June 2018 DOMINICAN REPUBLIC Systematic Country Diagnostic Dominican Republic Systematic Country Diagnostic Front Cover: Nighttime Lights Composite, 2013. Credits: Earth Observation Group, NOAA National Geophysical Data Center. Back Cover: Nighttime Lights Composite, 1992. Credits: Earth Observation Group, NOAA National Geophysical Data Center. Acknowledgements We thank the members of the Dominican Republic Country Team from all Global Practices, IFC, and MIGA, as well as all the partners and stakeholders in the Dominican Republic, who contributed throughout the preparation of this Systematic Country Diagnostic (SCD). We are grateful for their inputs, knowledge and advice. The team was co-led by Juan Carlos Parra (Senior Economist, GPVDR) and Samuel Pienknagura (Senior Econ- omist, GMTLC), under the guidance of Tahseen Sayed (Country Director, LCC3C). The team has collaborat- ed closely with the task leads from IFC, Guillermo Villanueva (Resident Representative DR), and Bexi Francina Jiménez Mota (Risk Management Officer) from MIGA. The core team that prepared this report included Alessandro Legrottaglie (Country Manager, LCCDO), Oscar Ishizawa (Senior Disaster Risk Management Specialist, GSURR), Juan Barón (Senior Education Economist, GED), Melissa Adelman (Senior Education Economist, GED), Rémi Trier (Senior Water Resources Specialist, GFA), John Anderson (Senior Economist, GFCLC), Raha Shahidsaless (Senior Private Sector Specialist, GFCLC), Daniel Reyes (Senior Economist, GMTCI), Edouard Al-Dahdah (Senior Public Sector Specialist, GGO), Marco Larizza (Senior Public Sector Specialist, GGO), Daniela Felcman (Public Sector Specialist, GGO), Roland Clarke (Lead Public Sector Specialist, GGOLP), Davide Zucchini (Senior Public Sector Specialist, GGOLP), Denny Mahalia Lewis-Bynoe (Senior Economist, CCECE), Luis de Pool (Associate Operations Officer, CLADH), Antonio Skarica (Strategy Analyst, CCECE), Maritza Rodríguez (Senior Financial Management Specialist, GGOLF), Christine Lao Peña (Senior Human Development Economist, GHN), Miriam Montenegro (Senior Social Protection Specialist, GSP), Christian Gómez Cañón (Consultant, GPV), Marco Co- banera (Consultant, GMTLC), Cecilia Briceño-Garmendia (Lead Economist, GTD), Felipe Targa (Senior Urban Transport Specialist, GTD), Christian Borja-Vega (Economist, GWA), Craig Kullman (Senior Water Supply and Sanitation Specialist, GWAGP), Marco Antonio Aguero (Senior Water Supply and Sanitation Specialist, GWAGP), Rafael Pardo (Financial Sector Specialist, GFCLC), Rodrigo Martínez Fernández (Natural Resources Economist, GEN), and Laura Calderón Etter (Consultant, GEN). The team received guidance and inputs for the preparation of this SCD from Tahseen Sayed (Country Director, LCC3C), Oscar Calvo-González (Practice Manager, GPVDR), Pablo Saavedra (Country Director, LCC1C), Ste- fano Curto (Senior Economist, GMTLC), Timothy Johnston (Program Leader, ECCWB), Abha Prasad (Program Leader, LCC3C), Vickram Cuttaree (Program Leader, EACPF), Cecile Niang (Lead Economist, GFCLC), Francis- co Carneiro (Practice Manager, GMTA1), Humberto López (Director Strategy and Operations, LCRVP), Carlos Végh (Chief Economist, LCRCE), Jorge Araujo (Economic Adviser, LCRVP), Daniel Lederman (Lead Economist, MNACE), and McDonald Benjamin (Adviser, LCROS). The team also received inputs from Alejandra de la Paz (Communications Associate, LCREC), Christelle Chapoy (Senior Communications Officer, LCREC), Maria Pia Cravero (Counsel, LEGLE), Carmen Amaro (Consultant, GGOLP), Gero Verheyen (Underwriter, MIGFR), Rafael Van der Borght, Paul Blanchard (Consultant, GSURR), Juan José Miranda (Environmental Economist, GENGE), Soufiane Hadji (Temporary, GSURR), Joaquín Muñoz Díaz (Consultant, GSURR), Xijie Lv (Consultant, GSURR), Andrés González (Consultant, GSURR), Karina Brito (Operations Analyst, GPV), Tambi Matambo (Senior Environmental Specialist, GCCRA), and Yunziyi Lang (An- alyst, GCCRA). The team received editorial support from Alexis Bernaut. The report was designed by Florencia Micheltorena. The team thanks the peer reviewers Samuel Freije-Rodríguez (Lead Economist, GPVDR), Tania Dmytraczenko (Program Leader, LCC2C), and Miguel Eduardo Sánchez Martín (Senior Economist, GMTDR), for their com- ments. Table of Contents 1. Overview 1 Geography, favorable external conditions, and the structural reforms of the nineties bolstered twenty-five years of high growth and social progress 1 Challenges to become a high-income country by 2030 and to improve social outcomes 3 Priority areas 7 2. Economic Performance and Growth 13 Twenty-five years of fast economic growth and convergence 13 Fast firm-level growth but little evidence of quality upgrading and weak links to exports 22 Challenges to competitiveness and inclusive growth 27 3. Drivers of inclusion 39 Recent progress on poverty reduction, shared prosperity and equality of opportunities 39 Migration and inclusion 55 Social policy and inclusion 58 4. Challenges to sustainability 67 Raising fiscal revenue to meet social demands and secure fiscal sustainability 67 Resilience to disasters, climate-related risks, and other shocks 71 Environmental sustainability 74 Social accountability 78 Social cohesion 80 5. Prioritization and knowledge gaps 83 Priority areas 83 Knowledge gaps 89 Annexes Annex 1. Economic and Social indicators 91 Annex 2. Benchmarking the Dominican Republic 93 Annex 3. Monetary policy in the DR 97 Annex 4. Characterizing Dominican exports 99 Annex 5. A closer look at SEZs, tourism, and agricultural products 105 Annex 6. A characterization of poverty by sex 109 Annex 7. Consultation process 111 Annex 8. Map of the Dominican Republic 113 References 114 List of figures Figure 1. Over the past 25 years, the Dominican Republic has converged to the living standards of high-income countries at a faster pace than the average country in LAC and the World 14 Figure 2. The DR’s recent growth episode has been accompanied by historically low levels of volatility 16 Figure 3. Drivers of growth in the DR 17 Figure 4. FDI inflows into the Dominican Republic 18 Figure 5. The service sector has been the main driving force of GDP growth and employment in the DR 20 Figure 6. The DR has a larger share of services in value added and employment compared to peers and predictions 21 Figure 7. Productivity has increased in all sectors, but employment flowed to lower productivity sectors 21 Figure 8. Sales and employment growth rates in the DR have been high compared to peers 23 Figure 9. Dominican firms lag those of other countries in terms quality upgrading, propensity to innovate and backwards linkages 24 Figure 10. Dominican firms create less backwards linkages compared to similar firms in other countries and exporters lag in quality 25 Figure 11. The DR has a higher export entry rate than comparators, but underperforms in the terms of survival 26 Figure 12. The DR suffers from frequent and long power outages, which have large economic costs 28 Figure 13. Real lending rates in the DR are higher than in comparators, potentially due to spillovers from government debt 32 Figure 14. The DR has a larger gap between prices at the destination and prices in the point of origin, even after controlling for factor affecting this gap 34 Figure 15. Checks on corruption and accountability institutions for middle-income countries and the DR 38 Figure B3.16. Indices for private consumption (from national accounts) and household income (from household survey) in real terms 40 Figure 17. Monetary poverty rates by area of residence 41 Figure 18. Growth incidence and shared prosperity 42 Figure 19. Gini index by income source in the DR and for LAC countries, 2008-2016 42 Figure 20. Poor, vulnerable and middle-class distribution in LAC and DR (2008-2016) 43 Figure 21. Distribution of income among poor and vulnerable populations, 2008 and 2016 44 Figure B4.22. Poverty rates by age group and level of education, by sex in 2016, and ratio of female and male poverty rates 46 Figure 23. Upward mobility and average annual GDP growth, 2008-2016 47 Figure 24. Moderate poverty rate by region, 2016 48 Figure 25. Human Opportunity Index 49 Figure 26. Contribution to the inequality of opportunities 50 Figure 27. Contributions to moderate and extreme poverty reduction, 2008-2013 and 2013-2016 50 Figure 28. Productivity and wages 52 Figure 29. Share of total employment by sector and sex, 2016 53 Figure 30. Dominican immigrants in the US and migration rates in LAC 55 Figure 31. Incidence of remittances by decile, 2016 56 Figure 32. Share of population by highest level of education achieved - Dominicans ages 25+ living in the DR or in the US, 2016 57 Figure 33. Share of households receiving government transfers and value of the transfers for recipient households, by income group 2008-2016 59 Figure 34. Concentration coefficients and budget shares by program, 2013 60 Figure 35. Share of students overage or failing school (dropout + not promoted), 2014 62 Figure 36. Percentage of individuals who consulted a public health facility 30 days before the survey, by year and income group 63 Figure 37. Maternal care and mortality, 2014: high coverage but lagging outcomes 64 Figure 38. Public and private health expenditures (as percent of GDP), 2014 65 Figure 39. Share of households paying for private electricity and education, by decile 2016 66 Figure 40. The DR has low levels of fiscal revenues which has contributed to an increase in public debt 69 Figure 41. The DR’s fiscal deficits and debt levels are low compared to regional peers 69 Figure 42. Government expenditure in the DR is relatively rigid, which may contribute to the recent uptick in debt levels 70 Figure 43. Natural disasters could increase sovereign borrowing costs in the DR 72 Figure 44. Frequency of natural disasters in LAC and number of disaster in the DR 73 Figure 45. Sectoral distribution of losses and potential maximum losses 74 Figure 46. Urbanization rates in the DR, LAC and upper-middle income countries (UMI), 1991-2016 76 Figure 47. Growth of forest capital and GDP, 1995-2014 (average annual growth rate, percent) 78 Figure 48. Governance at a glance: The DR in comparative perspective and over time. 79 Figure 50. Perception of insecurity in the neighborhoods 82 Figure A51. The composition of the DR’s exports stands closer to Central American countries than it does to other Caribbean islands 100 Figure A52. The DR’s export basket has moved towards intermediate goods of higher technological content and longer quality ladders 101 Figure A53. The DR has a lower incidence of trade on GDP compared to regional peers and benchmarks and trade is concentrated in a few trading partners 102 List of tables Table 1. Priorities and their impact on the Dominican Republic’s development challenges 11 Table 2. Profile by income groups, 2008 and 2016 45 Table 3. Coverage rates and D-index in urban and rural areas 49 Table 4. Priorities and their impact on the Dominican Republic’s development challenges 89 Table A5. Rankings according to distance to “best” performer 94 List of boxes Box 1. The structure and problems of the electricity sector 29 Box 2. Fiscal incentives and FDI attraction: Results from a global survey 37 Box 3. How do national accounts and household surveys compare in the DR? 40 Box 4. Gender and poverty in the DR 46 Box 5. The Participatory Anti-Corruption Initiative (IPAC) 81 Box A1. Opportunities and challenges from stronger DR-Haiti trade relations 104 Box A2. Global trends in the tourism sector 107 List of acronyms ACS American Community Survey ADESS Administradora de Subsidios Sociales API Access to Public Information BCRD Banco Central de la República Dominicana BEC Broad Economic Categories BOP Balance of Payments BVRD Bolsa de Valores de República Dominicana CA Central America CAFTA Central America Free Trade Agreement CCT Conditional Cash Transfer CDEEE Corporación Dominicana de Empresas Eléctricas CEI-RD Centro de Exportación e Inversión de la República Dominicana CEPAL Comisión Económica para América Latina y el Caribe CEPII Centre d’Études Prospectives et d’Informations Internationales CIF Cost Insurance and Freight CIT Corporate Income Tax CNE Comisión Nacional de Energía CPF Country Partnership Framework CORAA Corporación de Acueducto y Alcantarillado DGII Dirección General de Impuestos Internos ENFT Encuesta Nacional de Fuerza de Trabajo EU European Union EXPY Export Basket Productivity FAO Food and Agriculture Organization FDI Foreign Direct Investment FOB Free On Board FTA Free Trade Agreement GCI Global Competitiveness Index GDP Gross Domestic Product GHG Greenhouse Gas GNI Gross National Income GSD Gran Santo Domingo GVC Global Value Chains HOI Human Opportunity Index HPV Human Papilloma Virus ICT Information and Communications Technology IDB Inter-American Development Bank IFC International Finance Corporation IMF International Monetary Fund IPAC Iniciativas Participativas Anti-Corrupción IVACC Indice de Vulnerabilidad Ante Choques Climáticos JCE Junta Central Electoral LAC Latin America and the Caribbean LAPOP Latin America Public Opinion Project MEPYD Ministerio de Economía, Planificación y Desarrollo MESCYT Ministerio de Educación Superior, Ciencia y Tecnología MFA Multi-Fiber Agreement MFI Micro-Finance Institution MIGA Multilateral Investment Guarantee Agency MINERD Ministerio de Educación de la República Dominicana NA National accounts NDC Nationally Determined Contribution NFPS Non-Financial Public Sector NGO Non-Governmental Organization OECD Organisation for Economic Co-operation and Development ONE Oficina Nacional de Estadística PEFA Public Expenditure and Financial Accountability PISA Programme for International Student Assessment PR Public Relations PROSOLI Progresando con Solidaridad RTI Right to Information SCD Systematic Country Diagnostic SEDLAC Socio-Economic Database for Latin America and the Caribbean SEZ Special Economic Zones SIUBEN Sistema Unico de Beneficiarios SME Small and Medium Enterprises TERCE Tercer Estudio Regional Comparativo y Explicativo TFP Total Factor Productivity UMI Upper Middle Income UN United Nations UNESCO United Nations Educational, Scientific and Cultural Organization VAT Value-Added Tax WB World Bank WDI World Development Indicators WDR World Development Report WHO World Health Organization WTO World Trade Organization WBES World Bank Enterprise Survey WEO World Economic Outlook WYSE World Youth Student and Educational CHAPTER 1 Overview Geography, favorable external including the liberalization of foreign exchange trans- conditions, and the structural reforms actions, the introduction of new trade agreements, the of the nineties bolstered twenty-five elimination of price controls and of restrictions on for- years of high growth and social progress eign direct investment (FDI) in almost every sector, as well as trade liberalization. More recently, the country A fter a severe economic recession in 1990, the implemented an inflation targeting regime that has Dominican Republic (DR) has experienced a helped keep inflation at historically low levels. Broad- remarkable period of high economic growth based structural reforms were accompanied by a series over the past 25 years. The economy expanded by an of laws that provided incentives aimed at attracting average growth rate of 5.3 percent per year from 1993 FDI, and favoring specific sectors and firms. Following to 2017, allowing the country to reduce the gap with the enactment of law No. 16-95 in 1995, which opened respect to the living standards of high-income countries the possibility for foreign investments in almost every and making it one of the top performers in both Lat- economic sector, the country introduced a series of laws in America and the Caribbean (LAC), and the world. providing tax incentives to foreign investment. The Moreover, the country has experienced a relatively sta- pro-growth effects of economic reforms were amplified ble growth. After spiking in the 1980s, growth volatility by favorable external conditions and an advantageous gradually fell throughout the 1993-2017 period, even geographic location. The DR was favored by the Multi- amid the country’s exposure to natural disasters and cop- fiber Agreement (MFA), which imposed quotas to the ing with two subsequent crises: a domestic banking crisis exports of textiles from developing countries to high-in- in 2003 and the global financial crisis of 2009. come markets, and protected Dominican exports of textiles to the US against competition from countries High growth was the result of both a series of key like China. Proximity to North American markets and economic reforms implemented in the 1990s and natural endowment of beaches and access to the ocean, 2000s and favorable geographic and external eco- were also pivotal factors shaping economic develop- nomic conditions. The DR pursued economic reforms ment in the DR in recent years. 2 | Dominican Republic: Systematic Country Diagnostic Proximity to the US and Canadian markets and pro- Economic growth has been accompanied by signifi- ducers1 and economic reforms have also shaped FDI cant progress in terms of poverty reduction, growth inflows.2 The US is one of the top sources of FDI in- of the middle class, and access to services—especial- flows into the Dominican Republic, accounting for 25 ly in the last decade. Using the national poverty lines, percent of total inflows between 1993 and 2016. Can- after increasing from 32 to 50 percent between 2002 ada surpassed the US as the main source of FDI since and 2004 in the aftermath of the banking crisis, mon- 2010 by a small margin.3 Propelled by inflows of FDI, etary poverty has declined steadily since 2004, to reach investment has played a leading role as an engine of 29 percent in 2016. Using the LAC regional lines, be- growth over the past 25 years and physical capital accu- tween 2008 and 2016, poverty was reduced from 35 mulation was the biggest contributor to growth. Invest- to 21 percent, and the proportion of the middle class ment was the fastest growing component of domestic rose from 24 to 37 percent, outnumbering the poor for demand between 1992 and 2016 and the incidence of the first time in 2014.4 The vulnerable population is investment on GDP increased from 16 percent in 1991 the largest income group in the country (41 percent), to 24 percent in 2017. with risks of falling back into poverty if shocks materi- alize. Growth over the 2008-2016 period was slightly The DR’s geography, with its beaches and easy pro-poor, with per-capita income for individuals in the ocean access, is also a natural advantage for a strong bottom 40 percent growing at an average annual rate touristic sector, a shared feature with other Carib- of 4.2 percent, compared to 3.8 percent per year for bean and some Central American countries. Since the top 60 percent. The Gini coefficient decreased by the onset of the current growth episode, tourism has 2 points from 49.6 in 2008 to 47.1 in 2016, below re- played an important role in the Dominican economy. gional inequality levels throughout the period. Access The Hotels, Bars and Restaurants sector is the fifth to basic education, water, sanitation, and key assets has fastest growing in the country, averaging 6.4 percent improved since the early 2000s, resulting in a decrease per year over the 1991-2016 period. This sector’s rapid of the share of households with deprivations, a proxy growth was the product of supply and demand factors. for multidimensional poverty. An explicit policy push to attract FDI expanded the supply of available rooms by 244 percent in the last This SCD presents a storyline on the connection 26 years. This increase in supply has been met with between economic growth and poverty that departs strong demand, resulting in an increase in the occu- from previous narratives in other studies, including pancy rate of 10 percentage points since 1991 (from those conducted by the World Bank. Those studies 68 percent in 1991 to 78 percent in 2016). The ex- found a weak connection between the historically high pansion of tourism has also generated foreign currency rates of economic growth and poverty reduction in the for the Dominican economy. The net balance of trade country. Analyzing the consistency between data from in services increased steadily from 6 percent of GDP national accounts and the household survey that is used in 1994 to close to 9 percent of GDP in 2007. It has to measure poverty (Encuesta Nacional de Fuerza de Tra- since hovered around 6 percent of GDP, with a recent bajo – ENFT), we find large differences in the annual uptick in 2016. growth rates of private consumption from national ac- counts and household income from the ENFT before 1  Beyond geographic proximity, the linkages between the US and the DR have historic roots, and were reinforced after the US occupied the Dominican Republic from 1916 to 1924 and again in the mid-1960s. 2  On average, FDI inflows from North America accounted for approximately 45 percent of total FDI inflows and FDI in tourism accounted for 20 percent. 3  The recent uptick in FDI from Canada is partly explained by the boom in the mining sector in the DR, with a large presence of Canadian companies. One of these firms is the Pueblo Viejo Dominicana Corporation/Barrick Gold, the largest foreign investment in the history of the DR. As will be discussed in more detail below, mining is the fastest growing economic sector both in terms of value added and FDI inflows. 4  To compare income distributions and income groups across LAC countries, we use the SEDLAC harmonized datasets, with income expressed in US dollars PPP 2011. These income groups are defined using the LAC regional lines as follows: an individual is poor if her income is lower than US$5.5 a day; vulnerable if her income is between US$5.5 and US$13 a day, middle class if her income is between US$13 and US$70 a day, and upper class if her income is greater than US$70 a day. These values are expressed in PPP 2011. Dominican Republic: Systematic Country Diagnostic | 3 2008. These differences become smaller after the na- 2008 to 2.5 percent in 2015.6 This increase in expen- tional accounts were rebased and some methodological diture has allowed the country to rapidly increase its improvements were implemented for the ENFT. This health insurance coverage, but not enough to improve is the reason why this SCD uses 2008 as the starting significantly health outcomes, as it was compounded point for the analysis of the relationship between GDP by allocative inefficiencies. With education, likewise, growth and poverty reduction, which yields a higher increases in expenditure have enabled enrollment rates growth elasticity of poverty than the estimated in pre- to increase, but with the worst learning outcomes in vious studies. This observation does not invalidate the regional and international assessments. Moreover, de- conundrum of the DR having social outcomes compa- spite increasing in recent years, expenditure in edu- rable to those observed in countries with much lower cation in the DR stands below the LAC average and levels of development despite very high rates of eco- other middle-income countries in terms of GDP.7 Low nomic growth. These issues are discussed in chapter 3. investments in human capital have potential implica- tions not only on the welfare and lives of individuals, but also on productivity and competitiveness. Poor Challenges to become a high-income learning outcomes actually result in lower employabil- country by 2030 and to improve ity, productivity and wages for workers, and might be social outcomes associated with high school dropout rates for teenagers. Unemployment rates for ages 15-24 are almost three The generous and widespread tax incentives provid- times as high as rates for ages 25-54. ed to the private sector and high informality rates constrain the country’s capacity to generate fiscal Dominican firms typically exhibit low quality, low revenue which, combined with growing and rigid survival rates in export markets, and low levels of public expenditure, have put public debt in an in- backward linkages by export-oriented firms, out- creasing trajectory. Government revenue has fluctuat- comes that are likely shaped by economic policies. ed around 15 percent of GDP over the last 10 years, Firms are less innovative8 than similar firms in compara- one of the lowest levels in LAC. Tax expenditures are tor countries, which partly explains a relatively low pro- extremely high (estimated 6.4 percent of GDP in 2017) pensity to export. Even exporters display lower quality resulting from tax exemptions, and the tax base is small and survival rates relative to comparators. Low quality because of firm informality.5 While tax exemptions have from domestic firms, combined with tariff exemptions the potential of creating new jobs as new firms start op- and relatively easy access to high-quality inputs from erations to make use of these incentives, they also result the US, hinder the development of backward linkages in forgone fiscal revenue. Low government revenues, by export-oriented firms.9 This cycle is rooted in a sys- combined with growing and rigid expenditure, have led tem of exemptions that discourage linkages, in a lack to a rapid increase in public debt and interest payments. of incentives to adopt quality improvements, and in a business environment that hampers competitiveness. The low tax collection constrains the provision and Targeted economic incentives to local firms and spe- quality of services in education and health. For in- cial economic zones (SEZs) create an uneven playing stance, public health spending is one of the lowest in field which provides market power to a minority of eco- LAC despite an increase from 2.1 percent of GDP in nomic agents and fosters an enclave-type private sector, 5  See World Bank (2017c). 6  Data from the Global Health Expenditure Database compiled by the World Health Organization (WHO). The numbers represent the do- mestic general government health expenditure. 7  The DR’s expenditure in education as a share of GDP was 0.75 percentage points lower compared the average expenditure of countries in LAC with available data. 8  Innovativeness is defined in a broad sense: adoption of new processes, holding internationally recognized quality certifications, and on-the- job worker training. 9  Access to high-quality inputs from the US has a positive effect in the competitiveness of firms that use these inputs intensively. However, they seem to prevent stronger linkages between foreign firms and domestic firms, limiting the scope for knowledge spillovers. 4 | Dominican Republic: Systematic Country Diagnostic discouraging quality improvements. An inefficient elec- having some female ownership, below the LAC average tricity sector combined with high real interest rates (the at 40 percent. highest in LAC behind Brazil) further erode the com- petitiveness of Dominican firms and prevents greater The lack of opportunities in the labor market and connectivity between export-oriented and local firms. the poor quality of public services are push factors Additionally, those firms seeking to invest in quality im- for Dominicans to migrate. Out of 10.1 million peo- provements face a shortage of skilled workers.10 ple living in the country in 2016, more than 1.5 million lived abroad, with 1.1 million in the US. Migrants to An uneven playing field, along with distortions the US are more educated than those who stay in the from tax incentives, inhibit competition across and country, with 66 percent of migrants 25 or older having within sectors and limits both the competitiveness completed secondary education or higher, compared of domestic firms and the creation of more and to only 40 percent of Dominicans living in the DR. better jobs. Labor markets in the country are char- The Dominican diaspora sends remittances amounting acterized by unresponsive unemployment rates, high to close to 8 percent of GDP, lowering poverty by an levels of informality, and low wages. In the period estimated 3 percentage points, and reducing the partic- 2008-2016, the number of unemployed workers grew ipation of recipients in the labor market. by 10,000-odd people every year, with 1 out of every 6 workers willing to work additional hours in 2016. The Among countries with similar population size in labor market is the main connector between econom- LAC, the DR has a unique combination of relative- ic growth and poverty in the country, accounting for ly high immigration and emigration rates. Sharing a over fifty percent of poverty reduction between 2008 border with Haiti-a much poorer country, has facilitated and 2016, and offers the greatest potential to sustain migration of Haitians to the DR for economic reasons. and deepen poverty reduction going forward. After According to the latest survey of immigrants (2017) in 2013 the labor market has performed better and the the DR, almost 500,000 of them are Haitians, with 70 number of unemployed was reduced by 15,000 every percent having at most primary education with an aver- year until 2016. age of 4.1 years of schooling (compared to 9 years for Do- minican-born), and highly clustered in unskilled work, Women face challenges in labor markets. Labor force especially in agriculture, construction, and commerce. participation rates for women is much lower than for men (59 vs 82 percent), and women account for 83 per- The downside to the country’s geographic location cent of the economically inactive population. At 20.6 is its high exposure to disaster and climate-related percent the unemployment rate more than doubles that risks. Being in the pathway of major hurricanes in the for men (8.6 percent). Furthermore, once employed, Caribbean Basin and on the edge of a tectonic plate, women face high sectoral segregation, with women ac- makes the DR one of the most exposed countries in counting for less than 5 percent of the jobs in some the world to adverse natural events. The northeastern sectors (construction and agriculture) and accounting region is exposed to floods and mudslides caused by for much of the jobs in other sectors (social and health severe storms, while arid parts of the northwest experi- services, and education). Women also face a wage gap ence increasing temperatures leading to more drought, with average hourly wages around 15 percent lower which in turn reduces crop yields and water supplies. than men’s. Some of the outcomes for women might A global comparison suggests that the DR stands out be partly explained by one of the highest adolescent fer- as one of the most affected countries by natural disas- tility rates in the world and by high rates of violence ters in the last 20 years. The country ranks 10th world- against women. They are also underrepresented among wide in terms of human and economic losses due to firm owners, with 30 percent of the firms in the country weather-related events between 1997 and 2016.11 10  See the 2016 World Bank Enterprise Survey Dominican Republic Country Profile (https://www.enterprisesurveys.org/~/media/GIAWB/ EnterpriseSurveys/Documents/Profiles/English/Dominican-Republic-2016.pdf ) and 2017/2018 Global Competitiveness Report. 11  Germanwatch’s Global Climate Risk Index (https://germanwatch.org/en/download/20432.pdf ) Dominican Republic: Systematic Country Diagnostic | 5 From 1980 to 2008, 40 different disasters affected 2.65 of the population live in urban areas, agricultural activ- million people, almost a quarter of the entire country’s ities use 83 percent of the available water, mostly from population. In the past 40 years, disasters caused aver- superficial sources—with little efficiency. Only 1 in 4 of age annual losses amounting to 0.7 percent of GDP per the fully-functioning wastewater treatment plants were year, and catastrophic events can potentially destroy a at an adequate operating level, creating conditions for greater share of assets in the DR than in the average of the proliferation of water- and sanitation-borne diseas- Central American and Caribbean countries. The coun- es. Moreover, deterioration of natural and fresh water try’s increasing debt trajectory could be significantly af- resources increases the risks and cost to deliver safe wa- fected if disaster or climate-related risks materialize, as ter and sanitation with quality and availability. These they would impact fiscal deficits and potentially result conditions also exacerbate other risks related to climate in large increases in financing costs. Shocks from natu- change impacts and public health. Santo Domingo’s ral events have regressive distributional effects insofar current infrastructure, for instance, can only collect 14 poorest households are more vulnerable. Extreme winds percent of the generated waste water, with the excess are associated with increases in poverty of 1-2 percent- contaminating rivers and watersheds, posing environ- age points. mental risks and threats to the health of communities. The contamination of water sources has the potential to Climate change is anticipated to increase the fre- affect touristic activity, which is one of the main sources quency and severity of hydrometeorological haz- of growth and jobs in the country. ards in the country, notably sea level rise and the variability of slow-onset  natural events, particular- Increasing perceptions of insecurity in the coun- ly droughts. Temperature variability in tropical regions try affect both households and firms. Three is highly correlated with the occurrence of cyclonic and out of four Dominicans identify crime and vio- associated events such as floods and landslides. In the lence as the most important challenge facing the Caribbean, climate change is expected to increase the country, with seven out of ten reporting it be- intensity of excess rainfall, storm surges, and hurri- ing a conversation topic in the last two weeks.13 canes, with relatively higher impacts in the DR. Pre- This issue not only affects households, but almost 60 dicted impacts include flooding in coastal zones and percent of firms spend money on security, potentially low-lying areas, with potential damages and economic affecting investment decisions.14 losses projected to reach nearly 17 percent of GDP.12 The increased frequency of extreme events will have a The potential vulnerability and exclusion of Haitian negative impact and damage hotels, coastal infrastruc- migrants are threats to social cohesion. Achieving so- ture and beaches, leading to significant loss of tourism cial cohesion in the face of large cross-border popula- revenues; and will damage fish nursery areas and coral tion movements is a policy challenge in the DR, with reefs. Likewise, climate change is expected to increase most of the 500,000 Haitian migrants lacking a regu- the incidence and reproduction of mosquito-related lar migration status which would make them less vul- diseases such as dengue, chikungunya and Zika. nerable to exploitation and discrimination, and allow them to access better paid jobs.15 Since mid-2014, the Rapid urbanization, the lack of infrastructure to Dominican government has been implementing three accommodate it, and poor natural resource manage- parallel processes: i) regularization of foreigners with ment have increased pressure on natural resources, irregular migration status; ii) partial restoration of iden- environmental degradation and the DR’s vulnerabil- tity documents for descendants born in the DR; and iii) ity to climate change. Despite the fact that 80 percent registration of unregistered descendants.16 12 https://www.gfdrr.org/en/dominican-republic 13  ONE (2017). 14  Jaitman (2017). 15  Wooding and Riveros (2017). 16  Wooding (2018). 6 | Dominican Republic: Systematic Country Diagnostic This SCD identifies governance issues which hamper quality upgrade, and a labor market that does not gen- the country’s development, and underlie many of the erate enough good jobs. Policies in the DR have fostered aforementioned challenges.17 According to the 2016 an enclave economy where dynamic, export-oriented Enterprise Survey, 1 out of every 5 firms in the coun- firms do not connect with other firms in the economy, try report corruption as the biggest obstacle to doing discouraging the latter to improve the quality of their business (the highest proportion of answers), resulting production. Disincentives and barriers to innovation in an inefficient allocation of resources. This is consis- are further augmented by the market power enjoyed by tent with the statements of the government’s Estrategia firms operating in non-tradable sectors and flaws in the Nacional de Desarrollo (END), which candidly iden- institutional design of governmental agencies that have tifies corruption, clientelism and patrimonialism, as the goal of facilitating the flow of information and pro- phenomena that weaken state institutions and account- moting innovation. There is also a tight link between ability mechanisms, limiting the government’s effec- firm-level behavior and labor market outcomes. The tiveness. This SCD argues that these governance issues dearth of innovation and quality upgrading observed in might become binding constraints for the country to the DR prevents firms from generating enough quality graduate to high-income by 2030. jobs, discouraging the incentives of workers to invest in skills. Finally, a limited fiscal space and the poor pro- Many of the governance challenges identified in this vision of quality services affects the competitiveness of SCD are not new; rather, they are rooted in history. Dominican firms. First, the DR’s growing public debt For example, during the mid-20th century the coun- and the relatively high interest payments by the govern- try went through the early stages of an industrializa- ment spillover to high lending rates charged to the pri- tion process in part thanks to policies that protected vate sector. Also, Dominican firms’ ability to innovate, certain industries. At the same time, these protections adopt quality standards, and move up the value chains favored the emergence of a number of monopolies that is constrained by an inadequately educated workforce. over time increased their capacity to influence policy in their favor. The situation has improved since, yet, given initial conditions, there is still progress to be made. Priority areas Moreover, many of the DR’s development challeng- This SCD identifies five key and interrelated chal- es are tightly linked and reinforce each other, which lenges the DR needs to overcome to achieve a more means that changes to the status quo will require inclusive and sustainable growth. First, despite high tackling these challenges in tandem. There is a vicious economic growth rates, firms in the DR have been un- circle of limited fiscal space, low provision and poor able to achieve substantial quality improvements, and quality of public services, and weak governance. Low linkages between export-oriented firms and local firms levels of tax revenue erode the government’s ability to are weak. These productive challenges have hampered expand the coverage and quality of public services. A integration in international markets and prevented a small and inefficient state has led many Dominicans to more inclusive growth across firms. Second, the coun- opt out from the public system and purchase, to the ex- try has not generated sufficient tax revenues to provide tent they can afford it, privately-provided services. This, public services or higher quality, and as a result has seen together with the perception of corruption, weakens a deterioration of fiscal position. Third, due to lack of people’s incentives to pay taxes, which further limits fis- resources and an uneven incidence of spending, the cal revenue and weakens the governments’ accountabil- country has been unable to afford much-needed invest- ity. There is another vicious circle of policies that limit ments in human capital. A fourth challenge stems from competition, firms that have few incentives to invest in the degradation of the country’s natural capital, mainly 17  See Ministerio de Economía, Planificación y Desarrollo (2016), page 15. “Se parte del entendimiento de que nuestro actual modelo de desarrollo presenta falencias en materia de la calidad de la acción estatal, que se manifiestan en fuertes deficiencias en los servicios públicos que deberían garantizar los derechos constitucionales de la población […], así como patrones de actuación que entorpecen el desarrollo nacional, tales como corrupción, clientelismo y patrimonialismo.” Estrategia Nacional de Desarrollo 2030. Dominican Republic: Systematic Country Diagnostic | 7 caused by a lack of territorial and urban planning in (LAC and UMI - see Annex 2); ii) the validation of the face of rapid urbanization, and a lack of supporting findings using both existing literature and the new anal- infrastructure. The aforementioned challenges are am- ysis conducted for this SCD; iii) how a given priority plified by the country’s exposure to natural disasters, the may contribute to making progress in other areas; and fifth challenge, which impacts economic growth, mac- iv) a robust process of consultations with Bank staff and roeconomic stability, and households’ welfare. in-country stakeholders (see Annex 7). To address the challenges described above, the SCD The benchmarking exercise highlighted a set of is- identifies six priority areas for the country to over- sues where the country’s underperformance sets it come its development challenges and achieve its goal apart from its peers. This exercise provided a first long to become an inclusive and high-income economy list of issues, that was refined using the diagnostics in by 2030. First, credibly improving fiscal balance by in- the chapters to produce a shorter list of priorities. This creasing revenue and improving the quality of fiscal ex- short list was used during the consultation process with penditure. Second, enhancing the accumulation of hu- more than 200 participants in country from a wide man capital. Third, pursuing policies that create a level range of stakeholders including government, civil soci- playing field and a business environment that fosters ety, academia, private sector, and development partners. investment in quality upgrading. Fourth, improving At every meeting during the consultation process we the management and conservation of natural resources. asked participants to pick a few areas they considered Fifth, improving the country’s resilience to disasters and the most critical to address the country’s development climate-related risks. Finally, increasing contestability in challenges. The vast majority of priorities selected by the policy making process. participants are embedded in the six priorities identified in the SCD. The impact of dealing with these priorities will in- crease to the extent that they are tackled in tandem Improving fiscal balance by developing and that they incorporate the linkages between the a credible fiscal path that supports the different challenges. For example, direct efforts to provision of high-quality public services raise fiscal revenues will probably have a limited effect if they are not accompanied by explicit efforts to improve The DR’s narrow fiscal space and growing public debt the provision of public goods and to fight corruption. limits the country’s spending ability in much-need- Similarly, the impact of initiatives to foster innovation ed social programs and erodes the private sector’s or better processes of production will be muted to the competitiveness. These problems are compounded by extent that they do not address market power and oth- the country’s exposure to natural disasters, which both er disincentives to innovation. Moreover, reforms will impacts fiscal accounts by lowering tax collection, and have to take into account the political economy con- calls for fiscal buffers to respond when shocks material- straints that they could face. ize. Raising revenues and delineating a credible medi- um-term fiscal path with prudent fiscal rules are crucial These priorities were identified based on the follow- elements to open up fiscal space and ensure fiscal sus- ing criteria. i) A benchmarking exercise that compares tainability in the DR. the DR’s performance with the performance of coun- tries in LAC and the group of upper-middle income To strengthen fiscal sustainability, the DR needs countries (UMI). The indicators used for the compari- to tackle the revenue and expenditure side of fiscal son cover a wide range of topics including growth, fiscal accounts. Raising tax revenue will hinge on rational- accounts, business environment, competitiveness, hu- izing tax exemptions and expanding the tax base. Tax man development, natural resources, and governance. exemptions and informality have been important fac- The DR’s performance in each indicator is compared tors behind the DR’s relatively low tax revenues. The with the best performers in each comparison group limitations of the DR’s tax system in producing higher 8 | Dominican Republic: Systematic Country Diagnostic revenues and the importance of addressing these limita- levels, education outcomes are poor, both in regional tions are recognized in the country’s National Develop- and international assessments. The country consistently ment Strategy 2030. The latter outlines a plan aimed at ranks at the bottom of international assessments, and consolidating all existing tax expenditure schemes into a its own standardized assessments indicate low levels of single section of the tax code and establishing a coherent competencies. Cognitive skills and the quality of learn- and sustainable approach to tax expenditures which re- ing outcomes have potentially large implications for duces their fiscal impact and minimizes their distortive economic growth as they bolster innovation, ultimately effect on economic incentives. In addition to tackling a resulting in higher productivity gains.19 Low levels of relatively low tax base, the country needs to address the learning are likely to contribute to high dropout rates, growing rigidity of government expenditure, especially especially in secondary education, and are consistent due to growing interest payments. In addition to this, in with complaints from employers that poorly educated a context of tight fiscal space, the country needs to focus workers are an important barrier to doing business. on mechanisms to improve the quality of public services According to diagnostics of the sector, among the en- and, in some cases, reduce its fiscal footprint. The DR try points to improve skills acquisition and learning has increased its expenditure in key areas such as educa- outcomes are early childhood development, improved tion and health. However, as discussed in chapter 3, the teacher skills and recruitment, curricular reform, and quality of public expenditure (measured by outcomes) improved school management. in both these areas is obviously low. Likewise, the public systems of electricity distribution and water and sanita- Close learning gaps and low enrollment in early tion not only provide an unreliable service of low quali- years and in secondary education. On school and ty, but their operational costs are high. sector management, it is important to align roles and responsibilities of all levels of the Ministry of Educa- Efforts towards raising revenues and making a more tion (MINERD) to respond to the two aforementioned efficient use of resources need to be accompanied by challenges in the system. New approaches to moderniz- the development of a credible medium-term fiscal ing and articulating links at all levels of the system are plan which anchors expectations and allows the DR promising and have the potential to improve outcomes. to build buffers. The country needs to strengthen its These include strengthening in-service teacher training efforts to contain the rising government debt and its in- through the strategic use of data and technology; pro- terest payments. To achieve this objective, the country fessional development for school principals, including must design a credible and clearly defined fiscal plan, linking it to teacher support, so that behavioral change anchored in fiscal rules, that stabilizes debt levels and in instruction in the classroom becomes a reality; using anchors market expectations. Indeed, a credible fiscal technology to better and strategically monitor service anchor could lead to lower borrowing costs, which in delivery; and empowering each level of the MINERD turn could reduce the potential economic impacts of to make change happens in schools, as exemplified by fiscal adjustments.18 Beyond aiming at stabilizing debt the performance-agreements for school districts that the levels, the fiscal anchor, should also seek to build fis- Ministry has recently implemented. cal buffers to respond to economic and natural shocks without affecting the trajectory of public debt. Development of socio-emotional, technical and vo- cational skills. Improving the relevance of secondary Enhancing the accumulation of human education by offering clear paths to higher education, capital and addressing gender imbalances links with employers, expanding on-the-job opportu- nities, and incorporating socio-emotional skill train- Improving skills acquisition and learning outcomes. ing in pre-tertiary education, can certainly enhance Despite significant gains in access to education at all the development of all types of skills, reduce dropout, 18  See IMF (2017) and Végh et al. (2018). 