The Distributional Impact of Taxes and Social Spending in Bulgaria with an application to Green Fiscal Policies The Distributional Impact of Taxes and Social Spending in Bulgaria with an application to Green Fiscal Policies Kristina Noelle Vaughan Maynor Vinicio Cabrera June 2022 Abstract The Distributional Impact of Taxes and Social Spending in Bulgaria with an application to Green Fiscal Policies 1,2  his paper uses methods developed by the Commitment to Equity Institute and data from the European T Union Statistics on Income and Living Conditions Survey as well as the Household Budget Survey to assess the impact of the fiscal system on poverty and inequality in Bulgaria. This paper presents the first detailed distributional analysis of the fiscal system in Bulgaria including an analysis of the contribution of the individual components of the system. Overall, we find that the fiscal system in Bulgaria contributes to inequality reduction. We find that the fiscal system is poverty reducing when using the lower US$5.50 2011 revised PPP poverty line, and poverty increasing when using the higher EU at-risk-of-poverty poverty line. The difference in results is due to different incidence of taxation along the income distribution. We find that most of the redistributive and poverty-reducing impacts are attributable to contributory pensions, though this is not likely to be sustainable in the context of a rapidly aging population and a dwindling contributory base. Direct taxes and transfers are relatively ineffective in reducing inequality and have one of the lowest redistributive impacts in the EU. As an application of the CEQ framework, we consider the poverty and inequality impacts of increasing carbon taxation and removing fossil fuel subsidies in line with Bulgaria’s decarbonization goals. We find that these measures lead to an increase in poverty but have negligible impacts on inequality. We find that recycling revenues through targeted lump-sum transfers has the greatest mitigating impact on poverty and leads to declines over baseline inequality. This paper is a product of the Poverty and Equity Global Practice Group. It is part of a larger effort by the World Bank to provide open access to its research and contribute to development policy discussions around the world. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. EL Codes: H22, H5, D31, I3 Q5 Keywords: Fiscal Incidence, Fiscal Policy and Inequality, Income Inequality, Poverty, Social J Spending, Social Assistance, Taxation, Environmental Economics, Carbon Taxation, Revenue Recycling, Bulgaria. 1 Prepared by Kristina Vaughan, Economist, Poverty and Equity Global Practice, The World Bank, and Maynor Cabrera, Associate Director for Latin America & the Caribbean, Commitment to Equity Institute. 2 This paper has benefitted from the input and guidance of the peer reviewers Samuel Freije-Rodriguez and Lucian Pop. We are grateful for the valuable insight and discussion from Gabriela Inchauste, Reena Badiani-Magnusson, Nga Thi Viet Nguyen, Salman Zaidi, Tobias Steinegger and the MFM Mod to Microsim CoP working group. All remaining errors are our own. The findings, interpretations, and conclusions in this research note are entirely those of the authors. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Contents v Contents Abstract iv Executive Summary vii 1. Introduction 1 2. The fiscal system in Bulgari 3 2.1 Revenue 3 2.2 Social protection system 4 3. Methodology 9 3.1 Commitment to Equity (CEQ) 9 3.2 Green fiscal analysis 11 4. Data and empirical strategy for the analysis 15 4.1 Household survey data 15 4.2 Fiscal interventions 16 4.3 Empirical strategies and assumptions 18 4.4 Macro validation 21 5. Results: Distributional impact of the fiscal system in Bulgari 22 5.1 Impacts on inequality 22 5.2 Impacts on poverty 24 6. Incidence, progressivity, and marginal contributions of taxes and social spending 25 6.1 Distributional profile of the fiscal system 25 6.2 Direct taxes, indirect taxes, and social insurance contributions 28 6.3 Direct transfers, subsidies, and In-kind transfers 30 7. Welfare impacts of green fiscal policie 33 7.1 Current and planned efforts to reduce carbon dioxide emissions in Bulgaria 33 7.2 Welfare impacts of proposed carbon taxation policies 34 7.3 Revenue recycling measures 37 8. Conclusions and policy insights 40 Appendix A. Technical Appendix 42 Appendix B. Emissions per euro of COICOP item 44 Appendix C. Alternate CEQ analyses 46 Appendix D. Macro validation 50 Appendix E. Detailed description of the fiscal system 52 References 77 vi Figures Figure 1. CEQ income concepts: Pensions as deferred income scenario (PDI) 10 Contents Figure 2. Inequality from market to final income 23 Figure 3. Cross-country comparison: Change in inequality 23 Figure 4. Change in inequality from MIPP to disposable income, EU averages and Bulgaria 23 Figure 5. Poverty and poverty gap change 24 Figure 6. Taxes and benefits by decile, MIPP 25 Figure 7. Taxes and social insurance contributions by income quintile 29 Figure 8. Direct cash transfers by income quintile 30 Figure 9. Indirect subsidies by income quintile 31 Figure 10. Concentration of In-kind transfers 32 Appendix Figure 1. Diagrammatic Representation of Progressivity of Transfers 43 Tables Table 1. Estimates of millions of tonnes of CO2 emissions by source of final demand 14 Table 2. Bulgaria: Revenues 16 Table 3. Bulgaria: Expenditures 17 Table 4. Progressivity and marginal contributions of fiscal instruments 27 Progressivity and marginal contributions of original, low and high carbon taxation Table 5.  scenarios 36 Poverty and inequality at baseline and under the low and high scenarios Table 6.  36 Table 7. Poverty and inequality under various revenue recycling scenarios 38 Appendix table 1. Emissions per euro of COICOP item. Emissions per euro of COICOP item 44 Appendix table 2. PDI Scenario results (adult equivalent, AROP poverty line) 46 Appendix table 3. PGT Scenario results 47 Appendix table 4. Macro validation 50 Appendix table 5. DMIH Thresholds by category of person, 2018 59 Appendix table 6. DMI threshold by category of person, 2018 60 Appendix table 7. Contribution rates by pension fund 68 Appendix table 8. Tax allowances, 2018 73 Appendix table 9. Personal income tax schedule, 2018 73 Appendix table 10. Itemized tax base from main income sources 73 Appendix table 11. Excise duty schedule. 2018 75 THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Executive Summary vii Executive Summary Bulgaria is persistently one of the poorest and most unequal countries in Europe. Over the period 2010–2019, though Bulgaria has managed to lift almost half of its population out of poverty as measured by the US$5.50 per day 2011 revised PPP threshold, in 2019 Bulgaria was the second poorest country in the EU, surpassed only by Romania. While Bulgaria has made gains in reducing poverty, inequality, as measured by the Gini coefficient on disposable income, has widened, rising from 35 in 2010 to 40 in 2019, consistently rendering Bulgaria the most unequal country in the EU. Persistently high poverty and inequality raises concerns about the effectiveness of the fiscal system as a redistributive mechanism and a poverty reduction tool. In the 2021 Systematic Country Diagnostic (SCD), gaps in coverage, targeting and adequacy, particularly for non-contributory programs were identified as one of the main constraints to the efficacy of the social welfare system. As evidence of this, in 2018, 30 percent of the poorest 20 percent of households were not covered by any social assistance program, and beneficiaries in the bottom income quintile received less than a third of the total amount of social assistance benefits disbursed (World Bank, 2021). This report assesses the degree to which the Bulgarian fiscal system is contributing to or mitigating against persistently high levels of poverty and inequality. Using the Commitment to Equity (CEQ) framework, this report examines the degree to which the fiscal system as a singular construct, and each component individually, impacts poverty and inequality in Bulgaria. In doing so, this analysis expands upon prior work undertaken in the SCD by broadening the analysis to the entire fiscal system. Conducting the analysis in the standardized CEQ framework provides the additional benefit of allowing the performance of Bulgaria’s fiscal system to be assessed relative to its aspirational peers. Overall, Bulgaria’s fiscal system leads to a decline in inequality. When pensions are treated as a part of market income, rather than a direct government transfer, the fiscal system as a whole—direct taxes and transfers (excluding contributory pensions), indirect taxes and transfers, and in-kind transfers— leads to a decline in the Gini coefficient from 0.451 for market income plus pensions to 0.327 for final income. The greatest decline in the Gini occurs through the impact of in-kind benefits (such as health and education) decreasing it by 0.06 points, whereas indirect taxes and subsidies account for the lowest decline in the Gini, decreasing it by only 0.02 points. Direct taxes and transfer account for a 0.04 decline in the Gini. However, the redistributive impact of direct taxes and transfers is among the lowest in Europe. Second only to Hungary, Bulgaria is one of the countries that achieves the least redistribution through direct taxes and transfers (excluding contributory pensions). The Gini coefficient declines by 0.04 due to the impact of direct taxes and transfers (excluding contributory pensions). By comparison, in Ireland, the country that has the most redistributive impact of direct taxes and transfers (excluding contributory pensions), the Gini declines by 0.16 points. The limited role for direct taxes and transfers other than pensions could help to explain the persistently high levels of inequality prevalent in Bulgaria. viii When measuring poverty using the at-risk-of-poverty (AROP) poverty line, the measure used throughout the EU, the fiscal system is poverty increasing. The impact of direct and indirect taxes and transfers results in an increase in poverty from 25.4 percent to 27.7 percent when Executive Summary poverty is measured using the AROP poverty line (equivalent to US$16.06 2011 revised PPP per day in 2018). Direct taxes and transfers result in a 2.2 percentage point decline in poverty, but this is surpassed by the increase of 4.5 percentage points due to indirect taxes and transfers, resulting in a net increase in poverty of 2.3 percentage points. Conversely, when poverty is measured using the US$5.50 2011 revised PPP per day, the fiscal system is poverty reducing. Poverty using the US$5.50 2011 revised PPP per day poverty line declines from 9.6 percent to 7.8 percent. Direct taxes and transfers lead to a 2.9 percentage point poverty reduction, surpassing the 1.1 percentage point poverty increase as a result of the impact of indirect taxes and transfers. The result is a net decrease in poverty of 1.8 percentage points. The difference in results across the two lines is largely attributable to the different net cash positions of households along different parts of the income distribution. Households in lower income deciles are on average net beneficiaries of the fiscal system, however, as we move up the income deciles, households become net payers into the fiscal system. Since the US$5.50 poverty line is a substantially lower line, it disproportionately captures households who are net beneficiaries of the system whereas the AROP poverty line (equivalent to US$16.06 2011 revised PPP per day in 2018) is a higher poverty line and as such captures more beneficiaries who are net contributors of the fiscal system. Alternate analyses considering contributory pensions as a direct transfer suggest that contributory pensions account for most of the poverty and inequality reducing impacts of the fiscal system. When contributory pensions are treated as a direct government transfer, they contribute to a 0.101 decline in the Gini coefficient and a 12.8 percentage point decline in poverty as measured using the US$5.50 2011 revised PPP poverty line. To put this in context, after contributory pensions, the greatest poverty reduction impact of any one program is 1.3 percentage points (child and family means-tested), and 0.0332 Gini points (in-kind health transfers) underscoring the magnitude of the impact of contributory pensions. Furthermore, the poverty-reduction impact of contributory pensions is 3.4 times higher than the combined impact of direct transfers. The strong poverty and inequality reducing impact of pensions stems in part from pensions representing a relatively large proportion of income at 14.2 percent, exceeded only by VAT (15.5 percent). The role of contributory old-age pensions in fiscal redistribution and poverty r eduction i s l ikely to be increasingly undermined by adverse demographic trends. Globally, Bulgaria’s population is among the top-10 oldest, reflecting rapid outward migration of the working-age population as well as broader demographic trends such as declining birth rates and increasing life expectancy (World Bank, 2021). Bulgaria’s population is scheduled to shrink by 28 percent by 2070 with declines concentrated among the 25-49 age cohort (World Bank, 2021). As the contributory base for pensions shrinks and the number of pensioners expands, the sustainability of pensions in general and as a source of inequality and poverty reduction will increasingly be undermined. While Bulgaria has taken steps to improve the sustainability of the pension system, such as steadily increasing the retirement age and increasing the minimum period of pension contributions required to claim pensions, these measures are likely to be inadequate to counteract the adverse demographic trends. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES While direct transfers as a group have sizeable impacts on poverty and inequality, the marginal ix impact of each program on poverty is limited. As an aggregate group, direct transfers accounted for a 0.0344 decline in the Gini and a 4.7 percentage point reduction in poverty. Though large in Executive Summary aggregate, this is the combined impact of seven groups of transfers, of which only disability and child and family means-tested benefits have an impact on poverty reduction of more than 1 percentage point. Individually, disability benefits account for the largest impacts on poverty (1.2 percentage points) and inequality (0.009 Gini points) and old-age non-contributory benefits account for the least impacts on poverty (0.05 percentage points) and inequality (0.002 Gini points). The limited role for direct transfers echoes previous concerns about the efficacy of the social assistance system. The “other social assistance category, which includes the main last resort poverty reduction program, the monthly lump-sum social assistance allowance, result in a decline in the Gini of 0.007 points and a decline in poverty of less than 1 percentage point. The findings here corroborate previous concerns surrounding the targeting, adequacy, and generosity of the social assistance system, Bulgaria’s low level of expenditure on periodic means-tested cash benefits, and lack of systemic indexation of key eligibility parameters such as the GMI (Word Bank, 2022). In-kind benefits such as health and education have strong redistributive impacts. In-kind health benefits lead to a reduction of 0.032 Gini points whereas in-kind education benefits lead to a reduction of 0.024 Gini points. These measures have the strongest redistributive impact among direct taxes and transfers excluding contributory pensions. Though the amount of in-kind health and education benefits are relatively evenly distributed throughout the population, the relatively high magnitude as a proportion of income at 8.9 percent and 5.9 percent, respectively, coupled with the strong progressivity of these benefits contributes to the strong redistributive impacts. Indirect taxes, driven chiefly by VAT, have strong inequality reducing impacts but contribute to increases in poverty. As a group, indirect taxes account for a decline of 0.017 Gini points, the majority of which is coming through the impact of VAT (0.0133 percentage points). Indirect taxes contribute to increases in poverty of 2.6 percentage points, of which VAT accounts for 1.9 percentage points. The strong poverty increasing impact of indirect taxes is largely due to VAT representing a relatively high proportion of income, particularly for those in lower income deciles. Overall, the assessment of the fiscal system suggests that it is at odds with the goals of sustainable poverty and inequality reduction and in need of reform. The fiscal system should be restructured to enable direct transfers to play a greater role in poverty and inequality reduction, especially considering worsening demographic trends. Prior suggestions such as an improvement in the coverage and adequacy of the last resort poverty reduction program, automatic indexation of the GMI, ensuring the coefficients that determine eligibility and benefit amounts are better aligned to reflect the vulnerability of the poverty risks of various households and socio-economic groups, and a review of the asset qualifying criteria (World Bank, 2021) should be implemented. Bulgaria is one of the most carbon-intensive economies in Europe and it will have to undertake ambitious carbon policies to meet its decarbonization goals. Bulgaria ranks 36th in carbon intensity in the world, as measured by the CO2 per unit of GDP (World Bank, 2021). This is largely due to the high proportion of fossil fuels in its energy mix and high levels of energy intensity (World Bank, 2021). Along x its current decarbonization trajectory, Bulgaria is not on target to meet is decarbonization goals of a 55 percent reduction of 1995 emission levels by 2030 or carbon neutrality by 2050 as stipulated under the terms of the European Green Deal. Executive Summary Increased levels of carbon taxation and the removal of fossil fuel subsidies are the primary mechanisms being proposed for Bulgaria to meet its decarbonization goals. While Bulgaria already imposes carbon taxes on the most carbon intensive goods as a signatory to the Emissions Trading Scheme (ETS), the current levels are likely to be inadequate if Bulgaria is to meet its decarbonization goals. Prior work conducted by the World Bank, has suggested that Bulgaria will need to increase the levels of taxation of ETS goods and expand taxation to non-ETS sectors to meet its decarbonization goals (World Bank, 2022). The removal of fossil fuel subsidies, though a necessary precursor to Bulgaria meeting its decarbonization goals, is complicated by the fact that Bulgaria does not publicly report its fossil fuel subsidies, though explicit electricity and gas subsidies have been estimated to equivalent to some 3 percent of GDP (IMF, 2021). Carbon taxes are regressive but have negligible impacts on poverty and inequality. Since carbon tax as an indirect tax is a fiscal instrument, the marginal impacts can be assessed in the CEQ framework. The fiscal incidence of taxes analysis shows that just under 50 percent of carbon taxes are borne by the top 20 percent of the income distribution. In comparison, the lowest 20 percent of the income distribution incur only 7 percent of taxes. The limited impact of carbon taxes on poverty and inequality is due in part to them representing a relatively low proportion of income at 0.8 percent. Increasing carbon taxes to ambitious levels leads to carbon taxes becoming less regressive and more equalizing. Increasing the level of ETS carbon-taxation 4.5-fold and equalizing carbon taxation among ETS and non-ETS sectors leads to carbon taxes becoming less regressive, having more, though still negligible, equalizing impacts, and leads to a higher, though still small, poverty-increasing impact. Part of the reason is that under the ambitious scenario, the carbon tax as a percent of income increases from an initial 0.8 percent of income to 3.4 percent, over a four-fold increase. Direct effects of natural gas subsidies are weakly regressive, whereas indirect effects are progressive. Direct effects of natural gas subsidies (those paid directly to households) represented 0.1 percent of income whereas indirect effects of natural gas subsidies (those accruing to households through reduced prices for fuel supplied by firms) represented 0.4 percent of income. Given the low levels relative to income, neither direct effects nor indirect effects had substantial impacts on poverty or inequality. The imposition of higher levels of carbon taxation and the removal of fossil fuel subsidies are projected to lead to increases in poverty but have negligible impacts on inequality. Using the AROP line, increasing carbon taxation in line with Bulgaria’s decarbonization goals and removing fossil fuel subsidies will lead to a 1.0 percentage point increase in poverty under a conservative scenario and a 2.8 percentage point increase in poverty under an ambitious scenario. Similar effects, occur when measuring poverty using the US$5.50 per day 2011 revised PPP poverty line of the magnitude 0.2 and 0.6 percentage points, respectively. