Poverty and Distributional Impact of Fiscal Policy in Dominican Republic World Bank Group - Poverty and Equity Global Practice © [2023] International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy, completeness, or currency of the data included in this work and does not assume responsibility for any errors, omissions, or discrepancies in the information, or liability with respect to the use of or failure to use the information, methods, processes, or conclusions set forth. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org 1 Contents Executive summary ............................................................................................................................................................... 8 I. Introduction ...................................................................................................................................................................... 15 II. The tax system and government spending in the Dominican Republic in 2018 ........................... 17 A. In spite of several reforms to the tax system, the DR’s public revenues are relatively low 17 B. Collection of direct taxes and social security contributions can be improved .................... 20 C. Indirect taxes are the main source of public income ...................................................................... 20 D. Government spending designed for the poor and vulnerable populations .......................... 21 E. Subsidies and contributions in utilities ................................................................................................... 23 F. Social protection spending........................................................................................................................... 25 G. Education .............................................................................................................................................................. 28 H. Health ..................................................................................................................................................................... 29 III. Methodology, data and assumptions.................................................................................................................. 30 A. Methodology........................................................................................................................................................ 30 B. Data.......................................................................................................................................................................... 34 C. Administrative Data ......................................................................................................................................... 35 D. Allocation Overview..................................................................................................................................................... 35 IV. Main results...................................................................................................................................................................... 39 A. The fiscal system reduces inequality, but the impact could be improved ............................ 39 B. The impact of fiscal policy on poverty is modest .............................................................................. 42 C. Taxes are progressive, some more than others ........................................................................................... 46 i. Direct taxes are highly progressive ......................................................................................................... 48 ii. Indirect taxes are progressive .................................................................................................................... 49 D. Social spending and subsidies .............................................................................................................................. 54 i. Direct transfers are pro-poor....................................................................................................................... 55 ii. Subsidies reach the poor and non-poor alike..................................................................................... 57 iii. Education is the main contributor to inequality reduction ............................................................. 60 iv. Health insurance is reaching the poor, but other programs are doing less......................... 62 V. Conclusion: a progressive fiscal system with space for improvement .............................................. 62 VI. References ...................................................................................................................................................................... 65 2 List of figures Figure 1. Selected countries: Fiscal policy inequality reduction (circa 2015) ............................................... 9 Figure 2. The Dominican Republic: Net payers/beneficiaries (% of market income) ................................. 14 Figure 3. Evolution of Poverty, 2001-2020.................................................................................................. 15 Figure 4. Latin America: tax revenue including social security (2018), as a % of GDP.............................. 20 Figure 5. Latin America: social expenditure (% of GDP) ............................................................................ 22 Figure 6. The Dominican Republic: Electricity tariff schedule 2018 (RD$ per kWh) .................................. 24 Figure 7. Central America and the Dominican Republic: residence electricity tariffs by company ............. 25 Figure 8. Definition of CEQ Income Concepts ............................................................................................ 33 Figure 9. Inequality (Gini coefficient) at pre- and post-fiscal income concepts .......................................... 39 Figure 10. Inequality (Gini coefficient) at pre- and post-fiscal income concepts for national, urban and rural populations .................................................................................................................................................. 40 Figure 11. Selected countries: Fiscal policy inequality reduction ............................................................... 41 Figure 12. The Dominican Republic: Net payers/beneficiaries (% of market income) ............................... 44 Figure 13. Marginal contributions to poverty .............................................................................................. 45 Figure 14. Poverty reduction (US$ 5.5 PPP line): from market to consumable income ............................. 45 Figure 15. Direct and indirect taxes: Concentration curves and share of benefits in poor population, ranked by market income ........................................................................................................................................ 46 Figure 16. Incidence of taxes (% of market income) .................................................................................. 47 Figure 17. Indirect taxes: concentration curves and share of benefits in poor population ......................... 50 Figure 18. Indirect taxes: concentration shares by decile of market income .............................................. 51 Figure 19. Progressivity and redistributive effect of direct and indirect taxes in Latin America .................. 53 Figure 20. Incidence and concentration curves of transfers and subsidies ................................................ 55 Figure 21. Direct transfers: concentration curves and share of benefits in poor population ...................... 56 Figure 22. Direct transfers: Incidence and distribution of beneficiaries ...................................................... 57 Figure 23. Indirect subsidies: concentration curves and share of benefits in poor population ................... 58 Figure 24. Indirect subsidies: incidence and concentration of benefits per decile...................................... 59 Figure 25. Selected countries: progressivity and redistributive effect of subsidies .................................... 59 Figure 26. Education: Distribution of beneficiaries and incidence .............................................................. 60 Figure 27. Education: concentration curves and share of benefits in poor population ............................... 61 Figure 28. Health: concentration curves and share of benefits in poor population ..................................... 62 3 List of tables Table 1. Total revenue of the Dominican government in 2018 ..................................................................... 18 Table 2.Total expenditure of the Dominican government in 2018............................................................... 21 Table 3. Changes in poverty........................................................................................................................................... 42 Table 4. Population and income shares of market income by income groups, US$ PPP .............. 43 Table 5. Market income: share and accumulated share by decile before and after taxes and transfers .................................................................................................................................................................................... 43 Table 6. Incidence of Direct Taxes as % of Market Income ........................................................................... 49 Table 7. Incidence of indirect taxes, income groups (as % of market income) .................................... 50 Table 8. Dominican Republic: concentration of expenditures according to VAT exemption and place of purchase ................................................................................................................................................................. 52 4 Abbreviations ADESS Administrator for Social Subsidies ARS Health Risk Administrators BEEP Student Progress Bonus (Bono Escolar Estudiando Progreso) CAASD Corporation of Aqueduct and Sewage System of Santo Domingo CCT Conditional Cash Transfer CDEEE Corporation of State Electrical Companies CEQ Commitment to Equity CORAASAN Corporation of Aqueduct and Sewage System of Santiago CPI Consumer Price Index DGII General Directorate of Internal Taxation DR Dominican Republic ECLAC Economic Commission for Latin America and the Caribbean EDE Dominican Republic Power Distribution Company ENDESA Demographic and Health Survey ENCFT National Continuous Labor Force Survey ENGIH National Household Expenditure and Income Survey 2018 ENIGH National Survey of Household Income and Expenditure 2006-07 GDP Gross Domestic Product GNI Gross National Income ICV Life Quality Index IDSS Dominican Institute of Social Security IES Incentive to Tertiary Education Program (Incentivo a la Educación Superior) ILAE Incentive for School Attendance Program Regions INABIMA National Teacher Welfare Institute (public pension) INAPA National Institute of Water and Sanitation ITBIS Tax on the Transfer of Industrialized Goods and Services 5 LAC Latin American and the Caribbean LPG Liquefied Petroleum Gas MESCyT Ministry of Education, Science and Technology (Ministerio de Educación Superior, Ciencia y Tecnología) OECD Organisation for Economic Co-operation and Development OMSA Metropolitan Office of Bus Services ONE National Office of Statistics OPRET Office for the Reorganization of Transportation (in charge of metro and cable car services) PAE School Food Program (Programa de Alimentación Escolar) PIAMG Marine Officials Incentive Program PIPP Preventing Police Incentive Program (Programa Incentivo de la Policía Preventiva) PPP Purchasing power parity PROMESE/CAL Program of Essential Drugs, logistics support PROSOLI Solidarity program RD$ Dominican Republic Pesos SENASA National Health Insurance Authority SIE Electricity Superintendence SIPEN Pension Superintendence SISALRIL Bureau of Occupational Health and Safety SIUBEN Unified Beneficiary Selection System SNS National Service of Health VAT Value Added Tax 6 Acknowledgements This report was led by Alejandro de la Fuente and prepared by Maynor Cabrera. The team is grateful for the many inputs and comments provided by Jaime Aristy-Escuder, Gabriela Inchauste, Christian Gomez Canon, Haydeeliz Carrasco, Pedro Rodriguez, James Sampi and Johannes Herderschee. The report was prepared at the request of the Government of the Dominican Republic and was partially funded by the French Development Agency (Agence Française de Développement, AFD). 7 Executive summary This report assesses the impact of fiscal policy, both revenue and expenditure, on inequality and poverty in the Dominican Republic. On the revenue side, the analysis focuses on the personal income tax, the value added tax (tax on the transfer of industrialized goods and services, known as ITBIS in the Dominican Republic for its initials in Spanish) and excise taxes on alcoholic beverages, cigarettes, fuel products and telecommunication services. These taxes combined accounted for 7.8 percent of GDP in 2018, equivalent to 60 percent of total tax revenues. On the expenditure side, the analysis focuses on social protection benefits like direct cash and near-cash transfers (e.g., the school food-program and the school uniforms and supplies program), indirect subsidies (energy, water and public transport), and in-kind benefits on education and health, which together account for 39.2 percent of total government expenditures and 85.9 percent of social expenditures. To assess the fiscal system’s poverty and inequality impacts, the study estimated “pre- fiscal� and “post-fiscal� income measures. It follows the methodology developed by the Commitment to Equity (CEQ) Institute to assess fiscal policy (Lustig, 2018). To examine the amount of redistribution accomplished (among others) and, therefore, the fiscal system’s impact on poverty and inequality, the study creates measures of pre-fiscal and post-fiscal incomes. The pre-fiscal income measure (market income) is constructed by totaling for an individual all its private sources of income: labor income, private transfers receipts, capital income, imputed value of own housing and self-consumption, before any direct transfers -in-kind benefits or subsidies- and payment of taxes by a given household. “Post-fiscal� income takes pre-fiscal or market income and adjusts it by the resulting income after taking into account a series of revenue policies (including personal income taxes, the value added tax (VAT), and selected excise taxes) and expenditure programs, including public spending on health and education, selected subsidies, and unconditional cash transfers. Poverty and inequality measures are then derived for scenarios with pre- and post-fiscal interventions. The primary data used for calculating pre- and post-fiscal incomes comes from the 2018 National Household Expenditure and Income Survey (ENGIH). This report also uses health data to calculate and impute the propensity to visit public healthcare providers. For this, shares of all recorded medical visits by household type comes from the 2013 Demographic and Health Survey (ENDESA). Additional administrative data comes from the Ministry of Finance, the General Directorate of Internal Taxation (DGII), the Ministry of Education, the Bureau of Occupational Health and Safety (SISALRIL), the Central Bank and other official sources (Section III). The fiscal system’s net impact on inequality is positive, but could be improved The report found that in 2018 Fiscal Policy reduced income inequality in the Dominican Republic. Taxes, indirect subsidies, and transfers (including the monetized value of education and health) benefited more the population that was either poor or vulnerable to poverty relative to 8 those that were non-poor.1 Measured through the Gini coefficient, fiscal policy reduced inequality by 8.6 points, from 0.459 to 0.374, with greater benefits accruing to rural populations. A reduction in income inequality was still evident after excluding in-kind benefits (education and health services), with the Gini coefficient falling by 2.8 points. Of all the fiscal measures considered in this report, primary and secondary education, the national health system, the school food program (Desayuno Escolar), and the personal income tax are the interventions which reduce inequality the most. Primary and secondary education transfers reduce the Gini coefficient by 2.1 and 1.1 points, respectively. Health transfers (through the National Health System) also have a significant equalizing effect, reducing the Gini coefficient by 1 point. The school feeding program and the personal income tax reduce the Gini coefficient by 0.7 and 0.6 points, respectively. The excise tax on fuel (primarily consumed directly by the rich) and the ITBIS also reduce inequality, but to a lesser extent. On the other hand, subsidies to metro and water services mainly benefit the middle class and, as a result, do not decrease inequality at the national level. The overall reduction on inequality achieved by fiscal policy in the Dominican Republic is similar to the regional average in Latin America and the Caribbean (LAC). Fiscal systems reduce inequality in most developing countries, with some exceptions in Sub-Saharan Africa., In terms of the redistributive impact of its fiscal policy (including the monetized value of education and health services), the Dominican Republic ranks 9th out of 18 countries in Latin American countries. Many countries in LAC display a Gini similar to the one from Dominican Republic. In Ecuador, for instance, inequality drops by 8 points when education and health insurance are included. In others like Bolivia and Peru inequality gets reduced to a lesser degree. Figure 1. Selected countries: Effect of fiscal policy on inequality reduction (circa 2015) 1Those whose income per capita per day falls between US$11.50 and US$57.60 PPP based on estimations made by CEQ Institute using the methodology developed by López-Calva and Ortiz-Juarez (2014). 