DOMINICAN REPUBLIC COUNTRY ECONOMIC MEMORANDUM: SUSTAINING ECONOMIC GROWTH 1 Table of Contents Acknowledgments ........................................................................................................................................ 4 List of abbreviations ..................................................................................................................................... 5 Executive summary ...................................................................................................................................... 6 Chapter 1. Exceptionally strong, yet uneven economic growth ............................................................... 14 a. Introduction ........................................................................................................................................ 14 b. The Dominican Republic’s growth performance ............................................................................... 15 c. Yet, growth has been uneven ............................................................................................................ 21 d. Higher productivity growth is needed to sustain growth ................................................................. 24 e. Conclusions ......................................................................................................................................... 26 f. Appendix ......................................................................................................................................... 27as Appendix 1.1. Country peers ..................................................................................................................... 27 Appendix 1.2. Calculating Backward linkages ........................................................................................... 28 Chapter 2. Reversing the productivity decline .......................................................................................... 29 a. Introduction ........................................................................................................................................ 29 b. Productivity dynamics ........................................................................................................................ 30 c. Factors constraining productivity growth ......................................................................................... 34 i) Labor market skills inefficiencies and demand-supply mismatches ...................................................... 34 ii) Non-creative innovation ....................................................................................................................... 39 d. The implications of special tax regimes (STRs).................................................................................. 41 iii) Climate change ..................................................................................................................................... 42 e. Summary: closing the growth and income gaps ............................................................................... 45 f. Appendix ............................................................................................................................................. 47 Appendix 2.1. Calculating sectoral distortions .......................................................................................... 47 Appendix 2.2. Computable General Equilibrium (CGE) model ................................................................. 48 Chapter 3. Enhancing market competition to boost growth .................................................................... 51 b. Introduction ........................................................................................................................................ 51 c. Understanding the competition environment in the Dominican Republic ...................................... 52 d. Barriers to competition in the Dominican Republic .......................................................................... 55 i) Direct channel: Government as a market player (SOEs) ........................................................................ 58 ii) Indirect channel: Government as a regulator ....................................................................................... 61 iii) Indirect channel: Government as a referee .......................................................................................... 65 2 e. Conclusions ......................................................................................................................................... 70 f. Appendix ............................................................................................................................................. 71 Appendix 3.1. Product market regulation (PMR) indicators .................................................................... 71 Chapter 4. Reducing expenditure inefficiencies to open up fiscal space for productivity-enhancing investments ................................................................................................................................................ 73 a. Introduction ........................................................................................................................................ 73 b. Improving procurement efficiency .................................................................................................... 75 g. Conclusions ......................................................................................................................................... 81 Conclusion .................................................................................................................................................. 82 3 Acknowledgments consisted of Sebastian Galiani (Professor, University of Maryland), Jorge Thompson Araujo (Senior The Dominican Republic Country Economic Research Fellow, University of Brasilia and Consultant Memorandum (CEM) was prepared with the ELCMU), Paola Brens (Consultant, ELCMU), Milagros leadership and support of the Country Management Deza (Consultant, ELCMU) and Ivana Benzaquen Unit (CMU) led by Michel Kerf (Country Director, (Consultant, ELCMU), who provided inputs for the LCC2C) and Alexandria Valerio (Resident macroeconomic analysis in the Chapter 1, while the Representative, LCCDO). The team benefited from regional GDP estimates was led by Jose Pablo Valdez technical support and guidance from Doerte (Senior Economist, DECES) and Charl Jooste (Senior Doemeland (Practice Manager, ELCMU) and Pedro L. Economist, EMFMD) provided advice on the macro- Rodriguez (Lead Country Economist, ELCDR). structural forecasts; Dong Dao (Consultant, ELCMU), and Alexis Cruz (MEPyD) contributed in the Constructive feedback and discussions were received productivity analysis of Chapter 2; while Hasan Dudu from counterparts from the Ministry of Economy, (Senior Economist, EMFMD), Martin Christensen Planning and Development (MEPyD), led by Mr. Pavel (Economist, EMFMD) and Carmen Estrades Isa (Minister), and included: Miguel Ceara (former (Consultant, ELCMU) led the model simulations, with minister), Alexis Cruz-Rodriguez (Vice Minister of contributions from Huong Nguyen (Energy Specialist, Economic and Social Analysis), Evalina Gomez ILCE1) for the energy scenario. The competition policy (Director of Sectoral Economic Analysis) and Rosa analysis in chapter 3 was led by Denis Sanchez Cañete (Director of Poverty, Inequality and (Economist, ETIMT), with contributions from Seidu Democratic Culture); likewise, the National Dauda (Economist, ETIMT) and Soulange Gramegna Competitiveness Council (CNC) significantly (Consultant, ETIMT), under the guidance of Martha contributed in the development of the narrative, with Licetti (Practice Manager, ETIMT) and Graciela comments provided by the CNC president, Peter A. Miralles (Senior Economist, ETIMT), while Maria Prazmowski, and Katherine Javier (Director of the Scheker (Procurement Specialist, ELCRU), Paola Brens Competitiveness Intelligence unit); in addition, (Consultant, ELCMU) and Milagros Deza (Consultant, relevant suggestions were provided by the Analysis ELCMU) contributed for the procurement analysis of and Fiscal Policy unit from the Ministry of Finance, led Chapter 4. by Camila Hernandez, the Economic and Taxes Study unit led by Patricia Gil, as well as other members of Anjali Shahani (Operations Officer, ELCMU), Giselle the Ministry of Finance. The National Statistical Office Velasquez (Program Assistant, ELCMU), Benjamin (ONE) as coordinated by Augusto de los Santos, Vuilleminroy (Program Assistant, ELCMU), and Maria played an important role by providing relevant J. Hermann (Executive Assistant, LCCDO) provided information for several chapters of the report. The valuable administrative and logistical support. CEM benefited from consultations from private sector organizations, and academia, specifically Raul The team thanks the CEM peer reviewers for their Hernandez, director of the national pension helpful suggestions and insights: Marc Schiffbauer association and advisor of private sector companies, (Senior Economist, EECM2), Ana Cusolito (Senior provided insightful comments for Chapter 2. Economist, ETIMT), Guillermo Vuletin (Senior Economist, LCRCE), Daniel Riera-Crichton (Senior The task team was led by James Sampi (Economist, Economist, LCRCE), Samuel Pienknagura (Senior ELCMU) and Johannes Herderschee (until April 2022, Economist, IMF) and Fausto Patino (Economist, Senior Economist, CELCE). Gabriel Zaourak (Senior ELCFN). Economist, ELCMU) finalized the report. The team 4 List of abbreviations Abbreviation Definition BCRD Central Bank of the Dominican Republic BTI Bertelsmann Transformation Index CAFTA-DR The Dominican Republic-Central America Free Trade Agreement CGE Computable General Equilibrium Model DEE Directory of Companies and Establishments DGII General direction of domestic tax revenues DR Dominican Republic ECLAC Economic Commission for Latin America and the Caribbean FDI Foreign Direct Investment GDP Gross Domestic Product ILO International Labour Organization IMF International Monetary Fund LAC Latin American and the Caribbean MEPyD Ministry of Economic, Planning and Development MFA Multifiber Agreement MFMod Macro-Fiscal Model of the World Bank ONAPI National Office of Industrial Property ONE National Statistics Office PMR Product Market Regulation PPL Public Procurement Law PPP Purchasing Power Parity PWT Penn World Table SEZs Special Economic Zones SSP Shared Socioeconomic Pathways scenario RCPs The Representative Concentration Pathways TFP Total Factor Productivity WBES World Bank Enterprise Survey WBG World Bank Group WBG MCT World Bank Markets, Competition, and Technology WDI World Development Indicators 5 Abstract The Dominican Republic’s growth model has led to exceptional economic expansion, far exceeding the LAC average over the past two decades, but it is reaching its limits due to slow productivity growth. While impressive, the Dominican Republic’s growth performance has been based on factor accumulation rather than productivity growth and associated with regional disparities and wage stagnation. Productivity growth has been hindered by policy distortions, climate-related natural disasters as well as inadequate human capital. A new round of structural reforms will boost productivity growth, including by: (i) unleashing its human capital potential; (ii) fomenting competitive markets; (iii) revamping the innovation strategy; (iv) reducing public expenditure inefficiencies; and (v) strengthening resilience against external shocks and climate change. Executive summary The Dominican Republic’s growth model has led to exceptional economic expansion, but it is reaching its limits due to slow productivity growth The Dominican Republic’s macroeconomic performance exceeded the regional Latin America and the Caribben (LAC) average over the past two decades, leading to a rapid convergence towards the US in per capita levels. The Dominican Republic pursued economic reforms between the 1990s and early 2000s—including the liberalization of foreign exchange transactions and trade, and elimination of price controls and restrictions to foreign direct investment (FDI). These reforms accelerated economic growth in the following decades, which averaged 5.8 during the 2005-2019 period. Favorable geographical and external economic conditions also boosted growth, as did a combination of economic incentives that attracted investment. However, the Dominican Republic’s recent growth model has been based on factor accumulation rather than productivity growth . Capital formation has been the main contributor of economic growth, accounting for 3.7 percentage points of GDP growth on average between 2005 and 2019. Employment was responsible for a non-negligible 1.6 percentage-point rise in GDP over the same period, with labor quality only contributing 0.4 percentage points. After the banking crisis of 2003/2004, Total Factor Productivity’s (TFP) contribution to GDP growth has been positive—unlike that in most of LAC countries. However, it has not increased in lockstep with per capita income. Productivity in the Dominican Republic caught up with the U.S. by the mid-1990s. However, due to the absence of further reforms, it receded back to 60 percent by 2019, well below both structural and aspirational peer economies such as Costa Rica and Uruguay, respectively. The Dominican Republic’s growth model has also been associated with wage stagnation and regional disparities 6 Lagging productivity contributed to depressed real wages in a country with already high regional disparities. Although GDP per capita in the country is among the highest in the LAC region, average monthly earnings are among the lowest— the average wage in countries with a similar income level, such as Costa Rica, is nearly three times higher (Figure i). Despite rapid economic growth, the share of the population that is considered vulnerable (i.e., with an income between US$5.50 and US$13 per day in PPP) increased from 38.1 percent in 2005 to 43.7 percent in 2019.1 Similarly, depressed real wages contributed to consumption welfare inequality with growth benefits accruing only to a handful of regions. The richest 1 percent of households accounts for nearly 10 percent of national income – nearly 4 percentage points higher compared to countries such as Uruguay. Furthermore, income gaps by gender are sizable. Income per capita in the regions of Ozama or Yuma is significantly higher than in poorer regions such as Enriquillo or El Valle. Figure i. Average monthly earnings and GDP per capita in 2019, in constant 2017 US$ PPP US$ 2017 PPP US$ 2017 PPP GDP per capita 35,000 1,400 Average monthly earnings (RHS) 30,000 1,200 25,000 1,000 20,000 800 15,000 600 10,000 400 5,000 200 0 0 Source: International Labor Organization (ILO) and World Development Indicators (WDI). World Bank Staff’s calculations. Faster productivity would fuel economic growth in the medium and long run, especially for the poorest households. A one percentage point increase in TFP above its historical average would lift GDP growth to nearly 7 percent, enabling a rapid convergence with the living standards of high-income countries. Productivity enhancing reforms would increase the average income of the poorest households by nearly 17 percent by 2030, and almost double the income by 2050. Productivity growth has been hindered by policy distortions, climate-related natural disasters and inadequate human capital Lagging productivity seems to be caused by domestic distortions rather than external distortions and is more pronounced in the non-tradable sectors. A calibrated model for the 1 The World Bank LAC Equity Lab: https://www.worldbank.org/en/topic/poverty/lac-equity-lab1/poverty/head-count 7 Dominican Republic indicates that if the level of domestic distortions2 were reduced by 1 percentage point, GDP could rise by 1.46 percentage points above the current trend, while a 1 percentage point fall in the level of external distortions would only boost GDP by 0.21 percentage points. The level of internal distortion in non-tradable sectors can be twice as high relative to the tradable sectors. As an example, reducing the level of distortion in the telecommunications sector by 1 percentage point can add 0.21 percentage points to the current GDP growth trend. Similarly, removing distortions in the energy sector can potentially unlock another 0.09 percentage-points increase in GDP, by reducing blackouts that disrupt manufacturing activity. Productivity growth has been hindered by policy distortions, climate-related natural disasters as well as inadequate human capital. Using a household survey database, it is estimated that workers joining the labor market had an average learning gap of four years relative to their US peers. Despite having been in school for a similar length of time, their skill levels were behind that of workers educated in the US. In addition, a national survey of the private sector reveals that nearly 10 percent of firms consider that inadequate human capital represents a constraint for business development, the fourth main constraint in the survey. About 40 percent of those firms refer to a lack of technical skills as a major constraint for filling vacancies, a problem that is more predominant in large firms. Meanwhile, 28 percent of those firms indicate that labor market regulations and obstacles for obtaining a license are also important constraints for business development, while 11 percent refer to access to electricity as an impediment to development. Competitive pressures in the economy seem to have softened, reducing incentives to innovate. Moreover, the country has suffered a sharp decline in the number of patent registrations over the last 10 years, reflective of low innovation carried out by large firms. On a similar note, only 1.7 percent of firms that reported performing innovation-related activities introduced a product that could be considered novel in the country, while 96 percent of them introduced products already available in the province where they operate. These bottlenecks and other constraints partly explain why productivity of large firms has lagged that of small and medium-sized firms. A detailed analysis on how competition related factors could explain the productivity discrepancy between large and smaller firms is explored in Chapter 3. Tax exemptions may have contributed to low productivity growth through resource misallocation in sectors associated with low-sophistication manufacturing products. Formal manufacturing firms benefiting from special tax regimes (STR) – e.g., tax-exempted firms producing textiles, clothing and metal products in Special Economic Zones (SEZs) – were found to have nearly 30 percent lower productivity than peers outside the SEZs that are subject to the general tax regime (GTR). This does not necessarily mean that aggregate productivity in SEZs is lower than in the rest of the economy, since a number of firms operating in the former do not have counterparts in the latter. For example, medical device firms can only be found in the SEZs – while accounting for only 5 percent of the total number of SEZ firms, they generate nearly 30 2Distortions are measured as the gap between a frictionless simulated economy and current state of the Dominican economy using Input-Output (IO) tables. The Appendix 2.1. provides a detailed technical description on distortions calculation. 8 per cent of SEZ exports.3 Highly productive firms pay 0.4 percentage points more in income tax than firms in the low-productivity segment in the Dominican Republic, which are predominantly large firms. Firms that register under STR remain in the market longer than those that do not, despite relatively lower productivity levels. As an example, during the period 2007-2016, the exit rate of unproductive firms under the GTR is roughly 9 percent, while being 7 percent under the STR. Similarly, the entry of productive firms under GR is 2.3 percent, while only 0.6 percent for firms under STR. Climate change has also contributed to declining productivity.4 The 2019 Global Climate Risk Index ranked the Dominican Republic 12th among the countries most affected by natural disasters between 1998 and 2017, with hurricanes and tropical storms causing high and recurrent human and economic losses. The estimated annual average loss caused by hurricanes on building stock is valued at US$345 million (0.48 percent of GDP). Average annual temperatures have increased by approximately 0.45°C since 1960, with a concurrent rise in the average number of hot days and nights, which impacts economic activity negatively. Notably, rising temperatures between 2015 and 2020 were associated with productivity losses of between 2 and 9 percent among manufacturing firms located in poor regions of the country.5 A new round of structural reforms will boost productivity growth Sustaining high growth in the long run will require a new round of economic reforms . In addition to maintaining a sound macroeconomic framework, moving forward with critical reforms would be essential to keeping strong growth in the long run. They include: (i) unleashing its human capital potential; (ii) fomenting competitive markets; (iii) revamping the innovation strategy; (iv) reducing public expenditure inefficiencies; and (v) strengthening resilience against external shocks. Strengthening the Dominican Republic’s human capital can be achieved through a two-pronged strategy. First, matching educational outcomes with market needs, by modernizing curricula at high school level, encouraging continuing education for adults, facilitating the upskilling and re- skilling of workers through more relevant active labor market policies, allocating enough financial incentives to attract and retain high-quality teachers in rural and disadvantaged areas, involving when appropriate the private sector into the financing of public colleges, and implementing a learning diagnostic system for elementary schools at yearly basis. Second, reducing inequality of opportunities across genders and between rural and urban areas , by implementing monetary and non-monetary incentives for parental sharing of caretaking during the first year of birth, monitoring students drop out rates, introducing national scholarships for 3 Table 13 in: https://www.cnzfe.gob.do/transparencia/phocadownload/Estadisticas/Informes_Estadisticos_Anuales/Informe%20CNZFE%202 019_web.pdf 4 An exhaustive analysis of the growth and development challenges from climate change is under preparation in the Country Climate and Development Report (CCDR) for the Dominican Republic, which is expected to be finalized during the second half of 2023. 5 Association measured as the partial correlation between firm’s productivity and temperatures across regions. 9 university programs in rural areas for outstanding school students and supporting skills building for the vulnerable population through financial assistance programs. Competition policy reforms have the potential to support productivity in the Dominican Republic by reducing entry and expansion barriers. Markets are perceived as less open to competition and operate under highly concentrated structures, even in sectors where competition is viable. Product Market Regulation (PMR) indicators suggest two main sources of barriers to competition originating in the public sector:6 direct involvement of the government in the form of SOEs and indirectly due to an inadequate regulatory role. SOE presence is still strong in sectors where the economic rationale of state ownership is less clear, with firms seemingly exiting more frequently in those sectors. On the other hand, complex regulations and sectoral provisions protect incumbents. Production and export quotas, as well as price controls, seem to deter firms from entering and expanding. The delayed implementation of the Competition Law limited the active regulatory role of the government for generating competitive markets. The Law has only been implemented in 2017, despite being enacted in 2008. Since 2017, ProCompetencia has investigated potential anti- competitive practices in a wide range of sectors. However, 15 out of the 18 investigations initiated by ProCompetencia have been dismissed (as of March 2022) and only three decisions have been published. In addition, the Competition Law does not include the obligation by the relevant companies to notify and review the competition impact of mergers and acquisitions to obtain approval from ProCompetencia. Similarly, the ability of ProCompetencia to effectively monitor and investigate anticompetitive behavior could be strengthened by addressing financial resource constraints and staffing limitations. Incentives for private firms to innovate can be boosted through the modernization of the national innovation system. This could be achieved by introducing a system of business advisory and technology extension services, building managerial and organizational capabilities at the firm level, developing basic infrastructure for innovation, and implementing a national grant program for small and medium firms that have not benefitted from tax incentives. In addition, eliminating barriers to physical and human capital accumulation would help facilitate innovation. Substantial fiscal savings – opening fiscal space for productivity-enhancing public investments – can result from revenue reforms and from improved expenditure efficiency. On the revenue side, removing tax exemptions and widening the tax base remain a top priority in the country with strong political support. These reforms are even more important in the context of inflationary pressures during a pre-electoral year. Reforms on the expenditure efficiency side can also help address productivity bottlenecks. For example, improving the efficiency of public spending in education would help improve the quality of human capital and reduce skills mismatches through reallocating more resources to initiatives such as modernizing the curricula at high school level, better aligning college education with business needs (especially in the IT sector), and implementing early warning systems to prevent students from dropping out. 6 Barriers to entrepreneurship explains 39 percent of the PMR score for Dominican Republic, while State Control 34 percent. 