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Revenue recycling measures can help offset some of the adverse impacts, but the type and structure xi matters. Though there is a consensus on the need to mitigate adverse impacts using revenue generated from carbon taxation, there is less of a consensus on how best to do so. In this paper, we took an Executive Summary empirical approach comparing the mitigating impacts of revenue-neutral revenue recycling through targeted (to the bottom 20 percent of households) and untargeted lump-sum transfers, expansion of the existing social protection system, and reductions in the VAT and PIT rates. Targeted lump-sum transfers were most effective in mitigating the adverse impacts of carbon taxation. Under the more conservative scenario, lump-sum transfers targeted to the bottom 20 percent fully mitigated the increase in poverty due to carbon taxation and led to reductions in baseline poverty of approximately 2.5 percentage points using the AROP poverty line and 2.9 percentage points using the US$5.50 poverty line. Recycling revenues through the targeted lump-sum transfer also led to a decline of 0.02 in the Gini coefficient. Results under the more ambitious scenario displayed a similar pattern, though the impacts were of a greater magnitude. Universal lump-sum transfers and the expansion of the existing benefits of the social protection system also mitigated the adverse impacts of carbon taxation, but to a lower degree. Under the more conservative scenario, the universal lump-sum transfer fully mitigated the increase in poverty due to carbon taxation and resulted in a decline in baseline poverty of 0.8 percentage points using the AROP poverty line and 0.4 percentage points using the US $5.50 poverty line. The corresponding impacts for an expansion of the social protection system was a 0.6 percentage point reduction over baseline poverty using the AROP poverty line, and a 1.0 percentage point reduction over baseline poverty using the US$5.50 poverty line. Both measures also led to declines over baseline inequality of approximately 0.007 Gini points. Results under the more ambitious scenario displayed a similar pattern, though the impacts were of a greater magnitude. Reductions in VAT and PIT (labour) rates were generally insufficient to offset increases in poverty and led to increases over baseline inequality. Under the more conservative scenario, reducing VAT and PIT (labour) rates only partly offset the increase in poverty due to carbon taxation with poverty remaining 0.2 and 0.5 percentage points, respectively above baseline poverty using the AROP poverty line. While the reduction in VAT fully mitigated the increase in poverty due to increased carbon taxation using the US$5.50 poverty line, poverty remained 0.1 percentage points above baseline under the PIT (labour) scenario. Both reductions in the VAT and the PIT (labour) led to increased levels of inequality with the Gini 0.002 and 0.003 points, respectively, above baseline levels. Results under the more ambitious scenario displayed a similar pattern, though the impacts were of a greater magnitude. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES 1. Introduction 1 1. Introduction The effectiveness of the Bulgarian fiscal system as a poverty reduction tool and a redistributive mechanism has been the subject of debate given the persistently high levels of poverty and inequality prevalent in Bulgaria. Though Bulgaria has made some progress on poverty reduction, almost halving the number of absolute poor in a little under a decade from 11.4 percent in 2010 to 6.2 percent in 2019 as measured by the US$5.50 per day 2011 revised PPP poverty line, it remains one of the poorest countries in the EU. While progress has been made on poverty reduction, inequality, as measured by the Gini coefficient, has consistently widened over time, rising from 35.0 in 2010 to 40.0 in 2019 rendering Bulgaria the most unequal country in the EU. One of the main goals of this paper is to assess the impact of the fiscal system in its entirety on the welfare of Bulgarian households. In addition, as the fiscal system is a complex interplay of individual policies, we will also speak to the nuances of the system by analyzing the marginal impact of the individual components on household welfare. In doing so, we seek to better understand the role of each policy and how policies interact to contribute to the overall poverty, inequality and distributional characteristics of the fiscal system. In addition to assessing the effectiveness of the fiscal system, we analyze the impact of hypothetical carbon taxation scenarios that would allow Bulgaria to meet its decarbonization commitments under the terms of the European Green Deal through the lens of the current fiscal system. We also assess the mitigating impact of recycling the revenues collected from the imposition of carbon taxes using various policy options available in the existing fiscal system such as reducing direct and indirect taxes and increasing social transfers. We use the standard incidence analysis to assess the structure of the Bulgarian fiscal system in 2018 in the context of distributional, poverty and inequality impacts. The incidence analysis is conducted within the Commitment to Equity (CEQ) framework (Lustig et al., 2018) and allows us to answer four key questions: (i) How much income redistribution is achieved in Bulgaria through the current fiscal system? (ii) Who bears the burden of taxation and who receives the benefits? (iii) How equitable and pro-poor is each fiscal instrument? and (iv) What welfare impacts will the hypothetical green fiscal policies have on Bulgarian households? Analyzing the Bulgarian fiscal system within the standard CEQ framework has the advantage of facilitating cross-country comparisons on the efficiency of tax systems with other countries that have been analyzed in the same framework. We use two measures of poverty to assess the impact of the current fiscal system and the hypothetical green fiscal policies: (i) the internationally comparable US$5.50 2011 revised PPP poverty line (US$5.50), and (ii) the at-risk-of-poverty (AROP) poverty line used by countries in the EU, set at 60 percent of the national median equivalized disposable income after social transfers and is equivalent to US$16.06 2011 revised PPP poverty line per day in 2018. For our measure of inequality, we use the standard Gini coefficient. We summarize 5 key findings:(i) though the Bulgarian fiscal system reduces inequality, direct taxes and transfers have one of the least redistributive impacts in Europe; (ii) the fiscal system is poverty reducing when using the lower US$5.50 line, but poverty increasing when using the higher AROP poverty line; (iii) direct transfers as a group have sizeable impacts on poverty reduction, though 2 the marginal impact of any one group is limited; (iv) carbon taxes are poverty increasing but have negligible impacts on inequality; and (v) the efficacy of revenue recycling measures in mitigating increases in poverty due to carbon taxation depends on the design—lump-sum transfers are most 1. Introduction effective, whereas value added tax and personal income tax reductions do not fully offset increases in poverty and lead to increases in inequality. The rest of the paper is organized as follows. Section 2 discusses the fiscal system in Bulgaria, detailing the expenditure and revenue policies in place in 2018. Section 3 discusses the methodology for both the CEQ analysis and the green fiscal analysis. Section 4 discusses the data and empirical strategy for the analysis. Section 5 discusses the distributional impact of the current Bulgarian fiscal system. Section 6 discusses the incidence, progressivity and marginal contribution of each component of the fiscal system, Section 7 discusses the welfare impacts of green fiscal reforms and Section 8 concludes and provides policy insights. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES 2. The fiscal system in Bulgaria 3 2. The Fiscal System in Bulgaria Data on the tax system comes from the Ministry of Finance (MOF) of Bulgaria and the EUROMOD Bulgaria country report 2018–2021. Data on the particulars of the social protection system come from the EUROMOD Bulgaria country report 2018–2021 and European System of integrated Social PROtection Statistics (ESSPROS). The details of the fiscal system are covered in greater detail in Appendix E. All figures reported refer to 2018 values. 2.1 Revenue Bulgaria relies heavily on indirect taxes to fund its expenditure, with indirect taxes accounting for 17.0 percent of GDP and 50.7 percent of total receipts from tax and social contributions in 2018. Within indirect taxes, the bulk of revenue comes from value-added taxes (VAT) which account for 60 percent of indirect taxation revenues. By comparison, direct taxes account for much less revenue, accounting for only 6.8 percent of GDP or a fifth of total receipts from tax and social insurance contributions (SIC) in 2018. Within direct taxes, personal income tax (PIT) accounts for the bulk of direct taxes at 54 percent of direct tax revenue. Social contributions account for a further 9.7 percent of GDP or 30 percent of taxes and contributions collection, of which employers’ contributions account for 60 percent of the total. Given the difficulty with apportioning corporate income taxes to households, this study focuses solely on VAT, PIT, excise taxes, and SIC which together accounted for 72.5 percent of all tax revenue in 2018. Carbon taxes are included in the analysis but are treated as a separate indirect tax. Personal Income Tax PIT is levied on the income accruing to natural persons according to the schedule outlined in Appendix table 9 as governed by the Personal Incomes Taxes Act (PITA). Several sources of income are exempt from taxation including social assistance, pensions, and certain types of interest income (See Appendix E). The Bulgarian tax code also has tax allowances (deductions) which include deductions for permanently disabled persons, children, and bequests, among others (See Appendix table 8). The PIT duty is calculated on gross earnings net of social and health insurance contributions, and deductions (See Appendix table 10). Value Added Tax Value added tax (VAT) is levied on the sale of domestically produced goods and services as well as on the sale of products imported to the Bulgarian market, except for a list of VAT-exempted products (See Appendix E). VAT had a rate of 20 percent for almost all goods and services for 2018. Since 1 April 2011, the VAT rate for hotel accommodation has been 9 percent. 4 Excise duties Excise duties are charged as a percentage of the sales price or customs value or as a flat amount 2. The Fiscal System in Bulgaria in Bulgarian lev per unit or per other quantity measures on alcoholic beverages, cigarettes and energy and electricity products. The schedule of taxes by good is outlined in Appendix table 11. Carbon taxes Bulgaria, as a signatory to the EU’s Emission Trading Scheme (ETS) (See section 7.1 Current and planned efforts to reduce carbon dioxide emissions in Bulgaria) imposes an EU-wide carbon tax on the most carbon intensive sectors in that region. In 2018, the price of carbon (carbon tax) was set at 19.32 euros per tonne of CO2 emissions and applied to the following sectors: ›  carbon dioxide (CO2) from » e  lectricity and heat generation, »  energy-intensive industry sectors including oil refineries, steel works, and production of iron, aluminium, metals, cement, lime, glass, ceramics, pulp, paper, cardboard, acids and bulk organic chemicals, »  commercial aviation within the European Economic Area ›  nitrous oxide (N2O) from production of nitric, adipic and glyoxylic acids and glyoxal; ›  perfluorocarbons (PFCs) from production of aluminium 2.2 Social protection system › Cash transfers Sickness Sickness benefits are benefits granted upon medical referral. Depending on the benefit, a person may or may not have to satisfy a 6-month period of insurance. Employers pay the cash benefit for the first three working days of the temporary incapacity in the amount of 70 percent of the average daily gross salary for the month of occurrence of the incapacity. Thereafter, the benefit is 80 percent or 90 percent, depending on the benefit, of the average daily gross salary or of the contributory income if the gross salary is lower than the minimum contributory income. From the fourth day of illness until recovery of capacity to work, the benefit is paid for by the National Social Insurance Institute. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Disability benefits 5 These consist of different systems in operation for the provision of benefits for work-related 2. The Fiscal System in Bulgaria incapacity. The primary scheme is funded through contributory social insurance, which is compulsory for most class of workers, except pensioners who perform work without having entered into an employment relationship. The benefit is granted for the term of the invalidity or for life when disabled persons have attained pensionable age. In order to claim the benefit, individuals must be up to date on social insurance contributions. The amount of the benefit is determined by several factors including the number of years of insurance, age, degree of working incapacity or disability, pensions, earnings, average contributory income of the country, and average salary of the country. Old age Old age benefits are a combination of three parallel systems with their own characteristics: › Public pension insurance f  unctions as a pay-as-you-go system operating on a defined-benefit principle that is mandatory for all employees and self-employed individuals. The scheme is funded by contributions from both employer and employee, in the case of employees, and funded entirely by the self-employed in the case of self-employment. In order to claim this benefit, individuals must reach retirement age and have a fixed number of years of insurance. For men, the condition in 2018 was 64 years of age and 1 month and 38 years and 6 months of insurance and for women, the condition in 2018 was 61 years of age and 2 months and 35 years and 6 months of insurance. Benefits are calculated based on reference earnings, periods of insurance, monthly average contributory income for the country, and the individual coefficient of the claimant. In 2018, the rate of contribution was 8.78 percent for persons born before 1960 and 6.58 percent for persons born after 1959. › Supplementary compulsory pension is based on a defined contributory fully funded principle and consisting of two funds: the Universal pension fund and the Professional pension fund. » The Universal pension fund  is mandatory only for people born after 31 December 1959. The fund covers all employees and self-employed regardless of job category. Benefits are paid by the pension fund as a life-long pension. Benefits are generally paid out upon reaching the legal retirement age, but early retirement up to 5 years before the statutory retirement age is allowed under certain conditions. Both employers and employees make contributions to the universal pension fund. In 2018, the contribution rate was 5 percent, with the employer funding 2.8 percent and the employee funding 2.2 percent; self-employed persons must fund the entire 5 percent. » The Professional pension fund  is an early retirement fund for those who work in hazardous environments—labor category 1 and 2. Individuals covered under this scheme are entitled to fixed early-retirement pensions in advance of the statutory retirement age if they meet certain contributory conditions. Workers continue to receive early-retirement pensions until they become eligible for the public or universal pension schemes. Benefits are calculated 6 based on the accumulated sum in the individual’s capitalization account, bio-metric tables reflecting life expectancy and the approved technical interest rate. Only employers contribute to the funds at rates of 12 (1st labour category) or 7 percent (2nd labour category). 2. The Fiscal System in Bulgaria › Supplementary voluntary pension  insurance is based on a defined contributory fully funded principle and consists of two funds: the Supplementary Voluntary Pension Insurance fund and the Occupational Pension Fund. Both funds are optional for all persons having reached the age of 16. » The Supplementary Voluntary pension fund is normally based on a contract between the individual and a pension fund management company which sets out contribution levels. Benefits are paid out directly by the fund in the form of a life-long pension, a fixed-period pension, a lump-sum, or programmed withdrawals. » The Occupational pension fund  uses collective agreement or collective bargaining to determine the coverage. Under this scheme beneficiaries are eligible for a pension at the age of 60 or up to 5 years before according to the rules contained in the collective bargaining agreement. Family benefit Family benefits consist of over 15 programs including child benefits, parental benefits, childcare allowances and parental leave. Programs can be contributory, non-contributory, means-tested or non-means tested (See Appendix E). The main last-resort poverty reduction program is the monthly lump-sum social assistance benefit. The program is means-tested and non-contributory and grants low-income families who have income below the Differentiated Minimum Income (DMI) and meet other qualifying conditions access to this benefit. The DMI varies with the characteristics of each individual and is expressed as a percentage of the GMI (See Appendix table 6). The amount of the benefit is determined as a difference between the DMI of the household and the gross family income from all sources, net of allowed exceptions. In 2018, the monthly GMI was BGN 75 (US$88 2011 revised PPP) (annual BGN 900, US$1064.7 2011 revised PPP). Among other programs are maternity leave and parental leave benefits, child allowances, one- time grants for each newborn, and benefits for adoptive and foster parents. The contributory maternity (and paternity) benefit for pregnancy and childbirth affords qualifying working mothers 410 days of childbirth leave and grants fathers 15 days of leave subsequent to the birth of the child, provided they have been insured under the General Sickness and Maternity Fund for at least 12 months. Unemployment Unemployment allowance is a compulsory social insurance (contributory benefit) scheme provided to select unemployed persons. The duration of the benefit varies with the period of insurance contributions ranging from 4 months to a year. The amount of the unemployment benefit is 60 percent of the average gross contributory income calculated for the assessment period, which is the last 24 months preceding the termination of the employment. In 2018, the minimum amount of the unemployment benefit was BGN 9 (US$10.65 2011 revised PPP) per day and the maximum amount of the unemployment benefit was BGN 74.29 (US$87.9 2011 PPP) per day (approximately BGN 1560 per month). THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Monthly lump-sum social assistance allowance 7 This is a non-contributory allowance granted to households with low incomes. The allowance 2. The Fiscal System in Bulgaria is granted to households whose average gross income per household member is below the Differentiated Minimum Income (DMI) threshold. The DMI threshold varies based on the age, health, family and educational status of each individual and is calculated based on a percentage of the GMI (See Appendix table 6). The amount of the benefit is determined as the difference between the DMI of the household and the gross family income from all sources net of certain exceptions (See Appendix F). The GMI amount is determined by the Council of Ministers and was set at BGN 75 (US$88 2011 revised PPP) per month in 2018. Social pension This is a non-contributory benefit granted to a person who has turned 70 years of age. In order to qualify for the benefit, the person should not be in receipt of any other pension. The benefit is paid to individuals living alone or in households with gross annual income per household member which is below a “Guaranteed Minimum Income” (GMI) threshold (See Appendix E). In 2018, the GMI was BGN 75 (US$88 2011 revised PPP) monthly or BGN 900 (US$1064.7 2011 revised PPP) annually. In 2018, the social old-age pension was on average BGN 123.28 (US$145.84 revised PPP) per month. Heating allowances This is a non-contributory allowance granted to lone persons or households with low incomes. Persons with income lower than the Differentiated Minimum Income for Heating (DMIH) have the right to a targeted heating allowance, where the DMIH for heating varies based on family demographics (See Appendix table 5). The amount of the benefit is determined by the Minister of Labour and Social Policy for each heating season (5 months from 1 November to 30 March). For the 2017/2018 heating season, the benefit was BGN 73.02 (US$86.4 2011 revised PPP) per month of the heating season. For the 2018/2019 heating season, the corresponding benefit was BGN 74.83 (US$88.5 2011 revised PPP) per month. Housing allowances Persons whose income for the preceding month is less than 250 percent of the differentiated minimum income are entitled to a targeted monthly allowance for the payment of rents for municipality lodgings. The housing allowance is granted to orphans under the age of 25, lone elderly people over 70 and single parents. Survivor’s benefit Survivor’s benefits are social insurance contributory benefits provided to surviving spouses, children and parents of deceased insured persons. The survivor’s pension is calculated as a percentage of the pension the deceased received or would have received at the time of their death. The survivors’ pension is determined as a percentage of the personal pension of the deceased insured person at 50 percent, 75 percent and 100 percent if there are one, two, or three or more survivors, respectively. 8 › In-kind transfers Healthcare 2. The Fiscal System in Bulgaria The public healthcare system is a combination of a compulsory social insurance scheme (the General Sickness and Maternity Fund) financed by contributions and a scheme funded by taxes providing benefits in kind, other than those provided by the contribution-funded scheme. Patients under the contribution-funded scheme pay the physician, dentist, or health-care facility a user- fee for each visit which stood at BGN 2.90 in 2018. Medicines for certain diseases are paid for either partially or fully by the National Health Insurance Fund with the rest having to pay out of pocket (OOP). In 2018, OOP health expenditures accounted for 40.5 percent of current healthcare expenditure (World Bank, 2021). Education The education system in Bulgaria is predominantly publicly funded.3 The funding structure differs depending on the educational level. Kindergarten, primary and basic schools and general education schools are predominantly funded through the municipal budget, whereas vocational schools, which are mainly state owned are financed through the state budget. Kindergartens are also fee-based which helps provide some of the funding. Higher education institutions are funded through a combination of fees, state subsidies, and through other means such as donations, grants, and sponsorships among others. The education system is dominated by public institutions with few private institutions. In 2018, the proportion of enrollment in private institutions varies by level, registering 1.7 percent for up to the secondary level, 0.8 percent for private vocational schools, 18.3 percent for private colleges and 12.5 percent for universities and equivalent higher education schools.4 Expenditure on private institutions is highest at the tertiary level. 3 https://eacea.ec.europa.eu/national-policies/eurydice/content/funding-education12_en#:~:text=The%20education%20system%20 in%20Bulgaria,GDP%20was%20spent%20on%20education 4 Education in the Republic of Bulgaria for 2018/2019 School Year, Republic of Bulgaria National Statistics Institute. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES 3. Methodology 9 3. Methodology 3.1 Commitment to Equity (CEQ) To assess the distributional impact of the fiscal system in Bulgaria, the analysis follows the Commitment to Equity (CEQ) methodology developed by Lustig et al. (2018). The methodology is based on the estimation of income concepts that may or may not include specific fiscal interventions (Figure 1) to assess how the system and specific components (taxes and transfers) affect poverty and inequality. We apply this approach to answer the following questions: How much does the fiscal system contribute to changing market income inequality? Does it help reduce poverty? Which taxes and transfers are progressive or pro-poor? What would be the distributional impact of specific fiscal interventions? In this analysis, the following income concepts apply: › Market income  includes labor income (gross employee cash or near cash income, gross non- cash employee income, and employers’ social insurance contributions), income from capital assets (rent of property or land, interest, dividends, and profits), and private transfers (pensions from private pensions plans, and net regular inter-household cash transfers) and other income sources5. › Market income plus pensions (MIPP) is the sum of market income and contributory old-age pensions minus contributions to old-age pensions through the various pension funds. › Disposable income  is constructed by adding direct transfers and subtracting direct taxes and social contributions excluding contributory old-age pensions from market income plus pensions. Bulgaria’s direct taxes include personal income tax, and other social security contributions comprising the General Sickness & Maternity Fund, and Health Fund for employers, employees, and the self-employed; Unemployment Fund for employees and employers; and Work injury and occupational disease for employers. Direct transfers include child and family transfers, educational allowances, and social protection transfers like disability, survivors, unemployment, and housing. › Consumable income  adds subsidies and subtracts indirect taxes from disposable income. Bulgaria’s indirect taxes analyzed here are the VAT; excises on alcoholic beverages, tobacco, and fuel; and other indirect taxes not classified as VAT or excises. Subsidies included are electricity and gas. › Final income  adds to consumable income benefits in the form of social spending on health and education. 5 CEQ income aggregates include the value of goods produced for own consumption and the imputed value of the owner-occupied dwelling. However, to make the income aggregate more comparable with official poverty and World Bank’s international poverty estimates we exclude them from the market income construction. 10 As so much depends on the design and functions of the pension system in each country, there is no consensus in the literature on how to treat contributory pensions and related contributions. There are usually two scenarios. First is the Pension as Deferred Income (PDI) scenario where old-age 3. Methodology pension income is treated as deferred income and therefore added to market income, and pension contributions are treated as savings. Second is the pensions as government transfers scenario (PGT), in which pension income is treated as government transfers and pension contributions as taxes and thus subtracted from market income in line with the standard EU measurement of disposable income. The results from PDI scenario are analyzed here, and the results of the PGT are presented as an alternate analysis in Appendix table 3 in Appendix C. Figure 1. CEQ income concepts: Pensions as deferred income scenario (PDI) Market income plus pensions Labor income, capital income, private transfers, other income (Market income) PLUS Contributory pensions MINUS Contributions to Old-age pensions Direct Transfers Direct Taxes Child and family, social Personal income tax protection, and other transfers Social contributions Disposable Income Indirect subsidies Indirect taxes VAT, Excises (tobacco, Electricity and gas alcoholic beverages, fuel) Consumable Income In-kind transfers Education and health Final Income Source: Authors’ interpretation based on Lustig (2018). To unpack the distributional impact of the fiscal system, the following standard indicators6 are used. We include a visual depiction of these measures in Appendix Figure 1. › Concentration coefficient: T he coefficient of concentration (a quasi-Gini) is an index summarizing the concentration curve of a tax or transfer, which ranges from -1 to 1. A negative Concentration Coefficient indicates that the transfer is concentrated among the poorest, that is that it is pro- poor or absolute progressive in the CEQ terminology. While a positive Concentration Coefficient but lower than the Gini coefficient indicates that a transfer (tax) is progressive (regressive) and if it is higher than the Gini Coefficient, the transfer (tax) is regressive (progressive). The curves from which the measure is derived cover the cumulative percentage of households, from poor to rich ranked by MIPP, on the horizontal axis and the cumulative percentage of tax paid and transfers received by each centile. This is used together with the pre-fiscal income Gini (before any taxes and transfers) to construct the Kakwani index. 6 For more details, see Lustig 2018. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES › Kakwani index:  This is a useful measure for determining progressivity of taxes or transfers. The 11 Kakwani index for taxes is defined as the difference between the concentration coefficient of the tax and the Gini for pre-fiscal income; for transfers, it is defined as the difference between the Gini 3. Methodology for pre-fiscal income and the concentration coefficient of the transfer. A Kakwani index rating will be positive if a tax is progressive, negative if it is regressive and 0 if it is proportional in relative terms. The Kakwani index is positive if a transfer is progressive in relative terms. If the Kakwani is positive and the concentration coefficient is negative, the transfer is pro-poor.7 › Redistributive effect:  This indicator captures the marginal contribution of the fiscal system to the Gini coefficient of inequality. The marginal contribution is understood as the difference between the Gini coefficient with and without the tax or transfer.8 If positive, it captures a redistributive effect, corresponding to a lower Gini after the tax or transfer. › Poverty reduction effect: T  his captures the marginal contribution of the fiscal system element(s) to poverty defined at a certain poverty line. Again, the marginal contribution is understood as the difference between the poverty rate with and without the tax or transfer. If positive, it captures a poverty reduction effect, corresponding to a lower poverty rate after the tax or transfer. For comparability across countries, we present the results using the internationally comparable US5.50 per day 2011 revised PPP poverty line and present measures using the AROP poverty line (equivalent to US$16.06 per day 2011 revised PPP) in Error! Reference source not found. in Appendix C. Note that a progressive tax does not necessarily reduce inequality,9 which would instead require a positive redistribution effect, or poverty, which would require a positive poverty reduction effect. Similarly, it is possible for a tax to reduce inequality but at the same time increase poverty.10 For a complete picture, measures of progressivity must therefore be combined with marginal contributions to evaluate the effects of fiscal interventions on poverty and inequality. 3.2 Green fiscal analysis One of the main goals of this paper is to assess the distributional impact of carbon taxation and revenue recycling measures, here referred to as green fiscal analysis. In order to do so, we must first determine the carbon footprint of each household, that is, the total CO2 emissions associated with each household’s consumption profile. This is computed as the sum of CO2 emissions per amount of 7 A transfer is progressive in relative terms if the proportion received in relation to pre-fiscal income increases as income rises. This happens when the concentration curve lies between the pre-fiscal income Lorenz curve and the 45-degree line. A transfer is pro-poor when the concentration curve lies above the 45-degree line as this means that the amount of the transfer is higher in absolute terms for lower-income populations. 8 Note that there is path dependence in estimating these marginal contributions, since the order in which each intervention is considered matters for the magnitude of the estimated marginal contribution. The estimation approach uses a Shapely decomposition to address this issue, which involves estimating marginal contributions in every possible path and then taking the average. 9 In the literature, this is known as the Lambert conundrum (Lambert 2001; Lustig & Higgins, 2018). Taxes, for instance, can be regressive according to the Kakwani index but when combined with transfers make the system more equalizing than without the regressive taxes. For a thorough discussion see Enami, Lustig, and Aranda (2018). 10 See Higgins and Lustig (2018). 12 each product consumed. The CO2 emissions associated with each product is determined by multiplying the CO2 emissions embodied per unit of expenditure, here, CO2 emissions per euro, by the euros of expenditure for each of the 44 3-digit level COICOP products covered in the Household Budget Survey 3. Methodology (HBS)11. Whereas the consumption expenditure for each item is recorded in the HBS (in Bulgarian lev converted to euros using the exchange rate), the CO2 emissions per euro of expenditure for each of the 44 3-digit COICOP items must be computed. The approach to estimate the CO2 emissions per unit of expenditure largely follows the approach in Ivanova and Wood (2020), except here we do not adjust household consumption expenditure levels reported in the HBS to match those reported in the National Accounts (NA) data and we use the OECD’s emissions embodied in trade database (TECO2) as our source of CO2 emissions data rather than EXIOBASE as it has more recent emissions data. Mapping emissions data to household survey data is a non-trivial process, requiring a number of adjustments to arrive at the final estimate. We proceed systematically in the following steps: 1. Map COICOP categories to CPA categories. Databases containing emissions data linked to products or industries are usually based on the Classification of Products by Activity (CPA) product classification system, aligned with most input-output tables, whereas household consumption data is based on the Classification of Individual Consumption According to Purpose (COICOP) product classification which is aligned with most household expenditure surveys. In order to reconcile the difference between these two classification systems, we rely on contingency/bridge matrices which map COICOP products to CPA products or industries and vice versa. Contingency/ bridge matrices are produced by statistics offices as part of the NA but are rarely made available to the public except in the case of Austria, the Czech Republic, Denmark, Estonia, Finland, Slovakia and Sweden. Here we rely on matrices estimated for Bulgaria previously in Cazcarro et al. (2022). Since a contingency/bridge matrix is not available for Bulgaria, we use the contingency tables of the Czech Republic identified and provided in Cazcarro et al. (2022) as the country which is likely to have the contingency/bridge matrix most comparable to Bulgaria based on similarities along select key indicators of development. We adapt the 2016 contingency/bridge matrix from Cazcarro et al. (2022) for the purposes of this analysis where the resulting contingency/bridge matrix has a dimension 44 (COICOP products) x 43 (CPA industries). Each COICOP product i is mapped to one or more CPA industries j. The element xi , j of the matrix is the share of expenditure associated with product i from the household survey that is mapped to industry j in the NA data and rows of the matrix sum to 1, i.e. ∑ J j =1 xij = 1. Here we rely on two key assumptions—that the distribution of COICOP expenditure across CPA industries has stayed the same since 2016 and that the Czech Republic is sufficiently comparable to Bulgaria to have similar contingency tables/bridge matrices. 2. Get emissions data from the OECD in tonnes of CO2 per euro. Emissions data for 2018 is from the OECDs database on carbon dioxide emissions embodied in international trade (TECO2) database. Specifically, we use total CO2 emissions embodied in domestic final demand, that is, CO2 emitted and consumed domestically plus CO2 emitted abroad and embodied in final demand (CO2 imports). The database is constructed by combining the OECD’s Inter-Country Input-Output (ICIO) tables from 2018 with data from CO2 emissions from fuel combustion and other industry statistics provided by the International Energy Agency (IEA). We use these to estimate demand- based carbon dioxide emissions, for each of the 43 CPA industries for select regions of the world. 11 Expenditure on narcotics, prostitution and imputed rent are not reported in the Bulgaria Household Budget Survey. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Since we need emissions per euro of domestic final demand, we construct an aggregate of final 13 demand using the OECD’s 2018 Inter-Country Input Output (ICIO) tables. From here we construct CO2 emissions per euro of final demand for each of the 43 CPA industries by dividing the total CO2 3. Methodology emissions embodied in domestic final demand (millions of tonnes of CO2 emissions) by total final demand (reported in millions of USD but converted to millions of Euros) for each of the 43 CPA industries. The result is tonnes of CO2 emission per euro for each of the 43 CPA industries. One limitation here is that we cannot account for CO2 emitted directly by households largely due to utilities and personal transport. These are presented as an aggregate amount and are not mapped to or disaggregated by any sector, precluding their inclusion. Emissions from domestic household use in Bulgaria accounted for 4.1 million tonnes of CO2 emissions of the total 32.1 million tonnes of CO2 embodied in final demand in Bulgaria in 2018 (OECD, 2021). 3. Rebase emissions per euro from purchaser prices to basic prices. Information on emissions per euro from the 43 CPA industries is reported in basic prices, exclusive of trade margins (realized by wholesalers and retailers), transport margins (realized by the transport sector) and taxation (accrued to the government). Similarly, in the OECD emissions database, emissions associated with the trade and transport of goods is assigned to the trade and transport sectors, rather than being assigned to the individual sectors associated with each good (e.g. clothing). However, in the HBS information on consumption expenditure in euros for each of the 44 COICOP products is reported in purchaser prices which are inclusive of trade margins, transport margins, and VAT taxes. If we do not deduct those margins from expenditure, we risk understating the emissions associated with the trade and transport sector, and overstating the emissions associated with goods from the other sectors (e.g. clothing).12 Similarly, having taxes included in consumption expenditure risks artificially inflating the emissions data, since emissions data is reported in tonnes of CO2 emissions per euro of expenditure. Since the tax and trade margins are mapped to the 43 CPA industries, in order to take into account this difference in prices between the HBS and the emissions data, we deduct the proportion of tax, trade and transport margins in final consumption expenditure in emissions per euro for each of the 43 CPA industries using data from the OECD on tax, transport, and trade margins in 2010. Here we assume that the tax, trade and transport proportions in final consumption expenditure have remained the same since 2010 since more recent data is not available. 4. Convert CPA emissions per euro to COICOP emissions per euro. Using the contingency tables from step 1, emissions per euro reported in modified basic prices from the 43 CPA industries are mapped back to emissions per euro for each of the 44 COICOP products. Emission per euro in modified basic prices for each COICOP product is a weighted average of the emissions per euro in modified basic prices of each of the 43 CPA industries determined in step 2 where the weights are the shares of expenditure on COICOP product j attributed to each of the 43 CPA industries from 12 The example traditionally given is the purchase of a pair of shoes (Steen-Olsen et al., 2016). In the HBS, the purchase of shoes for 1,000 BGN would be entirely assigned to the COICOP category Footwear. Due to VAT rates, 20 percent of this would be deducted from the price at which it was purchased, leading to a price before taxes of 800 BGN. Now given the trade and transport margins, this 800 BGN is actually distributed over several sectors for instance 400 to Footwear, 200 to the Trade sector and 200 to the Transport sector. Since different sectors have different emissions associated with them, and since including VAT in the purchase price artificially inflates the emissions associated with products, assigning the 1000 BGN entirely to the Footwear sector would give a different emissions profile if tax, trade, and transport margins were not appropriately assigned. 14 the contingency/bridge matrix in step 1. The emissions associated with each of the 44 COICOP products are shown in Appendix B. 3. Methodology 5. Determine the carbon footprint of each household. The carbon footprint of each household is then determined by multiplying total expenditure in euros for each of the 43 COICOP categories by the emissions per euro at modified basic prices calculated in step 4 for each of the 43 COICOP categories to determine the emissions profile of each household. With these caveats in mind, in 2018 based on HBS data on consumption patterns, 17.5 million tonnes of CO2 emissions were attributable to Bulgarian households. When comparing this to the figure of 32.1 million tonnes of CO2 emissions attributed to final demand in the OECD database, it is important to contextualize the discrepancy. First, we are only considering emissions attributed to Household Final Consumption Expenditure in the HBS, whereas the 32.1 million tonnes of CO2 emissions cited includes emissions from total final demand—Household Final Consumption Expenditure from NA, Direct CO2 emissions by Households, Non-Profit Institutions Serving Households, General Government Fixed Capital Formation, Gross Fixed Capital Formation, Changes in Inventories and Valuables, and Direct purchases abroad by residents (See Table 1). Additionally, Household Final Consumption Expenditure in the HBS is often much lower, and especially lower in Bulgaria, than Household Final Consumption Expenditure in the NA, resulting in a lower level of emissions for the HBS as compared with the NA (See Table 1). The discrepancy is due to well-known issues with underreporting in the HBS compared with the NA. Second, we cannot accurately account for the 4.1 million tonnes of CO2 emissions directly emitted by households as this is presented as an aggregate amount and not assigned to the individual CPA products or industries, hence these emissions which are included and assigned to households in the 32.1 million tonnes of CO2 aggregate are excluded from our aggregates. Table 1. Estimates of millions of tonnes of CO 2 emissions by source of final deman Estimated Included in Source millions of tonnes the analysis of CO2 emissions Household Final Consumption Expenditure (HBS estimate) 17.5 Y Household Final Consumption Expenditure from NA 23.7 N Direct CO2 emissions by Households 4.1 N Non-Profit Institutions Serving Households 0.0 N General Government Fixed Capital Formation 0.6 N Gross Fixed Capital Formation 2.0 N Changes in Inventories and Valuables 1.2 N Direct purchases abroad by residents 0.4 N Source: Authors’ computations based on OECD ICIO and TECO2 databases. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES 4. Data and empirical strategy 15 for the analysis 4. Data and Empirical Strategy for the Analysis 4.1 Household survey data The analysis uses data from the Bulgaria HBS for 2018 which captures consumption expenditure and income and the European Union Statistics on Income and Living Conditions Survey (EU-SILC or SILC for short) 2019 which captures information on household income in 2018. The HBS is carried out annually and captures information on consumption, expenditure, income and other key socioeconomic indicators as well as demographic information. The 2018 HBS covered 2931 households and is representative of the resident population outside of hospitals, residential care facilities, prisons, military bases and other such institutions. Although the HBS captures household income and benefits, it is not as detailed as the SILC, given that it is primarily a household expenditure survey and not an income survey. The 2019 SILC (capturing 2018 income) captures more detailed information on sources of income such as wage income of employees, income of self-employed persons, property and capital income, old age pensions and a more detailed disaggregation of social assistance benefits. Moreover, the SILC has a larger sample size, sampling 7350 households in 2019, which may better capture variations across households allowing for better analysis on how fiscal interventions affect the distribution of income. Given the limitations of the HBS, and the fact that welfare is measured by income and not consumption in Bulgaria, this analysis will primarily be based on the 2019 SILC and uses income as a basic measure for analysis. We estimate market income using the income components in the 2019 SILC (capturing income year 2018): labor income, private transfers, and other income13. Market Income is the original income for the Pensions as a Government Transfer Scenario (PGT) and includes non-government old-age pensions that are part of private transfers. In our benchmark scenario or pensions as deferred income (PDI) scenario, contributory old-age pensions are treated as deferred income, and thus included in market income plus pensions, now labeled MIPP, market income plus pensions, that is net of social security contributions to old-age pensions. From MIPP, we obtain disposable income by removing direct taxes and non-contributory pension social insurance contributions and adding direct transfers, excluding contributory old-age pensions. To construct consumable income, we deduct indirect taxes (VAT and excises) from disposable income. However, indirect taxes are based on consumption, data for which are available only in the 2018 HBS. We, therefore, simulate indirect taxes and subsidies in the HBS and then apply survey-to-survey imputation to assign them to each SILC household. Finally, we add social spending on health and education, which are both imputed, to obtain the final income. The analysis is complemented by data from national and other public finance accounts from the National Statistical Institute of Bulgaria (NSI), Eurostat, IMF, and OECD. This includes information on consolidated government budgets, detailed revenue collection, and detailed information on 13 This analysis does not include imputed value of the use of dwelling by the owner and self-consumption. Those figures are included in SILC, but to produce poverty and inequality figures comparable with national (NSI) and international estimates (World Bank), we exclude them from the analysis. 