9 Source: CEQ Data Center (https://commitmentoequity.org/datacenter/) The impact of fiscal policy on poverty is modest Fiscal policy reduces poverty, albeit modestly (by 1.0 percentage point, using the National Poverty Line). Poverty reduction is driven primarily by the Desayuno Escolar, the cash program Comer es Primero and subsidies to electricity. While the ITBIS and fuel excise taxes reduce inequality, they increase poverty slightly as the vulnerable and poor populations still pay ITBIS for the purchase of goods and services and consume fuel (by purchasing in oil stations but mainly indirectly through the purchase of goods and services that use oil as an input). By international standards, the share of the Dominican population lifted from poverty (using the US$5.5 PPP 2011 threshold) as a result of fiscal policy interventions is also small (1.3 percentage points), relative to the impact of fiscal policy in other Latin American countries. The share of Dominicans lifted from poverty (using the US$6.85 PPP 2017 threshold) as a result of fiscal policy interventions is also small (0.9 percentage points).The Dominican government can improve the capacity of its fiscal system to reduce poverty, mainly through an increase in public revenue, which could then be transferred to the poor and vulnerable populations. Taxes are progressive, some more than others Both direct and indirect taxes reduce inequality. Direct taxes (i.e., taxes on personal income, interest, and dividends) fall mainly on taxpayers in upper- and middle-class households and are highly progressive since they affect the poor less than other segments of the population. However, the impact of direct taxation on inequality is small: It represents only 1.4 percent of average household market income, and a small portion of the government’s total tax revenues (9.4 percent). Indirect taxes are the main source of revenue in the Dominican Republic, representing 68 percent of total tax revenues. The report analyzed the impact of the ITBIS (VAT), and of excise 10 taxes on alcoholic beverages, cigarettes, fuel products and telecommunication services on inequality. Indirect taxes are progressive because they affect the poor less than other segments of the population. Nevertheless, indirect taxes reduce the population’s market income by an average of 9.3 percent, driving some people into poverty. The main tax driving the impact is ITBIS which is paid by everyone. Social spending and subsidies reduce inequality In terms of government spending, most direct transfers are pro-poor because the population living below the poverty line receives more benefits in absolute terms than other groups. The conditional cash transfer (CCT) for tertiary education (known as IES), contributory pensions, and the bonus for drivers (Bonogas Chofer) are progressive (i.e., the benefits for lower income households as a share of their total income are higher than for better- off households). But, they are not pro-poor. Indirect subsidies are not pro-poor in Dominican Republic because they are proportional to income. All subsidized services in the country have tariffs that do not fully cover their costs. Electricity and water also have implicit subsidies for irregular connections, fraud and nonpayment. Companies receive transfers from the central government to cover these costs and keep tariffs low. In the case of electricity, there is a tariff scheme that favors the low consumption of electricity. The water subsidy is regressive. The subsidy on metro services (offered by the public entity known as OPRET) benefits the urban population living in Distrito Nacional and Santo Domingo and is neutral. As the poorest populations in rural and other urban centers do not use this service, they do not benefit from this indirect subsidy. However, when only the population of Santo Domingo is considered, the subsidy is progressive. In contrast, the subsidy on urban bus services (offered by the public entity known as OMSA) is progressive. Even though this program operates mainly in urban centers, it benefits the lower and middle deciles of the population, and their services also reach Santiago’s population (Cibao Region). The electricity subsidy is slightly progressive. When compared with counterparts in Latin America and the Caribbean, the Dominican Republic’s indirect subsidies have a lower redistributive impact. Education transfers are pro-poor, as is the subsidized family health insurance. The benefits of education are pro-poor for all educational levels, except in tertiary education. At the tertiary level, education transfers are progressive as benefits as a share of the market income are higher for the poor than for the rich. Primary education is the most pro-poor intervention, followed by pre- primary and the literacy program, Quisqueya Aprende Contigo. Secondary education was also pro-poor in 2018. Health expenditures are less progressive than education and only the subsidized family health insurance was found to be pro-poor2. 2According to Aristy-Escuder, et al. (2018) “Progressivity benefits the poorest segments of population, but it could be an indicator of other social trends in education and health care. Those with higher incomes might be opting out for private education and, in the case of health, participate in contributive health insurance schemes �. Sánchez-Martín 11 and Senderowitsch (2012, p. 13) explained that “the education sector in the DR presents faulty public service delivery, which originates a private offer that is more of a reactive upshot to deficiencies in state education than a high quality alternative (at least not in every case).� See a discussion about this topic in Sánchez-Martin and Senderowitsch (2012) 12 Summary of Fiscal Policy Income Distribution and Poverty Impacts National USD 5.5 PPP National Poverty Extreme Poverty Line Poverty Line Line Market Income 22.1% 4.3% 16.6% Consumable Income 21.1% 2.9% 15.2% Poverty reduction/1 1.0% 1.3% 1.3% Of which (marginal contribution)/2 Desayuno escolar 1.9% 1.0% 1.4% Comer es primero 0.9% 0.2% 0.7% Indirect Tax ITBIS -1.6% -1.3% -1.2% Gini Market Income 0.459 Final Income 0.431 Inequality reduction 0.086 Of which (marginal contribution) /2 In-Kind education benefits: Primary 0.021 In-Kind education benefits: Secondary 0.011 In-Kind health benefits: National Health System 0.009 Desayuno escolar 0.008 Comer es primero 0.003 Indirect Tax ITBIS 0.004 Source: estimates based on ENGIH 2018 and official sources Notes: 1/ CEQ methodology does not estimate headcount poverty rates for Final Income 2/ Positive marginal contributions mean that the fiscal intervention reduces poverty or inequality. When marginal contributions are negative the effect is the opposite. Overall, most Dominicans are net beneficiaries of the country’s fiscal policy. Most of the population—i.e., first eight income deciles—are net beneficiaries of fiscal policy when all benefits and taxes are applied (Figure 2, right panel). The benefits are higher for the poor and the highest impact comes from educational and health in-kind transfers. The net benefits are not significant for the highest deciles (deciles 9-10), which comprise the upper middle class and wealthy income groups. The impact of direct transfers is only noticeable for the first five deciles, which comprise the poor and vulnerable income groups. The fiscal system as a whole has a neutral impact on the middle-classes, as taxes are compensated exactly by the benefits received. However, when the benefits from health and education are excluded, only the first three deciles of the population are net beneficiaries of the country’s fiscal system (Figure 2, left panel). Upper deciles become net payers. 13 Figure 2. The Dominican Republic: Net payers/beneficiaries (% of market income) Net incidence, all taxes and transfers except health and education benefits Net incidence, all taxes and transfers 25% 150% 20% (% of market income) 15% (% of market income) 100% 10% 5% 50% 0% -5% 0% -10% -15% -50% -20% 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Decile Decile All Direct transfers All Direct Taxes All Education All Health All Indirect Taxes All Indirect Subsidies All Indirect Subsidies All Indirect Taxes All Direct Taxes All Direct transfers All net transfers and subsidies All net transfers and subsidies Source: Authors' estimates based on ENGIH 2018 14 I. Introduction The Dominican Republic (DR) experienced strong and sustained growth, coupled with declining poverty rates, for more than a quarter century—until the economic crisis unleashed by the COVID-19 pandemic. The country grew at an average rate of 5.3 percent per year from 1993-2018. Meanwhile, the rate of poverty dropped from 43 percent in 2008 (the first year with comparable data available) to just 29 percent in 2016, and the rate of extreme poverty dropped from 13 percent to 6 percent over the same period.3 Improvements were more pronounced in rural areas, where poverty decreased from 46 percent in 2001 to 38 percent in 2016. Starting in 2017, national poverty figures experienced a break in comparability for methodological reasons. Nonetheless, recent comparable figures between 2017 and 2019 also show continued reductions in general and extreme poverty, as they fell 4.6 and 1.1 percentage points, respectively (Figure 3). At the same time, income inequality improved: the Gini coefficient dropped from 0.498 to 0.471 (-2.7 Gini points) between 2001 and 2016 and then from 0.438 to 0.431 (-0.7 points) between 2017 and 2019. Figure 3. Evolution of Poverty, 2001-2020 70 60 50 45.8 Percentage 40 32.7 30 30.1 24.0 23.4 24.9 23.2 20 10 0 2005 2016 2001 2002 2003 2004 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Official National Official Rural Official Urban 5.5 USD PPP11 line Source: ONE, MEPYD, Central Bank and World Bank estimates Note: 2020 figures are official preliminary estimate based on the Poverty Bulletin No. 8 by MEPyD. 3Poverty jumped from 32.4 percent in 2002 to 49.5 percent in 2004 due to financial crisis in 2003, which reduced real wages and employment. 15 The COVID-19 crisis is halting the progress on poverty reduction and may reverse the gains made over the past decade. Preliminary estimates indicate that the ongoing economic crisis caused the loss of around 280,000 jobs, representing up to 6 percent of the total employment. Based on official preliminary estimates, the spike in unemployment and economic inactivity increased poverty from 21 percent to 23.4 percent and extreme poverty from 2.7 percent to 43.5 percent. This entailed an increase in of 270,000 in the number of people living below the poverty line, even after mitigation measures were enacted.4 In the context of the crisis, fiscal policy could help attenuate the negative impacts on poverty and inequality.5 This could be achieved both through the government’s overall fiscal position—for instance, both by avoiding a loose fiscal policy that could lead to macroeconomic instability and higher inflation which is the worst tax on the poor and through tax policy and public spending that is carefully targeted for its distributional implications. Experience shows that public spending that is targeted to poor and vulnerable households may achieve a broader distribution of the benefits of economic growth.6 For instance, taxes and transfers have played a significant redistributive role in high-income countries and have somewhat attenuated high inequality. Fiscal policy could also help to attenuate the negative distributional impacts of the current crisis by, for instance, increasing existing social programs. At the same time, as in most of LAC, a policy imperative for DR will be to carry out some form of fiscal adjustment coming out of the crisis. Some of the fiscal policy tools that the government has at hand could be relevant for this purpose. The release of the 2018 National Household Expenditure and Income Survey (ENGIH) in 2020 provides an opportunity to assess the welfare impact of fiscal policy in 2018. The 2020 poverty estimates already reflect the impact of policy measures adopted by government last year to attenuate the impact of the crisis. Analyzing some of those policy instrument--even those taken in 2018 --could shed some light on their progressivity and inform the government as to whether the policies they adopted were appropriate. At a later stage, the data from 2018 could also help to project into 2020 the income aggregate and various fiscal policy instruments. Projections such a these could serve as baseline scenarios for simulating policy measures that attenuate the impact that the current crisis is having on fiscal accounts while protecting the poor. However, this type of simulation work is beyond the scope of this report. Data on the expenses and income of a sample of 8,881 Dominican households were obtained from the 2018 ENGIH. The analysis in this report is a significant advance in relation to the previous analysis of its kind, which had applied the 2013 tax structure to the database provided by the 2007 ENIGH. This new assessment for the DR has some improvements in comparison with the latest previous study of this kind (Aristy-Escuder, Cabrera, Moreno-Dodson, & Sanchez-Martin, 2018): while the last fiscal incidence analysis applied estimations based in 2006/2007 survey to fiscal concepts in 4 MEPyD (2021) Poverty Bulletin No. 8. 5 A similar study on the poverty and distributional implications of the fiscal system in the country conducted in 2014 using the 2007 ENGIH concluded that the fiscal system was progressive overall, but its impact on poverty reduction was modest. 6 Defined as those vulnerable to falling into poverty and the lower bound of the middle class in Latin American countries based on estimations by López-Calva and Ortiz-Juarez (2014). Those whose annual income per capita falls between US$11.50 and US$57.60 PPP based on estimations made by CEQ Institute. 16 2013, the analysis in this report applies the household survey data to the fiscal concepts from the same year, including the indirect effects for indirect taxes and the fuel tax; an enhanced approach to estimate evasion and exemption incidence of ITBIS; and the estimation of the effect of subsidies for metro, water and urban transport. This report follows the methodology developed by the Commitment to Equity (CEQ) Institute (Lustig, 2018). “Pre-fiscal� and “post-fiscal� income measures are estimated for this analysis to then compare poverty and inequality measures. As discussed below (Section III, Box 1), the pre-fiscal measures comprise market income before any transfers or taxes are applied. “Post-fiscal� income takes pre-fiscal income and adds to it a subset of fiscal policies: subsidies and direct transfers received, direct and indirect taxes paid, and in-kind transfers received through use of services. The CEQ methodology applied in this report corresponds to data from 2018, but its approach could set the foundations for carrying out policy simulations on reforms that the government may be considering to weather the impacts of the ongoing crisis. However, this type of exercise and its outputs would be distinct from the technical background work provided in this report. The remainder of this report is organized as follows: Section II describes the Dominican Republic’s tax systems and government spending in 2018 and compares them with those of selected Latin American countries. Section III includes a description of the data, methodology and assumptions made in carrying out the analysis in this report. The main results are provided in Section IV, starting with fiscal policy’s net impact on inequality, followed by its impact on poverty incidence. A comparison with other countries is then provided. Section IV also includes a detailed analysis of the distributional impact of taxes, social spending, and subsidies, to demonstrate their impact on the welfare of the poor. The report’s main conclusions are presented in Section V. II. The tax system and government spending in the Dominican Republic in 2018 The fiscal system comprises a set of taxes and social expenditures. On the tax side, the analysis in this report considers fiscal instruments such as personal income taxes, the value-added tax (tax on the transfer of industrialized goods and services known as ITBIS in the DR for its initials in Spanish), and excise taxes on alcoholic beverages, cigarettes, fuel products and telecommunication services. On the government expenditure side, the benefits to households that are analyzed include social protection benefits like direct cash and near-cash transfers (e.g., the school food-program and school uniforms and supplies program), indirect subsidies (water, public transport and electricity), and in-kind benefits on education and health. An overview of the size and composition of each of these main fiscal tools, as of 2018, is provided below. A. In spite of several reforms to the tax system, the DR’s public revenues are relatively low Total fiscal revenue in 2018 amounted to 14.2 percent of GDP. Table 1 provides a snapshot of public revenue sources in the 2018 fiscal year. Tax revenues are the most important source of revenue (close to 91 percent of total revenues) followed by non-tax revenues (8 percent of total 17 revenues). Despite several tax reforms that were implemented between 2001 to 2012, the 2018 tax-revenue collection amounted to RD$549.9 billion (13.0 percent of GDP), which is one of the lowest revenue collection as a percentage of GDP in LAC (OECD, 2022).7 This low collection is explained by high levels of evasion of both the ITBIS and income tax. According to the World Bank, the Dominican Republic's "compliance gap is the single largest factor explaining its large potential efficiency gap� (World Bank, 2017, p. x). When viewing direct taxes paid only, collections total RD$176.1 billion, equal to 4.2 percent of GDP in 2018 (Table 1, Column 2). In addition, indirect taxes comprise RD$373.7 billion, equal to 8.8 percent of GDP. The government still collects social security contributions (0.1 percent of GDP), even though the private sector runs health and old-age pension systems. Table 1. Total revenue of the Dominican government in 2018 Total in Portion of fiscal accounts Fiscal Accounts ENGIH analyzed Ratio between Survey survey total Included and fiscal accounts RD$ Million % GDP RD$ Million % GDP % of total RD$ Million analyzed (%) Total Revenue & Grants 601,120 14.2 332,472 7.85 55.3 163,074.0 27.1 Taxes 549,942 13.0 332,472 7.85 60.5 163,074.0 29.7 Direct Taxes 176,153 4.2 51,425 1.21 29.2 22,330.2 12.7 Personal Income Tax Yes 51,425 1.2 51,425 1.21 100.0 22,330.2 43.4 Others (Corporate Income Tax, No 124,727 2.9 0 - Property Tax) Indirect Taxes 373,713 8.8 281,046 6.63 75.2 140,743.8 37.7 ITBIS Yes 194,725 4.6 194,725 4.60 100.0 81,573.5 41.9 Excises Yes 93,731 2.2 86,322 2.04 92.1 59,170.3 63.1 Custom Duties No 39,269 0.9 Other indirect taxes No 45,989 1.1 Social Contributions No 2,514 0.1 Non-Tax Revenue No 47,699 1.1 Grants No 965 0.0 Source: Ministry of Finance (Annual Statement of operations), General Tax Directorate (Statistical Bulletin), National Statistical Office (ONE), Central Bank, Internal calculations based on the 2018 ENGIH Notes: Excises include taxes on fuel products, alcoholic beverages, beer, tobacco, telecommunications. This report focuses on the impacts of the personal income tax, the ITBIS and excise taxes. Government revenues considered in this report’s analysis amount to 7.85 percent of GDP and 60.5 percent of tax revenue in 2018.8 The most important source of tax revenue is the ITBIS, which accounts for approximately 35 percent of tax revenue. The final column of Table 1 presents the ratio between the total collections obtained in the household survey and the one reported by 7According to the OECD (2022), Guatemala and the Dominican Republic had the lowest tax burdens in the region in 2018, below the regional average (22.5 percent). The tax collection in the Dominican Republic was 13.2 percent of GDP, and includes taxes and social contributions but excludes local government revenues. 8Corporate income taxes, as well as import taxes, were not included because it is difficult to assign the economic burdens to any single individual or household. 18 the administrative data. One limitation of this analysis is that the ratios are significantly lower than 100 since the survey captures only taxes paid by households while the administrative data consider collections from other organizations, such as governments, NGOs, etc. Also, like other household surveys, the ENGIH likely does not capture the top of the income distribution. This may partly explain the low collection rate reported above. The Dominican Republic has one of the lowest tax burdens in the region, below the average among LAC countries of 22.5 percent of GDP in 2018 (OECD, 2022). Direct tax collection is one of the lowest and indirect tax collection is below the regional average (11.5 percent of GDP in year 2018, according to OECD). When social contributions are excluded, the gap is lower: the average among LAC countries is 19.1percent compared with 13.0 percent in the Dominican Republic. Tax revenues are lower compared with countries that have similar GNI per capita PPP levels, such as Mexico and Colombia. 