10 Improving the efficiency of the procurement process can lead to fiscal savings of up to RD$4.6- 5.3 billion (equivalent to 0.1 percent of GDP) and greater competition in product markets. The potential efficiency gains of utilizing more competitive procurement processes in government purchases between 2018 and 2019 could have generated savings equivalent to building 79 new primary schools, hiring over 7,000 elementary or high school teachers, expanding the national fiber optic network in line with the Digital Republic Plan or purchasing more than 100,000 laptops for a training program for recent graduate students. Resilience against external shocks and natural disasters could be enhanced through additional fiscal reforms and novel climate risk management strategies . On the fiscal side, the Dominican Republic would benefit from developing fiscal risk strategies to reduce budgetary uncertainty and enhance financial protection. On the climate risk management side, it could adopt programs for greening of public institutions and create flexible mitigation packages that offset temporary adjustment costs/externalities of decarbonization at regional, company and household level. Table i. Policy recommendations: key policy areas to accelerate growth Time Reform area Reform measure horizon 1/ Matching educational outcomes with market needs - Modernize curricula at high school level aimed at providing technical skills for an easy intersection to the labor market - Encourage continuing education for adults via flexible programs that allows to customize the courses needed at workplace - Facilitate the upskilling and re-skilling of workers through more relevant active labor market policies - Allocate enough financial incentives to attract and retain high-quality teachers in rural and disadvantaged areas - Consider the involvement of private sector into the financing of public colleges, such Unleashing that curricula reflect private sector skills needs Medium human capital - Implement a learning diagnostic system for elementary schools at yearly basis to long potential term 2/ Reducing inequality of opportunities across genders and between rural and urban areas - Consider monetary and non-monetary incentives for parental sharing of caretaking during the first year of birth - Monitor students dropping out – with special emphasis on teenage girls - Consider the introduction of national scholarships for university programs in rural areas for outstanding school students - Support skills building for the vulnerable population through financial assistance programs 1/ Mitigating potential distortions from the direct role of government as SOEs in Fomenting markets Short to competitive medium markets - Assess whether the direct participation of SOEs in markets complies with the term subsidiarity role of the State in commercial sectors. 11 - Explore options to bring-in private sector participation through PPPs or management arrangements where SOEs operate in highly concentrated markets that are viable for competition. 2/ Promoting pro-competition regulation - Leverage initiatives like the single window to reduce entry barriers for firms (e.g., reduce number of days to set a new business) - Telecom: Review licencing requirements to facilitate entry of new operators; Incorporate rules that discipline large market players with significant market power, with a particular focus on potentially abusive practices such as bundling and margin squeeze; and improve enforcement to implement existing regulation to facilitate access to essential facilities. - Gas: Develop third party access regulation - Broadcasting: Review requirement of holding a Dominican nationality to control service provision - Sugar: Consider eliminating production and export quotas - Electricity: Review tariff schemes in the electricity sector to reflect real production, transmission, and distribution costs - Retail: Review the rationale for price controls in retail products and propose alternative, less distortive, mechanisms. 3/ Strengthening competition policy enforcement - Consider developing a merger control framework to limit negative effects of market consolidation. - Include the leniency program as part of the mechanisms to deter cartels in the Competition Law. - Eliminate the maximum duration of investigations (12 months). - Eliminate the prescription of 12 months for anti-competitive conducts to be prosecuted and investigated by the Competition Agency. - Increase the level of sanctions to foster deterrence: calculate sanctions on the basis of turnover/eliminate caps for sanctions. - Implement the ability of ProCompetencia to develop effective coordination and regulation jointly with sector regulators and enhance inter-institutional cooperation mechanisms. - Increase the resources for ProCompetencia including more staff, independence to conduct down-raids, and technical laboratories for processing digital evidence. - Implement measures to mitigate conflict of interest of the competition agency (e.g., cool off periods). Develop tools to identify market distortions of state aid: Implement an inventory of state aid, develop analytical tools to understand competition impact of state aid. 1/ Modernizing the National Innovation System (NIS) - Develop a system of business advisory and technology extension services - Building managerial and organizational capabilities through customized courses for Revamping the local entrepreneurs Medium innovation - Develop basic infrastructure- National quality infrastructure (NQI) and Incubation & term strategy Accelerators - Eliminate barriers to physical and human capital accumulation - Develop a national grant program for small and medium firms that have not benefitted from tax incentives Reducing - Limit the use of exception procedures and under the threshold procedures in the expenditure national procurement processes, and incentive the use of more competitive Short inefficiencies procedures term in public - Consider the introduction of a national catalogue for standardized products procurement 12 - Consider implementing an Early-Warning system for assessing the efficiency of public procurement - Consider introducing a Total Cost of Ownership (TCO), Life Cycle Costing (LLC) and Value for Money (VM) law clauses into the current Procurement Law. - 1/ Reducing fiscal exposure to natural hazards events Strengthening - Develop a fiscal risk strategy aimed at reducing budgetary uncertainty economic - Develop a fiscal strategy for financial protection Short to resilience medium against shocks 2/ Enhancing climate adaptation and mitigation response term and climate change - Adopt programs for greening of public institutions - Create flexible mitigation packages that offset temporary adjustment costs/externalities of decarbonization at regional, company and household level. 13 Chapter 1. Exceptionally strong, yet uneven economic growth a. Introduction 1. The Dominican Republic has experienced strong and stable economic growth over the last two decades, surpassing most economies in Latin America and the Caribbean (LAC). Economic growth averaged 5.8 percent per year from 2005 to 2019 period, more than double compared the LAC average (2.6 percent). Among the country’s regional, structural, and aspirational comparators,7 only Panama experienced faster growth (6.7 percent per year) throughout this period. The aftermath of the pandemic has demonstrated the resilience of the Dominican economy, whose strong recovery has outpaced many regional peers. Growth was also highly resilient, only disrupted since 2005 by the Global Financial Crisis (GFC) in 2008, the European sovereign-debt crisis in 2012, and the COVID-19 pandemic in 2020. 2. The economic growth engines are reaching their limits due to slow productivity growth, contributing to depressed real wages. The growth decomposition exercise indicates that capital formation accounted for 3.7 percentage points of GDP growth on average between 2005 and 2019. Labor quantity was responsible for a non-negligible 1.6 percentage-point rise in GDP over the same period, but labor quality only contributed 0.4 percentage points. After the banking crisis of 2003/2004, total factor productivity (TFP)’s contribution to GDP growth has been close to zero. Relative to the real GDP levels in aspirational peers, the Dominican Republic’s real wages are, on average, 70 percent lower. Faster TFP growth would thus have the double benefit of reducing the country’s dependence on factor accumulation to sustain its high growth rate and addressing its real wage stagnation problem. 3. This chapter analyzes the growth dynamics of the Dominican economy, while highlighting the main bottlenecks that could be preventing a sustained growth model. Productivity growth has lagged in recent years, while improvements in labor quality have not matched the performance in GDP growth. Although robust economic growth has promoted job creation in manufacturing and services, real wages remain low, exacerbating income inequality and limiting the benefits of growth. International evidence shows that productivity can boost real wages, but economic growth in the Dominican Republic has primarily stemmed from the accumulation of production factors—labor and capital—not from rising productivity. If TFP — defined as the efficiency in increasing production while input factors, i.e., labor and capital, remain constant—increased by 1 percentage point above its historical average, the average GDP growth rate could rise to 7 percent in the long run. Conversely, a 1 percentage point decline in TFP growth would push GDP growth down to 3 percent, compromising the country’s development objectives. 7 Thestructural comparators are compressed by Tunisia, Bulgaria and Costa Rica, while aspirational by Uruguay, Croatia and New Zealand. The Appendix 1.1. provides a detailed description of structural and aspirational peer comparators. 14 b. The Dominican Republic’s growth performance 4. The Dominican Republic (DR) has been one of the fastest-growing economies in Latin America and the Caribbean. The economy expanded by an average of 5.8 percent in 2005 - 2019, as foreign direct investment (FDI) inflows, averaging about 4 percent of GDP over the last two decades, transformed the economy, and fueled tourism, services, manufacturing, construction, and mining. A combination of market-oriented structural reforms in the early 1990s and favorable external conditions supported economic growth (see Box 1.1). Prudent monetary and fiscal policy has contributed to macroeconomic stability, despite the increasing public debt level. 5. The Dominican Republic’s per capita income has converged steadily towards US per capita income levels. After a temporary increase in poverty rates8 in 2003/2004, the share of the population living in poverty declined from 48 percent in 2005 to nearly 20 percent in 2019. Such success was accompanied by a strong convergence towards the US per capita income level. In 2000, the Dominican GDP per capita represented 19 percent of the US, and almost doubled to 30 percent by 2019. Meanwhile, the average LAC GDP per capita remained flat at 25 percent during the same period. Overall, the Dominican economy has been a success story in the region over the last 15 years, although there is still ground to cover to catch up with aspirational peers such as Uruguay or Croatia. Figure 1.1. GDP Growth, 2000-2021 Figure 1.2. GDP Per Capita (y-o-y percentage change)9 (PPP, constant 2017 international US$) Structural Central America DR/USA LAC/USA Dominican Republic Latin America & Caribbean 35.0% Aspirational 14.0 29.5% 12.0 30.0% 10.0 8.0 24.7% 6.0 25.0% 25.3% 25.7% 4.0 2.0 - 20.0% -2.0 -4.0 18.6% -6.0 15.0% -8.0 -10.0 10.0% 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Source: World Bank, “The Dominican Republic: An Overview of Macroeconomic Performance”, background note 6. Strong private consumption and gross fixed capital accumulation sustained rapid economic growth. Between 2005 and 2019, on average, private consumption accounted for 72 percent of GDP and a nearly 4 percent rise in GDP, while gross fixed capital formation (GFCF) accounted for 25 percent of GDP and a rise in GDP of approximately 2 percent. The 1990s market- oriented reforms, specifically the price and exchange rate liberalization, coupled with increasing 8 Rate based on the upper-middle-income poverty line: US$6.85 per person/day in 2017 PPP. 9 Structural peers: Tunisia, Bulgaria, and Costa Rica. Aspirational peers: Uruguay, Croatia, and New Zealand. 15 access to finance through the decontrol of interest rates– e.g., the credit to private sector increased from 19.6 percent of GDP in 2005 to 28.2 percent in 2019- supported private consumption. Meanwhile, net exports had a negative impact on growth, consistently with a widening trade deficit after the expiration of the ATC and despite a rise in mining exports. 7. Market liberalization also promoted tourism activities, which became a major source of external revenue and job creation. Between 2000 and 2019, tourism activity accounted for 9.6 percent of GDP on average, while job creation in the sector doubled. In 2019, the tourism sector created around 358,000 new jobs, of which more than 70 percent through indirect job creation. About 7 million tourist arrivals by air were recorded in the country in 2019, twice as many as in 2000. In the past 20 years, more than 90 percent of tourists have been non-residents, of which more than 80 percent have been foreigners. Figure 1.3. Tourism Revenues Figure 1.4. Job Creation (in thousand) % of GDP USD$ Million Direct Indirect 8,000 16.0% 400 358 7,000 14.0% 350 6,000 12.0% 300 5,000 10.0% 250 4,000 8.0% 200 167 3,000 6.0% 150 2,000 4.0% 100 1,000 2.0% 50 0 0.0% - Source: BCRD. World Bank’s staff calculations. 8. Rising consumption boosted domestic and import demand for manufactured goods and services. Between 2005 and 2019, on average, services accounted for 60 percent of GDP, and a 3 percent rise in GDP. Notably, transport, commerce, and hospitality made up a combined 24 percent of GDP. Construction and manufacturing contributed, on average, 0.7 and 0.6 percentage points to GDP growth, respectively. Within manufacturing, local activities contributed more to economic growth (0.5 percentage points) than Special Economic Zones (SEZ) activities (0.1 percentage points), consistently with the declining contribution of exports—local manufacturing focuses on supplying domestic consumption, while SEZ firms focus mostly on exports. Effectively, consumption growth uplifted local manufacturing, as well as demand for imported merchandise. 16 Figure 1.5. Demand-Side Figure 1.6. Supply-Side Contribution to GDP Growth Contribution to GDP Growth (Percentage, by component) (Percentage, by component) Private Consumption Public Consumption Agriculture Manufacturing Construction Hotels & rest. GFCF Net Exports 10.0 Trans & Comm. Other Serv. GDP Growth Finance Commerce 6.4 8.0 Mining GDP Growth 6.1 8.0 6.4 5.5 6.1 6.0 5.5 3.7 6.0 4.0 3.7 4.0 2.0 2.0 0.0 0.0 -2.0 -2.0 2000-2005 2005-2010 2010-2015 2015-2019 2000-2005 2005-2010 2010-2015 2015-2019 Source: World Bank, “The Dominican Republic: An Overview of Macroeconomic Performance”, background note Box 1.1. Structural reforms and stabilization policies have been key to economic growth The Dominican Republic pursued major economic reforms since the late 20 th century, such as liberalizing foreign exchange transactions and trade, eliminating price controls, and jettisoning restrictions to foreign direct investment (FDI). A new Foreign Investment Law, approved in November 1995, opened up key sectors, including the banking sector, to foreign investment, extended to foreign participants the guarantees granted to domestic investors, and eliminated all restrictions on profit remittances and capital repatriation. In addition, the government put in place major reforms to the banking system to strengthen the financial system and eliminate distortions in credit markets. In late 1991, the structure of reserve requirements for commercial banks was unified at 20 percent for all deposits and selective portfolio requirements were abolished. Furthermore, beginning in 1993, significant progress was made in developing banking supervision and prudential regulation. On a similar note, favorable geographical and external economic conditions also propelled growth. The Dominican Republic’s economy is highly responsive to growth in the G7 countries,10 through remittance inflows that fuel consumption, and tourist arrivals that boost demand for hotels and related services. Indeed, proximity to the North American markets and a natural endowment of ocean beaches have been at the root of the development of a strong tourism sector. Moreover, the textiles industry—and especially manufacturers established in Special Economic Zones (SEZs)—benefited from the Multi-Fiber Arrangement (MFA) and its successor, the Agreement on Textiles and Clothing (ATC) (see Box 1.2. for more details on SEZs in the Dominican Republic). However, such benefits vanished after the ATC’s expiration in 2005. Despite the introduction of the Dominican Republic-Central America Free Trade Agreement (CAFTA-DR) in 2006, which fostered exports of medical devices, footwear, pharmaceuticals, and plastic products, the value of trade flows (exports plus imports as percentage of GDP) fell from 62 percent of GDP in 2005 to 51 percent in 2019. 10The G7 are Canada, France, Germany, Italy, Japan, the UK, and the US. The Dominican Republic’s estimated growth elasticity to growth in G7 economies is 1.7, the highest in the LAC. Conversely, its estimated elasticity to growth in China is only 0.1. 17 Figure 1.1.1. Growth Elasticities Figure 1.1.2. Business Cycle Synchronization (Response on one percent increase on GDP growth of (Hodrick-Prescott Cycle, 2000Q1-2019Q4) China and G-7, 2000Q1-2019Q4) 1.60 USA_Cycle DOM_Cycle 1.40 With respect to real GDP growth in 6.0% 1.20 ARG 4.0% 1.00 2.0% 0.80 URY PER BRA 0.0% China 0.60 NIC PRY DOM -2.0% 0.40 COL CHL CRI SLV HND -4.0% 0.20 GTM BOL -6.0% 0.00 MEX -8.0% -0.20 2000Q1 2001Q2 2002Q3 2003Q4 2005Q1 2006Q2 2007Q3 2008Q4 2010Q1 2011Q2 2012Q3 2013Q4 2015Q1 2016Q2 2017Q3 2018Q4 -1.00 -0.80 -0.60 -0.40 -0.20 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80 With respect to real GDP growth in G7 countries Source: Haver, WDI, Bloomberg, FRED and BCRD. World Bank's staff calculation. References IMF (2002). The Dominican Republic. Stabilization, Reform, and Growth 9. Domestic consumption sustained imports of consumer goods and commodities. Total merchandise exports declined from 17 percent of GDP in 2005 to 13 percent in 2019, primarily because of the expiration of the MFA/ATC. During the same period, mining exports increased from 6 percent to 18 percent of total exports, while “national” exports (i.e., excluding exports from SEZs and the mining sector) remained flat at 3 percent of GDP. On the other hand, increasing demand for intermediate inputs from local manufacturers translated into higher imports of commodities, which rose from 4 percent of GDP in 2005 to 9 percent in 2019. Therefore, total merchandise imports increased from 21 percent of GDP in 2005 to 23 percent in 2019. Figure 1.7. Merchandise exports Figure 1.8. Merchandise imports (Percentage of GDP) (Percentageof GDP) percentage of GDP percentage of GDP 1 2 30 16 2 8 14 1 1 2 25 6 8 1 3 9 7 6 12 0 2 7 7 4 3 4 0 2 2 2 2 20 4 4 10 8 7 5 13 12 8 7 11 11 11 10 9 8 8 8 15 6 6 10 10 10 9 9 9 8 8 9 8 7 7 7 5 7 4 6 10 4 4 4 3 4 4 4 3 4 3 3 4 3 3 3 3 5 5 6 5 5 2 3 4 3 4 4 6 8 8 8 7 7 7 7 6 7 7 7 7 7 7 3 3 3 3 3 3 0 0 National Special Economic Zones Mining Consumption Capital Commodities Petroleum Source: BCRD. World Bank's staff calculation. 18 10. An attractive return on capital and a high labor force entry rate accelerated the rapid accumulation of input factors that has supported growth. A combination of special tax regimes, tariff exemptions, reduced import tariffs, and relatively low capital stock compared with advanced economies contributed to a rapid increase in the internal rate of return on capital stock—from 5.1 percent on average during the 2000s, to 16 percent in recent years— attracting FDI into the country. These economic incentives were introduced, in part, by the structural reforms of the late 1990s and early 2000s (see Box 1.1.). As a result, FDI rose from 1.9 percent of GDP in 1990 to 3.2 percent in 2019. Meanwhile, the labor force entry rate (i.e., the net entry of workers) averaged 2.8 percent between 2000 and 2019, significantly higher than the LAC average, and more than twice as high as in the U.S. (0.8 percent). Foreign labor is an important component of the workforce, as immigrants—87 percent of them from Haiti—accounted for 5.6 percent of the country’s population in 2017.11 Figure 1.9. Real Internal Rate of Return of Figure 1.10. Labor Force Entry Rate Capital Stock (Change/Employ, percentage) (Percentage) 1990-2000 2000-10 2010-19 1990-2000 2000-2010 2010-2019 18% 3.2% 15% 12% 2.4% 9% 1.6% 6% 0.8% 3% 0% 0.0% Source: World Bank, “The Dominican Republic: An Overview of Macroeconomic Performance”, background note 11. Economic activity in SEZ promoted job creation and attracted FDI but has had limited spillovers to the rest of the economy. SEZs concentrated, on average, 3.8 percent of total employment over 2010-19, and provided 176,555 direct jobs as of 2019 (CNZFE, 2019). FDI into SEZs increased from 4 percent of total FDI in 2010 to 9 percent in 2019. Total sales from SEZs accounted, on average, for nearly 3 percent of GDP over 2010-19, but the contribution of SEZs to growth and job creation in the wider economy has been limited. A 1 percent increase in SEZ output leads to an estimated average increase in output in the rest of the economy of only 0.6 percent – calculated as the average value of the backward linkages of SEZs, similar calculation applies for other sectoral categories such as services or local manufacturing-, and to a 0.5 percent increase in indirect employment creation. Meanwhile, local manufacturing has had a greater impact on growth and job creation. As an example, a 1 percent increase in local manufacturing production leads to an estimated average increase in production in the rest of economy of 1.4 percent, and to a 1.7 percent increase in indirect job creation. 11 ONE (2018). “Segunda Encuesta Nacional de Inmigrantes: ENI-2017” 19 Figure 1.11. Backward linkages effects on Figure 1.12. Backward linkages effects on economic growth, 2016 employment creation, 2016 4 4 3 3 ln (employment share/median of ln (employment share/median of 2 2 employment share) employment share) 1 1 0 0 -1 -1 -2 -2 -3 -3 -4 -4 -5 -5 0 0.5 1 1.5 2 2.5 0 0.5 1 1.5 2 Backward linkages, indirect effects Backward linkages, indirect effects Agri Local Manufacturing Services SEZs Agri Local Manufacturing Services SEZs Source: Gomez and Majluta (2021) and Input Output Tables from the Central Reserve Bank of Dominican Republic (BCRD). World Bank calculations. Note: A backward-linkage value higher than one implies that a 1 percent rise in production in a given sector would increase production or employment in the wider economy by more than 1 percent, in average. See Appendix 1.1 for technical description of backward linkages calculation. Box 1.2. Special Economic Zones (SEZs) in the Dominican Republic: An historical review The SEZs are geographically bounded areas in which customs, tax, and investment regulations are more favorable than in the rest of the economy. Broadly, the promotion of SEZs attempts to reduce the cost of doing business by addressing infrastructure deficiencies, bureaucratic complexity, and other barriers to business stemming from trade, fiscal, tariff, and labor policies.12 The first Dominican SEZ—La Romana—was established in 1969 to attract local and foreign investment, provide training, promote the transfer of technology and know-how, and ultimately create jobs in an economically disadvantaged area on the border with Haiti. To help achieve these objectives, dedicated legislation granted generous tax incentives to companies located in SEZs. During the 1980s and 1990s, Dominican SEZs benefited from geographical proximity to the US, preferential trade agreements, and investment incentives. In the early 2000s, SEZ-produced textiles, favored by quotas under the ATC, accounted for a third of Dominican exports. The phasing out of the ATC, completed at the beginning of 2005, led to a decline in the Dominican textile industry. By 2020, the Dominican Republic counted 75 SEZs (Parques de Zona Franca) hosting 551 firms. The SEZs are primarily located in and around the provinces of Santiago (27 SEZs in the province itself, and seven in neighboring provinces) and Santo Domingo (18 SEZs in the province itself, and 11 in neighboring provinces). Therefore, the benefits from the establishment of SEZs have been geographically concentrated in two out of the country’s 10 states—Ozama and Cibao Norte. 12 Aggrwal, A. (2006). Special Economic Zones: revisiting the policy debate. Economic and Political Weekly, 4533-4536 20 Figure 1.2.1. SEZ Firms’ distribution in the territory, percentage of total firms per province in 2019 Source: Directory of Companies and Establishments (Directorio de empresas y establecimientos, DEE), 2019. References The World Bank (2022). Paving the way for prosperous cities and territories. Urbanization and Territorial Review of the Dominican Republic. c. Yet, growth has been uneven 12. While growth has led to poverty reduction, income inequality increased, and real wages stagnated. The top 1 percent of the household income distribution accounted for more than 30 percent of GNI in 2019, nearly 10 percentage points more than in Costa Rica, and 17 percentage points more than in Uruguay. This indicator of income inequality has deteriorated over time, increasing from 27 percent in 2012. Moreover, average real wages have flattened in recent years, increasing by nearly 2.7 percent per year between 2000 and 2010, but slowing down to an annual growth of 1.4 percent between 2010 and 2019.13 Figure 1.13. LAC Share of Gross National Figure 1.14. GNI concentration, 2012-2019 Income (GNI) by Percentile Group, 2019 (Percentage) (Percentage, selected countries) < Percentile 50 Percentile 50 - 90 Percentile 90-99 > Percentile 99 11.4% 9.8% 9.5% 9.0% 9.3% 6% 9% 12% 7% 12% 10% 11% 10% 8.5% 8.4% 24% 24% 29% 27% 26% 25% 30% 30% 6.6% 47% 46% 44% 45% 42% 43% 42% 41% 24% 23% 21% 20% 22% 18% 16% 16% 2012 2013 2014 2015 2016 2017 2018 2019 Source: ECLAC based Household Surveys. Note. Chile corresponds to 2017 and Mexico to 2018. 