16 expenditure and social benefits, like ESPROSS data. OECD is the source for the input-output table for the Bulgarian economy for the year 2010. Since more recent input-output tables for Bulgaria are not available, we revert to using the most recently available input-output tables under the assumption 4. Data and Empirical Strategy for the Analysis that the domestic inter-industry transfers have remained the same or largely similar since 2010. Given this is a strong assumption, these results should be interpreted with caution. To estimate the impacts of fiscal interventions on poverty, we use two poverty lines: the international line at US$5.50 per day at the 2011 revised PPP, and the EU at-risk-of-poverty (AROP) poverty line (approximately US$16.06 per day at the 2011 revised PPP), which is set at 60 percent of the median equivalized household disposable income. 4.2 Fiscal interventions For revenues, the analysis covers PIT, SIC, and indirect taxes. Within indirect taxes, we include VAT and excises on tobacco products, alcohol beverages, and oil derivatives (petrol, diesel, etc.). Carbon taxes are treated as a separate indirect tax and only the marginal impacts are assessed. Other indirect taxes and corporate income taxes are excluded due to a lack of data or because of the limitations of this methodology14. Items analyzed here account for about 63 percent of total government revenue. Table 2. Bulgaria: Revenues Fiscal Portion of fiscal Total in Accounts accounts analyzed Household survey Included Share of total Ratio survey to BGN BGN % of BGN Government % GDP fiscal accounts Million Million total Million Revenue (%) Total Revenue & Grants 42,522 100.0 38.7 28,091 66.1 26,635 62.6 Taxes 32,624 76.7 29.7 28,091 86.1 26,635 81.6 Direct Taxes 6,829 16.1 6.2 3,676 53.8 4,061 59.5 Personal Income Tax Yes 3,676 8.6 3.3 3,676 100.0 4,061 110.5 Others (Corporate Income Tax, Property No 3,153 7.4 2.9 0 - - - Tax) Indirect Taxes 16,281 38.3 14.8 14,902 91.5 11,962 73.5 VAT Yes 10,030 23.6 9.1 10,030 100.0 8,588 85.6 Excises Yes 4,909 11.5 4.5 4,872 99.2 3,374 68.7 Carbon taxesa Yes - - - - - 418 - Custom Duties No 236 0.6 0.2 0 - - - Other indirect taxes No 1,106 2.6 1.0 0 - - - Social Contributions Yes 10,614 25.0 9.7 10,614 100.0 10,613 100.0 Non-Tax Revenue No 9,898 23.3 9.0 0 - Note: a. Information on carbon tax revenues is not available in fiscal accounts. Source: Authors based on IMF (GFS Main Aggregates, GFS Revenue), OECD (Tax revenue Statistics) and Eurostat (gov_10a_main, gov_10a_ taxag). 14 The CEQ methodology does not estimate corporate income taxes and import duties incidence. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES On spending, the analysis covers social protection programs like old-age pensions, child and family 17 benefits, unemployment insurance, disability benefits, and other social programs, plus indirect subsidies (electricity and gas) and spending on health and education. In 2019 these accounted for 4. Data and Empirical Strategy for the Analysis 60 percent of total spending. One complication is that Bulgaria does not publicly report expenditure on fossil fuel subsidies, here we rely on prior work conducted by the IMF on explicit fuel subsidies from their database15. The European Commission has conducted similar work on the estimation of fuel subsidies within the EU16. Table 3. Bulgaria: Expenditures Fiscal Portion of fiscal Total in Accounts accounts analyzed Household survey Included Share of total Ratio survey to BGN BGN % of BGN Government % GDP fiscal accounts Million Million total Million Expenditure (%) Total Expenditures 40,623 100.0 37.0 24,737 60.9 24,200 59.6 Social protection 12,668 31.2 11.5 11,438 90.3 11,173 88.2 Contributory pensions Yes 7,738 19.0 7.0 7,695 99.4 7,881 74.9 Conditional & unconditional cash 4,930 12.1 4.5 3,744 75.9 3,293 66.8 transfers Family & Children Yes 1,869 4.6 1.7 1,170 62.6 1,323 70.8 Non-contributory Yes 8 0.0 0.0 8 100.0 8 101.8 pensions Disability Yes 1,313 3.2 1.2 1,058 80.5 1,034 78.7 Unemployment fund Yes 550 1.4 0.5 518 94.2 445 80.9 Other (survivors, social Yes 1,189 2.9 1.1 990 83.2 40.6 exclusion) 483 Education 3,876 9.5 3.5 3,394 87.6 3,350 86.4 Pre-school & Primary Yes 800 2.0 0.7 747 93.3 749 93.6 Secondary Yes 2,061 5.1 1.9 1,979 96.0 1,933 93.8 Tertiary Yes 689 1.7 0.6 668 96.9 666 96.7 Other n.e.c. No 326 0.8 0.3 0 - - - Health 5,449 13.4 5.0 5,063 92.9 5,063 92.9 Contributory Yes 3,930 9.7 3.6 3,930 100.0 3,930 100.0 Non-contributory Yes 1,211 3.0 1.1 1,133 93.6 1,133 93.5 Other No 308 0.8 0.3 0 - - - Housing No 1,156 2.8 1.1 0 - - - Subsidies (Total) 4,842 11.9 4.4 4,842 100.0 4,615 95.3 Electricity subsidies Yes 3,628 8.9 3.3 3,628 100.0 4,334 119.5 Natural gas subsidies Yes 1,214 3.0 1.1 1,214 100.0 281 23.2 Other expenses No 12,632 31.1 11.5 0 - - - Note: n.e.c. stands for Not Elsewhere Classified. Source: Authors based on IMF (GFS Main Aggregates, GFS Expenditure GFS Cofog), and Eurostat (gov10a_main, gov10a_tax, ESPROSS spr_ exp_eur, spr_exp_ffa, spr_exp_fol, spr_exp_fdi, spr_exp_fun). 15 The database is available at https://www.imf.org/external/np/fad/subsidies/data/subsidiestemplate.xlsx 16 State of the energy union report, EC (2021) https://ec.europa.eu/energy/sites/default/files/state_of_the_energy_union_report_2021. pdf 18 4.3 Empirical strategies and assumptions As noted previously in section 4.1 we use data from the 2018 HBS and SILC 2019 (capturing 2018 4. Data and Empirical Strategy for the Analysis income) for our analysis. Since more recent data was not available at the time, one possibility was to use statistical techniques to update the model to more recent years. Nowcasting techniques have been used in CEQs in Mexico, Argentina and Armenia (Scott, 2013; Rossignolo, 2017; younger and Khachatryan, 2014). We elected against using a nowcasting approach for several reasons (i) Detailed information on income and consumption growth by decile was not available at the time of undertaking this analysis, without detailed information by decile, researchers often resort to using GDP growth and assuming a neutral distribution which assumes everyone achieves the same level of growth independent of their position in the income distribution, an assumption which is unlikely to hold in practice; (ii) Nowcasting is particularly difficult in this context given the impacts of COVID-19 and now the impact of the war in Ukraine; and (iii) It is inadvisable to use years where countries have experienced severe shocks as representative years for fiscal analysis since they tend to be atypical and generally not representative. Given these limitations, we base our model in 2018, which reflects the distributional impacts of the fiscal system in 2018. The empirical strategy discussed below details the assumptions and limitations undertaken in this analysis. Future extensions of this work should involve nowcasting to understand the impact of the fiscal system post COVID-19 as data becomes available. › Personal Income Taxes and social security contributions For PIT, we follow the EUROMOD methodology (Tosheva et al., 2020). We estimated the gross income from employment (gross employee cash or near cash income and fringe benefits) and self-employment (gross cash benefits or losses from self-employment, including royalties) for labor income. Wage- earners with income less than the minimum wage are classified as informal and excluded from PIT calculations. Taxable income is computed by subtracting social insurance contributions and income deductions from gross income (Appendix table 10). We also deduct the normatively recognized expenditures from the tax base (NRE) for the self-employed (See Appendix table 10). For the income tax on capital income, we considered the interest, dividend, and profits from capital as taxable income. We apply the statutory rates outlined in Appendix table 9 to taxable income from labor and capital income. For social security contributions, we apply statutory rates for pension funds according to age (for persons born before 1960, and different rates plus Supplementary Mandatory Pensions Insurance Fund for persons born after 1960), General Sickness and Maternity Fund (GSMF), Unemployment fund (UF) contributions for wage-earners (See Appendix E).17 We also assume that social contributions to the pension fund, GSMF, UF, and Work Injury and Occupational Disease fund (WIDF). We also assume that both the employer and employee social contributions to the Pension Funds, GSMF, UF and WIDF are paid by employees. For WID, we applied the average rate. We assumed that all the employees belong to the 3rd category of labor and used the taxes for that category. As Tosheva et al. (2020) explained, there was not enough information to identify separate category of workers. We estimated 17 Based on prior experience with other CEQ assessments, using the reported contributions or direct taxes may generate inconsistent results, that is, contributions that are not in line with taxable income and often result in negative results for net market income. We also default to this to make the paper comparable with the methodology used in the corresponding EUROMOD paper. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES pension fund and GSMF contributions for the self-employed. For all the contributions, we consider 19 lower and upper thresholds, but when the income reported was lower than the minimum wage, we assume the workers were informal and therefore not paying social security contributions. 4. Data and Empirical Strategy for the Analysis › Pensions and social protection spending The analysis covers both the old-age contributory pensions and income from most of the social protection programs. Old-age pension income is treated as deferred income and pension contributions are treated as savings. Thus, pension income is part of the MIPP, as are pension contributions. However, we undertake sensitivity analysis in which pension income is considered government transfers and present the results in Appendix table 3 in Appendix C. The direct transfers comprise the cash benefits from social protection programs. This analysis covers disability, family & children (newborn child, maternal and parental allowances, among others), non- contributory pensions, and other cash transfers. For each program, we use the amount reported in the SILC data. However, the SILC does not provide enough information to get comparable figures with administrative accounts for education-related, housing-related, sickness benefits, survivor pensions, and other social exclusion benefits. For these programs, we included them in the aggregate category called “other direct transfers” using the amounts reported in the SILC data.18 › Indirect taxes The indirect taxes analyzed here are VAT and excises (fuel, alcoholic beverages, and tobacco). The SILC data does not contain household consumption expenditures which are the basis for estimating indirect taxes. Hence, the VAT and excises are first simulated using expenditure information in the HBS data, then imputed into the SILC using a survey-to-survey imputation method19. For VAT, we applied the statutory rates and considered the exempted items as outlined in Appendix E. In the case of excises, we estimated an implicit rate based on the product’s final price. For example, for alcoholic beverages, we converted the units to prices, using published average prices for the year by the WHO, and considering the alcoholic content. In the case of fuel, we estimated prices using the NSI figures about the total and units consumed in the year 2018. Hence, we calculated an average cost to convert the units consumed as a ratio of the final price. Additionally, we use a study published by EU on the effective rate as a proportion of the final price for the Tobacco rate, including taxes (EC, 201820). 18 The data was not sufficient to identify which transfer was part of each individual program. The most important transfers could be identified but only to include them in aggregate categories such as contributory, non-contributory, means-tested, non-means tested. Survivor pensions were identified but the total amount from the survey was different than the amount reported in administrative results. Without further information to resolve the discrepancy, we defaulted to including them in the “other” category. 19 First, a ratio of taxes to disposable income was estimated. Then, the median ratio for each percentile was calculated. Finally, a lowess smoothing was applied (Locally Weighted Scatterplot Smoothing). This procedure creates a smooth line through the ratio for each percentile. Using the smoothed ratio for each percentile, we merged this ratio with the same disposable income percentile in SILC. Finally, we multiply the ratio by the disposable income to get the indirect taxes paid. 20 European Commission (2018), “Excise duty tables: Excise duty - Tobacco products (EU harmonised)”, updated on 2019/05/14. 20 For both the VAT and fuel taxes, the exercise captures the direct effects—the amounts paid directly by households when purchasing locally manufactured and specific imported goods subject to these taxes—plus the indirect effects that taxes on oil derivates may have on product prices. The indirect 4. Data and Empirical Strategy for the Analysis effects are estimated using an input-output matrix for the Bulgarian economy for 2010 to map household consumption from the HBS to the input-output production sectors. The analysis does not consider the possibility of evasion of either VAT or excises as we do not observe place of purchase. We consider carbon taxes as a separate entity from the other indirect taxes and only assess the marginal impact since we do not have comparable macroeconomic information to validate the estimates. The estimated carbon tax incidence of each household is computed by first identifying the carbon emissions associated with each good in the HBS using the method outlined in section 3.2 Green fiscal analysis and then applying the respective carbon tax rate that prevailed in 2018 for each good. Here we differentiate between goods that are subject to carbon taxation by virtue of classifications according to the Emissions Trading Scheme and by coming from within the EU, and goods that are not using the method outlined in Appendix A. › Subsidies This report includes indirect subsidies for electricity and gas consumption. In both cases, the HBS has consumption information for 2018 to estimate them. For the electricity subsidy direct effects, the current tariffs for the second half of 2018, including all taxes, were used to obtain the consumption in kilowatt-hours per household. Then, the subsidy estimate equals the product of the kWh consumed by the difference between the price paid by households and the cost of production. The source of the production cost was the IMF’s Fuel subsidies template that has a data collection of fuel subsidies for 192 countries that cover 2015 to 2025. The input-output matrix is the source of information for indirect effects combined with the industries subsidy ratio from IMF compilation and Jellema and Inchauste’s (2018) approach to estimating indirect effects. In the case of gas, we used the subsidy ratio according to the consumer price estimated by the IMF in the computation of subsidy figures, estimating the indirect effects with the rate applied to industries and the direct effect resulting from the application of the explicit subsidy to households. › Social spending on health and education The SILC provides information on the number of children in each household attending school and at which level of education (pre-primary and primary; upper and lower secondary; and tertiary). However, there is a share of students that go to private schools. Private school enrollment rates are anticipated to be 1 percent at the primary level, 5 percent at the secondary level and 13 percent in 2018 (World Bank World Development Indicators). Because there were no other sources with information about the household profile of students in private schools, we randomly selected the private students from the 4th to 10th income deciles, assuming that the bottom 30 percent of the population does not use these services. The number of enrolled students in 2018 and Classification of the Functions of Government (COFOG) government spending comes from Eurostat data. Thus, we estimate the number of education benefits equivalent to government spending per student by level for each student. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES For health, we obtain government spending by type of services, using the COFOG classification to 21 disaggregate health spending (medical products, appliances, and equipment; outpatient services; and hospital services) between social security and the rest of the general government. For social health 4. Data and Empirical Strategy for the Analysis insurance, we include those contributing to that insurance, the under 18 population, and retirees. A per capita income for all insured populations was imputed. For the whole population, the rest of the general government spending was imputed to each Bulgarian citizen or EU national. The analysis, as well as the education, does not capture differences in the quality of the services provided. Nor does it reflect variation in the value or quality of these services across the income distribution. 4.4 Macro validation We assess the performance of the model and its assumptions by comparing the aggregate amount of each tax and transfer captured in the analysis with the same category from the official statistics (see Appendix table 4 in Appendix D for details). It is critical to evaluate whether the relative magnitude of the fiscal instruments represented by the SILC and the HBS is comparable to that in the economy. This exercise shows that the model performs relatively well. For all fiscal instruments together, the average ratio of the simulated to the actual amount is 0.59; in other words, the model captures on average 59 percent of the value of all revenue and expenditure in the official statistics. The ratio of consumption of HBS compared with National Accounts is one of the lowest in the European Union, it was higher only for Lithuania, Romania, and Slovakia. According to Eurostat21, the main reason for the difference is the difference between consumption in NA according to the domestic and national concepts. If only Social Expenditure and Subsidies is considered, the model covers 92 percent, and for personal income tax, VAT, excises, and social contributions the model covers 94 percent. The model performs particularly well for such important interventions as social insurance contributions and PIT, but less so on excises due to incomplete data. The estimations on Health and Education perform relatively better than social protection benefits and in the case of the indirect subsidies. Due to a lack of comparable data, the performance of the model on carbon taxation could not be assessed. 21 See https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Concepts_for_household_consumption_comparison_ between_micro_and_macro_approach#Relevance_and_coverage_rates_for_household_consumption 22 5. Results: Distributional impact of the fiscal system in Bulgaria 5. Results: Distributional Impact of the Fiscal System in Bulgaria 5.1 Impacts on inequality The fiscal system contributes to inequality reduction in Bulgaria. Figure 2 shows the changes in Gini coefficients from Original Income to Final income. For the PDI scenario, the original income, i.e., prior to any fiscal interventions, is the Market Income plus Pensions (MIPP) and the Gini coefficient stands at 0.451 Once all taxes, contributions, and transfers are accounted for, the Gini coefficient decreases to 0.327. Including only the effects of direct taxes, direct transfers, indirect subsidies, and indirect taxes, i.e., the net cash position, the Gini reduces from 0.451 for original income to 0.388 for consumable income22. The original income for the PGT scenario is the market income, and the effect of receiving contributory old-age pensions and paying contributions to old-age pensions makes a substantial difference as the Gini of the market income is close to 0.09 points higher without those interventions. This result is consistent with previous assessments of the fiscal system in Bulgaria, where direct taxes and transfers are limited in their redistributive impact (World Bank, 2021) and pensions are responsible for most of the redistribution within Bulgaria. For selected countries using the CEQ approach, Figure 3 shows a cross-country comparison of the decline in the Gini Coefficients from MIPP to Disposable income (Panel A), MIPP to consumable income (Panel B), and from MIPP to final income (Panel C). In this figure, we used the PDI scenario. After considering direct taxes and transfers, in 2018, the equalizing effects in Bulgaria are similar to Montenegro and Romania, and the redistributive effect is higher than in Turkey, Belarus, and Russia. When including the effects of indirect taxes and subsidies, the inequality reduction in Bulgaria is comparable to Poland but higher than in the rest of the countries in the sample. The redistributive effect when education and health are considered is almost the same as in Poland and higher than in other countries. In order to compare Bulgaria with a more significant number of EU countries, we use the results of EUROMOD, the tax-benefit model available for all EU countries. Figure 4 compares the change in MIPP inequality with that of disposable income in all EU countries. The equalizing effects of Bulgaria’s tax system are similar when the approach is based on CEQ or EUROMOD. Compared to other EU countries, Bulgaria is one of the countries that achieves the least redistribution in MIPP inequality to disposable income, that is of direct taxes and transfers other than contributory social pensions. The closest country to Bulgaria is Hungary. 