19  Hanushek and Woessmann (2010). Dominican Republic: Systematic Country Diagnostic | 9 increase employability, and increase productivity in Enabling human capital accumulation and enhanc- the labor market. This would reduce the skills mis- ing risk protection for the poor and vulnerable match which the private sector reports as an import- through social protection and labor market interven- ant barrier to doing business. To achieve this, it is also tions. Sustaining the recent gains on poverty reduction important that the MINERD works hand in hand will require further efforts to build the human capital of with the Ministry of Higher Education, Science and poor and vulnerable group, and to protect poor house- Technology (MESCYT), as the latter oversees techni- holds from either economic risks or disasters. However, cal programs in the country that are offered by univer- the largest fiscal expenditures are poorly targeted, while sities and institutes. the programs that are well targeted are relatively small. Both the incidence and average per capita value of gov- Improving health care financing and expenditure ef- ernment transfers have increased for all socioeconomic ficiency to ensure the quality of health care provision groups since 2008, including the middle class, with per and equitable access to health services. The quality capita transfers increasing along with average per capita of care and health outcomes is clearly lagging behind household income. These facts suggest the need to fur- the gains in coverage. The country faces a double bur- ther improve the targeting and adequacy of conditional den of high maternal and child mortality rates and an cash transfers. increasing share of noncommunicable diseases. Poor health outcomes do not only affect worker productivity Enhancing protection for the poor and vulnerable and children’s learning outcomes, but cost human lives from either economic risks or disasters. Connecting in the country. At 1.7 percent of GDP in 2016, the poor households to productive inclusion initiatives, and health public expenditures as a share of GDP are one preparing the social protection system and instruments of the lowest in LAC. While the sector requires more suited to disaster preparedness and response, will con- public resources to keep on improving the quality of tribute to deepen and sustain the recent gains on pov- services and expand financial protection, there is room erty reduction. The country should continue its efforts to improve both efficiency and the cost-effectiveness to provide identification documents to all citizens to re- of spending. Sector budgeting and resource allocation move barriers of access to social programs, financial ser- have merely followed previous allocations and spending vices, political participation, and secondary education. is heavily skewed toward curative care and hospitals at the expense of primary health care and public health Addressing gender imbalances. Women in the coun- services. try not only face high rates of violence and adolescent fertility, but also lower labor force participation, much The health sector also exhibits inadequate account- higher unemployment, and high employment segrega- ability mechanisms. In addition, the limited capacity tion across sector and lower wages. For instance, po- of the Ministry of Public Health to monitor, regulate tential productivity gains from incorporating women and enforce, contribute to inefficiencies in spending. who completed some level of education but then do Enhancing the quality of health services can save lives, not enter the labor market are about 4 percent of GDP improve the quality of life, and increase worker produc- in the DR.21 Higher female labor force participation is tivity. Not all necessary quality improvements require associated with lower infant mortality rates and higher large increases in resources. For instance, most mater- life expectancy. It also has positive effects on children’s nal deaths, too high a number for the country’s cover- development and well-being, increasing their future age of institutional birth and antenatal care, could be earnings capability, which may reduce the intergenera- prevented having properly trained personnel adhere to tional transmission of poverty.22 standards and follow protocols and quality standards.20 20  Castro (2016). 21  These estimates assume that current returns to each education level remain unchanged. See Mateo Díaz and Rodríguez-Chamussy (2016). 22 Ibid. 10 | Dominican Republic: Systematic Country Diagnostic Promoting a level playing field for firms territory is critically sensitive to desertification which and a business environment which is mainly caused by intense and inadequate land use fosters quality improvements and practices. Increased pressure on water resources from competitiveness urbanization and land use change, threatens the main engines of growth in the country. Creating institutional The Dominican economy is characterized by rela- and policy incentives for providers to improve efficiency tively high perceptions of market power and a reg- such as increasing metering, billing and collection, and ulatory divide which unbalance the playing field for reducing non-revenue water23 would not only increase firms and lowers incentives to innovate. The country resilience to natural disasters and improve service qual- stands out in LAC and worldwide in terms of perceived ity, but also increase commercial viability (reducing de- market dominance. Moreover, perceptions about mar- pendence on central government subsidies) and reduce ket dominance in the DR have worsened. The SEZ’s the degradation of water resources. Increasing access to regime continued focus on specific industries such as private sector financing and reducing subsidies will be textiles, footwear and leather goods, could entrench required to finance critical investments to close regional economic distortions and perpetuate the country’s dual disparities in access to water supply and sanitation in- export structure. The DR has made progress towards frastructure (contributing to human capital), and to re- greater competition and generating a more level play- duce the degradation of water resources stemming from ing field (such as the full implementation of the DR’s urbanization through increased investments in sanita- pro-competition law), but the country needs is penal- tion and wastewater reuse. Slowing down watershed ized by distortions such as targeted tax exemptions, that degradation will contribute to increase the country’s re- generate market power. Beyond the lack of competi- silience to climate change, reduce the spread of gastro- tion, the quality of exports and inputs is affected by a intestinal and respiratory diseases, and prevent negative weak regulatory and institutional capacity. effects on touristic activity, a pillar of economic growth and employment generation. Firm performance could also benefit from a deeper and more competitive financial market. Dominican Strengthening integrated water resources manage- firms face higher real interest rates than firms in most ment (IWM) to ensure stewardship of freshwater LAC countries or from comparator countries, and real resources to meet growing demands. Even though 80 interest rates have been on the rise since 2010. High percent of the population live in urban areas, agricultur- real interest rates may be one factor behind the coun- al activities use over 80 percent of the available volume, try’s relatively low credit volumes, one of the lowest in mostly from superficial sources, with little efficiency. To LAC. Capital markets are also shallow, with stock mar- address water use imbalances, the country should keep ket capitalization to GDP ranking among the lowest in up building on its recent efforts in basin level planning the region. for municipal, industrial, and agricultural water con- sumption. This implies continued focus on better coor- Improving the management of natural dination between the water and agricultural sectors to resources create conditions for water reuse for industrial and ag- ricultural activities with the aim to reducing freshwater Improving the efficiency of water and sanitation pro- exploitation as well, particularly in more arid parts of vision to increase resilience, improve service quality, the country. Stewardship and sustainability of freshwa- and reduce the degradation of water resources. En- ter resources through better planning and allocation are vironmental degradation and increased vulnerability critical to meet the growing competing demands from to climate change were caused by inadequate resource tourism,24 municipalities, industry, and the agriculture management. About two thirds of the Dominican sector. 23  Non-revenue water is defined as water which is pumped and then lost or unaccounted for. 24  Recent literature (Lim 2017) highlights the relationship between polluted water sources and beaches and tourism activity. On the one hand, Dominican Republic: Systematic Country Diagnostic | 11 Improving resilience to disasters and resilience by ensuring the safe location of population climate-related risks and managing issues of population relocation when re- quired, by building disaster and climate-risk informed Improving resilience to disaster and climate-relat- territorial and land use instruments, and through a ed risks would have positive effects on economic comprehensive management of natural resources. growth, fiscal sustainability, and households’ wel- fare. Between 1961 and 2014, average annual econom- Increasing transparency and ic losses caused by disasters in the DR amount to about accountability 0.69 percent of GDP, and the country exhibits a unique combination of similar and relatively high potential Improvements in transparency in the allocation and losses from catastrophic hurricanes and earthquakes execution of resources by the government, including among Central America and the Caribbean countries. the procurement of goods and works, are needed Accelerating the design and implementation of disas- to increase accountability, and to gradually enhance ter and climate-related risks management strategy, well the effectiveness of coalitions for social dialogue such integrated with the overall management of fiscal risks, as the Pacto Eléctrico and the forthcoming Pacto Fiscal. would not only contribute to fiscal and macroeconomic They will also gradually help to level the playing field sustainability, but also improve the government’s fiscal for the private sector, which will require to strengthen capacity to respond and recover quickly in the after- both supply of and demand for information. On the math of a large disaster. Increasing the resilience to di- supply side, the implementation of the access to public saster and climate-related risks of poor and vulnerable information framework should be strengthened, as well households through targeted programs would contrib- as the availability of government data. On the demand ute to sustain the gains in poverty reduction. Estima- side, efforts by the civil society, the media and academia tions indicate that extreme wind events are associated to effectively use available information to monitor gov- with increases in poverty equal to the average annual ernment’s performance and hold it accountable should poverty reduction between 2004 and 2012. Territorial be explored; and the private sector will have to be en- and urban planning will play a key role in increasing couraged in identifying business opportunities. TABLE 1. Priorities and their impact on the Dominican Republic’s development challenges Priority Area Growth Inclusion and human Sustainability challenges development challenges challenges Improving fiscal balance Indirect Indirect effect Direct effect effect Enhancing the accumulation of human capital Indirect Direct effect Indirect effect effect Promoting a level playing field and a better business Direct effect Indirect effect Indirect effect environment Improving the management of natural resources Indirect Indirect effect Direct effect effect Improving resilience to disasters and climate-related risks Indirect Indirect effect Direct effect effect Increasing transparency and accountability in the policy Indirect Indirect effect Indirect effect making process effect water effluents from industry, cruise ships and human activity that are untreated flow into natural water and oceans cause eutrophication: the ultra-high concentration of nutrients like nitrogen and phosphorus in a body of water. This results in a massive blooming of algae, covering ponds and streams, affecting the oceans and water bodies used by tourists in the country. Also, more tourism in the Dominican Republic implies more traffic from cruise ships. Cruise ships are one of the largest contributors to the problem of effluents and waste in the ocean. Integrating wastewater, port-efficiency and effluent treatment can have multiplicative and positive effects to reduce ocean pollution and make tourism more sustainable. CHAPTER 2 Economic Performance and Growth T he DR has benefitted from a long period of sustained high-growth. Policy reforms and geograph- ic endowments facilitated FDI attraction and investment growth, the main driver of GDP growth. The country’s private sector has been equally dynamic, with formal firm growth exceeding benchmarks. These positive developments notwithstanding, there is still untapped development poten- tial. Some of the same factors that fostered investment hinder the incentives of Dominican firms to engage in quality upgrading, which leads to uneven export performance and limits the country’s ability to connect fast-growing sectors with other parts of the economy and its potential to create high-quality jobs. To fully unleash the country’s growth potential, the country needs to address structural factors dis- torting firms’ incentives, such as improving the provision of inputs like electricity and skilled workers, reducing the relatively high financing costs faced by Dominican firms, developing a relatively shallow financial market, and creating a level playing field. Twenty-five years of fast economic The rapid growth of the last 25 years allowed the growth and convergence country’s real per capita income to increase substan- tially and helped the country reduce the gap with The DR has enjoyed 25 years of rapid economic respect to the living standards of high-income coun- growth. The economy expanded by an average growth tries at a faster pace than the average country in LAC rate of 5.3 percent per year from 1993 to 2017, one and the world. Real gross domestic product (GDP) per of the highest in the LAC region. This episode of eco- capita in 1993 stood at US$3,000 (in 2010 dollars), or nomic dynamism was ignited by a series of structural 7.5 percent of the GDP per capita of the US (Figure 1). reforms undertaken in the early 1990s, which allowed By 2016, the country’s real income was US$7,000, or the country to grow at an average rate of 6 percent per 13.5 percent of the GDP per capita of the US. Thus, the year between 1993 and 2000. Growth in the country DR closed the gap with respect to the income of the US remained high in the 2001-2017 period, averaging a by 6 percentage points (pp) over the last 25 years. By yearly rate of 4.9 percent. contrast, the median country in LAC and in the world, 14 | Dominican Republic: Systematic Country Diagnostic FIGURE 1: Over the past 25 years, the Dominican Republic has converged to the living standards of high-income countries at a faster pace than the average country in LAC and the World 14 2016 13 Conv r in Sh r of US p r c pit GDP 12 11 2010 10 1975 2000 9 1980 2005 8 1960 1985 Div r in 7 1995 1970 1990 6 1965 5 1000 2000 3000 4000 5000 6000 7000 GDP p r c pit , const nt 2010 US$ Source: World Bank staff’s calculation based on WDI data. closed their income gap with respect to the US by 2 and The relatively high GDP growth rate was bolstered 4 percentage points, respectively, over the same time- by a combination of factors, including a series of frame. Moreover, during the 1993-2017 period, the structural reforms and stabilization policies that country saw a sharper reduction in its income gap with started in the early 1990s. As many other countries in respect to the US than 86 percent of countries in LAC LAC, the process of structural reforms started in 1991 and 87 percent of countries in the world. and accelerated in the period between 1995 and 1999. This process included reforms to trade policy, to foreign The recent episode of high growth followed a decade investment policies, to the financial sector, and reforms of economic convergence in the 1970s and a decade to the tax and labor codes. This is documented in Lora of stagnation in the 1980s. The country experienced (2012), who constructs an index of structural reforms more than a decade of income convergence in the late that quantifies the intensity of reforms across LAC 1960s and the 1970s, after the country overcame the countries. According to this index, the country had a political and economic turmoil that followed the assas- relatively closed and economically repressed econom- sination of dictator Rafael Trujillo in 1961, and pursued ic system in 1986, with an index of structural reforms a series of reforms that boosted economic activity, such (0.3) that stood behind the regional average (0.35). By as the creation of special economic zones (SEZs). In the 1999 the process of structural reforms in the country 15 years between 1965 and 1980 average growth in the peaked, with an intensity that surpassed that of many country exceeded 6 percent per year and real income of its regional peers, and the regional average. Between per capita doubled. As in most countries in LAC, the 1999 and 2009 the pace of structural reforms in the DR economic progress of the 1970s came to a halt in the and LAC slowed down. Structural reforms and stabili- 1980s, in part because of the debt crisis that hit the zation policies were the largest contributors to the GDP region during that period. GDP per capita remained growth acceleration in the country in the 1990s,25 but stagnant throughout the 1980s and the country’s in- had no impact on growth in the 2000s.26 come diverged relative to that of the US, falling from 9 percent in 1980 to 7 percent in 1990. It took the Linkages between economic reforms and growth country 26 years to return to the levels of real income have been accentuated by the country’s geographic relative to the US observed in 1980. location and the resulting strong ties with the US. 25  See Loayza, Fjanzylber and Calderon (2005). 26  This could be the result of a slowdown in the pace of reforms or the types of reforms that were pursued in the 2000s. See Araujo et al. (2016). Dominican Republic: Systematic Country Diagnostic | 15 The Dominican Republic is highly responsive to growth ic expansion since the early 1990s, and it has been in the US27, which suggests that the strong growth per- accompanied by total factor productivity (TFP) formance of the US during the 1990s also contribut- growth. The contribution of physical capital has been ed to the growth acceleration observed in the DR. The relatively stable over 5-year periods, hovering around 50 high elasticity of the country’s growth with respect to percent of total growth and stands out as the largest US growth is also anecdotally illustrated in recent years, among comparator countries30, except for Peru (Figure with the DR’s growth over the past two years exceeding 3, panels A and B). The relative stability of the contri- its average at the same time as growth in the US accel- bution of capital contrasts with that of TFP and labor. erated. Labor accounted for approximately 50 percent of the growth between 1990 and 2000. Since then, its con- In spite of facing two crises and two major hurri- tribution has remained more modest, at about 30-40 canes and tropical storms,28 the remarkable growth percent of total growth. The share of growth explained performance of the last 25 years was accompanied by TFP growth over the last 25 years was 13 percent, by low levels of volatility, both in historical terms but it varied significantly from period to period.31 The and compared to other countries. After spiking in the positive contribution of TFP to total growth in the DR 1980s, the coefficient of variation of growth, a com- stands in contrast to what is observed for regional peers, mon measure of volatility, fell by half after 1993 (Figure where TFP growth has been negative. 2, panel A). This happened even as the country coped with a domestic banking crisis in 2003 and the global The predominant role of capital accumulation in financial crisis of 2009, and despite exposure to natural the growth process is explained by high investment disasters. The newfound resilience of the Dominican growth rates. Investment has played a leading role as economy is in part the product of the structural and an engine of growth over the past 25 years. It was the stabilization policies, all of which made the country fastest growing component of domestic demand be- nimbler.29 Compared to other countries, the DR stands tween 1991 and 2017, followed by consumption and out for having a combination of high growth rates and exports. Because of its robust growth, the incidence of low levels of volatility (Figure 2, panel B). investment on GDP increased from 16 percent in 1991 to 24 percent in 2017. Moreover, between 1991 and 2016, investment growth outpaced the levels observed Investment and capital accumulation in a set of comparator countries, except for Peru (Figure have been the main drivers of 3, panel C).32 growth, and they were accompanied by productivity growth In part, investment was propelled by inflows of for- eign direct investment (FDI), especially since the Capital accumulation has been the main driver of mid-1990s. FDI as a share of GDP rose sharply after growth, accounting for half of the country’s econom- 1996, from an average 1.2 percent of GDP between 27  Estimations presented in Végh, Lederman, and Bennett (2017), show that GDP growth in the DR is highly elastic to the growth rate of G7 countries, of which the US represents a big part. More specifically, the authors use quarterly data to estimate the elasticity of year-on-year growth rates of different LAC countries with respect to external factors (growth of G7 countries, China’s growth, commodity prices, and the US 10-year treasury rate). Their findings show that, after controlling for other external factors likely affected by US growth, the elasticity of the DR’s growth (in terms of deviations from the country’s average growth) with respect to G7 growth is positive and significant, and is the fourth largest among LAC countries with available data. 28  Hurricane Georges in 1998 and Tropical Storms Noel and Olga in 2007. 29  The reforms of the 1990s were mainly structural reforms that opened the Dominican economy to trade and foreign capital flows, and reforms to specific sectors. The reforms of the 2000s, such as the adoption of an inflation targeting regime by the Central Bank, sought to strengthen the country’s macro policy framework and financial stability. 30  See Annex 2 for a description of the country comparators. 31  One should be cautious in interpreting these results because TFP growth typically captures everything affecting growth that is not captured through capital accumulation and labor growth. These measurement problems are even larger when analyzing short periods of time. 32  For comparability, the benchmarking exercise uses UN National Accounts estimates. In some cases, the numbers defer from national sourc- es. For example, in the case of the DR, UN data points to investment growing at an average 7.2 percent between 1991 and 2016, while data from the Dominican Republic’s Central Bank points to higher growth closer to 9 percent. 16 | Dominican Republic: Systematic Country Diagnostic FIGURE 2: The DR’s recent growth episode has been accompanied by historically low levels of volatility Panel A: Coefficient of variation of real GDP per capita growth in the DR, over time 1,4 1,2 1,0 0,8 0,6 0,4 0,2 0,0 1965-1980 1981-1992 1993-2004 2005-2016 Panel B: Coefficient of variation of real GDP per capita growth (1992-2016) 12 10 8 6 4 2 0 Y Y Y L GT L T A DR V L L R A A A A R EX I M N N C O U DM I D D U B G G M S N CR FJ BO CH CO BE SL UR PR VC GU BH LK BR LC AT KN BR AR NI PE SU HN GR TT PA TU RO EC VE JA M Source: World Bank staff’s calculation based on data from WDI. 1980 and 1995 to an average 3.8 percent of GDP after sharp increase in FDI as a share of GDP occurred be- 1995 (Figure 4, panel A). FDI also increased noticeably tween 1996 and 1999, a period when the process of as a share of total investment over the past 25 years. It structural reforms accelerated in the country and tax accounted for 8 percent of total investment in 1991, incentives entered into force. and by 2016 that number had doubled to close to 16 percent. The country’s geographic characteristics have also fa- vored FDI inflows. Proximity to the US and Canadian Two policy factors that contributed to the sharp in- markets and producers33, the country’s Caribbean loca- crease in FDI inflows were the structural reforms and tion, with its beaches and easy access to the ocean, and stabilization policies of the 1990s and the explicit its newfound wealth of natural resources are three geo- efforts to attract foreign investments. Broad-based graphic attributes that have shaped the composition of structural reforms, like the enactment of law No. 16- FDI inflows, both in terms of origin and sectors of eco- 95 in 1995, which opened the possibility for foreign nomic activity. From 1993 until 2010, over 50 percent investments in almost every economic sector, were ac- of total FDI inflows came from North America. That companied by a series of laws providing tax incentives share fell in the aftermath of global financial crisis, but (CIT and VAT exemptions) to foreign investment. The still represented 34 percent of total FDI inflows in the 33  Beyond geographic proximity, the linkages between the US and the DR have historic roots, and were reinforced after the US occupied the Dominican Republic from 1916 to 1924 and again in the mid-1960s. Dominican Republic: Systematic Country Diagnostic | 17 FIGURE 3: Drivers of growth in the DR Panel A: Growth decomposition Panel B: Comparing drivers by sub-periods of growth across countries 7 6 6 5 contribution contribution 5 4 4 3 3 2 2 1 P rc nt P rc nt 1 0 0 -1 1990 1996 2001 2007 2012 PER LKA DR ROU ECU GTM TUN -1 -1995 -2000 -2006 -2011 -2016 -2 L bor C pit l TFP R l GDP rowth L bor C pit l TFP R l GDP rowth Panel: Investment and consumption growth in the DR and comparators, 1991-2016 10% 8% 6% 4% 2% 0% Tunisia Romania Guatemala LAC Honduras Ecuador Costa Rica Fiji Sri Dominican Peru median Lanka Republic Av r consumption rowth Av r inv stm nt rowth Sources: World Bank staff’s calculation based on data from The Conference Board, DR Central Bank (BCRD) and UN national accounts estimates. 2010-2016 period (Figure 4, panel B). The influence FDI, which stood at around 40 percent in the 1990s, of capital from North America on the DR’s economy fell to less than 15 percent in the 2011-2016 period. In can be more clearly seen in the special economic zones contrast, mining and real estate have gone from incipi- (SEZs), with a third of the firms being of US origin.34 ent levels in the 1990s, to representing over 30 percent This makes the US the largest foreign presence by num- in the 2010-2016 period. ber of firms, followed by Germany with only 2.5 per- cent of the total number of firms in SEZs. In terms of Changes in the growth and composition of FDI in- sectoral composition, two sectors that have a relatively flows are related to changes in the composition of de- large and stable participation in total FDI are industry mand and supply in the Dominican economy. On the and commerce (accounting for approximately 30 per- aggregate demand side, the growth rate of investment cent) and tourism (hovering around 20 percent) (Fig- fell in the 2000s (from an annual average of 13 percent ure 4, panel C). The rest of sectors have seen significant between 1992 and 2002 to 7 percent between 2004 and changes over time. The participation of utilities in total 2016). The deceleration in investment growth of the 34  See Consejo Nacional de Zonas Francas de Exportación (2016). 18 | Dominican Republic: Systematic Country Diagnostic FIGURE 4: FDI inflows into the Dominican Republic Panel A: Net inflows of FDI into the DR as a share of GDP 7 6 5 P rc nt of GDP 4 3 2 1 0 -1 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 2010 2013 2016 Panel B: FDI composition, Panel C: FDI composition, by sector by country/region of origin 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% 1993-2000 2001-2010 2011-2017 1993-2000 2001-2010 2011-2016 North Am ric Sp in R st of th World Comm rc /Industr /SEZs Tourism T l communic tions Oth r Europ n countri s LAC El ctricit Fin nc Minin R l st t Tr nsport Source: World Bank staff’s calculation based on data from WDI and BCRD. 2000s occurred while the country navigated through A look at the supply side: the rising two economic crises—the 2003 banking crisis and the role of services 2009 global financial crisis, which translates in a more stable participation in total GDP. This can be seen in Services has been the leading sector, consistently add- the specific case of FDI, which has stabilized at around ing 2.5-3 percentage points of growth per year over 4 percent of GDP since the late 1990s. Consumption, the past 25 years. Between 1992 and 2017, the broad on the other hand, accelerated from an average growth services sector35 grew at an annual average of 5.3 per- rate of 4.8 percent in the 1992-2002 period to 5.4 cent, exceeding the average annual growth of total value percent between 2004 and 2016. On the supply side, added of 5.1 percent (Figure 5, panel A), and contribut- FDI data shows an increase in the participation of ex- ing between 50 and 65 percent of the growth. Construc- tractive sectors and some services, like financial services. tion, which grew at an average annual rate that exceeded The next section looks in more detail at the supply side 8 percent, has contributed close to one percentage point of the Dominican economy and its implications for of GDP growth over the period, although its contribu- growth, productivity, and employment. tion declined in the 2000s. Agriculture and manufac- 35  These are all sectors except for agriculture, manufacturing, mining, and construction. Dominican Republic: Systematic Country Diagnostic | 19 turing have grown at an average of around 4 percent per size. The DR has a larger share of value added and em- year, and their joint contribution to growth has fallen ployment in services in 2016 than most of its structural from close to 30 percent of total growth in the 1990s peers. Moreover, a simple benchmarking exercise of val- (approximately 2 percentage points of GDP growth) to ue added in services as a share of GDP in 2016 and the 20 percent after 2000 (less than 1 percentage point). As share of employment in services, show that these two a result, the country has experienced noticeable changes numbers exceed what the country’s GDP and popula- in the sectorial composition of its GDP since the early tion predict (Figure 6). 1990s, with services increasing its weight in the econ- omy over time (52 percent of GDP in 1992 to over 65 Amid these changes in the structure of the econo- percent in 2017) and manufacturing and the primary my, labor productivity, defined as value added per sector declining by 12 and 6 percentage points between worker, has increased between 2000 and 2016, con- 1992 and 2017, respectively (Figure 5, panel B).36 sistent with TFP growth. Aggregate labor productiv- ity increased by 52 percent over the 2000-2016 peri- The service sector has also played a rising role in job od, which is equivalent to a 2.5 percent average yearly creation and employment growth. The total labor growth rate (Figure 7). Labor productivity has increased force increased at an average annual rate that has ranged in all sectors, but growth has been particularly strong in between 1.7 percent and 2.3 percent in different peri- the primary sector and construction. In general, labor ods over the past 15 years. Of this growth, employment productivity growth can be due to both within-sector growth in the broad services sector has been the largest productivity growth as well as reallocation of workers contributor, adding approximately 2 percentage points across sectors. Within-sector productivity accounted of total employment growth in each sub-period (Fig- for about 2.6 percentage points of productivity growth ure 5, panel C). The growth of total employment and per year, while reallocation effects had a small negative of employment in services has been particularly strong contribution of about 0.26 percentage points per year. since 2012. In contrast, the primary sector has seen an absolute contraction in the number of people working Productivity growth could be associated with the in the sector and the manufacturing sector has had an large presence of foreign firms and the country’s uneven performance.37 As a result, the share of employ- policy efforts to integrate into global markets since ment in the broad services sector in total employment the mid-1990s.38 In addition to tax incentives to for- increased by 10 percentage points between 2000 and eign investment, the country significantly reduced 2016 (from 60 to 70 percent) (Figure 5, panel D). In most-favored-nation (MFN) tariffs (non-preferential contrast, manufacturing and the primary sector have tariffs) since joining the World Trade Organization lost ground in terms of their incidence in total employ- (WTO) in 1995. From 1995 to 2014, the coun- ment, with declines of 3 and 7 percentage points be- try slashed its average MFN tariff by 7.5 percentage tween 2000 and 2016, respectively. points, the second largest reduction in LAC after Peru (see Bown et al. 2017). In 2006 the DR gained pref- Both in terms of employment and value added, the erential access to the US market when the Domini- services sector in the country exceeds that of its peers can Republic-Central America Free Trade Agreement and what is predicted by its level of development and (CAFTA-DR) with the US. DR-CAFTA has created a 36  Within the primary sector, most of the decline is explained by a fall in the share of agriculture. In contrast, mining has doubled its partici- pation from to 2 percent of total value added. 37  Manufacturing employment fell between 2000 and 2011, amid the DR’s banking crisis, a decline in demand for Dominican manufacturing due to the end of the multifiber agreement (MFA), and the global financial crisis. Since 2012, however, it has recovered, and has contributed positively to employment growth. 38  There are at least two channels through which foreign firms can have positive direct benefits in the local economy. First, they are typically highly productive from (high-income) countries closer to the technological frontier. Data show that multinational firms are highly innovative, representing 70 percent of business R&D in the world and with a high propensity to patent; see Javorcik (2010) and Lederman et al. (2014) for a discussion. Second, foreign firms create new jobs and pay on average higher wages compared to local firms. For evidence of externalities and spillovers through input-output linkages, see Javorcik (2004). For an assessment of the direct and reallocation effects see Alfaro and Chen (2012). For evidence on the positive impact of worker turnover from foreign firms to domestic firms, see Poole (2013). 20 | Dominican Republic: Systematic Country Diagnostic FIGURE 5: The service sector has been the main driving force of GDP growth and employment in the DR Panel A: Growth contribution, by sector Panel B: Evolution of the sectoral composition of value added 7 100% Sh r of tot l V lu Add d 6 contribution 80% 5 4 60% 3 40% P rc nt 2 20% 1 0 0% 1992- 1996- 2001- 2007- 2012- 1992 2002 2017 1995 2000 2006 2011 2016 Prim r M nuf cturin Construction S rvic s M nuf cturin Comm rc S rvic s T x s GDP Prim r Prim r Panel C: Sectoral contribution Panel D: Evolution of the sectoral of employment growth composition of employment of tot l mplo m nt 3,0% 100% 2,5% 80% 2,0% 1,5% 60% 1,0% 0,5% 40% P rc nt 0,0% 20% -0,5% -1,0% 0% 2001-2006 2007-2011 2012-2016 2000 2016 Prim r M nuf cturin Construction Prim r M nuf cturin Construction Comm rc S rvic s Tot l mplo m nt Comm rc S rvic s Source: World Bank staff’s calculation based on data from BCRD. wider and safer market for DR exports thereby having (Melitz 2003) and learning-by-exporting (Atkin, Khan- a positive overall impact for the economy (Calvo and delwal, and Osman 2017), but also because of pro- Cruz 2016 and Alemany 2011). ductivity-enhancing actions taken prior to becoming exporters (Labanca, Molina and Muendler 2013). FDI Trade openness and FDI can have a positive impact can increase a country’s productivity directly and can on a country’s productivity through several chan- yield productivity improvements in the local economy nels. Trade can give firms in a country access to higher through positive externalities to local firms and tech- quality inputs, which can in turn lead to productivity nological spillovers. These productivity-enhancing ef- gains. Import competition can force firms in a country fects can occur through direct market transactions (for to innovate (Amiti and Khandelwal 2013) and can force example, business relations between the foreign firm unproductive firms to exit the market. Also, exporting and domestic suppliers), worker turnover (that is, the firms have been found to be systematically more pro- movement of workers from foreign firms to local firms), ductive than non-exporters, in part due to selection through the reallocation of resources towards the most Dominican Republic: Systematic Country Diagnostic | 21 FIGURE 6: The DR has a larger share of services in value added and employment compared to peers and predictions Panel A: Benchmarking share of services in value added, 2016 80 70 60 50 40 30 20 10 0 Ecuador Nicaragua Honduras Peru Guatemala Sri El Salvador Romania Tunisia Dominican Fiji Costa Lanka Republic Rica S rvic s v lu dd d 2016 Pr dict d v lu s rvic s v lu dd d 2016 Panel B: Benchmarking share of services in employment, 2016 80 70 60 50 40 30 20 10 0 Romania Sri Honduras Guatemala Ecuador Peru Tunisia El Nicaragua Costa Dominican Fiji Lanka Salvador Rica Republic Emplo m nt s rvic s 2016 Pr dict d v lu s mplo m nt s rvic s 2016 Source: World Bank staff’s calculation based on data from WDI. Note: Predicted values are obtained from a regression of the share of services in either employment or value added in a specific year on log GDP and log population. FIGURE 7: Productivity has increased in all sectors, but employment flowed to lower productivity sectors Aggregate labor productivity growth and across sectors 90% rowth 80% 70% 2016-2000 l bor productivt 60% 50% 40% 30% 20% 10% 0% -10% -20% Primary Manufacturing Construction Services Value Added Growth in l bor productivit S ctori l r lloc tion ff ct Within s ctor compon nt Source: World Bank Staff’s calculation based on data from BCRD. 22 | Dominican Republic: Systematic Country Diagnostic productive firms, or indirect exposure to better technol- There is, however, little evidence of ogies (see Lederman et al. 2014 and references therein). quality upgrades and linkages between Dominican firms and firms in export- While there are no rigorous evaluations of the impact oriented sectors, suggesting that there are of FDI and trade on the country’s productivity and unexploited growth opportunities growth, a systematic analysis of the performance of Dominican firms and exports and the links between Despite their dynamism, Dominican firms lag those foreign firms and local firms, could provide indirect in other countries in terms of quality upgrading and evidence and shed some light on these potential im- their propensity to export. Firms in the DR have, on pacts. In what follows the chapter turns to this analysis.39 average, a lower propensity to hold internationally rec- ognized certifications compared to firms of similar size, age and sector of operation from other countries. Similar Fast firm-level growth but little gaps are observed in terms of in-the-job worker training evidence of quality upgrading and and process innovation (Figure 9, Panel A). Dominican weak links to exports firms have a lower probability to export compared to those with similar characteristics in terms of age, size, Dominican firms tend to grow faster geographic proximity to large markets, and other coun- compared to similar firms from other try characteristics (Figure 9, Panel A). Moreover, the countries gap between a firm’s observed probability of exporting and the level predicted by firm and country character- The economic dynamism of the last 25 years is clear- istics is larger in the DR than in comparator countries. ly reflected at the firm level. On average, data from the The observed gaps in terms of quality upgrading may World Bank’s Enterprise Surveys (WBES) show that for- be one of the factors explaining the low propensity to mal Dominican firms have grown at a relatively strong export observed for Dominican firms. Clearly these pace both in terms of employment and sales. Consistent averages mark a large degree of heterogeneity in firm with GDP data, employment and sales growth were performance, as many firms, especially those in SEZs, higher in the 2013-2016 period compared to the 2007- are competitive in global markets. Nevertheless, these 2010 period, and relatively higher than those observed patterns are indicative of the broad challenges of the in comparator countries (Figure 8, panel A). private sector. In fact, sales and employment growth rates of Do- The relatively small size of the DR’s economy, the minican firms were higher compared to firms of low propensity of Dominican firms to have interna- similar characteristics from other countries. Differ- tional recognized certifications, and proximity to the ences in growth rates between Dominican firms and US market, partly explain the low levels of backward those of other countries could be potentially driven linkages observed in the country, but there are ad- by firm characteristics (such as size, age or exporter ditional factors. Evidence presented in Sanchez-Mar- status), the sector in which the firm operates, integra- tin, de Pinies and Antoine (2015) shows that countries tion into global markets, and the year of the sample.40 of small size (measured by GDP) have lower levels of Data show that, on average, firms from the DR have backward linkages (measured by the share of inputs a higher growth rate compared to firms of similar indi- bought domestically). Additional factors that may affect vidual, sectorial and geographic characteristics (Figure a country’s decision between buying inputs domestical- 8, panel B).41 39  A detailed description of trade patterns and the performance of SEZs, tourism and agriculture, three outward-oriented sectors, is presented in Annex 4 and Annex 5, respectively. 40  On average, younger firms grow faster than older firms; see Lederman et al. (2014) and references therein. Similarly, there could be differ- ences in sectoral growth due to global demand and supply forces. Integration with specific markets and the year of the survey could also partly explain differences in growth rates. 41  These differences in employment and sales growth are statistically significant at the 90 percent significance level. Dominican Republic: Systematic Country Diagnostic | 23 FIGURE 8: Sales and employment growth rates in the DR have been high compared to peers Panel A: 3-year sales and employment growth across countries 0,40 r rowth 0,35 0,30 0,25 DHS S l s nd Emplo m nt 3- 0,20 0,15 0,10 0,05 0,00 Sri Lanka (2011) Sri Lanka (2011) Tunisia (2013) Colombia (2010) Guatemala (2010) Jamaica (2010) St. Lucia (2010) Romania (2013) Suriname (2010) Nicaragua (2016) Costarica (2010) Guyana (2010) Fiji (2009) DR (2016) DR (2010) Peru (2010) Panama (2010) Tunisia (2013) Costarica (2010) Suriname (2010) St. Lucia (2010) Jamaica (2010) Romania (2013) DR (2010) Guyana (2010) Nicaragua (2016) Guatemala (2010) Peru (2010) Fiji (2009) DR (2016) Panama (2010) Colombia (2010) -0,05 3- r mplo m nt rowth 3- r s l s rowth Panel B: Excess growth by DR firm compared to firms of similar characteristics 0,30 DR xc ss rowth r l tiv to b nchm rk 0,25 0,20 0,15 0,10 0,05 0,00 Controls: size, Controls: size, Controls: size, Controls: size, Controls: size, Controls: size, age, age, sector age, sector, age, sector, age, sector age, sector, sector, proximity proximity proximity and year proximity and year 3- r s l s rowth 3- r mplo m nt rowth Source: Authors' calculations using the World Bank’s Enterprise Service. Note: Growth is measured using the Davis, Haltinwanger, and Schu (1996) growth rates, which allows to calculate growth for firms with 0 initial values. ly or in international markets include the availability linkages and predicted levels is positive or non-signif- of local high-quality inputs and the costs of importing icant in all comparator countries apart from Ecuador high-quality inputs (Figure 10, panel A).42 Using an (Figure 10, panel B). extension of the regression analysis in Sanchez-Martin, de Pinies and Antoine (2015), Dominican firms dis- Low levels of backward linkages and low levels of play lower levels of backward linkages with local firms quality prevent the country from reaping larger than those predicted by observable factors such as firm benefits from the structural changes observed in characteristics, sectoral characteristics, country size, the export basket. Since the early 2000s the export and geographic proximity to high-income countries. In basket has tilted towards goods of higher technolog- contrast, the gap between observed levels of backward ical content, and of longer quality ladders, and has 42  The costs of importing are affected by distance and a country’s trade policy. The results in Javorcik and Spatareanu (2011) suggest that Multinationals have a higher probability of buying locally the farther away they are from their home country. 