19 Figure 4. Latin America: tax revenue including social security (2018), as a % of GDP 35.0 35000 30.0 30000 25.0 25000 20.0 20000 15.0 15000 10.0 10000 5.0 5000 0.0 0 Direct taxes Indirect taxes Other taxes Social securitiy contributions GNI per capita PPP Source: OECD-Stat and World Bank Data B. Collection of direct taxes and social security contributions can be improved Income taxpayers, as defined by tax code, are individuals, legal entities or non-resident, non- domiciled entities with a permanent establishment within the Dominican territory. Personal income tax, defined by Law 11-92, is applicable to wages, income or net business income of individuals, and applies a set of progressive rates ranging from 15 to 25 percent, depending on the value of net taxable income. For salaried workers, some deductions are allowed, such as a personal allowance or a basic tax exemption limit (close to US$8,400 in year 2018), social security contributions, Christmas bonus (“regalia pascual�); other than employment activities other expenses specified in the Tax Code may be deducted. There are also taxes for income obtained for prizes, housing rent, interest, and dividends. The social security system was reformed in 2001, “passing from a pay-as-you-go (PAYG) model to an individual capitalization account model including a subsidized regime for the most vulnerable� (Lavigne & Vargas, 2013, pp. 8). In 2018, around 3.9 million individuals, 37 percent of the total population, were affiliated with the pension systems. Only 6 percent of them belonged to public schemes—like the ones offered by the Central Bank, the pension program for teachers (INABIMA for its Spanish initials), the Ministry of Finance, or the Reserve Bank (BanReservas)— but only 48.2 percent of those affiliated with these pensions were contributors (SIPEN, 2019). In addition, the Dominican government still collects contributions to the old pay-as-you-go system, but, as of 2018, they were as low as RD$42.6 million (close to US$0.9 million). C. Indirect taxes are the main source of public income The value-added tax (ITBIS) is the single largest contributor to tax revenue in the Dominican Republic. The ITBIS that is collected by customs authorities represents 44.4 percent of the total 20 ITBIS revenues during 2017-2019, while the rest is collected via domestic transactions (average of 54.6 percent in 2017-2019). The general rate for this VAT is 18 percent, and the reduced rate is 16 percent.9 Goods and services that are exempt from this tax include goods such as: fuel oils, medicine, educational materials, living animals, and a variety of food items, among others; and services such as financial services (including insurances), pensions plans, transport, electricity, water, garbage collection, health, education, cultural activities, funeral services, beauty parlors, and housing rentals. The tax expenditure from ITBIS was 2.9 percent of GDP in 2018 (Ministerio de Hacienda, 2017, pp. 27). The excise taxes applied to consumption (known as ISC for their initials in Spanish) include levies on alcoholic beverages, tobacco, telecommunications, oil derivate products, insurance and some financial services. Rates applied to those products are ad valorem for insurance (16 percent), telecommunications (10 percent), and there is a debit tax on 0.15 percent of the value of each payment (RD$0.0015 for each peso) made through the financial system (checks and electronic payments). Alcoholic beverages have both an ad-valorem tax rate (10 percent) and a fixed rate10. Tobacco pays an ad-valorem rate (20 percent) plus a fixed tax of RD$51.97 on a pack of 20 cigarettes or RD$25.98 on a pack of 10, on average during 2018. For fuel taxes, there are one ad valorem tax (16 percent) applied on fossil fuels and oil derivates, and a two specific rates applied to fuel consumption11. D. Government spending designed for the poor and vulnerable populations The total fiscal expenditure of the non-financial public sector in 2018 was RD$699.5 billion, which represents about 16.5 percent of GDP (Table 2, Column 2). Table 2 presents the composition of the Dominican government's expenditures in 2018. Public social spending totaled around 7.5 percent of GDP. Drilling down, at 3.7 percent of GDP, education accounted for almost half of that spending. Meanwhile, health and social protection followed in importance, accounting for 1.6 percent and 1.5 percent of GDP, respectively. Together, these expenses represented 41.4 percent of the total social expenditures (health 21.7 percent and social protection 19.7 percent). Social security insurance on pensions and health were privatized in 2001 and have been since phased out or complemented by private capitalization pension systems and family health insurance. Finally, spending in other sectors—such as housing, community services, sports, recreational, culture and religion activities—amounted to 6.7 percent of total social expenditures and 0.5 percent of GDP. Table 2.Total expenditure of the Dominican government in 2018 Total in HHD Ratio between Fiscal Accounts Portion of fiscal accounts analyzed Survey survey total and 9 The reduced rate is applicable to certain type of yogurts, margarine, coffee, shortenings and food oils, sugars, and chocolates. 10 The fixed rate is adjusted every three months. The average rate during 2018 was RD$614.33, close to US$12.42 11 The fixed rate for each gallon of regular gasoline was RD$63.85 (around US$1.29) and RD$71.85 for premium gasoline (close to US$1.45). Those fixed rates include a specific rate applied to fuel consumption (RD$2 per gallon). There are different rates for diesel, fuel oil, kerosene, and jet fuel oil. 21 fiscal accounts Included RD$ Million % GDP RD$ Million % GDP % of total RD$ Million analyzed (%) Total Expenditures 699,545 16.5 274,456 6.5 39.2 240,365.9 34.36 Social protection 62,641 1.5 62,641 1.5 100.0 36,141.3 57.70 Contributory pensions Yes 29,343 0.7 29,343 0.7 100.0 2,963.0 10.10 Conditional & unconditional cash transfers 15,010 0.4 15,010 0.4 100.0 14,885.4 99.2 Comer es primero Yes 8,316 0.2 8,316 0.2 100.0 8,318.8 100.0 ILAE Yes 709 0.0 709 0.0 100.0 699.8 98.6 IES Yes 127 0.0 127 0.0 100.0 126.6 99.7 Bonogas Hogar Yes 2,587 0.1 2,587 0.1 100.0 2,487.3 96.1 Bonogas Chofer Yes 603 0.0 603 0.0 100.0 585.6 97.2 PIPP Yes 218 0.0 218 0.0 100.0 199.9 91.7 Bonoluz Yes 2,053 0.0 2,053 0.0 100.0 2,052.6 100.0 PROVEE Yes 398 0.0 398 0.0 100.0 396.7 99.8 Other Direct Transfers Yes n.a. n.a. n.a. n.a. n.a. 17.9 Near Cash Transfers 18,287 0.4 18,287 0.4 100.0 18,293.0 100.0 School food program (Desayuno escolar) Yes 16,372 0.4 16,372 0.4 100.0 16,375.6 100.0 School uniforms and supplies program Yes 1,915 0.0 1,915 0.0 100.0 1,917.3 100.1 Education 158,244 3.7 117,295 2.8 74.1 109,704.7 69.3 Pre-school Yes 9,283 0.2 6,760 0.2 72.8 7,373.0 79.4 Primary Yes 71,571 1.7 53,169 1.3 74.3 57,942.3 81.0 Secondary Yes 47,124 1.1 40,736 1.0 86.4 44,389.5 94.2 Post-secondary Non-Tertiary No 441 0.0 0 - Tertiary Yes 16,630 0.4 16,630 0.4 100.0 16,634.5 100.0 Other n.p.c. No 13,195 0.3 0 - Health 69,015 1.6 69,015 1.6 100.0 69,014.9 100.0 National Health System Yes 37,570 0.9 37,570 0.9 100.0 37,570.1 100.0 Family Health Insurance - Subsidized Yes 9,467 0.2 9,467 0.2 100.0 9,467.0 100.0 IDSS pensioned Yes 575 0.0 575 0.0 100.0 574.6 100.0 Promese-CAL Yes 2,761 0.1 2,761 0.1 100.0 2,760.5 100.0 Other Yes 18,643 0.4 18,643 0.4 100.0 18,642.6 100.0 Subsidies 26,320 0.6 25,505 0.6 96.9 25,505.0 96.90 Electricity Yes 17,068 0.4 17,068 0.4 100.0 17,068.2 100.0 Water Yes 4,612 0.1 3,797 0.1 82.3 3,797.1 82.3 Public Transport Yes 4,640 0.1 4,640 0.1 100.0 4,639.6 100.0 Other No 0 - - Other expenses No 6,023 0.1 0 - Sources: Ministry of Finance (Annual Statement of operations), ONE, Central Bank, internal calculations based on the 2018 ENGIH. Note: Other direct transfers include Bono Escolar Estudiando Progreso (BEEP) and Programa de Incentivo a la Marina de Guerra. BEEP was not distributed in 2018, but some households reported benefits in survey. The Dominican Republic's social expenditure levels are lower compared with most other Latin American countries, except for those in Guatemala that are lower, and in Peru that are comparable. Spending falls below the levels seen in countries with similar GNI per capita PPP, like Colombia and Mexico. Drilling down, education and indirect subsidy expenditures are close to regional averages (Figure 5). Health, social assistance (direct transfers) and contributory pension expenditures are lower with respect to other countries in LAC. However, contributory pensions are lower than those of other countries with privatized pensions systems, like Chile and El Salvador, and health expenditures are lower when compared with other countries with mandatory private health insurance schemes, like Colombia and Chile. Figure 5. Latin America: social expenditure (% of GDP) 22 30.0% 35000 25.0% 30000 25000 20.0% 20000 15.0% 15000 10.0% 10000 5.0% 5000 0.0% 0 Education Direct transfers Health Contributory pensions Subsidies GNI per capita PPP Source: CEQ Institute’s Standard Indicators and official estimates using data from the Dominican Republic Government E. Subsidies and contributions in utilities Electricity Electricity in the Dominican Republic is highly subsidized. In 2018, the government transferred US$850.5 million to the electricity sector, of which US$385.6 million were current transfers12. Over the past 18 years, the electricity sector's financial deficit, which has hovered around 1.4 percent of GDP a year, explains 50 percent of the country's increase in public debt13. This deficit is explained by the combination of historically high energy costs and low collections by its Electricity Distribution Companies (known as EDEs for their initials in Spanish). With the support of international financial organizations (e.g., the World Bank and the Inter-American Development Bank), the government has invested in several programs to increase collections and reduce the EDEs' losses. At the same time, between 2014 and 2020, the government built a new, 752 MW thermoelectric coal power plant (Central Termoeléctrica Punta Catalina) to reduce the generation price of its electricity. This power plant started operating at full capacity in May 2020, cutting electricity prices in the spot market by half. It is expected that this project will reduce the electricity sector's financial deficit to about 0.5 percent of GDP. The Dominican Republic’s power supply stood at approximately 19.651 GWh In 2018.14 There is an inverted block tariff scheme where low-consumption households pay a subsidized tariff. However, the government pays for services provided to a group of households that do not have metered connections. The estimated subsidy rate for the consumer within the 0-200 kWh range 12 See Ministerio de Hacienda (2018, p.32 and p.35) 13 The electricity sector deficit represented 1 to 2.3 percent of GDP during 2014-2021, see International Monetary Fund, July 2022 Article IV Consultation, Country Report No. 22/217. / 14 See https://www.cne.gob.do/estadisticas-energeticas/informacion-estadisticas/ 23 amounts to 47 percent of the estimated technical tariff (RD$10.25 per kWh)15 and falls to 17 percent for the consumer within the 601-700 kWh range. The technical tariff is the electricity price that would allow the EDEs to cover the full cost of electricity supplied to consumers. Electricity tariffs for the Dominican Republic are the lowest in Central America, according to ECLAC (2019), followed by Panama and Costa Rica (see Figure 6). Figure 6. The Dominican Republic: Electricity tariff schedule 2018 (RD$ per kWh) Source: EDE 15The technical tariff was obtained by adding the average 2018 energy purchase cost of the three EDEs, which was RD$6.61 per kWh, to the average distribution value added for that year (US$0.0735 per kWh, equivalent to RD$3.64 per kWh), established by resolutions SIE-061-2017-TF and SIE-087-2018-TF. 24 Figure 7. Central America and the Dominican Republic: residence electricity tariffs by company (US cents per kWh for a monthly consumption of 200 kWh) (2018) Source: (ECLAC, 2019) Water and other subsidies Water services are delivered by public companies that charge tariffs that are too low to permit recovery of all costs and capital depreciation. During the 2010-2015 period, supplier entities barely covered a third (34 percent) of their operating costs.16 In addition to energy and water, there are subsidies for public transportation—including buses, metro and cable car urban transit systems. In 2018, the government transferred to the Metropolitan Office of Bus Services (OMSA for its initials in Spanish) a total of RD$1,387.9 million to cover the difference between its income and expenses,17 supplying the service to 31 million passengers, of which 22.1 million paid fares18. In the same year, the government transferred RD$4,512 million to the office in charge of metro and cable car services (OPRET), which served a total of 89.6 million passengers. F. Social protection spending Social spending amounted to RD$62.6 billion or 1.5 percent of GDP in 2018. Following is a brief description of the components of social spending and to the programs included in this report’s incidence analysis. Pension system As previously mentioned, the DR’s pay-as-you-go pension system was privatized in 2001. However, the Dominican government, through its Ministry of Finance, is still paying pensions from the Dominican Institute for Social Security (IDSS). 16 (World Bank, 2016, p. 15). 17 Decree 448-97 establishes that annual budget lines will be assigned in the General Budget of the Nation to cover the difference between their income and expenses. 18 The information about the number of free and paid travels came from OMSA transparency portal. 25 Conditional and Unconditional Direct Cash Transfer Programs The main social safety net program is PROSOLI (Progresando con Solidaridad). PROSOLI's provides conditional cash transfers for food (Comer es Primero), incentives for school attendance (known as ILAE for its initials in Spanish), bonuses for progress in studying (known as BEEP for Bono Escolar Estudiando Progreso), incentives for Higher Education (Incentivo a la Educación Superior, IES), unconditional cash transfers to elders (known as PROVEE for Protección a la Vejez en Extrema Pobreza), and unconditional cash transfers to households so they can afford liquefied petroleum gas (known as Bonogas Hogar) and electrical energy consumption (known as Bonoluz Hogar).19 For PROSOLI's programs, the government applies a Life Quality Index (ICV for its initials in Spanish) that categorizes households into four groups: Extreme Poor (ICV-I), Moderate Poor (ICV-II), Non-Poor I (ICV-III) and Non-Poor II (ICV-IV). Finally, there is an important group of unconditional cash transfer programs, including a gas-consumption transfer for working drivers (called Bonogas Chofer), an incentive program for preventive police (known as PIPP for its initials in Spanish), and an enlisted navy incentive program (known as PIAMG for its initials in Spanish). The main transfer programs are described briefly below. Comer es Primero. This conditional cash transfer (CCT) program grants monthly financial aid of RD$825 (close to US$16.7 in 2018) to heads of households living in poverty (ICVs I and II) to purchase food. The target population consists of households in a situation of vulnerability and households with pregnant women or children under 5 years old. The future mothers must attend clinical check-ups to monitor their pregnancies and the household's children under 5 must attend health controls. ILAE. This program aims to promote school attendance at the elementary (primary) school level. The CCT program grants RD$150 (US$3) per month per student in elementary school for up to four children between 6 and 16 years old in households living in poverty (ICVs I and II). Conditionalities are school enrollment and an attendance record of at least 80 percent. BEEP. This program offers a conditional cash transfer of RD$500 (US$10) each month (paid bimonthly) for each young person (up to 21 years of age) that is enrolled and attending the first or second year of high school (secondary education), RD$750 (US$15) for those who are enrolled and attending the third and fourth years of high school, and RD$1,000 (US$20) for those who are in these final courses in the technical-professional track. The transfer is designed so that students’ families can purchase the equivalent of a basic food basket, thereby helping to reduce school dropout rates in poor households (ICVs I and II). Conditionalities are school enrollment and an attendance record of at least 80 percent.20 PROVEE. PROVEE is a non-contributory pension for poor families benefiting from Comer es Primero, belonging to ICVs I and II, with elderly people who do not receive a pension or are not working, so they can receive cash transfers without having to meet any specific conditions. The program is administered by the National Council of the Aged Person (CONAPE) and is executed 19(Carrasco, Garcia, Parodi, & Vasquez, 2016, pp. 16-17). 20The government is increasing the BEEP transfers to RD$1,200 from 1st to 4th grade and to RD$1,400 in technical- professional mode. Now, six grades in secondary school are covered, instead of four in year 2018. 26 by the Administrator of Social Subsidies (ADESS). Its objective is to complement the nutritional support of elderly people living in poverty in order to help them maintain good nutrition. Households receive monthly aid of RD$400 (US$8) for each elder member. Incentive for Higher Education (Incentivo a la Educación Superior, IES). The Ministry of Education is responsible for this program, which targets poor students enrolled in the Autonomous University of Santo Domingo. The transfer amounts to RD$500 (US$10) monthly to cover tuition fees, the purchase of books, and other materials. The program’s conditionalities are to ensure continuation to undergraduate studies.21 Bonogas Hogar. The aim is to protect poorer families from the high cost of fuel for cooking. The transfer is carried out via a magnetic card (Solidarity Card) and is equal to 6 gallons of gas of 25 pounds (5.5 kilos) per month. The average transfer was RD$228 (US$5) per month in 2019. Target population of this program includes ICVs I, II and III.22 Bonogas Chofer. The gas consumption subsidy for working drivers consists of a monthly transfer of RD$3,420 (US$69) to buy LPG paid to public bus and car drivers. The conditionality is to not increase the price of tickets or fare. Bonoluz. Households receiving the subsidy Bonogas and identified by the SIUBEN selection system in ICVs I, II, and III. The transfer amount is variable, between RD$4.44 (US$0.09) to RD$444 (US$9) per month and corresponds to the value of the energy that the beneficiary household consumes subject to a maximum of 100 kilowatts. Other programs. The Incentive Program for Preventive Police (known as PIPP for its initials in Spanish) is under the direct responsibility of the Dominican National Police and targets policemen offering voluntary street surveillance. The Enlisted Navy Incentive Program (PIAMG) is managed by the Navy. Both programs consist of a monthly transfer valid for purchasing food, which was RD$ 928 (US$19).23 Near cash transfers School food program (Desayuno Escolar). The purpose of the school food program is to guarantee food and nutritional care to young students in the DR’s educational system, thereby reducing absenteeism and dropout rates. The program covers part of students' daily nutritional needs and has four different modalities. First, the Extended School Day Modality includes three food deliveries: breakfast, lunch, and a snack (juice or dairy). Second, the Border school food program Modality (programa de alimentación escolar -PAE- fronterizo) is implemented in the communities of the southwest and northwest—covering approximately 3.2 percent of all the program’s beneficiary centers—and its implementation involves the parents of the student beneficiaries. Third, the PAE REAL Modality (School Rations with Local Foods) is implemented in educational centers in remote areas, as well as in places of difficult access, and benefits 21 Lavigne & Vargas, 2013, pp. 21. 22 (Carrasco, Garcia, Parodi, & Vasquez, 2016, pp. 18). 23 (Lavigne & Vargas, 2013, pp. 21). 27 preschool and primary school students. Working with the school, students' families are involved in the preparation, cooking, and distribution of the local food provided, respecting the set menu for each day of the week. Finally, the Urban PAE Modality is implemented in the cities' educational centers, covering about 92 percent of existing public educational centers.24 The school uniforms and supplies program. This program delivers the necessary materials and uniforms for students attending public schools, including a set of uniforms, a pair of shoes, socks, a box of colored pencils, charcoal pencils, coloring books, boxes of putty, an apron, a pencil sharpener, tempera boxes, safety scissors, erasers, a backpack, a geometric set, and a periodic table, according to the National Institute of Student Welfare (INABIE). G. Education Education expenditures amount to 3.7 percent of GDP and represent 49.5 percent of the central government’s public social spending25. The education system includes pre-primary, primary, secondary, and higher education levels. The pre-primary level is for children under six. It is not compulsory, and the Dominican government is responsible only for the last year of this initial cycle. Next, the primary level is compulsory for children between 6 and 14 years old and consists of two cycles: the first covers grades 1 to 4 and the second grades 5 to 8. The secondary level offers four grades for adolescents between the ages of 14 and 18 years.26 According to the General Education Law, the secondary level is free but not compulsory. It has two cycles: the first offers general and compulsory education needed to graduate, while the second offers the choice of general, vocational/technical or arts. The following level is higher education. There are also education subsystems, such as adult education (literacy, primary and secondary) and special needs education for children with physical disabilities or other special needs. Finally, there is professional training provided by National Institute of Professional and Technical Training INFOTEP27. Pre-primary through secondary public education is administered by the Ministry of Education. In 2017-2018, most primary and secondary level students were enrolled in public schools—75 percent and 80 percent of the total, respectively, according to the National Statistics Office (ONE). Enrollment in private schools represented 23 percent and 17 percent of students in primary and secondary levels, respectively; the remaining students attend semi-official schools. Higher Education is administered by the Ministry of Education, Science and Technology (MESCyT). There are three types of higher education institutions: universities, specialized institutes of higher education and technical institutes of higher education. In 2018, most of the students of higher education were enrolled in universities (97 percent, according to information 24 (CELAC, n.d.). 