22 Note that these are different income concepts from the disposable income concept used to compute the Gini as reported by Eurostat. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Figure 2. Inequality from market to final incom 23 Gini coefficient 5. Results: Distributional Impact of the Fiscal System in Bulgaria 0.60 0.543 0.55 0.50 0.45 0.410 0.451 0.388 0.40 0.327 0.35 0.30 0.25 Original income Disposable income Consumable income Final income ▬ Pensions as a government transfer ▬ Pensions as deferred income Note: Original or pre-fiscal income refers to the income before any taxes and transfers are accounted for. For the PDI scenario, the original income is Market Income plus pensions and for the PGT scenario the original income is Market Income. Source: Authors’ estimates. Figure 3. Cross-country comparison: Change in inequality Market income to Panel A:  Market income to Panel B:  Market income to final Panel C:  disposable income consumable income income Change in Gini coefficient Montenegro (2015) Montenegro (2015) Montenegro (2015) Romania (2016) Romania (2016) Romania (2016) Bulgaria (2019) Bulgaria (2019) Bulgaria (2019) Belarus (2015) Belarus (2015) Belarus (2015) Croatia (2014) Croatia (2014) Croatia (2014) Poland (2014) Poland (2014) Poland (2014) Russia (2014) Russia (2014) Russia (2014) Turkey (2016) Turkey (2016) Turkey (2016) 0 -0.02 -0.04 -0.06 -0.08 -0.10 -0.12 -0.14 Source: Authors’ estimates; other countries: World Bank CEQ repository. Figure 4. Change in inequality from MIPP to disposable income, EU averages and Bulgaria Change in Gini coefficient Netherlands Luxembourg (EUROMOD) Lithuania Denmark Germany Romania Bulgaria Bulgaria Slovenia Portugal Hungary Slovakia Belgium Sweden Czechia Finland Estonia Ireland Croatia Austria Poland Greece Cyprus France Latvia Spain Malta (CEQ) Italy 0 -0.02 -0.04 -0.06 -0.08 -0.10 -0.12 -0.14 -0.16 Source: Authors’ estimates for the CEQ-based figure for Bulgaria; EUROMOD statistics on Distribution and Decomposition of Disposable Income, accessed at https://euromod-web.jrc.ec.europa.eu/resources/statistics using EUROMOD version no. I4.0+ 24 5.2 Impacts on poverty However, how the fiscal system affects poverty differs from that of inequality depending on which 5. Results: Distributional Impact of the Fiscal System in Bulgaria poverty line is used. Poverty increases when using the AROP poverty line, i.e., the threshold of 60 percent of median equivalized disposable income (equivalent to US$16.06 per day 2011 revised PPP) and decreases when using the US$5.50 per day 2011 revised PPP poverty line. This result can be seen in Figure 5, in which poverty, according to the MIPP, is 25.4 percent but increases because of paying indirect taxes to 27.7 percent. On the contrary, if the international poverty line is used (US $5.50/day according to the 2011 revised PPP), poverty and the poverty gap are both reduced, as can be seen in Figure 5. The difference is likely due to the US$5.50 a day 2011 revised PPP being a lower poverty line than the AROP poverty line and direct taxes, social security contributions, and transfers having different incidences along different parts of the distribution. In both cases, the payment of indirect taxes reduces or eliminates the improvements in poverty resulting from direct transfers (social protection benefits). Three findings emerge: (i) Direct taxes, social security contributions, and direct transfers help reduce poverty and the poverty gap. Starting from a poverty rate of 25.4 percent, with an MIPP below the AROP threshold, the combined effects of the three interventions reduce poverty to 23.2 percent. The trend is the same using the international poverty line (Figure 5). (ii) The combined effect of indirect taxes and subsidies increases poverty for consumable income using the two poverty lines. The poverty rate jumps to 27.7 percent using the AROP poverty line and to 7.8 percent using the US$5.50 poverty line. (iii) However, from MIPP to consumable income, poverty increases only using the AROP poverty line. In other words, the reduction achieved with direct taxes, social security contributions, and direct transfers is offset by the increase caused by indirect taxes. Figure 5. Poverty and poverty gap change Panel A. Poverty, MIPP to Consumable income Panel B. Poverty gap, MIPP to Consumable Income Percent Percent 30 27.7 30 25.4 25 23.2 25 20 20 15 15 9.6 10.6 9.2 10 7.8 10 7.6 6.7 5.0 5 5 2.2 2.7 0 0 Market income Disposable Consumable Market income Disposable Consumable + pensions income income + pensions income income ▬ USD 5.5 PPP ▬ AROP poverty line ▬ USD 5.5 PPP ▬ AROP poverty line Source: Authors’ estimates. Source: Authors’ estimates. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES 6. Incidence, progressivity, and 25 marginal contributions of taxes 6. Incidence, Progressivity, and Marginal Contributions of Taxes and Social Spending and social spending This section examines the impact of the fiscal system on the welfare distribution of households. In general, an instrument (a tax or transfer) has a significant distributional impact if it targets those at the bottom of the income distribution (as captured by the concentration coefficients) and if it is relatively large compared to the recipient’s income. In addition to the distributional welfare impacts, we analyze the progressivity of each tax or transfer, and the marginal effect23 of each instrument on inequality (the redistributive effect) as measured by the Gini coefficient and poverty (the poverty reduction effect) as measured using the US$5.50 poverty line. Results using the AROP line and the PGT scenario are presented in Appendix table 2 and Appendix table 3, respectively in Appendix C. 6.1 Distributional profile of the fiscal system We first assess the distributional impacts of the tax system as a whole and then assess each instrument individually. We first divide the MIPP into deciles, the population in the first decile being the poorest and those in the tenth decile the richest. Figure 6 shows the share in the MIPP of the different components of the tax and benefit system by decile. Figure 6. Taxes and benefits by decile, MIP Percent of market income plus pensions 300 250 200 150 100 50 0 -50 -100 1 2 3 4 5 6 7 8 9 10 Market income + pensions decile J Direct transfers J Direct taxes J Indirect subsidies J Indirect taxes J Carbon taxes J Health J Education ▬ Net total position ⁝  Net cash position Notes: SIC: Social Insurance Contributions. The net total position is the difference between final income and MIPP. The net cash position is the difference between consumable income and MIPP. Source: Authors’ estimates. 23 The marginal effect is the change in MIPP by adding or subtracting only the specific fiscal intervention analyzed. 26 Most components of the fiscal system are progressive, and the poorest are net recipients of social benefits, as evidenced by their positive net cash position24. The share of each component (tax or transfer) is large for households in the bottom decile because the MIPP of the poorest individuals is 6. Incidence, Progressivity, and Marginal Contributions of Taxes and Social Spending very low. However, from the third decile on, Bulgarians are net payers into the fiscal system: direct and indirect taxes paid exceed cash benefits received. In summary, the net cash position is negative for all but the poorest 20 percent of the population. Nevertheless, when health and education are considered, the lower half of the income distribution are net recipients because the proportion of cash and in-kind benefits together exceeds tax payments. The total net position curve (the blue line in Figure 6) shows a decreasing trend, from the poorest receiving net benefits equal to 245 percent of their income to the wealthiest paying about 25 percent of their income into the fiscal system. Carbon taxes are regressive with lower income deciles paying a greater share of their MIPP. For policy and planning purposes, it is useful to determine whether a specific instrument or a combination of them is equalizing since the specific impact of one instrument may be different from that of the whole system. The analysis is based on the concepts defined in Section 3: (i) progressivity (Kakwani index); (ii) marginal contributions to inequality (redistributive effect); and (iii) marginal contributions to poverty (poverty-reducing effect). If there were only one fiscal instrument, using the Kakwani index would be enough to determine its effect on inequality. However, there are often multiple fiscal instruments, so the one-to-one relationship between the progressivity of an intervention and its effect on inequality no longer holds, as each instrument interacts with all the others in the wider system. In this case, we can use marginal contributions to inequality to examine the marginal effect of a particular tax instrument on inequality. Table 4 shows the Kakwani progressivity index for each tax and transfer alongside the magnitude of each instrument as a proportion of a household’s MIPP per capita as well as each instrument’s marginal contribution to inequality and poverty reduction in the PDI scenario (See Appendix table 3 in Appendix C for the results of the PGT scenario). We briefly discuss the marginal impacts of contributory pensions when comparing the PDI and PGT scenarios. 24 The net cash position captures the difference between MIPP and consumable income (equivalent to the net payment of direct taxes and indirect taxes, and cash benefits) as a share of MIPP. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Table 4. Progressivity and marginal contributions of fiscal instrument 27 Marginal Contributions 6. Incidence, Progressivity, and Marginal Contributions of Taxes and Social Spending Concentration Kakwani Redistributive Poverty reduction Percent of MIPP coefficien coefficien effect effect (+ = US$5.50 (+ = equalizing) poverty decline) (1) (2) (3) (4) (5) Disposable Income Direct transfers 5.8% -0.1926 0.6440 0.0344 0.0470 Child & family contributory 1.2% -0.0687 0.5201 0.0046 0.0046 Child & family means-tested 0.8% -0.4171 0.8685 0.0074 0.0108 Child & family non-contributory 0.3% -0.3464 0.7978 0.0022 0.0034 Disability 1.8% -0.1018 0.5532 0.0087 0.0118 Old-age non-contributory 0.0% -0.7250 1.1764 0.0002 0.0005 Unemployment 0.8% -0.0279 0.4792 0.0030 0.0042 Other 0.8% -0.4378 0.8891 0.0066 0.0086 Direct taxes -7.1% 0.5439 0.0984 0.0080 -0.0063 Personal Income Tax - interest -0.2% 0.9193 0.4679 0.0009 0.0000 Personal Income Tax - labor -7.0% 0.5348 0.0834 0.0079 -0.0063 income Social contributions (non-pensions) -7.4% 0.4594 -0.0462 -0.0041 -0.0065 General Sickness & Maternity Fund -0.8% 0.4563 -0.0505 -0.0004 -0.0014 employee Unemployment Fund employee -0.2% 0.4563 -0.0505 -0.0001 -0.0002 Health employee -1.7% 0.4563 -0.0505 -0.0008 -0.0023 General Sickness & Maternity Fund -1.1% 0.4563 -0.0505 -0.0006 -0.0014 employer Unemployment Fund employer -0.3% 0.4563 -0.0505 -0.0002 -0.0002 Work Injury & Occupational -0.4% 0.4563 -0.0505 -0.0002 -0.0002 Disease Fund Health employer -2.6% 0.4563 -0.0505 -0.0012 -0.0029 General Sickness & Maternity Fund 0.0000 -0.1% 0.5731 0.0701 0.0001 self-employed Health self-employed -0.2% 0.5731 0.0701 0.0002 0.0000 Consumable Income Indirect subsidies 8.1% 0.3424 0.1090 0.0046 0.0171 Electricity direct effects 3.1% 0.2385 0.2129 0.0059 0.0102 Electricity indirect effects 4.6% 0.4041 0.0473 -0.0014 0.0082 Gas direct effects 0.1% 0.4650 -0.0136 -0.0001 0.0000 Gas indirect effects 0.4% 0.4041 0.0473 -0.0001 0.0006 Indirect taxes -21.0% 0.4607 0.0094 0.0172 -0.0255 VAT -15.1% 0.4620 0.0106 0.0133 -0.0189 Oil derivates direct effects -1.5% 0.5196 0.0682 0.0026 -0.0011 Oil derivates indirect effects -2.0% 0.4386 -0.0128 0.0015 -0.0021 Excise Alcoholic Beverages -0.1% 0.4363 -0.0151 0.0001 -0.0001 Excise Tobacco -2.3% 0.4361 -0.0152 0.0016 -0.0024 Carbon emissions tax -0.8% 0.3636 -0.0877 0.0002 -0.0014 28 Table 4. Progressivity and marginal contributions of fiscal instruments (continued Marginal Contributions 6. Incidence, Progressivity, and Marginal Contributions of Taxes and Social Spending Concentration Kakwani Redistributive Poverty reduction Percent of MIPP coefficien coefficien effect effect (+ = US$5.50 (+ = equalizing) poverty decline) (1) (2) (3) (4) (5) Final Income In-kind (health + education) 14.8% -0.0164 0.4677 0.0610 Health 8.9% 0.0307 0.4206 0.0318 Social Security 6.9% 0.0396 0.4118 0.0235 Public health 2.0% 0.0000 0.4514 0.0072 Education 5.9% -0.0877 0.5390 0.0241 Pre-school 0.6% -0.1036 0.5550 0.0028 Primary level 0.7% -0.1752 0.6266 0.0038 Lower Secondary 1.4% -0.2483 0.6996 0.0077 Upper Secondary 2.0% - 6 0.163 0.6150 0.0089 Tertiary 1.2% 0.2979 0.1534 -0.0003 Notes: MIPP: Market income plus pensions. As is customary, the poverty reduction effects of in-kind benefits are not considered. Source: Authors’ estimates. 6.2 Direct taxes, indirect taxes, and social insurance contributions › Direct taxes Direct taxes are progressive, as evidenced by the positive Kakwani coefficient, and have equalizing effects but have limited impacts on poverty. PIT on interest is more progressive than PIT on labor income with Kakwani coefficients of 0.4679 and 0.0834, respectively (Table 4, column 2) The PIT on interest is also highly concentrated among the wealthier as evidenced by the concentration coefficient being close to 1. Relatively wealthier individuals pay a larger share of the total amount collected: 57 percent comes from individuals in the top quintile compared with only 3 from individuals in the bottom quintile (Figure 7). Direct taxes reduce inequality by 0.008 Gini points (Table 4, column 3). However, their impact on poverty is limited, with an impact of less than one percentage point increase (Table 4, column 5). Moreover, the poorest pay few direct taxes. Compared to the PIT on labor, the PIT on interest has negligible redistributive and poverty effects, in part because they represent only 0.2 percent of the MIPP compared to 7.0 percent of the PIT on labour (Table 4, column 1). THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Figure 7. Taxes and social insurance contributions by income quintile 29 6. Incidence, Progressivity, and Marginal Contributions of Taxes and Social Spending Indirect taxes Carbon tax Excises VAT All Direct taxes & contributions Pension SIC Non-pension SIC PIT & interest tax All 0 10 20 30 40 50 60 70 80 90 100 Percent of revenue before fiscal interventions J Poorest J Q2 J Q3 J Q4 J Richest Notes: PIT: Personal income tax; SIC: Social insurance contributions; VAT: Value-added tax. Source: Authors’ calculations. › Indirect taxes Indirect taxes are proportional and contribute to significant increases in poverty. However, even though the overall concentration of those taxes is proportional, the fiscal incidence of the first decile is high compared to their MIPP. The reason for this is that because these households have a relatively low MIPP, some are entitled to social protection benefits, which are then used to purchase goods and services that are taxed at tax rates that are relatively high compared with their (low) MIPP. Most of the observed impact comes from VAT, which is 15.1 percent of the MIPP. Both the VAT and the direct effects of oil taxes have positive but close to zero Kakwani coefficients (Table 4, column 2) implying that they are weakly progressive. These measures also reduce inequality by 0.013 and 0.003 Gini points, respectively (Table 4, column 4). On the contrary, indirect effects from Oil and Tax Excises from Alcoholic Beverages, and Tobacco have Kakwani coefficients negative and close to zero, suggesting that they are weakly regressive (Table 4, column 2). The poverty-increasing effect of VAT is significant accounting for a 1.9 percentage point increase. Taken together, VAT and excise taxes increase poverty by 2.6 percentage points (Table 4, column 5). In other words, after paying indirect taxes, an additional 2.6 percent of Bulgarian citizens became poor. Carbon taxes are regressive as evidenced by a negative Kakwani coefficient. Along the income distribution, the top decile accounts for 44 percent of revenues collected compared with only 7 percent for the poorest decile. Carbon taxes do not contribute significantly to inequality reduction and have negligible impacts on poverty. This is likely due to these taxes being a relatively small proportion of MIPP (0.8 percent). › Social insurance contributions Social security contributions (excluding contributions to pensions) are regressive overall, as evidenced  see Table 4, column 3). The reason for this regressivity is that there by the negative Kakwani coefficient ( is an upper threshold for contributory income (See Appendix E). This means that social security taxes are only levied on income up to a certain threshold, hence the relationship between income and social security contributions is non-monotonic. However, in the case of the self-employed, contributions are progressive because this group is dominated by higher-income contributors (top 3 deciles). At 30 the same time, the proportion of individuals in lower-income deciles is higher for employees, and the top income groups are paying less than the self-employed. Therefore, employee and employer contributions are regressive. Because employee contributions collection in the aggregate is higher 6. Incidence, Progressivity, and Marginal Contributions of Taxes and Social Spending than for own-account workers, the regressive effect prevails and social contributions are regressive overall. Contributions to the various funds do not significantly impact poverty and inequality (see Table 4, columns 4 and 5). 6.3 Direct transfers, subsidies, and In-kind transfers › Direct transfers All direct transfers are progressive to varying degrees, as evidenced by the positive Kakwani coefficients, and contribute to poverty and inequality reduction. Disability benefits and Child and family contributory benefits account for the largest share of household income (Table 4, Column 1), with other individual programs contributing less than 1.0 percent. Direct transfers disproportionately accrue to the poor: households in the bottom quintile receive 35 percent of direct transfers compared with, only 15 percent for the top quintile. Child & Family non-contributory, and other social security benefits have 50 percent or more of the total benefits accruing to the first quintile. Nevertheless, the distribution varies by program (Figure 8). Non-contributory benefits, other benefits and means-tested benefits are the most progressive, whereas the unemployment benefit is the least progressive. Despite this variation, the concentration coefficients are all negative (Table 4, Column 2), indicating that the benefits are concentrated among relatively poorer individuals. Taken together, direct transfers reduce inequality by 0.035 Gini points (Table 4, Column 4), and poverty by 4.6 pp (Table 4, Column 5). Figure 8. Direct cash transfers by income quintile Total Child & family contributory Child & family means-tested Child & family non-contributory Disability Old age non- contributory Unemployment Other 0 10 20 30 40 50 60 70 80 90 100 Percent of total J Poorest J Q2 J Q3 J Q4 J Richest Note: MIPP: Market income plus pensions. Source: Authors’ calculations. Disability benefits (1.2 pp), means-tested child and family benefits (1.1 pp), and other transfers (0.9 pp) contribute to reductions in poverty but with relatively low impacts (Table 4, Column 5). Similarly, these benefits contribute to reductions in inequality, but the impacts are relatively low—disability benefits (0.0087), means-tested child and family benefits (0.0074) and other transfers, which includes the lump-sum monthly social assistance program (0.0066) (Table 4, Column 5). THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES The limited role played by direct taxes and transfers in redistribution, and the limited impact of the 31 individual programs on poverty reduction corroborates previous findings by the World Bank  (World Bank, 2021). The prior findings pointed to the limited role of the lump-sum monthly social assistance 6. Incidence, Progressivity, and Marginal Contributions of Taxes and Social Spending allowance, Bulgaria’s last resort poverty reduction program, in poverty and inequality reduction due to issues with inadequate coverage, targeting, and generosity of benefits (World Bank, 2021). › Indirect subsidies Indirect subsidies as a group are progressive and contribute to reductions in the Gini coefficient (0.0046) and poverty (1.7 pp). The electricity subsidy received directly by households is the most progressive whereas the direct natural gas subsidy is weakly regressive. The direct electricity subsidy accounts for 3.1 percent of the MIPP, reduces the Gini coefficient by 0.006 points, and poverty by 1.0 percentage point. By comparison, the direct natural gas subsidy accounts for 0.1 percent of the MIPP, has negligible impacts on the Gini coefficient, and poverty. For gas subsidies and indirect electricity subsidies, close to half of the benefits accrue to the top income quintile and less than 7 percent to the poorest income quintile. For the direct electricity subsidy, 34 percent of benefits go to the top income quintile and 10 percent go to the bottom income quintile. Figure 9. Indirect subsidies by income quintile Gas indirect effects Gas direct effects Electricity indirect effects Electricity direct effects All 0 10 20 30 40 50 60 70 80 90 100 Percent of total J Poorest J Q2 J Q3 J Q4 J Richest Source: Authors’ estimates. › In-kind benefits Public expenditure on health and education is progressive and equalizing. Public health and education benefits together account for 14.8 percent of MIPP, the largest contribution of any group of programs. A positive Kakwani coefficient indicates that these in-kind benefits are progressive. In-kind benefits also help reduce inequality by an estimated 0.0610 Gini points. However, because disposable income, used for poverty estimates, does not include in-kind benefits, their poverty-reducing effect is not assessed. Health spending, which constitutes 8.9 percent of household income, contributes to a reduction of 0.032 Gini points. Within health, Public health spending is progressive and leveling, thus, by construction, the Kakwani index equals the Gini coefficient for MIPP. Education spending accounts for 5.9 percent of household income and reduces the Gini coefficient by 0.024 points. Figure 10 shows 32 that education spending is relatively equally distributed across quintiles, implying widespread use of public education services in Bulgaria. Although the Kakwani index is positive across education levels, spending on pre-primary, primary, secondary, and post-secondary education is more progressive 6. Incidence, Progressivity, and Marginal Contributions of Taxes and Social Spending than on tertiary education. The concentration ratios of spending on all education levels except tertiary are negative, suggesting that a larger share of these benefits goes to students at the bottom of the income distribution (Figure 10). Figure 10. Concentration of In-kind transfers Tertiary Upper secondary Education Lower secondary Primary level Pre-school All Public health Health Social security All In-kind - All 0 10 20 30 40 50 60 70 80 90 100 Percent of total J Poorest J Q2 J Q3 J Q4 J Richest Source: Authors’ estimates. › Comparison to PGT scenario The most striking difference between the PDI and PGT scenarios is the redistributive and poverty- reduction role of pensions. Under the PGT scenario, pensions have the largest poverty reduction impact accounting for a 12.8 pp decline, and the largest impact on inequality with a 0.101 decline the Gini coefficient. To put this in context, after contributory pensions, the greatest poverty reduction impact of any one program is 1.3 percentage points (child and family means-tested benefits), and 0.0318 Gini points (in-kind health transfers). Furthermore, the poverty- reduction impact of contributory pensions is 3.4 times higher than the combined impact of direct transfers. The strong poverty and inequality reducing impact of pensions stems in part from pensions representing a relatively large proportion of income at 14.2 percent, exceeded only by VAT (15.1 percent). The finding that contributory old-age pensions are driving most of the redistributive and poverty- reducing impact is consistent with previous findings (EUROMOD, 2021; World Bank, 2021). Bulgaria has one of the oldest populations in the world driven by increased life expectancy, high rates of outward migration among the working-age population and declining fertility rates (World Bank, 2021). Bulgaria is on track for its population to age further with significant declines among the 25–49 age group projected to occur by 2070 (World Bank, 2021). As the population aging process intensifies, the role of old-age pensions in poverty and inequality reduction will become increasingly compromised as the contributory base shrinks and the number of pensioners increases. Bulgaria has taken steps to increase the sustainability of pensions by steadily increasing the retirement age and the minimum period of contributions before which individuals can be eligible for pensions. However, in the context of such rapid population aging, these measures in isolation are likely to be insufficient. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES 7. Welfare impacts of green fiscal 33 policies 7. Welfare Impacts of Green fiscal policies 7.1 Current and planned efforts to reduce carbon dioxide emissions in Bulgaria Bulgaria has the most carbon-intensive economy in Europe, over three times as high as the EU- average (World Bank, 2021). The high carbon intensity is attributed to the dominance of fossil fuels in Bulgaria’s primary energy supply (29 percent coal, 23 percent oil, 14 percent natural gas), and Bulgaria having the highest energy intensity among EU member states. While overall greenhouse gas emissions have stabilized at 60 percent of their 1990 levels since 2009, this is insufficient to meet commitments of reducing emissions by least 55 percent relative to 1990 levels by 2030 as agreed upon in the European Green Deal (World Bank, 2021). In the sections that follow, we discuss some potential mechanisms to reduce Bulgaria’s carbon emissions in line with the terms of the European Green Deal. › Emissions trading scheme: An international carbon market Bulgaria, as a member of the EU, is a signatory to the EU’s Emission Trading Scheme (ETS), the primary mechanism in place to reduce carbon dioxide emissions. The EU ETS is a ‘cap and trade’ system whereby a cap is put on the total amount of desirable emissions and within the confines of that cap, entities buy or receive emission allowances, which they can trade in a market, where the total number of allowances is fixed to maintain value. If at the end of a year, an entity has surplus allowances relative to its emissions, it can keep them for future use or it can opt to sell them to the market. If, however, an entity has a shortage of allowances relative to its emissions, heavy fines are imposed. In 2018, the price of carbon was set at 19.32 euros, and through the price of allowances and fines, the ETS generated 18,817.9 million euros in revenue. To date, the EU ETS has been an effective tool, reducing emissions from ETS sectors by about 35 percent between 2005, when it was first implemented, and 2019. Currently, the ETS considers the most carbon intensive sectors, with plans for expansion to other sectors.25 › Carbon border adjustment mechanism: Combatting carbon leakage One of the major concerns and critiques with the current ETS system is the potential for carbon leakage, that is, EU-based companies could shift production processes abroad to avoid paying for carbon  emissions by taking advantage of disparities in carbon policies across jurisdictions, or, consumers could substitute away from EU products to relatively cheaper carbon-intensive imports as imports outside of the EU are not currently subject to carbon-pricing policies. The Carbon Border Adjustment Mechanism (CBAM) was designed to reduce the propensity for carbon leakage by requiring importers to purchase carbon certificates corresponding to the carbon price that would have been paid, had the 25 https://ec.europa.eu/clima/eu-action/eu-emissions-trading-system-eu-ets_en 34 goods been produced under the EU’s carbon pricing policies26. The goal of the CBAM is to equalize the price of carbon within and outside of the EU so as not to undermine competitiveness of EU companies or to encourage the shifting of emissions from the EU to other countries, as well as to encourage the 7. Welfare Impacts of Green fiscal policies greening of production processes abroad and generate revenues for the government. The system is intended to be phased in with importers required to report the emissions embedded in their products starting from 2023 and having to pay a financial adjustment in 2026 with importers paying the same tax for carbon as domestic producers under the EU ETS. Initially, the CBAM will only apply to the sectors covered by the ETS with later plans for expansion. › Fossil-fuel subsidies In addition to the imposition of carbon taxes, the phasing out of fossil fuel subsidies is an integral part of the green transition and part of Bulgaria’s commitment under the terms of the European Green Deal. The use of fossil fuel subsidies has been critiqued as inefficient for several reasons. First, fossil fuels subsidies distort the price of carbon-intensive fuels by artificially lowering their price, leading to a greater than socially optimal amount consumed, particularly since substantial negative externalities associated with their use are often not captured in the price. Second, fossil fuel subsidies distort incentives to switch to cleaner forms of energy by keeping the price of carbon- intensive fossil fuels low relative to cleaner fuels with less externalities. Third, expenditure on fossil fuels often constitutes a significant amount of government expenditure, money which could be put to more productive uses such as expenditure on health and education. Though official estimates of expenditures on fossil-fuel subsidies for Bulgaria are not publicly reported, efforts have been made to estimate these. Most recently, the IMF has estimated the amount Bulgaria spends on subsidies, differentiating between explicit and implicit energy subsidies. Explicit subsidies occur when the price paid by firms or households is below the cost of supplying energy (IMF, 2021). Implicit subsidies occur when the retail price fails to include externalities associated with environmental damage and/or there are preferential consumption tax rates on energy (IMF, 2021). The so called "price-gap" approach used by the IMF and in this paper should be considered a lower bound as it does not capture many producer subsidies. Further limitations of the price-gap approach are detailed elsewhere (World Bank, 2015). By contrast, the OECD uses an "inventory" approach which constructs an inventory of government actions benefiting the production and consumption of fossil fuels. As Bulgaria was not a member of the OECD at the time of writing, comparable estimates using this methodology were not available. Future work should explore the sensitivity of these results to alternate methodologies used to compute fossil fuel subsidies and the results of this analysis should be considered in the context of these limitations. 7.2 Welfare impacts of proposed carbon taxation policies In this section we estimate the distributional impact and impacts on poverty and inequality of potential carbon policies, including the two mechanisms discussed in the previous two sections 26 https://ec.europa.eu/commission/presscorner/detail/en/qanda_21_3661 THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES in conjunction with the removal of fossil fuel subsidies. It should be noted that these simulations 35 should be interpreted as immediate short-term effects on the welfare of households as a result of a change in prices. Because we assume that these are short-term immediate impacts, we do not 7. Welfare Impacts of Green fiscal policies account for behavioural responses like consumption switching across similar goods, cost-saving efforts by consumers, or reductions in the demand for goods in response to an increase in prices. Due to data limitations, we also do not model the welfare impacts due to potential changes in wages or employment from the firm side as a result of increased carbon taxes. Additionally, for all simulations, we assume that there is a full pass-through of the carbon taxes from producers to consumers and abstract away from tax evasion. The results should be interpreted in this context of these stipulations. We consider two scenarios, a conservative low scenario and a very ambitious high scenario, to determine the distributional impact and impacts on poverty and inequality of potential carbon policies. Each scenario is guided by: (i) the current mechanisms in place to reduce emissions (ETS), (ii) the planned mechanisms in place to reduce emissions (CBAM), (iii) necessary adjustments to the price of carbon including the removal of fossil fuel subsidies and the restructuring of carbon taxation required over time in order for Bulgaria to meet its decarbonization goals. The latter is guided by estimates of the structure of fossil fuel subsidies that prevailed in 2018 as estimated by the IMF and the necessary carbon adjustments necessary for Bulgaria to meet its decarbonization goals as estimated in previous work conducted by the World Bank (World Bank, 2022). Since the EU ETS sector already had a carbon price of EUR 19.32 in 2018, any changes to that sector will be increases over that price. For the purposes of this analysis, we assume the ETS sectors based on the CPA industry codes are Mining and quarrying; Paper products and printing; Coke and refined petroleum product; Chemicals and chemical products; Other non-metallic mineral products; Basic metals; Electricity, gas, steam and air conditioning supply; and Air transport. Scenario 1 (Low): Introducing carbon taxation to non-ETS sectors within the EU, increasing the price of carbon in ETS sectors within the EU › EUR 10/tCO2 in non-ETS sectors within the EU › EUR 45/ tCO2 in ETS sectors within the EU › Removal of explicit natural gas subsidies Scenario 2 (High):  Equalizing and increasing carbon taxation across non-ETS sectors ETS sectors within and outside of the EU › EUR 90/tCO2 in ETS and non-ETS sectors within and outside the EU › Removal of explicit natural gas subsidies The increase in carbon tax based on the two scenarios changes the marginal impact of the carbon tax in the fiscal system under the high scenario whereas the marginal impact under the low scenario remains largely comparable to the original carbon tax. For ease of comparability, we reproduce the marginal impacts of the carbon tax from Table 4 above (original carbon tax) and add the marginal impacts under the low and high scenario in Table 5 below. Comparing the original scenario with the high scenario, the carbon tax becomes less regressive, has more equalizing impacts, but also has a greater poverty increasing impact. The change in the marginal impact of the carbon tax across these two scenarios is in part driven by carbon taxes under the high scenario accounting for 3.4 percent of MIPP, more than 4 times higher than that of the original carbon tax. 36 Table 5.  Progressivity and marginal contributions of original, low and high carbon taxation scenarios 7. Welfare Impacts of Green fiscal policies Poverty reduction effect Percent of Concentration Kakwani Redistributive effect Carbon tax (+ = US$5.50 poverty MIPP coefficien coefficien (+ = equalizing) decline) Original -0.8% 0.3636 -0.0877 0.0002 -0.0014 Low -1.1% 0.3655 -0.0859 0.0002 -0.0014 High -3.4% 0.3722 -0.0791 0.0004 -0.0049 Source: Authors’ computations. We now turn to the poverty and inequality impacts of changing carbon taxes and removing fossil fuel subsidies according to the two scenarios within the complete fiscal system which may differ from the marginal impacts shown in Table 5 due to interactions with other instruments. Since carbon taxes are an indirect tax, we consider changes in poverty and inequality using the consumable income concept. Poverty in the EU is measured using the at-risk-of-poverty (AROP) concept where an individual is considered poor if they have equivalized disposable income (after social transfers) below 60 percent of the national median equivalized disposable income (after social transfers) in that year. For the purposes of this analysis, we consider impacts using three poverty lines: (i) the fixed at-risk-of-poverty line (AROP-fixed) where the poverty line is fixed under the baseline scenario (prior to any carbon taxes or revenue recycling measures), (ii) the flexible at-risk-of-poverty line (AROP-flexible) where the line is allowed to vary across scenarios based on changes to the disposable median income as a result of the carbon tax and revenue recycling measures and (iii) the internationally comparable US$5.50 per day poverty line (US$5.50). With the imposition of additional carbon taxes and the removal of explicit natural gas subsidies, poverty increases from 27.7 percent at baseline for both the AROP-fixed and AROP-flexible poverty line to 28.7 percent under the low scenario and 30.5 percent under the high scenario (Table 6). Estimates using the US$5.50 per day poverty line shows a similar increase, poverty increases from 7.8 percent in baseline to 8.0 percent under the low scenario and 8.4 percent under the high scenario with the imposition of carbon taxes and the removal of natural gas subsidies (see Table 6. Poverty and Inequality at baseline and under the low and high scenarios). Inequality, as measured by the Gini coefficient, remained largely unchanged across scenarios. Table 6.  Poverty and inequality at baseline and under the low and high scenarios Poverty line Baseline Low High AROP-fixed 27.7 28.7 30.5 AROP-flexible 27.7 28.7 30.5 US$5.50 7.8 8.0 8.4 Gini 0.3882 0.3882 0.3935 Source: Authors’ computations. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES 7.3 Revenue recycling measures 37 Recycling revenues from carbon taxation can help to offset distributional concerns as a result of 7. Welfare Impacts of Green fiscal policies the imposition of carbon taxes and the removal of fossil fuel subsidies. While there is largely a consensus that efforts should be made to protect the most vulnerable from the adverse impacts of carbon taxation, there is less of a consensus on how best to do so with the choice largely dependent on the individual circumstances of each country and its governments priorities (World Bank, 2022). Among measures that have been considered or implemented are transfers to households (Canada), reductions in income taxes (Ireland), reductions in employee social contributions (Nordic countries), increases in public spending (Portugal), investment in clean technology, and transition support to industry (World Bank, 2022). We only consider taxes collected from indirect taxation on household consumption based on the CO2 content of each good for our revenue recycling measures and the results should be interpreted in this context. Under the low scenario, and assuming full compliance and an efficient revenue collection system, it is estimated that the Bulgarian government would collect just under 512 million euros, 69.7 percent of which would come from the imposition of higher carbon taxes and 30.3 percent which would come from the removal of explicit natural gas subsidies. The comparable figures under the high scenario are 1219 million euros, 87.3 percent of which would come from the imposition of higher carbon taxes and 12.7 percent of which would come from the removal of explicit gas subsidies. It should again be stressed that these revenue amounts are specific to 2018. Here, we take an empirical approach and analyze the distributional, poverty and inequality impacts of several potential revenue recycling measures for the revenue generated for the high and low scenarios. It should again be noted here that the impacts of these revenue recycling measures do not account for any behavioural impacts in response to the policies on the part of households of firms such as labor supply or labor demand responses in response to changes in PIT tax rates, or changes in consumption patterns as a result of changes to VAT rates, and the results should be interpreted in that context. We consider five measures discussed below which all fully exhaust the tax revenues: › Uniform transfer to all households: All households receive a lump-sum transfer equal to the total amount of additional resources resulting from the emissions tax and the elimination of the natural gas subsidy divided by the total number of households. The lump-sum amounts to 162 euros in the low scenario to 386 in the high scenario. › Targeted lump-sum transfer to the bottom 20 percent of households: All households in the bottom 20 percent of the income distribution receive a lump-sum transfer equal to the total amount of additional resources resulting from the emissions tax and the elimination of the natural gas subsidy divided by the total number of households. The lump-sum amounts to 1,014 euros in the low scenario and 2,413 euros in the high scenario. › Social protection benefits proportional increase: In this case, the benefits received for family, disability, old age, unemployment, and other benefits identified in the survey increase proportionally. The benefits increment range goes from an additional 28.1 percent in the low tax scenario to 66.9 percent in the high scenario. › Reduction in the VAT rate: The VAT tax reduces from 20 percent to 17.84 percent under the low scenario and 20 percent to 14.86 percent under the high scenario. 38 › Personal income tax rate rebate: T  he PIT rate is lowered from 10 percent to 7.66 percent under the low scenario and 10 percent to 4.43 percent under the high scenario. 7. Welfare Impacts of Green fiscal policies Results under the fixed AROP line under the low scenario suggests that the impacts of the carbon tax can be fully mitigated and reductions over baseline poverty can be realized to varying degrees under the universal lump-sum transfer, the targeted lump-sum transfer and the social protection increase scenarios (See Table 7). The targeted lump-sum transfer results in the greatest reduction over baseline poverty (2.5 percentage points), followed by the universal lump-sum transfer (0.8 percentage points) and then the social protection increase (0.6 percentage points). Since the targeted lump-sum transfer directs the most resources to poorer households it dominates the other two scenarios. By contrast, VAT reductions and PIT rebates only partly offset the increase in poverty due to the imposition of carbon taxes. VAT reductions slightly dominate PIT rebates as VAT reductions would impact a wider share of poorer households rather than only those with members with formal workforce attachment (Table 7). It should again be noted here that these results do not account for labor supply responses which may bring some added poverty reduction benefits under the PIT rebate scenario. Table 7. Poverty and inequality under various revenue recycling scenarios Universal Targeted Social VAT Poverty line and scenario Baseline Carbon tax lump-sum lump-sum protection PIT rebate reduction transfer transfer increase Low AROP-fixed 27.7 28.7 26.9 25.2 27.1 27.9 28.2 AROP-flexible 27.7 28.7 27.7 25.2 28.2 27.9 29.1 US$5.50 7.8 8.0 7.4 4.9 6.8 7.8 7.9 Gini 0.3882 0.3882 0.3811 0.3654 0.3810 0.3899 0.3908 Reduction in poverty gap per Euro - - 0.05 0.21 0.06 0.01 0.01 (US$5.50) High AROP-fixed 27.7 30.5 25.4 19.6 26.2 28.2 28.9 AROP-flexible 27.7 30.5 27.8 19.8 28.8 28.2 30.5 US$5.50 7.8 8.4 6.6 1.9 5.9 7.9 8.2 Gini 0.3882 0.3883 0.3720 0.3424 0.3735 0.3923 0.4000 Reduction in poverty gap per Euro - - 0.05 0.14 0.06 0.01 0.01 (US$5.50) Source: Authors’ computations. Results using the US$5.50 poverty line are largely consistent with the fixed AROP poverty line, except VAT reductions fully offset the impact of the carbon taxes and the poverty impacts are greater given that the US$5.50 is a lower poverty line. Under the flexible AROP poverty line, where the line varies with changes to the median income, the impact of the carbon tax is fully offset from the universal lump- sum transfer and reductions in baseline poverty are realized from the targeted lump-sum transfer. By contrast, social protection increases, reductions in VAT and the PIT rebate are insufficient to offset the increase in poverty due to the carbon taxes. Results across all 3 poverty lines under the high THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES scenario largely mirror that of the low scenario, though the magnitudes differ. Across all measures, 39 the targeted lump-sum transfer is the most cost-effective, resulting in the largest reduction in the poverty gap per Euro spent. 7. Welfare Impacts of Green fiscal policies Though inequality remained largely unchanged with the imposition of the carbon tax, the universal lump-sum transfer, the targeted lump-sum transfer and the increase in social protection all led to declines over baseline inequality, with the targeted lump-sum transfer having the greatest redistributive impact. By contrast, the VAT reduction and the PIT rebate scenarios led to increases over baseline inequality, suggesting that these measures disproportionately benefit relatively wealthier groups. 