24 | Dominican Republic: Systematic Country Diagnostic FIGURE 9: Dominican firms lag those of other countries in terms of quality upgrading and likelihood to export Panel A: Benchmarking quality certifications, Panel B: Benchmarking process innovation, and employee training the likelihoodof exporting 0,4 Prob bilit of 0,3 tr inin work rs 0,2 0,1 Prob bilit of introducin n w proc ss s 0,0 -0,1 -0,2 Prob bilit of holdin qu lit c rtific tions -0,3 DR Ecu dor P ru Hondur s Fiji SriL nk Gu t m l Tunisi -0,4 -0,16 -0,14 -0,12 -0,10 -0,08 -0,06 -0,04 -0,02 0,00 Source: World Bank Enterprise Surveys (latest available year). Note: All results come from linear regressions of the dependent variable controlling for the firm’s size, age, sector and year fixed effect. In panel A additional controls include log of GDP, log of population, and the minimum distance to a G7 country or Russia. In panels B and D additional controls include a dummy variable taking value 1 if the firm has a quality certification, a dummy taking value 1 if the firm trains its workers, log GDP, share of firms in the country with quality certifications, log population, the minimum distance to a G7 country or Russia, a dummy variable if the country has an FTA with a G7 country or Russia, and dummy variables for the DR and its comparators. moved from final goods to intermediates and capital and geographic characteristics (Figure 10, panel D).44 goods (see Annex 4), changes that have the potential Moreover, quality upgrading occurs at a slower pace of fostering quality upgrading in the exports and larg- compared to what is observed in countries exporting er knowledge spillovers from exporters to local firms.43 similar products and with similar initial quality levels However, materializing these potential hinges in the as the DR.45 country’s ability to foster stronger ties between export- ers and domestic firms. Moreover, the growth of the Problems of perceived low quality are particularly export quality has been lower compared to countries evident for Dominican agro-exports. Rejection rates that export similar products and are at similar stages of at the US border are higher for Dominican fruits and development as the DR, limiting the country’s ability vegetables than those from other countries of the Cen- to take advantage of the changes in its export basket. tral America Free Trade Agreement (CAFTA). The Do- The average quality of Dominican exports, proxied by minican Republic has historically received favorable US the relative unit value of exports, has grown in recent market access due to strong economic ties and favor- years (Figure 10, panel C). In 2011 the average ranking able CAFTA-DR trading rules, but its ability to comply of an export product from the DR stood at the 18th with US Sanitary and Phytosanitary (SPS) regulation percentile of the quality distribution of that product. is low. In addition, evidence suggests that Dominican In 2016, the average ranking increased to the 30th products may be subject to more frequent and stringent percentile. Despite this improvement, the country’s inspections due to past compliance problems (Jouan- relative quality remains below that of structural peers jean et al. 2012). For more details about the perfor- and is below the level predicted by its size, income, mance of the agricultural sector see Annex 5. 43  Technological intensity is measured according to Lall (2000). Quality ladder estimates come from Khandelwal (2010). Export classification into intermediate, capital, final goods is from the BEC classification. Importantly, the attributes mentioned above are associated with higher growth (de la Torre et al. 2015). 44  The result refers to a regression of the median relative quality of a good exported by a country against the log of GDP, the log of population and the log average distance to trading partners. 45  A regression of the change in the relative quality of a country on initial quality and sectoral fixed effects shows that the DR’s change is significantly lower compared to benchmarks. Dominican Republic: Systematic Country Diagnostic | 25 FIGURE 10: Dominican firms create less backwards linkages compared to similar firms in other countries and exporters lag in quality Panel A: Correlates of backwards linkages Panel B: Benchmarking across countries and with comparators backward linkages 0,35 0,4 s 0,30 0,3 Diff r nc in b ckw rd link r l tiv to b nchm rk 0,25 0,2 0,20 Co ffici nt 0,1 0,15 0,0 0,10 -0,1 0,05 0,00 -0,2 -0,05 GDP Popul tion Dist nc FTA with Sh r -0,3 DR PER LKA FJI HND ECU TUN GTM to rowth rowth of firms pol s pol with qu lit c rtific tions St tistic ll si nific nt (10%) Not st tistic ll si nific nt St tistic ll si nific nt (10%) Not st tistic ll si nific nt Panel C: Evolution of median Panel D: Benchmarking median export quality in DR export quality in 2016 0,7 0,7 0,6 0,6 0,5 0,5 0,4 0,4 P rc ntil 0,3 0,3 0,2 0,2 0,1 0,1 0,0 0,0 SLV DR LKA PER 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 HND GTM CRI ECU TUN FJI 25th p rc ntil M di n 75th p rc ntil M di n Export Qu lit Pr dict d Source: World Bank staff calculations based on World Bank Enterprise Surveys (latest available year) and in CEPII. Note: Panel A and B are results from linear regressions of the dependent variable controlling for the firm’s size, age, sector and year fixed effect. In panel Aadditional controls include log of GDP, log of population, and the minimum distance to a G7 country or Russia. In panel Aadditional controls include log GDP, log population, the minimum distance to a G7 country or Russia, a dummy variable if the country has an FTA with a G7 country or Russia, and share of firms in the country with quality certifications. Predicted value are obtained from a regression of the country’s 2016 log median export quality on log GDP, log Population, and the average distance from trading partners. Poor performance in terms of quality upgrading may vival rate among LAC countries (Figure 11, panel B). A be one factor why, despite having relatively high en- new exporter in the country has a 5 percent probability try rates, Dominican exporters have lower survival of survival after three years, half the LAC average. This rates than any country in the region. The country ex- rate is also lower than the levels predicted by using a hibits entry rates (defined as the ratio of new exporters model like the one described above. to incumbent exporters) that exceed the level predicted by a regression model that controls for economic size Thus, policies and structural factors beyond size, and geographic variables (Figure 11, panel A). Howev- the sectoral composition of the economy, and prox- er, Dominican exporters exhibit the lowest 3-year sur- imity to the US, appear to deter quality upgrading 26 | Dominican Republic: Systematic Country Diagnostic FIGURE 11: The DR has a higher export entry rate than comparators, but underperforms in the terms of survival Panel A: Observed and predicted entry across countries 0,90 0,85 Firm Entr R t , pr dict d 0,80 0,75 BOL PRY GTM SLV NIC 0,70 ECU DOM 0,65 COL URY CRI MEX PER 0,60 CHL 0,55 0,50 0,3 0,4 0,5 0,6 0,7 0,8 0,9 Firm Entr R t , obs rv d Obs rv d vs pr dict d v lu s 45 d r lin Panel B: Observed and Predicted 3-year survival rates across countries Entr nt 3rd Y r Surviv l R t , pr dict d 0,14 0,12 0,10 MEX 0,08 COL CHL PER URY 0,06 CRI DOM ECU GTM SLV 0,04 BOL PRY NIC 0,02 0,00 0,00 0,02 0,04 0,06 0,08 0,10 0,12 0,14 0,16 0,18 0,20 Entr nt 3rd Y r Surviv l R t , obs rv d Obs rv d vs pr dict d v lu s 45 d r lin Source: World Bank staff calculations based on Export Dynamics Database. Data is available for a panel of countries between 1997 and 2014. Note: Predictions are the predicted value of a linear regression of the variable of interest on log GDP, log population, the average distance of the country to its trading partners, year and product fixed effects. and more integration between Dominican firms, from improving their quality is crucial to securing the hampering the country’s ability to achieve higher country’s future growth. The creation of more and bet- growth and create more and better jobs.46 Moreover, ter jobs could also help the country reducing its stub- the ability to reap the full benefits of the presence of bornly high level of informality, a topic that is further foreign firms depends on the ability of local firms to discussed in chapter 4.48 The rest of this chapter studies connect more robustly to local markets.47 This implies the potential causes preventing Dominican firms from that addressing the factors impeding Dominican firms becoming more outward-oriented and dynamic. 46  Evidence in Lederman et al. (2014) shows that employment growth and wages paid by firms are positively correlated with size and quality upgrading. 47  Fostering backward linkages is a necessary condition for knowledge spillovers from foreign firms to local firms, but it does guarantee these spillovers. 48  Lederman et al. (2014), using a framework proposed by Poschke (2013), argue that LAC’s high level of informality responds to the lack of dynamism of the region’s formal firms, as a potential worker’s decision to enter informality is associated with the scarcity of high-quality jobs in the economy. Dominican Republic: Systematic Country Diagnostic | 27 Challenges to competitiveness and critical to connect local suppliers and to connect Do- inclusive growth minican exporters to world markets. As described throughout this chapter, the country Infrastructure has improved but challenges remain. has benefitted from a relatively long period of eco- According to the World Bank’s Logistic Performance nomic progress, which is explained in part by a dy- Indicator (LPI), the quality of the trade and trans- namic private sector. The dynamism of the economy port-related infrastructure improved between 2007 and has been facilitated by private sector reforms. Based on 2016—the indicator improved by approximately 5 per- the World Bank Group’s Doing Business indicator the cent (from 2.18 to 2.29) and the DR cut the distance to Dominican Republic ranked above the regional aver- the best performers in the world. These improvements age for LAC, vis-à-vis business settings, at 99th out of notwithstanding, the country ranks 111th in a sample 190 countries. This compares favorably with its 2016 of 160 countries, and it lags regional peers, the LAC ranking of 103rd. The enhanced business climate is and Central American averages (2.45 and 2.43, respec- also evident in its international trade performance with tively). Similarly, according to the 2017-2018 Global strong growth in exports, largely in sugar, coffee, to- competitiveness Index (GCI) the country scores 3.3 out bacco, textiles, tourism, medical devices and electronics of 7 (101st place out of 137 countries worldwide) on (see World Bank 2015 and World Bank 2017). the quality of the overall infrastructure. ICT infrastruc- ture also appears to lag, with the country ranking 94th Nevertheless, there are still factors preventing the out of 137 countries in terms of internet bandwidth country from achieving higher and more inclusive (measured as kb/s/users), behind all LAC countries growth across sectors and stronger export perfor- with available data except for Paraguay and Venezuela.49 mance. Four characteristics appear to undermine the DR’s ability to achieve these objectives. Poor infrastruc- The relatively poor quality of infrastructure is most ture, especially in electricity, poor availability of an ad- visible in the electricity sector. According to the GCI, equately prepared labor force, high financing costs, and the quality of the supply of electricity ranks 125th out an uneven playing field, appear as the largest constraints of 137 countries in the world, a lower standing com- to a more robust process of economic development, pared to overall infrastructure. According to the World based on perceptions and international benchmarking Bank’s Enterprise Survey conducted in 2016, 14 per- of data. cent of Dominican firms rank electricity as the main barrier to growth, making it the second most common Despite improvements, the country’s response. This contrasts with other LAC countries, poor infrastructure constrains private where only 7 percent of firms report electricity as the sector development and export main constraint to firm growth.50 The poor quality of performance electricity provision is reflected in power outages due to rationing. While the electricity access rate in the DR is As an island state highly dependent on trade, good 98 percent, higher than the average for Latin America, road and port infrastructure is critical for success- it is estimated that only 52 percent of the electric con- fully creating functional domestic supply chains, to sumers have 24-hour service. These interruptions affect increase productivity and competitiveness, and to the production costs of Dominican firms and increases connect domestic SMEs into exports and global val- spending on alternative sources of energy. The country ue chains. The impact of trade on competitiveness is has the fourth highest frequency and third highest dura- clearly illustrated by oil. Oil products are imported and tion of power system interruptions among LAC coun- over 70 percent of the electricity generated depends on tries and structural peers (Figure 12, panel A). However, diesels and gas imports. Road and port infrastructure is contrary to Central American countries suffering from 49  See Global Competitiveness Report 2017-2018. 50  Enterprise Surveys: 2016 Dominican Republic Country Profile. 28 | Dominican Republic: Systematic Country Diagnostic FIGURE 12: The DR suffers from frequent and long power outages, which have large economic costs Panel A: Frequency and duration of system interruptions across countries, 2016 300 120 250 100 200 80 SAIDI SAIFI 150 60 100 40 50 20 0 0 CRI LCA MEXDOM PAN JAM ECU ROU LKA TUN CHI GTM ARG URY COL ATG BOL SLV TTO FJI PER BRA NIC DOM GUY HND S st m Av r Int rruption Dur tion Ind x (SAIDI) S st m Av r Int rruption Fr qu nc Ind x (SAIFI) Panel B: Cost of power outages across countries 9 s s sh r of s l s Th cost of pow r out 6 3 0 LCA JAM SUR ROU CRI ECU COL PAN GTM PER LKA GUY NIC TUN DOM FJI HND Source: World Bank staff estimates based on data from World Bank, Doing Business 2018, and World Bank Enterprise Surveys. low quality of electricity provision, the DR as an island and undermines business competitiveness, with month- economy has a limited ability to trade electricity and ly spending on backup generation increasing with firm connect its grid with nearby countries. size. Beyond power outages, the country’s electric sys- tem features the third highest industrial electricity The economic impacts of the problems of the elec- prices in the region at about 0.3 US$/kWh (Beylis and tricity sector are large. On average, Dominican firms Cunha 2017). Lower oil prices and changes in the com- lose 5 percent of their sales because of power outages position of the energy matrix may help lower the costs (Figure 12, panel B). This is the second largest value in of electricity. Nevertheless, the structural constraints of LAC, and the third largest value if one considers LAC the sector suggest that the additional costs induced by countries and the country’s structural peers. To mini- the sector’s problems will persist even if the efforts to mize the costs of outages, businesses must make large change the generation matrix materialize (Box 1). investments in back-up equipment, to be able to carry out their day-to-day activities. Approximately 95 per- The economic costs associated with problems in the cent of large industries have electric generators as well electricity sector are disproportionately larger for as 78 percent of small ones.51 This situation increases small firms. Analysis using data from the WBES shows the cost of production and goods and services provision that the relative losses associated with power outages 51  “Impact of the Crisis in the Electric Sector in the Dominican Economy,” IDB/INTEC. Dominican Republic: Systematic Country Diagnostic | 29 BOX 1. The structure and problems of the electricity sector The electricity sector is probably the paradigm of malfunctioning public goods in the Dominican Republic. High losses, high cost of energy generation, a weak regulatory environment, and the lack of transparency and account- ability around the use of government transfers to the sector, are persistent challenges despite the creation of a modern legal and regulatory framework. To improve the reliability and continuity of electricity service, the Dominican Republic (DR) began a series of structural power sector reforms at the end of the 1990s. This included allowing private sector participation in the generation and distribution sectors to expand the sector and improve the efficiency of service provision. In 2001 an Electricity Law was passed that created a modern legal and regulatory framework, including the creation of a new policy making institution, the National Energy Commission (CNE), and an energy sector regulatory institution, the Superintendence of Energy (SIE). Three regional distribution companies (EdeNorte, EdeSur and EdeEste) were also created to supply electricity service. These companies were initially privatized but later sold back to the State between 2003 and 2009. The Dominican Corporation of State-Owned Electricity Companies (CDEEE) was set up to control and coordinate all of the country’s state-owned electricity companies. This first wave of power sector reforms led to increased investment, technical capacity, and improved systems operations, but has not ensured the market’s financial sustainability. The Dominican experience with opening the electricity sector resulted in increased competition and private sector participation and was successful in in- creasing investment in generation capacity, improving operation of the National Interconnection Electricity System (SENI), strengthening the technical capacity of sector agencies and improving commercial practices. This invest- ment has assisted in diversifying the generation matrix of the country, transitioning from a matrix supported main- ly by fuel oil and hydropower to a more diversified one that also includes natural gas, coal, and other renewables. Despite reforms, the electricity sector still faces challenges that range from financial sustainability to prob- lems in the regulatory framework and the governance of the sector. At the point of the spear of the financial problems are the problems with distribution companies. The continuous cash deficits of the EDEs compromise the financial stability and viability of the entire sector and represent a large fiscal burden for the government. There are numerous factors behind the high financial deficit, including poor administration and commercial management, a tariff scheme that does not reflect the costs of delivering the service, and high technical and non-technical losses (due mainly to the poor condition of distribution networks, un-metered supply, theft, fraud, and mismanagement). Generation companies have been typically affected by payment uncertainty. Since the inception of the SENI there has been a significant debt owed by the distribution companies to the generation companies, in large part due to the gap between the applied and cost recovery tariff level as well as the high distribution losses. At the end of 2010 the debt represented around US$418 million, rising to US$781 million in 2014 prior to the fall in oil prices after which levels fell to US$202 million in 2015. These payment delays limit the cash flow of generators and their ability to generate electricity, and limits competition by deterring the entry of new companies. Similar problems of payment uncertainty also afflict transmission companies. While the past electricity reforms created a modern legal and regulatory framework, implementation of this framework has been less effective. The sector continues to suffer from a duplicity of functions among the main institutions. There is a duplication of functions among the newly established Ministry of Energy and Mines (MEM), CNE, and CDEEE with regard to the design of policies and planning for the power sector. There is no a single entity responsible for developing a long-term integrated strategic plan for the sector. Moreover, there is little continuity in the execution of the plans when there are changes in the administration. There is also ambiguity regarding the roles of the public versus private sector, distortions which hinders investment in the sector and effective and independent regulation. The persistence of the challenges described above suggests the sector is in an inefficient and stable equi- librium. According to Ruffín et al. (2014), traditionally reforms in the electricity sector have mostly focused on the careful design of technical solutions, ignoring the underlying political economy of policy implementation. Even well-designed technical solutions have produced limited results (investments in distribution in exchange for a com- 30 | Dominican Republic: Systematic Country Diagnostic mitment from the community to pay the energy bill), were reversed (pre-paid meters) or not implemented for years (a technical tariff scheme was prepared in 2010). An institutional and governance review for the country1 concludes that reforms are more likely to be blocked when the costs are concentrated in a small number of well-organized stakeholders while the benefits are widely shared among groups that are not well organized. On the contrary, re- forms are likely to succeed when the costs are widely shared in the society and benefits are concentrated in small groups (for instance, when the DR liberalized telecommunications at the end of the 90s). Given that most reforms required in the electricity sector will have costs concentrated in a small number of powerful stakeholders, the coun- try embarked in an open process, known as “Pacto Eléctrico”, to reach a consensus on a set of reforms in the sector. All the sector’s stakeholders were invited to participate in the Pacto, including generation companies, both public and private, and distribution companies (public sector) which would affect power purchase agreements in the years to come. The Pacto seeks to find a final solution to the issues of the electricity sector, so that the coun- try can count on “reliable, competitive and sustainable electricity for all”. Deliberations in the Pacto began in 2015, and after more than two years of deliberations, has not been signed and has been criticized by academics and some private sector associations for not addressing the structural issues of the sector. The apparent shortcomings of the Pacto illustrate the difficulties to break the long-standing inefficient equilibrium in the sector. Source: Carneiro et al. (2016). 1  See World Bank (2013). decrease with the firm’s size, even after considering sec- er expenditure needs that could enhance private sector tor characteristics and a firm’s age. This is reflected in competitiveness and meet social demands. It also adds the observed differences in responses by firms of dif- to the recent increase in government debt. ferent size regarding their main business environment constraint. Electricity is the most common response by small firms and the second most common response by Increasing human capital and forging medium-sized firms. In contrast, electricity does not ap- a more capable labor force will pear in the top 3 constraints of large firms. be crucial to the country’s ability to reap the benefits of structural The consequences of the problems of the electricity transformation sector may go beyond direct output losses and in- creases in production costs, as they distort quality The trade basket has moved to goods that demand upgrading decisions and exert fiscal costs on the Do- more skilled workers. The change in the relative de- minican government. Cross country evidence from the mand for skills in the exporting sector can be seen in WBES suggests that the likelihood of a firm holding SEZs. Since 2007, the skill composition of employment an internationally recognized quality certification de- in SEZs changed towards more skilled positions. Before creases with losses due to power outages, even after con- 2007, blue-collar workers accounted for approximate- trolling for size, age and sector of operation. In addi- ly 90 percent of total employment in SEZ firms. After tion, the financial losses of the electricity sector, arising 2007 the share of blue-collar jobs has steadily fallen, from operating and capital costs, are partly covered by reaching 80 percent in 2016 (Consejo Nacional de Zo- the Dominican government. Over the past 10 years the nas Francas de Exportación (CNZFE), 2016). More Dominican government spent on average 1.3 percent of broadly, the trade basket has changed in a way that re- GDP per year, or almost 10 percent of tax revenues. As sembles more those observed in high-income countries is discussed in chapter 4, this adds to the limited fiscal (the EXPY index has increased), which are also coun- space and limits the government’s ability to address oth- tries that rely more heavily on high-skilled workers.52 52  World Bank (2015). Dominican Republic: Systematic Country Diagnostic | 31 While the country has made noticeable progress in sources such as family and friends by connecting firms expanding access to education, the quality of educa- that are creditworthy to a broad range of lenders and tion remains a constraint to growth. The recent push investors. to increase public funds allocated to education has re- sulted in an expansion of access to education. This has Compared to other LAC countries, a larger share of resulted in a change in the composition of the labor formal Dominican firms uses bank credit to finance force, towards higher educational attainment. Despite operations and investment. Data from the 2016 this positive sign, the country underperforms in terms WBES show that approximately 57 percent of Domini- of standardized tests, as discussed in detail in chapter can firms have a bank loan or line of credit. In contrast, 3. The poor quality of education is also perceived by the share of firms that have a bank loan or line of credit private sector actors. The GCI ranks the DR 123rd out in the average LAC country is 47 percent. In addition, of 137 countries in an index of quality of education, Dominican firms finance a larger percentage of their putting the country behind all LAC countries except for purchases of fixed assets through banks compared to Nicaragua, El Salvador, Haiti and Paraguay. firms in other LAC countries (28 percent compared to 20 percent). As a result, DR firms rely less on internal Lacking adequately prepared workers constraints funds for the purchase of fixed assets compared to firms the growth potential of large exporting firms, which from other LAC countries (55 percent compared to 64 have the largest potential to create high quality jobs. percent).54 An inadequately prepared labor force is perceived as major business constraint by 31 percent of firms in the The country also stands above or close to regional 2016 WBES. It is the 3rd most common response for numbers in measures of access to finance for individ- large firms, with 14 percent of all large firms reporting uals. Data from the Financial Inclusion Dataset (Findex) it as the main business obstacle, and the 4th most com- shows that a larger share of Dominicans ages 15+ had mon response among exporters, with 11.5 percent re- savings at a financial institution compared to other LAC porting it as the main obstacle to growth. This, in turn, countries (26.5 percent compared to 13.5 percent), and implies that a limited supply of appropriate skills limits borrowed from a financial institution (18 percent com- the overall ability of the country to create well-paying pared to 10 percent). As in LAC, close to half of Domin- jobs and reduce informality, since exporters and large icans ages 15+ have a bank account, and over 60 percent firms tend to pay higher wages, on average, compared of those who borrowed from financial institutions used to small, non-exporting firms.53 those funds for a farm or business venture. The relative good standing in terms of access to fi- The DR compares favorably to other nance is also reflected in firm perceptions, although countries in the region in terms of there is some variation across firms of different char- access to finance, but capital markets acteristics. Approximately 5 percent of firms list lack are shallow, financial costs are high, of access to finance as the main business obstacle in and the country lacks a robust the country, compared to 14 percent of firms in LAC. insurance market Non-exporting firms are more likely to report lack of access to finance as the main obstacle compared to ex- Access to financial markets allow firms to fund the porting firms (5.5 percent compared to 2 percent). Sim- purchase of fixed assets – such as buildings, land, ilarly, smaller firms tend to view problems of access to machinery, and equipment – and working capital finance as larger hurdles compared to larger firms—6.3 necessary to operate and grow. Efficient financial mar- percent of small firms report it as the main business ob- kets reduce the reliance on internal funds or informal stacle compared to 3 percent of medium and large firms. 53  In addition, countries where the growth rate of large firms is higher experience larger reductions in informality. See Lederman et al. (2014), and Lederman, Maloney and Messina (2011). 54  Other sources of financing include equity and supplier credit. 32 | Dominican Republic: Systematic Country Diagnostic FIGURE 14: The DR has a larger gap between prices at the destination and prices in the point of origin, even after controlling for factor affecting this gap 0,5 0,4 Co ffici nt DR 0,3 0,2 0,1 0,0 No controls GDP and GDP, population GDP, population, GDP, population, (***) population (***) and distance (***) distance and distance, infrastracture infrastracture (***) and FE products (***) Source: World Bank’s staff calculation based on CEPII (Trade Unit Values); WDI, World Bank and Global Competitiveness Report. Note: The DR coefficient is the coefficient of a dummy variable taking value 1 if the exporter is the DR in a regression of the log difference between CIF and FOB prices on a series of controls described in the x-axis. Despite having access to finance, Dominican firms that real lending rates decrease with GDP, increase with face higher real interest rates than firms in most government debt (a proxy for economic risk), increase countries in LAC or from comparator countries, and with the incidence of interest payments on public debt real interest rates have been on the rise since 2010. (a proxy for the opportunity cost of banks), and in- The average real lending rate charged stood at around crease with natural disasters (all else equal, Caribbean 14 percent in 2016, more than 6 percentage points and Central American countries have larger real inter- higher than the regional average (Figure 13, panel A). est rates than other countries in the world). In the case The average real lending rate is the third highest rate of the DR, while bank concentration stands below the among LAC countries and comparator countries from regional average, exposure to natural disasters and the other regions (Figure 13, panel B). The real lending rate country’s relatively high interest burden of government in the country has been on the rise since 2010 while the debt appear to be larger than in other countries. The in- LAC average has hovered around its average. cidence of interest payments on gross debt in the coun- try was the highest among comparator countries from High real interest rates may be one factor behind the LAC and other regions.55 relatively low credit volumes. Total credit to private sector and total deposits have one of the lowest pen- Government debt also appears to crowd-out private etrations relative to GDP of all LAC countries. Total investment in capital markets. Capital markets in credit to the private sector in the DR was 28 percent of Dominican Republic are shallow, featured by a crowd- GDP, the third lowest value in LAC. Commercial loans ing-out effect between public/private sectors, and fi- to SMEs increased from around 12 percent of GDP in nancing for large businesses provided mainly by banks. 2007 to 17 percent in 2015, but average loan amounts Equity shares are not publicly traded, and there are only remain low and maturities are typically short. around 20 corporate debt issuers in the local stock ex- change “Bolsa de Valores de la República Dominicana” The high real lending rates appear to be related to (BVRD). Stock market capitalization to GDP is one of the high interest burden of government debt, the ex- the lowest in the region. The government is the main posure to natural disasters, and other country risks. issuer, and public debt accounts for the bulk of trades. A cross-country regression of real lending rates shows Returns on these instruments are higher than commer- 55  The DR has the second highest interest payment as a share of government revenue in LAC, behind Brazil. Dominican Republic: Systematic Country Diagnostic | 33 cial rates, and will deter inter-bank-rate competition market dominance in the country have worsened—the and corporate debt issuance. The fact that the govern- indicator decreased from 2.8 in the 2007-2008 report ment issues long-term debt through two different en- to 2.6 in the 2017-2018 one.56 Market dominance may tities, the Central Bank and the Ministry of Finance, be one of the forces driving the country’s limited qual- with a substantial rate spread for the same underlying ity upgrading. Using data from WBES, Lederman et risk and tenor, will continue to send mixed signals to al. (2014) show that higher levels of competition are the market and undermine price discovery. positively associated with a variety of innovation and quality indicators. Market power, institutions, and Market power erodes the country’s competitiveness fiscal incentives limit the country’s in international markets, which is crucial for an is- ability to produce a level and well- land economy that aims at integrating robustly in functioning business environment, the global stage. First, lack of quality upgrading and hampering competitiveness and innovation by Dominican firms makes Dominican ex- reducing incentives to innovate ports, and the inputs used in these exports, less compet- itive vis-à-vis those of other countries. In addition, firm Having a level and well-functioning business envi- concentration and monopolistic practices in non-trade- ronment, where resources are efficiently allocated able sectors used intensively by exporters could hamper and firms are given incentives to innovate and invest, the country’s competitiveness. In fact, like many other is crucial to a country’s ability to generate inclusive countries in LAC, the DR has non-tradeable sectors that growth. All else equal, a level playing field allows firms are more concentrated compared to other global peers to compete for scarce resources and to innovate to try and to what their income and size predicts. (Lederman to outperform their competitors (Aghion et al. 2005). et al. 2014). Similarly, having contractual certainty and fair treat- ment encourages investments. Creating a business envi- In the specific case of ground transportation, mar- ronment that generates a level playing field and fosters ket power, compounded by a lack of scale, leads to quality upgrading is crucial for the competitiveness and higher prices and discourages investments, making ability to strengthen backward linkages in an economy it more difficult for exporters to connect with local that is integrated in global markets. A well-function- suppliers. The country shows much lower fleet produc- ing business environment also relies on institutions that tivity than countries in Central America, driven by av- facilitate the flow of information among as many pri- erage trucking tariffs that at 4.75 USD/km, more than vate agents as possible, in such a way that they make triple the average of Mesoamerica; very old and inef- decisions with as much information as possible. Thus, ficient truck fleet with an average age of 21 years (by well-functioning markets take into account the public comparison on average LAC trucking fleet is 15 years good nature of information. old and OECD is 8 years old);57 and average distances travelled do not allow to take advantage of the econo- The country stands out as one of the worst perform- mies of scale of the sector (the DR’s average mileage is ers in LAC and in the world in terms of perceived 9,000 km/truck/year much lower than the 60,000 in market dominance. The 2017-2018 Global Compet- Panama, and 42,000 in Salvador). itiveness Report ranks the country 7th in the world in terms of market dominance, making it the second high- In addition, firm concentration and monopolistic est in LAC after Haiti. Moreover, a comparison over practices in non-tradeable sectors used intensively time shows that in the past 10 years perceptions about by exporters could erode the country’s natural geo- 56  The Global Competitiveness Report rates countries in a scale from 1 to 7 in several dimensions based on surveys and data. In the specific case of market dominance, the survey question asks “In your country, how do you characterize corporate activity? [1=dominated by a few business groups;7=spread among many firms]”. 57  See Barbero and Guerrero (2017). 34 | Dominican Republic: Systematic Country Diagnostic graphic advantage to compete in US markets. An SPS and pesticides. Existing institutions in charge of analysis of the difference between export prices in des- the surveillance of animal health and control of plant tination markets (CIF prices) and prices at the port of pests could also benefit from capacity building. In ad- origin (FOB prices), shows that Dominican exporters dition, vaccination campaigns and wider dissemina- have a larger gap between these two prices even after tion of risk-mitigation methodologies could be orga- controlling for the GDP and population of the export- nized. er, the GDP and population of the importer, the dis- tance between markets, the type of products exported, The country’s institutional arrangement also presents and the quality of infrastructure (Figure 14). These re- problems of design, constraining the efficient flow sults suggest that lack of competition among shipping of information among private agents. In the case of and insurance companies may increase the costs of ex- the institutional infrastructure supporting internation- porting in the DR. Indeed, Mesquita Moreira, Volpe al trade, there is a marked duality and fragmentation, and Blyde (2008) present evidence that low levels of which is likely to result in suboptimal export promotion competition among shipper lines is an important de- and entrepreneur support efforts (World Bank, 2015). terminant explaining the difference between the price The National Council for SEZs and the center for ex- of Dominican exports in the US and the price of Euro- ports and investment (CEI-RD) are de facto in charge pean exports. of investment attraction inside and outside SEZs. A similar duality is found in private exporter associations. Despite recent progress, the DR lags other coun- In spite of commendable coordination efforts, duplic- tries in the region regarding competition agencies. ities are likely to appear, and achieving economies of The 2017-2018 Global Competitiveness Report ranks scale in supporting exporters would be more difficult. the country next to last among 137 regarding the ef- Under its current structure, CEI-RD may not be suffi- fectiveness of anti-monopolistic policies, and has the ciently empowered to efficiently perform its mandate to third lowest score among LAC countries. The lack of promote exports and attract investment. Its budget per effectiveness of pro-competition policies is partly due employee is the lowest in Latin America, indicating a to delays in the full establishment of Pro-Competencia, heavy staffing burden. This limits space for commercial the pro-competition agency. Efforts to pass competition support and technical assistance as well as its presence laws in the date back to 1997 and the competition law abroad. Most Dominican trade representatives abroad was signed in 2008, but failure to appoint an Executive are in fact diplomats belonging to the Ministry of For- Director until early 2017 blocked the enforcement of eign Affairs who devote only part of their time to export the Competition law and prevented any legal investiga- promotion. tions from being launched.58 An additional factor creating an unbalanced play- Beyond lack of competition, the quality of exports ing field is the presence of special tax regimes. and inputs is affected by a weak regulatory and in- According to the Ministry of Finance, at the end of stitutional capacity, clearly evidenced in the case of 2016 there were 8 tax laws that generated tax expen- agricultural export rejections. As argued in World ditures, and 31 regulations granting tax exemptions Bank (2015), there is a need to create standardization to different sectors and for different reasons.59 These bodies, system certifications, conformity assessments, tax regimes grant exemptions on a variety of taxes, in- and accreditation mechanisms. A deeper involvement cluding corporate income tax, value added tax, and in international SPS standard setting institutions as import tariffs. The rationale for these tax exemptions well as the WTO SPS committee could ensure that is the creation of new jobs by firms that get established the Dominican support infrastructure conforms inter- in the country because of these incentives. Howev- national norms and prompts updates of regulations on er, the use of incentives results in forgone fiscal rev- 58  Article 67 of the competition law (42-08) stipulates that the law would become effective as soon as the Board of Directors and the Executive Director take office. 59  Ministerio de Hacienda (2016). Dominican Republic: Systematic Country Diagnostic | 35 BOX 2. Fiscal incentives and FDI attraction: Results from a global survey Fiscal incentives are a common tool used by developing countries to attract FDI. Developing countries pur- sue policies to attract multinational corporations (MNCs) with the goal that these firms create jobs and generate knowledge spillovers to domestic firms. However, many countries seek the attraction of a limited number of firms, leading to competition in the incentives offered to these firms. As a result, between 50 and 70 percent of develop- ing countries with available data offer corporate tax incentives to attract multinational corporations operating in specific sectors into the country (World Bank 2018). Tax incentives are an important consideration for foreign firms in deciding the country where they ultimately operate. The 2017/2018 Global Investment Competitiveness Report (World Bank 2018) shows that 59 percent of foreign firms of a global sample of MNCs say that low taxes are an important factor in their decision to invest in a country. The importance of low taxes appears to be greater for efficiency-seeking FDI (foreign investment that seeks to save costs in international production networks), compared to other types of foreign investment (invest- ment seeking market access or assets specific to a country)—64 percent of efficiency-seeking firms rank low taxes as an important factor compared to 52 percent of non-efficient seeking foreign firms. However, tax incentives are not the only factor affecting foreign firms’ location decisions. Political stability and a strong legal and regulatory environment are important factors affecting MNCs’ decisions to locate in specific countries—approximately 85 percent of foreign firms (both efficiency-seeking and non-efficiency seeking ones) reported these factors as important determinants of location decisions. Similarly, the availability of a talented and skilled workforce is another important factor affecting FDI decisions, especially for efficiency-seeking firms (78 percent of efficiency-seeking firms and 68 percent of non-efficiency seeking MNCs report it as an important determinant of location decisions). Thus, the country’s ability to raise revenue by means of rationalizing tax exemptions in the medium term will hinge on the country offering political and institutional stability, as well as a well-prepared labor force. Tax exemptions are not the only path to attract FDI. However, the viability of an FDI attraction model that reduces or eliminates tax exemptions in the medium-term will depend on the country’s ability to improve the quality of mac- roeconomic, political and regulatory institutions and producing a better qualified workforce. enue. In addition to the fiscal implications, that are ers. In addition to tax benefits, many special regimes discussed in chapter 4, special tax regimes create an (like SEZs and certain touristic developments) do not unlevel playing field that reduces the tax burden for face problems in the provision of electricity, since they certain economic actors and sectors. The differences rely on distribution by private companies or buy elec- in the tax burden across types of firms are clearly evi- tricity from public distribution companies at interna- denced in firm perceptions reported in the WBES. In tionally competitive prices.60 Tax exemptions have a the 2016 survey, non-exporters were twice as likely as long history in the DR and have been an important exporters to list taxes as the main business obstacle (12 instrument in the country’s industrialization and de- percent compared to 6 percent). Taxes were the 3rd velopment strategy. In recent decades, the country has most common response among non-exporters, while established clearer procedures to grant these exemp- it was the 8th most common response among export- tions. However, this has not prevented private actors 60  For example, the Consorcio Energético Punta Cana-Macao (CEPM) is a privately-owned utility company that generates, transmits, dis- tributes and commercializes energy in the touristic areas of Punta Cana, Bávaro and Bayahíbe. Many SEZs are located in industrial parks labeled as energy non-regulated users (usuarios no regulados) and can have direct power-purchase agreements (PPA contracts) with energy generating companies to satisfy their electricity demands at lower prices and better reliability than when purchasing electricity from EDES (government energy distribution companies). 