25 The figure from Ministry of Finance is 3.9% of GDP, and the difference correspond to School Food Program and the school uniforms and supplies programs that have been classified as direct transfers in this report. 26 The last year with a quarter average as the last grade of that level was the 2018-2019 school year. Starting in 2019- 2020, the structure of pre-university education changed to: Initial level 0-5 years. Primary: 6 to 12 years old, Secondary: 13-18 years old. 27 (OECD, 2008, pp. 29-36). 28 provided by ONE). Of these, 38 percent were enrolled in the public Autonomous University of Santo Domingo (UASD). The National Literacy Plan (Quisqueya aprende contigo). The goal of the National Literacy Plan is to overcome illiteracy in young people and adults (15 years and older) throughout the country. It is a component of the Quisqueya Sin Miseria strategy to combat poverty. This program has three basic related components: in addition to Quisqueya Aprende Contigo, it includes the National Plan of Comprehensive Care for Early Childhood (Quisqueya Empieza Contigo) and the Local Integrated Development Plan (Quisqueya Somos Todos). H. Health The central government's health expenditure was equivalent to 1.6 percent of GDP in 2018. The healthcare system has been evolving since the approval of Law 87-01 and Health Law 42-01 which introduced an “entirely new institutional framework to exercise the functions of stewardship, financing, insurance, and provision of services�28. Family Health Insurance (Seguro Familiar de Salud) is the main scheme providing health protection in the Dominican Republic; its mandate is to offer protection to the entire population. The Health Risk Administrators (Administradoras de Riesgos de Salud, ARS) are divided into two regimes for affiliates, like the pension system.29 Under the contributory regime, affiliated formal workers contribute 10 percent of their of wages (3 percent funded by the employee and 7 percent funded by the employer). The fully subsidized regime is funded by the state.30 As of 2018, 7.7 million people were affiliated to the Family Health Insurance (close to 72 percent of population), of which 3.6 million participated in the subsidized regime and 4.2 in the contributory regime. The uninsured population includes some with contributory capacity (small business and independent professionals), low-income, informal- sector workers and undocumented persons.31 The creation of the National Health Service (Servicio Nacional de Salud, SNS), through Law 123- 15 was an important milestone in the public provision of health services. This entity coordinates the Regional Health Services, which provide a strengthened first level of care with geographic affiliation as a gateway—or first point of entry—for accessing health services in the public network. “This means that the functions of stewardship and provision of services have been legally separated and are now under the Ministry of Health and National Health Services�.32 The Essential Medicines Program/Central Logistics Support (PROMESE/CAL, for its initials in Spanish) is an institution dedicated to guaranteeing the operation and consolidation of a supply system of medicines, products, sanitary supplies and laboratory reagents. It acts as the only supply center to fulfill the demands of the National Public Health System. One of its programs 28 (Rathe, 2018, pp. 15). 29 The Health Risk Administrators (ARS) are public or private institutions that ensure health protection to their affiliates through the services of the Health Services Providers (Prestadoras de Servicios de Salud, PSS). PSS are public, private or semi-public entities or physical persons who provide medical attention and rehabilitation services (Lavigne & Vargas, 2013, p. 16). 30 (Lavigne & Vargas, 2013, pp. 16). 31 (Rathe, 2018, pp. 19). 32 (Rathe, 2018, pp. 24). 29 involves dispensing subsidized medicines to citizens through the network of 550 drugstores Farmacias del Pueblo.33 III. Methodology, data and assumptions A. Methodology This report estimates the impact of fiscal policy on micro-level welfare indicators by allocating fiscal policy elements, programs, expenditures, or revenue collections to individuals and households appearing in the 2018 ENGIH. The methodology framework for allocation and post- allocation analysis follows the methodology developed by the Commitment to Equity (CEQ) Institute to assess fiscal policy.34 The analysis creates measures of income––or “Income Concepts� –– that exclude (“pre-fiscal�) and include (“post-fiscal�) policy elements to examine the amount of redistribution accomplished (among others) and, therefore, the impact of the fiscal system on poverty and inequality. Figure 8 summarizes the construction of these income concepts. Market income, or pre-fiscal income, is constructed by totaling all private sources of an individual’s income: monetary labor income, non-monetary labor income, private transfers receipts, capital income, imputed value of own housing and self-consumption.35 This income aggregate has the same components as National Continuous Labor Force Survey (ENCFT) used by the Central Bank, ONE, MEPYD and World Bank for poverty estimation 36. The income concepts appearing in the blue boxes below market income in Figure 8 include different elements of publicly provided transfers (or publicly mandated revenue collections). Consequently, each is a post-fiscal income concept. The measures of poverty and inequality were calculated after the allocations were made. CEQ methodology considers two scenarios: pensions as deferred income and pensions as government transfers (Lustig, 2018). The Dominican Republic capitalization pension system has not started paying contributory pensions and, therefore, this system was not included in the analysis. As of 2018, public pensions schemes were not collecting contributions and were only paying contributory pensions funded by the government. So, only the pensions as government transfers were included in the analysis. Box 1: CEQ Terminology and Application in the Dominican Republic Taxes and transfers, and fiscal policy more generally, are powerful instruments at the disposal of the state for reducing extreme forms of material deprivation and narrowing the gap between economic elites and the rest. They can also help to equalize opportunities, through public education, for example, and thus increase social mobility and the productive potential of the underprivileged. To assess whether governments are using these tools effectively, it is important to be able to quantify how inequality and poverty change before and after the application of these fiscal policies. 33 (PROMESE/CAL, 2020). 34 (Lustig, 2018) 35 ENGIH survey question D402 specifies that income is gross (before any discounts or taxes). 36 Because Central Bank estimated all the income concepts in monthly basis, to annualize we multiplied everything by 12. 30 To quantify the impact that fiscal policies have on income (or purchasing power, or welfare), the team first must estimate a (counterfactual) income state that would be experienced before the transfers, benefits, and burdens generated by the fiscal system are received or imposed. As a proxy for this state, the team defines pre-fiscal income Ih as the cumulative income received from wages and salaries (that is, from labor market transactions) plus the market value of auto-production and auto-consumption; from capital (including real estate); and from private transfers (such as remittances from family members working abroad). The h subscript indexes a set of households (but equally could index individuals). The team then assembles a set of taxes and transfers T i to examine for example, Ti in the Dominican Republic might include the personal income tax. For each household h, the team then uses the micro-data (for the Dominican Republic, the National Household Income and Expenditure Survey or ENGIH), to allocate shares (Sih) of each program i= 1,….I in Ti to each household h. With the estimated shares, the team generates post-fiscal income at the household level Yh such that: Yh = Ih - Σi Ti Sih. (1) Figure 8 provides a schematic of the equation above and contains only one pre-fiscal income concept (Market Income) and several post-fiscal income concepts (Disposable Income, Consumable Income, Final Income). To determine the impact of the fiscal system on either poverty or inequality, the team takes the difference between its preferred measures of poverty or inequality over the pre- and post-fiscal distributions. Naturally, the extent of the fiscal system under consideration limits the team’s choice of the post -fiscal income concept. The impact of a fiscal system that includes only two elements must be estimated over a post-fiscal income concept that includes only these two elements. To determine the impact of single tax or transfer (or a subset of taxes and transfers), the team takes the difference in inequality (or poverty) at the post-fiscal income concept excluding the item in question (but including everything else in the team’s fiscal system) and the post-fiscal income concept including the item in question (and including everything else in the team’s fisc al system). A single tax or transfer (or a fiscal system) is inequality reducing when the addition of the fiscal item in question to an income concept reduces measured inequality. A transfer is pro-poor if the population living below the poverty line receives more benefits in absolute terms than other groups, implying that the per capita government spending on the transfer tends to fall with pre-fiscal income. In a Lorenz curve figure, pro-poor transfers’ concentration curve would lie above and to the left of the 45-degree line. Pro-poor transfer 100 Cumulative share of income and transfers 80 60 40 20 0 0 10 20 30 40 50 60 70 80 90 100 Cumulative share of population Pre-fiscal income Pro-poor transfer 45 degree line A transfer (tax) is progressive if, when households are ranked by pre-fiscal income levels, the cumulative household shares of the transfer (tax) are greater (less) than the cumulative household shares of pre- fiscal income. In a Lorenz curve figure, a progressive transfer’s (tax’s) conc entration curve would lie above and to the left (below and to the right) of the Lorenz curve for pre-fiscal income. The team calls a transfer 31 (tax) progressive when the transfers received, measured as a share or fraction of pre-transfer income, decline (rise) with income. Progressive tax and progressive transfer 100 Cumulative share of income, tax, and transfers 80 60 40 20 0 0 10 20 30 40 50 60 70 80 90 100 Cumulative share of population Pre-fiscal income Progressive Transfer 45 degree line Progressive Tax A transfer (tax) is regressive if, when households are ranked by pre-fiscal income levels, the cumulative household shares of the transfer (tax) are less (greater) than the cumulative household shares of pre- fiscal income. In a Lorenz curve figure, a progressive transfer’s (tax’s) conc entration curve would lie above and to the left (below and to the right) of the Lorenz curve for pre-fiscal income. The team calls a transfer (tax) regressive when the transfers received, measured as a share or fraction of pre-transfer income, rise (decline) with income. Regressive tax and regressive transfer 100 Cumulative share of income, tax, and transfers 80 60 40 20 0 0 10 20 30 40 50 60 70 80 90 100 Cumulative share of population Pre-fiscal income Regressive Transfer 45 degree line Regressive Tax Because taxes always reduce purchasing power, the team refrains from labeling taxes “pro -poor,� although when taxes paid (measured as a share of pre-tax income) increase with income levels, they are by definition progressive. In everyday usage, for example, the team often calls a marginal income tax rate schedule that has increasing marginal rates by taxable income bracket a “progressive� income tax. Two indicators used to understand how a fiscal policy element is progressive or regressive are the concentration shares and the incidence of a fiscal policy. Concentration shares calculate the share of the value of fiscal policy captured by (or imposed on) a subset of the population such as the poorest 10 percent of individuals or the richest 10 percent of individuals. For example, if the richest 10 percent of Dominicans pay 75 percent of the total personal income taxes collected in a given year, then the richest decile’s concentration share of personal income taxes is 75 percent (and that in turn implies that the other 90 percent of Dominicans pay no more than 25 percent of total personal income revenues). The incidence of a fiscal 32 policy element calculates the value of a benefit captured (or a tax imposed) relative to the value of income before the benefit was received or before the tax was imposed. Figure 8. Definition of CEQ Income Concepts Market Income • Income from work (wages/salary + in-kind benefits) • Income from capital • Self-provision of goods/services • Private transfers/remittances/alimony • Private pensions • Imputed rent + Contributory Pensions Market Income plus Pensions Direct Taxes and Contributions - • Personal Income Tax Net Market Income Direct Transfers + • Conditional or unconditional cash transfers • Near-cash transfers Disposable Income + Indirect Subsidies - • Electricity Indirect Taxes • Value-added tax (ITBIS) • Water • Oil derivates • Public transport (Metro and OMSA) • Excise taxes: beverages, alcohol, communications, and tobacco • Consumable Income In-Kind Transfers + • Monetized value of education and health services Final Income 33 Source: Excerpted and adapted from Lustig (2018) B. Data Household surveys: National Expenditure and Income Household Survey (ENGIH) and Demographic and Health Survey (ENDESA). The ENGIH is a nationally representative survey conducted by the Central Bank of the Dominican Republic. The fundamental purposes of the survey are to update the: (i) basic basket of goods and services, as well as the weighting coefficients, to calculate the consumer price index (CPI) and the average cost of living; (ii) income levels of the population and caloric requirements for the poverty calculations; and (iii) structure of household consumption within the framework of the National Accounts.37 The ENGIH provides detailed information on income, expenditures and benefits provided by the various public programs analyzed in this report. The survey includes multiple modules covering health, education, economic and labor market activity, household consumption expenditure, agricultural production, and rent (or, for owner-occupied housing, imputed rent). The ENGIH also provides a roster by household of individual, demographic, and dwelling characteristics. All of these features make the survey highly convenient for conducting a comprehensive assessment of the impacts of fiscal policy on welfare. However, the Dominican Republic's official poverty estimates are produced based on the National Labor Force Survey (ENCFT for its Spanish acronym) and not on the ENGIH. Unfortunately, the ENCFT does not provide information about expenditures, which is required to allocate subsidies and indirect taxes; that information is only available in the ENGIH. The income aggregates from the ENGIH comprise all labor and non-labor sources of income, including private transfers and remittances. The income aggregates compiled by the Central Bank were further refined in coordination with the Economic and Social Analysis Unit (UAAES) at the Ministry of Economy, Planning and Development (MEPyD) to adopt the same methodological decisions applied in the construction of the income aggregate employed for the official estimation of poverty. The ENGIH sample includes 11,652 selected dwellings, of which 8,881 were effective respondents. So, the response rate was greater than 75 percent. Field work was carried out over a whole year (2018), allowing for the collection of the different frequencies of the spending and consumption habits of a seasonal nature. This study focuses solely on fiscal year 2018 because that is the latest year for which the ENGIH is available. The target population is composed of those households residing in private, primary dwellings throughout the country. The domain estimation considers three categories of representativeness: (i) the country’s four major geographical regions: Greater Santo Domingo, North or Cibao, East and South; (ii) the whole 37 Dominican Republic Central Bank (2020, pp. 17-18). 34 country, urban and rural areas; (iii) Greater Santo Domingo, the remaining urban areas, and the remaining rural areas.38 The 2013 Demographic and Health Survey (ENDESA) is used to impute a propensity to visit public health care providers as well as to calculate shares of all recorded visits by household type. The 2013 ENDESA includes information on healthcare use––over a 12 month recall period and not conditional upon illness or injury––in a nationally representative survey sample of women and men of reproductive age. Propensity scores and shares of total clinic and hospital-level visits are estimated within the ENDESA to generate a household’s expected public healthcare benefit received within the ENGIH 2018. C. Administrative Data Additional information was required to complement the household survey information. Some of the sources applied in this analysis are: (i) government finance statistics and budget reports from the Ministry of Finance; (ii) cash transfer beneficiaries and amounts disbursed from ADESS; (iii) tax laws, explanatory documents on the tax system, tax collection and effective rates from General Internal Taxes Directorate (DGII); (iii) information on energy costs and tariffs (CDEEE and SIE); (iv) estimations of the cost of education by level from the Ministry of Education; (v) health information from Ministry of Health and National Health System; (vi) information on affiliation to Health Family Insurance from Bureau of Occupational Health and Safety (SISALRIL); (vii) information about affiliation to pensions systems from SIPEN; (viii) the Central Bank’s 2013 input-output matrix; (ix) ICV estimations made by World Bank; (x) statistics on population, education, health, poverty, and labor indicators from ONE; (xi) information on water subsidies (CAASD and INAPA);and (xii) information on transport subsidies (OMSA and OPRET). D. Allocation Overview Where possible, the report allocates fiscal policy elements to individuals or households based on direct observation or direct identification from ENGIH responses. For example, when an individual queried in a socioeconomic survey is asked to recall how much she has received from a cash transfer program like Bonogas Hogar, the team directly “observes� the benefits received from this program. These Bonogas Hogar receipts recorded by households are then assumed to be the same Bonogas Hogar listed in the executive, administrative, and other budget reporting for the same year. In the ENGIH, however, few fiscal policy elements could be allocated via direct observation. Instead, the analysis uses imputation and simulation (sometimes in combination with direct observation). Imputation is used when a survey unit’s benefit recipient (taxpayer) status must be inferred (rather than directly identified), or the amount received (paid) is retrieved from administrative records or program (tax) rules (rather than directly recorded in the survey), or both. Simulation is available when neither direct identification nor imputation can be used, so that the beneficiaries (taxpayers) and the amount received (paid) are simulated based on the program 38 (Dominican Republic Central Bank, 2020, pp. 17-18). 