13 Real wages approximated as the deflated wage compensation divided by number of employees. 21 13. Job creation occurred mostly in low-salary sectors. The employment rate climbed from 54.6 percent in 2000 to 61 percent in 2019, while the share of formal employment rose from nearly 48 percent of total employment in 2000 to 51 percent in 2019. However, formal employment—for which information is available—concentrated in commerce (16 percent of total formal employment in 2019), manufacturing (14 percent), hotels & restaurants (6 percent), and other services (15 percent)—sectors whose combined average wage is 12 percent lower than the average salary in the formal labor market. Figure 1.15. Employment and Unemployment Figure 1.16. Wages vs Share of Total Formal Rates Employment (Percentage of labor force) (2019, by economic activity) Employment Unemployement (RHS) 25,000 23,000 Mining 62 8 60 7 21,000 Average wage in RD$ Transport & Communication 58 6 19,000 5 Finance 56 17,000 Other serv. 4 Construction 54 15,000 3 Manufacturing 52 2 13,000 Edu. & Health Commerce 50 1 11,000 Agriculture Hotels & rest. 48 0 9,000 2000 2003 2006 2009 2012 2015 2018 -2 3 8 13 18 Employment participation share Source: Directory of Companies and Establishments (Directorio de Empresas y Establecimientos, DEE). World Bank, The Dominican Republic: An Overview of Macroeconomic Performance, background note 14. A large male-female labor force participation gap persists with negative impacts on poverty and productivity. Although 35 percent of women entering the labor market had a bachelor’s degree as of 2019, versus 17 percent of men, the labor force participation rate among women is 25 percentage points lower than among men. This gender participation gap mirrors a wage gap. After controlling for years of schooling, age, work experience, and household conditions—among other observable factors—the wage gap between men and women reached an estimated 16 percent in 2019, and 27 percent in the informal labor market. Closing the wage gap could potentially reduce the poverty rate by more than 7 percentage points. 14 14 MEPyD (2022). Probabilidad de Transición y Trampas de la Pobreza. Forthcoming. 22 Figure 1.17. Employment rate by gender Figure 1.18. Education Attained by gender, (Percentage) 2019 (Percentage of the labor force by gender) Male Female Male Female 74.7 73.2 70.2 69.2 71.1 40% 38% 36% 35% 48.6 25% 38.8 41.8 35.7 35.9 17% 5% 3% 2000 2005 2010 2015 2019 Elementary High School Bachelor None Source: Background note prepared for the report “The Dominican Republic: An Overview of Macroeconomic Performance” 15. Growth has also been spatially uneven. The country’s growth model has favored the region of Ozama, where the capital city, Santo Domingo, is located. Estimates of subnational GDP15 show that on average, the regions of Ozama and Cibao Norte accounted for nearly 56 percent of national GDP between 2015 and 2019. Meanwhile, Enriquillo and El Valle contributed just 2 and 3 percent to national GDP, respectively. The estimated GDP per capita in Ozama is roughly twice as high as in Enriquillo, further evidencing sizable regional disparities. The case of the Yuma region, where the major tourist resort of Punta Cana is located, is notable: although the region is not among the largest contributors to GDP, it is as rich as Ozama in per capita terms. Figure 1.19. Average structure of GDP by Figure 1.20. GNI per capita, 2018 Region (Atlas method, current USD) (Percentage of national GDP, average 2017-2019) Ozama 41.2% current USD Cibao Norte 14.9% Valdesia 8.8% 8,000 Cibao Sur 7.9% 6,000 Yuma 7.0% 4,000 Cibao Nordeste 5.5% 2,000 Higuamo 4.7% 0 Cibao Noroeste 3.3% El Valle 3.3% Enriquillo 2.4% Extra-region 1.0% 0.0% 10.0% 20.0% 30.0% 40.0% 50.0% Source: MEPyD (2022). “Aproximando el PIB a nivel regional en República Dominicana”. World Bank Staff’s calculations. 15 MEPyD (2022). “Aproximando el PIB a nivel regional en República Dominicana”. 23 d. Higher productivity growth is needed to sustain growth 16. Contributions to growth from productivity and quality of education have declined over time despite a rapid accumulation of capital and a growing labor force. Capital formation has been the main driver of economic growth, accounting for 3.7 percentage points of GDP growth on average between 2005 and 2019. Labor quantity was responsible for a non-negligible 1.6 percentage-point rise in GDP over the same period, but labor quality only contributed 0.4 percentage points. After the banking crisis of 2003/2004, total factor productivity (TFP)’s contribution to GDP growth has been non-negative—unlike in most of the LAC—but it has not increased in lockstep with per capita income. Figure 1.21. Growth Decomposition Figure 1.22. Output and TFP Evolution (Production factors, percentage points) (Index 2000=1) Labor force Labor quality Capital TFP 3.10 Output per Employed Person Output per Hour Worked 0.0 0.1 0.4 0.3 2.70 Per Capita Income 4.1 TFP 3.0 2.30 5.8 4.0 4.1 1.90 1.7 1.8 0.4 1.1 0.3 1.50 -2.7 1.10 0.70 2000-2005 2005-2010 2010-2015 2015-2019 2019 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Source: World Bank, “The Dominican Republic: An Overview of Macroeconomic Performance”, background note Box 1.3. Defining Productivity In the growth literature, economic productivity is approximated by Total Factor Productivity (TFP), typically known as the Solow residual. TFP represents efficiency in the use of resources (capital, labor, and intermediate inputs) in the production process. Effectively, it accounts for the share of growth that cannot be explained by growth in observable production factors, such as labor and capital. Therefore, TFP reflects unobservable factors that can boost growth, such as managerial skills, advanced technologies, and innovation, as well as knowledge spillovers from new companies entering the market. As an example, the productivity growth of a restaurant in Santo Domingo could be measured as the increase in the number of mofongos—a typical plantain-based dish—prepared without extending working hours or using more ingredients. What could explain the increase? The chef may have learned new waste-management techniques to optimize the use of ingredients, and/or he could have improved his performance thanks to an upgraded kitchen. References Syverson, Chad. “What Determines Productivity?” Journal of Economic Literature 49, no. 2 (2011): 326–65. 24 17. Higher TFP would have resulted in higher real wages. Global experience shows that TFP is positively correlated with real wages. Thus, economies with higher productivity also have higher real wages. Relative to the levels of aspirational peers, with higher productivity level, the Dominican Republic’s real wages are, on average, 70 percent lower. Faster TFP growth would thus have the double benefit of reducing the country’s dependence on factor accumulation to sustain its high growth rate and addressing its real wage stagnation problem. Figure 1.23. Real Wages and Total Factor Productivity (Historical correlation between 1990 and 2019) Source: Penn World table 10.0. World Bank’s staff calculations. 18. Despite the increasing accumulation of factors of production, a continued decline in productivity could slow down GDP growth to under 4 percent in the long run. The World Bank’s Macro-Fiscal Model (MFMOD) estimates that the current downward trend in TFP has the potential to thwart GDP growth. For example, with a 0.5 percentage point decrease in TFP growth, GDP growth could not maintain the current rate of approximately 5 percent per year. In a more negative scenario, an average TFP decline of 2 percent per year would lead to long-run growth below 3 percent by 2050. Conversely, an increase in TFP by 1 percentage point above the current average could lead to average GDP growth of 7 percent by 2050, enabling the country to rapidly catch up with aspirational peer economies. 25 Figure 1.24. GDP Growth under different scenarios (Percentage points) Baseline TFP Growth -0.5% 13 TFP Growth +1% TFP Growth -1% 12 TFP Growth -2% 11 10 9 8 7 6 5 4 3 2 1 0 2021 2025 2029 2033 2037 2041 2045 2049 Source: World Bank, “The Dominican Republic: An Overview of Macroeconomic Performance”, background note e. Conclusions 19. The Dominican Republic will need to accelerate productivity growth to reach US per capita income levels. The country’s rapid economic growth has been supported by factor accumulation (capital and labor) fueled by favorable investment incentives such as tax exemptions and a growing labor supply, partly fueled by regional migration. However, low productivity growth, in the absence of overdue structural reforms, have contributed to real wage stagnation. By accelerating TFP growth to one percentage point above historical average, the average GDP growth would increase up to 7 percent in the long run. Conversely, a one percentage point decline in TFP growth would drive GDP growth towards 3 percent, compromising the country development objectives. 20. Higher firm productivity in employment absorbers sectors complemented with upgrading of skills can result in higher wages. Low firm productivity growth constrains gains in real wages. At the same time, inadequate human capital prevents limits firm productivity growth which would induce higher real wages. While enhancing firm productivity, it is also important to unlock constrains blocking labor mobilization, which provide sufficient bargaining power to the labor force to be compensated for their productivity. In this regard, it is important to promote competitive markets across the territory to creates equal job opportunities. The following two sections provide a detailed analysis on these two points, increasing firm productivity and promoting competitive markets. 26 f. Appendix Appendix 1.1. Country peers The CEM uses the same definition of structural peer countries as in the SCD (Systematic Country Diagnostic) 2018 in order to compare Dominican’ performance, of which Bulgaria is included per its similarity in population size and GDP per capita. We narrow the list of structural peers to 3 countries by using the ‘CEM 2.0’ tool, which is mainly based on the World Development Indicators (WDI) database. Structural Peers Under this classification, the tool identifies countries with similar economic or demographic performance as Dominican Republic. The analysis relies on a proximity statistical method by using a pre-defined set of variables. The set of variables selected for the period 1991-2019 are – the following: i) GDP per capita (constant 2011 USD PPP dollars), ii) Foreign Direct Investment (net inflows, as percentage of GDP), and iii) Population size (in millions) The analysis suggests the following countries (in order of proximity): Tunisia, Costa Rica and Bulgaria. Aspirational Peers This classification presents countries that possess similar structural conditions but have evolved and overperformed when compared to Dominican Republic. Specifically, based on the aforementioned set of variables, the tool identifies the countries that has achieved a remarkable progress in a specific target variable. For the period 1991-2019, the target variable is: GDP per capita (constant 2011 USD PPP dollars). This classification delivers the following countries (in order of distance to the best performance): Uruguay, Croatia and New Zealand. 27 Appendix 1.2. Calculating Backward linkages Backward linkages (BL) represent the relationship of the industry “i” with its supply chain—i.e., the dependency of the industry on other sectors in its network. Therefore, BL incorporate direct, indirect, and induced effects. Specifically, BL include the effects from an increase in demand of input from sector “i” on the same sector “i”, as well as the effects from sector “j” to sector “i”. The BL for either output or employment effects can be calculated as follows: ∑ , (A.1.1) = ∑ (∑ ) , where , is the (i,j) element of the matrix M. BL provide a direct comparison among sectors in terms of potential capacity to create employment and growth. By using the diagonal matrix E, populated with number of jobs per unit of output, the employment multiplier is obtained: e M =EM (A.1.2) Notice that by ignoring the diagonal terms (i,i), BL would only capture indirect effects. 28 Chapter 2. Reversing the productivity decline a. Introduction 21. The Dominican per capita income has rapidly converged towards US per capita income but slowing productivity growth poses risks to sustain economic growth. Despite the rapid GDP growth in the last two decades, productivity growth has continuedly declined. The Dominican’s total factor productivity (TFP) level represented 70 percent of US levels in 1990, but in the absence of further reforms it declined by 10 percentage points by 2019. 22. Therefore, this chapter seeks to understand the productivity dynamics in the Dominican Republic by accounting for sectoral differences. There are three channels through which productivity can be affected, the within-firm (capabilities, innovation, etc.), the between (allocation of resources across firms in a sector) and net entry components (entry of productive and exiting of unproductive firms in the market). The importance of those channels varies across sectors. While the between component in the construction and transport sectors negatively contributed to productivity growth, a negative contribution emerges from the net entry component in the manufacturing and hospitality sectors, which offset the benefits from innovation. 23. Market distortions, including tax exceptions, have prevented the optimal allocation of resources across sectors, mainly affecting the productivity growth of large firms. The productivity outperformance of small and medium-sized firms in the services sector could indicate that lack of incentives to innovation, poor allocation of resources, and/or gaps in managerial capability are more severe in large firms. In addition, tax exceptions have benefitted, mostly, less productive large firms. A dataset of formal firms operating in similar activities but with some favored by special tax regimes (STRs), such as the ones located in Special Economic Zones (SEZs), is analyzed in this chapter. The key finding is that productivity of low-tech firms in SEZs is nearly 30 percent lower than comparable firms outside the SEZs. 24. Climate change also represents a risk to firm productivity. The Dominican Republic is the world’s 12th most-affected country by natural hazards over the 1998-2017 period, according to the 2019 Global Climate Risk Index, with hurricanes and tropical storms recurrently causing high human and economic losses. 25. Finally, this chapter estimates the growth benefits of improving human capital, fostering productive investment, increasing competition in SEZs, and promoting a transition to renewable energy. Such reforms may increase long-run growth by 1.4 percentage points above its current trend, boost the average income of the poorest households by nearly 17 percent by 2030, and almost double income by 2050, thus substantially reducing poverty. 29 b. Productivity dynamics 26. Productivity growth in the Dominican Republic benefited from the structural reforms introduced in the early 1990s, but subsequently declined. The Total Factor Productivity (TFP) in the Dominican Republic, which was equivalent to 70 percent of US levels as of 1990, caught up with them by the mid-1990s. However, due to the absence of further reforms and worsening economic distortions (see Box 2.2.), it fell back to 60 percent of US productivity levels by 2019, well below both structural and aspirational peers (see Box 2.1 for productivity estimation details). Alternative measures such as output per worker or output per hour worked relative to the US level, points out the subdue performance of productivity growth compared to the GDP per capita growth. 27. The channels through which firm-level productivity is affected, vary significantly per sector. Productivity changes can be broadly decomposed into three components: (1) within-firm, (2) between, and (3) net entry (see Melitz and Polanec, 2015; Maloney and Cusolito, 2018). The within-firm component refers to the increase in the capabilities of each firm, the between component focuses on the allocation of resources and market share to more productive firms, and net entry reflects the exit and entry of productive and/or unproductive firms. In the Dominican Republic during the 2015-19 period, the between component in the construction, hospitality and transport sectors negatively contributed to productivity growth. Meanwhile, a negative contribution from the net entry component offset the benefits from innovation in the manufacturing and hospitality sectors. In addition, the contribution of the within-firm component has been very limited across sectors, except in hospitality. Overall, the between component is predominately negative in services-related activities. Figure 2.1. Total Factor Productivity (TFP) Figure 2.2. Productivity decomposition, 2015- (as percentage of US productivity) 2019 (Contribution per component) Structural Aspirational Dominican Republic contribution 1.1 0.15 1 0.1 0.9 0.05 0 0.8 -0.05 0.7 -0.1 0.6 -0.15 0.5 Construction Hospitality Manufacturing Transport Within Between Net Entry Source: A background note prepared for the report “Understanding productivity growth in a small open economy: the case of the Dominican Republic”. Productivity decomposition follows the methodology proposed by Melitz and Polanec (2015). In addition, the deflated value-added per firm is considered output, while wage compensation represents the labor force weighted by human capital, and stock of capital by fixed assets – including machinery. 30 Box 2.1. Measuring Productivity To measure TFP, we impose a production function governed by certain conditions, such as increasing monotonicity between output production and factor accumulation. The literature typically considers the Cobb-Douglas and Constant Elasticity of Substitution (CES) production functions. In both cases, the production function can take the form of: = ( , , ) (B.2.1) is output production, the stock of capital, the number of employees, and the intermediate inputs, such as commodities. The remaining term captures TFP. While the empirical estimation of has been at the center of debate in growth literature, the methodology proposed by Ackerberg, Caves, and Frazer (2015) is the most widely accepted when utilizing firm-level data. References Ackerberg, Daniel A., Kevin Caves, and Garth Frazer. “Identification Properties of Recent Production Function Estimators.” Econometrica 83, no. 6 (2015): 2411–51. 28. Productivity growth of small and medium-sized formal firms have outperformed of large firms. Between 2015 and 2019, the average productivity advantage of small and medium- sized formal firms over large firms amounted to 35 percent in construction, 45 percent in hospitality, and 38 percent in transport.16 Factors such inadequate competition environment, mismatch of workers skills with the required by firms, human capital deficiencies, among others, might be responsible for such productivity dynamics. For example, a Dominican freight transport industry group complained in 2017 about alleged monopolistic practices imposed by the national federation of transport contractors (FENATRADO), which deliberately constrained the ability of contractors not affiliated with FENATRADO to participate in the market. 17 On the other hand, the presence of a productivity premium for small and medium-sized firms is less visible in the formal manufacturing sector, where economies of scale are an important natural advantage for large firms. The mechanism through which the aforementioned factors could have affected firm dynamics is further explored in the remainder of this section. 16 In the absence of price information, productivity estimates are based on revenues productivity (TFPR), which might differ from actual productivity when prices are known (TFPQ). See Cusolito and Maloney (2018) for a detailed discussion. 17 https://www.competitionpolicyinternational.com/r-dominicana-fenatrado-acusada-de-monopolizar-transporte/ 31 Figure 2.3. TFP distribution across formal firms by size, variables in logs (Probabilities restricted by similar firm’s characteristics and location) a. Construction b. Transport c. Hospitality d. Manufacturing Source: A background note prepared for the report “Understanding productivity growth in a small open economy: the case of the Dominican Republic”. Note. Size classification follows the legal classification of the country, while firm level information has been provided by the National Statistical Office (ONE) and represents only firms with tax ID (formal firms). Firm data is relevant at sector level, but not at regional level, preventing similar analysis from being performed per region. In addition, the deflated value-added per firm is considered an output, while wage compensation represents the labor force weighted by human capital, and stock of capital by fixed assets – including machinery. 29. Substantial growth benefits can emerge from addressing domestic market imperfections. A calibrated model for the Dominican Republic (see Appendix 2.1. for a detailed description) suggests that domestic distortions (monopolies and cartels, as well as market and bureaucratic barriers to entry—e.g., complex regulations, shortcomings in human capital and domestic transport, and high energy costs, among others, see Box 2.2.) played a greater role in the country’s productivity decline than external distortions (e.g., tariff and non-tariff barriers, subsidies, and trade costs). If the level of domestic distortion were reduced by 1 percentage 32 point, GDP could rise by 1.46 percentage points above the current level, while a 1 percentage point fall in the level of external distortion would only boost GDP by 0.21 percentage points. 30. Removing distortions in non-tradable sectors and backbone services18 could increase GDP by 1.13 percentage points, versus a 0.33 percentage-point rise from removing distortions in non-backbone tradable sectors. A recent World Bank analysis indicates that limited competition hinders the optimal use of telecommunication services in the Dominican Republic, which lags peer countries in mobile subscriptions and internet penetration. 19 Reducing the level of distortion in the telecommunications sectors (e.g., high infrastructure costs, suboptimal tariff regulation and entry regulatory constraints) by 1 percentage point can add 0.21 percentage points to the current GDP growth trend. Similarly, removing distortions in the energy sector can potentially unlock another 0.09 percentage-point increase in GDP, by reducing blackouts that disrupt manufacturing activity. The removal of distortions in non-backbone tradable sectors can also unlock growth benefits, but of a lower magnitude. For instance, a 1 percentage point reduction in distortions in the metals industry can boost aggregate GDP by about 0.12 percentage points, while the same reduction of distortions in textiles and clothing manufacturing would increase GDP by 0.03 percentage points. Figure 2.4. Estimated GDP elasticities with Figure 2.5. Estimated GDP elasticity from internal respect to the removal of economic sectoral distortions for the top 20 sectors in the distortions, 2016 economy, 2016 (in percentage points) 1.6 1.46 Telecommunications 0.2077 1.4 Base metal fabrication 0.1155 Energy 0.0985 1.2 Trade 0.0763 Financial and insurance activities 0.0683 1 Accommodation and Food and… 0.0682 Exploitation of mines and quarries 0.0640 0.8 Transport and storage 0.0635 Oil refining 0.0408 0.6 SEZ - Manufacture of textile products,… 0.0366 Manufacture of textiles, clothing 0.0366 0.4 Livestock, Forestry and Fishing 0.0318 0.21 SEZ - Manufacture of alcoholic and… 0.0205 0.2 Manufacture of alcoholic and non-… 0.0205 Health 0.0056 0 Education 0.0056 Elasticity external Elasticity internal Building 0.0037 distortions distortions Other Crops 0.0007 Cultivation of tobacco and beverages… 0.0007 Sugar cane cultivation 0.0007 0 0.05 0.1 0.15 0.2 0.25 Source: Zentner and Parro (2019). 18 Backbone services in general comprise energy supply, communication, distribution, transportation, finance and insurance services and form the foundation of the economy. 19 World Bank. 2021. Dominican Republic: Leveraging Competition in the Telecom Sector to Accelerate Economic Growth. World Bank, Washington, DC. 33 Box 2.2. Economic distortions Economic distortions have been widely associated with misallocation of resources, which prevents the optimal utilization of input factors—i.e., labor and/or capital. Distortions can be of a market or non- market nature. For example, elevated fixed costs that discourage the entry of innovative firms in the telecoms sector are a market barrier, hampering the flow of ideas into the sector. On the other hand, onerous bureaucratic obstacles to obtaining an operation license from a municipality are a non-market barrier to the entry of productive firms. Government interventions via state-owned enterprises (SOEs) can also cause misallocation of resources, as SOEs do not always need to be productive to survive in the market. Finally, informality can be another source of distortions. Busso, Fazzio, and Levy (2012) find that informal firms in Mexico are much less productive than formal firms, yet command a large share of resources, hence contributing to the very low overall productivity in the country. References Restuccia, Diego and Richard Rogerson. “The Causes and Costs of Misallocation” Journal of Economic Perspectives 31, no. 3 (2017): 151–74. c. Factors constraining productivity growth i) Labor market skills inefficiencies and demand-supply mismatches 31. Human capital indicators are lagging in the Dominican Republic, negatively affecting firm performance. The OECD’s 2018 PISA test showed poor educational outcomes in the Dominican Republic relative to both structural and aspirational peer countries, including those within the LAC region. The Dominican Republic ranked, on average, 23 positions below its structural comparators, and 43 positions below its aspirational comparators. Similarly, on the Human Capital Index (HCI), the Dominican Republic ranks below other Central American countries, apart from Honduras and Guatemala. Inadequate human capital has also affected the performance of firms located in SEZs, where 19 percent of firms considered human capital one of the main bottlenecks to private sector development.20 20The World Bank (2022). Paving the way for prosperous cities and territories. Urbanization and Territorial Review of the Dominican Republic. 