36 | Dominican Republic: Systematic Country Diagnostic from taking advantage of special tax regimes to sus- ing accountable institutions. The WDR 2017 Gover- tain an advantage over other sectors of the economy nance and the Law poses that the challenges that mid- (Daude, Gutiérrez, and Melguizo 2014). For example, dle-income countries face to leave the middle-income the end of export quotas without modifying tax in- trap go beyond policy choice to the challenge of power centives has increased the number of Dominican firms imbalances. The World Bank’s World Development into SEZs, which in 2016 accounted for approximate- Report (WDR) finds that at upper-middle-income lev- ly 35 percent of the total of firms operating in SEZs, els, legislative, judicial, media, and civil society checks and leads to some domestic firms enjoyed benefits that become increasingly important, with much larger im- others do not.61 provement for countries that escaped the middle-in- come trap (“escapees”) than for countries that stay in Special tax regimes also create a bias towards im- the trap. According to the data used in the WDR, the ports, which is amplified by the poor quality of country has made less progress on improving judicial Dominican inputs, discouraging backward linkages checks on the executive and guaranteeing a media free and limiting the scope for knowledge spillovers. The of influence than non-“escapees” (Figure 15, panels Central Bank of the Dominican Republic published in B-C). Consistent with the hypothesis that bigger mar- 2014 the results of a census of SEZ firms that shows kets and more complex business networks require a that these firms import 81 percent of their inputs. rules-based approach, the WDR finds that “escapees” The study also found significant variation between lower their levels of corruption significantly before be- the sourcing patterns of traditional zone industries coming high-income economies, whereas “non-escap- like textile and clothing, and footwear (which source ees” do not see an improvement. The DR has been less 28 percent and 22 percent of their inputs domestical- effective in reducing corruption than the set of coun- ly, respectively) and newer industries like medical and tries that remain stuck in the trap (Figure 15, panel A). surgical equipment, and electrical equipment (which This suggest that improvements in the control of cor- source less than 3 percent of their inputs domestical- ruption and accountability of institutions the country ly). These results show that the challenges of creating aims to reach high-income status by 2030, as stated in backward linkages will likely increase as the structure its national development strategy (Estrategia Nacional of SEZs changes. Tariff exemptions could be a force de Desarrollo - END). behind these patterns—the reliance on imported in- puts is three times as large for SEZ firms compared to non-SEZ firms (World Bank, 2017). Another fac- tor deterring higher levels of backward linkages could be the quality of Dominican inputs, an issue that was discussed earlier in the chapter. A similar bias towards imports is observed in the tourism sector, where all-in- clusive hotels import a large share of their inputs. In addition, the all-inclusive nature of tourism in the country limits the potential for tourists to consume domestic goods and services. More broadly, the enclave nature of SEZs and other special regimes, limits their positive spillover effects to the local economy. The creation of a level playing field for firms, one of the requirements to make growth more inclusive in the country, emphasizes the importance of hav- 61  The expiration of the deadline to make the SEZ program in the DR compatible with the WTO Agreement on Subsidies and Countervailing Measures led to the elimination of the export share requirements (ESR) to access to SEZs– firstly for producers of leather, textiles and apparel and leather, sectors designated as ‘national priority’ in 2007, and afterwards for all SEZs producers in 2011. Dominican Republic: Systematic Country Diagnostic | 37 FIGURE 15: Checks on corruption and accountability institutions for middle-income countries and the DR Panel A: Public sector corruption Panel B: Judicial constraints on the executive 0,0 0,2 in ind x scor in ind x scor -0,03 0,1 -0,05 ch n ch n 0,0 -0,08 Av r Av r -0,10 -0,1 DR Non- sc p s Esc p s DR Non- sc p s Esc p s Panel C: Government influence of media 1,5 in ind x scor 1,0 0,5 ch n 0,0 Av r -0,5 DR Non- sc p s Esc p s Source: WDR 2017 and own calculations using data from V-Dem 2015. CHAPTER 3 Drivers of inclusion T his chapter describes recent progress in poverty reduction, shared prosperity, and equality of oppor- tunities in the DR, followed by an analysis of the two main drivers of inclusion: labor markets and social policy. Due to the importance of both emigration and immigration, the issue is covered in one of the sections. The country has recently made important progress in poverty reduction and gains in the size of the middle class. However, there are large and persistent spatial and urban/rural disparities in terms of poverty, access to services, and equality of opportunities. Limited fiscal space, clientelistic prac- tices, and low internal and external accountability for service delivery have sustained low levels of public investment and shortcomings in human capital formation and the provision of public goods. Low pro- ductivity and human capital levels, added to high migration flows contribute to shaping labor markets that do not connect the labor force, especially women and youth, to more and better jobs. Despite overall improvement in most human development outcomes, deficits in human capital formation throughout the life cycle perpetuate poverty and reduce income earning potential. These challenges are important barriers for the country to achieve more inclusive growth and to sustain a high growth trajectory. To assess a country’s performance in terms of translating national poverty lines, moderate poverty was reduced its economic growth into monetary poverty reduction, it from 43.2 percent in 2008 to 28.9 percent in 2016, is important to cross-check relevant comparable aggregates while extreme poverty was reduced from 12.8 percent for households from national accounts and household in 2008 to 6 percent in 2016.62 Poverty has been per- surveys. Box 3 below describes the results of these checks. sistently higher in rural areas but the gap with urban areas is shrinking (Figure 17). Moderate poverty in ru- ral areas was 16 percentage points (pp) higher than in Recent progress on poverty urban areas in 2008 (55 vs 39 percent), and the differ- reduction, shared prosperity and ence was reduced to 11 pp in 2016 (38 vs 27 percent). equality of opportunities For extreme poverty the gap between the two areas de- creased from 11 pp in 2008 (21 vs 10 percent) to 4 The country has recently made important progress percent in 2016 (9 vs 5 percent). Using decomposition in poverty reduction, especially since 2014. Using techniques, only 9 percent of the reduction in moderate 62  Despite the reduction on extreme poverty, the country missed its MDG goal to bring it to 5.4 percent in 2015. 40 | Dominican Republic: Systematic Country Diagnostic BOX 3. How do national accounts and household surveys compare in the DR? When linking economic growth and poverty, it is important to check the consistency of these two data sources. The household survey used to measure poverty in the country is the labor force survey (Encuesta Nacional de Fuerza de Trabajo – ENFT), and poverty is measured using an income aggregate. Conceptually, the concept that is closest to household income in national accounts is private consumption given that i) financial penetration in the country is low, limiting the ability of households to smooth consumption; and ii) the ENFT fails to capture returns to capital. Private consumption includes final consumption by households and by non-profit institutions serving households. According to the supply-use tables for the period 2007-2012 (latest available), households account for more than 99 percent of private consumption. The base year for the national accounts in the country was changed to 2007. The new year better reflects the structure of the Dominican economy and uses inputs that were not available in the construction of previous Na- tional Accounts series. As part of the project, the country collected for the first time a national census of economic activity and a directory of formal establishments, and it started the systematic production of a SEZs census and a survey of the tourism sector. The labor force survey has also seen improvements over time. Its sampling frame was updated in 2003 and 2008 to reflect population figures from the 2002 Census and a more accurate cartography for the capital city Santo Domingo. In 2005 questions were added and changed to better capture household income. Improvements in both data sources seem to have resulted in a better alignment between private consumption from national accounts and household income from the household survey. To illustrate this point, indices for these two variables are constructed using different base years. For instance, using 2002 as base year, the private consump- tion index in 2015 is 70 percent higher than the household income index. This difference declines with the base year and it becomes negligible when 2008 is used as base year (see Figure B3.16 for indices using 2005 and 2008 as base years). Due to a better alignment between the household aggregates from national accounts and the labor force survey we limit the analysis linking economic growth and monetary poverty to the period starting in 2008. FIGURE B3.16: Indices for private consumption (from national accounts) and household income (from household survey) in real terms Panel A: Base year 2005 Panel B: Base year 2008 1,8 1,4 1,6 1,2 1,4 1,2 1,0 1,0 0,8 0,8 2005 2007 2009 2011 2013 2015 2008 2009 2010 2011 2012 2013 2014 2015 HH surv incom ind x Priv t consumption (NA) ind x HH surv incom ind x Priv t consumption (NA) ind x Source: Authors’ calculations using data from national accounts (Central Bank of the DR) and ENFT. Dominican Republic: Systematic Country Diagnostic | 41 Figure 17: Monetary poverty rates by area of residence Panel A: Moderate poverty rate Panel B: Extreme poverty rate 60 60 50 50 40 40 30 30 20 20 10 10 0 0 2001 2006 2011 2016 2001 2006 2011 2016 N tion l Rur l Urb n N tion l Rur l Urb n Source: Authors' calculations using ENFT. poverty and 13 percent of the reduction in extreme pov- 1,000 a decade earlier). But child mortality remains erty are explained by rural-urban migration. three times higher for the poorest quintile (42 deaths per 1,000 births) compared to the wealthiest quintile Access to services and most human development (17 deaths per 1,000 births). In addition, performance indicators have improved, but with persistent in- in math and language in 3rd grade national assessments equalities. Non-monetary poverty, measured as the is substantially lower for children from poorer house- share of the population with no access to key basic holds.66 services, steadily decreased from 44 percent with no access to two services in 2000 to 25 percent in 2016, Growth over the 2008-2016 period was slight- while the share lacking access to four services declined ly pro-poor according to incidence analysis. from 16 percent to 6 percent in the same period.63 Per-capita income for individuals in the bottom 40 Between 2000 and 2016 there was a notable expansion percent grew at an average annual rate of 4.2 per- in sanitation, especially in rural areas, where the share of cent, compared to 3.8 percent per year for the top the population with no access to sanitation was reduced 60 percent (Figure 18, panel A). Changes in labor by 30 percentage points.64 Most human development income explain most of the growth in total in- indicators also improved for the population as a whole, come. In regional context the growth of per capita and for the poor, although inequalities remain. Chron- income for the bottom 40 percent in the DR ranks ic malnutrition among children (under age 5) declined tenth among 17 countries in Latin America for the from 11 percent in 2002 to 7 percent in 2013, yet 11 2008-2016 period, despite having the fourth high- percent of the poorest quintile were chronically mal- est average GDP growth rate (Figure 18, panel B). nourished in 2013.65 Child mortality declined modestly in the past decade (31 deaths of children under age five The country also shows improvements in terms of in 2013 per 1,000 births, compared to about 42 per having a slightly less unequal income distribution. 63  Indicators of access to services include school attendance, level of schooling, drinking water, sanitation, quality of housing, overcrowding, and ownership of assets. 64  Access to sanitation is defined as connection to the public sewage system or septic tanks. 65  On the other hand, there are more overweight children (11 percent) in the highest income quintile compared to 3.6 percent in the lowest income quintile. Being overweight also presents health risks for noncommunicable diseases such as cardiovascular disease and diabetes. 66  MINERD (2018). Resultados de la evaluación diagnóstica nacional de tercer grado de primaria: informe nacional. Santo Domingo, República Dominicana. 42 | Dominican Republic: Systematic Country Diagnostic Figure 18: Growth incidence and shared prosperity Panel A: Growth incidence in the DR 2008-2016 6 rowth 4.2 p rc nt 5 3.8 p rc nt 4 nnu l p rc nt 3 2 1 0 Av r 1 2 3 4 5 6 7 8 9 10 -1 D cil s L bor incom Non l bor incom Source: Authors' calculations using ENFT. Panel B: Shared prosperity in LAC countries 2008-2016 9 rowth r t 6 3 Annu li d incom 0 -3 PAN CHI NIC COL BOL PER URY ECU BRA DR PY SLV ARG CRI MEX HND GTM -6 Ov r ll incom rowth Bottom 40 Incom rowth Source: LAC Equity Lab. Figure 19: Gini index by income source in the DR and for LAC countries, 2008-2016 Panel A: Gini index by income source, DR Panel B: Gini index for LAC countries, 2008 and 2016 0,8 0,6 0,7 0,5 0,6 0,4 0,5 0,4 0,3 2008 2009 2010 2011 2012 2013 2014 2015 2016 SLV BOL HND URY ECU COL PER PAN PRY BRA DR ARG L bor incom Non-l bor incom For i n incom Gini 2008 Gini 2016 Source: Authors' calculations using ENFT. Source: SEDLAC. Note: countries are sorted in descending order according to the change in the Gini index between 2008 and 2016. Dominican Republic: Systematic Country Diagnostic | 43 Figure 20: Poor, vulnerable and middle-class distribution in LAC and DR (2008-2016) Panel A: Population distribution in the DR Panel B: Population distribution in LAC 0,5 0,5 Sh r of th popul tion Sh r of th popul tion 0,4 0,4 0,3 0,3 0,2 0,2 0,1 0,1 2008 2009 2010 2011 2012 2013 2014 2015 2016 2008 2009 2010 2011 2012 2013 2014 2015 2016 Poor Vuln r bl Middl cl ss Poor Vuln r bl Middl cl ss Source: Authors' calculations using SEDLAC. The Gini coefficient decreased by 2 points from 49.6 Mobility across income groups between 2008 and in 2008 to 47.1 in 2016, below regional inequality 2016 was accompanied by the poor shifting closer to levels throughout the period. Inequality in the dis- the poverty line and the vulnerable shifting further tribution of all income sources declined in this peri- away from it. As seen in Figure 21, panel A, a high- od, with labor income being more equally distribut- er share of the poor (area under the density curve) is ed than non-labor income and income from abroad close to the poverty line of US$5.5 a day in 2016 com- (Figure 19, panel A). Among 12 countries with Gini pared to 2008, which reduces the income gains needed coefficients for 2008 and 2016, the DR ranks second to leave poverty. Something similar is observed for the to last in terms of inequality reduction during that pe- group of vulnerable, with a lower share being close to riod (Figure 19, panel B).67 the poverty line in 2016, and consequently, a higher share closer to the vulnerability line of US$13 a day Between 2008 and 2016, the middle class increased (Figure 21, panel B). as poverty was reduced.68 This not only made the middle class outnumber the poor for the first time in The profiles of households and individuals classified 2014, but also brought the size of the country’s middle as poor, vulnerable, or middle-class show progress class (as a share of the population) to exceed regional along different dimensions since 2008. The average levels in 2016 (Figure 20). The difference between the years of education for the heads of poor households vulnerable population and the middle class decreased increased by 0.8 years between 2008 and 2016, reduc- from 16 percent in 2008 to 4 percentage points in ing the gap with better-off households; and the share 2016, while in the LAC region it went down from 7 of poor heads of household with complete secondary to 2 percentage points. Nevertheless, at 40.8 percent, education or higher increased from 12.3 to 18 per- the share of vulnerable population in the DR in 2016 cent in that period. There are a few notable changes in is relatively high compared to other countries in the the labor market. Female labor force participation in region, and constitutes the largest income group in the poor households increased by 5 percentage points since country. 2008, closing the gap with men. The share of jobs with 67  These countries have harmonized household survey data in the LAC Equity Lab. 68  To compare income distributions and socioeconomic groups across LAC countries, we use the SEDLAC harmonized datasets, with income expressed in US dollars PPP 2011. These income groups are defined using the LAC regional lines as follows: an individual is poor if her income is lower than US$5.5 a day; vulnerable if her income is between US$5.5 and US$13 a day, middle class if her income is between US$13 and US$70 a day, and upper class if her income is higher than US$70 a day. These values are expressed in PPP 2011. 44 | Dominican Republic: Systematic Country Diagnostic Figure 21: Income distribution among poor and vulnerable populations, 2008 and 2016 Panel A: Distribution among Panel B: Distribution among the poor (density) the vulnerable (density) .25 .3 .2 .2 .15 .1 .1 .05 0 0 2 4 6 6 8 10 12 14 HH per capita income HH per capita income 2008 2016 2008 2016 Source: Authors' calculations using data from SEDLAC. a signed contract increased by 9, 10 and 15 percent- tion in both Chile and the DR moved to a higher in- age points for individuals in poor, vulnerable and mid- come group, the highest among 13 LAC countries with dle-class households, as did the share of salaried jobs in comparable data for the period (Figure 23). For each all groups, hinting at increased formalization of jobs. point of average annual GDP growth, 6.4 percent of the Finally, the share of middle-class individuals with a job population in the country moved to a higher income in the public sector increased from 12 to 18 percent in group compared to 9.8 percent in Chile, and the fifth the period (Table 2). This increase in public sector em- lowest among the same group of LAC countries. ployment tracks the increase in size of the middle-class in the country. Despite the rapid reduction in poverty, there are large and persistent inequalities between rural and urban Poverty rates for women are higher than for men areas and between the different regions of the coun- across several dimensions. For the entire population, try. Monetary poverty in rural areas was 25 percentage poverty rates for women (30.4 percent) exceed those points higher than in urban areas in 2000. This differ- for men (27.4 percent). Once the population is split by ence fell to 15 points during the banking crisis when sex and dimensions like age, area of residence (urban/ urban poverty increased more than rural poverty, and rural), occupation, education, and marital status, wom- has hovered between 10 and 15 points since then. The en have higher poverty rates than men in almost every greatest concentration of poverty in the border areas subgroup. The only exceptions are self-employed and with Haiti has not changed in the last 15 years, nor has unpaid workers (see Annex 6). Box 4 below describes the fact that the National District and North Central some of the differences. regions have the lowest poverty rates (Figure 24). The high concentration of public investment does not con- Since 2008, the share of the population moving to a tribute to reducing these inequalities, with 75 percent of higher income group in the DR is one of the highest the investment in the period 2012-2015 concentrated in in LAC; but when compared to its high economic nine provinces that represent 54 percent of the country’s growth rates, it is lower than expected.69 Between population. Investment is also concentrated in provinces 2008 and 2016, more than 31 percent of the popula- with relatively low poverty rates and in only a few sectors. 69  Upward mobility is defined as the sum of the shares of the population that moved to a higher income group in the period, from poor to Dominican Republic: Systematic Country Diagnostic | 45 TABLE 2. Profile by income groups, 2008 and 2016 2008 2016 Poor Vulnerable Middle Poor Vulnerable Middle class class Household profile Share of households 28.7 39.0 30.6 16.0 38.1 43.4 Female head 32.1 37.1 28.8 19.2 39.2 39.3 Male head 27.1 39.9 31.4 14.1 37.4 45.9 Average household size 4.4 3.7 2.9 4.3 3.5 2.8 Demographic dependency 1.1 0.6 0.4 1.1 0.6 0.4 Economic dependency 2.2 1.2 0.7 2.0 1.0 0.5 Household head characteristics Years of education head 5.5 6.8 9.6 6.3 6.9 9.3 Share with complete secondary 12.3 21.6 42.1 18.0 23.5 41.2 education or higher Head cohabitation 17.8 26.6 44.5 15.4 23.0 39.2 Shares of household total income Labor income 62.5 70.9 72.0 57.7 66.4 74.0 Non-labor income 19.3 13.2 10.2 21.6 15.1 10.5 Government transfers 2.9 0.8 0.2 5.7 2.9 0.9 Remittances 2.8 4.6 6.4 3.0 4.4 4.7 Individual profile Share of individuals 35.2 39.9 24.1 21.0 40.8 36.8 Ages 0-14 41.0 28.4 18.0 41.4 26.3 15.2 Labor Force Labor force participation 43.9 56.6 66.9 44.7 40.8 36.8 Female labor force participation 39.5 48.2 59.4 44.6 48.5 61.6 Employment Labor Contract 40.5 47.8 58.6 49.9 61.0 73.2 Informal employment 68.5 56.2 42.5 66.9 56.3 41.7 Part time work (<30 hours) 23.7 18.1 14.6 22.4 16.8 10.9 Salaried worker 44.1 52.4 55.5 49.0 54.3 60.3 Large firm 20.3 29.5 37.6 20.2 29.7 33.1 Small firm 69.9 59.6 50.3 68.6 59.6 48.6 Public sector employment 9.8 10.9 12.1 11.2 10.8 18.3 Source: Authors’ calculations using data from SEDLAC. 46 | Dominican Republic: Systematic Country Diagnostic BOX 4. Gender and poverty in the DR Incidence of poverty status across the life cycle in the DR has similar trends for both women and men. For the entire population and for all age groups (0-14, 15-24, 25-40, 41-55, and 56+ in Annex 6), female poverty rates are higher than male’s. Poverty rates decrease with age but at different speeds for men and women. Starting at a sim- ilar level during childhood (0-14 years) poverty decreases sharply, especially for men, and by age 15-24 the share of poor men is six percentage points lower than the share of poor women. The gap increases slightly for the age group 25-40, and shrinks to under three percentage points for older groups (see Figure B4.22, panel A). As expected, female and male poverty rates decline as education increases, but differences in poverty rates between women and men increase with education (Figure 2X). Poverty rates are higher for women than for men at all education levels, and the share of women among the poor also increases with education. Women account for the largest share among individuals with complete secondary education or higher, but are also overrepresented among the poor individuals with those levels of education. For instance, women account for 52.5 percent of the in- dividuals with complete secondary education, and for 61.3 percent among the poor with that level of education (see Figure B4.22, panel B). In terms of escaping poverty, the returns to education are lower for women than for men. Marital status is correlated with poverty rates for women in the DR. Around 38 percent of households in DR has a woman as head of household, 84 percent of whom does not cohabit with a partner (44 percent are divorced, 22 percent are widowed and 18 are married with no husband present). Divorce/separation is the main route to house- hold headship for women, with 44 percent of female heads in that status. Furthermore, women are overrepresented among the divorced household heads (with 71 percent) and among the divorced household heads that are poor (with 90 percent). While the poverty rate for women heads in other marital statuses is around 20-25 percent, it is 31.5 percent for divorced female heads (see Annex 6). More research is needed on the gender aspects of poverty in the country. Figure B4.22: Poverty rates by age group and level of education, by sex in 2016, and ratio of female and male poverty rates Panel A: Poverty rates and ratio by age group Panel B: Poverty rates and ratio by level of education 50 2,0 50 2,0 Poverty rate ratio (Female/Male) 40 1,6 40 1,6 Poverty rate ratio (Female/Male) Poverty rate (%) Poverty rate (%) 30 1,2 30 1,2 20 0,8 20 0,8 10 0,4 10 0,4 0 0 0 0 0-15 15-24 25-40 41-55 56 and Never Complete Complete older attended primary secondary M l F m l R tio F m l /M l M l F m l R tio F m l /M l Source: Authors' calculations using ENFT. Dominican Republic: Systematic Country Diagnostic | 47 FIGURE 23: Upward mobility and average annual GDP growth, 2008-2016 8 PAN 7 nnu l GDP rowth 6 PER PRY 5 URY DOM COL ECU 4 CHL HND 3 MEX ARG BRA 2 SLV Av r 1 0 10 15 20 25 30 35 Sh r of popul tion th t mov d to hi h r incom roup Source: Authors' calculations using LAC Equity Lab. Despite progress in coverage and equality across Both average coverage rates and the similarity of population groups, large differences between ur- coverage rates across groups have increased in ur- ban and rural areas are also observed in terms of ban and rural areas, with room to improve cover- opportunities for children. To measure equality of age in rural areas. Coverage of sanitation in rural opportunities we use a Human Opportunity Index areas is only 50 percent, 58 percent in water and 67 (HOI) that combines: i) the level of coverage of ba- percent in assets ownership (Table 3). The dissimi- sic opportunities necessary for human development, larity of coverage rates for groups defined by circum- including primary education, water, sanitation, asset stance characteristics, or D-index, is more than twice ownership, and good housing;70 and ii) the degree to as high in rural areas for asset ownership and more which the distribution of those opportunities depends than three times as high for sanitation. The value of on circumstances not under the children’s control such this index can be understood as the share of oppor- as gender, household income, or household charac- tunities of children achieving each dimension that teristics (knows as D-index or dissimilarity index).71 would have to be reallocated to eliminate the differ- A higher value of the HOI denotes a higher equal- ence across circumstance groups. For instance, a value ity of opportunities. The HOI increases with cover- for the D-index of 16.4 in sanitation in rural areas in age rates and decreases with the dissimilarity index. 2016 means that 16.4 percent of the opportunities of The large differences in the equality of opportunities children having access to sanitation need to be real- between urban and rural areas are the result of differ- located to eliminate the differences in coverage rates ences in coverage rates -higher in urban areas for all across groups defined by gender, parental education, dimensions- and the dissimilarity in access rates for household income, etc. groups defined by circumstance characteristics (e.g., gender, parental education, household composition) Despite significant improvements in service cover- lower for all dimensions in urban areas. Consequently, age and similarity in coverage rates, a few charac- the value of the HOI is consistently higher in urban teristics (education of parents, living in urban areas, areas (see Figure 25). and household income) explain a large share of the inequality of opportunities for children at birth. vulnerable, from vulnerable to middle class, or from middle class to upper class, using the LAC regional lines. In the absence of panel datasets for most countries in the period, we use synthetic panels to estimate mobility at the household level using the non-parametric method proposed by Lucchetti (2017). 70  Asset ownership is defined as lacking no more than one of the following assets: refrigerator, telephone, and electricity. Good housing (solid floor in the figure) is defined as having floor material in the dwelling different from dirt. School attendance is measured for children ages 10-14. 71  Paes de Barros et al. (2009). 48 | Dominican Republic: Systematic Country Diagnostic FIGURE 24: Moderate poverty rate by region, 2016 Source: Authors' calculations using ENFT. For instance, these three characteristics account for 93 Growth has been the main driver behind the reduc- percent of the remaining inequality of opportunities in tions in moderate and extreme poverty, especially in asset ownership, for more than 85 percent in sanitation rural areas. Between 2008 and 2016, income growth and water, and 70 percent in completion of 6th grade accounts for 86 percent of the reduction in moderate on time, with the education of parents alone accounting poverty (76 percent in urban areas and 100 percent for more than 40 percent of the inequalities in com- in rural areas) and for 75 percent or the reduction in pletion of 6th grade (Figure 26). The prominent role extreme poverty (72 percent in urban and 88 percent of these circumstances in explaining the inequality of in rural areas). Consequently, redistribution (changes opportunities for children at birth hamper intergener- in income distribution) plays a more important role in ational mobility. For instance, the education of moth- urban areas. ers is particularly important for the nutritional status of young children: under-five children of mothers with The bulk of the reduction in moderate poverty after less than 5 years of education have more than double 2008 was achieved after 2013, with labor income as the stunting rates (10 and 9 percent respectively) of the most important channel connecting growth and under-five children of mothers who attended universi- poverty reduction. Moderate poverty was reduced by ty (4.3 percent).72 According to Galasso and Wagstaff 3.5 percentage points (pp) in 2008-2013, and by 10.5 (2016), the per capita income penalty a country incurs pp in 2013-2016. According to decomposition anal- for not having eliminated stunting when today’s work- ysis, labor income explained only 10 percent of the ers were children, ranges from 7 to 10 percent of GDP poverty reduction in 2008-2013 compared to 62 per- per capita. cent of the poverty reduction in 2013-2016, consis- 72  Stunting rates between under-five children in rural (16.9 percent) and urban (16.3 percent) areas were not significantly different. Dominican Republic: Systematic Country Diagnostic | 49 Figure 25: Human Opportunity Index Panel A: Rural areas Panel B: Urban areas 110 110 100 90 90 70 80 50 70 60 30 50 10 40 2000 2002 2004 2006 2008 2010 2012 2014 2016 2000 2002 2004 2006 2008 2010 2012 2014 2016 HOI 6th Gr d on Tim School tt nd nc HOI 6th Gr d on Tim School tt nd nc W t r S nit tion Solid floor Ass t Own rship W t r S nit tion Solid floor Ass t Own rship Source: Authors' calculations using ENFT. TABLE 3. Coverage rates and D-index in urban and rural areas 6th grade Asset own- School Solid floor Sanitation Water on time ership attendance Urban coverage rates 2000 59.1 83.9 98.2 96.9 69.8 84.9 2016 83.4 86.8 96.8 99.4 87.6 85.0 Urban D-index 2000 13.4 5.9 0.7 2.1 12.6 3.3 2016 5.5 4.5 1.1 0.4 4.9 3.9 Rural coverage rates 2000 40.6 51.3 96.6 83.9 19.7 47.2 2016 77.3 67.1 96.0 94.6 50.4 57.8 Rural D-index 2000 20.0 12.9 0.9 6.5 30.4 11.4 2016 7.1 9.4 1.4 2.4 16.4 6.4 Source: Authors’ calculations using ENFT. tent with a higher growth of job creation and increases the observed increases in both participation and em- in real wages in the latter subperiod. The contribution ployment rates for women. of women’s occupation rates and labor earnings was similar across the two subperiods at around 30 per- The reduction in extreme poverty was similar in cent, while men’s labor contributed to an increase in 2008-2013 and 2013-2016, with labor income al- poverty of 0.7 pp in 2008-2013, and to a decrease in most tripling its contribution in 2013-2016 com- poverty of 4.1 pp in 2013-2016 (Figure 27, panel A). pared to 2008-2013. Labor income contributed 20 The contribution of women’s occupation rate is higher percent of the reduction in extreme poverty in 2008- than the contribution of men’s and is consistent with 2013 compared to 61 percent in 2013-2016, with labor 50 | Dominican Republic: Systematic Country Diagnostic Figure 26: Contribution to the inequality of opportunities Contribution to th in qu lit 100 80 of opportuniti s 60 40 20 0 Asset ownership Sanitation Water Solid floor School attendance 6th grade on time P r nts' duc tion Urb n / rur l F mil incom Numb r of siblin s G nd r of child G nd r of hous hold h d Pr s nc of both p r nts Source: Authors' calculations using ENFT. Figure 27: Contributions to moderate and extreme poverty reduction, 2008-2013 and 2013-2016 Panel A: Moderate poverty in pov rt (%) 10 -10 Contribution to tot l ch n -30 -50 -70 Share of Share of Women labor Share of Share of Men labor Non-labor women occupied women income men occupied men income income 2008-2013 2013-2016 Panel B: Extreme poverty in pov rt (%) 10 -10 Contribution to tot l ch n -30 -50 -70 Share of Share of Women Labor Share of Share of Men Labor Non-labor women occupied women income men occupied men income income 2008-2013 2013-2016 Source: Authors' calculations using ENFT. Dominican Republic: Systematic Country Diagnostic | 51 earnings for both women and men accounting for 50 ployment affects women and youths disproportionately. of the 61 percentage points (Figure 27, panel B). This The unemployment rate for women, at 20.6 percent, is finding is consistent with the increase in real wages doc- more than twice that for men at 8.6 percent, with high- umented later in this chapter. er rates for youth (ages 15-24) of both sexes, at 41 and 21 percent for females and males, respectively. There The contribution of non-labor income to the reduc- are big differences in unemployment rates per educa- tion in both moderate and extreme poverty decreased tion levels and sex. The unemployment rate for wom- between 2008-2013 and 2013-2016. The way infor- en with no education is 7.7 times higher in 2016 than mation on non-labor income is collected in the ENFT the rate for comparable men (23.7 vs 3.1 percent), 3.3 does not allow us to distinguish between government times higher in the case of women with primary educa- and private transfers (which we refer to as domestic tion, and 2.2 times higher for women with secondary transfers), but only between transfers, and income from education. Unemployment is highest for people with abroad (remittances, pensions, interest payments) and secondary education or incomplete tertiary (between rent income. The contribution of income from abroad 17 and 19 percent) -despite similar participation rates- and rent income to the reduction in moderate poverty compared to people with lower levels of education (9 remained around 14 percent across subperiods, while percent for people with no education and around 12 the contribution of domestic transfers was decreased percent for people with some primary education). This from 52 percent in 2008-2013 to 9 percent in 2013- hints at segmentation in the labor market with higher 2016. In the case of extreme poverty the contribution demand for jobs that require either low or high levels of income from abroad and rents more than doubled of education (and skills), and can be related to the high from 11 percent in 2008-2013 to 23 percent in 2013- dropout rates in the first years of secondary education. 2016, while the contribution of domestic transfers was reduced from 55 to 15 percent across subperiods. Half of the jobs in the Dominican economy in 2016 were in the informal sector, with the informality rate rising after 2000 and decreasing in the last 5 years.74 Labor markets as links between This decrease was driven by the growth of formal jobs in economic growth and households’ commerce and public sector employment, especially in welfare education and health. The relatively high level of infor- mality pushes workers outside the social security system, Labor markets in the country are characterized by strips the government of potential tax resources, and persistent unemployment rates, high levels of infor- limits the participation of workers in organized unions.75 mality, and low wages. Unemployment rates have had The historical weakness of labor unions in the Domin- a muted response to the high rates of economic growth, ican Republic may partly explain the observed decrease increasing from 4.8 percent in 2008 to 5.5 percent in in the real return to labor, despite the rising productivi- 2016, thus signaling the limited ability of the econo- ty, as they do not enjoy a strong bargaining position vis- my to link the labor force to jobs, and that of workers à-vis entrepreneurs when negotiating minimum wages. to convert their skills into income.73 In those 8 years the working-age population has increased by 148,000 Real wages are persistently low in the DR while labor people per year, with 109,000 being added to the la- productivity has grown strongly. Non-wage labor bor force every year. The economy created an average of costs in the country are below regional levels. Wages 99,000 new jobs per year, while adding around 10,000 lost over 30 percent of their purchasing power in the people to the number of unemployed every year. Unem- wake of the 2003 banking crisis and despite rebound- 73  World Bank (2017b). 74  The official definition of informality in the DR using the ENFT includes: salaried workers in firms with 5 or fewer employees, own account workers and employers in the following occupational groups: agricultural workers and farmers, clerical support workers, drivers, craftsmen, sales workers, and unskilled workers. Domestic and unpaid workers are also included. 75  Carneiro and Sirtaine (2017). 52 | Dominican Republic: Systematic Country Diagnostic Figure 28: Productivity and wages Panel A: Product and wages. Indices in real Panel B: Hourly wages in US$ for salaried terms per hour worked (2008=100) employees, LAC countries 2015 2,0 6 5 Sh r of th popul tion 1,5 4 3 1,0 2 1 0,5 0 2008 2009 2010 2011 2012 2013 2014 2015 2016 URY PRY CHL BOL COL SLV BRA PER MEX CRI PAN ARG ECU DOM GTM HND NIC Product p r hour work d M di n hourl w s Source: Authors' calculations based on national accounts and ENFT. Source: SEDLAC and World Development Indicators. ing in its aftermath, have not reached pre-crisis levels in the top 60 (62.1 vs. 75.5 percent in 2016). There are terms of real purchasing power. Productivity per hour also differences by education level, with participation worked in real terms grew by 74 percent between 2008 increasing with education. In 2016, 63.5 percent of the and 2016, while real hourly wages increased by 21 per- working-age population with no education participated cent (Figure 28, panel A). Average hourly wages in the in the labor force, compared to 67.2 percent of peo- country are among the lowest in the LAC region (Figure ple with complete primary, 76.7 percent for those with 28, panel B). Non-wage labor costs in the DR, at 45 per- complete secondary, and 87.7 percent of those with cent, are slightly below LAC regional levels, with 20 of complete tertiary. the 45 points coming from mandatory contributions.76 When measuring the actual cost for a firm should it The largest differences in labor force participation want to formalize an informal salaried worker, the cost are observed between men and women. Although in the DR increases to 59 percent, still below the LAC women account for 65 percent of university enrollment regional level of 76 percent. If the components of the and 80 percent of those who graduate with honors, only average non-wage cost of salaried labor are combined 59 percent of working-age women participate in the la- with the nominal restriction that wages cannot be lower bor market while 82 percent of men do. Participation of than the minimum wage, and this is expressed as a share men has remained practically constant since 2006 while of GDP per worker, the country has the second lowest female participation increased by 4 percentage points.78 cost in the region at 18 percent. There are big differences between the participation rates of men and women by education level. For people Labor force participation is around the regional av- with no education, the participation rate for men is 2.3 erage, with large differences across income groups, times that of women, around 1.6 times for people with education levels, and sex.77 Despite small increases incomplete or complete primary, 1.4 times for people in the participation of both groups since 2006, par- with complete or incomplete secondary and 1.1 times ticipation of the bottom 40 percent is lower than for for people with tertiary. 76  The estimates of non-wage labor costs come from Alaimo et al. (2017), and include three components; (i) the additional mandated ben- efits (such as bonuses and paid leave) that a salaried worker is entitled to receive according to the legislation, (ii) the mandatory contributions to finance social insurance systems and other benefits (such as training, housing, among others), and (iii) a monetized version of job security provisions. 77  Labor force participation is computed using population between ages 15-65. 78  The increase in female participation rates was accompanied by an increase in unemployment rates from 16.6 percent in 2008 to 20.3 percent in 2016. Female occupation rates only increased from 22.9 to 24.3 during the period. Dominican Republic: Systematic Country Diagnostic | 53 Figure 29: Share of total employment by sector and sex, 2016 Construction 96 4 Agriculture, Livestock, Hunting, Forestry and Fishing 95 5 Transportation, Storage and Communication 89 11 Real Estate Activities, and bussinnes activities 70 30 Manufacturing industries 79 31 Public Administration and Defense 65 35 Commerce 64 36 Other sectors 61 39 Hotels and restaurants 47 53 Other Community Service Activities and extraterritorial bodies 37 63 Education 32 68 Social and Health Services 25 75 Domestic Service 8 92 0% 20% 40% 60% 80% 100% M l F m l Source: Authors' calculations using ENFT. In 2016, women accounted for 83 percent of the in- Sectoral gender segregation is a strong feature of the active population, and 3 in 10 women between the Dominican labor market. Sectors like construction, ages of 25-60 were inactive. The latter share decreases agriculture, and transportation are heavily dominated sharply with the level of education, with 54 percent by men, while domestic service, as well as social and of women ages 25-60 with no education in 2016 be- health services are dominated by women (Figure 29). ing inactive, compared to 12 percent for women with Given that specific innate talents for most sectors (if university education. The main reasons for inactivity any) are unlikely to differ across men and women, this in women are household responsibilities (76 percent) sectoral segregation suggests that many workers are not –a responsibility which is almost exclusively on wom- pursuing their comparative advantage. en–. The importance of household responsibilities to explain inactivity might hint at challenges which Average hourly wages for women are around 15 per- women face to balance work and family in the DR. cent lower than for men between 2008 and 2016, More generally, the lower participation and high in- and the difference is twice as high for women in the activity rates for women might suggest the existence bottom 40 percent of the income distribution. These of barriers of access to the labor market or the lack of differences are not explained by observable characteris- opportunities and incentives. On the other hand, 10 tics.79 We use a matching technique developed by Ñopo percent of the total inactive population over 15 years (2008) to compare hourly wages of women and men of age reports temporary or permanent disabilities as who share the same observable characteristics including their main reasons of inactivity. When focusing on age, education, area of residence, presence of children in men between the ages 25-60, this share rises to 47 per- the household, industry of employment, occupational cent. Seventeen percent reported to be inactive due to category, and job tenure. As an example, the earnings temporary disabilities which could be associated with of a female worker, 30 years old, with complete second- transitory health problems that directly affect not only ary education, living in an urban area, no children in the participation rate but potentially productivity in her household, and who has worked as an employee in the labor market. The other 30 percent were inactive manufacturing for 6 months, will be compared to the due to a permanent disability, indicating challenges of earnings of a male worker with the exact same charac- productive inclusion for this group. teristics (30 years old, complete secondary education, living in an urban area, no children in his household, 79  The gap was computed using workers ages 20-55, and working more than 30 hours a week. 54 | Dominican Republic: Systematic Country Diagnostic who has worked as an employee in manufacturing for relationship have experienced some form of violence 6 months). We then estimate to which extent the gap perpetrated by their partners in the last 5 years. The is explained by the observable characteristics, and find country has the third highest rate of female homicides that it is negligible, meaning that the explanation of the in LAC, and around a quarter of all women ages 15-49 gender wage gap in the country is beyond the set of have experienced physical violence.82 observables that typically influence differences in pay across sexes. Women are underrepresented among entrepreneurs. Only 14 percent of small and medium-sized enterprises Youth also face high unemployment rates. While are fully owned by women.83 When considering firms of unemployment rate for adults ages 25-54 decreased all sizes with at least some female ownership, this mea- from 13.5 to 10.6 percent between 2001 and 2016, the sure increases to 30 percent, below the 40 percent av- rate for youth ages 15-28 remained around 30 percent erage across LAC. The percentage of firms in the coun- during the period, the second highest in LAC and only try with a female top manager is half the LAC average comparable to Jamaica’s (32.8 percent in 2015). Young (Enterprise Surveys). Data on microenterprises reflect women’s unemployment rate is twice as high as men’s the tendency of women-owned firms to be engaged in (43.3 vs. 21.6 percent) despite a much lower partici- lower-productivity retail and other service activities.84 pation rate (41 percent for women vs. 61 percent for Only 5.4 percent of firms managed by women in the men). manufacturing and services sectors in the Dominican Republic engage in export sales, significantly below Some of the inferior labor market and educational the 14.7 percent average for firms managed by women outcomes for women might be partly explained by across LAC and less than the 13.4 percent of male-man- one of the highest rates of adolescent fertility in the aged Dominican firms that export. Analyses at the re- world and by high rates of violence against women. gional level have found access to finance to be one key These phenomena usually constraint women’s ability constraint to female entrepreneurship – 70 percent of to make their own choices and reach their goals. For women-owned SMEs in LAC are unserved or under- instance, different studies find a significant correlation served financially.85 Other constraints identified at the between teenage pregnancy and childbearing, lower ed- regional level include lack of access to information, e.g., ucational achievement, and poorer labor market out- about market or investment opportunities, and less comes for women.80 At 97.3 births per 1,000 women access to business advisory services and other capacity ages 15-19, adolescent fertility rate in the DR is the building activities such as coaching and mentoring.86 highest in LAC and the third highest among upper middle-income countries, surpassed only by Equatorial As in other immigrant-receiving countries, challeng- Guinea and Gabon. Many teenagers are forced to drop es in the Dominican labor market have resulted in their studies to attend the pregnancy, with consequent concerns that immigration may be leading to de- lower levels of education reducing labor income in creased employment opportunities for local work- adulthood by an estimated 17 percent.81 Gender-based ers—and thus contributing to stagnant wages and violence is one of the main mortality causes in the low poverty reduction.87 Given the country’s unique country for women ages 15-49, and around 30 percent combination of high inflow and outflow of migrants of women in this age group that have ever been in a (for its population size - see Figure 30, panel B), the 80  World Bank (2012). 