35 (tax). The subheadings below provide a summary of the allocation assumptions and decisions for various fiscal policy elements contained in this analysis. More details about the report’s methodological approach are provided in Annex II. Pension Contributions The analysis assumes that contributions to old-age pensions are practically nil, and they are not estimated. In the case of perceived old-age pensions from the pay-as-you-go system, it assumes that all but the highest 5 percent of the distribution of pensions correspond to this scheme. Personal Income Taxes Direct taxes were estimated using statutory rates applicable in year 2018. For the salaried regime, progressive income brackets and the applicable rate (10 percent) for interest and dividend income are applied. To consider tax evasion the analysis assumes that employees from government, private companies and free zones are formal, and self-employed workers, domestic labor and employer were informal. Direct Transfers For most direct transfers, the figures reported in surveys for the number of beneficiaries were applied subject to some corrections when significant deviations from administrative figures were found for beneficiaries from ADESS. To estimate benefits, an annual average per beneficiary was applied by dividing the total budget by the number of beneficiaries, taking information from administrative records. Considering that the number of beneficiaries could be under- or overrepresented, the characteristics of beneficiaries according to program rules (i.e., the age of children) were compared. If differences were not completely corrected, beneficiaries were randomly assigned or excluded using the ICV categories to approximate coverage from administrative records. Indirect Taxes For indirect taxes (ITBIS and excise taxes), the analysis assumes that consumers pay exclusively the burden of indirect taxes and a combination of the methodology used in (Aristy-Escuder, Cabrera, Moreno-Dodson, & Sanchez-Martin, 2018) and (Inchauste, Goraus, & Carrasco, 2019) was applied. The embedded rate is then estimated using an input-output matrix and information from the places of purchase and the goods and services with a high propensity for tax evasion. For goods with products on which the ITBIS was paid as a condition of the place of purchase the embedded rate was applied and exempt goods paid embedded rate.39,40 For excise taxes, the effective rate identified in the consumption in household surveys provided by DGII was applied.41 For oil derivative taxes, the direct and indirect effect was estimated using an input-output matrix, based on cost-push model.42 Box 2. ITBIS estimation. 39 See Annex 1. 40 A do file from Inchauste, et al. (2019) was adapted for this estimation. 41 Insurance tax, bank debit tax andlLuxury goods excise taxes were not included (see Tax Code: article 383, article 382 and article 375, paragraph 1). 42 (Coady, 2008) 36 The estimation of the ITBIS in this report follows a methodology based on Aristy-Escuder, et al. (2018) and Inchauste, et al. (2019). Value-added tax (VAT) evasion is a problem in the Dominican Republic. According to DGII (2018), tax gap with respect to potential collection was around 31.7 percent and 44.5 percent between 2007-2017. Therefore, it was important to include an adjustment for evasion in the estimations. Assumptions of tax evasion were made for the products based on the place of purchase and the urban or rural location of the household. Goods and services were clustered in the following two groups, and two categories for the second group: 1. Exempted 2. Non-exempted goods and services (General rate at 18 percent and reduced rate at 16 percent) 2.a Purchased in formal places. 2.b Purchased in informal points of sale. For group 2, the place of purchase is considered as informal or formal. Aristy-Escuder, et al. (2018) assume as exempted (i.e., not paying any ITBIS at all) the goods and services purchased in informal points of sale. In this analysis, Inchauste’s methodology is applied, which considers that the effective rate for exempted and informal place of purchases is the embedded rate, not zero as was considered in Aristy-Escuder, et al. (2018). The embedded rate includes the equivalent cost of ITBIS paid by sellers of exempted or ‘informal’ sales for the necessary inputs to produce them. The effective rate for non-exempted goods purchased in formal place of purchase is the statutory rate. To make these adjustments, two auxiliary files were created. The first includes each of the goods contained in the ENGIH 2018 that were classified in one of the two groups and two categories described above. The second defines whether the place of purchase is informal or not. Statutory rates were applied for goods purchased in formal places that belong to group 2.a (18 percent and a reduced rate of 16 percent), and the estimated embedded rate for exempted goods and services (group 1) and for non-exempted goods purchased in informal places (group 2.b). Water, transport, and energy subsidies Using the same methodology as (Aristy-Escuder, Cabrera, Moreno-Dodson, & Sanchez-Martin, 2018), implicit electricity transfers were calculated by applying existing tariffs. We estimated Implicit kWh consumed by each household were estimated and the subsidy was applied to users consuming less than 700 kWh a month. For those in the ENGIH survey who consume electricity but declare that they do not pay electricity bills, a standard implicit subsidy is calculated. In terms of transport, subsidies for metro and cable car services were estimated using information from OPRET. All transfers from the central government were assumed to be equal to the amount subsidized, and then the average subsidy per trip during 2018 was estimated. After that, the ratio 37 between the subsidies and the tariff paid was estimated to arrive at the subsidy by multiplying this ratio (1.82) by the amount paid, as declared in the survey. For OMSA bus subsidies, estimates were made for those who reported paying to use the bus and those who traveled free. For the first group, the cost per trip was estimated, and then the tariff paid was subtracted to get an estimated subsidy per trip and then this result was multiplied by the number of daily trips for the household. For the second group (not paying), the estimated cost per trip was imputed by the number of daily trips. For water, the methodology used by World Bank (2021) was applied. The cost per cubic meter was estimated and the price paid according to tariff schedules was subtracted from this number using data from the public institutions in charge of providing drinking water (INAPA and CAASD). For those consumers not paying for water service, the median cubic meter consumption for non- payers was imputed to arrive at estimates. In-kind transfers (education and health) Receipt of in-kind benefits is based on directly identified utilization of the public education or public health care system.43 The ENGIH records how many household members are enrolled in the public education system (and at what levels) and whether any household members recently visited a public health care facility. The monetized value of the in-kind transfer is based on the “government cost� approach. For example, total education expenditures are divided by the total number of users (students) to get a uniform per-user cost of producing and delivering the service. This per-user cost then is defined as the value of the transfer received. This cost represents what the utilizing household would have to pay to acquire the service at the government’s cost. The education benefit is based on the total cost by level, estimated by UNESCO and the Ministry of Education, then divided by the enrolled population in public schools according to the survey. In-kind health transfers are accounted for by estimating only the impact of the subsidized social security regime which is free for the poor and vulnerable and not the contributory regime, which works as a private insurance. The ENGIH is used to determine whether individuals with health insurance belong to the social security's subsidized regime. For the uninsured, the analysis identifies only those who use the services of public hospitals or ambulatory centers. It is also possible to identify those who are insured by the Dominican Institute of Social Security (IDSS). Finally, public spending under the Essential Medicines Program (PROMESE/CAL) is also computed. For health, beneficiaries for IDSS and the subsidized regime of Family Health Insurance were identified according to affiliation response in the ENGIH survey. For beneficiaries of the subsidized regime, the average transfer by insured (per capita) from the government to SENASA was imputed. For IDSS affiliates, an average benefit was estimated by dividing the government transfer by the total number of insured. For National Health System, ENDESA 2013 was used to identify each ENGIH household’s propensity to visit a public health care provider. Using ENDESA 43We used education per capita benefits estimated by Ministry of Education ( formulario Unesco), which uses national averages, and it was not adjusted for quality in any way. We did not apply scaling-up or scaling-down for health and education benefits. 38 2013 share of public health facilities, households were selected based on those with a higher propensity to match the same share of households using public facilities in ENGIH household. For those identified as users of public facilities, an average cost per user for National Health System facilities which was estimated by dividing the total expenditure (transfer to National Health System, Ministry of Health collective services and high-cost illness) by users of health public services in the ENGIH survey, identified using matching-score analysis from ENDESA 2013. Once the beneficiaries of PROMESE were selected using propensity score, an average benefit was estimated by dividing the program’s expenditures in 2018 by the same share of population of users reported in ENDESA 2013. As with education, the ratio of health expenditure to disposable income under the survey is adjusted to match the ratio calculated using national accounts. IV. Main results A. The fiscal system reduces inequality, but the impact could be improved The fiscal policy implemented in the Dominican Republic in 2018 reduced income inequality. Figure 9 shows that fiscal policy in 2018 reduced the market income Gini coefficient from 0.459 to 0.374, a decline of 8.6 Gini points. This means that taxes, indirect subsidies, and transfers (including the monetized value of education and health) tend to benefit the poor and vulnerable populations more in relative terms. Excluding education and health services, a reduction in inequality is still evident as the Gini coefficient is reduced by 2.8 points. Figure 9. Inequality (Gini coefficient) at pre- and post-fiscal income concepts 0.70 0.65 0.60 Gini coefficient 0.55 0.50 0.459 0.453 0.439 0.431 0.45 0.40 0.374 0.35 0.30 0.25 Market Net Market Disposable Consumable Final Income Income Income Income Income Source: estimates based on ENGIH 2018 and official sources 39 Fiscal policy benefits poorer populations in both urban and rural areas by reducing income inequality. Figure 10 shows that taxes, indirect subsidies, and other transfers (including the monetized value of education and health services) reduces inequality among the rural population, as seen in the 9-point decline in the Gini coefficient between market income inequality (with a Gini coefficient of 0.40) and final income (0.31). At the same time, fiscal policy also reduces inequality for the urban population (with the Gini coefficient moving from 0.46 to 0.38, i.e., 8 Gini points). Figure 10. Inequality (Gini coefficient) at pre- and post-fiscal income concepts for national, urban and rural populations 0.50 0.46 0.44 0.45 0.46 0.43 0.44 0.44 0.40 0.37 0.40 0.38 0.38 0.35 0.37 0.30 0.31 0.25 Market Income Disposable Income Consumable Income Final Income Urban Rural National Source: estimates based on ENGIH 2018 and official sources The program Desayuno Escolar and the personal income tax are the government’s two most inequality-reducing interventions. The excise tax on oil products and the ITBIS also have significant impacts on consumable income. Indirect subsidies (on metro and water services), which benefit non-poor populations do not decrease inequality.44 When final income is the end-income concept, primary and secondary education transfers are the most equalizing public interventions. Health transfers (through the National Health System, the Ministry of Health Services, and the subsidized family health insurance) also has a significant equalizing effect. Indirect subsidies in general are not equalizing. When compared to other countries in LAC, the Dominican Republic stands in the middle of the group in terms of the redistributive impact of its fiscal policy (including the monetized value of education and health services).45 Argentina, Brazil, and Uruguay have interventions that reduce 44 The metro subsidy may have a contribution that is not captured by CEQ analysis, associated with its reduction of both traffic congestion and environmental pollution. 45 The results must be compared cautiously, because some cases include a scaling-down effect. 40 inequality by more than 15 Gini points. The Dominican Republic’s fiscal policy has a lower redistributive impact, relatively speaking, because its 8.6-Gini-point decline is subtracted from an initial Gini coefficient that is lower, compared with a regional average of 10.4 Gini point decline. Many countries in LAC have a Gini for Market Income similar to the Dominican Republic. Some like Ecuador reduce inequality up to 8 Gini points when education and health insurance are included. Others like Bolivia and Peru reduce inequality to a lesser degree. When the effects of health and education are not considered, the change from market to consumable income is still in the middle of the region with a Gini coefficient falling by 2.8 points, slightly lower than regional average (3.3 Gini point reduction). The corresponding reduction in inequality for Argentina, Brazil, and Uruguay is more than 5 Gini points. Figure 11. Selected countries: Fiscal policy inequality reduction From market income to final income (circa 2015) Source: CEQ Data Center (https://commitmentoequity.org/datacenter/) 41 B. The impact of fiscal policy on poverty is modest Table 3 shows that fiscal policy reduces slightly poverty incidence (1.0 percentage point, using the National Poverty Line).46 As expected, applying direct and indirect taxes increases poverty, and applying direct transfers and indirect subsidies reduces it. Table 3. Changes in poverty National National Extreme USD 5.5 USD 6.85 Poverty Line Poverty Line PPP 2011 PPP 2017 Market Income 22.1% 4.3% 16.6% 24.0% Net Market Income 22.1% 4.3% 16.6% 24,0% = Market (-) Direct Taxes Gross Income 19.0% 2.6% 13.6% 21,2% = Market + Direct Transfers + Pensions Disposable Income = Market + Direct Transfers + Pensions 19.0% 2.6% 13.6% 21.2% – Direct Taxes Consumable Income =Disposable + 21.1% 2.9% 15.2% 23.1% Subsidies– Indirect Taxes Source: estimates based on ENGIH 2018 and official sources Note: In CEQ methodology, poverty is not estimated for Final Income Table 4 reflects how the Dominican Republic's income distribution changes with fiscal policy applied in 2018, using US$ PPP. The share of the measured population living below the poverty line (at US$5.5 2011 PPP per day) is 16.6 percent and this group accounts for only 3.7 percent of market income. Fiscal policy reduced poverty by around one percentage point: after taxes and transfers, the share of the population that is considered poor is reduced from 16.6 percent to 15.2 percent. In addition, the income share of those below the poverty line (US$5.5 PPP per day) increases from 3.7 percent to 3.8 percent after fiscal policy is applied. The share of the population deemed vulnerable to poverty comprises around a third (33.2 percent) of the total before fiscal policy is applied. After taxes and transfers, the vulnerable group increases (36.6 percent) because some are elevated from poverty while others are from the middle class. In terms of population, the middle-class accounts for close to half (46.8 percent) and concentrates almost 60 percent of total market income. The results are similar when using the new international poverty line for upper middle-income countries, equivalent to US$6.85 in 2017 PPP per day. In this case, poverty is reduced by 0.9 percentage points, from 24 percent according to market income to 23.1 according to consumable income. The 2.9-point reduction in poverty due to direct transfers, i.e., from market income to 46 This result is consistent with Aristy-Escuder, et al. (2018). Even though the two CEQ analysis are not fully comparable (in the 2016 study the methodology applied the tax and public expenditure structure of 2013 to ENIGH 2007), we can note that fiscal policy in 2018 is more pro-poor. Fiscal Policy of 2013 reduced extreme poverty (calculated using the national extreme poverty line) from 13.8% to 13.1%, whereas poverty remains slightly higher (41.2% vs. 42.3%). Also, fiscal policy in 2013 reduced the market income Gini coefficient from 0.514 to 0.458, when direct and indirect taxes and government transfers (including the monetized value of education and health services) and subsidies are considered. 42 consumable income, is offset by the net effect of indirect taxes and subsidies, which increase poverty47. Table 4. Population and income shares of market income by income groups, US$ PPP 2011 Income share Population shares Market After taxes and Market After taxes and Income transfers Income transfers Share (Consumable (Consumable income) income) Extreme Poor (0-3.2) 0.7 0.5 5.1 3.6 Moderate Poor (3.2-5.5) 3.0 3.3 11.5 11.6 Vulnerable (5.5-11.5) 16.4 19.4 33.2 36.6 Middle Class (11.5- 61.9 45.7 60.2 46.8 57.6) Wealthy (57.6+) 19.8 14.8 3.5 2.4 Total 100.0 100.0 100.0 Source: authors' estimates based on ENGIH 2018 The first four deciles (the poorest 40 percent) have 14.0 percent of income, and the 9th and 10th deciles (richest 20 percent) possess half of the market income in the Dominican Republic. For each of the deciles 1 to 8, income share increases after fiscal policy. Meanwhile, the top earners (deciles 9 and 10) see their income share reduced after paying all their taxes (Table 5). Table 5. Market income: share and accumulated share by decile before and after taxes and transfers Income share Decile After taxes and transfers Market Income (Consumable income) 1 1.8 2.1 2 3.1 3.5 3 4.1 4.4 4 5.1 5.4 5 6.2 6.5 6 7.5 7.7 7 9.1 9.3 8 11.6 11.6 9 15.6 15.3 10 35.9 34.1 Total 100.0 100.0 Source: author’s estimates based on ENGIH 2018 47 See https://blogs.worldbank.org/opendata/updating-international-poverty-line-2017-ppps 43 Most Dominicans are overall net beneficiaries of their country's fiscal policy. Most of the population—i.e., first eight income deciles—are net beneficiaries of fiscal policy when all benefits and taxes are applied (Figure 12, right panel). The benefits are higher for the poor and the highest impact comes from educational and health in-kind transfers. The net benefits are not significant for the highest deciles (deciles 9-10), which comprise the middle class and wealthy income groups. The impact of direct transfers is only noticeable for the first five deciles, which comprise the poor and vulnerable income groups. There is a neutral impact on the middle-class, as taxes are compensated exactly by the benefits received. However, when the benefits from health and education are excluded, only the first three deciles of the population are net beneficiaries of the country’s fiscal system (Figure 12, left panel). Upper deciles become net payers. Figure 12. The Dominican Republic: Net payers/beneficiaries (% of market income) Net incidence, all taxes and transfers Except education and health benefits Net incidence, all taxes and transfers 140% 25% 120% 20% 100% 15% (% of market income) (% of market income) 80% 10% 60% 5% 40% 0% 20% -5% 0% -10% -20% -15% -40% 1 2 3 4 5 6 7 8 9 10 -20% 1 2 3 4 5 6 7 8 9 10 Decile Decile All Education All Health All Direct transfers All Indirect Taxes All Indirect Subsidies All Direct Taxes All Direct Taxes All Direct transfers All Indirect Subsidies All Indirect Taxes All net transfers and subsidies All net transfers and subsidies Source: authors' estimates based on ENGIH 2018 Direct transfers (Desayuno Escolar, Comer es Primero) and indirect subsidies (electricity) tend to reduce poverty the most (Figure 13). Indirect taxes (ITBIS and fuel excise taxes) increase poverty. Nevertheless, the combination of taxes and transfers has a net positive effect on welfare. This means that a simultaneous increase in taxes and social expenditure will further reduce poverty in the Dominican Republic. While the ITBIS and fuel excise taxes reduce inequality, they increase poverty slightly as the vulnerable and poor populations still pay ITBIS for the purchase of goods and services and consume fuel by purchasing in oil stations but mainly indirectly through the purchase of goods and services that use oil as an input. 