34 Table 2.1. PISA 2018, Global Rankings and Scores Figure 2.6. Human Capital Index (HCI)21 (Score=average score across mathematics, science, (Central America, 2018) and reading) Ranking Score 0.65 Dominican Republic 77 334 0.60 Costa Rica 58 414 0.55 Structural Bulgaria 50 426 0.50 Peers Tunisia N/A N/A 0.45 New Zealand 15 502 0.40 Aspirational Dom. Rep Honduras Panama Nicaragua Costa Rica Guatemala El Salvador Croatia 37 417 Peers Uruguay 52 423 Source: OECD and World Bank Human Capital Index database. World Bank Staff’s calculations. 32. While the share of workers with a secondary degree is at par with other peers, firms cannot hire workers with the skills they need. Approximately 70 percent of the labor force has completed secondary school, a share 3 percentage points higher than structural peer countries, and only 2 percentage points below aspirational peers. However, for services firms where more than half of the labor force holds a bachelor’s degree, labor productivity is roughly half the level of aspirational peers and 44 percent lower than in structural peers. A national survey of the private sector reveals that nearly 10 percent of firms consider that inadequate human capital represents a constraint for business development, the fourth main constraint in the survey. More importantly, nearly 70 percent of firms indicate that their labor force requires training to strengthen technical skills to improve job performance. The percentage of firms requiring skills upgrading of their labor force is 82 percent in electricity, 81 in the communication sector, 74 percent in the retail trade sector (or commerce), and 73 percent in hospitality. The absent of adequate skills in the labor market, complicates the hiring process. Interestingly, those constraints are more predominant in large firms compared to small and medium-sized firms filling. Specifically, within firms with more than 250 employees, 54 percent of firms report having difficulties to fill a vacancy, while only 41 percent within micro firms (those with number of employees ranging between 16 to 19). 21The HCI measures the contributions of health and education to a worker’s productivity. The index score ranges from zero to one and indicates the future productivity of a child born today relative to the potential productivity achievable if he/she enjoyed full health and complete education. 35 Figure 2.7. Labor productivity of firms with at Figure 2.8. Percentage of firms reporting to least 50 percenf of labor force educated require skills training, per sector, 2020 Structural Aspirational Dominican Republic Yes No 11.5 59.5% Agriculture Transport 63.9% 11 Construction 64.6% Labor Productivity (Log) 10.5 Manufacturing Industries 68.8% Water 73.2% 10 Hotels, Bars,&Rest 73.3% 9.5 Mining 74.0% Trade 74.4% 9 Communications 81.0% Electricity 82.1% 8.5 All sectors Manufacturing Services 0.0% 50.0% 100.0% Source: Enterprise Survey Database and National report of skills and employment (ENDHACE). World Bank Staffs calculations. Figure 2.9. Firms difficulty to fill a vacancy Figure 2.10. Firms with difficulties to fill a vacancy by size Yes No 16 to 29 40.5% 59.5% 30 to 49 36.2% 63.8% Employees 50 to 99 45.8% 54.2% 100 to 249 48.4% 51.6% 250 or more 54.2% 45.8% 0.0% 50.0% 100.0% Source: National report of skills and employment (ENDHACE). World Bank Staffs calculations. 33. In particular, skilled operators, craftsmen, and professionals are scarce, particularly for large firms. The absence of necessary skills to carry out the task required by the position and lack of manual dexterity are the main reasons preventing the adequate recruitment of new workers. Likewise, 11 percent of firms indicates that absence of skills for solving complex problems configure a constraint for filling a vacancy, while nearly 10 percent points out to factors such as reading comprehension and ICT basic abilities. Such constraints exacerbate the difficulties for recruiting adequate professionals and operators required by the firms, which is significantly worse for large firms. 36 Figure 2.11. Skills required and missed for filling Figure 2.12. Difficulties to fill a vacancy by a vacancy occupation percentage 100 7 6 13 7 12 80 60 58 48 53 40 52 40 16 13 19 20 19 11 18 19 22 23 9 0 16 to 29 30 to 49 50 to 99 100 to 250 or 249 more Management Professionals & techical Adm.& assistants Operators and craftsmen Elementary occupations Agricultural workers Source: National report of skills and employment (ENDHACE). World Bank Staffs calculations. 34. The difficulties of finding the right candidates affect the workload of current staff, as well as the probability of introducing new products in the market. Nearly 32 percent of firms point out that the difficulty for hiring the right candidates generates an increasing workload for current staff, affecting labor productivity of such workers; while 14 percent of firms indicate that an inadequate recruitment process prevents the development and introduction of new products in the market. Similarly, 9 percent of firms suggest that such constraints affect the quality of the products offered to consumers. About 18 percent of the surveyed firms indicate that the absence of relevant training programs in the country is the main reason preventing workers’ skills upgrading, while 15 percent highlight insufficient time within working hours to conduct training sessions, and 12 percent point to the lack of relevant courses in local institutions and universities. Figure 2.13. Consequences of inadequate Figure 2.14. Main reasons preventing adequate recruitment training percentage percentage Increasing workload 31.8 No relevant training program 17.9 Increasing business costs 16.1 Absence of time for the training 14.9 No introduction of new products 13.7 Absence of availability of courses 12 Monetary losses 9.9 No need for training 10.1 Product quality for consumers 9.2 No availability of relevant courses 9.4 Introduction of managerial… 7.2 Conflict with courses schedules 7.6 Defficiency consumer support 4.5 No budget 6.6 Outsourcing of personnel 3.4 Workload preventing training 3.4 No development of products 1.7 Absence of relevant information 3.3 No new technologies 0.4 No a priority for the firm 2.7 0 20 40 0 10 20 Source: National report of skills and employment (ENDHACE). World Bank Staffs calculations. 37 35. Despite missing skills, the labor force participation rate has significantly increased over time, but workers show significant learning gaps relative to their US peers. The number of active workers increased by 8 percent per year, on average, between 2005 and 2019, but those workers had an average learning gap of four years relative to their US peers—i.e., despite having been in school for a similar length of time, their skill level was behind that of workers educated in the US by the equivalent of four years of schooling. The gaps are smaller for agricultural workers and machinery operators, both occupations requiring a specific and limited skillset. The low skill level of workers entering the labor market may have enhanced the bargaining power of employers and contributed to the slow growth of wages in the country (Jobs Diagnostics, 2020). On the other hand, 10 percent of surveyed firms in a national census report that education inefficiencies are one of the main constraints for business development and is the top constraint on secondary constraints for business development. Overall, labor market deficiencies can be explained by both significant learning gaps in the current labor supply, and mismatches between knowledge and skills acquired during education training during school and college, and skills required by firms to improve business performance. Whether the root of the education problem lies only on the protracted learning gap or mismatch of skills between supply and demand of labor goes beyond the scope of this report. Figure 1.15. Learning Gap22 (Estimated as the difference between USA years of education and DR quality adjusted years of education) 9.0 8.0 Professionals 7.0 Average Growth 2005-2019 Managers Agricultural 6.0 Technicians Workers Clerical 5.0 4.0 Craft traders Machine 3.0 operators 2.0 1.0 Elementary Services Workers workers - -2 -1 0 1 2 3 4 5 6 Learning Gap Source: Background note prepared for the report “Sectoral, Spatial and Labor Market Aspects of Structural Transformation in the Dominican Republic” 22Learning-adjusted years of schooling for DR was calculated based on the aggregate results showed in the HCI 2020 update. On which expected years of schooling is 11.9 and learning adjusted 6.6. While to calculate the years of education in the USA, the educational levels variable had to be transformed into years assuming Postsecondary nondegree awards (8 years); high school diploma or equivalent (12 years); some college, no degree (13 years); Associate degree (14 years); bachelor’s (16 years); master's (18 years) and doctoral or another professional degree (21 years) 38 Figure 1.16. Obtsacles constraining business development Main obstacles Secondary obstacles Employees with low academic level 33.5% Tax burden 25.0% 29.7% Competitors in the informal sector 14.7% Political instability 25.9% 25.0% Energy 11.2% Labor Mkt regulations 18.5% 17.8% Employees with low academic level 9.7% Crimes, robberies and disorder 17.1% Secondary 10.0% Tax payment procedures 6.9% Tax payment procedures 9.9% 9.6% Other Obstacles 6.1% Administrative obstacles 6.3% 6.2% Procedures for obtaining licenses… 4.2% Access to financing 5.1% Crimes, robberies and disorder 4.1% 4.4% Transport access for the company 3.8% Principal Customs regulations and foreign trade 3.7% 3.4% Judicial system 2.0% Political instability 3.6% 0.0% 10.0% 20.0% 30.0% 40.0% Land access 2.5% Transportation and logistics of… 1.9% Administrative obstacles 1.7% Access to financing 1.6% Transport access for the company 1.4% Labor Mkt regulations 1.2% Judicial system 0.4% 0.0% 10.0% 20.0% 30.0% Source: National report of skills and employment (ENDHACE), 2020. National Statistical Office (ONE). World Bank Staff’s elaboration. ii) Non-creative innovation 36. Although the majority of Dominican firms report conducting innovation-related activities, the resulting innovations are not novel. Nearly half of small firms and 78 percent of large firms reported innovation-related activities between 2018 and 2019. Nevertheless, approximately 96 percent of the innovations introduced cannot be considered novel, either in the country or in the relevant market (the figure is similar when only considering product innovations). In fact, the number of patent registrations has been falling since 2010, when they reached the highest level since 2000. Specifically, the number of patents registered per year stood at 312 in 2019, down from 406 in 2010. The decline is broadly explained by the fall in patents of international origin, which dropped from 371 in 2010, to 270 in 2019. Overall, innovation in the country seems to focus narrowly on imitation, which has limited productivity spillovers, compounded by a declining inflow of innovative ideas from abroad. 39 Figure 2.17. Percentage of firms that have Figure 2.18. The type of innovation performed, reported innovated by size, 2018-2019 2018-2020 (Percentage of total firms per bracket) (Percentage of total innovation) 2.7 1.7 percentage of firms 90 78.3 80 68.3 70 64.8 58.1 60 50.2 50 40 30 20 95.5 10 New for firm but not the market 0 16 to 29 30 to 49 50 to 99 100 to 249 250 above New for the market but not in country employees New for the country but not in the world Figure 2.19. Patents registration by origin Figure 2.20. The type of product innovation (Number of patents) performed, 2018-2020 (Percentage of total product innovation) Number of patents 4.2 7.4 400 350 300 250 200 150 100 50 88.3 0 New for firm but not the market 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 New for the market but not in country Of domestic origin Of international origin New for the country but not in the world Source: ONE (2021) and ONAPI. World Bank Staff’s calculations. 37. Inadequate information about new technologies prevents technology adoption. About 83 percent of the surveyed firms financed innovation-related activities out of their own resources during 2018-2020, while 13 percent resorted to borrowing from the banking sector. However, inadequate information regarding the technology intended to be adopted, organization rigidities such as structure and procedures, and the scarcity of qualified workers are reported to be the main bottlenecks for adopting new technologies. The absent of relevant information can also reflect limited managerial capabilities. In both cases, there is a room for the government to ensure correct dissemination of technologies and improving workers’ skills. 40 Figure 2.21. Sources of financing for innovation Figure 2.22. Main reasons preventing innovation, efforts, 2018-2020 2018-2020 percentage percentage 0.6 1.8 1.8 Absence of information 21.1 Organization rigidities 18.3 12.5 Excessive competition 16.7 No qualified workers 14 Elevated costs 9.2 Uncertainty on outcomes 8.2 Access to legal services 5.7 83.3 No government support 4.4 Inadequate financing sources 2.5 Own resources Banking credit Business union loans Government credit 0 5 10 15 20 25 Others Source: ONE (2021). World Bank Staff’s calculations. d. The implications of special tax regimes (STRs) 38. Evidence suggests that larger firms – not necessarily the most productive - are the main beneficiaries of tax incentives, while smaller and more productive firms are subject to a higher tax burden. Highly productive firms pay 0.4 percentage points more in income tax than firms in the low-productivity segment in the Dominican Republic. Consequently, firms that register in special tax regimes remain in the market longer than those that do not, despite their productivity level. As an example, during the period 2007-2016, the exit rate of low productivity firms under general tax regime (GR) is roughly 9 percent, while under special tax regime (STR) for firms located in SZE is 7 percent. Meanwhile, the entry of productive firms under the GR is 2.3 percent, while only 0.6 percent for firms under the STR. Figure 2.23. Effective Income Tax Rate by Figure 2.24. Entry and Exit Rate by Productivity Productivity Segment, 2017 Segment (2007–2016) Entry rate GR Entry rate STR 10.0% Exit rate GR Exit rate STR 8.0% 6.0% 4.0% 2.0% 0.0% Low High Low High productivity productivity productivity productivity 2007 2016 41 Source: Azuero et al (2019), “Productivity, Misallocation, and Special Tax Regimes in the Dominican Republic”. Note. GR means General tax regime, while STR, Special Tax Regime applied to firms located in SZEs. Productivity measured as TFPR (Total Factor Productivity Revenue) 39. As a result, formal firms in low-tech industries granted with heterogenous tax incentives have lower productivity. By comparing productivity differences of formal firms located in Special Economic Zones (SEZs) that benefit from special tax regimes (STR), with similar peers located outside SEZs, such as low-tech producers of textiles & clothing and metal products, it is found that those located in SEZs tend to accumulate production factors, such as labor and capital, instead of growing based on efficiency gains. Specifically, productivity in those industries is nearly 30 percent lower compared to peers that are under the general tax regimes (GTRs). This does not necessarily mean that aggregate productivity in SEZs is lower than in the rest of the economy, since a number of firms operating in the former do not have counterparts in the latter. For example, medical device firms can only be found in the SEZs – while accounting for only 5 percent of the total number of SEZ firms, they generate nearly 30 per cent of SEZ exports. Figure 2.25. TFP kernel distribution for Manufacturing Firms, SEZ vs non-SEZ (Probabilities restricted by similar firm’s characteristics and location) a. TFP (in logs) b. Real wages (in logs) Source: A background note prepared for the report “Understanding productivity growth in a small open economy: the case of the Dominican Republic”. TFP calculated by assuming trans-log production function. The y-axis represents the density, while x- axis the productivity level. In addition, the deflated value-added per firm is considered output, while wage compensation represents the labor force weighted by human capital, and stock of capital by fixed assets – including machinery. iii) Climate change 40. Due to its geographic position and island status, the Dominican Republic is highly vulnerable to natural disasters such as droughts, flooding, hurricanes, landslides, and wildfires. The 2019 Global Climate Risk Index ranked the Dominican Republic 12 th among the countries 42 most affected by natural disasters between 1998 and 2017, with hurricanes and tropical storms causing high and recurrent human and economic losses. Key assets such as electricity generation, transmission, and distribution infrastructure are exposed to disaster risk, while 35 percent of the country’s transport network is vulnerable to extreme weather events.23 The estimated annual average loss (AAL) imposed by hurricanes on the country’s building stock is valued at US$345 million (0.5 percent of GDP).24 The vulnerability level in 13 provinces (around 40 percent of the total) is high or very high.25 26 The Santo Domingo metropolitan area has both the country’s largest built-up surface in flood-prone areas, and the highest rate of growth of built-up surface in such areas,27 which make it particularly vulnerable to weather-related disasters and climate- related events. Such events, including hurricanes, have a regressive distributional effect, as they affect most heavily the poor and vulnerable. Figure 2.26. Hurricane risk in the Dominican Figure 2.27. Vulnerability to climate change Republic. in the DR excluding the tourism sector. Source: The World Bank (2018) and GoDR (2018). Note: The larger the circle, the higher the Annual Average Losses (AAL) that the province could potentially incur over the long term (absolute risk). The darker the color, the higher the ratio of AAL/Province Exposure (relative risk). 41. Climate change has raised average temperatures and sea level, while reducing precipitations. Since the 1960s, the average annual temperature in the Dominican Republic has increased by approximately 0.45C, with a significant rise in the number of hot days and hot nights. Between 1960 and 2003, the average annual number of hot days increased by 63 days, while the average annual number of hot nights increased by 48 nights. By mid-century, the 23 World Bank (2022). Consolidating the Recovery: Seizing Green Growth Opportunities. Semiannual Report for Latin America and the Caribbean. 24World Bank (2018). Hurricanes and Earthquakes Risk Profile, Dominican Republic. 25Elements deemed to increase vulnerability to climate change include exposure to hydrometeorological events, low adaptive capacity, and degradation of the territory. 26 Government of the Dominican Republic (2018). Third National Communication to the United Nations Framework Convention on Climate Change. 27 World Bank (2022). Paving the way for prosperous cities and territories. Urbanization and Territorial Review of the Dominican Republic. 43 country’s average annual temperature is projected to rise from 24.5°C to 25.9°C in a high emissions scenario (Shared Socioeconomic Pathways scenario, SSP5-8.5). By the end of the century, the average temperature is expected to increase by between 2.5 and 5 C in a high emissions scenario, with the number of days hotter than 35 C in a month rising sharply from the 2080s—primarily between May and October. The DR has also experienced a slight reduction in total precipitation. By mid-century, the country is likely to suffer a further decrease in rainfall and an increase in the number of consecutive dry days, primarily from May to August. 28 At the same time, climate change is expected to cause higher-intensity windspeed and storm surge. 29 By 2050, coastal areas are likely to experience a sea level rise of +0.3 meters in a high emissions scenario (Representative Concentration Pathways scenario, RCP8.5), with effects including coastal erosion, ecosystem degradation, population displacement, water pollution, and water-supply disruption.30 42. Rising temperatures are already affecting firm productivity in manufacturing. 31 Historical information reveals that rising temperatures between 2015 and 2020 were associated with productivity losses of between 2 and 9 percent among manufacturing firms located in the Dominican Republic’s poorest regions. As GDP per capita in those regions is roughly one-third of level in the richest regions, climate change is further complicating the path to shared prosperity by widening income inequality. Inter-sectoral linkages can amplify microeconomic shocks,32 either through backward or forward linkages.33 As an example, rising temperatures can increase the nominal cost of cotton production, an important component for the clothes industry. Such higher cost could result in either a reduction of the quantity produced of clothes, or in input misallocation causing labor or capital wedge costs, and dampening productivity. Table 2.2. Historical productivity effects in manufacturing firms from increasing temperatures, period 2015-2020 Log of Output per worker Log of Firm productivity (TFP) (1) (2) (3) (4) Avg log of temperature 0.137 0.223 -0.396 -1.737* (1.573) (1.574) (0.529) (0.915) Log of temp*poor -0.124*** -0.300*** -0.0185** -0.0906*** (0.0308) (0.0339) (0.00725) (0.0266) Log of precipitation 0.0613 0.339 0.0907 0.166 (0.336) (0.369) (0.0803) (0.148) Constant 12.46** 10.54 15.55*** 18.93*** (6.170) (7.397) (2.134) (2.997) Observations 808 1,119 808 1,119 R2 0.124 0.102 0.030 0.020 28 World Bank (2022). Climate Change Knowledge Portal. 29 Holland, G., Bruyère, C.L (2014). Recent intense hurricane response to global climate change. Clim Dyn 42, 617–627. https://doi.org/10.1007/s00382-013-1713-0 30 UNFCCC and WHO (2021).Health and Climate Change. Country Profile 2021. 31 An exhaustive analysis of the growth and development challenges from climate change is under preparation in the Country Climate and Development Report (CCDR) for the Dominican Republic, which is expected to be finalized during the second half of 2023. 