81  Strictly speaking this effect is attributable to child marriage, which is probably the cause of 3 out of 4 births from mothers younger than 18 in the DR. See Wodon et al. (2017). 82  Observatorio de igualdad de género CEPAL, and ENDESA 2013. 83  ONE (2015). 84 Ibid. 85  World Bank (2016). LAC Regional Gender Action Plan FY2016-19. 86  infoDev (2015). 87  World Bank (2017b). Dominican Republic: Systematic Country Diagnostic | 55 next section analyzes the potential effects of both immi- both countries (for instance, the DR’s GNI is 8 times gration and emigration on the labor markets, and more higher than Haiti’s), the fact that they share a porous broadly, of migration on inclusion. land border (each country’s only land border) and his- torical ties. Dominican emigration is much more con- centrated in destinations relative to other Caribbean Migration and inclusion countries (the Dominican diaspora is the seventh most concentrated diaspora in LAC and the second most For a country of its size, the DR has a unique com- concentrated in the Caribbean). bination of high emigration and immigration rates. Close to 1.1 million Dominicans lived in the US Remittances have grown over time with the stock of in 2016, and more than 400,000 in other countries, migrants, and represent 7.7 percent of GDP in 2016. i.e., 15 percent of the total population residing in the Around 18 percent of households receive remittances country. Large-scale migration to the US began in the from abroad, with a very similar incidence of remit- 1960s after the economic and political turbulence that tances across deciles (constructed using income before followed the assassination of dictator Rafael Trujillo remittances), except for households in the lowest decile, (Figure 30, panel A). The country also has a relatively a third of which are recipients (Figure 31). Remittances large immigration rate. According to the latest survey of can affect household level decisions such as labor force immigrants, more than 570,000 foreigners, equivalent participation and education choices. to 5.6 percent of the country’s population, lived in the country in 2017.88 As with emigration, immigration Moderate poverty would increase by around 2-3 into the country comes mainly from one source, Haiti, percentage points and extreme poverty by 1-2 pp in with 87 percent of the 570,000 immigrants. The large the absence of remittances to households. These esti- flow of migrants from Haiti is unsurprising given the mates for the 2008-2016 period account for the poten- large gaps in terms of levels of development between tial income that the migrant may have made at home, Figure 30: Dominican immigrants in the US and migration rates in LAC Panel A: Number of Dominican immigrants Panel B: Migration and population in the US, 1960-2016 in LAC, 2013 1.200.000 20 BHS 1.000.000 BLZ 800.000 10 Immi r tion R t CRI SUR DOM 600.000 ARG PRY TTO GUY JAM 400.000 0 MEX SLV 200.000 0 -10 1950 1960 1970 1980 1990 2000 2010 2020 -20 0 20 40 60 80 Emi r tion R t Source: Migration Policy Institute (MPI) tabulation of data from US Census Source: Migration and Remittances Factbook, 2016 Bureau, 2010 and 2016 American Community Surveys (ACS), and 2000 Notes: The immigration rate is calculated as the ratio of immigrants over Decennial Census. Data for 1960 to 1990 are from Campbell J. Gibson and total residents in the host country. Emigration is defined as the ratio Emily Lennon, "Historical Census Statistics on the Foreign-Born Population between emigrants and the number of residents in the home country. The of the United States: 1850 to 1990" (Working Paper No. 29, US Census size of the bubble is proportional to the country’s population Bureau, Washington, DC, 1999). 88  ONE (2018). 56 | Dominican Republic: Systematic Country Diagnostic Figure 31: Incidence of remittances by decile, 2016 40 35 32,5 Sh r of hous holds 30 25 18,6 19,1 18,3 18,6 20 16,1 15,8 16,4 16,5 17,0 15 10 5 0 1 2 3 4 5 6 7 8 9 10 D cil Source: Authors' calculations using ENFT. and take into account that households with migrants ticularly strong for adults with less schooling. Among may not be randomly selected from the population.89 adults who did not complete primary school, partici- Among household receiving remittances, around 50 pation is 7.3 percentage points lower if remittances are percent have female heads, compared to 35 percent received, 2.5 percentage points for adults who finished with female heads among households that do not re- secondary school or have some tertiary schooling, and 2 ceive them. Something similar is observed on remit- percentage points for adults with complete tertiary edu- tances and the inactivity of the household head in the cation. While receipt of other transfers (including pub- labor market.90 Among recipient households, 37 per- lic transfers) is also associated with lower participation, cent of households have an inactive head, while among this effect is smaller than that associated with remittanc- non-recipient households, the share with inactive heads es. This difference may be due to how remittances and is only 20 percent. public transfers are allocated across households: while public transfers are targeted at poor households, typi- Access to alternative sources of income, such as re- cally with children or retirees, remittances are often the mittances, has the effect of decreasing labor force result of joint household decisions wherein a household participation by increasing reservation wages.91 On member migrates specifically to increase household in- average, adults in households that received remittances come, and at times, to decrease the need for other mem- in 2012 were 6 percentage points less likely to be in bers to work.92 the labor force than similar adults without remittances, and the relationship between remittances and participa- High unemployment and emigration rates of work- tion is about the same in households in the bottom 40 ers with more schooling imply insufficient demand percent and those in the top 60 percent. However, it is for skilled labor, and firm surveys suggest this is due significantly greater for women than for men (7.5 per- to skills mismatch. As of 2016, 66 percent of Domin- centage points for women and 1.3 for men), and par- ican immigrants in the US, 25 or older, had complete 89  These estimations use the method proposed by Acosta, Fajnzylber, and Lopez (2007) as well as data from the ENFT to construct a coun- terfactual situation without migration and correcting for the potential income which migrants might have had, had they not left their country. 90  A person is considered inactive in the labor market if she does not belong to the economically active population, which includes people who are working, and those unemployed who are looking for jobs. 91  These results come from World Bank (2017b) and do not address the potential endogeneity of receipt of remittances and labor force partici- pation. Migration may have been planned around the labor supply decisions of non-migrating household members. For instance, individuals may migrate to compensate for other household members not being able to work – or may have a direct effect on other members’ decision to work. 92  World Bank (2017b). Dominican Republic: Systematic Country Diagnostic | 57 secondary education or higher, compared to only 40 The vast majority of Haitian immigrants in the Do- percent of Dominicans living in the DR (Figure 32).93 minican Republic are escaping poverty and looking Along with high unemployment rates for skilled labor, for better work opportunities.97 The first large-scale the relatively high emigration of skilled labor further migratory waves of Haitians started back in the 1920s, supports the possibility that the local labor market has to work in government-owned sugar cane fields.98 insufficient demand for it. In meantime, 31 percent Around 70 percent of Haitian migrants have at most of firms in 2016 identified an inadequately educated primary education with an average of 4.1 years of workforce as a major constraint for business in Domin- schooling (compared to 9 years for Dominican-born), ican Republic,94 suggesting the presence of skills mis- and are highly clustered in unskilled work, especially in match, with educated workers not possessing the skills agriculture, construction, and commerce.99 that firms require.95 In addition, the economy does not generate the number of full-time jobs that work- There are concerns that Haitian immigration may be ers would be willing to take. One in six workers in the contributing to the stagnant wages in the country, country was willing to work additional hours in 2016, but the evidence is mixed. The extent to which immi- the highest rate among 16 countries in LAC.96 The rate gration affects wages in local labor markets is in large is higher for workers in the bottom 40 percent of the part determined by whether the immigrants’ skills are income distribution with more than one in five workers substitutes or complements to that of the local work- willing to work additional hours. ers. If they are substitutes, this can result in increased Figure 32: Share of population by highest level of education achieved - Dominicans ages 25+ living in the DR or in the US, 2016 35 30 25 20 15 10 5 0 No Incomplete Complete Incomplete Complete Incomplete Complete education primary primary secondary secondary tertiary tertiary In th DR In th US Source: Authors' calculations using ENFT and US Census Bureau ACS 2016. Notes: The calculation using the ACS includes only people who report being born in the DR. The number of years of education in the ACS were used to determine the equivalent highest level achieved according to the DR education system. 93  ACS 2016 data obtained from the IPUMS database version 7.0. See Ruggles et al. (2017). 94  Enterprise Surveys (http://www.enterprisesurveys.org), The World Bank. 95  World Bank (2017b). 96  Source: ILOSTAT. Persons in time-related underemployment comprise all persons in employment, who satisfy the following three criteria during the reference period: a) are willing to work additional hours; b) are available to work additional hours i.e., are ready, within a specified subsequent period, to work additional hours, given opportunities for additional work; and c) worked less than a threshold relating to working time i.e., persons whose hours actually worked in all jobs during the reference period were below a threshold, to be chosen according to national circumstances. The time-related underemployment rate is calculated by dividing the number of persons in time-related underemployment by the total number of persons in employment. 97  According to the most recent survey of migrants. See ONE (2018). 98  See Ferguson (2003). 99  ONE (2018). 58 | Dominican Republic: Systematic Country Diagnostic competition for jobs while complementary skills can showed that CCT households were positively associated lead to increased productivity for local workers. The with lower risk of catastrophic expenditures in health, evidence suggests that there is no negative relationship increased height-for-age of children 4 or older, lower between the proportion of the local labor force that is risk of adolescent pregnancy, and lower risk of repeating Haitian-born and the wages of local labor once individ- grades in secondary school.101 Responding to the 2003 ual characteristics are taken into account, and that since macroeconomic and banking crisis, monetary transfers Haitian immigrant labor is largely limited to unskilled started being provided to incentivize the formation of and informal employment in agriculture, construction, human capital through education, health and nutrition. and commerce, Haitian immigrants are more likely to To increase efficiency and transparency of the sector, two be complements than substitutes to both capital as well separate institutions were created for the targeting (SIU- as to the relatively more skilled Dominican workers.100 BEN, Sistema Único de Beneficiarios) and payment of social subsidies (ADESS, Administradora de Subsidios Sociales). In the past decade, noncontributory programs Social policy and inclusion have expanded coverage significantly, with a particular focus on reaching the poorest in the last five years. De- Social policy has potentially important roles in sustain- spite the increased pro-poor coverage using the SIUBEN ing the recent gains in poverty reduction, closing spatial targeting mechanism, there are gaps (particularly those gaps in poverty and access to services, and in increasing related to lack of documentation) that need to be ad- equality of opportunities. This section covers social in- dressed to reach the entire population. To respond to surance programs, education, and health, and how they the multidimensional nature of extreme poverty, the are related to poverty reduction and to the country’s abil- Government recently launched the ‘Progresando Uni- ity to sustain high rates of economic growth. It also iden- dos’ strategy, aimed at extreme poor households using tifies where social policies could better address key hu- geographical targeting prioritizing 14 among the poorest man development and inclusion challenges throughout provinces. This initiative would provide the poorest with the lifecycle including pregnancy and birth; early child a combination of services including technical-vocational development; childhood and adolescence; transition to training, entrepreneurship opportunities, financial in- labor market and adulthood; and retirement/aging. clusion services and housing improvements. Social insurance Social accountability mechanisms have been key tools to improve service delivery to CCT recipients Social protection programs have contributed to the in education, health care and other social services at reduction of poverty, specifically through public the local level. Starting as a pilot in 2010, the CCT transfers. The Dominican Republic has made substan- ‘Reportes Comunitarios’ (community scorecards) is a tial progress in the establishment of a social protection community monitoring scheme for both beneficiaries system in the past 10 years, including rationalizing of and providers, aimed at identifying problems in the de- spending in social welfare by reducing the dispersion of livery of services to CCT households by allowing those and strengthening the design of programs such as Pro- problems to be addressed at the community level. The gresando con Solidaridad (PROSOLI). Cash transfers baseline survey of an impact evaluation showed that in the country are generally well focused and propor- the scorecards helped solve problems in the communi- tionally do more to help the poor and at-risk popula- ty pertaining to public service delivery, by encouraging tion. According to an IDB study, without public cash beneficiaries to express their opinions and suggest solu- transfers poverty would have been 2.3 percentage points tions. The scorecards were scaled up at national level higher (or 6 percent) in 2014, and the Gini coefficient since 2014 and particularly adapted to the 14 poorest would have been 1 point higher. The evaluation also provinces in the country since 2016. In 2013, comple- 100  See Sousa, Sánchez, and Baez (2017). 101  Carrasco et al. (2014). The study assumes that households’ behavior and other sources of income would be unchanged in the absence of the cash transfers, potentially overestimating the effect of the transfers on poverty and inequality. Dominican Republic: Systematic Country Diagnostic | 59 Figure 33: Share of households receiving government transfers and the value of the transfers for recipient households, by income group 2008-2016 Panel A: Share of households receiving Panel B: Value of per-capita daily transfers for government transfers recipients (US$ per capita per day, PPP2011) 50 1,0 40 0,8 Sh r of th popul tion US$ PPP2011 30 0,6 20 0,4 10 0,2 0 0,0 2008 2010 2012 2014 2016 2008 2010 2012 2014 2016 Extr m poor Poor Vuln r bl Middl cl ss Extr m poor Poor Vuln r bl Middl cl ss Source: Authors' calculations using LAC Equity Lab. menting the scorecards, the program implemented the income); for poor recipients, it increased from US$0.17 ‘Puntos Solidarios’, one-stop direct service points where to US$0.34 a day (and from 2.5 to 5 percent of total beneficiaries can present grievances regarding adminis- income); for vulnerable recipients, from US$0.44 to trative aspects handled by either the program, ADESS US$0.64 a day; and for middle-class households, from and/or SIUBEN. The use of these strategies ensures US$0.6 to US$0.83 a day, all expressed in PPP 2011 greater inclusion and accessibility of the poor to social (Figure 33, panel B). The fact that increasing shares of protection services. both vulnerable and middle-class households receive government transfers, and that average values increase The incidence of government transfers has increased with the income of the socioeconomic group in ques- for all socioeconomic groups (defined using the LAC tion suggest the need to revise the targeting of benefi- regional lines) since 2008.102 The share of extreme ciaries to maximize poverty reduction and the transfers’ poor households receiving government transfers in- redistributive potential. creased from 30 to 41 percent between 2008 and 2016; for poor households, the share increased from 25 to Moreover, the largest fiscal expenditures are poorly 40 percent; for vulnerable households, from 15 to 38 targeted. While many government expenditure items percent; and the share for middle class households rose (e.g., education and public health) are estimated to be from 5 to 21 percent in the period (Figure 33, panel progressive in absolute terms, i.e., per capita benefits A). This increase in incidence has been accompanied decline with pre-benefit income, their incidence as a by an increase in amounts for recipients, with the av- share of GDP is relatively small. By contrast, items erage value of the transfers increasing with the income that are regressive in absolute terms, i.e., per capita of the socioeconomic group under study. The per capita benefits increase with pre-benefit income, such as average transfers to extreme poor recipients increased subsidies to electricity distribution or tax expendi- from US$0.31 to US$0.44 a day between 2008 and tures,103 account for a large share of government ex- 2016 (and from 4.1 to 7.6 percent of total household penses (Figure 34). 102  A household is considered extreme poor if its per capita income is under US$3.2 a day, poor if its per capita income is under US$5.5 a day, vulnerable if the per capita household income is between US$5.5 and US$13 a day, middle class if the per capita household income is between US$13 and US$70 a day, and high income if the per capita household income is higher than US$70 a day, all expressed in PPP 2011. 103  Cabrera et al. (2015) show that the bulk of total tax expenditures derived from ITBIS exemptions benefits non-poor households (88 percent). 60 | Dominican Republic: Systematic Country Diagnostic Figure 34: Concentration coefficients and budget shares by program, 2013 Mor pro r ssiv Mor r r ssiv -0,6 -0,4 -0,2 0,0 0,2 0,4 0,6 Incentive to school attendance Primary Education Spending School supplies School Food Program LPG subsidy for drivers (Bonogas) Pre-school Education Spending Non-contributive ambulatory health Non-contributive health Hospitals Health subsidized regime Senasa Eat First (Comer es primero) Lower Secondary Education Spending Food supplement for senior citizens Quisqueya Aprende Promese LPG subsidy for HH (Bonogas) Electricity subsidy for low-income HH Upper Secondary Education Spending Subsidy to electricity distribution Tertiary Education Spending Incentive to the Navy Tax Expenditures 0 2 4 6 8 10 12 14 16 18 20 Sh r of GDP Conc ntr tion co ffici nt Source: Authors' calculations using Cabrera et al. 2016a. A low share of workers actively contributing to the can expect to receive an old-age pension, less than 30 pension system means that only 30 percent of the la- percent of the labor force.105 bor force can expect to receive contributory pensions, one of the lowest shares in Latin America. Like in The low density of contributions and the parameters other Latin American countries, a public defined-ben- of the pension system will result in inadequate ben- efit system was replaced with a private defined-contri- efits at retirement and potentially high levels of pov- bution scheme (Social Security Law 87–01). The same erty among the elderly. The adequacy ratio, measured legislation also introduced a semi-contributory mini- as the starting pension relative to average wages in the mum pension scheme (subsidized-contributive regime) economy, is projected to reach a maximum of less than and foresaw a noncontributory social assistance scheme 20 percent of earnings when the defined-contribution (subsidized regime), which has not yet been implement- system matures in 2045 for an average wage earner with ed.104 Participation in the contributory pension system a contribution density of 75 percent. In addition, the is mandatory for all public and private sector employees, subsidized regime should provide a pension of 60 per- as well as for self-employed persons. About 2.9 million cent of the public sector minimum wage to all needy participants are registered in the defined-contribution residents ages 60 and older, but this system has not system and only about 200,000 individuals remain in come into effect.106 Consequently, high shares of the the old defined-benefit system or in special regimes for population must continue to work well beyond the of- some public-sector employees, out of 4.2 million work- ficial pension age of 60–65 years. The labor market par- ing individuals and an economically active population ticipation rates in 2016 are estimated to be 42 percent of 4.9 million. Less than half of the participants (1.4 in the 65–69 age group, 30 percent in the 70–74 age million) are actively contributing to the system -and group, and almost 20 percent in the 75–79 age group. 104  Figliuoli (2018). Only three other countries in LAC do not have non-contributory pensions as part of their pension systems: Nicaragua, Haiti, and Honduras (CEPAL, 2018). 105  Figliuoli (2018). 106  CEPAL (2018). Dominican Republic: Systematic Country Diagnostic | 61 The lack of identification documents is an important ucation. The net enrollment rate in initial education, issue to close the inequality gap in Dominican Re- which support children ages 5-6 was 47.2 percent in public. Undocumented individuals107 are not only pre- 2015/16 (up from 35.5 percent in 2007/08). This vented from participating in social protection programs, compares sharply with rates of over 95 percent in basic but also cannot open a bank account, hold land titles, education.108 In 2014, children ages 3-5 in households marry, work in the formal sector, vote, obtain a driver’s from the top income quintile almost doubled the en- license or passport, or attend school beyond 7th grade. rolment rate of children in the lowest quintile (25.4 vs. These individuals are not included in the group of legal 48.2 percent, respectively). International evidence has and illegal Haitians residing in the country, but rather shown that quality preschool and early child develop- are poor Dominicans who never obtained their birth ment programs can have a significant impact on future certificates or let their old “cédulas” expire. Government school performance and earnings, with the highest im- policies have been designed and implemented in the last pact on children from low-income families. There is a 10 years to document poor individuals. Between 2011 gap of information in the country on the measurement and 2015, the Social Cabinet and the Central Elector- of the quality of these pre-school and initial level edu- al Committee (Junta Central Electoral - JCE) provided cation services.109 identification documents to more than 255,000 poor individuals registered in SIUBEN, and the JCE provid- High dropout rates and low levels of learning and ed documents to an additional 1,165,725 individuals skills acquisition which affect not only access to em- not registered in SIUBEN. By 2015, out of the total ployment but also wages, are the biggest challenges number of individuals registered in the SIUBEN data- facing the education system. For the eight-year basic base, 3.3 percent were undocumented (285,646 indi- education program, out of every 100 students who be- viduals). The SIUBEN database is currently under up- gin school in first grade, only 75 complete the fourth date and by the end of 2018 will produce more recent grade, 63 complete the sixth grade, and 52 complete figures of undocumented individuals. the program on time. Despite substantial progress in learning by Dominican students between 2006 and Education 2013 as captured by regional assessments, Domini- can students still perform poorly and lag behind other The country has also made significant gains in access LAC countries, reflecting the weakness of the system to education at all levels in the last 15 years, includ- in helping students develop basic cognitive skills (e.g., ing closing the gap in educational achievement be- reading and math), particularly in early grades. In the tween the bottom 40 percent and the top 60 percent. latest regional assessment conducted in 2013, 74 and The average number of years of schooling for individ- 85 percent of third graders performed at the lowest level uals 25 years and older in the bottom 40 percent of in reading and math, respectively.110 These results are the population increased by 43 percent, from 4.9 years substantially worse than those of comparable countries, in 2001 to 7 years in 2016, while for the top 60 per- and represent the highest proportion of such results cent it increased by 24 percent, from 7.8 to 9.7 years in among 15 countries in the region. Poor performance in the period. For both groups the increase was higher for these assessments is observed at all socioeconomic lev- women, whose average years of schooling was identical els, although students with higher socioeconomic levels to the men’s in 2001, and was 0.6 years higher in 2016. obtained better results than those with lower socioeco- nomic levels. In addition, students with the lowest so- Despite progress, equitable access to early childhood cioeconomic levels in the region performed better, on education services is still low compared to basic ed- average, than the Dominican students with the highest 107  Undocumented means (i) an individual, 16 years or older, who does not have a national identity card or birth certificate; or a (ii) a child younger than 16 years of age who lacks a birth certificate. 108  Source: MINERD, Plan Estratégico 2017-2020, p. 39. 109  EDUCA (2015), Informe de Progreso Educativo. 110  Produced by the Latin American Laboratory for Assessment of the Quality of Education, it includes 15 countries: Argentina, Brazil, Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, Guatemala, Honduras, Mexico, Nicaragua, Panama, Paraguay, Peru, and Uruguay. 62 | Dominican Republic: Systematic Country Diagnostic Figure 35: Share of overage or failing school students (dropout + not promoted), 2014 Panel A: Share of overage students by grade Panel B: School dropouts and not-promoted 50 20 40 15 Sh r of nroll d 30 10 20 5 10 0 0 1 2 3 4 5 6 7 8 1 2 3 4 1 2 3 4 5 6 7 8 1 2 3 4 B sico M di B sico M di On r ov r d 2+ rs ov r d R p tition Dropout Source: Authors' calculations with information from the Sistema de Gestión de Centros Educativos, MINERD, DR. socioeconomic levels.111 In 2015, the country partici- of being overage in later grades, exacerbating an existent pated for the first time in the PISA worldwide assess- problem, and more importantly, of dropping out (see ment which evaluates education systems by testing the Figure 35). Cognitive skills and the quality of learning skills and knowledge of 15-year-old students, and had outcomes have potentially large implications for eco- the worst performance among 70 countries in math and nomic growth through higher productivity gains.112 For science, with the highest share of low performers in all these productivity gains to improve the livelihoods of subjects. workers, they must be translated into salary increases, a situation which has not taken place in the Dominican Moreover, the country’s own standardized assess- labor market. ments in 3rd and 12th grades also highlight sys- tem-level inefficiencies which affect student learning. Health In 2017, only 12 percent of all 3rd graders achieved satisfactory levels of competencies in Spanish and 27 In the field of health, the country faces a similar sit- percent in mathematics, while the largest share of stu- uation, with gains in coverage but lagging quality. dents achieved only a basic level in both subjects (50 Coverage of health services and insurance is relatively percent in Spanish and 44 percent in mathematics). A high, with significant increases since 2007. The most similar situation was observed in the 12th grade, where recent statistics show coverage of health insurance for 32 of every 100 students could not finish their stud- 73 percent of total population in 2017, with full cov- ies, creating an educational dead-end which is costly erage of the poor enrolled in the Unique Beneficiary for students due to lower expected labor income, as System (SIUBEN). At the same time, the coverage well as for the education system. Young men, students gaps by income level and residence area (urban/rural) in public schools, and those in rural areas tend to un- have decreased since 2007. derperform in the 12th grade standardized assessment compared to young women, students in private schools, The quality of care and health outcomes is clearly and in urban areas. Grade repetition increases the risk lagging. The country faces a double burden of high 111  The index of socioeconomic level is a composite variable that includes information on education, income and labor market outcomes for parents, assets and books ownership, and characteristics of the dwelling. 112  Hanushek and Woessmann (2010). Dominican Republic: Systematic Country Diagnostic | 63 Figure 36: Percentage of individuals who consulted a public health facility 30 days before the survey, by year and income group 100 80 60 P rc nt 40 20 0 1 2 3 4 5 Quintil 2007 2013 Source: Demographic and Health Survey (DHS) 2007 and 2013 surveys, Dominican Republic. Note: For 2007, public health facilities include hospitals from the public network of the Secretaría de Estado de Salud Pública y Asistencia Social (SESPA) and rural clinics from SESPAS. For 2013, these include the public network of the Ministry of Public Health (MSP) and rural clinics and primary care centers from the MSP. SESPAS was renamed MSP. maternal and child mortality rates, and an increas- mortality, as well as preventable deaths from non- ing share of noncommunicable diseases. Health out- communicable diseases such as cervical cancer. De- comes are especially poor for mothers and young chil- spite high coverage and utilization rates of pre-natal dren, both groups with high mortality rates, and more controls and attention of births by qualified personnel generally, noncommunicable diseases (NCDs) are on -comparable to those in LAC and UMI countries, the the rise and now account for the highest share of the maternal mortality rate in the DR, at 92 per 100,000 burden of disease.113 The poor health outcomes do not births, is more than twice the rate for UMI countries only affect worker productivity and children’s learn- (41) and 33 percent higher than in LAC (69) (Figure ing outcomes, but cost human lives in the country. 37). As a result, the country did not achieve the ma- The slow improvements in outcome indicators leave ternal and child mortality MDGs, with infant and un- the country well behind the Latin America region and der-five mortality rates being the third highest in LAC the group of upper-middle-income countries (UMI). after Haiti and Bolivia, and maternal mortality above A reflection of the quality of public services is the the LAC and UMI averages. Various assessments indi- large share of the population using private facilities, cate that the majority of maternal deaths in the country especially among the higher income quintiles (Figure could have been prevented through a combination of 36). While the majority of the population used public factors including adherence to norms, protocols, and facilities for outpatient consultations (59 percent) and quality standards by properly trained personnel; ensur- inpatient services (51 percent), these shares are signifi- ing access to required equipment and inputs (depen- cantly lower than in other upper middle-income LAC dent on having sufficient funding); and enforcement countries such as Costa Rica (at least 80 percent) and of monitoring and supervision and regulatory arrange- Panama (at least 70 percent).114 ments. Moreover, neonatal mortality – which is partic- ularly sensitive to the quality of health care – has not The gap between coverage and quality of health is improved significantly in the past 15 years (21 deaths clearly illustrated in the case of maternal and child per 1,000 births in 2013, compared to 24/1,000 in 113  According to IHME (2017), the share of disability-adjusted life years (DALYs) accounted for by NCDs has increased from 40 percent in 1990 to 62 percent in 2016. 114  World Bank (2015). 64 | Dominican Republic: Systematic Country Diagnostic Figure 37: Maternal care and mortality, 2014: high coverage but lagging outcomes 41 Maternal mortality rate 69 (maternal deaths per 100,000 live births) 92 98 Births attended 94 by qualified personnel (%) 98 96 Pre-natal control (%) 97 98 0% 20% 40% 60% 80% 100% UMI LAC DR Source: World Development Indicators. 1999).115 Another sign that the quality of care is poor is While the country’s public spending on health’s share of that cervical cancer is the most frequent type of cancer GDP increased from 1.4 percent in 2004 to 2.9 percent among women between 15 to 44 years of age, account- in 2014, it remained lower than UMIC and LAC aver- ing for 11.8 percent of total cancer deaths among wom- ages of 3.4 and 3.7 percent, respectively, and the third en.116 Since cervical cancer is preventable and treatable lowest in LAC. when diagnosed at an early stage, the relatively high mortality rate suggests shortcomings in the screening While more public resources are needed in the health process and a lack in continuity of care. sector to keep on improving the quality of services and expand financial protection, there is room to Lack of sufficient funding is often cited as a major improve efficiency and cost-effectiveness of sector constraint to investments in health sector inputs, spending. One major rigidity in the sector is the criteria with total health expenditures in the country being for budget allocation. Despite expanding Results-based the lowest in LAC in 2014 (Figure 38). After falling financing (RBF) pilots in primary level care117 in more from 5.9 percent of GDP in 2000 to 4.3 percent of regions, health sector financing remains largely based GDP in 2004, total health expenditures as a share of on historical budgeting, and provides no incentives for GDP in the DR have been practically constant. Public quality or efficiency. Other inefficiencies result from pa- spending’s share of total health spending has increased, tients tending to bypass primary care centers and use leading to a significant reduction in the share of out of hospitals for basic care. Bypassing could also result from pocket spending, but both the country’s share of public weaknesses in the referral and counter-referral system, spending on health relative to its GDP and per capita and from the organization of services which emphasize public spending on health remain low relative to its in- curative care, resulting in insufficient resources at the come level. From 2004 to 2014, the public share of total primary level to adequately respond to patients’ needs health spending increased from 32 percent to 69 per- without going to hospitals. For instance, a study esti- cent while out of pocket spending’s share of total health mates that the Ministry of Health spent 62 percent of its expenditures decreased from 52 percent to 21 percent. recurrent budget on curative services and only 6 percent 115  Ministerio de Salud Pública, 2017. Situación de la Mortalidad Materna en la República Dominicana. 116  ICO Information Centre on HPV and Cancer. 2016 Dominican Republic: HPV and Related Cancers Fact Sheet. 117  A results-based adjusted capitation payment mechanism for primary health care facilities that was first piloted in regions VI, VII, and VIII has been expanded to pilots in at least five other regions. In this pilot, a portion of the capitation payments to private health care facilities is based on catchment population and a portion is based on attainment of targets in terms of coverage and adherence to care protocols for selected indicators. Financing of secondary and tertiary levels of care is based on fee for service. Dominican Republic: Systematic Country Diagnostic | 65 Figure 38: Public and private health expenditures (as percent of GDP), 2014 Paraguay Costa Rica Ecuador Nicaragua Honduras Uruguay Brazil Panama Chile Colombia El Salvador Bolivia Mexico Guatemala Peru Argentina Dominican Republic 0 2 4 6 8 10 H lth xp nditur , public (% of GDP) H lth xp nditur , priv t (% of GDP) Source: World Development Indicators. on primary care and health promotion in 2014.118 Large of the public system might weaken the incentives for disparities in the allocation of human resources have a individuals to pay taxes to contribute to the provision negative effect on the quality of services. Although the of public goods.120 country’s overall health-worker-to-population119 ratio increased from 19 in 2011 to 25 in 2015, exceeding the The 2012 National development strategy (END) minimum of 23 recommended by the WHO, there is a mentions clientelism as a behavior pattern that ham- large and persistent gap in health-worker-to-population pers the country’s development and that can be a ratios between urban and rural areas. The ratio in rural contributing factor to sustaining a low provision of areas increased from 17 percent of the ratio in urban public goods over time by weakening social account- areas in 2011 to 23 percent in 2015, far from its 2015 ability mechanisms.121 When this practice is prevalent, goal of 50 percent. voters pursue individual benefits rather than demand provision of goods with collective benefits. A number of studies emphasize the negative impact of this prac- Governance issues in the provision of tice on the provision of public goods.122 This practice public services been widely studied and is considered to be one rele- vant challenge in the country (see World Bank 2013, More broadly, in a setting of low quality of public Lozano 2013, Benito and Lozano 2012, Benito 2017, services, households have increasingly (with income Keefer and Vlaicu 2008, Keefer 2002, Mitchell 2009, and over time) opted out by adopting private solu- Morgan, Hartlyn and Espinal 2011, and Moya Pons tions to collective problems. For instance, almost 8 2013). International observers and local NGOs have percent of the poorest households had an electrical in- documented its practice in their reports of national verter in 2016, with the share increasing with income elections.123 Even though the extent of this strategy is to reach 48 percent of households in the richest decile. difficult to measure, the perceptions of its use are high In the case of education, those shares are 20 and 70 in the country.124 percent, respectively (Figure 39). In turn, opting out 118  Rathe and Hernandez (2015). 119  Health workers per 10,000 population. 120  See Sánchez and Senderowitsch (2012). 121  See Ministerio de Economía, Planificación y Desarrollo (2016), page 15. 122  World Development Report 2017: Governance and the Law. 123  See reports from the Organization of American States for most elections in the DR dating back to 1962 (http://www.oas.org/EOMDa- tabase). 124  See results from the LAPOP survey in Morgan, Espinal, and Zechmeister (2015). 66 | Dominican Republic: Systematic Country Diagnostic Figure 39: Share of students overage or failing school (dropout + not promoted), 2014 Panel A: Share of households Panel B: Share of households with with electrical inverter school-age kids paying for private school 80 80 60 60 P rc nt P rc nt 40 40 20 20 0 0 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 D cil D cil Source: Authors' calculations using ENFT. CHAPTER 4 Challenges to sustainability T here are signs of unsustainability in the country’s current development path. These include pres- sure on the management of natural resources from rapid urbanization that is creating problems of access to services; high vulnerability to disasters and climate-related risks with negative impacts on growth, fiscal accounts, and poverty. Agriculture expansion and the rapid growth of extractive indus- tries could put additional pressure on the country’s renewable natural capital assets. Preservation of these assets is critical to sustain the dynamism of pivotal sectors like tourism, more broadly, exports and to mitigate negative impacts from climate change. Tax incentives to attract FDI have limited the country’s ability to mobilize fiscal revenue and spend on human capital and other inclusion policies. More recently, growing social demands have led to narrowing fiscal space and renewed concerns about debt build-up. Broad governance concerns have led to mounting social discontent and might constraint the ability of the country to become a high-income economy by 2030. Raising fiscal revenue to meet social On the expenditure side, growing demands for demands and strengthen fiscal balance increased social spending and the fiscal costs of inefficiencies in the provision of public services Despite robust growth over the past 25 years, the have led to rising government expenditure. The country has struggled to maintain fiscal balance, es- country has been historically characterized by the pecially over the past 15 years. Over the 1990-2016 relative small size of its government, measured as period, the country averaged an overall fiscal deficit of the incidence of government expenditure as a share 2.4 percent of GDP. Fiscal deficits have become more of GDP. For example, in 1998 the incidence of prominent since the banking crisis of 2003—increasing government expenditure on GDP was 12 percent, from an average 1.3 percent of GDP prior to 2003, to while in the median LAC country it stood at 25 per- an average 2.8 percent of GDP after 2003. Beyond av- cent. Since then, the country’s public expenditure erages, the overall fiscal deficit has varied widely, going has risen and now stands at 17 percent of GDP. A from 1.3 to 8.0 percent of GDP. fraction of this increase came in response to social 68 | Dominican Republic: Systematic Country Diagnostic demands.125 For instance, public protests erupted in in 2015. Structural reforms have also supported the 2010 demanding government to fulfill a decade-old governments’ consolidation efforts and include (i) tax promise to spend 4 percent of GDP in education. administration improvements to reduce fraud and eva- These protests resulted in a national pact (Pacto Na- sion; (ii) public financial management reforms covering cional para la Reforma Educativa) whereby the gov- transparency of public accounts; and (iii) tighter public ernment committed to comply with the 4 percent debt management focused on extending the overall ma- rule. A fragile and inefficient electricity sector is the turity of public debt obligations on the back of favor- recipient of significant transfers from government, able external financing conditions. Since 2015, howev- which heightens pressures on an already tight fiscal er, fiscal deficits have widened, reaching 3.2 percent of envelope. Government has transferred on average GDP, and are expected to stand above 3.5 percent over 1.7 percent of GDP per year (12 percent of its rev- the next three years. enue) to the electricity sector over the past 10 years to cover for its losses (see chapter 2). In the face of recurrent fiscal deficits, the stock of public debt has been on the rise since 2007 and is Ongoing efforts to increase revenue notwithstand- projected to rise over the medium term. The non- ing, tax exemptions and a small tax base limit the financial public-sector debt more than doubled between country’s ability to sufficiently increase revenue. 