44 Figure 13. Marginal contributions to poverty Increment Reduction All direct transfers incl contributory pensions 3.3% 3.5% NCT Desayuno Escolar (per capita) 1.4% 1.9% CCT comer es primero (per capita) 0.7% 0.9% All indirect subsidies 0.5% 0.5% All direct taxes 0.0% 0.0% Indirect Tax ISC Oil Taxes - Ind eff (per capita) -0.3% -0.5% Indirect Tax ISC Oil Taxes - Dir eff (per capita) -0.4% -0.6% Indirect Tax ITBIS (per capita) -1.2% -1.6% All taxes -2.6% -2.1% All indirect taxes -2.6% -2.1% -0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04 USD 5.5 PPP National Moderate Poverty Line Source: estimates based on ENGIH 2018 and official sources At 1.3 percent, the share of the Dominican population lifted from poverty (using the US$5.5 PPP threshold), as a result of fiscal policy interventions, is relatively small when compared with other Latin American countries' fiscal policy impacts. Argentina had the most pro-poor fiscal policy in the region over the last five years. The Dominican government can improve the capacity of its fiscal system to reduce poverty, mainly through an increase in public revenue (e.g., taxes), which could be transferred to the poor and vulnerable populations48. Figure 14. Poverty reduction (US$ 5.5 PPP line): from market to consumable income 48 The quality of public services has a significant impact on poverty and inequality reduction, but it is beyond the scope of this analysis. 45 6 3.4 3.6 4 1.7 2 0.2 0.2 0.3 0 -2 -0.6 -1.7 -1.3 -4 -3.2 -3.7 -6 -4.9 -8 -8.2 -10 -12 -11.8 -14 ARG VEN ECU PAN CHI COL DOM MEX HND DOM PER BOL GTM SLV 2017 2013 2011 2016 2013 2014 2018 2014 2011 2013 2011 2015 2014 2017 Source: CEQ Data Center and estimates based on ENGIH 2018 and official sources C. Taxes are progressive, some more than others The results of the analysis show that both direct and indirect taxes are progressive.49 The concentration curves for direct and indirect taxes lie below the Lorenz curve for market income (Figure 15). As expected, direct taxes are much more progressive than indirect taxes, meaning that the biggest proportion are paid by the rich (income above US$50 PPP a day). The upper class, which receives 19.8 percent of total market income, pays 64.6 percent of the total direct taxes and 24.3 percent of the indirect taxes. In terms of market income deciles, individuals in decile 10 receive 35.9 percent of total market income, account for 85 percent of the direct taxes and 42.7 percent of the indirect taxes paid. Figure 15. Direct and indirect taxes: Concentration curves and share of benefits in poor population, ranked by market income Concentration curves, ranked by market income Concentration shares by income group 49 A tax is everywhere progressive (regressive) if its concentration curve lies everywhere below (above) the market income Lorenz curve. 46 100 Direct taxes Indirect taxes 64.6 23.3 Cum proportion of fiscal intervention 80 % of total tax payments 61.4 60 40 35.2 20 12.8 0 0.0 2.4 0 10 20 30 40 50 60 70 80 90 100 Cum proportion of population Upper class (57.6+ US$ PPP) Lorenz curve (Market Income) Middle Class (11.5-57.6 US$ PPP) Direct Taxes Vulnerable (5.5-11.5 US$ PPP) Indirect Taxes Poor (0-3.2 US$ PPP) Source: authors' estimates based on ENGIH 2018 Figure 16. Incidence of taxes (% of market income) 12.0 12.0 10.7 (% reduction of market income) (% reduction of market income) 10.0 10.0 8.8 8.0 8.0 6.5 5.1 5.4 6.0 4.5 6.0 4.0 4.0 2.0 0.8 0.0 0.0 0.0 2.0 0.0 Extreme Poor Vulnerable Middle Wealthy 0.0 Poor (0-3.2 (3.2-5.5 poor (5.5- Class (10- (50+ US$ 1 2 3 4 5 6 7 8 9 10 US$ PPP) US$ PPP) 10 US$ 50 US$ PPP) Decile PPP) PPP) Direct Indirect Direct Taxes Indirect Taxes Taxes Taxes Source: estimates based on ENGIH 2018 and official sources 47 Indirect taxes (ITBIS and excise taxes) reduce the population’s market income by 9.3%, on average. Indirect taxes affect poor people but are still progressive in relative terms, that is, the amount of indirect taxes collected expressed as a share of market income is lower for the poor than for other groups of the population. The market income of the upper class is reduced by 10.7%, while the income of the extreme poor (below the US$3.2 PPP a day threshold) drops 5.1% after paying indirect taxes. Because no value-added taxes are paid for goods in an exempted basic consumption basket (ITBIS exemptions) and a significant share of purchases are made in informal stores,50 the progressivity of indirect taxes is increased, since that basket represents a higher proportion of the poor's consumption of the poor. i. Direct taxes are highly progressive Direct taxes are highly progressive, but they only represent 1 percent of market income for the average taxpayer. Direct taxes fall mainly on the upper and middle classes.51 Direct taxes reduce the market income of the wealthiest by 4.5 percent, while they reduce the average market income of the total population by 1.4 percent. The top decile of the population pays 85 percent of the direct taxes received by the government. For the top decile, total direct taxes decrease their market income by 3.3 percent, while they reduce the market income of the seventh decile by only 0.1 percent—and they do not affect the income of the first four deciles at all. In terms of socioeconomic groups, middle-class households (per capita income of between US$10 and US$50 a day) pay 35.2 percent of total direct taxes, while those above US$50 a day pay 64.6 percent. Personal income taxes, which account for 98.7 percent of the direct taxes included in the analysis, are highly progressive. These taxes reduce the market income of the tenth decile by 3.2 percent and 0.9 percent for the ninth decile. In terms of socioeconomic groups, personal income taxes reduce the average market income of the middle class by 1.4 percent and the wealthiest segment of the population by 4.4 percent. The middle class accounts for 35.5 percent of total personal income tax payments, and the highest income group captured in the survey represents 64.4 percente. This distribution is explained by a high-exemption threshold for the personal income tax and a high level of labor informality—the informal sector accounts for 57 percent of income earners. The dividend and interest tax represent 1.3 percent of total direct revenues for the government included in this analysis. These taxes reduce the market income of the population by just 0.02 percent. The wealthiest segment of the population pays 63.4 percent of the dividend tax and 93.2 percenteof the interest tax. The middle-class accounts for 35.1 percent and 6.2 percent of the dividend and interest tax, respectively. The market income of the wealthiest segment is reduced by 0.08 percent by those taxes. 51One caveat of the analysis of direct taxes is that collection from the survey does not capture all sources of income because like other household surveys, the ENGIH likely does not capture the top of the distribution or top income earners. So, the income captured by the survey is lower compared with national accounts, as was noticed by (DGII, 2018, p. 38), so estimated personal income tax is around 43 percent of personal income tax collection. 48 Table 6. Incidence of Direct Taxes as % of Market Income Interest + Dividend Wages 1 0.00 -0.01 2 0.00 0.00 3 0.00 0.00 4 0.00 -0.01 5 0.00 -0.03 6 0.00 -0.06 7 0.00 -0.12 8 0.00 -0.41 9 0.00 -0.89 10 -0.05 -3.21 Total -0.02 -1.36 Source: author’s estimates based on ENGIH 2018 ii. Indirect taxes are progressive Indirect taxes are the main source of tax income in the Dominican Republic. They represent 68% of total taxes. The indirect taxes included in the analysis are: ITBIS; excise taxes on alcoholic beverages, cigarettes, fuel products and telecommunication services. Indirect taxes (ITBIS and excise taxes) reduce the population’s market income by 9.3 percent, on average. Indirect taxes affect poor people, but are still progressive in relative terms. that is, the amount of indirect taxes collected expressed as a share of market income is lower for the poor than for other groups of the population. The market income of the upper class is reduced by 10.7 percent, while the income of the extreme poor (below the US$3.2 PPP a day threshold) drops 5.1 percent after paying indirect taxes. Because no value-added taxes are paid for goods in an exempted basic consumption basket (ITBIS exemptions) and a significant share of purchases are made in informal stores, the progressivity of indirect taxes is increased, since that basket represents a higher proportion of the poor's consumption of the poor .52 Indirect taxes are also progressive as they reduce the market income across all deciles, but the most significant reduction is on the top decile. These taxes reduce the market income of the first decile by 5.0 percent, compared to 10.3 percent for the top decile. In terms of socioeconomic groups, indirect taxes reduce the middle-class market income by 8.8 percent. In terms of concentration, the share of indirect tax payments for the first nine deciles (57.3 percent) is below their share of market income (64.1 percent). The middle class has a share in indirect taxes (61.3 percent) comparable to their share of the market income (60.2 percent). The upper 52 The Dominican economy has a high degree of informality. Around 55 percent of employment originates in informal activities, with a lower effective tax rate. If we assume that poor people buy a higher proportion of their consumption in the informal sector, then it also makes the indirect taxes more progressive. See Bachas, Gadenne, & Anders (2020) for an explanation of how consumption taxes are redistributive, lowering inequality. 49 class pays 24.3 percent of all indirect taxes, which is higher than their market income share (19.8 percent). Figure 17. Indirect taxes: concentration curves and share of benefits in poor population Concentration curves, ranked by market income Concentration shares by income group 100 Cum proportion of fiscal intervention 80 22.4 19.9 26.6 23.0 18.2 10.5 39.0 69.0 60 60.9 65.6 60.8 63.4 % of total tax payments 62.3 40 53.4 20 0 0 10 20 30 40 50 60 70 80 90 100 Cum proportion of population 15.7 18.1 13.9 14.2 9.6 12.1 Lorenz curve (Market Income) 6.5 2.9 3.5 1.4 1.4 2.0 2.4 1.1 ITBIS Oil Indirect Oil Direct Poor (0-3.2 US$ PPP) Vulnerable (5.5-11.5 US$ PPP) Alcoholic Bev Beer Middle Class (11.5-57.6 US$ PPP) Upper class (57.6+ US$ PPP) Tobacco Source: authors' estimates based on ENGIH 2018 In terms of tax revenue, the ITBIS is the most important indirect tax, representing 53.0 percent of total indirect taxes included in this analysis. It has an important set of exempted basic consumption goods and services, which makes it a slightly progressive tax, even though it is a value added tax. Up to the eighth decile there is a lower share in ITBIS payments than market income. In terms of socioeconomic groups, the middle class, and the wealthiest pay 83.2 percent of ITBIS while benefiting from 79.9 percent of market income. The ITBIS reduces market income by almost 4.9 percent, on average. The income of the top decile earners is reduced by 5.5 percent while the first decile sees market income reduced by 3.7 percent. The fuel tax is the second most important indirect tax, accounting for 37.4 percent of total indirect taxes analyzed. The direct effect of fuel tax (purchases in oil stations) is more progressive than the ITBIS. The top two deciles contribute a share of total direct fuel tax payments (64.4 percent) which is higher than their share of total market income (51.6 percent). The wealthiest group pays 26.6 percent of total fuel taxes, which reduces their market incomes by 3.5 percent including direct and indirect effects (versus -1.7 percent for the extreme poor). On the other hand, the indirect effects of the fuel tax (through the purchase of goods and services that use oil as an input) are neutral or proportional to market income. Table 7. Incidence of indirect taxes, income groups (as % of market income) 50 Fuel Alcoholic ITBIS Beer Tobacco Communications Indirect Direct Beverages Extreme Poor (0-3.2) -3.73 -0.96 -0.85 -0.13 -0.20 -0.02 -0.03 Poor (3.2-5.5) -3.93 -0.96 -1.03 -0.12 -0.22 -0.02 -0.06 Vulnerable (5.5-10) -4.11 -0.99 -1.27 -0.23 -0.42 -0.04 -0.07 Middle Class (10-50) -4.17 -0.99 -1.36 -0.26 -0.35 -0.02 -0.06 Wealthy (50+) -4.35 -1.02 -1.65 -0.16 -0.36 -0.03 -0.08 Source: own estimates based on ENGIH 2018 and official sources Indirect taxes on alcoholic beverages, beer and tobacco, account for 7.8 percent of indirect taxes. Among them, the excise taxes on beer (4.7 percentage points of indirect taxes) are the most important, followed by the excise taxes on other alcoholic beverages (3 percentage points). The beer tax reduces market income by 0.41 percent and alcoholic beverage tax reduces it by 0.28 percent. The middle class and the vulnerable population pay 63.4 percent and 12.1 percent, respectively, of beer tax. The highest incidence of the beer tax is on the middle class (0.45 percentage points), while the alcoholic tax impacts the wealthiest by reducing market income by 0.38 percent. This top decile pays 44 percent of the excise tax on alcoholic beverages and almost 36 percent of beer taxes. The excise tax on tobacco products is not progressive. After the fourth decile, the share of excise tax on tobacco products is higher than the share of the market income. To compare, this tax reduces the market income of the extreme poor by 0.03 percent and the income of the richest population by 0.01 percent.53 The excise tax on communications is highly progressive. The wealthiest pay 39.0 percent of the tax, which is significantly higher than its share of market income (19.8 percent). This tax reduces the income of the wealthiest by 0.33 percent and the income of the extreme poor by just 0.03 percent. Figure 18. Indirect taxes: concentration shares by decile of market income 53While tobacco taxes are regressive in the short run, there are international evidence that states that tobacco taxes have positive and pro-poor lifetime distributional effects (see Fuchs, 2019). 51 70.0 60.0 (% of total tax payments) 50.0 40.0 30.0 20.0 10.0 0.0 1 2 3 4 5 6 7 8 9 10 Decile ITBIS Oil Indirect Oil Direct Alcoholic Bev Beer Tobacco Communication Source: own estimates based on ENGIH 2018 and official sources The progressivity of the ITBIS is explained by the exemptions and by purchases in informal establishments. Table 8 shows the distribution of consumption according to the type of good (exempted or taxed) and the place of purchase (formal or informal). First, total consumption is more concentrated in the highest deciles, since 26.8 percent of consumption is carried out by households in the tenth decile. Second, consumption of exempted goods is over-represented in the total consumption of lower-income households. For instance, the exempted consumption of the first decile represented 4.2 percent of total exempted consumption while the share of the total expenditure of the first decile is 3.7 percent of total expenditure. Third, the purchases for lower-income households are more frequent in informal places compared with those of higher income deciles.54 Table 8. Dominican Republic: concentration of expenditures according to VAT exemption and place of purchase Exempted Non-exempt Total Expenditure Total Formal Informal Total Formal Informal 1 3.7% 4.1 2.7 6.7 3.1 2.1 4.5 2 4.7% 5.1 3.5 7.8 4.1 3.2 5.4 54 Bachas et al. (2020) states that “the informal sector thus makes consumption taxes progressive: households in the richest quintile face an effective tax rate that is twice that of the poorest quintile .� 52 3 5.6% 6.0 4.7 8.2 5.1 3.9 6.5 4 6.3% 6.7 5.4 8.9 5.8 4.6 7.3 5 7.7% 8.2 6.9 10.3 7.1 6.1 8.3 6 8.9% 9.2 8.3 10.7 8.4 7.7 9.3 7 9.6% 9.8 9.4 10.5 9.5 8.7 10.5 8 11.6% 11.6 11.7 11.3 11.6 11.1 12.3 9 15.1% 14.9 15.8 13.2 15.4 17.1 13.2 10 26.8% 24.6 31.6 12.4 29.8 35.4 22.6 100% 100 100.0 100.0 100 100 100 Source: own estimates based on ENGIH 2018 and official sources The redistributive effect of the tax system in the Dominican Republic is similar to the tax systems of other Latin American countries (Figure 19). The Kakwani index—which is equal to the difference between the concentration coefficients of a particular tax system and the Gini coefficient of the reference income—is positive for the Dominican Republic, which indicates that its tax system is more progressive. Nevertheless, the Dominican tax system's marginal contribution to changing the Gini coefficient is lower than in the other selected countries, probably due to the country's lower tax collection. For example, the Dominican Republic has a high exemption threshold for the income tax, which reduces the amount of direct taxes as percent of GDP. Even though indirect taxes are progressive, they are less progressive than direct taxes, and contribute to an increase in poverty. These indirect taxes reduce the population’s market income by an average of 9.3 percent, thus affecting the poor and vulnerable population by increasing poverty. This is especially relevant in the case of the ITBIS that is paid by all population strata. Nevertheless, indirect taxes are progressive because they affect the poor less than other segments of the population. Further, the Dominican Republic’s indirect taxes still have a relatively high Kakwani, which means that the system is more progressive than most of those of the other countries included in the sample. This is probably explained by the progressivity of excise taxes and the exemptions to the value added tax (ITBIS).55 Figure 19. Progressivity and redistributive effect of direct and indirect taxes in Latin America Direct Taxes Indirect Taxes 55As stated previously, the high degree of informality of the Dominican Republic can increase the progressivity of the indirect taxes. See Bachas et al. (2020). 53 0.50 0.018 0.10 0.010 0.45 0.016 MARGINAL CONTRIBUTION SCALE 0.05 MARGINAL CONTRIBUTION SCALE 0.008 0.40 0.014 0.00 0.35 KAKWANI SCALE 0.006 KAKWANI SCALE 0.012 0.30 -0.05 0.010 0.25 0.004 0.008 -0.10 0.20 0.002 0.006 -0.15 0.15 0.10 0.004 0.000 -0.20 0.05 0.002 -0.25 -0.002 0.00 0.000 ARG COL CHI GTM SLV DOM MEX DOM ARG COL CHL GTM SLV DOM MEX DOM 2017 2014 2013 2014 2017 2013 2014 2018 2017 2014 2013 2014 2017 2013 2014 2018 Marginal Contribution Kakwani Marginal Contribution Kakwani Source: CEQ Institute’s Standard Indicators, own estimates using ENGIH and official sources D. Social spending and subsidies Social spending and subsidies include direct transfers (cash and in-kind); indirect subsidies to public services; in-kind education and in-kind health transfers. According to Figure 20 (left), education is higher more valuable (relative to income) than the other benefits included, while subsidies are lower. Besides, the value with respect to market income (incidence) decreases moving up the deciles for direct transfers, education, and health, but incidence barely changes for indirect subsidies. The concentration curve for fiscal interventions, shows that direct transfers (including contributory pensions) and education are pro-poor, health is relatively progressive, and indirect subsidies are proportional to market income (Figure 20, right). The effects of each of these on inequality and poverty are described below. 