40 8. Conclusions and policy insights 8. Conclusions and Policy Insights This paper examined the fiscal incidence of the tax-benefit system in Bulgaria and its contribution to poverty and inequality reduction. The analysis reveals that when measuring poverty using the AROP poverty measure, the fiscal system is poverty reinforcing. Although Bulgaria’s fiscal system is inequality- reducing overall, the impact of direct taxes and transfers and social contributions is among the least redistributive in the EU. Most of the redistribution and poverty reducing impacts of the Bulgarian fiscal system are coming through the role of contributory pensions. This is particularly concerning as Bulgaria is a rapidly aging society and the sustainability of pensions as a source of poverty reduction and fiscal redistributor is increasingly being undermined. Direct taxes and transfers, including means- tested child and family benefits, play a limited role in poverty reduction. Complementary analysis conducted by the World Bank has pointed specifically to gaps in the coverage, targeting and adequacy of the social assistance (non-contributory) system as the main source of the limited role for the fiscal system in redistribution and poverty reduction (World Bank, 2021). The limited redistributive impact and the poverty-reinforcing impact of the fiscal system may help to explain the persistently high levels of poverty and inequality prevalent in Bulgaria. Taken together, these facts signal the need for a reform of the fiscal system that is more conducive to the goals of reducing poverty and inequality. Direct taxes and transfers, particularly, means-tested benefits, should play a greater role in poverty and inequality reduction, particularly in the context of reduced sustainability of the pension system. In line with prior suggestions made by the World Bank, Bulgaria needs to address low targeting, adequacy, and generosity of the social assistance system, low levels of expenditure on periodic means-tested cash benefits, misalignment of coefficients determining benefit amount and eligibility with the characteristics of those most at risk of poverty, and a lack of systemic indexation of key eligibility parameters such as the GMI. Assessing the welfare impact of green fiscal policies shows that poorer households would have higher carbon tax expenditures as a proportion of their income as compared with wealthier households, though wealthier households would pay a larger share of carbon taxes. The assessment also showed that in the absence of any mitigating policies, poverty is expected to increase though inequality would remain largely unchanged. Revenue recycling measures could mitigate the impacts of the increase in poverty, though, it mattered through which fiscal channel revenues were recycled and the structure of the policy. Targeted lump-sum transfers to poorer households are simulated to have the greatest impacts on poverty and inequality, surpassing an expansion in the existing social protection system, likely due to the issues mentioned above. By contrast, VAT cuts and PIT rebates were insufficient to offset increases in poverty due to carbon taxation and had inequality-reinforcing impacts. As Bulgaria will likely need to increase the level of carbon taxation to meet its decarbonization goals and commitments under the European Green Deal, an analysis of the welfare impacts of proposed carbon taxation reform and the effectiveness of various revenue recycling measures in mitigating welfare impacts, should be part of the consideration on how best to structure green fiscal policy. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES The analysis of the Bulgarian fiscal system carried out in this paper underscores the importance of 41 governments conducting a periodic assessment of their countries’ fiscal systems to ensure that the 8. Conclusions and Policy Insights structure and performance of the fiscal system is aligned with the country’s social and economic goals. The specific application to green fiscal analysis underscores the importance of having a distributional assessment as a standard part of the broader analysis when governments are considering fiscal reforms in order to quantify the impact on vulnerable populations and to assist in the design of appropriate mitigating policies. Amidst the ongoing food and energy crisis following on from the war in Ukraine, the Bulgarian government has implemented a raft of fiscal measures aimed at protecting Bulgarians from the rising cost of living. Among the measures instituted are freezing power and heating prices for households and providing electricity and gas subsidies for businesses. Other measures being considered are reduced VAT on basic food items and reduced fuel excises. Future work should analyze the distributional, fiscal and welfare impacts of the current social protection package as well as potential alternative policies. Another impact of the war has been to compromise the energy security of Bulgaria, forcing Bulgaria to rely more heavily on domestic energy sources and to look for alternate external energy suppliers. With the impacts of the war still unfolding, it is too early to determine what impact, if any, this might have on Bulgaria's decarbonization trajectory. As more becomes evident, the green fiscal analysis undertaken in this paper should be updated to reflect Bulgaria's new reality. 42 Appendix A. Technical Appendix Appendix Total emissions embodied in final domestic demand per euro for industry j is computed as: ET,j EBG R, j + EROE, j + EROW , j = FDT,j FDBG R, j + FDROE, j + FDROW , j where: Ex,j are millions of tonnes of CO2 emissios embodied in domestic final demand for industry j in region x FDx,j is final demand in millions of euros for industry j in region x Thus, the share of emissions embodied in final domestic demand per euro for each region x for industry j is computed as: 1 sBGR,j = FDROE, j + FDROW , j 1+ ( FDBG R, j ) 1 sROE ,j = FDBG R, j + FDROW , j 1+ ( FDROE, j ) 1 sROW,j = FDBG R, j + FDROE, j 1+ ( FDROW , j ) Hence, we can derive the share of emissions per euro of final domestic demand attributable to Bulgaria, the Rest of the EU and the rest of the World. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Appendix Figure 1. Diagrammatic Representation of Progressivity of Transfers 43 Appendix Source: Lustig et. al, 2018 44 Appendix B. Emissions per euro of COICOP item Appendix Appendix table 1. Emissions per euro of COICOP item. Emissions per euro of COICOP item Domestic and COICOP category imported emissions kt CO2 per Euro Food 0.08 Non-alcoholic beverages 0.03 Alcoholic beverages 0.04 Tobacco 0.04 Clothing 0.03 Footwear 0.03 Actual rentals for housing 0.01 Maintenance and repair of the dwelling 0.17 Water supply and miscellaneous services relating to the 0.07 dwelling Electricity, gas and other fuels 7.50 Furniture and furnishings, carpets and other floor coverings 0.21 Household textiles 0.03 Household appliances 0.02 Glassware, tableware and household utensils 1.35 Tools and equipment for house and garden 0.02 Goods and services for routine household maintenance 0.54 Medical products, appliances and equipment 0.19 Out-patient services 0.01 Hospital services 0.01 Purchase of vehicles 0.01 Operation of personal transport equipment 0.23 Transport services 1.00 Postal services 0.13 Telephone and telefax equipment 0.02 Telephone and telefax services 0.02 Audio-visual, photographic and information processing 0.02 equipment Other major durables for recreation and culture 0.08 Other recreational items and equipment, gardens and pets 0.28 Recreational and cultural services 0.01 Newspapers, books and stationery 0.07 Package holidays 0.07 Pre-primary and primary education 0.01 Secondary education 0.01 THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Appendix table 1. Emissions per euro of COICOP item. Emissions per 45 euro of COICOP item (continued) Appendix Domestic and COICOP category imported emissions kt CO2 per Euro Post-secondary non-tertiary education 0.01 Tertiary education 0.01 Education not definable by level 0.01 Catering services 0.01 Accommodation services 0.01 Personal care 0.41 Personal effects n.e.c 0.14 Social protection 0.01 Insurance 0.04 Financial services n.e.c 0.04 Other services n.e.c 0.03 Source: Authors’ computations. 46 Appendix C. Alternate CEQ analyses Appendix Appendix table 2. PDI Scenario results (adult equivalent, AROP poverty line) Marginal Contributions Percent Concentration Kakwani Redistributive Poverty reduction of MIPP coefficien coefficien effect effect (+ = AROP (+ = equalizing) poverty decline) (1) (2) (3) (4) (5) Disposable Income Direct transfers 5.8% -0.1524 0.6000 0.0341 0.0666 Child & family contributory 1.2% 0.0334 0.4142 0.0037 0.0108 Child & family means-tested 0.8% -0.3097 0.7573 0.0076 0.0149 Child & family non-contributory 0.3% -0.2788 0.7264 0.0021 0.0045 Disability 1.8% -0.1036 0.5512 0.0086 0.0169 Old-age non-contributory 0.0% -0.6181 1.0657 0.0001 0.0000 Unemployment 0.8% -0.0277 0.4753 0.0030 0.0062 Other 0.8% -0.4646 0.9122 0.0066 0.0089 Direct taxes -7.1% 0.5500 0.0984 0.0097 -0.0124 Personal Income Tax - interest -0.2% 0.9236 0.4760 0.0010 0.0002 Personal Income Tax -7.0% 0.5355 0.0879 0.0087 -0.0124 Social contributions (non-pensions) -7.4% 0.3807 -0.0292 -0.0009 -0.0070 Contributions Gral Sickness & -0.8% 0.4580 0.0104 -0.0003 -0.0015 Maternity Fund employee Contributions Unemployment Fund -0.2% 0.4580 0.0104 -0.0001 -0.0005 employee Contributions Health employee -1.7% 0.4580 0.0104 -0.0007 -0.0030 Contributions Gral Sickness & -1.1% 0.4580 0.0104 -0.0004 -0.0019 Maternity Fund employer Contributions Unemployment Fund -0.3% 0.4580 0.0104 -0.0001 -0.0005 employer Contributions Work Injury & Occup. -0.4% 0.4580 0.0104 -0.0001 -0.0010 Disease Fund Contributions Health employer -2.6% 0.4580 0.0104 -0.0010 -0.0049 Contributions Gral Sickness & 0.1450 0.0001 -0.1% 0.5926 -0.0000 Maternity Fund self-employed Contributions Health self-employed -0.2% 0.5926 0.1450 0.0003 -0.0004 Consumable Income Indirect subsidies 8.1% 0.4000 0.0476 -0.0001 0.0602 Electricity direct effects 3.1% 0.2411 0.2065 0.0056 0.0281 Electricity indirect effects 4.6% 0.4000 0.0476 -0.0014 0.0284 Gas direct effects 0.1% 0.4637 -0.0161 -0.0001 0.0005 Gas indirect effects 0.4% 0.4000 0.0476 -0.0001 0.0024 THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Appendix table 2. PDI Scenario results (adult equivalent, AROP poverty line) (continued) 47 Marginal Contributions Appendix Percent Concentration Kakwani Redistributive Poverty reduction of MIPP coefficient coefficient effect effect (+ = AROP (+ = equalizing) poverty decline) (1) (2) (3) (4) (5) Indirect taxes -21.0% 0.4543 0.0067 0.0163 -0.0830 VAT -15.1% 0.4554 0.0078 0.0126 -0.0649 Oil derivates direct effects -1.5% 0.5137 0.0661 0.0025 -0.0055 Oil derivates indirect effects -2.0% 0.4327 -0.0149 0.0014 -0.0089 Excise Alcoholic Beverages -0.1% 0.4311 -0.0165 0.0001 -0.0002 Excise Tobacco -2.3% 0.4302 -0.0174 0.0015 -0.0098 Final Income Health 8.9% 0.0221 0.4295 0.0332 Social Security 6.9% 0.0438 0.4038 0.0221 Public health 2.0% 0.0044 0.4432 0.0067 Education 5.9% -0.0280 0.4756 0.0192 Pre-school 0.6% 0.0288 0.4188 0.0018 Primary level 0.7% -0.0784 0.5260 0.0029 Lower Secondary 1.4% -0.1569 0.6045 0.0059 Upper Secondary 2.0% -0.1203 0.5679 0.0079 Tertiary 1.2% 0.3223 0.1253 -0.0005 Source: Authors’ estimates. Appendix table 3. PGT Scenario results Marginal Contributions Percent Concentration Kakwani Redistributive Poverty reduction of MIPP coefficient coefficient effect effect (+ = US$5.50 (+ = equalizing) poverty decline) (1) (2) (3) (4) (5) Disposable Income All contributory pensions 14.2% -0.3766 0.9200 0.1010 0.1283 All direct transfers incl contributory pensions 20.2% -0.2913 0.8347 0.1400 0.1822 All direct transfers excl contributory -0.0720 pensions 6.0% 0.6154 0.0344 0.0380 Direct transfers: child & family 0.0853 1.2% 0.4581 0.0046 0.0042 contributory (Per capita) Direct transfers: child & family 0.8% -0.2058 0.7492 0.0074 0.0125 means-tested (Per capita) Direct transfers: child & family nc 0.3% -0.1573 0.7007 0.0022 0.0017 (Per capita) Direct transfers: disability (Per capita) 1.9% -0.0432 0.5866 0.0086 0.0072 48 Appendix table 3. PGT Scenario results (continued) Marginal Contributions Appendix Percent Concentration Kakwani Redistributive Poverty reduction of MIPP coefficien coefficien effect effect (+ = US$5.50 (+ = equalizing) poverty decline) (1) (2) (3) (4) (5) Direct transfers: old-age nc (Per 0.0% -0.5672 1.1106 0.0002 0.0004 capita) Direct transfers: unemployment (Per capita) 0.8% 0.0595 0.4839 0.0030 0.0020 Direct transfers: other (Per capita) 0.9% -0.3040 0.8474 0.0066 0.0076 All direct taxes -15.0% 0.5438 0.0004 0.0110 -0.0010 All contributions -11.5% 0.5092 -0.0342 0.0041 -0.0010 All direct taxes and contributions -26.5% 0.5288 -0.0146 0.0157 -0.0027 Personal Income Tax - interest -0.2% 0.9085 0.3651 0.0010 0.0000 Personal Income Tax -7.2% 0.5724 0.0290 0.0078 -0.0002 Contributions Gral Sickness & Maternity Fund employee -0.8% 0.5062 -0.0372 0.0002 -0.0001 Contributions Unemployment Fund employee -0.2% 0.5062 -0.0372 0.0000 -0.0001 Contributions Health employee -1.8% 0.5062 -0.0372 0.0004 -0.0001 Contributions Gral Sickness & Maternity Fund employer -1.2% 0.5062 -0.0372 0.0003 -0.0001 Contributions Unemployment Fund employer -0.3% 0.5062 -0.0372 0.0001 -0.0001 Contributions Work Injury & Occup. Disease Fund -0.4% 0.5062 -0.0372 0.0001 -0.0001 Contributions Health employer -2.7% 0.5062 -0.0372 0.0006 -0.0001 Contributions Gral Sickness & 0.0001 0.0000 Maternity Fund self-employed -0.1% 0.6062 0.0628 Contributions Health self- -0.1% 0.6062 0.0628 0.0002 0.0000 employed Contributions to pensions employee -5.0% 0.5062 -0.0372 0.0013 -0.0001 Contributions to pensions employer -6.2% 0.5062 -0.0372 0.0017 -0.0003 Contributions to pensions self- -0.5% 0.6062 0.0628 0.0006 0.0000 employed Consumable Income All indirect subsidies 8.3% 0.3104 0.2330 0.0046 0.0171 Indirect subsidies: Electricity direct 3.1% 0.2117 0.3316 0.0059 0.0102 effects Indirect subsidies: Electricity indirect effects 4.7% 0.3692 0.1742 -0.0014 0.0082 Indirect subsidies: Gas direct 0.1% 0.4205 0.1229 -0.0001 0.0000 effects Indirect subsidies: Gas indirect 0.4% 0.3692 0.1742 -0.0001 0.0006 effects THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Appendix table 3. PGT Scenario results (continued) 49 Marginal Contributions Appendix Percent Concentration Kakwani Redistributive Poverty reduction of MIPP coefficient coefficient effect effect (+ = US$5.50 (+ = equalizing) poverty decline) (1) (2) (3) (4) (5) All indirect taxes -21.6% 0.4206 -0.1228 0.0172 -0.0255 Indirect taxes: VAT -15.5% 0.4225 -0.1208 0.0133 -0.0179 Indirect taxes: Oil derivates direct -1.5% 0.4722 -0.0712 0.0026 -0.0009 effects Indirect taxes: Oil derivates indirect -2.1% 0.3997 -0.1437 0.0015 -0.0022 effects Indirect taxes: Excise Alcoholic -0.1% 0.3950 -0.1484 0.0001 -0.0004 Beverages Indirect taxes: Excise Tobacco -2.4% 0.3949 -0.1485 0.0016 -0.0023 Final Income Net health transfers 9.1% 0.0236 0.5198 0.0332 In-Kind health benefits: Social 7.1% 0.0304 0.5130 0.0248 Security In_kind health benefits: Public health 2.0% 0.0000 0.5434 0.0072 Net education transfers 6.0% 0.0593 0.4841 0.0243 In-Kind education benefits: Pre- 0.6% 0.0538 0.4896 0.0029 school In_kind education benefits primary 0.7% -0.0077 0.5511 0.0039 level In-Kind education benefits: Lower 1.5% -0.0712 0.6146 0.0077 Secondary In-Kind education benefits: Upper -0.0021 Secondary 2.0% 0.5454 0.0090 In-Kind education benefits: Tertiary 1.2% 0.3662 0.1772 -0.0002 Source: Authors’ estimates. 50 Appendix D. Macro validation Appendix Appendix table 4. Macro validation CEQ estimations Actual CEQ estimations (% Actual of disposable (% of private Ratio (EUR million) (EUR million) income) consumption) [1] [2] [3] [4] [1] / [3] Old-age contributory pensions 4,029.0 15.2% 4,066.7 11.2% 0.99 Social security contributions 5,427.0 20.5% 4,864.5 13.4% 1.12 Contributions to social security for old-age pensions 3,261.0 12.3% 2,576.2 7.1% 1.27 Contributions to social security 8.2% 2,288.3 6.3% 0.95 (other than pensions) 2,166.0 Direct Taxes 2,076.0 7.8% 1,879.4 5.2% 1.10 Social Protection 1,683.0 6.4% 1,914.2 5.3% 0.88 Social Assistance 927.0 3.5% 1,108.5 3.0% 0.84 Conditional and Unconditional Cash 923.0 3.5% 1,104.3 3.0% 0.84 Transfers Child & family contributory 352.6 1.3% 320.7 0.9% 1.10 Child & family non-contributory 95.0 0.4% 93.9 0.3% 1.02 Child & family means-tested 228.4 0.9% 183.6 0.5% 1.24 Other direct transfers 246.7 0.9% 506.0 1.4% 0.49 Non-contributory Pensions 4.3 0.0% 4.2 0.0% 1.02 Social Insurance 756.0 2.9% 805.7 2.2% 0.94 Disability 528.6 2.0% 540.7 1.5% 0.98 Unemployment 227.5 0.9% 265.0 0.7% 0.86 Indirect taxes 6,115.6 23.1% 7,619.2 21.0% 0.80 VAT 4,390.5 16.6% 5,128.3 14.1% 0.86 Sales Tax - Excise Taxes 1,725.2 6.5% 2,490.9 6.9% 0.69 Oil direct 426.0 1.6% 1,117.1 3.1% 0.38 Oil indirect 592.1 2.2% Alcoholic Beverages 31.1 0.1% 150.6 0.4% 0.21 Tobacco 676.1 2.6% 1,223.2 3.4% 0.55 Customs Duties - 0.0% Other - 0.0% Indirect subsidies 2,359.9 8.9% 3,096.6 8.5% 0.76 Electricity 2,216.0 8.4% 2,476 6.8% 0.90 Direct 889.7 3.4% Indirect 1,326.3 5.0% Gas 143.9 0.5% 621 1.7% 0.23 Direct 26.8 0.1% Indirect 117.1 0.4% THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Appendix table 4. Macro validation (continued) 51 CEQ estimations Actual CEQ estimations (% Actual of disposable (% of private Ratio Appendix (EUR million) (EUR million) income) consumption) [1] [2] [3] [4] [1] / [3] Education 1,712.9 6.5% 0.0 0.0% 987220.82 Pre-school & Primary 383.1 1.4% 381.8 1.0% 1.00 Pre-school 174.2 0.7% Primary 208.9 0.8% Secondary 988.5 3.7% 1,012.0 2.8% 0.98 Lower secondary 413.8 1.6% Upper secondary 574.6 2.2% Tertiary 341.4 1.3% 341.3 0.9% 1.00 Health 2,588.56 9.8% 2,588.7 7.1% 1.00 Contributory 2,009.3 7.6% 2,009.3 5.5% 1.00 Noncontributory 579.3 2.2% 579.4 1.6% 1.00 Source: Authors’ estimates, Eurostat, OECD, and NSI. 52 Appendix E. Detailed description of the fiscal syste Appendix The main data source for the details of the fiscal system is the Bulgaria EUROMOD Country report and ESSPROS. › Social Benefits system The details of the social benefit system are summarized by category in the tables and text below. Gross monthly income refers to: All components of the original (market) income which are defined as taxable by PITA (See section (1)  on taxes). (2) All pensions: › pension for insurance and old age; › pensions for invalidity due to general sickness; › pension for invalidity due to employment injury or occupational disease; › social old-age pension; › social invalidity pension; › survivor’s pension; › pension supplement in case of deceased spouse; › pension from voluntary pension insurance. (3) All contributory benefits: › unemployment benefit (incl. the benefit for long-term unemployment); › cash benefit for sickness due to General Sickness › cash benefit for sickness due to employment injury of occupational disease › cash benefit for pregnancy and childbirth › cash benefit for bringing up a child up to the age of . (4) All family benefits: › monthly allowance for bringing up child up to age of 1; › monthly child allowance for bringing up a child until completion of secondary education but not longer than 20 years of age; › monthly child allowance for bringing up a permanently disabled child up to 18 years of age and until completion of secondary education but not longer than 20 years of age; › targeted allowance for pupils; › lump sum allowance for raising a child until 1 year of age for mothers who are regular tertiary students; › lump sum allowance for raising twins until 1 year of age; › lump sum allowance in case of adoption. (5) Irregular income sources (e.g. lottery winnings, bequests, etc.). THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Exceptions: 53 (1) All benefits granted under LSA: ›  monthly social assistance allowance; Appendix ›  targeted allowance for heating. (2) Supplement for care by assistant (3) All allowances for social integration of disabled people (4) Lump sum pregnancy grant (5) Lump sum grant upon childbirth. › Child and Family Benefits Benefit:  Cash benefit for pregnancy and childbirth Income test:  Contributory, Non-means tested Eligibility conditions:  This is a benefit granted to mothers upon a medical referral and under the condition that the mother has been insured at least for 12 months. The benefit is paid for a period of 410 days, 45 days of which—before the anticipated date of delivery. Fathers have a right to a benefit for a period of 15 days immediately after the childbirth. The same eligibility conditions apply. Moreover, the benefit can be received by the fathers (instead of mothers) from the 6th month until the 1st year of the child and upon the mother’s approval. The benefit amounts 90% of the contributory income for the last 24 months. As of 1 June 2017, mothers can receive 50% of the benefit if they do not use the rights that they are entitled to (if eligibility conditions are fulfilled), or if they interrupt the receipt of the benefit due returning to work. In some cases (in case of mother’s death or if she has been denied paternity rights) this opportunity can be used by the father (if he fulfills the eligibility conditions). As of 1 June 2017, if some of the following events occur before the 42th day after the childbirth: (1) the child is born dead or deceases before; (2) the child is accommodated to a family of relatives or foster families as a child protection measure; (3) the child is included in the list of children available to be fostered, mothers can receive a maternity and family benefit until the 42nd day following the childbirth. If her working capacity is not fully recovered, the duration of receipt can be prolonged upon a decision of competent medical authorities. If the mentioned events occur after the 42nd day following the childbirth, the benefit receipt shall be terminated if mother’s working capacity is fully recovered. Amount in 2018 (BGN): 9  0% of the contributory income for the last 24 months, not less than the minimum daily wage and not more than the average daily renumeration for the period based on which the benefit has been calculated. Benefit:  Cash benefit for bringing up a child up to age of 2 Income test:  Contributory, Non-means-tested Eligibility conditions:  This benefit granted upon a medical referral and under the condition that the person has been insured at least for 12 months. The mother can receive 50% of the benefit if she does not use the right which she is entitled to (in case that she fulfills all eligibility conditions), or