32 Carvalho and Gabaix, 2013 33 Acemoglu, Carvalho, Ozdaglar and Tahbaz-Salehi, 2012 44 Year fixed effect Yes Yes Yes Yes No Yes No Yes Region fixed effect Source: ENAE database. World Bank calculations. The regression is based on firm-level information, where productivity is estimated using the trans-log production function approach of Ackerberg, Caves, and Frazer (2015), and matched with temperatures at regional level. The regression follows a fixed-effect specification, with regional and time-varying dummies. Note: Poor regions include Enriquillo, El Valle, and Valdesia. *** significance at 1 percent, ** significance at 5 percent and * significant at 10 percent. Standard deviation is regional, clustered via the bootstrapping method. e. Summary: closing the growth and income gaps 43. Structural reforms aiming to close the productivity gap in the Dominican Republic could boost economic growth by up to 1.4 percentage points over its baseline path by 2050. Relevant reforms include improvements to education and human capital, a boost to productive investment, enhancements to market competition and productivity in SEZs, and a transition to renewable energy sources (see Figure 2.28 for more details of the assumed scenarios). To analyze the effects of such reforms on the Dominican Republic’s growth path, a computable general equilibrium (CGE) model was constructed based on the country’s input-output matrix, social accounting matrix, and other relevant indicators. Figure 2.28. Proposed Scenarios •Educated workers productivity •The public investment increases by 0.5 pp improves by 0.2% per year, while 0.3% as percentage of GDP, while 2.5 pp in for non-educated workers in the private investment and 1.7 pp of FDI. The eperiod 2022-25. Then 0.4% and 0.6% TFP increases as per the associated in 2026-40, respectively for skilled and investment plan by 0.1%, year average, Education unskilled. Lastly from 2041-50 it Boosting Productive due to public investment and 9.5% due to Improvement increases 0.6% and 0.9% for skilled and Investment private investment. unskilled. •Equalizing productivity from different •Subsidies are faces out over 5 years sizes firms. A drop in markups is •The rise in TFP in production of introduced. renewables results in a decline in average Special Economic •By 2050 total factor productivity unit costs of renewables of 25% by 2040 Zones (SEZs) increase by 30 percent in Special Renewable Energy and by 50% by 2050 relative to baseline. Productivity Economic Zones. Transition Improvements 44. A bundle of coordinated reforms can make the greatest impact on economic growth, while helping to reduce income inequality. The implementation of the whole package of reforms described above could increase the income of the poorest workers by nearly 40 percent above the current trend by 2050, with a beneficial effect in terms of lower income inequality. Considering the impact of specific components of reform, education improvements and efforts to boost productive investment can each add 0.3 percentage points to GDP growth by 2030, 45 relative to the baseline scenario. SEZ productivity improvements would increase GDP growth by 0.1 percentage points above the baseline in the long run. The energy transition would have a neutral impact on economic growth under this model, while making the growth path more sustainable, while creating new employment opportunities and bringing health benefits to the population. Figure 2.29. GDP Impact Figure 2.30. GDP Growth Impact (Percentage change with respect to the baseline (Percentage change with respect to the baseline scenario of non-reform) scenario of non-reform) 2030 2050 2030 2050 0.300 1.2 0.250 1.0 0.200 0.8 0.150 0.6 0.100 0.4 0.050 0.2 0.000 0.0 Figure 2.31. Wages Impact, 2050 Figure 2.32. Income Distribution, 2050 (Percentage change with respect to the baseline (Percentage change with respect to the baseline scenario of non-reform) scenario of non-reform) Skilled formal labor Skilled informal labor First quintile Second quintile Unskilled formal labor Unskilled informal labor Third quintile Fourth quintile 0.80 Fifth quintile 0.60 0.4 0.40 0.3 0.20 0.2 0.00 0.1 -0.20 0.0 46 f. Appendix Appendix 2.1. Calculating sectoral distortions Following Caliendo, Parro, and Tsyvinski (2018), the production function for a country “i” in sector “j” is determined by the following Cobb-Douglas production function 1 (B.1.1) = where is TFP, the number of employees in sector “j”, and the intermediate inputs used by firms in the sector “j”. Meanwhile, is the share of value added on the nominal output. A Constant Elasticity Substitution (CES) production function aggregates the intermediate goods (1 )/ (B.1.2) = 1 , , where “n” indicates the country from where intermediate inputs are imported, while “k” indicates the specific sector from which inputs are imported. Finally, indicates the demand preferences. The consumer preferences in the country are given by 1 (B.1.3) 1 ( ) = where are the consumptions goods from sector “j” in country “i”, such that is the aggregated consumption basket in the economy with demand shifters affecting consumer preferences for each product “j” and assumed to be time invariant. Assuming a cost , associated with importing goods from country “n”, the share of inputs ( , ) from country “n” sector “k”, in the intermediate consumption of country “i”, sector “j”, can be expressed as follows (1 ) 1 (B.1.4) , , = / where the price paid for material inputs, and the average salary in the economy. Finally, distortions can be expressed by ,i (B.1.5) ,i = ,i References Caliendo, Lorenzo, Fernando Parro, and Aleh Tsyvinski. 2022. "Distortions and the Structure of the World Economy." American Economic Journal: Macroeconomics, 14 (4): 274-308. 47 Appendix 2.2. Computable General Equilibrium (CGE) model The Computable General Equilibrium (CGE) model used in this study is based on the Mitigation, Adaptation, and New Technologies Applied General Equilibrium (MANAGE) model developed by the World Bank. The MANAGE is a single-country CGE model that relies on a neo-classical structural modeling approach. Most of the model’s assumptions follow the standard CGE literature. Extended documentation and a user guide for the model can be found in Mensbrugghe (2019). This appendix briefly explains the main features of the MANAGE model. Production activities in the MANAGE model are profit maximizers under constant returns to scale technologies. They use labor, capital, land, and intermediate inputs to produce commodities and services (which we will be referred to as commodities from here on) for domestic and international markets. The production function is nested, with a constant elasticity of substitution (CES) production function in value added nests, and a Leontief technology at the intermediate input nest. The CES production function allows for substitution of factors in a specific nest, while the Leontief technology assumes a fixed ratio between factors. Thus, using a nested production structure allows for the use of different substitution elasticities among factors. In the top nest, value-added and aggregate non-energy intermediate inputs are combined, following a Leontief production technology. This creates a link between sectors, as the output of a sector is an input for others. At the second level, the composite intermediate input is obtained by combining all non-energy intermediate inputs with a Leontief technology. The value-added composite aggregates the capital composite factor and other factors of production (labor and land). The last nest combines energy and capital with a CES production function, making them substitutable. Demand for factors and intermediate inputs, as well as the output level, are determined according to the production nest. One of the novelties of the MANAGE model is the ability of production activities to determine the energy intensity of production endogenously, based on energy prices. This distinction is important when analyzing carbon pricing policies. Introducing carbon pricing is likely to raise the cost of energy, which in this framework would incentivize substituting capital with energy. The intuition behind this mechanism is that firms are likely to invest in energy-efficient technologies to use less energy, and hence substitute capital with energy. The MANAGE model also has a vintage capital structure, where old and new capital are treated differently in terms of substitutability with energy. New capital is substitutable with energy, while old capital is a near complement. That is, the vintage capital structure captures the semi-putty/putty relations across inputs, with more elastic long-run behavior as compared with the short-run. Energy production in this version of the MANAGE model accounts for five types of electricity generation activities: Coal, Gas, Oil, Hydro, and Renewables. The electricity generation mix is endogenously determined, based on the relative cost of each generation activity. Alternatively, the model allows for targeting of a specific energy generation mix, by adjusting the investment in each type of generation (e.g., increasing investment in renewables to follow a renewable energy target). All markets in the model are perfectly competitive, implying that prices are equal to marginal costs in the equilibrium. Thus, firms compete with each production activity compete with others in the factor markets to hire labor and capital. The model accounts for four types of labor (skilled formal, unskilled formal, skilled informal and unskilled informal), one type of capital, and one type of land in. Labor and land supply are determined by a supply function that is sensitive to average wage and land price, respectively. Labor 48 supply is also segmented across sector groups. Hence, movement of labor across those sector groups is limited. This is achieved by introducing a CET function which drives the supply of labor to the sector groups based on relative wages across sector groups and an elasticity of substitution. This enables the model to mimic labor market rigidities. Capital supply is the result of a capital accumulation process, where shrinking activities release capital that is added to “new” capital stock. New capital is fully mobile across sectors. This allows to mimic some rigidities in the capital market, as movement of capital from a declining sector to an expanding sector is limited. The rate of return on capital is the same across expanding sectors, and lower in declining sectors. The model encompasses ten representative household types, according to income deciles. Households are the owners of factors of production. They supply labor depending on real wages: higher wages induce more labor supply—thus, ignoring the wealth effect on labor supply, which would entail reducing the labor supply at very high levels of real wages. Income sources other than factor income for households are income and transfers from the government and the rest of the world. Households spend their income on consumption, savings, and direct taxes. The distribution of consumption across commodities is determined by a two-level utility function. At the first level, a Constant Difference in Elasticities (CDE) utility function determines the consumption of aggregated commodities. The use of CDE enables a better representation of income effects on household demand, by allowing consumption shares to change as income and prices change—unlike other functional forms such as Linear Expenditure System (LES) or CES demand functions, which assume that expenditure shares are constant and independent from household income. The aggregated commodities groups are food, manufacturing, energy, services, and transport. Effectively, the first-level utility function distributes household consumption spending across the broader categories. Then, a second level CES nest distributes spending on each aggregate consumption group among commodities in that group. For example, the energy group consists of coal, refined petroleum, coke, electricity, and natural gas. The government does not account for behavioral assumptions and is entirely neutral. It collects taxes and receives transfers from the rest of the world and domestic agents, then spends them on savings, government consumption and investment, and transfers to rest of the world. The government can borrow from domestic institutions or from the rest of the world but must pay interest on debt later. All tax rates are fixed at base-year levels. The volumes of the government’s current and investment spending are also fixed. This implies that government savings (primary balance) are endogenous and adjust to clear the government balance. The gap between government investment demand and public savings is satisfied through foreign and domestic borrowing. Alternative government closures can be considered for the simulations of fiscal reforms. For example, there can be a target for government budget balance, and a ‘swing’ fiscal instrument—such as personal income taxes—adjusts to achieve the target. The rest of the world (ROW) exports from and imports to the Dominican Republic according to Constant Elasticity of Transformation and Armington specifications, respectively.34 Both specifications assume that domestic commodities are not perfect substitutes for traded commodities. Thus, imports and exports are determined by the difference between domestic and world prices, which are considered to be fixed in line with the small open economy assumption. ROW also makes transfers to domestic agents, and receives 34This version of the model considers only one trade partner, the rest of the world. However, the wider model allows for additional trading partners to be added in a two-level nested structure. 49 transfers from them. Such transfers are assumed to be constant as a share of GDP. Lastly, ROW makes investments in the Dominican Republic, which correspond to foreign-exchange flows for investment purposes (e.g., FDI or short-term capital movements). The model follows a savings-driven closure where aggregate investment is flexible and equal to the available volume of savings. Foreign savings are exogenous and fixed as a share of GDP, while government and household savings are endogenous. In effect, the rate of return on capital adjusts to equalize investments and savings. Hence, the model reflects a crowding-out effect, whereby government investment displaces private investment. The model dynamics follows the neo-classical growth framework (Solow-Swan growth model), whereby the long-run growth rate of the economy is determined by three main factors: capital accumulation, labor supply growth, and increases in productivity. The stock of capital is endogenous, while the latter two are exogenously determined. The capital stock in each period is the sum of depreciated capital from the previous period and new investments. For each type of labor, the maximum stock of labor available in each period grows exogenously, based on population projections by age cohort and cohort-specific participation rates. The model is calibrated to a new Social Accounting Matrix (SAM), built for this report, which relies on 2016 Supply and Use tables.35 Macroeconomic indicators (e.g., national accounts, trade, government accounts) also refer to 2016. 35The latest version of the SAM published by the Dominican Republic’s central bank dates to 2013. However, in the context of the CEM, the World Bank worked with the Ministry of Economic Planning and Development (MEPyD) to update the SAM to 2016. 50 Chapter 3. Enhancing market competition to boost growth b. Introduction 45. Competition policy has the potential to support productivity and economic growth in the Dominican Republic. Competition fosters cost reductions, innovation, productivity, and economic growth (Bassanini and Ersnt, 2002; Acemoglu et al., 2007; Aghion and Griffith, 2008; Bloom, Draca and Von Reenen, 2011), shifts market share toward more efficient producers; and induces firms to become more efficient to be successful (Kitzmuller and Licetti, 2013; Buccirossi et al., 2009; Voigt, 2009). Competition might also support the functioning of three drivers for economic and inclusive growth by i) inducing firms to take productivity-enhancing actions, such as product or process upgrading – the “within-effect” (Aghion et al, 2004), ii) promoting a better allocation of resources towards more efficient producers and sectors – the “between-effect” (Syverson, 2011), and iii) improving market dynamics, allowing less efficient companies to exit the market while creating space for more efficient firms – the “selection effect” (Eslava et al., 2013). Some of these effects can also vary depending on the institutional environment (Aghion and Howitt, 2006), the quality of the legal system (Buccirossi, Ciari, Duso, Spagnolo, & Vitale, 2009), openness to trade and rigidities in labor markets (Aghion and Griffith, 2008), among others. 46. Domestic markets in the Dominican Republic are perceived as being less competitive than domestic markets in comparator countries; and dominated by few groups of firms . Based on the Bertelsmann Stiftung's Transformation Index (BTI) indicators,36 the perception of the degree of competition in domestic markets in the Dominican Republic is less favorable than in peer countries such as Bulgaria and Costa Rica (structural peers) or Uruguay and Serbia (aspirational peers). This is particularly the case regarding whether there is a competition- enabling environment as well as the strength of the antitrust policies in place. Furthermore, WEF Global Competitiveness Report data confirms that local competition is perceived to be low, and markets seem to be dominated by few firms. 47. This chapter analyzes the competition environment and potential sources of market distortions that can hinder competition. There are, potentially, two channels through which government interventions can shape the incentives for firms to enter, compete, and grow. First, the direct government participation in the market through the role of state-owned enterprises (SOEs). Second, the indirect channels where the government can foster or deter competition through two different roles. On one side, as a regulator setting the rules under which firms can enter and compete, and also as a market referee that disciplines markets and enforces antitrust rules to deter, prosecute, and sanction anticompetitive practices. 36The BTI is a perception indicator based on in-depth assessments of countries and is managed by Bertelsmann Stiftung. The responses reflect the situation in the country at the end of January 2021. 51 c. Understanding the competition environment in the Dominican Republic 48. Domestic markets in the Dominican Republic are perceived as being less competitive than domestic markets in comparator countries, albeit slightly above the LAC average. Based on the Bertelsmann Stiftung's Transformation Index (BTI) indicators,37 the perception of the degree of competition in domestic markets in the Dominican Republic is less favorable than in peer countries such as Bulgaria and Costa Rica (structural peers) or Uruguay and Serbia (aspirational peer). Figure 3.1. Competition perception using BTI 10 9 8 7 6 5 4 3 2 1 0 OECD avg. Dominican Republic Uruguay Croatia Costa Rica Tunisia Serbia Bulgaria LAC avg. Structural peers Aspirational peers Market-based competition (higher value = better competition-enabling environment) Anti-monopoly policy (Higher value = stronger policy in place) Source: World Bank staff elaboration based on Bertelsmann Stiftung's Transformation Index BTI, 2022. Note: The BTI is a perception indicator based on in-depth assessments of countries and is managed by the Bertelsmann Stiftung. The responses reflect the situation in the country at the end of January 2021. Score 1–10 (best). 49. In addition, markets in the manufacturing sector are perceived as relatively concentrated as compared to structural peers. Over 46 percent of the firms surveyed in the Dominican Republic stated that they operate in concentrated market structures (i.e., monopoly, duopoly, and oligopoly), compared to an average of 30 percent of firms in Uruguay and Croatia,. 38 As of 2016, the level of perceived concentration in the Dominican Republic was about three times the level perceived in structural peers such as Bulgaria (12.4 percent) and Tunisia (12.4 percent).39 37 The BTI is a perception indicator based on in-depth assessments of countries and is managed by Bertelsmann Stiftung. The responses reflect the situation in the country at the end of January 2021. 38 Estimates from the WB Enterprise Survey (2016). 39 It is important to note that while government regulations and interventions that disrupt the marketplace by limiting entry, facilitating dominance, or un-levelling the playing field may explain the high market concentration, concentrated market structures can also be due to natural barriers, small market size, or firms operating more efficiently because of economies of scale. 52 Figure 3.2. Market structure in LAC Figure 3.3. Competition perception in DR 60% 49.5% 50% 46.5% 40% 30% 20% 10% 0% 2010 2016 Source: World Bank staff elaboration based on World Bank’s Enterprise Survey for most recent years. Note: The shares reflect the percent of responding establishments that answered “None”, “One”, “2-5” or “More than 5” or whose responses were classified as such to the question “For fiscal year [indicated in parenthesis], for the main market in which this establishment sold its main product, how many competitors did this establishment’s main product/product line face?”, respectively. For example, “None” was coded as “Monopoly” and “One” as "Duopoly". Establishments with no answers to the question and establishments whose main market for its main product line is external are excluded. 50. Firm-level data from administrative records confirm that several activities in the manufacturing sector are highly concentrated in the Dominican Republic. Market concentration (Herfindahl-Hirschman index HHI) and dominance (Kwoka) indices show that manufacturing sectors are still highly concentrated and some show risks of dominant position.40 Interestingly, these correspond to manufacturing sectors that could be denoted as viable for competition based on the intrinsic market characteristics and in the absence of market failures. The most concentrated sectors in the Dominican Republic include manufacturing of oils of vegetable origin; basic iron and steel; dairy products; and beverages. In total, 11 out of 38 sectors analyzed are highly concentrated (HHI index above 2,500 points),41 and all are related to manufacturing activities. Furthermore, the Kwoka dominance index, which provides an estimate of the differences in size among competitors within a market suggests near-monopolistic structures in the manufacturing of petroleum and refinery products.42 However, high concentration does not necessarily translate into a lack of competition, as it might be related to natural market characteristics linked to the size of the markets and geographic location. 40 The Kwoka dominance index provides an estimate of the unequal size of the companies within a market to unveil the potential presence of monopolistic structures. This indicator ranges from 0 (highly competitive) to 1 (monopolistic structures). 41 The Department of Justice [WHICH COUNTRY?] establishes that markets with HHI index above 2,500 refer to highly concentrated market structures. 42 Estimates based on the tax administrative records provided by the DGII from the Dominican Republic. These estimates are conducted on the 3-digit sector classification as the maximum level of disaggregation available in the official data. These estimates are indicative of market concentration at the sectoral level, although do not necessarily correspond to a relevant product-market. Some of these markets are provided by the state as this note describes in the next section. 53 Figure 3.4. HHI- Concentration index 2017-2018 Figure 3.5. Kwoka dominance index 2017-2018 (0 to 1-monopolistic structure) Manufacture of Petroleum… Manufacture of Petroleum Refining… Production of Oils from Vegetable… Production of Oils from Vegetable… Manufacture of basic iron and steel Manufacture of basic iron and steel Manufacture of Dairy Products Manufacture of Dairy Products Manufacture of sugar Manufacture of sugar Manufacture of cocoa, chocolate… Manufacture of cocoa, chocolate and… Manufacture of beverages Manufacture of beverages Manufacture of ceramic products Manufacture of ceramic products Manufacturing of Soaps and… Manufacturing of Soaps and Detergents Manufacture of Cement, Lime and… Manufacture of Cement, Lime and… Exploitation of Mines and Quarries Exploitation of Mines and Quarries Communications Communications Manufacture of glass products Manufacture of glass products Conservation, Production and… Conservation, Production and… Manufacture of tobacco products Manufacture of tobacco products Banks, Associations, Credit Corp., Banks, Associations, Credit Corp., Exchange agents and/or… Exchange agents and/or remittance… Manufacturing of Chemical… Manufacturing of Chemical Substances Insurance Insurance Manufacture of grain mill products Manufacture of grain mill products Electricity, water and sewage Electricity, water and sewage Manufacture of Furniture Manufacture of Furniture Manufacturing of Pharmaceutical… Manufacturing of Pharmaceutical… Manufacture of Textiles and Apparel Manufacture of Textiles and Apparel Manufacture of Wood, Paper and… Manufacture of Wood, Paper and… Trade of fuels Trade of fuels Manufacture of baking products Manufacture of baking products Manufacture of plastics Manufacture of plastics Editing, Recording, Printing Editing, Recording, Printing Hotels and restaurants Hotels and restaurants Transportation and storage Transportation and storage Wholesale and retail of vehicles Wholesale and retail of vehicles Housing rental services Housing rental services Health services Health services Construction Construction Other services Other services 0 5,000 10,000 0.0 0.2 0.4 0.6 0.8 1.0 Source: WBG Markets, Competition, and Technology (MCT) using DGII administrative records 51. Highly concentrated markets tend to be more prone to anticompetitive practices and show relatively lower productivity. Although market concentration does not necessarily translate into market distortions, there is some evidence suggesting that highly concentrated markets in the Dominican Republic such as manufacturing of petroleum and refined products, oils, iron, and dairy are relatively less productive compared to markets where there is stronger market competition. This also shows that sectors with more restrictive regulations (in orange) also seem to be associated with higher concentration and lower productivity – suggesting reform opportunities through which enhanced competition could play a key role in driving productivity and economic growth.43 Moreover, fewer operators in these markets also raise the risks for 43Such potential relationship between productivity and market concentration is not causal. From the theoretical standpoint, competition may have both a positive and negative effect on productivity, as reflected in an r inverted-U shape relationship 54 potential anticompetitive conducts, both collusion and abuse of dominance, as well as potential negative effects of mergers and acquisitions. Figure 3.6. Sectoral market concentration HHI and firm productivity (TFP), manufacturing firms in 2017 1 Wholesale Log (Total Factor Revenue productivity) 0.8 0.6 ConstMnfMnftextMnf chem. furnit. Oth Serv Mnf tobac. Tranp. 0.4 Trade Mnf fuels pharm. Rem… Mnf soap Edit, print, Mnf wood Comm. Hotel. Mnf bak. Mnf iron Mnf glass Insurance Mnf plast. Mnf dairy 0.2 Proc. Elect, water, Meat gas Mnf bev. Mnf petrol. Mnf cement Mnf choc. Mining Housing Mnf grain 0 R² = 0.2156 0 1000 2000 3000 4000 Mnf 5000 ceramic 6000 7000 8000 9000 10000 Mnf oils -0.2 Market concentration (HHI) Note: Estimates of TFPR are based on the Hsieh and Klenow methodology (Hsieh & Klenow, 2009) using the administrative tax records from the DGII and following (Azuero, Bosch, Cardoza, & Sanchez-Navarro, 2019). Concentration indicators are estimated by the WBG MCT staff using DGII administrative records. In orange sectors with restrictive regulation based on the PMR results and sector specific assessments led by the WBG MCT staff. Source: WBG Markets, Competition, and Technology (MCT) using DGII administrative records d. Barriers to competition in the Dominican Republic 52. Government interventions can shape how markets operate and affect the incentives for firms to enter, compete, and grow through direct and indirect channels. Potential barriers to competition operate through two main channels. In the direct channel, the government acts as a market player through State-Owned Enterprises (SOEs). In the indirect channel, the government sets the market rules in its role as a regulator, and discipline markets, sanction, and deter anti-competitive behaviors in its role as a referee. This analysis is based on the OECD-WBG Product Market Regulation (PMR) indicators, evidence from the new WB Businesses of the State (BOS) database, and administrative information from the tax records in the Dominican Republic (DGII). (Aghion et. al, 2005). At lower levels of competition, firms closer to the technological frontier innovate to escape from competition, thereby increasing productivity. At higher levels of competition, this incentive fades. There is some evidence of a negative relationship between concentration and productivity in the Dominican Republic. 