54 Figure 20. Incidence and concentration curves of transfers and subsidies Concentration curve Incidence 100.0 80 Cum proportion of fiscal intervention 70 80.0 60 (% of market income) 60.0 50 40 40.0 30 20 20.0 10 0.0 0 0 10 20 30 40 50 60 70 80 90 100 1 2 3 4 5 6 7 8 9 10 Cum proportion of population Decile Lorenz curve (Market Income) Direct transfers Direct transfers Education Education Health Health Indirect subsidies Indirect subsidies Source: own estimates based on ENGIH 2018 and official sources i. Direct transfers are pro-poor Most direct transfers programs are pro-poor in the Dominican Republic (Figure 21, left). The only exceptions are the CCT for tertiary education (IES), contributory pensions, and the bonus for drivers (Bonogas Chofer) for which the benefits for lower income households are higher only as a share of their total income but not in absolute terms (i.e., they are progressive). The distributions of benefits from different programs vary (Figure 21, right). The elder-care program PROVEE (Protección a la Vejez en Pobreza Extrema), the CCT incentives to attend primary school (ILAE), and Desayuno Escolar are the most progressive programs—they are all pro-poor, and which have and have 30 percent of their expenditures going to the poor (US$5.5 PPP a day). Around 22 to 24 percent of the public expenditures under Bonogas Hogar, Bonoluz and Comer es Primero reach poor households (US$5.5 PPP a day). In contrast, just 12 percent or less of the total benefits from contributory pensions, IES and Bonogas Chofer go to poor households, which explains why those programs are deemed progressive and not pro-poor. Comer es Primero and the aggregate of the other direct transfers are pro-poor, since, apart from representing a larger share of market income for poor households than for non-poor ones, the total transferred amount in aggregate terms is also larger for the poor. The pro-poorness of some of these direct transfers could be linked to the fact that they are targeted via the Life Quality Index ICV.56 56Although the ICV is not a perfect match with monetary income, it has helped to improve transfers' progressivity (Carrasco, et.al. 2016)]. 55 Figure 21. Direct transfers: concentration curves and share of benefits in poor population Concentration curves, ranked by market income Concentration share by income group 100 Cum proportion of fiscal intervention 80 0.0 0.0 0.0 0.2 0.3 0.0 0.3 0.5 0.0 0.0 0.0 10.2 25.8 27.1 28.5 60 30.0 37.8 39.7 41.0 40.2 46.7 56.5 40 % of total benefits 39.6 73.0 39.5 42.7 41.9 20 39.6 39.5 38.5 32.7 55.0 0 35.0 0 10 20 30 40 50 60 70 80 90 100 Cum proportion of population 34.6 30.6 30.1 29.4 Lorenz Curve (Market Income) 22.3 20.6 20.5 20.0 16.8 Comer es primero 8.4 4.8 Bono gas Hogar Bono Luz Desayuno Escolar Poor (0-3.2 US$ PPP) Vulnerable (5.5-11.5 US$ PPP) School Supplies Middle Class (11.5-57.6 US$ PPP) Upper class (57.6+ US$ PPP) Other direct transfers Source: estimates based on ENGIH 2018 and official sources The value of the benefits from direct transfers with respect to market income (incidence) is higher for the poorest households. For the lower deciles, the incidence is more significative for Desayuno Escolar, followed by Comer es Primero and school supplies, but, for the rest of the direct transfer programs, it is much lower (Figure 22, left). Besides, the beneficiary households are concentrated in the lowest three deciles of market income (Figure 22, right), especially in cases of PROVEE, Desayuno Escolar, school supplies and the incentives to attend primary school (ILAE). The beneficiary households of Comer es Primero, Bonogas Hogar and the cash transfer BonoLuz are more concentrated in poorest deciles; even though the share of beneficiary households is higher for first decile, the importance of each decile decreases only gradually up to the sixth decile. Finally, the beneficiaries of the incentives for tertiary education (IES) benefits are concentrated between the fourth and seventh deciles of market income. 56 Figure 22. Direct transfers: Incidence and distribution of beneficiaries Incidence (deciles of market income) Distribution of beneficiaries per decile 30.0 25.0 25.0 20.0 20.0 (% of market income) (% total beneficiaries) 15.0 10.0 15.0 5.0 10.0 0.0 5.0 1 2 3 4 5 6 7 8 9 10 Decile 0.0 1 2 3 4 5 6 7 8 9 10 Desayuno Escolar Comer es primero School supplies Bono gas hogar Bono luz Pensions Comer es primero ILAE IESU Bonogas Hogar ILAE PROVEE IESU Bonoluz PROVEE PIPP Bono gas choferes Other Transfers Desayuno escolar School supplies Source: estimates based on ENGIH 2018 and official sources The aggregate impact of direct transfers in the Dominican Republic is as progressive as it is in Chile and Uruguay in terms of the Kakwani coefficient (i.e., the concentration of benefits in poor households with respect to market income distribution). Coverage rates are high: close to 38 percent of poor households (US$5.5 PPP) in the Dominican Republic benefit from at least one direct transfer program, as do 24 percent of the total population. The Dominican Republic had the second highest coverage of CTs among countries in LAC in 2010, trailing only Ecuador and similar to Argentina and Brazil.57 However, the Dominican direct transfers' marginal contribution to inequality is lower than that of other Latin American countries' contributions because the amount allocated to direct transfers is only half as high as Chile's and a fifth of Uruguay's, for example. However, it is only slightly lower than Mexico's and higher than Colombia's, countries with similar per capita income. ii. Subsidies reach the poor and non-poor alike The overall effect all the indirect subsidies analyzed—electricity, urban transport, metro, and water— is neutral, since they are proportional to income and none of them is pro-poor. As shown in the next figure, The concentration curves for indirect subsidies lie below the Lorenz curve for market income, except the water and electricity subsidies, which are next to it (Figure 23). They represent, in total, 1.9 percente of market income. All of those subsidies have subsidized tariffs 57 Stampini & Tornarolli (2012). 57 (i.e., tariffs that are below costs). In the cases of electricity and water, there are also implicit subsidies related to irregular connections, fraud and nonpayment. Figure 23. Indirect subsidies: concentration curves and share of benefits in poor population Concentration curves, ranked by market income Concentration share by income group 100.0 Water Electricity Urban transport Metro 0.8 19.5 7.9 13.5 80.0 Cum proportion of intervention % of total benefits 60.0 69.4 77.9 65.7 64.3 40.0 20.0 23.1 17.4 14.0 12.3 0.0 6.7 2.1 1.9 3.5 0 10 20 30 40 50 60 70 80 90 100 Cum proportion of population Lorenz curve (Market Income) Water Poor (0-3.2 US$ PPP) Vulnerable (5.5-11.5 US$ PPP) Metro Electricity Middle Class (11.5-57.6 US$ PPP) Upper class (57.6+ US$ PPP) Urban transport Source: estimates based on ENGIH 2018 and official sources The water subsidy, which is regressive, accounts for 26.3 percent of indirect subsidies. Middle class and wealthy households receive 83.9 percent of this subsidy. Their market income increases by 0.24 percent, while the poor’s income increases by around 0.15 percent. The subsidy on metro services, which represents 11 percent of total indirect subsidies, is also regressive. However, Metro subsidy benefits are progressive if it is considered only the population living in Distrito Nacional and Santo Domingo, i.e. the benefits are proportionally higher for lower- income households. It benefits mainly the middle class, which receives 69.4 percent of this subsidy. This subsidy provides the ninth and tenth deciles an increase of 0.28 percent and 0.23 percent in their market incomes, respectively, while the first and second deciles have an increase of just 0.10 percent and 0.07 percent, respectively. This subsidy is not progressive because it only benefits urban population living in Distrito Nacional and Santo Domingo; the poorest population from rural and other urban centers do not use this service, so they do not benefit from this subsidy. The electricity subsidy represents 57.9 percent of the total indirect subsidies included in this analysis and is slightly progressive. Around 79.5 percent of this subsidy is received by non-poor consumers (i.e., middle class and wealthy) and another 17.4 percent benefits the vulnerable population. The market income of the vulnerable and middle class increases by 1.12 percent and 58 1.15 percent, respectively, as a result of this subsidy, but it increases the income of the poor by only between 0.91 percent and 1.02 percent. The subsidy on urban transport services (offered by OMSA) is progressive because it increases the poor's market income more, but it is not pro-poor. Even though this program operates mainly in urban centers, it benefits the lower and middle deciles population, and their services also reach Santiago’s population (Cibao Region). The top two deciles receive 32.4 percent of the subsidies, while the first two only receive 8.2 percent. This subsidy increases the income of the extreme poor and the poor by 0.2 percent, which is double the increase to the middle class (0.1 percent). Figure 24. Indirect subsidies: incidence and concentration of benefits per decile Per decile Concentration 1.4 45.0 (% of market income) (% of market income) 1.2 40.0 35.0 1.0 30.0 0.8 25.0 0.6 20.0 15.0 0.4 10.0 0.2 5.0 - - 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Decile Decile Water Metro Water Metro Electricity Urban transport Electricity Urban transport Source: estimates based on ENGIH 2018 and official sources Overall, subsidies are not the most efficient way to protect the poor in Dominican Republic, as evidenced by the fact that subsidies to water and metro are only slightly regressive, subsidies to electricity are proportional to income and only subsidies to urban transport are progressive but only in relative terms. All subsidies are neutral or proportional to market income. When compared with their Latin American counterparts, indirect subsidies in the Dominican Republic have a lower redistributive impact (Figure 25). This is explained by the lack of progressivity for the metro, electricity and water services. Other countries have a much higher Kakwani index, which means that they have implemented subsidies that benefit the poor and the vulnerable populations more.58 Figure 25. Selected countries: progressivity and redistributive effect of subsidies 58 Subsidies are not the most efficient way to protect the poor, as evidenced by the fact that some of them are regressive. However, these policies may have another objective, like the subsidy to metro. 59 0.01 0.600 MARGINAL CONTRIBUTION SCALE 0.01 0.500 0.01 0.400 KAKWANI SCALE 0.01 0.300 0.00 0.200 0.00 0.100 0.00 0.000 ARG COL CHI GTM SLV DOM MEX BOL DOM 0.00 2017 2014 2013 2014 2017 2013 2014 2015 2018 -0.100 Kakwani Marginal Contribution Source: estimates based on ENGIH 2018 and official sources iii. Education is the main contributor to inequality reduction Benefits supporting primary and secondary education represent more than 10 percent of market income for the poor and vulnerable populations (Figure 26, left), but those benefits are not significant as a percentage of market income for households in middle and upper classes (US$50 PPP or more). This is because the upper deciles are not using public education, at least until the tertiary (higher) level of education. For example, in pre-school and primary education, less than 12 percent of the beneficiaries come from the eighth to tenth deciles, and less than 20 percent of households that used public secondary education are in deciles 8 to 10. The case is quite different for tertiary education, where the three highest income deciles account for almost 40 percent of enrolled students. Furthermore, it is possible to find beneficiaries of the literacy program (Quisqueya Aprende Contigo) even in high income deciles. Figure 26. Education: Distribution of beneficiaries and incidence Distribution of beneficiaries per decile Incidence per income group 60 25.0 100.0 (% of market income( (% of total beneficiaries) 80.0 20.0 60.0 40.0 15.0 20.0 0.0 10.0 Quisqueya Pre-school Primary Secondary Tertiary aprende contigo 5.0 Extreme Poor (0-3.2) Moderate Poor (3.2-5.5) 0.0 Vulnerable (5.5-11.5) Middle Class (11.5-57.6) Pre-school Primary Secondary Tertiary Quisqueya aprende Wealthy (57.6+) 1 2 3 4 5 6 7 8 9 10 contigo Source: estimates based on ENGIH 2018 and official sources Education is pro-poor for all levels, except for tertiary (where it is only progressive because the benefits are higher as a share of the market income of the poor than of the rich). Programs with more benefits to poor population are primary school which has the highest progressivity in education, followed by the pre-primary level and the literacy program (Quisqueya Aprende Contigo). Although the CEQ analysis made with ENIGH 2007 concluded that secondary education was only progressive in 2013, secondary education had become pro-poor in 2018.59 In sum, more than 30 percent of the benefits from education at lower levels— i.e., pre-school, primary and the literacy program—go to poor households. Meanwhile, less than 20 percent of secondary-level benefits and only 8 percent of tertiary-level benefits go to poor households. Figure 27. Education: concentration curves and share of benefits in poor population Concentration curves, ranked by market income Concentration share by income group 100.0 Quisqueya Primary Pre-school Aprende Secondary Tertiary 90.0 0.2 0.0 0.2 0.4 1.1 Cum proportion of fiscal intervention 80.0 25.3 33.4 33.3 37.4 70.0 % of total benefits 60.0 41.6 61.7 50.0 38.2 42.2 43.0 40.0 30.0 20.0 28.8 10.0 32.9 28.4 24.3 0.0 19.2 0 10 20 30 40 50 60 70 80 90 100 8.5 Cum proportion of population Lorenz curve (Market Income) Quisqueya aprende contigo Poor (0-3.2 US$ PPP) Vulnerable (5.5-11.5 US$ PPP) Pre-school Primary Middle Class (11.5-57.6 US$ PPP) Upper class (57.6+ US$ PPP) Secondary Tertiary Source: estimates based on ENGIH 2018 and official sources 59 Aristy-Escuder, et al. (2018). 61 iv. Health insurance is reaching the poor, but other programs are doing less Health expenditures are less progressive than education, since only subsidized family health insurance is pro-poor. The rest of the health programs are progressive, but not pro-poor. IDSS health is progressive but less than other health programs (Figure 28). Furthermore, the services offered by PROMESE/CAL and IDSS are not more progressive than the proportional benefit of "other" expenditures made by the Ministry of Health which include health services received by the entire population. Only subsidized family health insurance and the National Health System offer a share of benefits greater than the share of poor population (17.4 percent). Meanwhile, under that threshold, only 13.5 percent of IDSS benefits, went to poor households. Figure 28. Health: concentration curves and share of benefits in poor population Concentration curves Spending share in poor population Subsidised National Health Other- Ministry 100.0 Insurance System of Health Promese - Cal IDSS 0.6 3.6 3.5 3.2 90.0 10.3 Cum proportion of fiscal intervention 80.0 70.0 36.7 45.8 46.8 48.1 60.0 % of total benefits 51.1 50.0 40.0 30.0 39.9 20.0 33.2 33.2 34.3 10.0 25.1 0.0 0 10 20 30 40 50 60 70 80 90 100 22.7 17.4 16.6 14.4 Cum proportion of population 13.5 Lorenz curve (Market Income) National Health System Poor (0-3.2 US$ PPP) Vulnerable (5.5-11.5 US$ PPP) Subsidised Health Insurance IDSS Middle Class (11.5-57.6 US$ PPP) Upper class (57.6+ US$ PPP) Promese-CAL Other Source: estimates based on ENGIH 2018 and official sources V. Conclusion: a progressive fiscal system with space for improvement The results of the analysis show that the fiscal policy implemented in the Dominican Republic in 2018 was pro-poor, thus reducing inequality. Taxes, indirect subsidies and transfers (including the monetized value of education and health) tend to benefit, in relative terms, the country's poor and vulnerable populations more. When measured through the Gini coefficient, the country’s fiscal policy reduces inequality by 8.6 Gini points, from 0.459 to 0.374. Excluding the ample in-kind benefits of education and health services, a reduction in inequality is still evident, with a 2.8 point reduction in the Gini coefficient. Of all the fiscal measures studied here, the school 62 food program Desayuno Escolar and the personal income tax are the government's two most equalizing interventions. The excise tax on oil products and the ITBIS also have significant impacts on consumable income. On the other hand, subsidies to metro and water services do not reduce inequality. When final income is the end-income concept, primary and secondary education transfers are the most equalizing public interventions. Health transfers—offered through the National Health System, the Ministry of Health Services and the subsidized family health insurance—also have significant equalizing effects. Fiscal policy reduces poverty slightly (by 1.0 percentage point, using the National Poverty Line). Desayuno Escolar and Comer es Primero tend to reduce poverty the most. On the other hand, the ITBIS and fuel excise taxes increase poverty. In terms of taxes (government revenue), the results of the CEQ analysis show that both direct and indirect taxes are progressive. Direct taxes fall mainly on taxpayers in upper- and middle-class households and are highly progressive, yet they represent only 1.4 percent of market income for households, on average. Indirect taxes are more important in terms of tax collection (accounting for 68 percent of tax revenues) and are also progressive, but to a lesser degree. In terms of benefits (government expenditures), most of the direct transfer programs are pro-poor. Direct transfers and education (in-kind) transfers are pro-poor, while health (in-kind) transfers are progressive. The only exceptions are the CCT for tertiary education (IES), contributory pensions, and the bonus for drivers (Bonogas Chofer) which are progressive, but not pro-poor. Indirect subsidies are generally not progressive: their benefits tend to be proportional to market income. More specifically, the electricity subsidy is progressive; the water subsidy is not progressive; the subsidy on urban transport services (via OMSA) is only progressive but not pro- poor. The metro services subsidy (via OPRET) is neutral for the population of Santo Domingo, where metro operates. When compared with those of other Latin American countries, indirect subsidies in the Dominican Republic have less of a redistributive impact. In terms of education and heath transfers, both are progressive, but education more so. Education is pro-poor for all levels—except for tertiary education, which is only progressive. Primary school is the program with highest progressivity, followed by pre-primary and the literacy program Quisqueya Aprende Contigo. Secondary education was also pro-poor in 2018, while it was only progressive back in 2013. Health expenditures are less progressive: only the subsidized family health insurance is truly pro-poor. IDSS health is the least progressive of all health programs. Furthermore, only the subsidized family health insurance and the National Health System transfers offer the poor a greater share of benefits when compared with their share of the population based on market income (17.3 percent). Quality of education and health was not analyzed in this report but improving the quality of public services will be beneficial for poverty and inequality reduction. Overall, most Dominicans are net beneficiaries of the country's fiscal policy. Most of the population—i.e., first eight income deciles—are net beneficiaries of fiscal policy when all benefits and taxes are applied. The impact of benefits with respect to beneficiaries' market incomes, is 63 higher for poor and the highest impact comes from educational and health in-kind transfers. These same benefits are not significant for the highest deciles and for the middle class and wealthy income groups. The impact of direct transfers is only noticeable for first five deciles and for the poor and vulnerable income groups. There is a neutral impact on the middle-class, as taxes are exactly compensated by the benefits received. However, when the benefits from health and education are excluded, only the net cash position of the first three deciles is positive, while upper deciles become net payers. Fiscal policy’s impacts on inequality and poverty, while positive, could be improved. In general, when compared with other countries' fiscal systems using the same CEQ methodology60, the Dominican fiscal system has a positive impact—and yet has the capacity to improve: Additional spending could increase the transfers and subsidies that go directly to poor and vulnerable populations. That said, tax reforms to increase public income should evaluate tradeoffs between keeping the tax system's progressivity (and thus positively impacting the poor) and avoiding the distortionary effects of some taxes on investment, savings and labor supply (which could end up negatively impacting the poor). Any potential negative impact of a tax reform on the poor and vulnerable populations should be compensated by increasing the social expenditures and subsidies that benefit them the most (e.g., Comer es Primero). At the same time, increasing the quality and the coverage of public services—namely, education and health—can allow them to have a higher impact on economic progress and welfare. Some of the findings from this paper such as the fact that Comer es Primero, but not subsidies, tend to reduce poverty the most give some reassurance on the selection of the policy measures adopted last year by government to attenuate the impact of the crisis on poverty. Moreover, it is worth stressing that while the methodology applied in this paper corresponds to data from 2018, such approach could set the foundations for undertaking policy simulations on reforms that the government might be contemplating to achieve fiscal consolidation after the ongoing COVID crisis and without dismantling existing social programs. Such exercise and its outputs would be distinct from this technical background work, but remain highly interconnected given the current economic situation. 