2007 and 2017, going from 17.1 percent to 35 percent Over the past 10 years, government revenue has fluc- of GDP (Figure 40, panel B). The gross stock of debt tuated around 15 percent of GDP, while government of the Dominican government increases to 49 percent expenditure has averaged 17.4 percent of GDP over of GDP in 2017 if one includes the outstanding stock the same period. Ongoing efforts to fight evasion and of debt issued by the Central Bank for its recapitaliza- fraud are expected to raise revenue by 0.5 pp of GDP tion. In fact, the inclusion of the Central Bank’s deficit in coming years. These efforts notwithstanding, the adds approximately 1.5 percent of GDP to the govern- DR will still rank poorly in terms of revenue collec- ment’s deficit (quasi-fiscal deficit). Going forward, the tion. A cross-country comparison of fiscal revenues in gross external debt of the overall public sector (NFPS 2016 shows that the DR is one of the countries with and Central Bank) is projected to continue rising in the the lowest levels of revenue as a share of GDP in LAC medium-term, reaching 55 percent of GDP in the early (Figure 40, panel A). Low tax revenues can be partial- 2020s (IMF 2017). ly attributed to high tax expenditures (6.7 percent of GDP in 2016) resulting from numerous VAT and ex- Compared to regional peers, the country does not cise exemptions and other tax incentives provided to stand-out for its fiscal deficits or its public debt. In firms operating in special tax regimes, a relatively high fact, other LAC countries also have a propensity to pro- tax threshold (only 14 percent of formal workers pays duce fiscal deficits, which are in many cases higher than personal income taxes), and a low tax base because of those observed in the DR. The country’s average defi- informality (see World Bank 2017c). cit between 1990 and 2016 (2.4 percent of GDP) was lower than the LAC average (2.7 percent of GDP). A Policy makers have taken steps to rein in fiscal defi- finer comparison with sub-groups within LAC and its cits, but the results so far have been limited. Follow- structural peers reinforces this point—CAFTA-DR and ing the 2012 tax-reform package, fiscal consolidation Caribbean countries had an average fiscal deficit of 3.24 and a more favorable external environment (higher US percent of GDP and structural peers had an average growth and lower commodity prices) helped improve deficit of 2.9 percent of GDP (Figure 41, panel A). The macroeconomic balances temporarily, by gradually only sub-region that has a lower average fiscal deficit reducing the consolidated overall fiscal deficit from over the 1990-2016 is South America. In terms of debt, 6.9 percent of GDP in 2012 to 0.3 percent of GDP the gross stock of debt of the Dominican non-finan- 125  For example, data from the WHO and UNESCO show that between 2004 and 2014 public expenditure in education and health increased by 2.8 and 1.4 percentage points as a percentage of GDP, respectively. Dominican Republic: Systematic Country Diagnostic | 69 Figure 40: The DR has low levels of fiscal revenues which has contributed to an increase in public debt Panel A: Government revenue as a share of GDP Panel B: Gross public debt in the DR and comparator countries, 2016 in the DR as share of GDP 35 60% 30 50% 25 40% 20 30% 15 20% 10 5 10% 0 0% GTMLKA CRI DR SLV PER TUN NIC HND FJI ROUECU LAC 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017/1 Av r Source: WEO, IMF and BCRD. Non-fin nci l public s ctor C ntr l B nk Figure 41: The DR’s fiscal deficits and debt levels are low compared to regional peers Panel A: Overall fiscal deficits across Panel B: Gross public debt as share countries, 2010-2016 averages of GDP across countries, 2016 4 120 100 3 80 P rc nt of GDP P rc nt of GDP 2 60 40 1 20 0 0 Dominic n South LAC Structur l CAFTA C ribb n PRY URY BOL COL SLV PER DR SUR BRA MEX CHI GTM VEN NIC ECU PAN CRI LAC ARG JAM R p. Am ric P rs -DR nd C ribb n Source: WEO, IMF and BCRD. With the exception of Paraguay, Mexico and Uruguay, gross debt refers to the debt of the non-financial public sector. cial public sector was lower than 12 out of the 19 LAC ing fiscal accounts because of legal or political barriers. countries with available data and compared to the LAC Rigid expenditures include salaries, interest payments average (Figure 41, panel B). and social transfers. In the case of the DR, the compo- sition of government expenditure has progressively be- Despite comparatively low fiscal deficits and debt come more rigid—this type of government expenditure levels, the composition of government expenditure, accounted on average for 60 percent of total expendi- featuring a growing rigidity of expenditures because ture between 2000 and 2007, increased to an average of rising interest payments, and its pro-cyclical pat- 65 percent between 2008 and 2012, and reached an tern, adds complexities to the country’s fiscal prob- average 70 percent in the 2013-2017 period. The ris- lem. Countries where the composition of government ing degree of rigidity was mostly driven by an increase expenditure is more rigid have a harder time consolidat- in the expenses related to interest payments (Figure 42, 70 | Dominican Republic: Systematic Country Diagnostic Figure 42: Government expenditure in the DR is relatively rigid, which may contribute to the recent uptick in debt levels Panel A: Rigidity of government expenditure Panel B: Rigidity of government expenditure in the DR over time in comparator countries P rc nt of tot l ov rnm nt xp nditur s 100% 100 P rc nt of tot l ov rnm nt xp nditur 43 80% 80 57 57 51 48 60% 60 40% 40 66 60% 65% 70% 20% 20 43 43 49 52 0% 0 2000-2007 2008-2012 2013-2017 ECU PER NIC GTM CRI W s nd s l ri s Int r st Curr nt tr nsf rs Ri id xp nditur Non-ri id xp nditur Good nd s rvic s C pit l xp nditur s Panel C: Change in gross non-financial public-sector debt as share of GDP, 2010-2016 25 20 points of GDP 15 10 5 p rc nt 0 -5 -10 P ru Nic r u Gu t m l LAC v r Ecu dor DR (NFPS) Cost Ric El S lv dor ( xl. DR) Source: WEO, IMF and BCRD. Note: rigid expenditure is composed of compensation of employees, interest and social benefits; and non-rigid expenditure is composed of net acquisition of nonfinancial assets, not elsewhere classified and purchases/use of goods and services. The figure shows comparator countries for which data for all expenditure categories were available. panel A).126 A comparative look at the composition of Bennett 2017). The pro-cyclicality of fiscal policy in the government expenditure also shows that the country DR increases the country’s reliance on debt and limits has a more rigid structure compared to regional peers the country’s ability to construct buffers to respond to (Figure 42, panel B). In addition to a relatively rigid shocks (economic or otherwise). Finally, while falling composition of government expenditure, the country over time, the dollar denominated debt represents more displays a pro-cyclical government expenditure pat- than half of public debt, constraining the country’s abil- tern—spending more in times of economic bonanza ity to depreciate its currency (see Annex 3). and cutting expenditure during bad times. This pattern stands in contrast with other LAC countries, such as The underlying features of the fiscal process have Peru, Chile or El Salvador, which present a counter-cy- led to the accumulation of public debt at a faster clical pattern at least since 2007 (Végh, Lederman, and pace than its regional peers, raising concerns about 126  Social spending in education (4 percent) and health (2 percent) have been the largest areas of social spending (Oxfam 2017). These num- bers, however, have remained relatively stable since 2012. Up to December 2017, most of the Non-Financial Public Debt has a fixed interest rate (86.3 percent of total). Only 7.2 percent has a maturity less than one year. Dominican Republic: Systematic Country Diagnostic | 71 the current trajectory. As discussed above, the public to have riskier debt profiles, even after controlling for debt/GDP ratio in the country increased by approx- other characteristics everything else equal. imately 17 percentage points in the last 10 years.127 This period includes the aftermath of global financial crisis, when many countries relied on external funding Resilience to disasters, climate- to counteract the adverse effects of a sharp decline in related risks, and other shocks global growth. However, a look at the accumulation of debt across LAC shows that the DR had one of the While debt levels are expected to remain at manage- sharpest increases in its stock of public debt—it ex- able levels in the medium-term, the debt trajectory ceeded the LAC average and was higher than the in- could be significantly affected if adverse economic crease observed in all its regional comparators except or disaster and climate-related shocks materialize, for Costa Rica and El Salvador (Figure 42, panel C). as they would impact fiscal deficits and potentially increase financing costs. IMF (2017) estimates that Thus, the country appears to be in a cycle where- the country’s debt trajectory is highly susceptible to ad- by fiscal (and debt) dynamics put upward pressure verse growth shocks—it estimates that a 2.5-percentage on the country’s cost of financing, making expen- points slowdown in growth from baseline projections diture more rigid, which in turn limits fiscal space for two years could result in a 5-percentage points in- and increases the need for external financing. As crease in the country’s debt to GDP ratio by 2020. Pub- shown earlier, despite being one of the fastest growing lic debt is also sensitive to exchange rate movements, economies in LAC, the country has one of the highest since it is mostly dollar-denominated. Indeed, follow- burdens of government interest payments relative to ing the PetroCaribe debt buyback operation, the coun- debt levels in LAC, which increases the need for exter- try increased issuance of dollar-denominated bonds in nal financing. Moreover, the country’s sovereign debt 2015 and external public debt securities are now esti- is perceived to be riskier to what its current debt level mated to account for more than half of the total ex- predicts (Végh et al. 2018).128 Part of the DR’s relative- ternal debt (this portion was about a quarter in 2008). ly poor standing in terms of risk rating, and the asso- This puts pressures on the Central Bank to intervene in ciated relatively high cost of financing, is explained by forex markets and prevent sharp depreciations of the ex- the rapid increase in its debt levels and its exposure to change rate. Finally, debt dynamics could also be altered external shocks. An extension of the exercise in Végh et by weather-related shocks, which may depress econom- al. 2018 shows that, conditional on debt levels, GDP ic activity in the short-run and increase the country’s per capita, and the size of an economy, countries that cost of financing. Recent World Bank staff’s calculations experience larger accumulation of debt are perceived showed that the materialization of a counterfactual nat- to have riskier sovereign debt profiles compared to ural disaster in 2017 could have led to a 7-year long those that have stable debt to GDP ratios.129 Similarly, primary deficit increase of 1 to 1.5 percentage points countries in the Central America and the Caribbean as a share of GDP relative to baseline, depending on (including the DR), which are constantly exposed the intensity of the shock.131 This, in turn, would have to natural disasters and are highly exposed to shocks translated into a 4 to 12 percentage point increase in stemming from North America,130 are also perceived the debt to GDP ratio relative to the baseline level in 127  The increase in public debt includes a one-off issuance to recapitalize the Central Bank. 128  The analysis uses Fitch sovereign risk rating. 129  As in Végh et al. (2018), sovereign risk is proxied by Fitch ratings. 130  Estimations presented in Végh, Lederman, and Bennett (2017), show that the growth of the DR’s GDP is highly elastic to the growth rate of G7 countries, of which the US represents a big part. More specifically, the authors use quarterly data to estimate the elasticity of year-on-year growth rates of different LAC countries with respect to external factors (growth of G7 countries, China’s growth, commodity prices, and the US 10-year treasury rate). Their findings show that, after controlling for other external factors likely affected by US growth, the elasticity of the DR’s growth (in terms of deviations from the country’s average growth) with respect to G7 growth is positive and significant, and is the fourth largest among LAC countries with available data. 131  More precisely, the exercise simulates the economic impacts of natural of different magnitudes hitting the DR, one event of moderate intensity and one event of strong intensity. 72 | Dominican Republic: Systematic Country Diagnostic Figure 43: Natural disasters could increase sovereign borrowing costs in the DR Dominican Republic Vietman Taiwan, China Philippines Mexico Thailand 0,0 0,5 1,0 1,5 2,0 2,5 3,0 H poth tic l r tin down r d in notch s Source: Calculations from Standards and Poor’s. 2024. In addition, estimations by Standards and Poor’s generation, national system of protected areas, human show that the country’s credit rating is more sensitive settlements and tourism.134 to a large disaster, compared to other emerging markets (Figure 43), thus affecting its financing costs.132 Tropical storms and floods are the most frequent climate-related hazard in the DR. The northeastern The downside to the country’s geographic location region is exposed to floods and mudslides from severe is its high exposure to adverse natural events. The storms, while arid parts of the northwest are experienc- country is highly exposed to rapid weather-related di- ing increasing temperatures leading to more drought, sasters (tropical storms, hurricanes, cyclones, floods and which reduces crop yields and water supplies. More- landslides), climate-related events (drought), climate over, Hispaniola Island is in the center of a hurricane change impacts (sea-level rise and desertification), and belt, where intense storms have a negative impact dam- earthquakes. A global comparison suggests that the aging hotels, coastal infrastructure and beaches, leading DR stands out as one of the countries which has been to significant losses in tourism revenues, and damages most affected by natural disasters in the last 20 years, to fish nursery areas and coral reefs. Climate change with the country ranked 10th worldwide terms of hu- is expected to increase the risk and intensity of excess man and economic losses suffered from weather-related rainfall, storm surges, and hurricanes in the Caribbe- events between 1997 and 2016 (Germanwatch’s Global an, with relatively higher impacts in the Dominican Climate Risk Index 2018).133 From 1980 to 2008, 40 Republic. Predicted impacts include flooding in coastal disaster events affected 2.65 million people, almost a zones and low-lying areas, with potential damage and quarter of the country’s population. The DR’s first Na- economic losses projected to reach nearly 17 percent of tionally Determined Contribution (NDC) identifies GDP in the country.135 climate change adaptation as a constitutional priori- ty of the country, with the sectors identified as most Vulnerability to adverse natural events is exacer- climate-vulnerable including drinking water, energy bated by unplanned urban growth, land degrada- 132  So far, the DR has seen improvements in credit ratings and access to bond markets. In Dec. 2017, OECD improved the risk rating for the DR from 5 to 4. The DR has as improved access to international financial markets, evident from the latest issuance, at 8.9 percent for a 5-year bond, that was more than twice over-subscribed than the offered amount. However, as mentioned in the text, these improvements could be reversed in case a shock hits the country. 133 https://germanwatch.org/en/download/20432.pdf 134 http://www4.unfccc.int/ndcregistry/PublishedDocuments/Dominican%20Republic%20First/INDC-DR%20August%202015%20(un- official%20translation).pdf 135 https://www.gfdrr.org/en/dominican-republic Dominican Republic: Systematic Country Diagnostic | 73 tion, and weak enforcement of building codes and quently reported events during this period, the coun- zoning regulations. The DR is ranked 26 out of 171 try’s territory is marked by a long history of destructive countries in the United Nations’ World Risk Index,136 earthquakes.137 which covers four dimensions: i) exposure to natural hazards; ii) vulnerability as dependent on infrastruc- Between 1961 and 2014, disasters in the country – ture, nutrition, living conditions and economic cir- mainly of hydro-meteorological origin– have caused, cumstances; iii) coping capacities as dependent on on average, annual economic losses around 0.69 per- governance, preparedness and early warning measures, cent of 2015 GDP. At the sectoral level, approximately access to healthcare, social and material security; and iv) half of the direct and indirect economic losses (Figure adapting capacities with respect to impending natural 45, panel A) are concentrated in the agriculture (32 per- events, climate change and other challenges. cent of the losses) and transport sectors (18 percent). In addition, population growth and rapid urbanization Over the past 50 years, the country has experienced over the last 60 years have significantly increased the a higher number of disasters from adverse natural country’s total exposure and the value of assets at risk of events per year, relative to its land surface, than the adverse natural events. Recent estimates suggest that the average for both Central American countries and negative impacts of storms last up to 15 months after Caribbean countries. Out of all the disasters reported the strike, with an average reduction in GDP of about between 1967-2016, 93 percent are of hydro-meteoro- US$1.1 billion (4.5 percent of gross domestic product logical origin (Figure 44, panel A). The frequency of in 2000 and 1.5 percent in 2016).138 disaster events has increased significantly in the last 5 decades, shifting from one event every two years on av- Catastrophic events have the potential to destroy erage in the 1960s, to almost one event per year in the a greater share of assets in DR than the average of 1990s, and to 2.6 events per year in the 2000s (Figure Central American and Caribbean countries. The 44, panel B). Even though earthquakes are the least fre- country could experience an annual average loss Figure 44: Frequency of natural disasters in LAC and number of disasters in the DR Panel A: Number of natural disasters Panel B: Number of disasters events in the DR by event type LAC (1967-2016) by decade and event type (1960-2017) 12 Gr n d S int Kitts nd N vis Anti u nd B rbud 10 Dominic S int Luci Numb r of v nts B li 8 Av r C ribb n J m ic Av . C ntr l Am ric & C ribb n 6 El S lv dor P n m Av r C ntr l Am ric 4 Dominic n R public Nic r u Cost Ric 2 Hondur s Gu t m l H iti 0 0 20 40 60 80 100 1960-69 1970-79 1980-89 1990-99 2000-09 2010-17 Numb r of v nts Tropic l Storms Floods Drou ht H drom t orolo ic l G olo ic l Oth r E rthqu k Wildfir Source: EM-DAT: The Emergency Events Database. 136 http://weltrisikobericht.de/english/ 137  Notably, the city of Santiago has been devastated twice by earthquakes in 1562 and 1783. In 1946 a 8.0-magnitude earthquake struck the Samaná peninsula resulting in a considerable loss of lives and severe damages reported across the country. 138  Ishizawa, Miranda, and Strobl (2017). 74 | Dominican Republic: Systematic Country Diagnostic Figure 45: Sectoral distribution of losses and potential maximum losses Panel A: Sectoral distribution of direct and Panel B: Potential Maximum Loss (PML) from indirect losses caused by major disasters, hurricanes and earthquakes for countries in Central 1967-2017 (percent of total) America and the Caribbean, return period 250 years A ricultur Gr n d B li Tr nsport S int Luci Comm rc Av . C ribb n Housin J m ic Dominic n R public En r Av . CA nd C ribb n Tourism Hondur s W t r & S nit tion Av . C ntr l Am ric H iti Em r nc Nic r u Educ tion El S lv dor Environm nt Gu t m l Cost Ric H lth P n m 0% 5% 10% 15% 20% 25% 30% 35% 0 5 10 15 20 Sh r of tot l xpos d v lu Dir ct loss s Indir ct loss s PML hurric n s PML rthqu k s Source: Based on assessment conducted with the support of CEPAL and Source: World Bank Country Disaster Risk Profiles. Note: averages for MEPyD’s assessment. Central America and the Caribbean exclude the DR. from earthquake and hurricanes equivalent to about in chapter 3, the vulnerable population is the largest in- 4.3 percent of the budget for 2017.139 The coun- come group in the DR (41 percent). Individuals in this try has a unique combination of similar and rela- group remain at risk of falling back into poverty if af- tively high potential losses from catastrophic hurri- fected by shocks, including disasters, against which they canes and earthquakes among countries in Central cannot protect or insure themselves. High speed winds American and the Caribbean (Figure 45, panel B).140 during 2000-2012 are estimated to be associated with For a 250-year return period the DR’s Probable Maximum an increase of the moderate poverty rate of 1-2 percent- Loss (PML), expressed as a percentage of total building age points, and an increase between 0.5 and 1 pp of exposure, is higher than the Central America (CA) aver- the extreme poverty rate. At the global scale, it is esti- age for both hurricanes (DR 7.5 percent; CA 4.0 percent) mated that the impacts of disasters are more than twice and earthquakes (DR 7.3 percent; CA 6.6 percent).141 as significant for poor people than for anyone else.143 These results highlight the multi-hazard feature of the At the national scale, the climate shocks vulnerabil- country and the importance of addressing both hurri- ity index (IVACC, from its Spanish acronym), which cane and seismic risks. measures the likelihood that a household is vulnerable to the occurrence of hurricanes, storms, and flooding Disasters can hinder poverty reduction efforts and given certain socioeconomic and geographical charac- threaten advances in shared prosperity both through teristics, is twice as high for the poorest households and the triggered economic losses and through direct im- decreases as the standard of living increases. When trop- pacts. Shocks created by adverse natural events have re- ical cyclone Noel hit the country in 2007, 90 percent gressive distributional effects as vulnerability to climate of the directly affected persons were below the national shocks is higher for the poorest households.142 As seen poverty line.144 139  World Bank Country Disaster Risk Profile. 140  This set of countries includes the six countries in Central America and Belize, Grenada, Saint Lucia and Jamaica in the Caribbean, in which the World Bank has prepared Country Disaster Risk Profiles. 141  The 250-year return period PML indicates a level of losses that has a probability equal or superior to one to be exceeded in 250 years. In others words it is a level of loss that has an annual probability of 0.4 percent of being exceeded. 142  Báez, Fuchs, and Rodríguez-Castelán (2017). 143  Hallegatte et al. (2017). 144  CEPAL (2008). Dominican Republic: Systematic Country Diagnostic | 75 Environmental sustainability vere inequalities in access to economic opportunities by transport services.147 The condition is critical for transit The country is highly exposed and extremely vulnera- services that carry 70 percent of all trips in the city and ble to disasters and climate-related shocks. Vulnerabil- represents most of the times the only mobility option ity is being exacerbated by climate change, rapid and for the bottom 40 percent of the population. Despite unplanned urbanization, and natural resource degrada- the introduction of a metro line (heavy rail transit) and tion. the recent construction of a second line, the system only moves 7 percent of the total trips in the city. Data from Unplanned, rapid urbanization has created imbal- 2007 shows that average commuting times within the ances along different dimensions. The country has National District (core urban area in the GSD) was over urbanized faster than the LAC region and the group one hour, more than any other capital in the region, of upper middle income (UMI) countries in the last leaving expected figures for the areas outside the core 25 years. With an estimated 55 percent of the popu- urban area at worrying levels. Transit service provision lation living in urban areas back in 1991 (15 pp lower is dominated by carros concho (unregulated car sedans than LAC and 12 pp higher than UMI), the country is that operate as a paratransit service), with six times as now on par with the region at 80 percent (and 15 pp many vehicles as the total bus and minibus fleet in the higher than UMI – see Figure 46). While urbanization GSD. On top of the limited access to economic oppor- processes generate positive agglomeration externalities tunities by the low-income population due to conges- for economic growth, unplanned urban expansion also tion and poor transit service provision, road safety is generates negative externalities such as high pressure on a serious issue in the country. The DR is the western natural resources, pollution and congestion, which af- hemisphere’s most dangerous place to drive, and 15th fect human health and quality of life. worst in the world, with 29 out of every 100,000 people in DR dying in road accidents every year, compared to The rapid urbanization has also posed a challenge in 15.9 per 100,000 in LAC and 16.7 in upper middle-in- the Greater Santo Domingo (GSD) and secondary come countries.148 cities in terms of the location of new urban dwell- ers and their access to the labor markets by transit Some of the imbalances that have been exacerbated and other modes of transport, with a serious road by rapid urbanization relate to the access to water safety aspect. Total building exposure is estimated at and sanitation. The country lost half of its per capita US$153 billion, which, as a percentage of GDP is 40 water resources in the last 25 years due to population percent higher than the Central American average. For- and economic growth, and poor management. Even ty three percent of the total building exposure is con- though 80 percent of the population live in urban ar- centrated in Santo Domingo, with 42 percent in sec- eas, agricultural activities use -with low levels of effi- ondary cities outside the capital.145 Because of the rapid ciency- 83 percent of the available volume, mostly from economic growth, personal motor vehicle ownership superficial sources.149 The country has a high water scar- in the country has tripled in the last 15 years, most of city index, especially in the regions of the Yaque del Sur them exacerbating the already congested streets of the and Yaque del Norte river basins, where water demand GSD.146 For the 3.5 million-habitant GSD area, several reaches 86 percent and 66 percent of the available sup- unplanned peri-urban neighborhoods are presenting se- ply, respectively.150 145  World Bank Country Disaster Risk Profiles. A building exposure model is a building inventory (stock) database which captures important attributes such as geographical location, urban/rural classification, type of occupancy (e.g., residential and five types of non-residential use), structural typology (e.g., wood, concrete, masonry, etc.) and economic (replacement) value. For the asset value determination in the residential sector, unit costs of construction that depend on the type of building (structural vulnerability class) were derived. Globally, international estate agency reports also provide unit cost values for the higher-end of housing and non-residential sectors. 146  The national motor vehicle fleet grew from 1.2 million in 2000 to 4.1 million in 2017 (Dirección Nacional de Impuestos Internos, DGII). 147  Observatory of social policies and development, bulletin 7 (2017). 148  According to the WHO Global Status Report on Road Safety (2015). 149  Data for 2010 from FAO, 2016. AQUASTAT website http://www.fao.org/nr/water/aquastat/main/index.stm [accessed on 2018/01/04]. 150  Ministerio de Medio Ambiente y Recursos Naturales (2012). 76 | Dominican Republic: Systematic Country Diagnostic Figure 46: Urbanization rates in the DR, LAC and upper-middle income countries (UMI), 1991-2016 80 LAC 70 DOM P rc nt 60 UMI 50 40 1991 1996 2001 2006 2011 2016 Source: World Development Indicators. Access rates to water supply and sanitation remain activity, one of the main sources of growth and jobs in relatively high, but wastewater collection and treat- the country.153 ment continue to lag behind water supply. Only 53 percent of the water treatment plants and 26 percent Increased pressure on water resources from urban- of the fully-functioning wastewater treatment plants ization and land use change combined with reduced were estimated to be at an adequate operating level, cre- precipitation under a climate change scenario, ating conditions for the proliferation of water- and san- threatens the main engines of growth in the coun- itation-borne diseases. Untreated municipal wastewater try. Hydrogeological and climate change models pre- and fecal sludge from on-site sanitation is one of the dict a reduction in spatial distribution of rainfall by main contributors to the water resource degradation re- approximately 60 percent in the next 100 years and sulting from urbanization.151 Approximately 38 percent total runoff by 95 percent for the year 2100, affecting of the urban population has access to sewer networks, water availability, and aquifer recharges.154 In addition, but current trends suggest that it will drop to 25 per- demographic growth, accelerated environmental degra- cent, while usage of on-site septic tanks is expected to dation, deforestation, and lack of mitigation measures rise from 25 to 51 percent by 2030.152 A small percent- could intensify the pressure on water resources. Envi- age (10-20 percent) of the wastewater that is collected ronmental degradation, mainly from soil erosion, has through networks is properly treated. Treatment of fe- an important economic impact for the hydropower in- cal sludge is essentially non-existent. Estimates suggest dustry, as it reduces dam capacity, increases the costs re- that over 200,000 tons/year of sludge originating from lated to the maintenance of irrigation and water supply treatment plants, septic tanks, and latrines goes untreat- infrastructure, and worsens coastal water quality. ed contaminating surface water and groundwater. San- to Domingo, for instance, generates 13,680 liters per Inadequate resource management has led to envi- second (lps) of waste water, but current sewerage in- ronmental degradation and increased vulnerability frastructures are collecting only 14 percent (1,900 lps). to climate change. Estimates indicate that around 64 Excess wastewater contaminates rivers and watersheds, percent of the Dominican territory is critically sensitive posing environmental risks and threats to the health of to desertification mainly caused by intense and inad- communities, and has the potential to affect touristic equate land use practices,155 and nearly 40 percent of 151  World Bank (2004). 152  Joint Monitor Programme UNICEF/WHO. 2017. Estimates on the use of water, sanitation, and hygiene in the Dominican Republic 153  The Environmental Performance Index (2014) ranks Dominican Republic at 113 and 99 out of 178 countries in terms of access to drink- ing water and sanitation, and wastewater treatment, respectively. 154  Secretaría de Estado de Medioambiente y Recursos Naturales (2004). “Primera Comunicación Nacional” (PDF). UNDP. p. 44. 155  Izzo et al. (2013). Dominican Republic: Systematic Country Diagnostic | 77 soils are considered degraded. In the long run, soil deg- upgrading of the tourism sector, are increasing- radation can drive desertification processes, especially ly under pressure. The continued loss of coral reefs, under climate change scenarios. Information available mangroves, and wetlands that protect its 1,600 km of on levels and sources of pollution of coastal zones is of- beaches from storm surges are a major threat to coastal ten limited, but several studies indicate that watershed communities and the tourism industry. Virtually all of degradation in coastal areas (e.g., Puerto Plata or Boca the DR’s critical marine resources are at risk and 70- Chica) originates from inland activities and has led to 90 percent of coral reefs are already dead.161 Drivers of a decrease in the number of tourists in the country.156 change include overfishing, coastal development, pol- For instance, livestock production is a force of land use lution, and climate change. The decline of the marine change in the country and occupies the largest area assets may prevent ocean-economy sectors and indus- of hillside land in the upper and middle basins of the tries from reaching their potential, hampering econom- country. It is estimated that the pasture area is currently ic growth. The marine resources and ecosystems, if well five times larger than the optimal area for such use.157 managed, have the potential of helping with the devel- Estimates of climate change impacts indicate that to- opment of higher-value tourism, growing the DR’s blue tal precipitation in the country could decrease by 11 economy,162 ensuring that economic activity is in bal- percent between 2010 and 2030,158 and temperature is ance with the long-term capacity of these ecosystems, projected to increase between 0.5 and 1 degree Celsius and increasing people’s resilience to climate change. by 2030, which could have severe impacts on the ag- Ecosystem-based adaptation and increasing resilience to ricultural sector if climate smart agricultural practices coastal and marine systems are part of the elements of are not adopted. Finally, drier seasons imply more land the strategic planning approach identified in the NDC. prone to fires. During the 2003-2013 period, more than 2,000 forest fires occurred in the country, affecting Although the country has been successful at con- over 90,000 ha.159 trolling deforestation, pressures on forests from ag- riculture remain. The DR has increased its forest cover The Dominican Republic is showing relatively high- from 22 percent of the territory in 1973 to 39 percent er water footprints, that are polluting water resourc- in 2012.163 In addition, World Bank wealth estimates164 es, from specific economic and consumption activi- suggest that, between 1995 and 2014, DR’s remarkable ties, compared to Latin American average. Grey water GDP growth has been paired with a positive change footprints160 from domestic water supply in Dominican in forest land values (Figure 47). Despite progress in Republic (9 percent) triples the average of Latin Amer- reversing deforestation trends and increases of the forest ica (3 percent). This indicates the need to accelerate cover, the agricultural sector still drives 60 percent of wastewater treatment and increase the sustained use of deforestation, especially through the expansion of slash resources because it affects other segments of the econ- and burn agriculture and poor extensive livestock pro- omy and livelihoods of communities, particularly those duction practices in protected areas on the border with with higher vulnerability to natural hazards. Haiti. Other direct causes of deforestation and forest degradation include illegal logging, natural disasters, Marine assets, which form the country’s blue econ- forest fires, and infrastructure projects for mining, en- omy and are a crucial ingredient for the growth and ergy and tourism. In fact, recent unpublished estimates 156  See Page, Essex and Causevic (2014) and World Bank (2004). 157  Ministerio de Medio Ambiente y Recursos Naturales (2014). 158  Estrategia Nacional de Cambio Climático in Dominican Republic Policy Notes. 159  Forest Carbon Partnership Facility (FCPF) Carbon Fund (2015). 160  Grey water footprint is the amount of fresh water required to assimilate pollutants to meet specific water quality standards. The grey water footprint considers point-source pollution discharged to a freshwater resource directly through a pipe or indirectly through runoff or leaching from the soil, impervious surfaces, or other diffuse sources. 161  USAID (2013). 162  Patil et al. (2016). 163  Ministerio de Medio Ambiente y Recursos Naturales (2014) y FAO (2015). 164  See Lange, Wodon and Carey (2018). 78 | Dominican Republic: Systematic Country Diagnostic FIGURE 47: Growth of forest capital and GDP, 1995-2014 (average annual growth rate, percent) 3 DOM 2 For st c pit l rowth CRI 1 SLV MEX PER COL 0 0 1 2 3 4 5 6 ARG GTM -1 BOL PRY NIC -2 GDP rowth Source: World Bank (2017). The Changing Wealth of Nations. of forest cover and land use change by the Ministry of gion’s on political stability, policy design and imple- Environment and Natural Resources (MARN) suggest mentation (regulatory quality in Figure 48, panel A), that deforestation may have increased in recent years. freedom of expression and participation (voice and accountability), and the quality of contract enforce- Mangroves have paid the toll of tourism expansion, ment and property rights (rule of law). In addition to exacerbating coastal vulnerability. The expanding weaknesses in the control of corruption, perceptions area devoted to tourism infrastructure results in signif- of government effectiveness, which include the quality icant forest area losses, especially mangroves. The im- of public services, the civil service, and the credibil- pact of tourism on biodiversity is considered high in the ity of the government’s commitment to its policies, coastal forests, mangroves, and seagrass beds, which are are lower than in LAC. However, this comparison seriously threatened ecosystems. A study of the dynam- ignores improved perceptions of governance issues in ics of land use and coverage change, estimated that in the country since 2004. The country is perceived to be the 2000-2010 period about 436 hectares of mangrove more politically stable, its government more effective, forest were lost due to tourism activities.165 Mangroves with higher regulatory quality and enhanced rule of are not only home to a large variety of fish, crab, shrimp law (Figure 48, panel B). and mollusk species, but also provide coastal protec- tion through the stabilization of coastline, and prevents The National Development Strategy (END) high- damages from waves and storms. lights the importance to fight corruption, clientelism, and patrimonialism to achieve an effective national development.166 This is an appropriate area of focus in Social accountability the END as using the World Governance Indicators, the DR is the third lowest LAC country in terms of control Compared to LAC regional averages, the country of corruption, and the 2017-2018 World Economic Fo- fares relatively well in several dimensions of gover- rum ranks it 135th among 137 countries on ethics and nance, with improvements since 2004. Perceptions corruption. Promising initiatives include the Participa- of governance issues in the DR are similar to the re- tory Anti-Corruption Initiative (IPAC, see Box 5). 165  Forest Carbon Partnership Facility (FCPF) Carbon Fund (2015). 166  “Se parte del entendimiento de que nuestro actual modelo de desarrollo presenta falencias en materia de la calidad de la acción estatal, que se manifiestan en fuertes deficiencias en los servicios públicos que deberían garantizar los derechos constitucionales de la población […], así como patrones de actuación que entorpecen el desarrollo nacional, tales como corrupción, clientelismo y patrimonialismo.” Estrategia Nacional de Desarrollo 2030 (“We begin with the understanding that our current development model has shortcomings in the quality of state action, which are manifested in strong deficiencies in public services that should guarantee the constitutional rights of the population […], as well as patterns of action that hinder national development, such as corruption, clientelism and patrimonialism.” National Development Strategy 2030). Dominican Republic: Systematic Country Diagnostic | 79 Effective, efficient and transparent procurement of 101 out of 111 countries) in the 2017 Global Right to goods and services is critical to maximize the value Information (RTI) ranking, which analyses the quali- for money and minimize risks of corruption. Procure- ty of Access to Public Information (API) laws. Four- ment is a cross-cutting government function concerning teen years after the enactment of the API Law (Law almost all government institutions. Efficient and trans- 200 of 2004), the country still faces a few challenges parent procurement practices ensure the effective use of in terms of its implementation. For example, polit- resources with quality public services.167 According to the ical parties do not comply with the mandate to pro- World Bank’s Benchmarking Public Procurement Report vide information on the origin and destination of their for 2017, government providers wait about six months or funds, and many institutions are yet to disclose infor- more to receive payment,168 which reduces the spectrum mation on public officials or procurement processes. of companies that are able to do business with the govern- In terms of open data, the country scored 27 percent in ment, and raises the price of contracts to offset payment the Open Knowledge International’s 2016 Open Data delays. According to the latest PEFA report, there was a Index, which measures the openness of government significant discrepancy in 2015 between the information data, and ranked 57 out of 94 countries, with only 6 reported in the government’s procurement system and percent of government databases which are fully open. the data on budget execution in the Integrated Finan- In addition to the weak legal and institutional frame- cial Management System (SIGEF),169 pointing to serious work for API, an impact evaluation of the API pilot shortcomings with the registry of amounts contracted. project on empowerment of citizens in poor mu- nicipalities found no impact on the awareness and Further improvements in terms of transparency are use of information under the specific API rules. needed for the impact of citizen engagement and Moreover, despite progress in the disclosure of budget social accountability initiatives to fully materialize. information in recent years, the International Budget The country ranked lowest in the LAC region (and Partnership reports that the country still provides few Figure 48: Governance at a glance: The DR in comparative perspective and over time Panel A: LAC and the DR, 2016 Panel B: The DR in 2004 and 2016 Voic nd ccount bilit Voic nd ccount bilit Control of Politic l st bilit Control of Politic l st bilit curruption / no viol nc curruption / no viol nc Rul Gov rnm nt Rul Gov rnm nt of l w ff ctiv n ss of l w ff ctiv n ss R ul tor R ul tor LAC 2016 DR 2016 DR 2004 DR 2016 Source: Authors' calculations using the World Governance Indicators. 167  In 2015, Central Government institutions in the DR awarded contracts for RD$ 54,120 million, representing around 9 percent of nation- al expenditures for that year. See PEFA 2016 (Giussani et al. 2016). 168  See more at World Bank (2017e). 169  SIGEF data on budget execution corresponding to purchase of goods and services and fixed investment expenses and purchase of capital goods totaled RD$ 108,577 million in 2015, well above the RD$ 54,120 million reported in HYPERLINK “http://www.comprasdominicana. gob.do” www.comprasdominicana.gob.do. 80 | Dominican Republic: Systematic Country Diagnostic opportunities for the public to understand and question documentation by Haitian migrants increases their vul- the budget process (with a score of 17 out of 100). nerability to exploitation and discrimination, and limits their access to better paid jobs.170 Social cohesion Three Dominicans out of four identify crime and vi- olence as the most important challenge the country An issue affecting social cohesion in the country was faces, with unemployment being the second at 42 the discrimination against Haitian migrants and percent (Figure 50).171 The issue is very much present their descendants. An estimated 130,000 individuals in the daily lives of citizens, with seven out of 10 people of Haitian ancestry born in the Dominican Republic reporting it being a conversation topic in the last two were affected by a sentence by the Constitutional Court weeks; the share is even higher in urban areas. Victim- in 2013. The sentence 168-13 ordered the review of ization rates have increased in the last 10 years, with 30 all births registered in the country from 1929 to 2007 percent of people 12 and older reporting being a victim to ensure that no children of foreigners in transit had in 2015, up from 21 percent in 2005. Perceptions of mistakenly been registered as Dominican nationals. insecurity have also increased from 40 percent in 2008 The sentence applied the current definition of transit to 59 percent in 2016, the highest level since 2006, retroactively, with the effect of revoking the citizenship and the second highest in LAC -following Venezuela- rights of numerous persons born in the DR who would among 29 countries.172 have had a right to citizenship under earlier regulatory definitions of transit. Law 169-14 was passed to rem- Firms are also affected by crime and violence, with edy this situation by establishing a special regime for a higher share of firms that experienced losses due Dominican-born descendants of non-resident foreign- to theft or vandalism in the DR than the average ers who were registered between 1929 and 2007, re- for the Caribbean. Crime and violence are costly, instating their Dominican citizenship (about 55,000, with a potentially larger economic impact in coun- of which only 13,000 repossessed their documents), tries that rely on tourism, like the DR. Estimates and for those descendants who are not registered, who from the IDB for 17 countries in LAC find an aver- would be registered as foreigners (fewer than 9,000 ap- age cost of crime and violence of 0.5 percent of GDP. plications as of late-2017). Since mid-2014 the Domin- In the presence of crime and violence individuals ican government has been implementing three parallel change their behavior. Close to 60 percent of firms in processes to: i) regularize foreigners with irregular mi- the DR spent money on security (once again, higher gration status; ii) partially restore identity documents than the average for the Caribbean),173 and 85 per- for descendants born in the DR; and iii) register per- cent of households did something to avoid burglary.174 sons born in the country of foreign ancestry. As part Other costs include the reduction of investment by of the regularization process of foreigners with irregular firms, incurring productivity losses, and the allocation migration status, around 250,000 migrants were tem- of resources by governments to tackle the associated porarily regularized as non-residents, and are exempt problems.175 from deportation, with very few having access to the social security system. Fewer than 5,000 migrants have obtained residence as part of this process. The lack of 170  Wooding and Riveros (2017). 171  ONE (2017). 172  Espinal et al. (2017). The text of the question reads: “Talking about the neighborhood where you live and thinking about the possibility to be a victim of assault or theft, do you feel (a) very safe, (b) somewhat safe, (c) somewhat unsafe, or (d) very unsafe? [“Hablando del lugar o el barrio donde usted vive y pensando en la posibilidad de ser víctima de un asalto o robo, ¿usted se siente muy seguro(a), algo seguro(a), algo inseguro(a) o muy inseguro(a)?”] The variable was transformed to a scale from 0-100, where 0 means ‘very safe’ and 100 means ‘very unsafe’. 173  Jaitman (2017). 174  ONE, op. cit. Actions include reinforcing security of doors and windows, installing steel fences, and having guard dogs. 175  Jaitman, op. cit. Dominican Republic: Systematic Country Diagnostic | 81 BOX 5. The Participatory Anti-Corruption Initiative (IPAC) The IPAC was conceived as a response to a government request to the Donor Community to address the wide- spread perception of corruption in the Dominican Republic. It mainly focused on building a multi-stakeholder coali- tion comprised by civil society, government, and private sector to promote important public sector governance re- forms directed to reduce the scope of corruption. Following consultations and some preparatory work, the program started in mid-2010. IPAC was supported by the donor community, particularly by USAID, IDB, UNICEF, AECID, the European Union (EU), and the World Bank—which also had a coordinating role among donors. Following the preparatory work, the government set up ten working groups to identify potential reforms in 5 prior- ity sectors: water, energy, health, education, and infrastructure, and 5 cross-cutting areas: civil service, financial management, procurement and acquisition, access to information, and external control. In these working groups, representatives from the government outlined concrete steps that were needed to reform each area and to monitor the implementation of the reforms. The steps were intended to be implemented in the short term (most of them within one year, though some others indicated the need for a longer timeline for their implementation) with govern- ment funding and some support from the donor community. The recommendations were approved by the Cabinet during a day-long meeting. A Government focal point was appointed to ensure the implementation and monitoring of each recommendation. Main achievements IPAC made possible to have, for the first time, different public institutions collaborating on developing a working plan to reduce the risk of corruption in Dominican Republic. The public health working group was instrumental in accelerating a reform of the unified procurement system for drugs and medical equipment that had been stalling for years. As a result, the new procurement system led to a reduction of 65 percent of the cost of medicines and medical supplies, with a net saving for the government of US$25 million in 2013. From the education working group, a new and very powerful coalition (Dominican Initiative for Quality in Education, IDEC) was formed to tackle the root causes of poor service delivery in the sector. The initiative started working in 2013, and adopted a transparent monitoring process. To this date, IDEC has published 8 monitoring reports. Achievements in terms of transparency included the disclosure of data from the Electricity Corporation on service provision by individual circuits (this enabled users with an internet connection to check the schedules for service disruption due to network maintenance) and the online publication of the Central Government budget in a friendly format. In terms of public expenditure efficiency, the procurement and acquisition working group set the bases for the design of the transactional portal for public procurement which is now fully operational in 130 institutions. Source: Kaufmann, Gallina, and Senderowitsch (2015). 