55 `Figure 3.7. Pillars for fostering market competition PILLARS FOR FOSTERING MARKET COMPETITION 1. Direct role 2. Indirect role Government as a regulator (economy- Government as a market player Government as a market referee wide and sectoral) Legal provisions and regulations that Tackle cartel agreements or anti- strengthen dominance: restrictions to Involvement of the state in the markets competitive conducts that raise the the number of firms, statutory (Direct role) as a competitor or single costs of key inputs and final products monopolies, bans towards private provider of goods or services and reduce access to a broader investment, lack of access regulation for variety of products essential facilities Foster capacity building and Interventions that are conducive to independence of regulatory bodies collusive outcomes or increase the costs Competitive neutrality and rules that and the competition authority. of competing: controls on prices, FDI can tilt the playing field. Strengthen the antitrust and restrictions, and other market variables that increase business risks. institutional framework to combat anticompetitive conduct (abuse of dominance) Source: WBG Global Markets, Competition and Technology Unit 53. Product-Market Regulation (PMR) indicators shed light on the extent to which government interventions restrict competition. The PMR database collected by the World Bank and the OECD offers internationally comparable indicators that measure the degree to which regulations (on the books) foster or limit firm entry and competition in areas of the product market where competition is viable. The PMR was collected for the first time in the Dominican Republic in 2014. As of 2014, according to the PMR results, the Dominican Republic had more restrictive product market regulations than most countries analyzed, albeit below the LAC average.44 The PMR shows that potential entrants and incumbent firms in DR faced more regulatory restrictions than aspirational peers such as Croatia and Bulgaria, although less so than in structural peers as Costa Rica. This in turn limits the extent of competitive pressure and contestability in key Dominican markets. The regulatory constraints are even more noticeable when the Dominican Republic is compared to the best-performing OECD countries: the economywide product market regulations are almost twice as restrictive than the top 5 OECD countries’ average. 44It is important to note that the PMR indicators do not reflect how laws and regulations are enforced. Hence, a country that has competition-friendly laws “on the books”, but that does not enforce such laws, would still obtain a favorable score. The PMR index is based on qualitative information about more than 1,000 features of economywide and sectoral regulations. The information is collected using a standardized questionnaire, scored from 0 (least restrictive) to 6 (most restrictive), and aggregated using standardized weights. 56 Figure 3.8. Dominican Republic PMR and peer economies, 2013-2017 Structural peers Aspirational peers 3.50 Upper middle income avg. = 2.28 3.00 High income avg. = 1.54 2.50 2.00 1.50 1.00 0.50 0.00 Costa Rica China Uruguay Mexico Ukraine Switzerland Spain Dominican Republic Slovak Republic Luxembourg Belgium United Kingdom Ecuador El Salvador United States Philippines Argentina Netherlands Honduras Kenya Panama Estonia Australia Bolivia Egypt Turkey Croatia Peru Sweden Lithuania Japan Italy Indonesia Rwanda Russia South Africa Guatemala Nicaragua Korea Romania Bulgaria Malta Ireland Finland Denmark Iceland Austria India Jamaica Paraguay Colombia Slovenia Poland Latvia Norway Canada Greece Chile Hungary Germany New Zealand Brazil LCR avg. France Czech Republic Portugal Israel Cyprus Senegal Source: PMR questionnaire for the Dominican Republic and OECD and WBG-OECD Product Market Regulation Database, 2013- 2017. Note: Absolute values from 0 to 6. Higher values are associated with more restrictive regulations. 54. PMR results suggest that about two thirds of the public sector-related barriers to competition stem from the indirect channel. According to PMR results, roughly 34 percent of the barriers to competition in the Dominican Republic are explained by the direct channel or state control in the form of State-owned enterprises (SOEs) in markets. Additionally, barriers to competition stem from the indirect involvement of the state in business operations in the form of complex and burdensome regulatory procedures and implicit barriers to trade and investment which insulate domestic incumbents from foreign competitive pressures. Figure 3.9. PMR in Dominican Republic Decomposition of PMR Score Decomposition of High Level Indicators 3.0 3.0 BTI: Other barriers to trade and investment 2.5 2.5 BTI: Barriers to Explicit barriers to trade and 20% trade and investment 2.0 investment 27% 2.0 BTE: 33% Regulatory protection of incumbents 1.5 Barriers to 30% BTE: entrepreneurs 1.5 Administrative burdens on 39% hip startups 1.0 66% BTE: 1.0 Complexity of regulatory State control procedures 0.5 67% SC: 34% 50% Involvement in business 0.5 operation 0.0 34% SC: Dominican Republic Public ownership 0.0 State control Barriers to Barriers to trade (SC) entrepreneurship and investment (BTE) (BTI) Source: PMR questionnaire for the Dominican Republic and OECD and WBG-OECD Product Market Regulation Database, 2013- 2017. 57 i) Direct channel: Government as a market player (SOEs) 55. The World Bank Businesses of the State (BOS) database shows the government plays an important role in the economy through State-Owned Enterprises (SOEs). 45 As of 2019, the government remained as a shareholder in 36 firms across a varied set of sectors, ranging from manufacturing of goods, refined petroleum, water utilities, electricity production and distribution to insurance and real estate activities. As of 2019, SOE revenues reached the equivalent of 7 percent of the GDP (USD 6.3 billion) and occupied over 2 percent of the formal employment (approximately 47,400 workers). 56. At least 40 percent of the SOEs in the Dominican Republic continue operating in activities that are viable for competition and private sector participation. SOE revenues in the Dominican Republic are relatively lower than the regional LAC average and below structural peers (e.g., Croatia and Uruguay) and structural peers (e.g., Costa Rica). However, at least 40 percent of these firms still operate in competitive sectors that could be viable for the private sector and where the risk of crowding out private competitors is high. Figure 3.10. Domestic SOE revenues of as percentage of 2019 GDP 25 Domestic revenues as % of GDP 20 15 10 5 0 Croatia Uruguay Paraguay Average LAC Costa Rica Dominican Peru Republic Structural peers Aspirational peers Note: Average for LAC includes only countries with over 60% coverage in revenues and employment information in the WB BOS database, namely, Bolivia, Uruguay, Paraguay, Costa Rica, Dominican Republic, Peru, and Paraguay. Only relevant aspirational or regional comparator countries are shown in the figure. Source: World Bank staff elaboration based on WB Businesses of the State (BOS) database 45The SOEs in the WB Businesses of the State (BOS) database refer to those entities with 10% or more direct or indirect state participation, that are legally separated entities, engaged in market production and operating for profit . For more information, see Dall'Olio, A., Goodwin, T., Martinez Licetti, M., Orlowski, J., Patino-Pena, F., Ratsimbazafy, F., & Sanchez-Navarro, D. Using ORBIS to Build a Global Database of Firms with State Participation (English). Policy Research working paper WPS 10261 WBG. http://documents.worldbank.org/curated/en/099800112132221252/IDU03d9586040b28504839081120922e33694f65 58 Figure 3.11. Distribution of SOEs in Dominican Republic by sector type, 2019 Note: Competitive sectors refer to competitive sectors characterized by small entry barriers; contestable sectors are characterized by moderate entry barriers, public goods, or externalities; and natural monopoly sectors are those that exhibit high entry barriers, economies of scale, or sub-additivity cost structures. Source: World Bank staff elaboration based on WB Businesses of the State (BOS) database and SOE sector taxonomy46 (Dall’ Olio et al., 2022b) 57. SOEs have a prominent role in key enabling sectors for growth in the Dominican Republic. SOEs are large and leading market players in partially contestable sectors in the Dominican Republic. For instance, the sector with the highest concentration index (HHI) corresponds to the manufacture of petroleum and refined products, where the SOE controls almost 98 percent of the market.47 SOEs are present across different network sectors, including segments that are typically opened to competition (e.g., electricity generation, mobile services).48 For example, most of the energy generation capacity is concentrated among the Dominican Consortium of State Electricity Companies (CDEEE).49 The rest of installed capacity 46 The sector-type is based on the economic rationale for SOE presence based on intrinsic market characteristics and market failures (e.g., externalities, natural monopolies). See Dall' Olio, A., Goodwin, T., Martinez Licetti, M., Alonso Soria, A., Drozd, M., Orlowski, J., Patino Pena, F., Sanchez-Navarro, D. Are All State-Owned Enterprises Equal? A Taxonomy of Economic Activities to Assess SOE Presence in the Economy (English). Policy Research working paper WBG 10262. http://documents.worldbank.org/curated/en/099951412132222133/IDU0e7057cdf09fa8044eb09d8a058ea7dfe2831 47 Estimates based on the revenues as of 2018 in the DGII tax records. 48 The total number of operators are based on the Registry of Enterprises (Directorio Economico, 2019) and the market shares are estimated based on the DGII tax records. 49 It includes the only transmission company, the three distribution companies, the monopoly on hydropower and the only completely public thermal generation company (1.9 MW). https://blogs.iadb.org/caribbean-dev-trends/en/dominican-republics- energy-market/ 59 (thermal generation) is distributed among 12 private and mixed companies, where the state retains 49 percent participation. 58. In addition, when SOEs participate in a sector, regulation often limits state divestment. According to the PMR results, there are legal or constitutional constraints that limit the sale of the stakes held by the government in network sectors such as electricity generation and transmission,50, postal,51 water supply and treatment,52 or even in commercial activities such as manufacture of tobacco.53 59. In this context, granting preferential treatment to SOEs in markets where they compete with private operators might limit contestability and hinder efficient driven dynamics. Competitive neutrality is a principle according to which all enterprises, public or private, domestic, or foreign, should face the same set of rules, and where government’s contact, ownership, or involvement in the marketplace, in fact or in law, does not confer an undue competitive advantage on any actual or potential market participant (OECD, 2015).54 Although most SOEs are corporatized and remain subject to the private law and the competition law, they might still benefit from legal and de facto advantages. As captured by the PMR sub-indicators, the weak governance of SOEs may further preclude public operators from market forces. For instance, the Minister of Treasury (Hacienda) chairs the Board of the largest bank in the country and second largest in Central America, the public Bank -BanReservas, which is a SOE in the Dominican Republic with access to financing not available for private operators. In procurement, risks are associated to lack of transparency and ample use of direct negotiation (International Trade Administration, 2021). 60. In addition, more firms tend to exit the markets in sectors with SOE presence and in those with more restrictive regulations to competition in the Dominican Republic. Using the DGII administrative tax records between 2007-2017 55 and the WB BOS database, we can estimate and differentiate the exit rate by the presence of SOEs in markets. The presence of SOEs seems to be related to a larger number of firms exiting the markets. Estimates suggest a relatively higher share of firms exit the markets in Dominican Republic in sectors with SOE presence, which can 50 The sale of any equity stake in the hydraulic generation and transmission sector is prohibited. The Article 138 of Law 125-01 and article 1 of the Presidential Decree No. 628-07 establishes that Hidroelectrical Generation Company is exclusively for the state ownership. Presidential Decree 628-07 concretizes this by creating the SOEs ETED and EGEHID. 51 There is a legal monopoly established in this segment. It is important to note that to sale or grant concessions in this sector, laws No. 307 and No. 40 should be modified. 52 On the water supply and treatment level, the laws that create the different entities provide the public (governmental) nature of such institutions. Therefore, under such legal frame, it is not possible to sale the stakes held by government since these are governmental entities (PMR questionnaire, 2013-2017). 53 If the private participation was done by a capitalization process, articles 13 and 14 of Law No. 141-97 for the Reform of Public Companies restrict the participation of the private sector in controlling these companies. According to the paragraph I of the Article 13, the number of shares subscribed through these new capital contributions may not, in any case, exceed 50% of the total shares effectively paid by the capitalized companies. 54 OECD, Round table on Competitive Neutrality. Issues paper by the Secretariat, 2015, p.4. 55 It compiles the universe of formal firms obliged to declare to the tax authority. 60 be indicative of potential risks of displacing private competitors. Furthermore, a higher rate of firms exiting the markets are evidenced in sectors with more restrictive regulations such as price controls, barriers to entry, and provisions that protect incumbents. Those are explained in more detail in the next section. Figure 3.12. Exit rate of firms in markets with Figure 3.13. Firms exit at a higher rate also in SOE presence is relatively higher (and more sectors with restrictive regulation to competition disperse) Note: Exit rate corresponds to the percentage of incumbent Note: Exit rate corresponds to the percentage of incumbent firms that closed operations in a specific year. The exit rate in firms that closed operations in a specific year. The exit rate in the graph is the annual sectoral median observed between the graph is the annual sectoral median observed between 2007-2017. Source: WBG Markets, Competition and 2007-2017. The definition of markets with restrictive Technology estimates based on DGII administrative tax regulation is based on the sector assessment and PMR results. records and WB BOS database. Source: WBG Markets, Competition and Technology estimates based on DGII administrative tax records, PMR and sector- specific assessment. ii) Indirect channel: Government as a regulator 61. New firms in the Dominican Republic, particularly those in services, face complex and burdensome processes and require up to 97 days to set up a business. According to the Enterprise Survey, it takes on average 41 days to set up a business in the Dominican Republic, almost twice the number of days of some aspirational peers (e.g., Bulgaria) and 14 percent more than regional peers (e.g., Costa Rica), while still slightly below the LAC average. This statistic is even more burdensome for firms operating in the service industry, which can take up to 97 days. Smaller firms also cope with a higher number of days to start a business (on average 69 days), which may hinder the ability of potential entrants to threaten established and large incumbents in the market. Recently, the Ministry of Trade, Industry, and SMEs launched new guidelines and promoted the use of the new single window (Formalizate.gob.do) as a new strategy to facilitate entry and formalization to foster the entry of potential investors and creation of new businesses 61 (Ministry of Industry, Trade, and SMEs, 2020). The impact of these new guidelines is yet to be evaluated.56 Figure 3.14. Complex procedures and administrative burden imposed on firms explain over half of the barriers to entrepreneurship in the Dominican Republic 3.00 competition-friednly regulation) Score (higher values show less 2.50 2.00 1.50 1.00 0.50 0.00 LCR Costa Rica Dominican Uruguay Croatia Bulgaria OECD OECD Top 5 Republic Complexity of regulatory procedures Administrative burdens on start-ups Regulatory protection of incumbents Source: OECD and OECD-WBG Product Market Regulation Database 2013-2017. Note: The index scale ranges from 0 to 6, from most to least competition-friendly regulation. Figure 3.15. Average day for obtaining a Figure 3.16. Days for obtaining a license per sector license and firm’s size 200 176 120 97.5 180 100 160 80 69.2 140 55.3 120 60 41.6 42.5 93 100 40 80 22 20 20.1 15 60 41 20 36 36 40 21 0 20 Services - Retail Small (5-19) Medium (20-99) Large (100+) Manufacturing Services - Other Services - All Non-exporter Domestic 0 Bulgaria Croatia Costa Rica Serbia Uruguay DR Source: World Bank Enterprise Survey (2016) 62. In network industries, despite recent improvements, regulatory restrictions remain in place and key regulatory tools to promote competition seem to be missing. Potential competitors face incumbents, either public or private, that appear to benefit from regulatory protection resulting in an unlevel playing field. The PMR indicators highlight relatively high barriers in network sectors compared to structural peers such as Bulgaria, which are mostly 56New guidelines for the procedure to create a new company in the Dominican Republic were published by the Ministry of Industry, Trade and SMEs in 2020. See: https://www.micm.gob.do/images/pdf/publicaciones/libros/guias/2020/Guia_Creando_mi_empresa.pdf 62 driven by restrictions in electricity, postal services, gas, and road transportation services. For instance, in air transport, authorities can deny (ex-ante or at any later time) foreign airlines the right to operate on international routes based on the sufficiency of existing supply.57 In gas, lack of regulation for third-party access (TPA) limits the entry of new players. In telecom, local loop unbundling, which allows competitors in concentrated markets to use parts of the incumbents’ network infrastructure, is regulated since 2017. However, it has not been fully implemented, and apparently incumbents still have de facto power to restrict access (World Bank, 2021a). Figure 3.17. Barriers in network sectors in the Figure 3.18. Barriers in network sectors by Dominican Republic economic activity 6.0 5.12 5.0 4.39 6% 4.07 3.97 7% Electricity 4.0 30% Post 3.0 2.69 11% 2.25 Gas 2.0 Road 1.0 21% Airlines 0.0 26% Telecom Tunisia Costa DR Uruguay Bulgaria Croatia Rica Source: OECD and OECD-WBG Product Market Regulation Database 2013-2017. Note: The index scale ranges from 0 to 6, from most to least competition-friendly regulation. 63. Regulation in transportation services facilitates anti-competitive behavior despite being prohibited by the competition law. In road transport, the trade association of truckers, FENATRADO (Trade Union of Transport Operators of the Dominican Republic) coordinates the business decisions of their members providing road freight. Even though some of these practices are deemed anti-competitive as per the Competition Law (Art. 5), limited enforcement has favored the continuity of these behaviors. The current framework of FENATRADO restricts competition as it fixes prices negotiating the rates for its members (both individual transport operators and transport firms), segments specific zones for operation or transport firms (i.e., market division), and establishes quotas such as minimum cargo requirements to be transported by affiliated operators to FENATRADO regarding import and export cargo (Mizrahi Alvo & Guzman, 2017). This may be associated with the low performance of DR -78 out of 168 countries- in the World Bank Logistics Performance Index as of 2018.58 57 Law 491-06 of Civil Aviation at Article 246 points a) and b). 58 Logistics performance index measures the easiness of arranging competitively priced shipments. 63 64. In telecommunications, stringent licensing procedures are hindering entry into sectors that are essential to foster digitalization and productivity gains. The information and communication technologies sector (ICT) is critical to driving digitalization and productivity in an economy that is growing through the expansion of services, like the Dominican Republic. However, the ICT sector is highly concentrated with two large and vertically integrated operators that remain in a position to exert significant market power.59 The largest operator is Claro, which controls over 70 percent of the market share in fixed services and a quasi-monopoly position in a significant share of municipal areas.60 Restrictive regulation limits entry of firms that could challenge these operators. Current rules require new operators to pass through a burdensome process to provide telecommunication services, which might block, delay or disincentivize market entry. Applicants are required to be incorporated or registered as a legal person per the DR companies’ Law.61 In the case of public broadcasting services (radio and television, excluding cable television), it is also required that the person who has control of the service provider (the majority shareholder) is a Dominican national, preventing entry of foreign investors. 62 65. In other non-network industries, such as sugar manufacturing, the setup of regulatory quotas and export permits affects the functioning of markets that could have otherwise been competitive. Some manufacturing industries that are highly concentrated operate under restrictive regulations that protect incumbents. For instance, restrictions in the sugar sector in the form of annual production and export quotas as well as export permits hinder the ability of entrants to compete.63 INAZUCAR, the Dominican Sugar Institute, determines a yearly amount of sugar harvest and allocates it across four sugar mills through government decrees. For example, Decree 745-21 established sugar cane and sugar production quotas for the year 2021.64 Additionally, sugar producers require an export permit issued by the Dominican Sugar Institute, (Art. 18 of Law No. 618 of 1965). Furthermore, these regulations also pose imminent risks of conflict of interest as the government intervenes simultaneously as the sectoral regulator through the National Sugar Institute and as a market player through the State Sugar Counsel (CEA in Spanish) that controls one of the largest sugar mills. 66. Price controls in retail and electricity can deter competition or even favor anticompetitive practices such as cartels. According to the PMR, there are neither regulations on opening hours for retail services nor specific regulations for large outlets or promotions and 59 Compañía Dominicana de Teléfonos, S.A. “Claro”, and Altice Hispaniola, “Altice”. 60 As of 2020, Claro owned most of the country's fixed broadband lines and fiber optic networks and provides most of xDSL and FTTX internet services. INDOTEL. https://openknowledge.worldbank.org/handle/10986/35859?locale-attribute=es 61 See Article 22 of the Telecommunications Law (No. 153-98). 62 See Article 73 of the Telecommunications Law (No. 153-98). Article 22 of the same law requires that to obtain a concession and the respective licenses to provide Public Broadcasting Services, the concessionaire must be incorporated in the Dominican Republic. In 2021, Indotel began a process to modify Law 153-98, but the process does not seem have concluded. 63 According to the Art. 18, Law 618 of 1965, the National Sugar Institute provides certificates to the sugar producers to export. https://do.vlex.com/vid/decreto-n-353-19-840890417 64 Art.2 of the Decree allocates production quotas as follows: 57% for Central Romana Corporacion, 25% for Cristobal Colon, 13% for Barahona and 4% for the SOE Porvenir. Similarly, Art. 4 establishes export quotas granting 63% to Central Romana Corporacion, 27% to Cristobal Colon, and 10% to Barahona. 64 discounts. However, price controls in retail services are still present in the Dominican Republic.65 Price controls and tariff schemes are also seen in the electricity sector. They controls can facilitate coordination and agreements among competitors and deter the entry of competitors, since they may not properly reflect real production and distribution costs. For instance, since 2001 the government introduced a tariff mechanism, which is published periodically by the sectoral regulator (Superintendencia de Electricidad -SIE). This regulated tariff does not reflect the real costs of generation, causing operative and financial losses to the distributors, and deterring potential entrants (The Economist, 2015).66 As a result, the electricity sector is heavily subsidized since 2003 by the central government through the Bonoluz program (formerly the Programa de Reducción de Apagones), credit lines, and a tariff stabilization fund (a reserve mechanism to cover fuel-price variations). iii) Indirect channel: Government as a referee 67. The Dominican Republic developed its competition law in 2008, but it was only operationalized in 2017. The General Law for the Defence of Competition - Law No. 42-08 - was approved in 2008.67 However, it was not in force for almost a decade until 2017, when the Executive Director68 of the Competition Authority (National Commission for the defence of Competition –ProCompetencia) was appointed and the law finally started to be enforced. 