60 As in other countries, this CEQ assessment does not include corporate income taxes or import taxes because it is difficult to assign the economic burdens to any single individual or household, and existing methodologies still need to develop a credible analytical framework (See Lustig, 2018). Fiscal reforms can also include changes to Corporate Income Tax, but the effect of this tax was not included in this report. 64 VI. References Aristy-Escuder, J., Cabrera, M., Moreno-Dodson, B., & Sanchez-Martin, M. (2018). The Dominican Republic: Fiscal Policy, Income Redistribution, and Poverty Reduction. In N. Lustig, Commitment to equity handbook: Estimating the impact of fiscal policy on inequality and poverty. Brookings Institution Press. Bachas, P., Gadenne, L., & Anders, J. (2020). Informality, Consumption Taxes and Redistribution. Washington, D.C.: World Bank. Carrasco, H., Garcia, E., Parodi, S., & Vasquez, M. (2016). ¿Cómo se redistribuyen los recursos públicos en República Dominicana? Washington, D.C.: IADB. Coady, D. (2008). The Distributional Impacts of Indirect Tax and Public Pricing Reforms: A Review of Methods and Empirical Evidence. In R. Gillingham, Poverty and Social Impact Analysis by the IMF : Review of Methodology and Selected Evidence. Washington, D.C.: IMF. DGII. (2018). Estimación del incumplimiento tributario en la República Dominicana. Santo Domingo. Dominican Republic Central Bank. (2020). Encuesta Nacional de Gastos e Ingresos de los Hogares ENGIH 2018. Principales resultados y síntesis metodológica. Santo Domingo, Dominican Republic. ECLAC. (2019). Estadísticas de produccion de electricidad de los países del Sistema de Integración Centroamericana (SICA). Datos preliminares a 2018. Mexico, D.F.: ECLAC. ECLAC. (2019). Estadísticas de producción de electricidad de los países del Sistema de la Integración Centroamericana (SICA). Datos preliminares a 2018. Mexico, D.F.: ECLAC. Fuchs, A., Gonzalez, F., & Paz, D. (2019). Distributional effect of tobacco taxation. A comparative analysis. Washington, D.C.: World Bank. Inchauste, G., Goraus, K., & Carrasco, H. (2019, May 10). Distributional incidence of VAT in the context of informality. World Bank. Lavigne, M., & Vargas, L. (2013). Social Protection Systems in Latin America and the Caribben: Dominican Republic,. Santiago, Chile: ECLAC. López-Calva, L., & Ortiz-Juarez, E. (2014). A vulnerability approach to the definition of the middle class. The Journal of Economic Inequality, 12(1), 23-47. Lustig, N. (2018). Commitment to equity handbook: Estimating the impact of fiscal policy on inequality and poverty. Brookings Institution Press. Ministerio de Hacienda. (2018). Gastos Tributarios en República Dominicana. Estimación para el Presupuesto General del Estado del año 2019. Santo Domingo, República Dominicana: Dirección General de Ingresos DGI, Ministerio de Hacienda. 65 Ministerio de Hacienda. (2018). Presupuesto Ejecutado. Tomo I. Santo Domingo: Dirección General de Presupuesto DIGEPRES, Ministerio de Hacienda. OECD. (2008). Reviews of National Policies for Education: Dominican Republic 2008, Reviews of National Policies for Education. Paris: OECD Publishing. doi: https://doi.org/10.1787/9789264040823-en OECD. (2020). Revenue Statistics in Latin America and the Caribbean 2020. Paris, France: OECD Publishing. OECD. (2022). Revenue Statistics in Latin America and the Caribbean 2022. Paris: OECD Publishing. doi:https://doi.org/10.1787/58a2dc35-en-es PROMESE/CAL. (2020). Memoria Institucional 2020. Santo Domingo, Dominican Republic: PROMESE/CAL. Rathe, M. (Dominican Republic: Implementing a health protection system that leaves no one behind). 2018. Washington, D.C.: Word Bank Group. Ravaillion, M. (2018, June). Inequality and Globalization: A review Essay. (A. E. Association, Ed.) Journal of Economic Literature, 620-642. Sanchez-Martin, M., & Senderowitsch, R. (2012). The political economy of the middle class in the Dominican Republic : individualization of public goods, lack of institutional trust and weak collective action. Washington, D.C.: World Bank. SIPEN. (2019). Resumen Estadístico Previsional al 31 de Diciembre de 2018. Santo Domingo, Dominican Republic: SIPEN. Stampini, M., & Tornarolli, L. (2012). The growth of conditional cash transfers in Latin America and the Caribbean: did they go too far? Washington, D.C.: Interamerican Development Bank. World Bank. (2016). Building a better future together. Washington, D.C.: World Bank. World Bank. (2016). Building a Better Future Together: Dominican Republic Policy Notes. Washington, D.C.: World Bank. World Bank. (2017). Gearing up for a more efficient tax system. Washington, D.C.: World Bank. World Bank. (2021). Public Expenditure Review, Chapter 3 Water Supply and Sanitation. Washington, D.C.: World Bank. 66 Annex 1 ITBIS evasion assumptions: place of purchase Not paying taxes when purchased in Paying taxes only in urban areas when purchased in Paying in both urban and rural areas Acueductos Almacenes de provisiones Agencia online de turismo Asociaciones de choferes /prestadores independientes Bares Agencias de viaje internacionales Ayuntamiento, recolectora de basura Botica popular Agencias de viaje nacionales Cabina de peaje Cafeterías Armería Carnicerías Car wash Aseguradoras Colmados, supercolmado y colmadones Centro deportivo (estadio, cancha) Asociación de ahorros y prestamos Comedor económico y popular Funerarias Banca comercial Comedor universitario Heladerías Banca de apuestas deportivas Comedores y/o fonda de comida Minimarkets Banca solidaria Cooperativas de ahorros y créditos Modistas y sastrerías Banco de ahorros y créditos Escuela pública Mueblerías Cadenas de comida rápida Estación del metro Panaderías y reposterías Centro de copia, impresión y encuadernación Estaciones, terminales, paradas de autobuses de empresas o sindicatos Reparadoras de zapatos Centros diagnóstico especial e imágenes Fantasías Salones de belleza, peluquerías y barberías Centros médicos, clínicas Hogares Talleres de mecánica en general, desabolladura y pintura Cines Iglesia Tienda de mascotas Colegios Instituciones gubernamentales Museo Compañías de servicios telefónicos y comunicaciones Instituciones sin fines de lucro Puesto de venta de carbón Compañías de taxis Junta de vecino, empresa administradora, condominio Empresa de alquiler de electrodomésticos y enseres domésticos Compañías de televisión pagada (Telecable) Mercado de pulga Fabrica de muebles Compra por internet Mercados Fabrica de muebles e inmuebles Constructora Pescaderías Fabrica de ropas Consultorios dentales Picapollos Taller de fabricación y reparación de inversores Consultorios médicos Puesto de rifa de aguante y lotería electrónica Financiera Corporaciones de créditos Puestos de leche Centro de uñas (Exclusivamente) Distribuidora de electricidad (EDESTE, EDESUR, EDENORTE, ETC.) Puestos de pollo Fábrica de embutidos Distribuidoras de gas Puestos de venta de empanadas, fritura, hamburguesa, hot dog Agencia de envios (remesas, paquetes, etc) Editoras de periódicos Puestos de venta de neumáticos y/o lubricantes Cafeteria / Comedor del trabajo Empresa de organización de eventos Puestos de venta de periódicos y revistas Puesto de venta de articulos pregrabados o para grabar (CD, DVD, USB, entre otros) Empresa de servicios de lavandería Puestos fábrica hielo y/o plantas de agua purificada Puesto de fabricación, venta y reparación de puertas y ventanas Empresas de servicio de agua en camiones Taller de herrería Empresa privada de Seguridad Empresas de servicio de pintura Talleres de ebanistería, carpintería Compra-venta Envasadoras de gas o GLP Tienda de ropa de paca segunda mano Taller de reparación de electrodomesticos Estacionamiento Vendedor ambulante Spa Estaciones de gasolina Ventorrillos Plataforma Web Estudios fotográficos Empresa o persona administradora de condominio Puesto de ventas para aerosoles y detergentes Farmacias y superfarmacias Centro de internet Tienda de Cigarros Electrónicos Ferreterías Taller de refrigeracion Puesto de venta de carne preparada Gimnasios Guardería privada Tienda de alquiler de prendas de vestir (trajes, vestidos, disfraces) Hipermercados (tiendas por departamentos con supermercado) Estancia Infantil pública Venta de embutidos Hospitales Colmado popular (INESPRE) Banco de sangre Hostales, moteles, cabañas, pensiones Puesto de venta de productos de belleza Tienda de adornos y accesorios de vehículos (auto adornos) Hoteles Puesto de venta de leña Distribuidora de huevo Importadoras de vehículos, Dealers Puesto de frutas Generadora de energía eléctrica privada Institutos comerciales Tienda de dulces Fábrica de calzados Joyerías Asociación de estudiantes Colchonería Laboratorios de análisis clínicos Centro de apuestas (gallera, casino, etc.) Fábrica de queso Librerías Botánica, espiritismo y santería Puesto de venta de pelo postizo Líneas aéreas Puesto de venta de comida (no incluye restaurante, comedor o fonda) Puesto de venta de pampers Oficinas de abogados Puesto de reparación de neumáticos (Gomero) Puesto de venta de café, té y chocolate preparado Opticas Persona Particular o Independiente Tienda de ventas de cerámicas Perfumerías Centro de rehabilitación Pizzerías Asilo Relojerías Cafetería dentro de un establecimiento Restaurantes Centro de educación especial público Supermercados Centro de educación especial privado Tienda de juguetes Centro de rehabilitación privado Tienda de licores, liquor store Tienda de accesorios y bisutería Tienda de venta de computadoras Instituto técnico superior Tienda deportiva Granja Tienda especializada en aparatos y accesorios de salud Escuela de actividades deportivas Tiendas de electrodomésticos Escuela de actividades recreativas (baile, musica, artes plásticas, etc.) Tiendas de repuestos de vehículos Puesto de venta de libros usados Tiendas de ropa Empresas de zonas francas Tiendas de zapatos Tiendas por departamentos Tour operadores internacionales Tour operadores nacionales Universidades Veterinaria Discoteca Colegio semi-privado Parque de atracciones (esparcimiento, juegos, recreación) Tienda de celulares Teatro Fabrica de detergentes y desintectantes Laboratorio de medicamentos Estación de radio/televisión Club de diversión y esparcimiento Club deportivo Organismos internacionales (consulados, embajadas, otros) Puesto de venta de gasolina Tienda de electrónica Tienda de venta de aparatos musicales Fábrica de aparatos y accesorios de salud Empresa de reciclaje de cartón Fábrica de cervezas Fábrica de plaguicidas, insecticidas, pesticidas y fungucidas 67 Annex II Description of methods for fiscal interventions Fiscal Intervention Description of Method (use taxonomy in Handbook) BE VERY DETAILED TO ENSURE REPLICATION Old-age contributory pensions Direct identification. It was assumed that these are practically nil, and they were not estimated. In the case of perceived old-age pensions from the pay-as-you-go system, all but the highest 5 percent of the distribution of pensions were assumed to correspond to this scheme Direct Taxes Personal Income Tax Simulation. Were estimated using statutory rates applicable in year 2018, with no informality or tax evasion assumption for salaried regime. It was assumed that own- account workers are not paying personal income tax. For the salaried regime, progressive income brackets were applied and for interest and dividend income the applicable rate was applied(10 percent). Conditional and Unconditional Cash Transfers Comer es primero Direct identification with adjustment. Beneficiaries and benefits were identified in the survey, but benefits in survey were higher than fiscal accounts, so reported values were scaled down proportionally to get the total in survey similar to fiscal accounts. Incentivo a la asistencia escolar Direct identification. Beneficiaries and benefits were identified in the survey. Other Direct Transfers Direct identification. Beneficiaries and benefits were identified in the survey for programs like "Bono Estudiando Progreso", however this program was not operational during 2018. These benefits were included in case some household received benefits from previous fiscal year. Incentivo a la educación superior Direct identification with adjustment. Benefits and beneficiaries were lower in survey than in Admin Accounts. Additional beneficiaries were selected for enrolled in tertiary education and target population according to SIUBEN (ICV 1, 2 & 3). Benefits for additional beneficiaries were assigned using statutory value of this transfer. Bonogas Hogar Direct identification with adjustment. Benefits and beneficiaries were lower in survey than in Admin Accounts. Additional beneficiaries were selected for target population according to SIUBEN (ICV 1, 2 & 3). Benefits for additional beneficiaries were assigned using statutory value of this transfer. Bonogas Chofer Direct identification with adjustment. Benefits and beneficiaries were lower in survey than in Admin Accounts. Additional beneficiaries were selected according to labor category reported and for target population according to SIUBEN (ICV 1 & 2). Benefits for additional beneficiaries were assigned using statutory value of this transfer. Incentivo a la policía preventiva Direct identification with adjustment. Benefits and beneficiaries were lower in survey than in Admin Accounts. Random selection using target population according to profession Bonoluz Direct identification with adjustment. More beneficiaries in survey than admin accounts. Correction of duplicated benefits per household and random de-selection of beneficiaries Programa incentivo a la marina de guerra Info not available in survey. Protección a la vejez en extrema pobreza Direct identification with adjustment. Benefits and beneficiaries were lower in survey than in Admin Accounts. Random selection using target population according to age. Noncontributory Pensions Near Cash Transfers (Food, School Uniforms, etc.) Alimentación escolar Imputation. Coverage was identified in the survey for those reporting receiving these benefits and attending public schools. Per student benefit reported by Ministry of Education were assigned to identified beneficiaries. School supplies Imputation. Coverage was identified in the survey for those reporting receiving these benefits and attending public schools. Per student benefit reported by Ministry of Education were assigned to identified beneficiaries. 68 Indirect taxes VAT Simulation. A combination of the methodology used in Aristy-Escuder, et al. (2018) and Inchauste, et al. (2019 was applied. Then the embedded rate was estimated using an input-output matrix and information from the places of purchase and the goods and services with a high propensity for tax evasion. For goods with (i) products on which the ITBIS was paid as a condition of the place of purchase the embedded rate was applied; and (ii) exempt goods paid embedded rate. Excise Taxes Fuel Taxes Simulation. For excise taxes, consumption was identified in household surveys, and the effective rate provided by DGII was applied. For oil derivative taxes, the direct and indirect effect was estimated using an input-output matrix. Alcoholic Beverages Simulation. For excise taxes, consumption was identified in household surveys, and the effective rate provided by DGII was applied. Beer Taxes Simulation. For excise taxes, consumption was identified in household surveys, and the effective rate provided by DGII was applied. Tobacco Taxes Simulation. For excise taxes, consumption was identified in household surveys, and the effective rate provided by DGII was applied. Communication Taxes Simulation. For excise taxes, consumption was identified in household surveys, and the effective rate provided by DGII was applied.. Indirect subsidies Electricity Simulation Method: The kwH consumption was estimated using the expenditure in electrical energy. With this consumption, the subsidy of tariff was derived as the difference between full tariff and subsidized tariff. Transport (metro) Simulation. All transfers from the central government are assumed equal to the amount subsidized, then the average subsidy per trip during 2018 was estimated. Transport (bus) Simulation. All transfers from the central government are assumed equal to the amount subsidized, then the average subsidy per trip during 2018 was estimated. For bus subsidies, estimates were made for those who report paying to use the bus and those who travel free. For the first group, the cost per trip was estimated, and then the tariff paid was subtracted to get an estimated subsidy per trip and this result was multiplied by the number of daily trips for the household. For the second group (not paying), the estimated cost per trip was imputed by the number of daily trips according to the information published by OMSA Transparency Portal and ENGHI survey. Water Simulation. The methodology used by Kullman, et al. (Pending citation) was adapted. With data from the public institutions in charge of providing drinking water (INAPA and CAASD), the cost per cubic meter was estimated, and the price paid according to tariff schedules nd subtracted. For those consumers not paying for water service, the median cubic meter consumption for non-payers was imputed to make estimates. Education Imputation. The survey reports whether the individual attends school, the level of Pre-school education and if the school is public or private. The education benefit was based on estimations of per capita expenditure by UNICEF and Ministry of Education. Imputation. The survey reports whether the individual attends school, the level of Primary education and if the school is public or private. The education benefit was based on estimations of per capita expenditure by UNICEF and Ministry of Education. Imputation. The survey reports whether the individual attends school, the level of Secondary education and if the school is public or private. The education benefit was based on estimations of per capita expenditure by UNICEF and Ministry of Education. Imputation. The survey reports whether the individual attends school, the level of Tertiary education and if the school is public or private. The education benefit was based on estimations of per capita expenditure by UNICEF and Ministry of Education. Quisqueya aprende contigo Imputation. The survey reports whether the individual attends school, the level of education and if the school is public or private. The education benefit was based on estimations of per capita expenditure by UNICEF and Ministry of Education. Health National Health System Imputation: Using Endesa 2013 information, a propensity matching score was applied to select beneficiaries in ENGHI survey. Subsidized Family Health Insurance Imputation: The average value of insurance prime paid by the government was imputed for those reporting being part of subsidized family health insurance. 69 IDSS Imputation: The average value of government transfers to IDSS was imputed for those reporting being part of IDSS. Promese - CAL Imputation: Using Endesa 2013 information, a propensity matching score was applied to select beneficiaries in ENGHI survey. Other - Ministry of Health Imputation: For the rest of health expenditures, a per capita benefit was applied to the whole population. 70