82 | Dominican Republic: Systematic Country Diagnostic Figure 50: Perception of insecurity in the neighborhoods 60 50 40 30 20 10 0 GUY URY PRY VCT CHL SLV COL BOL KNA LCA DMA USA BRA PER DR MEX GRD CAN ATG NIC JAM CRI PAN HND ARG ECU GTM HTI VEN Source: Espinal et al. 2017, using LAPOP 2016. CHAPTER 5 Prioritization and knowledge gaps Priority areas the country has been unable to afford much-need- ed investments in human capital. A fourth challenge T he Dominican Republic could sustain a stems from the degradation of the country’s natural strong economic performance in the medi- capital, mainly caused by a lack of territorial and um-term, but there are many pending key urban planning in the face of rapid urbanization, a challenges to become a high-income and more eq- lack of accompanying infrastructure and a lack of uitable society by 2030. Although growth rates for planning in the management of natural resources. the DR are projected to be around 5 percent for 2018- The aforementioned challenges are amplified by the 2020, the country faces structural challenges to foster- country’s exposure to natural disasters, the fifth chal- ing a more dynamic private sector that can create better lenge, which impacts economic growth, macroeco- jobs and ensure social inclusion. nomic stability, and household welfare. This SCD identifies five key and interrelated chal- The SCD identifies six priority areas to meet these lenges which the DR needs to overcome to achieve challenges. First, credibly improving fiscal balance by a more inclusive and sustainable growth. First, increasing revenue and improving the quality of fiscal despite high rates of economic growth, firms in the expenditure. Second, enhancing the accumulation of DR have been unable to achieve substantial quality human capital. Third, pursuing policies that create a improvements, and linkages between export-oriented level playing field for firms and a business environment firms and local firms are weak. These productive chal- that fosters investment in quality upgrading. Fourth, lenges have hampered integration in international improving the management and conservation of nat- markets and prevented more inclusive growth across ural resources. Fifth, improving the country’s resilience firms. Second, the country has not generated suffi- to disasters and climate-related risks. Finally, increas- cient tax revenues to reach higher levels and quality ing both transparency and accountability in the policy of public services, and as a result it is jeopardizing its making process. fiscal sustainability. Third, due to a lack of resources, 84 | Dominican Republic: Systematic Country Diagnostic These priorities were identified based on the follow- Improving fiscal balance by developing ing criteria. i) A benchmarking exercise that compares a credible fiscal path that supports the the DR’s performance with the performance of coun- provision of high-quality public services tries in LAC and the group of upper-middle income countries (UMI). The indicators used for the compari- Growing government expenditure combined with a son cover a wide range of topics including growth, fiscal low revenue base, has resulted in fiscal deficits and accounts, business environment, competitiveness, hu- a growing debt trajectory. Government expenditure man development, natural resources, and governance. rose by 5 percentage points as a GDP share between The DR’s performance in each indicator is compared 2000 and 2017, from close to 13 percent to 18 per- with the best performers in each comparison group cent of GDP. Fiscal revenue, in contrast, has hovered (LAC and UMI - see Annex 2); ii) validation of the around 13 and 15 percent of GDP throughout that findings using existing literature and new analysis con- period. The combination of both these patterns has ducted for this SCD; iii) the capacity of a priority to led to a rapid increase of government debt and interest contribute to progress in other areas; and iv) a robust payments. consultation process with Bank staff and in-country stakeholders (see Annex 7). A narrow fiscal space and increasing debt dynamics limit the country’s ability to invest in priority pro- The benchmarking exercise highlighted a set of is- grams, and erode the private sector’s competitive- sues where the country’s underperformance sets it ness. These problems are compounded by the exposure apart from its peers. This exercise provided a first long to natural disasters, which impacts fiscal accounts by list of issues, that was refined using the diagnostics in lowering tax collection and triggering emergency expen- the chapters to produce a shorter list of priorities. This ditures, and calls for fiscal buffers to respond whenever short list was used during the consultation process with shocks materialize. Raising revenues and delineating a more than 200 participants in country from a wide credible medium-term fiscal path are crucial elements range of stakeholders including government, civil soci- to opening up fiscal space and ensuring long-term fiscal ety, academia, private sector, and development partners. sustainability in the DR. At every meeting during the consultation process we asked participants to pick a few areas they considered Tax exemptions and informality have been import- the most critical to address the country’s development ant factors behind the DR’s relatively low tax rev- challenges. The vast majority of priorities selected by enues. There are currently 11 special tax regimes in participants are embedded in the six priorities identified the DR, which in addition to creating an unleveled in the SCD. playing field, have considerable fiscal implications. In- deed, the foregone revenue from tax exemptions in the The impact of implementing these priorities will DR amounted to 6.6 percent of GDP in the 2014- increase to the extent that they are tackled in tan- 2016 period (World Bank, 2017). Moreover, some dem and that they incorporate the linkages between exemptions, such as VAT exemptions, are found to be the different challenges. For example, direct efforts to regressive (Cabrera et al., 2016). Similarly, labor infor- raise fiscal revenues will probably have a limited effect if mality costs the Dominican government an estimated they are not accompanied by explicit efforts to improve 1 percent of the annual budget in foregone revenue. the provision of public goods and to fight corruption. The limitations of the tax system in the DR in produc- Similarly, the impact of initiatives to foster innovation ing higher revenues and the importance of addressing or better processes of production will be muted to the these limitations are recognized in the DR’s National extent that they do not address market power and oth- Development Strategy 2030, which outlines a plan to er disincentives to innovation. Moreover, reforms will consolidate all existing tax expenditure schemes into have to take into account the political economy con- a single section of the tax code, and to establish a co- straints that they could face. herent and sustainable approach to tax expenditures Dominican Republic: Systematic Country Diagnostic | 85 which will reduce their fiscal impact and minimize country, with only 42 percent of children (ages 4-5 their distortive effect on economic incentives. years) in the poorest quintile attending pre-primary school, compared to 62 percent in the richest quintile. In a context of tight fiscal space, the country needs According to diagnostics of the sector, some of the entry to focus on mechanisms to improve the quality of points to improving skills acquisition and learning out- public services and, in some cases, reduce its fiscal comes include early childhood development, improve footprint. The DR has increased its expenditure in key teacher skills and recruitment, curricular reform, and areas such as education and health. However, as dis- improving school management. The country should cussed in chapter 3, the quality of public expenditure also keep on strengthening the link between industry (measured by outcomes) in both these areas appears to and school/university by improving technical and vo- be low. Similarly, the public systems of electricity distri- cational education and expanding on-the-job training bution and water and sanitation provide an unreliable opportunities. This would reduce the skills mismatch service of low quality, and have high operational costs. the private sector reports as an important barrier to do- ing business. Developing a credible medium-term fiscal plan could anchor expectations, reduce interest pay- Improving health care financing and efficiency of ex- ments, and enable the country to build buffers. The penditures to ensure quality of health care provision DR needs to strengthen efforts to contain the mounting and equitable access to health services. At 1.7 per- government debt and its interest payments. To achieve cent of GDP in 2016, the health public expenditures this objective, the country must design a credible and as a share of GDP ranks below the country’s 10-year clearly defined fiscal plan that stabilizes debt levels and health plan target of 4 percent, and is one of the low- anchors market expectations. Indeed, a credible fiscal est in LAC. Sector budgeting and resource allocation anchor could lead to lower borrowing costs, which in have mainly been carried out following previous allo- turn could reduce the potential economic impacts of cations and spending is heavily skewed toward curative fiscal adjustments (see IMF 2017 and Vegh et al. 2018). care and hospitals compared to primary health care and Beyond aiming at stabilizing debt levels, the fiscal an- public health services. The sector also exhibits inade- chor, should also seek to build fiscal buffers to respond quate accountability mechanisms which, along with the to economic and natural shocks without affecting the Ministry of Public Health’s limited capacity to moni- trajectory of public debt. tor, regulate and enforce, contribute to inefficiencies in spending. Improving the quality of health services Enhancing the accumulation of human can save lives, improve the quality of life, and increase capital workers’ productivity. Not all necessary quality im- provements require additional resources. For instance, Improving skills acquisition and learning outcomes. the majority of maternal deaths, too many for the coun- Despite modest progress in regional studies, the coun- try’s coverage of institutional birth and antenatal care, try ranks last in learning outcomes, by far, in both can be prevented with properly trained personnel ad- PISA and TERCE. Cognitive skills and the quality of hering to norms, and following protocols and quality learning outcomes have potentially large implications standards.177 for economic growth as they bolster innovation, and ultimately result in higher productivity gains.176 Early Enabling human capital accumulation and strength- childhood education would improve the youth’s transi- ening risk protection for the poor and vulnerable tion from school to work and provide them with assets through social protection. Sustaining the recent gains to be more competitive in the labor market. However, on poverty reduction and breaking the cycle of pover- access to early childhood education is unequal in the ty and deprivation will require further efforts to build 176  Hanushek and Woessmann (2010). 177  Castro (2016). 86 | Dominican Republic: Systematic Country Diagnostic the human capital of poor and vulnerable group, and ment that fosters competitiveness, notably via enhanced to protect poor households from either economic risks competition, the promotion of quality standards, and or disasters. This can be achieved by further improv- improved access to finance. ing the coverage, targeting and adequacy of conditional cash transfers, connecting poor households to produc- The country stands out in LAC and worldwide in tive inclusion initiatives, and preparing social protec- terms of perceived market dominance. The 2017- tion system and instruments for disaster preparedness 2018 Global Competitiveness Report ranks the country and response. The country should continue its efforts 7th in the world in terms of market dominance, making to provide identification documents to all citizens to re- it the second highest in LAC after Haiti. Moreover, a move barriers of access to social programs, financial ser- comparison over time shows that in the past 10 years vices, political participation, and secondary education. perceptions about market dominance in the country have worsened—the indicator decreased from 2.8 in the Addressing gender imbalances. Women in the country 2007-2008 report to 2.6 in the 2017-2018 one. Market not only face high rates of violence and adolescent fertil- dominance may be one of the reasons behind the DR’s ity, but also lower labor force participation, much higher limited quality upgrading (Lederman et al. 2014). The unemployment, and high employment segregation across full implementation of the DR’s pro-competition law is sector and lower wages. Higher female labor force par- a step towards generating a more competitive business ticipation is associated with lower infant mortality rates environment. However, further steps need to be taken and higher life expectancy. It also has positive effects on to remove distortions, such as selective tax exemptions, children’s development and well-being, increasing their to generate market power. future earnings capability, which may reduce the inter- generational transmission of poverty.178 The analysis pre- Reducing progressively the regulatory divide be- sented in this SCD suggests that gender equality gains tween SEZ and non-SEZ exporters is a complemen- that increase labor market opportunities for women and tary direction to foster competition and promote empower adolescent girls may be central to reducing competitiveness by alleviating distortions and pro- poverty, addressing disadvantages faced by female-head- moting economy-wide competitiveness. The con- ed households. Improving women’s sexual and reproduc- tinued focus of the SEZ regime on specific industries, tive health and rights to achieve better maternal health especially textiles, footwear and leather goods, could en- outcomes and lower adolescent fertility rates would trench economic distortions and perpetuate the coun- also produce better infant and child health outcomes try’s dual export structure. The country already phased which would in turn improve the returns to the early out export performance-based subsidies in SEZs, which childhood development programs mentioned above. is a positive first step. Over the medium-to-long term, harmonizing the regulatory system will require a pro- Promoting a level playing field and a gressive reduction of the number of “strategic sectors” business environment that fosters quality and their eventual elimination. Overall, exporters out- improvements and competitiveness side and inside SEZs should be receiving support from the Export Promotion Agency in the form of commer- Having a level and well-functioning business envi- cial actions, technical assistance, and the identification ronment, where resources are efficiently allocated of export opportunities. and firms have clear incentives to innovate and in- vest, is crucial to a country’s ability to generate in- Policies aimed at homogenizing regulations, coupled clusive growth. Thus, to break with the cycle of low with explicit policies, could foster stronger linkag- competitiveness and low levels of linkages between ex- es between fast-growing, export-oriented firms, and port-oriented firms and domestic firms the DR needs domestic firms. The DR’s large presence of foreign to promote a level playing field and a business environ- firms and, more generally, export-oriented firms cre- 178  Mateo Díaz and Rodríguez-Chamussy (2016). Dominican Republic: Systematic Country Diagnostic | 87 ates a great growth potential which can only be seized, interest rates have been on the rise since 2010. The aver- however, to the extent that domestic firms can connect age real lending rate charged in the DR stood at around to, and learn from these firms. To fully reap the growth 14 percent in 2016, more than 6 percentage points benefits of FDI attraction and global integration, the higher than the regional average. Moreover, the real DR could promote policies that foster linkages between lending rate in the DR has been on the rise since 2010 export-oriented firms and domestic firms. Achieving while the LAC average has hovered around its average. this goal would require creating an appropriate envi- ronment that leads domestic firms to meet the quality Lowering financing costs could help deepen the DR’s and reliability standards required by foreign firms. In relatively shallow financial markets. High real interest addition, a comprehensive policy intervention aimed at rates may be one factor behind the country’s relatively reducing search costs and tackling constraints of both low credit volumes. The total credit to private sector buyers and sellers produces more effective results than and total deposits penetration relative to GDP is one of a piecemeal approach. Pursuing such a policy interven- the lowest of all LAC countries. Total credit to the pri- tion would require a regulatory and institutional revi- vate sector in the DR amounted to 28 percent of GDP, sion of the support framework to encourage consistency the third lowest value in LAC. Moreover, commercial and coordination across public and private actors in- loans to SMEs increased from around 12 percent of volved in promoting SEZs. GDP in 2007 to 17 percent in 2015, but average loan amounts remain low and maturities are typically short. Strengthening regulatory and institutional capacity Capital markets are also shallow, with stock market can help improve the quality of exports and domestic capitalization to GDP being one of the lowest in the inputs, which currently lag behind comparators. As region. They are characterized by a crowding-out effect argued in World Bank (2015), there is a need to create between public/private sectors, and financing for large standardization bodies, system certifications, conformi- businesses provided mainly by banks. Equity shares are ty assessments, and accreditation mechanisms to create not publicly traded, and there are only around 20 cor- better incentives for quality improvements, and to get porate debt issuers in the local stock exchange “Bolsa a better flow of information on standards to exporters. de Valores de la República Dominicana” (BVRD). The Deeper commitment to international SPS standard set- government is the main issuer of debt instruments, and ting institutions, as well as to the WTO SPS committee, public debt accounts for the bulk of trades. Returns on could ensure that the Dominican support infrastructure these instruments are higher than commercial rates, and conforms to international norms and prompts regula- will deter inter-bank-rate competition and corporate tion updates on SPS and pesticides. Existing institu- debt issuance. The fact that the government issues long- tions in charge of the surveillance of animal health and term debt through two different entities, the Central control of plant pests could also benefit from capacity Bank and the Ministry of Finance, with a substantial building. In addition, vaccination campaigns and a wid- rate spread for the same underlying risk and tenor, will er dissemination of risk-mitigation methodologies could continue to send mixed signals to the market and un- be organized. This could help, for example, reduce the dermine price discovery. DR’s underperformance in terms of agricultural export rejections (see Chapter 2), and create opportunities for Improving the quality of public services, especial- the tourism sector by reducing health risks associated ly electricity, can help lower the costs of firms and with the consumption of domestic products. boost their competitiveness. As an island state highly dependent on trade, good infrastructure is critical for Efforts to reduce financing costs could complement successfully creating functional domestic supply chains, efforts to boost competition in facilitating competi- to increase productivity and competitiveness, and to tiveness-enhancing investments by firms. Dominican connect domestic SMEs into exports and global value firms face higher real interest rates than firms in most chains. Infrastructure has improved in the DR but chal- LAC countries or from comparator countries; and real lenges remain. These challenges are most visibly seen in 88 | Dominican Republic: Systematic Country Diagnostic an electricity sector characterized by constant interrup- water exploitation particularly in more arid parts of the tions and high tariffs. The economic impacts of the elec- country. Freshwater resources’ stewardship and sustain- tricity sector’s problems in the DR are considerable. On ability through better planning and allocation is critical average, Dominican firms lose 5 percent of their sales to meet the growing competing demands from tourism, due to power outages, the second largest value in LAC. municipalities, industry, and the agriculture sector. To minimize the costs of outages, businesses must make large investments in back-up equipment. Moreover, the Improving resilience to disasters and economic costs associated with problems in the electric- climate-related risks ity sector are disproportionately larger for small firms. Improving the country’s resilience to disaster and cli- Improving the management of natural mate-related risks would have positive effects on eco- resources nomic growth, fiscal sustainability, and household welfare. Accelerating the design and implementation of The improving efficiency of water and sanitation a strategy for managing disaster and climate-relate risks providers must be improved to enhance service qual- which is well integrated with the overall management ity, and reduce the degradation of water resources. of fiscal risks, would not only contribute to fiscal and Creating institutional and policy incentives for pro- macroeconomic sustainability, but improve the gov- viders to improve efficiency such as increasing meter- ernment’s fiscal capacity to quickly respond and recov- ing, billing and collection, and reducing non-revenue er quickly in the aftermath of a large event. Reducing water would not only improve service quality, but also the physical vulnerability of public infrastructure and increase commercial viability (reducing dependence on improving technical standards to ensure safety the re- central government subsidies) and reduce the degrada- quirements of housing and private construction will tion of water resources. Both increasing access to private ensure the business continuity of essential public ser- sector financing and reducing subsidies will be required vices as well as the sustainability of public spending and to finance critical investments to, on one hand, close investments. Developing targeted social programs for regional disparities in access to water supply and san- poor and vulnerable households to increase resilience itation infrastructure (contributing to human capital), to disaster and climate-related risks will contribute to and on the other hand, reduce water resource degrada- sustain the gains in poverty reduction. Territorial and tion stemming from urbanization through increased in- urban planning will play a key role in increasing resil- vestments in sanitation and wastewater reuse. Slowing ience by ensuring the safe location of population and by down watershed degradation will contribute to increase managing population relocation issues when required, the country’s resilience to climate change, reduce the by building disaster and climate-risk informed territori- spread of gastrointestinal and respiratory diseases, and al and land use instruments, and through a comprehen- prevent negative effects on touristic activity. sive management of natural resources. This is aligned with the urgent need identified in the NDCs to build Strengthening integrated water resources manage- adaptive capacity among vulnerable groups, especially ment (IWM) to ensure the stewardship of freshwater women and youth.180 resources to meet growing demands.179 The DR must keep on building on recent efforts in basin level plan- Increasing transparency and ning for municipal, industrial, and agricultural water accountability in the policy making consumption. This implies continued focus on better process coordination between the water and agricultural sectors to create conditions for water reuse for industrial and The government should increase efforts to improve agricultural activities with the aim to also reduce fresh- transparency and accountability, and gradually en- 179  This is one of the elements of the strategic planning approach to climate change adaptation as articulated in NDC. 180  The NDCs also acknowledge the significance of women’s participation in the transformation of society towards a low-carbon and resilient development. Dominican Republic: Systematic Country Diagnostic | 89 TABLE 4. Priorities and their impact on the Dominican Republic’s development challenges Priority Area Growth Inclusion and human Sustainability challenges development challenges challenges Improving fiscal balance Indirect effect Indirect effect Direct effect Enhancing the accumulation of human capital Indirect effect Direct effect Indirect effect Promoting a level playing field and a better Direct effect Indirect effect Indirect effect business environment Improving the management of natural resources Indirect effect Indirect effect Direct effect Improving resilience to disasters and climate- Indirect effect Indirect effect Direct effect related risks Increasing transparency and accountability in the Indirect effect Indirect effect Indirect effect policy making process hance the effectiveness of coalitions for social dialogue performance and hold it accountable, and for the pri- such as the Pacto Eléctrico and the forthcoming Pacto vate sector to identify business opportunities. Fiscal. They will also gradually help level the playing field for the private sector. This will require strength- ening both supply of and demand for information. Knowledge gaps On the supply side, the implementation of the access to public information framework needs to be strength- This SCD identifies data/knowledge gaps as they ened, as well as capacities for opening government data. relate to critical issues relevant to the achievement On the demand side, it will be important to support ef- of the twin goals and other development goals, and forts by civil society, media and academia to effectively proposed further research. use available information to monitor the government’s Topic Knowledge/data gaps and further research Agriculture Lack of recent data on productivity for subsistence and small farmers, land ownership, access to finance and technology. The last census of agriculture was collected in 1982. Gender The last comprehensive assessment of gender issues in the country by the WB was done in 2002. Migration Lack of information on immigration flows, a challenge to identify the impacts of immigration on labor markets. Social protection and labor Lack of a rigorous assessment of the impacts of social programs on poverty and livelihoods. Jobs diagnostic with analysis of labor demand. Analysis of the effect of labor market regulations on job creation, informal employment, and unemployment. Analysis of the underlying causes of low female labor force participation and high sectoral segregation by sex. 90 | Dominican Republic: Systematic Country Diagnostic Topic Knowledge/data gaps and further research Transport Characterize the logistics costs and their determinants. This includes the role of logistics costs in the productivity of different sectors, as well as the characterization of the trucking sector regulatory framework and performance in relation with domestic transport costs, connectivity to ports and airports, and CO2 emissions. Producing a reliable national GHG inventory in transport sector will play a vital role in supporting the country’s NDC implementation as transport is one of the priority sectors for reaching its conditional economy-wide mitigation target (25 percent reduction in GHG emission by 2030 compared to 2010 baseline). Urban accessibility analysis. Urban Urbanization review covering topics like spatial patterns of urbanization and determinants of spatial growth patterns; implications of the existing spatial distribution; dynamics of agglomeration economies and productivity growth in urban areas; land use and territorial development. Environment Information on greenhouse gas emissions in the agriculture sector to produce reliable national inventories of GHG is lacking. A national inventory has started in a small area of the country which needs to be scaled up and work towards Tier 3 information system. Establishing GHG inventories for agriculture as well as land use, land use change and forestry sectors will significantly contribute to supporting the country’s NDC conditional economy-wide mitigation target. Lack of reliable data and knowledge on air quality (emissions) and water effluents. Water Better understand the fiscal impacts of recent Government’s reductions of subsidies and transfers to CORAAS (water and sanitation providers). Making CORAAS commercially viable can spur water and sanitation investments. With better evidence about the impacts of subsidy reduction to CORAAS, the country could explore better targeting mechanisms to apply other subsidies in areas with lowest coverage and highest poverty rates with the aim to improve efficiency of public spending in reaching the poorest. Factor allocation across firms Lack of any rigorous assessment of factor allocation across firms. Hsie and Klenow (2008) and Pages et al. (2010) show that factor misallocation is one of the key factors explaining low TFP in developing countries, including LAC countries. Thus, efforts to collect data and assess factor allocation across firms in the DR would be a crucial input to understand the country’s growth prospects. Cost and benefits of FDI and SEZ Lack of rigorous analysis of the costs and benefits of FDI and SEZs. Given the relatively large presence of foreign firms in the DR, it would desirable to assess the economic impact of these firms in the local economy. Unfortunately, the country does not have the available data to tackle these empirical exercises. For example, they would require labor flows in and out of foreign firms (see Poole 2013) and TFP estimates of all firms in the DR. Trade Lack of an updated study of the impacts of CAFTA-DR on the Dominican economy. Jaramillo and Lederman (2006) provided an initial assessment of the potential trade and labor market impacts of CAFTA-DR on Central American economies and the DR at the time when the agreement was signed. Assessing these impacts is particularly important for an economy with such a substantial trade dependence with the US and with labor markets which do not appear to deliver social outcomes with sufficient vigor. Competition policy Lack of comprehensive, international comparable, data on concentration at the sectoral level. Dominican Republic: Systematic Country Diagnostic | 91 ANNEX 1 Economic and Social indicators TABLE A1. Dominican Republic - Basic Statistics 2002 2010 2016 Population, million 8,8 9,9 10,6 GDP, current US$ billion 26,6 54,0 71,6 Real GDP growth (annual %) 5,8 8,3 6,6 GDP per capita, current US$ 3008 5451 6722 International poverty rate ($1.9 day/2011 PPP) 6,0 2,6 1,6 Lower middle-income poverty rate ($3.2 day/2011 PPP) 14,6 12,4 6,2 Upper middle-income poverty rate ($5.5 day/2011 PPP) 31,6 32,5 21,0 Inflation, consumer prices (annual %) 5,2 6,3 1,6 Gini coefficient 50,0 46,9 45,3 Life expectancy at birth, years 71,0 72,7 73,9 Sources: World Bank, WDI. TABLE A1. Dominican Republic - Basic Statistics 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Real Sector* Agriculture (percent of GDP) 6,3 6,9 6,8 7,4 6,9 6,8 6,7 6,5 6,1 6,7 6,7 6,9 6,9 6,8 6,5 6,1 6,2 Industry (percent of GDP) 31,2 29,9 30,2 30,1 32,2 31,4 29,8 28,6 28,1 26,4 26,8 26,7 25,7 27,0 27,5 28,0 28,3 Services (percent of GDP) 55,8 57,8 57,9 59,0 58,7 58,1 56,9 56,9 58,0 58,8 58,4 58,3 59,1 58,2 57,1 56,6 56,4 Total consumption 80,4 80,2 80,0 78,4 78,9 81,0 81,0 80,1 81,6 85,3 83,8 84,2 84,0 82,2 79,3 78,0 76,7 (percent of GDP) Investment (percent of GDP) 25,4 26,6 25,3 19,1 22,5 24,8 24,8 24,1 28,4 22,1 26,4 23,3 23,4 22,0 23,8 26,5 26,5 Exports (percent of GDP) 34,7 31,4 30,3 35,1 35,7 31,5 28,4 26,8 24,7 22,4 24,0 25,0 25,7 26,7 26,7 25,4 25,4 92 | Dominican Republic: Systematic Country Diagnostic Imports (percent of GDP) 45,2 41,5 39,8 36,3 37,5 37,2 36,1 35,2 36,0 31,9 34,9 33,9 33,5 31,2 30,2 31,4 30,8 Fiscal Accounts Consolidated public sector debt 18,2 19,8 21,9 39,5 46,9 32,5 34,3 31,2 32,0 36,4 36,6 37,8 41,3 45,1 44,3 44,5 47,1 Non-Financial Public Sector -1,6 -2,1 -1,1 -4,8 -3,0 -0,3 -0,7 0,1 -3,6 -3,1 -2,8 -3,1 -6,9 -4,0 -3,6 -0,3 -2,8 Balance Monetary Sector Consumer price inflation 9,0 4,4 10,5 42,7 28,7 7,4 5,0 8,9 4,5 5,8 6,2 7,8 3,9 3,9 1,6 2,3 1,7 (end of period) Exchange rate RD$/US$ 16,2 16,7 17,5 29,4 41,4 30,2 33,2 33,1 34,5 35,9 36,8 38,1 39,3 41,8 43,5 45,0 46,0 (end of period) Monetary policy interest rate n/a n/a n/a n/a 7,0 10,0 8,0 7,0 9,5 4,0 5,0 6,8 5,0 6,3 6,3 5,0 5,5 (end of period) Notes: n/a Not available. *The non-additivity of the chain-linked indices used leads to the fact that the sum of each component of GDP may not add up to 100 percent for the period. Sources: Central Bank of Dominican Republic. Dominican Republic: Systematic Country Diagnostic | 93 ANNEX 2 Benchmarking the Dominican Republic B orrowing from the SCDs for Colombia and The relative distance is calculated using the formula: Chile, a large set of indicators was used to compare the DR’s performance with respect to other countries in two reference groups: the LAC region and the upper-middle income countries. For each in- dicator the average value in the 2012-2016 period was used to reduce the potential outlier effect of atypical Where s ∈ {LAC,UMI} denotes the reference group, years. i is the particular indicator for which the distance is being calculated, b(i,s) denotes the “best” performer We used a metric for relative distance to assess the coun- in indicator i among countries in s, w(i,s) denotes the try’s relative performance in each indicator. The relative “worst” performer in indicator i among countries in s, distance is calculated by dividing the distance between and V(i,c) denotes the value of indicator i in country c. the DR and the “best” performer in the reference group, by the distance between “best” and “worst” perform- Note that this definition of relative distance does not ers in the group. The “best” performer is defined as the use normative considerations about the optimal value country in the 95th percentile of the indicator, and the for each indicator but uses instead the best achiever in “worst” performer as the country in the 5th percentile. the reference group as the benchmark. However, for some of the indicators a higher value de- notes better performance (e.g., coverage of social insur- We organize the results in 4 groups defined by ranges of ance programs) while for others a higher value denotes relative distance to the “best” performer as follows: val- worse performance (e.g., maternal mortality rates). In ues of relative distance below 25 percent (i.e., the DR’s the latter cases, the best performer is defined as the coun- performance is a short distance away from the “best” try with the lowest value, and the worst performer as performer) are assigned low priority (1 ), those be- the country with the highest value of the indicator. This tween 25 and 50 percent are assigned medium-low pri- explains the use of absolute value in the formula below. ority (2 ), those between 50 and 75 percent are con- 94 | Dominican Republic: Systematic Country Diagnostic sidered medium-high priority (3 ), and finally those peers. The structural peers include countries that share higher than 75 percent are considered high priority (4 economic traits with the DR, such as their current ac- ). count deficit, educational attainment, current GDP levels, GNI per capita, labor force with basic education, Table A5 below presents the results in descending order total population, and poverty levels. Based on these cri- by level of priority. teria, the group of structural peers includes Sri Lanka, Tunisia, Ecuador, Peru, Guatemala, Honduras, Roma- To benchmark Dominican Republic’s performance, nia. Fiji is added as it is an island country in the Upper this report uses five groups of peers: CAFTA countries, Middle-Income group that also issues sovereign bonds. Latin American peers, upper middle-income countries, The group of structural peers provides an appropriate countries exposed to natural disasters, and structural benchmark for answering macro-relevant questions. TABLE A5. Rankings according to distance to “best” performer Priority Area UMI LAC 1st pillar: Institutions: 1.17 Ethical behavior of firms 1st pillar: Institutions: 2. Ethics and corruption 1st pillar: Institutions: 3. Undue Influence 2nd pillar: Infrastructure: 2.07 Quality of electricity supply 5th pillar: Higher Education and training: Quality of education 6th pillar: Good market efficiency: 6.03 Effectiveness of anti-monopoly policy Adequacy of social safety net programs (% of total welfare of beneficiary households) Coverage of social insurance programs (% of population) Benefit incidence of social safety net programs to poorest quintile (% of total safety net benefits) Adolescent fertility rate (births per 1,000 women ages 15-19) Mortality rate, neonatal (per 1,000 live births) Children in employment, wage workers (% of children in employment, ages 7-14) General government revenue (% of GDP) Health expenditure, total (% of GDP) International migrant stock (% of population) Tax revenue (% of GDP) Domestic credit to private sector (% of GDP) 5th pillar: Higher Education and training: On-the-job training School enrollment, secondary (% gross) 6th pillar: Good market efficiency: 6.02 Extent of market dominance 8th pillar: Financial market development: 8.05 Venture capital availability Informal employment (% of total non-agricultural employment) Subsidies and other transfers (% of expense) Coverage of social safety net programs (% of population) Mortality rate, infant (per 1,000 live births) Mortality rate, under-5 (per 1,000 live births) Dominican Republic: Systematic Country Diagnostic | 95 TABLE A5. Continued Priority Area UMI LAC Self-employed, total (% of total employment) Vulnerable employment, total (% of total employment) Wage and salaried workers, total (% of total employment) Change in General government gross debt 2007 -2016 (% of GDP) Foreign direct investment, net inflows (% of GDP) Gross fixed capital formation (% of GDP) Real interest rate (%) 3rd pillar: Macroeconomic Environment: 3.05 Country Credit Rating 8th pillar: Financial market development 11th pillar: Business sophistication: 11.05 Value chain breadth 9th pillar: Technological readiness 11th pillar: Business sophistication: 11.02 Local supplier quality 6th pillar: Good market efficiency: 6.04 Effect of taxation on incentives to invest 7th pillar: Labor market efficiency: 7.06 Pay and productivity 7th pillar: Labor market efficiency: 7.08 Country capacity to retain talent Age dependency ratio (% of working-age population) Improved water source (% of population with access) Primary completion rate, total (% of relevant age group) School enrollment, primary (% gross) School enrollment, tertiary (% gross) Improved water source (% of population with access) Primary completion rate, total (% of relevant age group) School enrollment, primary (% gross) School enrollment, tertiary (% gross) Electricity production from coal sources (% of total) Prevalence of underweight, weight for age (% of children under 5) Children out of school, primary 9th pillar: Technological readiness: 9.02 Firm-level technology absorption 1st pillar: Institutions: 4. Public-sector performance 8th pillar: Financial market development: 8.03 Financing through local equity market 11th pillar: Business sophistication: 11.03 State of cluster development 2nd pillar: Infrastructure: Transport Infrastructure 6th pillar: Good market efficiency: 1. Domestic competition 6th pillar: Good market efficiency: 2. Foreign competition 7th pillar: Labor market efficiency: 7.02 Flexibility of wage determination 7th pillar: Labor market efficiency: 7.09 Country capacity to attract talent 8th pillar: Financial market development: 8.04 Ease of access to loans 96 | Dominican Republic: Systematic Country Diagnostic TABLE A5. Continued Priority Area UMI LAC 8th pillar: Financial market development: 8.06 Soundness of banks Distance to frontier score (0=lowest performance to 100=frontier) Share of youth not in education, employment or training, total (% of youth population) Employment in services (% of total employment) Labor force participation rate, total (% of total population ages 15+) (modeled ILO estimate) Net migration Prevalence of HIV, total (% of population ages 15-49) Urban population (% of total) Adolescents out of school (% of lower secondary school age) GDP growth (annual %) Progression to secondary school (%) 6th pillar: Good market efficiency: 6.07 Time to start a business Cost to export (US$ per container) Fossil fuel energy consumption (% of total) Prevalence of overweight, weight for height (% of children under 5) Unemployment, total (% of total labor force) (modeled ILO estimate) Dominican Republic: Systematic Country Diagnostic | 97 ANNEX 3 Monetary policy in the DR F ollowing a period of high inflation in the The country’s below-target inflation is partly due aftermath of the 2003 domestic banking cri- to the Central Bank’s interventions to keep depre- sis, the country started a transition towards ciations at bay. Out of the 38 countries with inflation an inflation targeting (IT) regime in 2005, a regime targeting frameworks in the world, the DR is one of that was fully adopted in 2012. Inflation averaged 7 the two countries that doesn’t have a floating exchange percent in the between 1992 and 2002, and sharply in- rate regime (IMF 2016). The country has crawl-like ar- creased to 27 percent and 51 percent in 2003 and 2004, rangement (IMF 2016), that seeks to manage the vol- respectively. Following this, the Dominican govern- atility of the exchange rate around its long-term path. ment initiated the technical steps to adopt an inflation targeting regime with the advice of the IMF. In 2012 The choice of the exchange rate regime is partly mo- the Dominican Central Bank fully adopted an inflation tivated by a relatively high-pass through and by the targeting regime. composition of government debt. The country’s crawl- like arrangement minimizes the potential pass-through Since the adoption of the IT regime, the country from exchange rate depreciations (the process by which has been successful in maintaining the inflation rate exchange rate depreciations cause prices to increase in low, although it has usually stayed below the Central terms of local currency, thus causing inflation), which Bank’s target. In 2012 the Central Bank established a in the DR is still high.181 The downside to the country’s timeline to gradually reduce the inflation target range to- exchange rate stability is that it limits the role of the ex- wards its long-run goal of 4 percent (+/- 1 percent). The change rate as an external shock absorber. The composi- target was set at 5.5 percent (+/- 1 percent) in 2012 and tion of the public debt may create additional incentives the Central Bank decreased the target by 0.5 percentage to avoid sharp depreciations. The public external debt points per year. However, except for 2013, inflation has has increased in recent years in absolute terms and as a been consistently below the Central Bank’s target range. share of GDP. Moreover, interest payments have been 181  IMF 2016 Annual Report on Exchange Arrangements and Exchange Restrictions (AREAER), Dominican Republic country report. 98 | Dominican Republic: Systematic Country Diagnostic rising in tandem with external debt. Against this back- drop, exchange rate depreciations can lead to increases in the value of debt payments expressed in local curren- cy. It is worth pointing, however, that existing evidence suggests that the DR’s real exchange rates appear to be close to its long-term equilibrium level (Cruz-Rodri- guez 2016). Dominican Republic: Systematic Country Diagnostic | 99 ANNEX 4 Characterizing Dominican exports E xplicit policy efforts, geography, and natu- of total exports in the average South American coun- ral endowments, have bolstered Dominican try and 6 percent in Mexico. The DR, at 46 percent, exports over the past 25 years. Dominican stands closer to the Central American average (36 per- exports have grown at an average annual rate of 4.7 cent), and looks very much like countries as Panama or percent over the period. Exports are evenly distributed Costa Rica (Figure A51). The DR’s export of services between exports of goods and exports of services. The are mainly tied to tourism and travel (more than 80 dynamism of SEZs, geographic and cultural proximity percent of service exports according to 2013 UN esti- to the US and Canada, and the attractiveness of being mates). The observed differences between the DR and a relatively large Caribbean country have been factors other Caribbean countries may be explained by differ- favoring the country’s export growth. ences in size and labor costs, factors that can affect a country’s comparative advantage in producing manu- The composition of the export basket in terms of the factured goods (Lederman and Lesniak 2017). Indeed, participation of goods and services is more similar compared to other Caribbean countries, the DR is rel- to Central American countries than to other Carib- atively large182 (it is one of the three largest Caribbean bean islands or South American countries. On aver- countries in area and population) and has low income age, exports of services accounted for 67 percent of total levels (its GDP per capita lies below both the Caribbe- exports in 2016 in Caribbean countries (excluding the an’s average and median). DR). This share increases to 74 percent when consider- ing exclusively Caribbean islands other than the DR. In Exports have been subject to several shocks over the contrast, South American countries and Mexico have past 25 years, but they have proven to be resilient. The export baskets that are dominated by merchandise ex- first shock was the domestic banking crisis of 2003, the ports, with exports of services accounting for 14 percent second shock was the end of the Multifiber Agreement 182  In terms of size, the DR is closer to a Central American country than a Caribbean country (except Haiti and Cuba). The Dominican Republic has an extension that is comparable to those of Central American countries and has a larger population than all Central American countries with the exception of Guatemala. 