6970 Besides the competition agency, sectoral regulators also have mandates to mitigate anti- competitive practices as well as promote competition and well-functioning markets. 68. The law contains all key elements to tackle potential anti-competitive practices and promote competition. It applies to national and foreign, private, or public economic agents engaging in economic activities in the domestic market. It also applies to agreements or conducts that occur outside the Dominican Republic territory when such agreements or conducts have 65 The Presidential Decree No. 3366 of 1969 gives the Ministry of Industry and Commerce the faculty of price control to oil and its derivates. This Ministry determines on a weekly basis the prices by resolution. 66 The SIE does not disclose the criteria used to calculate the applied tariff, which undermines transparency and increases uncertainty (The Economist, 2015) 67 The Competition Act was part of the Market Ordering Code, which included several laws; but this Code was dismembered because the laws were approved one by one. 68 ProCompetencia’s Authorities are the Executive Director and the Board of Directors –composed of five members—. The Executive Director is responsible for the investigative phase of cases, assessing the merits of complaints and deciding to pursue an action or dismiss a complaint. The Executive Director may also initiate an investigation ex officio. The Board of Directors decides the cases and has the power to impose sanctions for anticompetitive behavior in breach of Law No. 42-08. The decisions of the Board of Directors are subject to a motion for reconsideration and appeal before the Superior Administrative Tribunal. In turn, the Tribunal’s decisions may be challenged before the Supreme Court of Justice. 69 Decree No. 5-17 of the Executive Power, issued on January 6, 2017, appointed the Executive Director of the National Commission for the Defense of Competition – ProCompetencia- https://procompetencia.gob.do/sobre-nosotros/historia/ 70 Although three members of the Board of National Commission for Protection of Competition were sworn in 2011 and a fourth member was appointed in early 2015. https://unctad.org/system/files/official-document/diaepcb2016d2_en.pdf 65 restrictive effects on competition.71 The law proscribes abuses of dominant position, concerted practices, and unfair competition. 69. However, lack of merger control powers hinders the ability of the competition agency to limit the negative effects of market consolidation, a key concern given the role and regulatory protection that often benefit large incumbents. The competition law does not include the obligation to notify and review the competition impact of mergers and acquisitions to obtain approval from ProCompetencia. Merger control is a key preventive tool, especially in the context of Dominican Republic , where the size of the economy and the presence of highly concentrated markets in key sectors may exacerbate the risks of market consolidation. As of 2022, only concentration operations that take place in specific regulated markets require the prior authorization of the respective regulators. Such is the case of the telecommunications 72, banking73; securities74; insurance75; electricity76; pension funds77 and health risk administration78 sectors. 70. While several transactions of mergers and acquisitions have taken place in the Dominican Republic over the past decade, only a few occurred in regulated sectors that require ex-ante authorizations to prevent anticompetitive impact. Based on information from the 71 For example, anticompetitive conducts that could affect prices of tradable goods or services that are provided in DR. This is the extraterritorial scope of the Competition Law. 72 Law No. 153-98 and the Regulation for Free and Fair Competition of the Telecommunications Market require that any transfer, assignment, lease or grant of the right to use any title or lien granted on concessions or licences may be performed only after obtaining the authorisation of the Dominican Institute of Telecommunications (Indotel). Likewise, the sale or assignment of shares resulting in the seller’s or transferor’s loss of control requires the authorisation of Indotel. Moreover, mergers and market concentrations in the telecommunications sector are expressly subject to the prior approval of Indotel, which can challenge the operation and/or request remedies. 73 The DR Monetary and Financial Law requires the prior authorization of the Monetary Board for mergers, share transfers of 30% or more of the paid-in capital, absorption, and substantial asset and liabilities transfers of any financial intermediation entity (Articles 9 and 35 of the Law). 74 The DR Companies Act (Law No. 479-08) states that public limited companies must submit M&A agreements to the Securities Superintendence, which will accept or reject the projected concentration within 15 days (Articles 386) 75 The DR Insurance and Bonds Law (Law 146-02, Articles, 174 to 184) allow insurance and reinsurance companies to merge with the prior authorisation of the Dominican Insurance Superintendence. The Superintendence may also recommend that an insurance company merges if its financial statements or the Superintendence’s examinations show that the company cannot guarantee the fulfilment of its obligations. 76 Paragraph II of Article 12 of the Rules for the Application of the Electricity General Law -Decree No.555-02, states that the Electricity Superintendence, before authorising the transfer of generation concessions, mergers or sales of shares involving generation companies, must determine whether the petitioners own generation centres with a total capacity representing a significant percentage of the maximum demand of the national interconnected electric system, and thus , in accordance with the criteria of the National Commission of Energy, whether the operation could constitute a threat to competition in the electric wholesale market. Article 82 of the Law establishes a similar provision for the transfer of concessions in other segments of the electricity market. 77 Article 93 of the Law of Social Security and Article 50 of Decree No. 969-02 that establishes the Pension Rules, require that mergers be approved by the Pension Superintendence. The Superintendence may require amendments to, or reject, the projected operation. 78 Article 153 of the Social Security Act mandates that health risk administrators and the National Health Insurance Scheme must obtain authorisation from the Health and Labour Risks Superintendence before completing a concentration operation. 66 Thomson Reuter, 54 mergers and acquisitions for a total of USD 1.7 billion occurred between 2010-2020 in the Dominican Republic. However, only 16 percent of those occurred in sectors where they require the prior authorization from the respective regulator. The remaining transactions occurred in non-regulated markets where consolidation strategies could lead to reduced competition. The latter include breweries, rice milling, accommodation services, travel agencies, manufacture of paper and basic metals, where neither the competition agency nor regulators have the mandate and power to assess and prevent potential implications for the markets. 71. Moreover, enforcement remains relatively weak as faculties to investigate prescribe within a year, and many investigations for anticompetitive practices have been opened, but only two instances of anticompetitive behaviors have been sanctioned. Since 2017, ProCompetencia has investigated potential anti-competitive practices in a diverse range of sectors including bottled water, beer beverages, pension fund administrators (AFP), oil, liquified petroleum gas, rebars, chicken eggs, family health insurance (ARS), wheat flour, rice, bread, used vehicle imports, milk, freight transport, and the drug market. 79 However, 15 out of the 18 investigations initiated by ProCompetencia have been dismissed (as of March 2022) and only three decisions have been published.80 Two of these decisions refer to conducts for abuse of dominance and one cartel, respectively. As compared to this enforcement record, the World Bank’s anti-cartel enforcement database (ACED) finds that, between 2017-2022, competition agencies in the LAC region sanctioned 14 cases of cartels, with an average of 2 cases per year. Limited anticartel enforcement is particularly problematic as cartels slow productivity growth, damage public trust in market economies and limit the role of the private sector as an engine of growth (World Bank, 2021). The third decision only pertains to false information provided by a firm (see Box 4.1.). This large dismissal of investigations could indicate challenges to building solid cases despite the ability of ProCompetencia to access direct evidence. 81 This could also be related to the limitation of maximum duration of the investigations (12 months) that may hinder the ability of ProCompetencia to build and close the cases in such limited time. Finally, the 79 Resolutions of the Executive Director can be found here: https://procompetencia.gob.do/resoluciones/direccion-ejecutiva/ and Resolutions of the Board of Directors can be found here: https://procompetencia.gob.do/resoluciones/consejo-directivo/ 80 i) Cervecería Nacional Dominicana (RESOLUCIÓN NÚM. 018-2018), imposition by the supplier of prices and other conditions of sale to its resellers, without commercial justification [article 6, letter b) of Law 42-08]- https://procompetencia.gob.do/wp- content/uploads/2020/02/cd018-2018.pdf; ii) Coastal Petroleum Dominicana, SA (RESOLUCIÓN NÚM. DE-26-201) admitted complaint filed by Refinería Dominicana de Petróleo PDV, SA (REFIDOMSA) against Coastal Petroleum Dominicana, SA, for abuse of dominant position in the market of import and commercialization of Liquefied Petroleum Gas (LPG) in the DR [article 6 of Law 42-08]- https://procompetencia.gob.do/wp-content/uploads/2020/02/de-026-19.pdf; iii) Profarma International, Sued & Fargesa, Gassó Gassó and Mercantil Farmacéutica (RESOLUCIÓN NÚM. 010-2021), for agreeing on prices, discounts, charges, and other conditions of sale and the exchange of information [article 5, letter a) of Law 42-08]- https://procompetencia.gob.do/wp- content/uploads/2021/06/resolucion-num-010-2021-version-publica.pdf 81 ProCompetencia can conduct dawn raids (see Art. 31 of the law) and grant leniency to the first cartel member to come forward with information. See the Regulation for the establishment of a Regime for Reduction of Sanctions for Collaboration during Investigations of cartel behaviors available at https://procompetencia.gob.do/wp-content/uploads/2021/06/cd-008-2021-que- aprueba-rrs-por-colaboracioin.pdf 67 competition authority faces limitations in terms of prosecution capacity as its investigative faculties prescribe within one year counted from the cease of the anti-competitive conduct. 82 Box 4.1. Decisions of ProCompetencia related to anticompetitive practices (2017-2022) Cervecería Nacional Dominicana (RESOLUCIÓN NÚM. 018-2018). ProCompetencia found the company in breach of the Competition Act for abusing its dominant position by imposing unfair trading conditions to its buyers [article 6, letter b) of law 42-08]. The Company was fined 3000 minimum wages (at the time, about USD 915,887). 83 Profarma International, Sued & Fargesa, Gassó Gassó and Mercantil Farmacéutica (RESOLUCIÓN NÚM. 010- 2021), ProCompetencia found the companies in breach of the Competition Act for collusive behaviour [article 5, letter a) of law 42-08]. The Companies were fined 315, 281, 38 and 167 minimum wages, respectively (at the time, the total fines amounted to about USD 245,531). The reconsideration request submitted by MERCANTIL FARMACÉUTICA, S.A. was dismissed.84 Sodetransp (RESOLUCIÓN NÚM. 001-2022), was fined for providing false information to ProCompetencia [article 61, letter d) of law 42-08] with an amount of 200 minimum wages (at the time, about USD 66,721). 85 Source: ProCompetencia 72. In addition, the level of sanctions remains too low to effectively deter anticompetitive practices, despite empirical evidence showing that sanctions are associated with increased productivity in the targeted industries. Using evidence from Mexico, Reed et al. (2022) show that the effect of antitrust sanctions on sales, wages, and productivity is positive within the industries under investigation. From a policy design perspective, and according to international best practices, antitrust fines should pursue a twofold policy objective: (i) to impose penalties on infringing undertakings; and (ii) to ensure that the threat of penalties will deter both the infringing undertakings and other undertakings that may be considering anti-competitive activities from engaging in them.86 The Competition Authority in DR requires the presence of the Prosecutor’s Office (Fiscalia Nacional), the police, and judges to conduct down-raids, which limits the ability to conduct down-raids in a timely and effective manner to collect evidence in situ. As a result, the competition agency has conducted only one down-raid. Additionally, the current Competition Law does not provide the legal basis for the implementation of the leniency program to destabilize cartels. The gravity of the sanction should always reflect the gains generated by the infringement to offset them with the amount of the fine. Furthermore, the sanctions in DR are not always reflecting the gains generated by the infringement. To this end, most jurisdictions set fines based on a maximum allowable amount based on the turnover of the ultimate corporate 82 Article 56 of the Competition Law 42-08. 83 The decision is available at: https://procompetencia.gob.do/wp-content/uploads/2020/02/cd018-2018.pdf 84 The decision is available at: https://procompetencia.gob.do/wp-content/uploads/2021/06/resolucion-num-010-2021-version- publica.pdf 85 The decision is available at: https://procompetencia.gob.do/wp-content/uploads/2022/01/cd-001-2022-firmada-y-sellada.pdf 86 For the UK, see the 2012 “OFT's guidance as to the appropriate amount of a penalty” at pp. 2-3. Available at http://www.oft.gov.uk/shared_oft/business_leaflets/ca98_guidelines/oft423.pdf. 68 parent. As compared to this, applicable fines of Law 42-08 of DR are based on a cap amount ranging from 20 to 3,000 minimum salaries depending on the type of infringement.87 73. Although ProCompetencia has a legal mandate to develop sectoral competition regulation with sectoral agencies, implementation has been slow . In recent years ProCompetencia has allocated significant resources to advocacy activities. The agency has published several guidelines and regulations and developed training programs for both its own officials as well as for employees of other government agencies88 as part of its advocacy strategies. Some guidelines include the "Guide on Competition for the Public Sector: the Role of the Dominican State in the Protection of Free Competition” (2021); the "Basic Guide on Competition for Business and Trade Associations", and "Guide for the Prevention and Detection of Collusion in Public Procurement".89 Since 2016, ProCompetencia has also conducted five market reports by the Market Observatory, including policy recommendations to foster pro- competition markets. However, these recommendations are not binding in most jurisdictions (Art. 14 of the Competition Law).90 Instead, ProCompetencia has a unique mandate that remains unexploited: on the one hand, Law 42-08 states that sector-specific regulations enforced by sectoral regulators supersede the competition law;91 on the other hand, the law mandates that, within two years from its entry into force, ProCompetencia must convene with the regulators to jointly review, propose, and dictate the competition regulation that will govern the operation of the respective regulated markets.92 Yet, ProCompetencia only held the first meeting of the Council for Regulatory Improvement in 2022 and has kept interinstitutional cooperation at the bilateral level. This issue seems especially relevant given the overlapping competition mandates between ProCompetencia and sectoral regulators. Therefore, strong coordination and clear separation of the powers between sectoral regulators and the competition agency is key to promoting an effective application and enforcement of competition law across different sectors. 74. In addition, the ability of ProCompetencia to identify distortive state aid or state support measures remains underutilized. Although controlling state support measures to SOEs is critical to minimize competition distortion and is under the mandate of ProCompetencia, it has not been systematically implemented. Beneficiaries that receive state support, either public or private, enjoy a comparative advantage over their competitors that is not necessarily associated with their efficiency. Control of state support measures is a necessary safeguard for effective competition, free trade, and efficient management of fiscal resources. The competition law (Art. 15) establishes an obligation for the state to not offer any support measure to SOEs or companies providing services delegated by the state that may create entry barriers or unlevel the playing field. The same article empowers ProCompetencia to analyze state aid and proposes measures 87 Article 61 – Law 42-08 88 For example, ProCompetencia officials, the Office of the Attorney General of the Republic (PGR), the Directorate of Public Procurement and Contracting and the Department of Investigations of Crimes and High Technology Crimes have participated in cartel investigation techniques trainings facilitated by experts from the United States (DOJ). The Inabie (University Student organization), and National School of Judiciary, as well as other government entities have received training on competition issues since 2017. 89 The Resolution enacting these Guides can be found here: https://procompetencia.gob.do/wp-content/uploads/2020/08/cd-007-2020-firmada- y-sellada.pdf 90 The market reports published by the Market Observatory can be found here: https://procompetencia.gob.do/observatorio-de-condiciones- de-mercado/#15-89-wpfd-ano-2016-1604515989 91 Article Nº20, Paragraph III– Law 42-08- 92 Article Nº69 – Law 42-08- Transitional Provisions. 69 to limit anticompetitive effects. This is very critical as a tool to minimize the risks of companies recipients of government support in the form of state aids (or subsidies) to gain an undue advantage over their competitors.93 In 2022, ProCompetencia presented the report “ The Impact of State Aid from a Competition Perspective: The Case of Law 28 -01 on Border Development”, which appears to be the first analysis on state aid undertaken by the agency. 94 In this sense, it would be key for ProCompetencia to develop the mechanisms to further articulate and develop its mandate on state aid control. e. Conclusions 75. Markets in the Dominican Republic are constrained by large state presence and distortive market regulations that prevent entry and foster market concentration. Markets in the Dominican Republic are perceived as less open to competition and operate under highly concentrated structures even in sectors where competition is viable. Product Market Regulation (PMR) indicators suggest two main sources of public sector-related barriers to competition in the Dominican Republic: direct involvement of the government in the form of SOEs and indirectly as the regulator. SOEs are still present in sectors where the economic rationale of state ownership is less clear, and firms seem to exit more frequently in those sectors. Complex regulation and sectoral provisions protect incumbents, as production and export quotas as well as price controls seem to deter the incentives for firms to enter and grow. 76. SOEs operate in markets that are viable for competition and private participation, thus reducing the incentives for productive firms entering the market or accelerating the exit of productive firms. SOE revenues in the Dominican Republic are relatively lower than the regional LAC average and below structural peers (e.g., Croatia and Uruguay) and structural peers (e.g., Costa Rica). However, at least 40 percent of these firms still operate in competitive sectors that could be viable for the private sector and where the risk of crowding out private competitors is high. In addition, more firms tend to exit the markets in sectors with SOE presence and in those with more restrictive regulations to competition in the Dominican Republic. 77. The delayed implementation of the competition Law in 2017, despite being enacted in 2008, limited the active regulatory role of the government in fomenting competitive markets. Since 2017, ProCompetencia has investigated potential anti-competitive practices in a diverse 93 This artificial advantage distorts the normal competitive process by altering the level playing field between competitors. Secondly, the granting of a subsidy to a particular company may deter other companies from investing in the market in which the beneficiary undertaking operates, due to the level of investments already made by the beneficiary undertaking with the public support (crowding out effect). Thirdly, companies benefiting from subsidies may use the public resources to engage in anticompetitive conducts. These conducts may take the form of predatory pricing or other conducts aimed at excluding present competitors or at preventing the entry of potential competitors to the markets in which they operate. In addition, subsidies entail a significant opportunity cost for public authorities and may generate negative incentives on domestic producers such as rent-seeking dynamics. See European Commission, Communication from the Commission to the European Parliament, the Council, and the European Economic and Social Committee and the Committee of the Regions, EU State Aid Modernization (SAM), of 8 May 2012, COM(2012) 209 final, 3 (paras 6 and 12). See also Lorenzo Coppi, ‘The role of economics in State aid analysis and the balancing test’, in Erika Szyszczak (ed), European Handbook on State Aid Law (Cheltenham: Elgar, 2011) 64– 89, at 76. 94 The report is available here https://procompetencia.gob.do/wpfd_file/estudio-sobre-el-impacto-de-las-ayudas-estatales/ 70 range of sectors, however, 15 out of the 18 investigations initiated by ProCompetencia have been dismissed (as of March 2022) and only three decisions have been published. In addition, the Competition Law does not include the obligation by relevant companies to notify and review the competition impact of mergers and acquisitions to obtain approval from ProCompetencia. As a result, 54 mergers and acquisitions for a total of USD 1.7 billion occurred between 2010-2020 in the Dominican Republic, with most of them not reviewed by ProCompetencia. f. Appendix Appendix 3.1. Product market regulation (PMR) indicators The Organization for Economic Co-operation and Development (OECD)—World Bank Group (WBG) Product Market Regulation (PMR) indicators for 2013-2017 are a comprehensive and internationally comparable set of indicators that measure the extent to which de jure policies promote or limit competition in areas of product markets where competition is viable. The indicators cover regulatory barriers to firm entry and competition at the economywide level and in key enabling sectors. The economywide indicators cover barriers to competition in three broad areas: state control (public ownership and involvement in business operation); barriers to entrepreneurship (complexity of regulatory procedures, administrative burdens on startups, and regulatory protection of incumbents); and barriers to trade and investment (explicit barriers to trade and investment and other barriers to trade and investment). The PMR indicators are based on over 800 similar qualitative information on regulatory structures and policies collected through the OECD and WBG through questionnaires, answers of which reflect the situation as of 1 January of the year of each country’s data (2014 for Dominican Republic). The questionnaire answers are first coded into quantitative scores (range from 0 to 6, with 6 indicating that regulations are least friendly to competition). The economywide indicators are grouped by regulatory domains. First, the coded scores are grouped into 18 low-level components of regulations using weights determined by the OECD methodology. The 18 components are then aggregated into 7 mid-level categories using equal weights. Then, the 7 mid-level categories are aggregated into 3 high-level components —state control, barriers to entrepreneurship, and barriers to trade and investment—, with the aggregations applying equal weights. Finally, these three broad categories are averaged to get the economywide PMR score. The sectoral indicators, which summarize the regulatory information by sector, cover barriers to competition in network sector—energy (electricity and gas), transport (rail, air, and road), and communications (post and telecom)—, professional services (accounting, legal, architect, and civil engineer), and retail distribution. The indicators are useful for monitoring the regulatory achievements of countries and for evaluating the progress of policies to limit regulatory gaps. Although the PMR indicators touch on a wide range of topics and sectors, barriers to competition are their focus and are not intended to comprehensively cover issues related to private sector development. 71 Economywide and Sectoral PMR Indicators A. Economywide indicators B. Sectoral indicators Source: OECD and OECD-WBG PMR Indicator Database 2013-2017. 