder (2015). The model builds on Melina and Xiong (2013), Melina, Yang, and Zanna (2014), Berg et al. (2013), and Buffie et al. (2012), which analyze the public investment and growth nexus together with debt sustainability and natural resource revenue management in low- and middle-income countries. In addition to the modeling framework developed in these papers, the framework developed for Kenya incorporates the analysis of expenditure composition in public investments, which involves human capital components such as health and education spending as well. In particular, the framework is a small-economy model with limited asset market participation to capture the presence of agents that do not have access to financial markets in low- and middle-income countries. The production side of the model exhibits (i) a traded goods sector featuring learning- by-doing externalities to capture the effects of the Dutch disease that may arise because of natural resource booms; (ii) a non-traded goods sector; and (iii) a natural resource sector. A typical firm in the traded (T) and non-traded (N) goods sectors produces output, yj,t , j = {T, N} according to the technology yj,t = zj `kj,t-1j1-αj `Aj,tLj,tjαj`kG,t-1jαG where zj is a total factor productivity scale parameter, kj,t is end-of-period private capital, kG,t-1 is end- of-period public capital, αj is the labor share of sectoral income, and αG represents the output elasticity with respect to public capital. Labor productivity Aj,t is given by: Aj,t = zα,jet βj,eht βj,h where et represents the average education of the labor force, and ht represents the average health status of the labor force. The model also features inefficiencies and absorptive capacity constraints for public investment and a time-varying depreciation rate of public capital to capture lack of maintenance, in line with the empirical literature for low- and middle-income economies (see Gupta et al. 2011, among others). To reflect this, effective investment is assumed to take a particular functional form to enable the deviations of government investment expenditure from the initial steady state more than a threshold lead to a decrease in efficiency of the additional investment proportional to the size of the deviation. This mechanism captures absorptive capacity constraints in Kenya and comparable countries. Saving/spending scenarios are chosen on the basis of the policy options. These are determined by the country’s plans as documented by strategic documents such as Vision 2030 and the Medium- Term Plans, as well as alternative scenarios that are commonly observed in other countries. Public investment can be frontloaded and the degree of the frontloading is linked to the degree of investment inefficiency. 126 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M As far as fiscal policy is concerned, the model has a fund where any positive difference between inflows (including natural resource revenue) and outflows (including investment expenditures) is saved and the lower bound of this fund is a policy choice. The fund is drawn down when such a difference is negative. However, when the fund reaches a chosen lower bound, then one or more fiscal instruments react to close it either instantaneously or by temporarily allowing accumulation of public debt and satisfying the government intertemporal budget constraint in the long run. In the case of Kenya—where natural resource exploitation is a recent phenomenon and virtually no fiscal buffers have been accumulated yet—a lower bound of zero is set for the fund, which effectively becomes a non-negativity constraint for government assets. The model allows four fiscal instruments to close the fiscal gap (consumption tax, labor income tax, government consumption, and government transfers). For simplicity, where needed, only the consumption tax is allowed to stabilize debt in the long run and the other instruments are left at their initial steady state. Although the use of other instruments, combined or in isolation, implies somewhat different macroeconomic dynamics, the bottom-line of the results outlined below is robust to such choices. For complete details of the model, derivation of the equilibrium conditions, and calibration to Kenya, see Levine, Melina, and Onder (2014. FROM ECONOMIC GROW TH TO JOBS AND SHARED PROSPERIT Y 127 Appendix C: Estimating the potential growth rate in Kenya based on the Cobb-Douglas production function The estimation of the potential growth rate is Kenya’s gross domestic product, the values of based on a Cobb-Douglas production function the parameters of the model, α, γ, returns to that estimates potential output (Y) as a function education, and capital depreciation rate, need to of the stock of physical capital, human capital– be assumed. In addition, for the forecasting period, adjusted labor (H), and a given technology (A), that the initial and final values of the input variables determines the total factor productivity (TFP). This (working age population ratio, participation rate, can be presented with the following equation: unemployment rate, TFP growth rate, gross capital formation, and average years of schooling), need Y = A(Kα H1- α)γ (C.1) to be exogenously determined. The initial values of the input variables refer to the last available The superscripts α and 1- α indicate the share of data for the variables used or a historic average physical and human capital–adjusted labor in the of a certain variable for a specified period of time. output, respectively, whereas the superscript γ The final values of the input variables refers to measures the returns to scale of the inputs. For their expected values at the end of the forecasting example, γ can be equal to 1, greater or less than period (2020), which are usually exogenously set. 1, suggesting constant, increasing or decreasing returns to scale, respectively. The human capital– The baseline scenario assumes no structural shift adjusted labor (H) is calculated as a function of in the economic relations (the parameters). The the labor force adjusted for employment and parameters of the model are assigned values that participation rates and the level of human capital are already pre-defined in the economic literature. that is estimated as a function of return on The pre-defined parameters and initial and final education and the average years of schooling. values of the input variables used are presented In estimating the potential growth rate of in table C.1. Table C.1. Assumed Values of the Parameters and Initial and Final Values of the Input Variables in the Model Initial value Final value Parametera Value Input variable (2015) (2020) Share of physical capital (α) 35% Working age population ratio 55% 56.7% Share of human-adjusted capital (1-α) 65% Participation rate 70.3% 70.3% Returns to scale of the inputs (γ) 1 Unemployment rate 13.4% 13.4% Return of education 5% TFP growth rate 1.7% 1.7% Capital depreciation ratio 5% Gross capital formation 29.2% 31.8% Average years of schooling 6.7 years 7 years a. The values of the parameters are taken from the economic literature in this area: Ghosh and Kraay (2000) and Kuepie, Nordman, and Roubaud (2009). 128 K E N YA C O U N T R Y E C O N O M I C M E M O R A N D U M The World Bank Delta Center Menengai Road, Upperhill PO Box 30577-0100 Nairobi, Kenya JOIN THE CONVERSATION! Tel: +254 20 2936000 www.facebook.com/worldbankkenya Fax: +254 20 2936382 Website: www.worldbank.org/kenya www.twitter.com/worldbankkenya