100 | Dominican Republic: Systematic Country Diagnostic Figure A51: The composition of the DR’s exports stands closer to Central American countries than it does to other Caribbean islands 1,0 Exports of s rvic s/Tot l Exports 0,8 0,6 0,4 0,2 0,0 PRY GUY URY SLV VCT CHL BOL COL SUR PER BRA DR LCA KNA DMA MEX VEN ECU TTO ARG GTM NIC HTI HND PAN CRI BLZ JAM BHS ATG GRD Source: World Bank staff’s calculation based on data from WDI, data for 2016. in 2005, and the third shock was the global financial and other natural resource based products, like rolled crisis of 2009 which originated in the US, the country’s cigars, have also gained prominence. Exports of pre- main trading partner. These three shocks contributed to cious metals and other metals accounted for 26 percent five years of export contractions. Since 2010, however, of merchandise exports in 2016, up from approximately exports recovered strongly growing at an average annual 10 percent of the country’s exports in 2003, and gold is rate of 7.3 percent (above the 5.6 average growth rate of the largest export product by value. The rising promi- GDP). Indeed, after the global financial crisis of 2009, nence of metals in the export basket is likely to continue exports of Dominican goods grew in both value and as new investments in the extractive sector materialize.183 volumes, and growth in quantities exceeded the world average (Gualier et al. 2013). Similarly, after dipping Changes in the attributes of the merchandise export following the global financial crisis, exports of services basket open opportunities for future growth. Since have grown steadily since 2010. the early 2000s the country’s export basket has tilted to- wards goods of higher technological content, and of lon- Zooming into merchandise exports, the shocks that ger quality ladders, and has moved from final goods to hit the country have led to important changes in the intermediates and capital goods (Figure A52).184 Longer composition of its export basket. The MFA, which im- quality ladders imply that the country’s export basket posed quotas to the exports of textiles from developing has now more scope for quality upgrading compared to countries to high-income markets, protected Domin- the past. Because of these changes, the country’s export ican exports of textiles to the US against competition basket is more technologically biased and has a longer from countries like China. Since the end of the MFA, average ladder length than most CAFTA-DR countries, the textile sector lost weight in the DR’s export basket. with exception of Costa Rica,185 but less technologically In contrast, export goods like medical equipment, met- biased than the average Caribbean country. als, and electrical appliances have gained terrain. Metals 183  There has been an uptick in FDI from Canada in recent years, which is partly explained by the boom in the mining sector in the DR, a sector with large Canadian presence. One example of the role played by Canadian firms in the Dominican mining sector is the Pueblo Viejo Dominicana Corporation/Barrick Gold, the largest foreign investment in the history of the DR. As will be discussed in more detail below, mining is the fastest growing economic sector both in terms of value added and FDI inflows. 184  Technological intensity is measured according to Lall (2000). Quality ladder estimates come from Khandelwal (2010). Export classification into intermediate, capital, final goods is from the BEC classification. Importantly, the attributes mentioned above are associated with higher growth (de la Torre et al. 2015). 185  World Bank (2015) also shows that the DR has an export basket that has become more complex over time, as measured by the EXPY index, which is associated with higher GDP growth. Dominican Republic: Systematic Country Diagnostic | 101 Figure A52: The DR’s export basket has moved towards intermediate goods of higher technological content and longer quality ladders Panel A: Share of merchandise exports that are intermediates, consumption and capital goods 80% 70% 60% 50% 40% 30% 20% 10% 0% 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 Consumption Int rm di t C pit l Panel B: Share of merchandise exports of Panel C: Average quality ladder lenght different technology intensity in the DR, by year of the DR´s export basket 100% 2,5 2010-2016 l dd r l n ht 80% 2,0 60% 1,5 40% 1,0 20% 0,5 Av r 0% 0,0 2001 2016 6 8 10 12 Prim r R sourc B s d Lo GDP p r c pit const nt 2010 USD Low t ch cont nt M dium nd Hi h t ch cont nts Panel D: Cross country comparison of the technological contet of export baskets, 2016 100 of tot l xports, 2016 80 60 40 20 P rc nt 0 SVL DR HND NIC CRI GTM Oth r C ribb n CAFTA Oth r LAC countri s Prim r Products R sourc B s d Low T ch M dium nd Hi h T ch Source: World Bank staff based on Khandelwal (2010), Lall (2000), and data from UN Comtrade. 102 | Dominican Republic: Systematic Country Diagnostic Figure A53: The DR has a lower incidence of trade on GDP compared to regional peers and benchmarks and trade is concentrated in a few trading partners Panel A: Benchmarking trade openness Panel B: Export and import market across countries, 2009-2016 concentration 90 0,35 80 0,30 70 LCA of GDP (%) 0,25 ATG 60 MEX HHI Imports 50 0,20 GTM BRB BLZ DR JAM PAN GUY CRI 40 0,15 HND SLV PRY P rc nt 30 PER CHL SUR 0,10 ARG COL ECU 20 URY BOL 10 0,05 BRA 0 0,00 LKA DR PER GTM CRI ECU SLV HND FJI TUN NIC 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 (X+M)/GDP v r 2009-2016, obs rv d Pr dict d v lu s HHI Exports Panel C: Benchmarking exports to the US 14 Lo of Tr d V lu US in millions 12 10 8 6 4 2 0 PEY GUY URY VCT BEL BOL SLV CHL COL DMA LCA KNA SUR DR PER BRA MEX GRD ATG BRB BHS JAM PAN HTI NIC TTO GTM CRI HND ARG ECU VEN Lo Tr d V lu US Lo Tr d V lu US pr dict d Panel D: Benchmarking exports to Haiti 16 14 Lo of Tr d V lu s US '000 12 10 8 6 4 2 0 BOL PRY BEL BRB URY ECU CHL JAM SUR ARG SLV HND GUY CRI BRA COL MEX PER GTM PAN DR Lo Tr d V lu US Lo Tr d V lu US pr dict d Source: World Bank’s calculations based on UN Comtrade, WDI, and IMF BOP data. Note: the Herfindahl Hirschman index of concentration in panel B is measured as the sum of the trade shares squared. The benchmarking exercise in panel A controls for log GDP, log population and the log average distance to trading partners. The benchmarks in the lower panels come from a gravity estimation of trade flows. Dominican Republic: Systematic Country Diagnostic | 103 In the specific case of SEZs, the emergence of more tance), its trade with the US and Haiti, the two largest sophisticated manufacturing processes has increased trading partners, are above the levels predicted by grav- the value addition (production stages) that is gener- ity variables (see Box A1 and Figure A53, panels C and ated in the country and opens the door for the coun- D). In the case of trade with the US, the DR shares this try to become more robustly inserted into Global pattern with most CAFTA-DR countries. The dispro- Value Chains (GVCs). The distance between the pro- portionately high levels of trade with the US and Haiti duction stages in between SEZ exports and imports has result in the country having the third most concen- increased over time (World Bank 2016). This obser- trated export basket and the fourth most concentrated vation points to the fact that more stages of produc- import basket in LAC (Figure A53, panel B). Export tion are taking place in the country, particularly in the concentration is associated with larger GDP volatility pharmaceuticals, footwear, and electrical sectors. Thus, (Lederman, Pienknagura and Rojas 2015 and references domestic value addition embedded in Dominican ex- therein). Export market concentration is mainly driv- ports has increased with the emergence of these medi- en by SEZ exporters, as they have a Herfindahl index um high-skill manufactures. This, in turn, could deepen (HHI) (a proxy for concentration) of 0.55, compared Dominican participation in GVCs, which at this point to an HHI of 0.12 for non-SEZ exporters.186 stands below most regional peers, including CAFTA- DR countries like El Salvador, Costa Rica, or Guate- In addition to export market concentration, there are mala (de la Torre et al. 2015). GVC participation could also stark differences between SEZ exports and non- foster bigger knowledge transfers and boost growth. SEZ exports along several dimensions, highlighting the duality of the export basket. The relatively good The incidence of trade flows in the economy remains performance in terms of sophistication and ingenui- low. The country ranks at the bottom of the regional ty embedded in the overall Dominican export basket distribution in terms of the incidence of trade flows on masks an important difference between the goods pro- GDP. The country’s trade share stands at 53 percent in duced inside and outside SEZs. Firms operating inside 2016, a number that is lower than any other CAFTA- SEZs tend to export goods with some level of tech- DR member except for Guatemala (45 percent). But nological transformation, whereas non-SEZ export- a simple ranking of trade openness misses important ers focus on primary and resource-based products.187 forces that shape trade flows and affect the incidence of As discussed later in the chapter, SEZs operate as en- trade on GDP, such as the size of an economy or geo- claves, which limits the potential growth benefits of the graphic remoteness. For example, Brazil, with a lower changes in the composition of exports. incidence of trade in total GDP compared to the DR, has a very large domestic market and is farther away from global growth poles. However, the DR’s underper- formance along this dimension persists even after con- trolling for factors such as GDP, population and average distance from trading partners (Figure A53, panel A). Moreover, trade is relatively concentrated along re- gional lines (the US and to a lesser extent Haiti), exposing the country to shocks stemming from its main trading partners. A similar benchmarking exer- cise as the one described above but at the bilateral level shows that, while the country has aggregate trade levels below what is predicted by gravity factors (size and dis- 186  World Bank (2015). 187 Ibid. 104 | Dominican Republic: Systematic Country Diagnostic BOX A1. Opportunities and challenges from stronger DR-Haiti trade relations The trade relationship between the Dominican Republic and its only neighboring country has always been in- tense, and it grew in an asymmetric way during the past decade. Dominican shipments to Haiti have increased from 3 percent of total exports at the beginning of the 2000s to around 16 percent in 2012, with Haiti becoming the second most important destination for Dominican exports. Dominican exports to Haiti have boomed, rising 16-fold from the decade’s beginning to 2012 (World Bank 2015). All exported products showed gains, although increases were relatively smaller in food products and metals, the most important sectors in the early 2000s. Trade integration with Haiti offers opportunities to Dominican exporters, and to the country as a whole (World Bank 2012). New Dominican exporters can learn-by-exporting in a market in which they have a geographic advan- tage due to proximity. Data from the Exporter Dynamics Dataset shows that firms that export to Haiti are more likely to survive in the market compared to firms that export similar products to other markets. They also tend to be smaller firms compared to those that export to other markets. The number of goods exported to Haiti has grown over time, contributing to export diversification. Out of the 5,300 HS6 products, the DR exported 659 to Haiti in 2006. By 2016, this number had grown to 1,877, almost 3 times the number in 2006. The pattern observed for exports to Haiti contrasts to that observed for exports to the US, where the number of products has remained relatively flat. In contrast to exports to the US, exports to Haiti are concentrated in resource-based and low-tech products, according to the classification proposed by Lall (2000). Exports to Haiti are mainly focused on food products, metals and textiles. Medium and high-tech goods comprised 10 percent of total exports to Haiti in 2016. By con- trast, exports to the US have led the transformation of the DR’s export basket, and exports of medium and high- tech goods corresponded to 34 percent of total exports to the US in 2016. To fully exploit the potential benefits of trade relations with Haiti, the countries would need to overcome long- standing political and economic tensions. Haiti occupied the DR in the 19th century. Then, in October 1937, under the command of Dictator Trujillo, the Dominican government killed thousands of Haitians in what has been coined as the “parsley massacre”.1 More recently, increased immigration enforcement and border control imposed by DR authorities have temporarily interrupted trade flows in several occasions affecting DR’s exports. 1  Estimates of the number of Haitians killed range widely from the low thousands to over 10,000. Dominican Republic: Systematic Country Diagnostic | 105 ANNEX 5 A closer look at SEZs, tourism, and agricultural products Special Economic Zones can be extended by the National Free Zones Council, the free-zone regulator. T he Dominican Republic is often considered an example of the successful implementa- Motivated by the end of the Multi-Fiber Arrangement tion of SEZs in the Western hemisphere. (MFA) in 2005 and by the signature of the CAFTA- The zones fueled economic growth during the 1980s DR agreement, SEZs in the country experienced a and 1990s and, while they experienced a sharp de- structural shift from clothing manufacturing to me- cline in export share due in part to the expiration of dium-high skill manufacturing, a shift that impacted the Multi-Fiber Agreement and stronger international employment in the sector. Historically, export compet- competition in the textile and apparel industry in 2005, itiveness in the DR has largely depended on the type exports from firms located in SEZs have recovered since of preferential market access the country receives from 2009. Exports from SEZs accounted for approximately the US. In the 1990s, the US assigned a specific import 55 percent of the country’s total exports of goods in quota of clothing and textile to the DR under the MFA, 2016. Surgical equipment, chemicals and plastics, and which created a mushrooming of maquilas operating footwear have recently emerged as the new drivers of under the special regimen. After 2005, the CAFTA- export dynamism in the zones (World Bank, 2015). DR provided broad preferences that, coupled with the country’s special tax regime, and its dynamic policy to As in many other countries, locating in SEZs attract Foreign Direct Investment, created the condi- opens the door to a variety of duty-free mecha- tions for the emergence of electrical, footwear, and med- nisms and fiscal exemptions. Exemptions include ical equipment products in SEZs. Amid these changes, full corporate tax exemptions, full exemption from employment in SEZs has been uneven. Employment in local VAT (ITBIS) or tax on assets, and exemption SEZs remained over 170,000 until 2004, when it began from any import tax, tariff or export tax. Exemptions falling and bottomed out in 2010 at 121,000 jobs. In are granted for 15 years—20 years if the company is 2016, SEZs employment reached 163,147 persons, ac- located near the border with Haiti—but the period counting for 3.7 percent of total employment. 106 | Dominican Republic: Systematic Country Diagnostic Going forward, structural change of SEZs firms will tion to employment, including jobs indirectly supported continue to create challenges and opportunities. by the industry was 15.9 percent of total employment On the one hand, the potential of SEZs to continue (678,000 jobs). In addition, visitor exports generated creating jobs at the same pace as before is likely to be approximately 39 percent of total exports in 2017. The reduced, because emerging industries are less labor in- Dominican Republic ranks 45th in the world (out of tensive (World Bank 2017). This situation may be par- 185 countries) as it relates to the relative importance of ticularly detrimental for women as they filled the bulk travel and tourism’s total contribution to GDP.188 of jobs available in the garment sector. On the other hand, the emergence of medium high-skill manufactur- Tourism in the Dominican Republic has been on a ing is correlated with higher wages as the demand for growth trend since 2010. Tourist arrivals in the coun- more skilled workers increases. More broadly, structural try by air grew by 25 percent between 2010 and 2014, change in SEZs has led to more production stages tak- adding more than one million visitors over the period.189 ing place in the country but linkages with firms outside Web-based travel service Expedia announced a growth of SEZs remain low. The more sophisticated nature of of 45 percent in travel demand to the Dominican Re- the manufacturing production taking place in the zones public in 2014 as compared to 2013.190 Punta Cana (footwear, pharmaceutical, and electrical products) en- alone saw over 50 percent growth in demand. In 2015, tails that many inputs are imported because there is not the Dominican Republic received 5.6 million tourist domestic availability or national products do not meet arrivals (non-resident air arrivals), making it the largest required standards. As of 2012, Dominican compa- of any tourism destination in the Caribbean, with the nies in SEZs were acquiring 81 percent of their inputs next largest being Cuba, with 3.5 million. Even outside abroad (World Bank 2017). the Caribbean, the Dominican Republic’s tourism per- formance, measured by tourism arrivals and spending, remains strong, with the country trailing Singapore and Tourism remaining ahead of destinations such as Seychelles, Fiji, Maldives and Costa Rica. The United States remained The country has natural endowments that give it an the country’s single largest source market, with 2 mil- advantage for the development of tourism. Its prox- lion US arrivals, followed by Canada (with 0.7 million imity to the US, its Caribbean location, its geographic arrivals) and Europe (with 1.1 million). Hotel occupan- diversity, and the relative large size and historic heritage cy rates have been rising since bottoming out in 2009 distinguish the country from other Caribbean islands (77 percent in 2017), even as the supply of hotel rooms and other LAC countries. Tourism in the Dominican Re- increased by 17 percent over the last 9 years. Many fac- public plays a significant role in the country’s economy. tors have contributed to this growth trend, including government investment in tourism-related infrastruc- Travel and tourism’s direct contribution to GDP in ture and marketing, plus private sector investment in 2017 was 5.4 percent of GDP and the sector’s to- hospitality projects. The government has set a goal of 10 tal contribution (including indirect effects through million annual visitors by 2023. backward and forward linkages) reached 17.2 percent of GDP. In 2017, travel and tourism directly supported The current competitive advantages of the Domini- more than 206,000 jobs (4.8 percent of total employ- can Republic are the larger size of the island191 and ment in the country), and the sector’s total contribu- associated ability to command very low prices of 188  World Travel Tourism and Council (2018). 189  The Dominican Republic’s cruise industry has also been growing. Cruise tourism closed the year 2015 with 574,000 passengers, excluding crew, which would be over 650,000 in total. This was a 100 percent increase compared to the previous year. Furthermore, in January 2016, when compared to the same month the previous year, there was a 53 percent increase, and in February, a 63.4 percent growth compared to the previous year. 190  http://www.godominicanrepublic.com/wp-content/uploads/2013/12/may-june-2015-media-memo-final1.pdf 191  Within the Caribbean region, with 48,445 square kilometers, the Dominican Republic is only smaller than Cuba (with 105,806 square kilometers) and is larger than Jamaica (with 11,188 square kilometers), Puerto Rico or Trinidad and Tobago. Dominican Republic: Systematic Country Diagnostic | 107 BOX A2. Global trends in the tourism sector Global trends seem to suggest a move towards the following: a. Developing local and authentic: There is a trend towards increasing demand for goods of local provenance. For example, based on a 2015 survey, only 54 percent of US international travelers shop at luxury brand stores while on vacation versus 86 percent who want to shop at stores featuring locally-made goods;1 b. Attracting Millenials: 1.8 billion out of 7 billion people worldwide belong to the Millennial generation (ages 18- 34). “Experiencing everyday life in another country” and “increasing their knowledge” are top travel motivations for Milennial travelers (WYSE Travel Confederation Milennial Traveller Report). For example, 53 percent of US Milennial travelers regularly or occasionally participate in athletic competitions on vacation (and 20 percent more would like to try);2 c. Mixing leisure and business: 94 percent of younger travelers are “more than” or “equally” likely to take a busi- ness-leisure trip in the next five years;3 d. Catering to senior travelers: World population ages 65+ will rise from 600 million in 2015 to more than 1 billion by 2030.4 In the US 66 percent of Senior Travelers define retirement as a “time to travel and explore new places.” 5 However, while they are a good market to tap into especially during off-season, these travelers travel closer to home and access for them is critical. e. Attracting adventure travelers: Adventure travel is defined as any tourist activity that includes two of the fol- lowing three components: a physical activity, a cultural exchange, or interaction and engagement with nature.6 Adventure travel has grown at a 65 percent yearly rate since 2009. The value of the global outbound adventure travel sector was more than $345 billion in 2012. Seventy one percent of US Milennials want “to participate in adventure activities that take me out of my comfort zone when on vacation.”7 Engaging with nature is one of the most popular vacation activities for US Seniors (73 percent). 1  2015 Resonance Report: Portrait of the US international Leisure Traveler. 2  2015 Resonance Report: Portrait of the US Milennial Traveler. 3  Bridgestreet Global Hospitality Bleisure Report, 2014. 4  United Nations Department of Economic and Social Affairs, World Population Prospects: 2012 Revision, June 2013. 5  2015 Resonance Report: Portrait of the US Senior Traveler 6  Adventure Travel Trade Association. 7  2015 Resonance Report: Portrait of the US Milennial Traveler. packaged tourism offerings. The main offering on the Going forward, the sector faces challenges and op- island is all-inclusive tourism, offered at exceptionally portunities that will be addressed in detail through- low prices for both Americans and Europeans. Ap- out this SCD. As mentioned earlier the sector’s main proximately 75 percent of the tourists visiting the Do- asset is the country’s natural endowments, making the minican Republic opt for the all-inclusive packages.192 preservation of the country’s beaches and ocean crucial A one-week beach travel to the Dominican Republic to the sector, a topic that will be discussed in detail in is cheaper than destinations within the Caribbean, and the sustainability chapter. The country could also find even cheaper than places such as Seychelles, Spain and ways to expand the sector to complement the all-inclu- Mexico (Cancun). Generally, destinations that could sive model by continuing to incentivize the non-all-in- compete with the Dominican Republic on price fall clusive tourism growth (e.g., community-based tour- outside the Latin America and the Caribbean region ism, adventure tourism, etc.). For example, in 2015 and are far from the Dominican Republic’s main source the construction of the port terminal “Amber Cove” in market, the United States. Puerto Plata has impacted enormously the north coast 192  https://www.traveltradecaribbean.com/tourists-dominican-republic-shorter-stays-spending 108 | Dominican Republic: Systematic Country Diagnostic economy with tourist arrivals expected to surpass 1 mil- tive quality of the agricultural export basket appears to lion visitors in 2017. Providing access, coupled with be decreasing over time, with Dominican agricultural improving the quality of tourism assets and services exports fetching lower prices than the same product ex- offered, and implementing the right marketing strat- ported by other countries. In addition, rejection rates egy, could attract higher spending tourists that would at the US border are higher for Dominican fruits and leave more money behind in the local economy, thereby vegetables than those from other CAFTA countries, due increasing spending by tourists and creating local em- to the country’s low ability to comply with US Sanitary ployment and business opportunities. Taking steps in and Phytosanitary (SPS) regulation. Evidence suggests this direction could put the country in a position to that Dominican products may be subject to more fre- continue to seize some of the opportunities that emerge quent and stringent inspections due to past compliance from changes in the global demand for tourism (Box problems (Jouanjean et al. 2012). Rejections of Do- A2). The sector’s vulnerability to external shocks (cli- minican products are mainly due to: (i) inappropriate mate change, changes in the world economy), will re- use of pesticides;193 (ii) Salmonella contamination; and quire steps to mitigate the risks posed to future growth (iii) problems with permits or registrations as well as of the sector and the wider economy. with shipments that do not comply with sanitary and phytosanitary requirements. Rejections are not limited to the US. An audit was undertaken in October 2015 Agricultural and foodstuff exports in response to continued interceptions in the European Union (EU) of consignments of plants originating in Agricultural products and foodstuffs have increased the DR, due to the presence of harmful organisms on its weight in the export basket and have the potential fruits and vegetables.194 Another audit dated February to reduce rural poverty. Dominican exports of vegeta- 2015 on food of non-animal origin for pesticides also ble products and foodstuffs increased from close to 14 mentions that the effectiveness of the system of pes- percent of merchandise exports in 2000 to 20 percent ticide authorization continues to be limited by: i) the in 201. Moreover, more than 50% of the rural popula- absence of a single, accurate, reliable, regularly updated tion is employed in the primary sector (1.2 million out and publicly available database of approved pesticides; of 2.2 million people). ii) absence of pesticide quality controls and formulation analysis; iii) insufficient improvement in laboratories; The agricultural sector also provides a barometer of and iv) lack of pre-export official and private sampling the competitiveness of Dominican exports by local and analysis to determine compliance with maximum firms outside SEZs. The agricultural sector mostly op- residue levels. erates outside tax incentive schemes, although 20 per- cent of vegetable exports come from firms in SEZs, and it is less likely to attract FDI. Moreover, Dominican exports of vegetable products and foodstuffs increased from close to 14 percent of merchandise exports in 2000 to 20 percent in 2016 These characteristics make agriculture a good reflection of the country’s competi- tive strengths and weaknesses. The quality of some of the main exports to the US stands below that of other CAFTA-DR countries, despite some recent improvements in organic vari- eties (World Bank 2015). More important, the rela- 193  The products that face the highest rate of rejection at the US border are eggplants, tamarind, bitter melon, frozen vegetables, and chilled beans. 194  http://ec.europa.eu/food/fvo/audit_reports/details.cfm?rep_id=3449 Dominican Republic: Systematic Country Diagnostic | 109 ANNEX 6 A characterization of poverty by sex, 2016 Poverty rate Share of poor in group Group Share of Share of Male Female Ratio Male Female population female Female / in group Male Total 100 50.3 27.4 30.4 1.11 47.1 52.9 Age 0 - 14 26.0 48.3 41.9 42.4 1.01 51.5 48.5 15 - 24 19.2 49.1 26.5 32.0 1.20 46.2 53.8 25 - 40 23.0 51.2 22.2 29.5 1.33 41.7 58.3 41 - 55 16.7 51.4 18.7 19.9 1.07 47.0 53.0 56 and older 15.2 52.9 19.4 22.1 1.14 43.9 56.1 Area Rural 20.5 47.9 35.6 40.4 1.14 48.9 51.1 Urban 79.5 51.0 25.2 27.9 1.11 46.4 53.6 Labor market (15 years and older) Employer 1.9 25.0 5.4 9.0 1.67 64.2 35.8 Salaried 33.5 47.2 15.2 15.5 1.02 52.3 47.7 Self-employment 20.7 25.7 21.6 19.1 0.88 76.5 23.5 Unpaid worker 0.8 56.5 27.4 26.4 0.96 44.4 55.6 Unemployed 2.2 59.8 37.9 39.9 1.05 38.9 61.1 110 | Dominican Republic: Systematic Country Diagnostic Poverty rate Share of poor in group Group Share of Share of Male Female Ratio Male Female population female Female / in group Male Education (15 years and older) Incomplete secondary 18.1 46.8 25.5 33.5 1.31 46.5 53.5 Complete secondary 18.6 52.5 16.4 23.5 1.43 38.7 61.3 Complete or incomplete tertiary 20.9 59.2 6.6 11.3 1.72 28.6 71.4 Marital status Cohabitation (married or living 35.8 50.3 23.8 24.3 1.02 49.1 50.9 together) Divorced/Separated 11.0 65.2 13.3 31.3 2.35 18.6 81.4 Widow/widower 3.9 83.6 22.6 24.0 1.06 15.6 84.4 Single 49.2 44.4 31.8 36.0 1.13 52.6 47.4 Household headship (marital status and cohabitation) Share of heads and poverty at the household level * No cohabitation (married) 4.3 (6.7) 57.6 8.1 20.3 2.50 22.7 77.3 Cohabitation (married or living 53.2 (18.0) 12.7 24.8 22.2 0.89 88.5 11.5 together) Divorced/Separated 23.4 (43.9) 70.5 8.7 31.5 3.60 10.4 89.6 Widow/widower 10.0 (22.1) 82.9 18.0 24.1 1.34 13.4 86.6 Single 9.1 (9.4) 39.0 8.3 23.2 2.80 35.8 64.2 Source: Authors’ calculations using ENFT. * Share of female heads in parentheses Dominican Republic: Systematic Country Diagnostic | 111 ANNEX 7 Consultation process T he SCD core team followed an inclusive pro- sector (industrial chambers, representative of SEZs, cess in the development of the final product representatives of the financial sector, and represen- and the elaboration of the diagnostics, that tatives of the tourism sector), civil society, academ- relied on internal discussions with World Bank, IFC ics (meeting with economists and academics from and MIGA staff, and discussions with a wide range of local universities), and development partners. stakeholders in the Dominican Republic, including • February 8, 2018: ROC Concept Note review government officials, private sector, academia, civil meeting. society, and development partners. In this annex, the • March-May 2018: Two rounds of consultations collaborative steps followed towards the preparation of with the regional front office. the draft document are detailed. • April 2018: Brainstorming meeting with GP focal points to discuss the decision meeting draft of the • July-October 2017: Initial outreach to Global Prac- storyline and identify priorities. tices to identify focal points and organize the work. • April 2018: Brainstorming meeting with Practice • October 2017: Country Team brainstorming on a Managers to discuss preliminary findings of the first draft of the storyline. SCD and proposed priorities. • November-December 2017: Country visit to pres- • May 2018: Follow-up meeting with Program Lead- ent a revised proposed storyline and identify key ers to refine storyline and priorities. challenges for the Dominican Republic. The week- • May 30, 2018: ROC Decision Meeting. long mission included more than 20 meetings • June 2018: Submission to Board of Directors. with more than 200 participants from government • September 2018 (planned): Dissemination of SCD (Ministry of Economy, Development and Planning, in country and launch of the CPF dialogue. Ministry of Finance, Central Bank, Ministry of In- dustry, Trade and Production, the National Direc- torate for Tax Collection (DGII), and the National Council for Competitiveness (CNC)), the private Dominican Republic: Systematic Country Diagnostic | 113 ANNEX 8 Map of the Dominican Republic 114 | Dominican Republic: Systematic Country Diagnostic References Acosta, P., Fajnzylber, P., and H. Lopez. 2007. “The Im- Beilys, G., and B. Cunha. 2017. Energy Pricing Policies pact of Remittances on Poverty and Human Capi- for Inclusive Growth in Latin America and the Carib- tal: Evidence from Latin American Household Sur- bean. Washington, DC: World Bank. veys.” World Bank Policy Research Working Paper Benito, A. B. 2017. “El Diputado Conseguidor: Rep- 4247. resentación Política Particularista en el Congreso de Alaimo, V., Bosch, M., Gualavisí, M., and J. M. Vil- la República Dominicana (1998-2010).” Revista Es- la. 2017. “Measuring the Cost of Salaried Labor pañola de Ciencia Política 44: 151-179. in Latin America and the Caribbean.” Technical Benito, A. B. 2014. “Caciques y facciones en República Note IDB-TN-1291, Inter-American Development Dominicana.” Revista Mexicana de Sociología 76 (4): Bank. 529-556. Alfaro, L., and M. Chen. 2002. “Selection and Market Benito, A. B., and W. Lozano. 2012. “Clientelistic Sta- Reallocation: Productivity Gains from Multination- bility and Institutional Fragility.” Latin American al Production.” NBER Working Paper 18207. Perspectives 39 (6): 198-211. Araujo, J., Vostroknutova, E., Brueckner, M, Clavijo, Cabrera, M., Aristy, J., Moreno-Dodson, B., and M. Sanchez. 2016. “Fiscal Policy and Redistribution in M., and K. Wacker. 2016. Beyond Commodities: The the Dominican Republic: An Analysis Based on the Growth Challenge of Latin America and the Caribbe- “Commitment to Equity” Methodology for 2013,” an. Latin American Development Forum, Washing- Washington, DC: World Bank. ton, DC: World Bank. Carneiro, F., Niang, C., Mele, G., Pena, C., Lambrides, Báez, J. E., Fuchs, A., and C. Rodríguez-Castelán. 2017. M., Berman, L., Parra, J. C., and J. Brault. 2016. Shaking Up Economic Progress: Aggregate Shocks in Building a Better Future Together: Dominican Republic Latin America and the Caribbean. Washington, DC: Policy Notes. Washington, DC: World Bank Group. World Bank. https://hubs.worldbank.org/docs/imagebank/pag- Banco Mundial. 2014. Cuando la Prosperidad no es es/docprofile.aspx?nodeid=27159212 Compartida. Banco Mundial: Washington, DC. Carneiro, F., and S. Sirtaine, editors. 2017. When Barbero, J. A., and P. Guerrero. 2017. El transporte au- Growth is not Enough: Explaining the Rigidity of Pov- tomotor de carga en América Latina: Soporte logístico erty in the Dominican Republic. Directions in Devel- opment – Poverty. Washington, DC: World Bank. de la producción y el comercio. Inter-American Devel- Carrasco, H., García, E., Parodi, S., and M. Vásquez. opment Bank. Dominican Republic: Systematic Country Diagnostic | 115 2016. Cómo se Redistribuyen los Recursos Públicos en Giussani, B., Guardiola, U., Ospina, J., and J. Shep- República Dominicana? Banco Interamericano de herd. 2016. PEFA República Dominicana 2016. AE- Desarrollo. COM International Development Europe SL. Castro, A. 2016. The Unresolved Problem of Maternal Hallegatte, S., Vogt-Schilb, A., Bangalore, M., and J. Mortality in the Dominican Republic. Draft Case Rozenberg. 2017. Unbreakable: Building the Re- Study. World Development Institute. silience of the Poor in the Face of Natural Disasters. CEPAL. 2008. “República Dominicana: Evolución de Climate Change and Development. World Bank. los impactos socioeconómicos de la tormenta Noel”, https://openknowledge.worldbank.org/han- CEPAL México. dle/10986/25335 Consejo Nacional de Zonas Francas de Exportación. Hanushek, E., and L. Woessmann. 2010. The High 2016. Informe Estadístico Sector Zonas Francas de Cost of Low Educational Performance: The Long- República Dominicana 2016. run Economic Impact of Improving PISA outcomes. Cord, L., Genoni, M.E., and C. Rodríguez-Castelán, Programme for International Student Assessment, editors. 2015. Shared Prosperity and Poverty Eradica- OECD. tion in Latin America and the Caribbean. Washing- Hausmann, R., Hidalgo, C., Jiménez, J., Lawrence, ton, DC: World Bank. R., Levy-Yeyati, E., Sabel, C., and D. Schydlowsky. Cruz-Rodriguez, A. 2016. “Tipo de cambio real en la 2012. “Building a Better Future for the Dominican República Dominicana: enfoques alternativos de Republic.” CID Working Paper, Harvard Universi- equilibrio y desalineamientos,” Ciencia y Sociedad, ty: Cambridge, MA. 41 (3): 589-616. infoDev. 2015. Profiling Caribbean Women Entrepre- Daude, C., Gutiérrez, H., and A. Melguizo. 2014. neurs. Washington, DC: The World Bank. “’Doctoring the Ball’: The Political Economy of Institute for Health Metrics and Evaluation - IHME. Tax Incentives for Investment in the Dominican 2017.  Global Burden of Disease Compare Data Republic” OECD Development Center Working Visualization. Seattle, WA: IHME, University of Paper No. 322. Washington. de la Torre, A., Ize, A., and S. Pienknagura. 2015. Lat- Ishizawa, O., Miranda, J., and E. Strobl (2017). “The in America Treads a Narrow Path to Growth: LAC Impact of Hurricane Strikes on Short-Term Local Semiannual Report, April 2015. Washington, DC: Economic Activity in the Dominican Republic: Ev- World Bank. idence from Nightlight Images”, World Bank Policy de la Torre, A., Levy-Yeyati, E., and S. Pienknagura. Research Working Paper 8275. 2013. Latin America’s Deceleration and the Exchange Izzo, M., Araujo, N., Aucelli, P., Maratea, A., and A. Rate Buffer: LAC Semiannual Report, October 2013. Sánchez. 2013. “Land Sensitivity to Desertification Washington, DC: World Bank. in the Dominican Republic: An Adaptation of the Díaz-Bazán, T. 2015. “Measuring Inequality from Top ESA Methodology.” Land Degradation & Develop- to Bottom.” World Bank Policy Research Working ment 24: 486-498. Paper No. 7237. Jaitman, L. (ed.). 2017. The costs of crime and violence: Espinal, R., Morgan, J., Boidi, M. F., and E. Zech- new evidence and insights in Latin America and the meister. 2017. Cultura política de la democracia en Caribbean. Inter-American Development Bank. la República Dominicana y en las Américas, 2016/17. Javorcik, B. 2004. “Does Foreign Direct Investment FAO. 2015. Evaluación de los Recursos Forestales Mundi- Increase the Productivity of Domestic Firms? In ales 2015 – Informe Nacional República Dominicana. Search of Spillovers through Backward Linkages.” Ferguson, J. 2003. “Migration in the Caribbean: Haiti, American Economic Review 94 (3): 605–27. the Dominican Republic and Beyond.” Javorcik, B. 2010. “Foreign Direct Investment and In- Ferreira, F., Messina, J., Rigolini, J., López-Calva, L. F., ternational Technology Transfer.” In Encyclopaedia Lugo, M. A., and R. Vakis. 2013. Economic Mobil- of Financial Globalization, edited by G. Caprio. ity and the Rise of the Latin American Middle Class. Amsterdam: Elsevier. Latin America and Caribbean Studies. Washington, Jouanjean, M-A., J-C. Maur, and B. Shepherd. 2012. DC: World Bank. “Reputation Matters: Spillover Effects in the En- Forest Carbon Partnership Facility (FCPF) Carbon forcement of US SPS Measures,” Policy Research Fund (2015). Dominican Republic Emissions Reduc- Working Paper Series 5935. Washington, DC: The tion Program Idea Note (2015). World Bank. Galasso, E., and A. Wagstaff. 2016. “The Economic Kaufmann, D., Gallina, A., and R. Senderowitsch. Costs of Stunting and How to Reduce Them.” Poli- 2015. “A Collective Action Approach Against Cor- cy Research Note. World Bank. ruption: The Case of the Dominican Republic.” 116 | Dominican Republic: Systematic Country Diagnostic GPSA Working Paper 3. World Bank: Washington, ra-suelo-2012.pdf ) DC. Ministerio de Medio Ambiente y Recursos Naturales. Keefer, P., and R. Vlaicu. 2008. “Democracy, Credibili- 2012. “Plan Estrategico Institutional del Ministerio ty, and Clientelism.” The Journal of Law, Economics, de Medio Ambiental y Recursos Naturales 2012- & Organization 24(2): 371-406. 2016.” Santo Domingo. Keefer, P. 2002. “The Political Economy of Public Mitchell, K. 2009. “Democratisation, External Expo- Spending Decisions in the Dominican Republic: sure and State Food Distribution in the Dominican Credibility, Clientelism and Political Institutions”, Republic.” Bulletin of Latin American Research 28 mimeo. (2): 204-226. Kitschelt, H., and D. Kselman. 2011. “Organizational Molinas, J. R., Paes de Barros, R., Saavedra, J., and M. Extensiveness and Political Clientelism: The Formal Giugale. 2012. Do Our Children Have a Chance? A and Informal ‘ties that bind’”. Working paper. Human Opportunity Report for Latin America and Lange, G., Wodon, Q., and K. Carey (eds). 2018. The the Caribbean. Washington, DC: World Bank. Changing Wealth of Nations 2018: Building a Sus- Morgan, J., Espinal, R., and E. J. Zechmeister. 2015. tainable Future. Washington, DC: World Bank. Cultura política de la democracia en República Do- Lederman, D., and J. Lesniak. 2018. Open and Nimble: minicana y en las Américas, 2014. USAID. Finding Stable Growth in Small Economies. Direc- Morgan, J., Hartlyn, J., and R. Espinal. 2011. “Do- tions in Development—Countries and Regions;. minican Party System Continuity amid Regional Washington, DC: World Bank. Transformations: Economic Policy, Clientelism, Lederman, D., Messina, J., Pienknagura, S. and J. and Migration Flows.” Latin American Politics and Rigolini. 2014. Latin American Entrepreneurs: Society 53 (1): 1-32. Many Firms but Little Innovation. World Bank Lat- Moya Pons, F. 2013. Manual de Historia Dominicana. in American and Caribbean Studies;. Washington, Ediciones Librería La Trinitaria. DC: World Bank. Moya Pons, F. 1992. Empresarios en Conflicto. FLAC- Lim, K. 2017. “The Role of the International Mari- SO. time Organization in Preventing the Pollution of Oficina Nacional de Estadística – ONE. 2017. Fascículo the World’s Oceans from Ships and Shipping.” UN I. Seguridad Ciudadana en la República Dominicana Quaterly 54(2): 52-54. (ENHOGAR-2015). Loayza, N., Fajnzylber, P., and C. Calderón. 2005. Eco- Oficina Nacional de Estadística – ONE. 2015. Obser- nomic Growth in Latin America and the Caribbean: vatorio MIMPYMES Boletín #2: Edición de Genero. Stylized Facts, Explanations, and Forecasts. Washing- Oficina Nacional de Estadística – ONE. 2012. Prim- ton, DC: World Bank. era encuesta nacional de inmigrantes en la República Lozano, W. 2013. La Razón Democrática. Editorial Dominicana. UNIBE, Centro Bonó, FLACSO, Fundación Frie- Paes de Barros, R., Ferreira, F., Molinas, J., and J. Saave- drich Ebert Stiftung. dra. 2009. Measuring Inequality of Opportunities Lucchetti, L. 2017. “Who escaped poverty and who was in Latin America and the Caribbean. Washington, left behind”, World Bank Policy Research Working D.C.: World Bank Group. Paper 8220. Page, S., Essex, S., and S. Causevic. 2014. “Tourist at- Mateo Díaz, M., and L. Rodríguez-Chamussy. 2016. titudes towards water use in the developing world: Cashing in on Education: Women, Childcare, and A comparative analysis”. Tourism Management Per- Prosperity in Latin America and the Caribbean. Latin spectives 10: 57-67. American Development Forum. Patil, P., Virdin, J., Roberts, J., Singh, A., S. Diez. 2016. Ministerio de Economía, Planificación y Desarrollo. Toward a blue economy: a promise for sustainable 2016. Ley 1-12 Estrategia Nacional de Desarrollo. growth in the Caribbean (Vol. 2): overview (English). Santo Domingo. (Available at http://economia.gob. Washington, D.C.: World Bank Group. do/mepyd/wp-content/uploads/archivos/transpar- Poole, J. 2013. “Knowledge Transfers from Multina- encia/base-legal/end_2030.pdf ). tional to Domestic Firms: Evidence from Worker Ministerio de Hacienda. 2016. Gastos Tributarios en Mobility.” Review of Economics and Statistics 95 (2): República Dominicana: Estimación para el Presu- 393–406. puesto General del Estado del año 2017. Pritchett, L., and E. Werker. 2012. “Developing the Ministerio de Medio Ambiente y Recursos Naturales. guts of a GUT (Grand Unified Theory): Elite Com- 2014. Estudio de uso y cobertura de suelo 2012. San- mitment and Inclusive Growth.” ESID Working to Domingo, R.D. (Available at http://ambiente. Paper 16/12. gob.do/wp-content/uploads/2016/12/uso-cobertu- Rufín, C., Zucchini, D., Senderowitsch, R., and M. E. Dominican Republic: Systematic Country Diagnostic | 117 Sánchez. 2014. “The Dominican Republic: Mov- Wooding, B. 2018. “Haitian Immigrants and Their ing from Exit to Voice – Shifting Incentives in the Descendants Born in the Dominican Republic,” Power Sector.” In Problem-Driven Political Economy Oxford Research Encyclopedia of Latin American Analysis – The World Bank Experience, Fritz, V., Levy, History. B. Levy, and R. Ort. Washington, DC: World Bank. Wooding, B. and N. Riveros. 2017. “Migración Lab- Ruggles, S., Genadek, K., Goeken, R., Grover, J., and oral Haitiana hacia República Dominicana: Reali- M. Sobek. 2017.  Integrated Public Use Microdata dad, Retos y Propuestas hacia una Gobernanza más Series: Version 7.0 [dataset]. Minneapolis: Universi- Efectiva,” Perspectivas No. 03-17, OBMICA. ty of Minnesota. https://doi.org/10.18128/D010. World Bank. 2018. Global Investment Competitiveness V7.0 Report 2017/2018: Foreign Investment Perspectives Sánchez, M. E., J. de Piniés, and K. Antoine. 2015. and Policy Implications. Washington, DC: World “Measuring the Determinants of Backward Linkag- Bank. es from FDI in Developing Countries: Is it a Mat- World Bank. 2017a. World Development Report 2017: ter of Size?” World Bank Policy Research Working Governance and the Law. Washington, DC: World Paper 7185. Bank. Sánchez, M. E., and R. Senderowitsch. 2012. “The Po- World Bank. 2017b. “Do Labor Markets Limit the In- litical Economy of the Middle Class in the Domin- clusiveness of Growth in the Dominican Republic?” ican Republic: Individualization of Public Goods, World Bank, 2017c. “An assessment of tax efficiency, Lack of Institutional Trust, and Weak Collective Ac- a cost-benefit analysis of tax expenditures, and an tion.” Policy Research Working Paper 6049, World exploration of labor informality and its tax implica- Bank. tions : gearing up for a more efficient tax system,” Skoufias, E., Narita, R., and A. Narayan. 2014. “Does Washington, DC: World Bank. Access to Information Empower the Poor? Evidence World Bank. 2017d. World Development Report 2017: from the Dominican Republic.” World Bank Policy Governance and the Law. Washington, DC: World Research Working Paper 6895. Bank. Sousa, L., Sánchez, D., and J. Baez. 2017. “Wage Effects World Bank. 2017e. Benchmarking Public Procurement of Haitian Migration in the Dominican Republic.” 2017. Washington, DC: World Bank. In When Growth is not Enough: Explaining the Ri- World Bank. 2016. “Special Economic Zones in the gidity of Poverty in the Dominican Republic, edited Dominican Republic: Policy Consideration for a by Francisco Galrao Carneiro and Sophie Sirtaine. more Competitive and Inclusive Sector,” Washing- Directions in Development – Poverty. Washington, ton, DC: World Bank. DC: World Bank. World Bank. 2015. “How to sustain export dynamism Stokes, S. 2009. “Political Clientelism” In The Oxford by reducing duality in the Dominican Republic : Handbook of Comparative Politics, edited by Carles a World Bank trade competitiveness diagnostic,” Boix and Susan C. Stokes, 604–27. Oxford Hand- Washington, DC: World Bank. books of Political Science Series. New York: Oxford World Bank. 2013. Patronage or Reform? Political Econ- University Press. omy of Policy Performance in the Dominican Republic. Transparency International. 2017. Las Personas y la Cor- Institutional and Governance Review Series. Wash- rupción: América Latina y el Caribe. ington, DC: World Bank. UNESCO. 2017. Análisis del Desarrollo Mediático en World Bank. 2012. Teenage Pregnancy and Opportuni- República Dominicana. UNESCO: La Habana. ties in Latin America and the Caribbean. Washing- USAID. 2013. Dominican Republic Climate Change ton, DC: World Bank. Vulnerability Assessment Report. African and Latin World Bank. 2006. Dominican Republic Poverty Assess- American Resilience to Climate Change (ARCC). ment: Achieving More Pro-poor Growth. World Bank: Végh, C., Lederman, D., and F. Bennett. 2017. Leaning Washington, DC. Against the Wind: Fiscal Policy in Latin America and World Bank. 2004. “Environmental Priorities and Stra- the Caribbean in a Historical Perspective. LAC Semi- tegic Options - Country Environmental Analysis,” annual Report, April 2017. World Bank, Washing- Washington, DC: World Bank. ton, DC. World Economic Forum. 2017. The Global Competi- Wodon, Q., Male, C., Onagoruwa, A., Savadogo, A., tiveness Report 2017-2018. and Yedan, A. 2017. “Impacto económico del ma- World Travel and Tourism Council. 2018. Travel & trimonio infantil y las uniones tempranas: Resumen Tourism – Economic Impact 2018 Dominican Repub- para la República Dominicana.” Washington, DC: lic. Banco Mundial y UNICEF.