72 Chapter 4. Reducing expenditure inefficiencies to open up fiscal space for productivity-enhancing investments a. Introduction 78. Despite a prudent fiscal policy, the financing of productivity-enhancing public investments faces significant constraints on the revenue side. The central government balance has fluctuated within a relatively narrow range since 2000—except for higher deficits recorded in 2003 (due to the Baninter banking crisis) and 2012. However, government revenues have fluctuated around 14 percent of GDP on average, significantly below several LAC and structural peers. The government’s tax revenues have only increased from 12.2 percent of GDP in 2000 to 13.3 percent in 2019. Figure 4.1. Central Government Fiscal Figure 4.2. General government tax Performance revenues (As a percentage of GDP, 2000-2019) (As a percentage of GDP, 2000-2019) 3.0 10.5 2019 Guatemala 2.0 2000 1.0 Dominican Republic 13.3 - 13.4 percentage -1.0 Costa Rica -2.0 Structural 16.9 -3.0 -4.0 17.6 Nicaragua -5.0 -6.0 El Salvador 18.1 -7.0 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Uruguay 18.2 Primary balance (RHS) Overall Balance (RHS) Aspirational 22.6 Structural Primary Balance (RHS) - 5.0 10.0 15.0 20.0 25.0 Source: A background note prepared for the report “The Dominican Republic: An Overview of Macroeconomic Performance” 79. Taxes that directly affect private consumption make up the vast majority of overall tax revenues, while exemptions are above the regional average. In 2019, taxes on goods, services, income, and profits accounted for nearly 92 percent of all tax revenues, versus 76 percent in the LAC and 66 percent in OECD countries, on average (notably, social contributions are not considered tax revenues in the Dominican Republic). In 2019, taxes collected from income and profits were equivalent to 4.3 percent of GDP (versus 6 percent in the LAC and 11.3 percent in the OECD), while those from goods and services represented 8 percent of GDP (versus 11.2 percent in the LAC and 10.8 percent in the OECD). Importantly, the tax collection effort is relatively low in the Dominican Republic, and tax exemptions are among the highest in the region—for example, the value of those pertaining to goods consumption reaches 3 percent of GDP, versus 2 percent in the LAC. 73 Figure 4.3. Tax revenue, composition Figure 4.4. Tax exemptions, composition (As a percentage of GDP, 2019) (As a percentage of GDP, 2019) General taxes on consumption Individual income taxes Good & Services Property Corporate income taxes Specific excises Social security Corporate income Foreign trade taxes Others 8 Personal Income and Profits International Trade 6 10.8 4 11.2 2 8.0 0 Dom. Rep. Peru Uruguay El Salvador Costa Rica LAC Colombia Nicaragua Jamaica Panama Guatemala Mexico Chile Argentina Bolivia Brazil Paraguay Ecuador 0 Dom. Rep. LAC OECD Source: A background note prepared for the report “The Dominican Republic: An Overview of Macroeconomic Performance” 80. Rising interest payments, rigid current spending and a relatively low tax-to-GDP ratio constrain the country’s fiscal space for public investments. Debt service absorbs nearly one-fifth of tax revenues, while the sum of wages and salaries, current transfers, and debt service equals more than 60 percent of public expenditures. Interest payments increased by 14.4 percentage points between 2000 and 2019, reaching a level equivalent to nearly 20 percent of total revenues, and squeezing budget allocations toward productive investment. Notably, public investment shrank to 2.3 percent of GDP in 2019, a level 1 percentage point lower than in 2000. Figure 4.5. Interest Payments Figure 4.6. Expenditure Composition (As a percentage of tax revenues, 2000-2019) (As a percentage of total, 2019) 25.0% Gross Investme 20.0% nt, 9% Wages Other , and 12% Salaries, 15.0% 26% Social Benefits, 10.0% 8% Donation… 5.0% Interest , 16% Goods and 0.0% Subsidies Services, 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 , 4% 11% Source: A background note prepared for the report “The Dominican Republic: An Overview of Macroeconomic Performance” 81. While public debt has been rising as a share of GDP, its trajectory is assessed to be sustainable, thanks to fast-paced GDP growth and relatively low primary deficits. The debt of the consolidated public sector, including state-owned enterprises and the central bank, grew from 18.2 percent to 50.5 percent of GDP between 2000 and 2019. About 64 percent of the 74 consolidated public debt is accrued by the central government, whose borrowings also cover the energy sector’s deficit and most of its investments (i.e., the energy sector is both a direct and a contingent fiscal liability). The central bank has run sizeable quasi-fiscal deficits since the Baninter crisis in 2003, as the latter was not recapitalized after the crisis, and could not recoup all the resources drained by it through seigniorage.95 From 2000 to 2004, public debt increased by 28.7 percentage points, with almost 50 percent of this increase stemming from the impact of the Baninter crisis. 82. Substantial fiscal savings – opening fiscal space for productivity-enhancing public investments – can result from revenue reforms and from improved expenditure efficiency. On the revenue side, removing tax exemptions and widening the tax base remain a top priority in the country with strong political support. These reforms are even more important in the context of inflationary pressures during a pre-electoral year. Reforms on the expenditure efficiency side can also help address productivity bottlenecks. For example, improving the efficiency of public spending in education would help improve the quality of human capital and reduce skills mismatches through reallocating more resources to initiatives such as modernizing the curricula at high school level, better aligning college education with business needs (especially in the IT sector), and implementing early warning systems to prevent students from dropping out. Figure 4.7. Consolidated Public Debt (As a percentage of GDP, 2000-2019) Central Bank Other Non-Financial Public Sector Central Government 60.0% 50.5% 50.0% 46.6% 41.2% 40.0% 33.8% 29.8% 30.0% 20.0% 10.0% 0.0% 2000-2005 2005-2010 2010-2015 2015-2019 2019 Source: A background note prepared for the report “The Dominican Republic: An Overview of Macroeconomic Performance” b. Improving procurement efficiency 80. Improving the efficiency of the procurement process can lead to significant fiscal savings and greater competition in product markets. The Dominican Republic is currently working on 95 In the past, the IMF provided technical assistance and support for the 2007 Recapitalization Law (the latter being a condition of the 2005-2008 SBA). 75 improving the Public Procurement Law to obtain gains from a more efficient public procurement system. The government’s proposal incorporates new methods that are aligned with good procurement practices around the world. These are Total Cost of Ownership (TCO), Life Cycle Costing (LLC), Value for Money (VfM), Sustainability, Framework Agreements, and requirement for awards to Micro, Small, and Medium enterprises (MSMEs). Additionally, in line with transparency best practices, the government launched “MapaInversiones” to consolidate evidence on procurement processes with options for suggestions and complaints. Currently, the platform includes only purchases related to the COVID-19 pandemic (2020) and Fiona Hurricane (2022). As of February 28, 2022, the Executive Branch submitted to the Congress a proposal to amend Law 340-06 and its amendments, which seeks to include the aforementioned concepts. 81. A detailed analysis of the procurement processes during 2018-19 period reveals the importance of less competitive processes – such as minor purchases, purchases under the relevant threshold, and exception processes in the budget.96 Nearly 80 percent of the procurement processes in 2018-2019 were used less competitive methods, while on average 13 percent of processes during this period used the exception method, which is intended for cases of emergency. Figure 4.8. Number of Procurement Figure 4.9. Number of Procurement Processes Processes (As a percentage of total, 2018) (As a percentage of total, 2019) Price Comparison Exception Price 5% processes Comparison 6% 6% National Exception Public National processes Tender Minor Public Minor purchases 19% 1% Tender purchases 26% 22% Purchases 1% Purchases under the under the threshold threshold 60% 53% Figure 4.10. Procurement Processes by Figure 4.11. Procurement Processes by Contracted Amount Contracted Amount (As a percentage of total, 2018) (As a percentage of total, 2019) 96The data comes from the Open Data platform from the System of Purchases and Public Contracting of Goods, Services, Works and Concessions: https://datosabiertos.dgcp.gob.do/opendata 76 Others Others 4% 1% Price Comparison 13% Price Comparison 16% Minor purchases Minor purchases 5% 6% Purchases Purchases under the Exception under the Exception threshold processes threshold 2% processes 1% 45% 57% National Public National Public Tender 0% Tender 2% Source: Background note prepared for the report “Procurement Process Characteristics of the Dominican Republic” 82. Procurement processes are primarily intended for the purchase of goods and services. Purchases below the threshold and minor purchases methods are mainly intended to purchase goods. Exception processes, on the contrary, are mainly used to purchase services. By comparing the procurement process between national entities versus subnational entities, procurement data shows that national entities allocate more than 20 percent of their contracted amount to exception processes. Figure 4.12. Commonly used procurement Figure 4.13. Contracted amount by processes by object procurement processes by Entity and Type of (As percentage of total, 2018-19) process97 (As percentage of total, 2018-19) Services Goods Competitive Exception Process Less-competitive 15% 33% 15% 6% 16% 8% 57% 55% 23% 1% 65% 65% 15% 28% 84% 67% 83% 70% 71% 64% 43% 45% 35% 35% Sub- National Sub- National Purchases Purchases purchases purchases Exception Exception under the under the threshold threshold Process Process Minor Minor national national 2018 2019 2018 Source: Background 2019 note prepared for the report “Procurement Process Characteristics of the Dominican Republic” 83. Five ministries contracted more than 50 percent of exception process (see table 4.1.). Less competitive processes were the most commonly used by the Ministry of Energy and Mines 97Competitive processes refer to tenders, raffles and price comparison, while less competitive process are minor purchases and purchases below the threshold. 77 and the Ministry of Foreign Affairs. More than half of contracted amounts by the remaining ministries were allocated to competitive processes, mostly national public tenders. Table 4.1. 2019 Procurement processes by Ministry: amount contracted by type of process (As percentage of total) Exception Total - Less Ministry Competitive Process Competitive Administrative Ministry of the Presidency 18% 76% 6% Ministry of Public Administration 63% 21% 16% Ministry of Agriculture 82% 12% 6% Ministry of Culture 53% 31% 16% Ministry of Defense 63% 25% 12% Ministry of Sports, Physical Education and Recreation 46% 13% 41% Ministry of Economy, Planning and Development 53% 7% 40% Ministry of Education 31% 65% 5% Ministry of Higher Education, Science and Technology 27% 45% 28% Ministry of Energy and Mines 24% 3% 73% Ministry of Finance 25% 65% 11% Ministry of Industry, Trade and MSMEs 42% 30% 28% Ministry of Interior and Police 79% 5% 16% Ministry of Youth 86% 0% 14% Ministry of Women 55% 1% 44% Ministry of the Presidency 59% 13% 27% Ministry of Environment 55% 1% 44% Ministry of Public Works and Communications 50% 50% 0% Ministry of Foreign Affairs 34% 14% 52% Ministry of Public Health and Social Assistance 17% 77% 6% Ministry of Labor 76% 3% 21% Ministry of Tourism 64% 10% 27% Source: Background note prepared for the report “Procurement Process Characteristics of the Dominican Republic” 84. Different procurement processes lead to different prices in most scenarios. Evidence from the Dominican Republic's public purchases shows that less competitive processes tend to have higher prices on average. In the case of rice bags of 125 pounds, on average, acquiring those products through a more competitive process could have saved the country RD$1,000 or approximately US$20 per item. Similarly, of the use of competitive processes for milk (unit) could have saved on average RD$108 or approximately US$2 per item. Finally, the price for gasoline purchased by the government is lower under competitive prices. Lower prices from competitive methods are also found in government purchases of hardware appliances, office appliances, and catering services. 78 Figure 4.14. Rice bags 125 pounds - Prices by Figure 4.15. Milk Unit Whole/Evaporated Type of Processes Prices by Type of Processes (Average, 2018-19) (Average, 2018-19) Price comparison Price comparison Minor purchases Minor purchases Purchases below the threshold Purchases below the threshold National public tender National Public Tender Exceptional Process Exceptional Process 3,463 3,408 2,800 599 2,500 2,378 2,240 2,260 278 260240 108 151 30 47 48 2018 2019 2018 2019 Figure 4.16. Regular Gasoline Prices by Type of Figure 4.17. Hardware Appliances: Nails Processes Prices by Type of Processes (Average, 2018-19) (Average, 2018-19) Price comparison Price comparison Minor purchases Minor purchases Purchases below the threshold National public tender Purchases below the threshold Exceptional Process 106.9 462 95.7 423 74.0 67.8 327 231 217235 45.9 47.7 201218 2018 2019 2018 2019 Figure 4.18. Office Appliances: All Bond Paper Figure 4.19. Catering Services for Lunch Prices by Type of Processes Prices by Type of Processes (Average, 2019) (Average, 2018) 79 Price comparison Price comparison Minor purchases Minor purchases Purchases below the threshold Purchases below the threshold Exceptional Process National public tender 692 401 341 561 300 500 250 406 2019 2018 Source: Background note prepared for the report “Procurement Process Characteristics of the Dominican Republic” 85. In addition to helping increase competition in the economy more broadly, using more competitive procurement processes could generate fiscal savings of around RD$4.6-5.3 billion. The potential efficiency gains of utilizing more competitive procurement processes in government purchases between 2018 and 2019 could have generated savings equivalent to building 79 new primary schools, hiring over 7,000 elementary or high school teachers. expanding the national fiber optic network in line with the Digital Republic Plan or purchasing more than 100,000 laptops for a training program for recent graduate students. Table 4.2. Procurement Processes Potential Savings of a Competitiveness Shift in Purchases Potential savings Aggressive Savings Conservative Scenario 2/ Savings Scenario 3/ In RD$ 5,356,106,059 4,699,264,112 As % of 2019 Health budget 7.1% 6.2% As % of 2019 Education budget 2.9% 2.5% Number of schools built 4/ 79 69 New teachers hired 5/ Elementary 7,740 6,791 High School 7,051 6,187 Digital Republic Plan (2018): Pillars & KPIs Pillar I: Education & Technology Number of laptops 6/ 112,120 98,370 Number of PCs 6/ 90,982 79,824 Pillar II: Broadband for everyone Expansion of National 165% 145% Fiber Optic Network 7/ Source: Background note prepared for the report “Procurement Process Characteristics of the Dominican Republic” Notes: 1/ Potential savings calculated as the difference between the actual amount spent on those articles and the amount that could have been spent if the procurement process with the lowest price (competitive process) would have been chosen and then extrapolated to the total amount of the database 2/ If the procurement process with the lowest price would have been chosen. 80 3/ If the procurement process with the second lowest price would have been chosen. 4/ Estimated cost of a basic school built in Santo Domingo (RD$68 million) based on data from the “Unidad de Fiscalización del Programa Nacional de Edificaciones” (2022). 5/ Number of primary and secondary education teachers that could be hired per year based on minimum monthly salary of RD$57,668 (primary) and RD$63,298 (secondary) in 2020. 6/ Calculations based on the average price paid for a laptop and personal computer (PC) through different procurement processes in 2018 7/ Calculations based on an upcoming investment of $54 mill by ETED to expand the national fiber optic network by 400 KM aimed at connecting the whole country. g. Conclusions 83. Limited tax collection and increasing interest payments constrain the country’s fiscal space for carrying out substantial productive investment. The tax collection effort is relatively low in the Dominican Republic, with tax exceptions are among the highest in the region, and debt services absorbing nearly one-fifth of tax revenues. Therefore, public investment shrank to 2.3 percent of GDP in 2019, a level 1 percentage point lower than in 2000 84. Substantial fiscal savings – opening fiscal space for productivity-enhancing public investments – can result from revenue reforms and from improved expenditure efficiency. On the revenue side, removing tax exemptions and widening the tax base remain a top priority in the country with strong political support. Reforms on the expenditure efficiency side can also help address productivity bottlenecks. For example, improving the efficiency of public spending in education would help improve the quality of human capital and reduce skills mismatches through reallocating more resources to initiatives such as modernizing the curricula at high school level, better aligning college education with business needs (especially in the IT sector), and implementing early warning systems to prevent students from dropping out. 85. Improving the efficiency of the procurement process can lead to fiscal savings of up to RD$4.6-5.3 billion (equivalent to 0.1 percent of GDP) and greater competition in product markets. The potential efficiency gains of utilizing more competitive procurement processes in government purchases between 2018 and 2019 could have generated savings equivalent to building 79 new primary schools, hiring over 7,000 elementary or high school teachers, expanding the national fiber optic network in line with the Digital Republic Plan or purchasing more than 100,000 laptops for a training program for recent graduate students. 81 Conclusion This Country Economic Memorandum examined the Dominican Republic’s recent economic growth experience from macroeconomic, sectoral and microeconomic perspectives . The report looked into the country’s economic growth performance at both the aggregate and firm levels. It also assessed the Dominican Republic’s market competition environment as well as its prospects to open up fiscal space in order to carry out much needed public investments. While the Dominican economy grew much faster than the vast majority of LAC countries over the past two decades, its current growth model might be reaching its limits . A key finding of this report is that the country’s economic growth has been based on factor accumulation rather than productivity gains. Furthermore, its growth model has been uneven and associated with significant regional disparities as well as wage stagnation, which does not favor inclusiveness. Most of the obstacles to faster productivity growth at the microeconomic and aggregate levels are associated with domestic policy distortions. They comprise an inadequate human capital performance, skills mismatches between labor supply and demand, an insufficiently competitive business environment, limited innovation capacity, and a constrained fiscal space. Sustaining high growth in the long run will require a new round of economic reforms . In addition to maintaining a sound macroeconomic framework, moving forward with critical reforms would be essential to keeping strong growth in the long run. They include: (i) unleashing its human capital potential; (ii) fomenting competitive markets; (iii) revamping the innovation strategy; (iv) reducing public expenditure inefficiencies; and (v) strengthening resilience against external shocks and climate change. Table ii summarizes the main policy actions that will help the Dominican Republic sustain a high growth trajectory going forward, while at the same time strengthening its resilience against external shocks, natural disasters, and climate change. Table ii. Policy recommendations: key policy areas to accelerate growth Time Reform area Reform measure horizon 1/ Matching educational outcomes with market needs - Modernize curricula at high school level aimed at providing technical skills for an easy intersection to the labor market - Encourage continuing education for adults via flexible programs that allows to Unleashing customize the courses needed at workplace Medium human capital - Facilitate the upskilling and re-skilling of workers through more relevant active labor to long potential market policies term - Allocate enough financial incentives to attract and retain high-quality teachers in rural and disadvantaged areas - Consider the involvement of private sector into the financing of public colleges, such that curricula reflect private sector skills needs - Implement a learning diagnostic system for elementary schools at yearly basis 82 2/ Reducing inequality of opportunities across genders and between rural and urban areas - Consider monetary and non-monetary incentives for parental sharing of caretaking during the first year of birth - Monitor students dropping out – with special emphasis on teenage girls - Consider the introduction of national scholarships for university programs in rural areas for outstanding school students - Support skills building for the vulnerable population through financial assistance programs 1/ Mitigating potential distortions from the direct role of government as SOEs in markets - Assess whether the direct participation of SOEs in markets complies with the subsidiarity role of the State in commercial sectors. - Explore options to bring-in private sector participation through PPPs or management arrangements where SOEs operate in highly concentrated markets that are viable for competition. 2/ Promoting pro-competition regulation - Leverage initiatives like the single window to reduce entry barriers for firms (e.g., reduce number of days to set a new business) - Telecom: Review licencing requirements to facilitate entry of new operators; Incorporate rules that discipline large market players with significant market power, with a particular focus on potentially abusive practices such as bundling and margin squeeze; and improve enforcement to implement existing regulation to facilitate access to essential facilities. - Gas: Develop third party access regulation - Broadcasting: Review requirement of holding a Dominican nationality to control service provision Fomenting - Sugar: Consider eliminating production and export quotas Short to competitive - Electricity: Review tariff schemes in the electricity sector to reflect real production, medium markets transmission, and distribution costs term - Retail: Review the rationale for price controls in retail products and propose alternative, less distortive, mechanisms. 3/ Strengthening competition policy enforcement - Consider developing a merger control framework to limit negative effects of market consolidation. - Include the leniency program as part of the mechanisms to deter cartels in the Competition Law. - Eliminate the maximum duration of investigations (12 months). - Eliminate the prescription of 12 months for anti-competitive conducts to be prosecuted and investigated by the Competition Agency. - Increase the level of sanctions to foster deterrence: calculate sanctions on the basis of turnover/eliminate caps for sanctions. - Implement the ability of ProCompetencia to develop effective coordination and regulation jointly with sector regulators and enhance inter-institutional cooperation mechanisms. - Increase the resources for ProCompetencia including more staff, independence to conduct down-raids, and technical laboratories for processing digital evidence. - Implement measures to mitigate conflict of interest of the competition agency (e.g., cool off periods). 83 Develop tools to identify market distortions of state aid: Implement an inventory of state aid, develop analytical tools to understand competition impact of state aid. 1/ Modernizing the National Innovation System (NIS) - Develop a system of business advisory and technology extension services - Building managerial and organizational capabilities through customized courses for Revamping the local entrepreneurs Medium innovation - Develop basic infrastructure- National quality infrastructure (NQI) and Incubation & term strategy Accelerators - Eliminate barriers to physical and human capital accumulation - Develop a national grant program for small and medium firms that have not benefitted from tax incentives - Limit the use of exception procedures and under the threshold procedures in the national procurement processes, and incentive the use of more competitive Reducing procedures expenditure - Consider the introduction of a national catalogue for standardized products Short inefficiencies - Consider implementing an Early-Warning system for assessing the efficiency of public procurement term in public procurement - Consider introducing a Total Cost of Ownership (TCO), Life Cycle Costing (LLC) and Value for Money (VM) law clauses into the current Procurement Law. - 1/ Reducing fiscal exposure to natural hazards events Strengthening - Develop a fiscal risk strategy aimed at reducing budgetary uncertainty economic - Develop a fiscal strategy for financial protection Short to resilience medium against shocks 2/ Enhancing climate adaptation and mitigation response term and climate change - Adopt programs for greening of public institutions - Create flexible mitigation packages that offset temporary adjustment costs/externalities of decarbonization at regional, company and household level. 84