if she 54 interrupts the receipt of the benefit due to returning to work. In some cases, the benefit can be paid to the father or, since June 2017, to one of the mother’s parents—if the following conditions are fulfilled: (1) the mother is severely diseased and father is either deceased or missing, and (2) the mother’s Appendix parent should be employed on a labour contract. The person whom the benefit is paid to has to use the paid-leave for bringing up a child up to age 2 according to the regulations of Bulgarian Labour Code The amount of the benefit is determined by the Public Social Insurance Budget Act Amount in 2018 (BGN):  BGN 380 (US$449.5 2011 revised PPP) Benefit:  Cash benefit in case of adoption of a child up to age of 5 Income test:  Contributory, Non-means-tested Eligibility conditions:  This benefit is given to people who have been insured against the sickness and maternity risks for at least 12 months and who are in paid leave in case of adoption of child between 2 and 5 years, according to the Labour Code. The duration of the benefit is 365 days but cannot be received after the 5th year of the child. Amount in 2018 (BGN): 9  0% of the contributory income for the last 24 months, not less than the minimum daily wage and not more than the average daily renumeration for the period based on which the benefit has been calculated. Benefit:  Monthly allowance for bringing up a child up to age of 1 Income test: Non-contributory, Means-tested: Average gross monthly income per family member for the last 12 months <= BGN 450 (US$532.2 2011 revised PPP) Eligibility conditions:  This is an allowance paid to mothers of children up to one year of age. To receive the benefit, mothers should not be insured and should not be in receipt of contributory benefit for pregnancy and childbirth, benefit for bringing up a child up to age 2 and benefit in case of adoption of a child between 2 and 5 years 9. The threshold level and the amount of the allowance are determined by the State Budget Act. The benefit is paid monthly until the child turns 1. Amount in 2018 (BGN):BGN 100 (US$118.3 2011 revised PPP) (monthly) Benefit:  Monthly allowances for a child until graduation from high school, but not after the age of 20 Income test:  Non-contributory, Means-tested: Income group 1:   Average gross monthly income per family member for the last 12 months <= BGN 400 (US$473.2 2011 revised PPP) Income group 2:   Average gross monthly income per family member for the last 12 months > BGN 400 (US$474.4 2011 revised PP) & <= BGN 500 (US$591.5 2011 revised PPP) THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Eligibility conditions:  This is an allowance paid to a parent bringing up a child until the completion 55 of secondary education, but no longer than the moment when the child reaches 20 of age. As of 1 January 2017, this allowance cannot be received simultaneously with the monthly allowance for Appendix permanently disabled children below 18 years of age and prior to finishing secondary school, but not later than reaching 20 years of age. This allowance should be terminated in case: (1) the child has been accommodated to other family as a child protection measure under the Child Protection Act; (2) when, within one month of the school year, a child attending a group for compulsory preschool education has more than three absence days and there are no valid reasons justifying the absence; (3) when, within one month of the school year, a child in school education has missed more than 5 school hours without valid reasons justifying the absence; (4) when the child receiving the allowance becomes a parent. The child who has become a parent is entitled to a lump-sum allowance which amount is equal to the amount of the suspended allowance for the period from the termination to the expiration of the period for which it was granted. In order to receive the lump-sum allowance, the child should continue to attend school and should not have missed, within a month, 5 school hours, for which there are no valid reasons. The threshold level and the amount of the allowance are determined by the State Budget Act. Amount in 2018 (BGN): BGN US$ 2011 revised PPP Family with one child 40 47.3 Family with two children 90 106.5 Family with three children 135 159.7 Each child after the fourth +20 +23.7 Twins (per child) 75 88.7 Income group 1: 100% of benefit Income group 2: 80% of benefit Benefit:  Monthly allowance for permanently disabled children below 18 years of age and prior to finishing secondary school, but not later than reaching 20 years of age Income test:  Non-contributory, Non-means tested Eligibility conditions:  This is an allowance paid to a parent of a child with permanent disability up to 18 years of age and until completion of secondary education but not later than 20 years of age. The allowance is paid regardless of the family income and upon a medical referral. As of 1 January 2017, this allowance cannot be received simultaneously with the monthly allowances for a child until graduation from high school, but not after the age of 20, and (2) monthly allowances for social integration under the Integration of People with Disabilities Act. The allowance amount depends on the severity of the permanent disability and in 2018 and on whether the child is placed in a family of relatives or foster care family (as a child protection measure). 56 Amount in 2018 (BGN): Per month, per child Appendix Own family (US$ 2011 Relatives/Foster care Severity of disability Own family (BGN) Relatives/Foster care revised PPP) (US$ 2011 revised PPP) 90%+ 930 1100.2 490 579.7 71%-90% 450 532.3 420 496.9 50-70% 350 414.0 350 414.0 Benefit:  Lump sum pregnancy grant Income test: Non-contributory, Means tested: Average gross monthly income per family member for the last 12 months <= BGN 350 (US$414.0 2011 revised PPP) Eligibility conditions:  This is a lump sum allowance paid to mothers who are not insured and do not receive a contributory maternity benefit. If the parents are divorced, only the income of the parent who has received the right to take care for the child shall be taken into account. The threshold level and the amount of the allowance are determined by the State Budget Act. Amount in 2018 (BGN):  BGN 150 (US$177.4 2011 revised PPP) Benefit:  One-off allowance upon childbirth Income test:  Non-contributory, Non-means tested Eligibility conditions: This is an allowance paid to mothers upon childbirth regardless of the family income. The right to a benefit occurs since the date of the childbirth. The amount of the benefit depends upon the number of the born children. It is determined by the State Budget Act. In case of twins, the benefit for each child shall be paid in the amount for the second child. If the parents are divorced, only the income of the parent who has received the right to take care for the children shall be taken into account. Amount in 2018 (BGN): BGN US$ 2011 revised PPP First child 250 295.7 Second child 600 709.8 Third child 300 354.9 Fourth and each subsequent child 200 236.6 THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Benefit:  One-off allowance for schoolchildren enrolled in 1st gradе at a state or municipal school 57 Income test: Non-contributory, Means-tested: Appendix Average gross monthly income per family member for the last 12 months <= BGN 450 (US$532.3 2011 revised PPP) Eligibility conditions:  This is a lump sum allowance paid to a family whose child is enrolled in 1st grade at state or municipal schools. The threshold level and the amount of the allowance are determined by the State Budget Act. Amount in 2018 (BGN):  BGN 250 for the school year 2017/2018 (US$295.74 2011 revised PPP) Benefit:  Lump sum allowance for raising a child until 1 year of age for mothers who are regular tertiary students Income test:  Non-contributory, Non-means tested Eligibility conditions:  This is a benefit for a mother of a child up to age 1 under the condition that she is a regular tertiary student. The benefit can be received if the child is not placed in a specialized child-care institution. In addition, the mother shall not be insured and shall not receive contributory maternity benefit for pregnancy and childbirth or contributory maternity benefit for bringing up a child up to the age of 2. Since 1st of January 2016, the benefit is paid on two instalments. The first instalment is paid immediately after the approval of the application and the second—after the mother proves that she has been enrolled in the next semester. Its amount is determined by the State Budget Act. Amount in 2018 (BGN):  BGN 2880 (US$3407.0 2011 revised PPP) Benefit:  Lump sum allowance for raising twins until 1 year of age Income test:  Non-contributory, Non-means tested Eligibility conditions:  This is a benefit for a mother of twins up to age 1. The benefit cannot be received if even one of the children is placed in a specialized child-care institution. Its amount is determined by the State Budget Act. Amount in 2018 (BGN):  BGN 1200 per child (US$1419.6 2011 revised PPP) Benefit:  Lump sum allowance in case of adoption Income test:  Non-contributory, Non-means tested Eligibility conditions:  This is a benefit granted to adoptive parents if they and the adopted child/ children reside in Bulgaria permanently, i.e. more than 183 days per year. The amount of the benefit is determined annually in the State Budget Act on the condition that it should not be lower than the 58 amount for the previous year and not higher than the amount of the cash benefit for pregnancy and childbirth. Appendix Amount in 2018 (BGN):  BGN 250 (US$295.7 2011 revised PPP) Benefit  Financial support for the prevention of the abandonment, bringing up of a child by relatives or foster family Income test:  Non-contributory, Means-tested Eligibility conditions:  These are a set of allowances granted in order to support the bringing up of a child at risk. Allowances could be paid on a monthly basis or as a one-off payment. Social workers assess the income of the family and the necessity the financial support to be provided. Amount in 2018 (BGN): One-off payments: Cannot be higher the 5-time amount of the GMI. Monthly payments: Can be up to 3 times the amount of the GMI (for children up to 7 years of age), up to 3.5 times the amount of the GMI (for children between 7 and 14 years of age) or up to 4 times the amount of the GMI (for children between 14 and 20 years of age). For children with disabilities, a supplement of 75% of the GMI should be paid to the monthly allowances. › Unemployment and sickness benefits Benefit:  Unemployment Income test:  Contributory, Non-means tested Eligibility conditions: This is a benefit granted to the involuntarily unemployed, provided that they: (1) are currently not entitled to an old-age pension; are not undertaking an economic activity requiring compulsory social insurance (i.e. employed (2)  or self-employed; registered agricultural producer) (3) became unemployed involuntarily; (4)  have been insured for at least 12 months amid the last 18 months before the termination of the employment; is registered at the Territorial Public Employment Office as an unemployed person and is ready (5)  to accept a proposed job. The duration of the benefit varies from 4 months to 1 year depending on the duration of contribution. Amount in 2018 (BGN): 60% of the average gross contributory income calculated for the assessment period which is the last 24 months preceding the month of the termination of the insurance. Minimum of BGN 9 (US$10.6 2011 revised PPP) per day, maximum of was 74.29 (US$87.9 2011 revised PPP) per day. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Benefit:  Cash benefit for sickness due to General Sickness 59 Income test:  Contributory, Non-means tested Appendix Eligibility conditions:  This benefit is granted for temporarily disabled individuals upon a medical referral and under the condition that the person has been insured for at least 6 months. The requirement for 6 months of insured length of service shall not apply to those under the age of 18. Amount in 2018 (BGN):  80% of the contributory income for the last 18 months (daily) Benefit:  Cash benefit for sickness due to an employment injury or occupational disease Income test:  Contributory, Non-means tested Eligibility conditions:  This benefit is granted for temporarily disabled individuals granted upon a medical referral. Notwithstanding that the benefit is contributory, no qualification period is required. Amount in 2018 (BGN):  90% of the contributory income for the last 18 months (daily) › Targeted benefits Benefit:  Targeted Heating Allowance Income test:  Non-contributory, Means-tested  ncome is below the so-called “Differentiated Minimum Income for Heating” (DMIH) based on the I respective category. Appendix table 5. DMIH Thresholds by category of person, 2018 Category of person DMIH threshold person living alone 233.08% person with reduced working capacity of 50% or more living alone 272.68% orphan child 219.88% lone parent with one or more children aged under 18 (under 20 for children attending 272.68% school) two spouses living together (per each spouse) 167.08% child aged between 0-18 (up to 20 when studying) 180.28% child with permanent disabilities 219.88% child accommodated in relatives’ family or in foster family (as a result of a child 224.68% protection measure) person cohabiting with other persons 224.68% pregnant woman 45 days before the term 206.68% parent taking care of a child under age of 3 206.68% person over the age of 70 206.68% 60 Appendix table 5. DMIH Thresholds by category of person, 2018 (continued) Category of person DMIH threshold Appendix person over the age of 75 living alone 297.88% person over the age of 65 living alone 311.08% person with reduced working capacity of 50% or more 206.68% person with reduced working capacity of 70% or more 246.28% person with reduced working capacity of 90% or more 297.88% person over the age of 65 living alone 311.08% person with reduced working capacity of 50% or more 206.68% Source: Bulgaria EUROMOD Country report, ESSPROS.  his benefit is granted to persons living alone or families whose income is below Eligibility conditions: T the so-called “Differentiated Heating Income” (DHI) and who also fulfill other eligibility criteria. DHI is calculated on the basis of the so-called “Basic Heating Income” (BHI) following a similar procedure as the DMI for the monthly social assistance allowance. but the percentages for the different categories of beneficiaries are higher. The amount of the allowance is determined by the Minister of Labour and Social Policy. It is granted for a period of 5 months—from November to March—the duration of Bulgaria’s heating season. Amount in 2018 BGN: For the 2017/2018 heating season, the benefit was BGN 73.02 (US$86.4 2011 revised PPP) per month of the heating season. For the 2018/2019 heating season, the corresponding benefit was BGN 74.83 per month (US$88.5 2011 revised PPP). Benefit:  Lump-sum social assistance allowance Income test: Non-contributory, Means-tested: Income is below the so-called “Differentiated Minimum Income” (DMI) based on the respective category. Appendix table 6. DMI threshold by category of person, 2018 Category of person DMI threshold person over the age of 75 living alone 165.0% person over the age of 65 living alone 140.0% person over the age of 65, cohabiting with other persons 100.0% person under the age of 65 living alone 73.0% person under the age of 65, cohabiting with other persons 66.0% two spouses living together (per each spouse) 66.0% person with reduced working capacity of 50% or more 100.0% person with reduced working capacity of 70% or more 125.0% child aged between 0 and 16 (up to 20 if attending school) 91.0% orphan child aged between 0 and 16 (up to 20 if attending school) N/A THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Appendix table 6. DMI threshold by category of person, 2018 (continued) 61 Category of person DMI threshold Appendix child accommodated to another family as a result of child protection measure aged N/A between 0 and 16 (up to 20 if attending school) child aged between 7 and 16 (up to 20 if attending school), if having 5 or more N/A unexcused absences from school child aged between 5 and 16 (up to 20 if attending school), if having 5 or more 30.0% unexcused absences from school within a month child aged between 7 and 16 and not attending school N/A child aged between 5 and 16 and not attending school 20.0% children (up to 20 when studying) if it is not proved that all mandatory immunisations 30.0% and preventive medical checkups and done child aged 16-18, not attending school and is registered at a local public employment 66.0% office orphan or child accommodated in a foster family 100.0% child with permanent disability up to 18 years of age 100.0% lone parent taking care of a child under age of 3 120.0% lone parent with one or more children aged under 16 (under 20 if attending school) 100.0% parent taking care of a child up to 16 years of age (up to 20 years if studying) N/A parent taking care of a child up to 18 years of age not attending school N/A parent taking care of a child aged between 16-18 not attending school—if the child is N/A registered at local employment office child aged between 5 and 16 (up to 20 if attending school), if having 5 or more unexcused absences from school within a month/ if missed 3 school days—for children N/A attending preschool education pregnant woman 45 days before the term 100.0% parent taking care of a child aged under 3 100.0% Source: Bulgaria EUROMOD Country report, ESSPROS. Eligibility conditions:  This is a non-contributory allowance granted to households with low incomes. However, there are particular categories of persons who are non-eligible for this benefit. If such persons are part of the household, then they should be removed from the assessment unit. These categories are: (1) persons who have relatives that are legally obliged to maintain them; (2) persons accommodated for more than 30 days in medical, social, military or other institutions; persons studying in high schools or in evening form of education, including pupils who are (3)  enrolled in private schools (exceptions: disabled persons, pregnant women, and parents caring for a child up to age 3); (4) persons who have refused cultivating state or municipal land suggested to them.  In addition, the right to monthly social assistance benefits depends on the following conditions: (1) the claimant should live in a lodging composed of maximum 1 room per household member; the claimant should not be registered as sole proprietor and should own shares of the capital (2)  of an economic undertaking; (3) the claimant should not have shares or equities above 500 BGN; 62 the claimant should not possess mobile and immobile assets that can be a source of income, (4)  except for the belongings that serve the usual needs of the family (this is evaluated by the social worker); Appendix the claimant should not have contracts for a transfer of property in return for the obligation (5)  for support and care (e.g. caring for lone elderly owners); (6)  the claimant should not have acquired residential or summerhouse property or ideal parts of such property through purchase or donation during the last 5 years; claimants who are unemployed people should have a registration at local public employment (7)  offices for at least 6 months prior the application and should not have declined a job or training offer (this requirement is not applicable for single parent taking care for a child below 3 years of age; a person with permanently reduced working capacity of 50% or more; people taking care for a severely diseased family member; people with mental disorders). Amount in 2018 BGN:  The amount of the benefit is determined as a difference between the DMI of the household and the gross family income from all sources, net of allowed exceptions. In 2018, the monthly GMI was BGN 75 (US$88.7 2011 revised PPP) (annual BGN 900, US$1064.7 revised PPP). Benefit:  Targeted allowance for paying rent for a municipal lodging Income test: Non-contributory, Means-tested: Income is below 150% of the “Differentiated Minimum Income” (DMI) based on the respective category. Eligibility conditions: This benefit is paid to persons accommodated in municipal lodgings whose income for the previous month are below 150% of the DMI (Appendix table 6). Only the following groups are entitled to such allowance: orphans up to 25 years of age who have graduated at a specialized institution providing (1)  vocational training for disabled children and children at risk—until November 2020; (2) persons over 70 years of age living alone and; (3) lone parents. Amount in 2018 (BGN):  The amount of the benefit depends on the actual expenditures made by the beneficiaries. Benefit:  Targeted allowance for issuing a personal ID Income test:  Non-contributory, Means-tested: Social worker assessment Eligibility conditions:  This benefit granted upon an assessment of needs made by social workers and its amount is determined by the social workers. Amount in 2018 (BGN):  Variable THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Benefit:  Targeted allowance for medical treatment abroad 63 Income test:  Non-contributory, Means-tested: Appendix Social worker assessment Eligibility conditions:  This benefit is granted upon an assessment of needs made by the social workers only to persons having received permission for medical treatment abroad from the Ministry of Health, at the expense of its budget. The financial support shall be granted as a lump-sum allowance to cover the personal expenses of the beneficiary and her/his relatives Amount in 2018 (BGN):  Variable Benefit:  Targeted allowance for free traveling Income test:  Non-contributory, Non-means tested Eligibility conditions:  This benefit is granted to mothers of many children, persons with reduced working capacity over 71%, and children up to age 16 with permanent disabilities. Mothers of many children are entitled to free travel once per year—1 return ticket by the railway or bus transport anywhere within the country. To facilitate the social adaptation individuals with reduced working capacity over 71%, children up to age 16 with a specific type and degree of disability of 50% or more, and war-disabled individuals have the right to free travel twice a year—2 return tickets by rail or bus transport anywhere within the country. Amount in 2018 (BGN):  Variable Benefit Lump-sum social assistance allowance Income test:  Non-contributory, Means-tested: Social worker assessment Eligibility conditions:  This is a benefit granted only once per year to cover unexpected health, educational, living and other emergencies. The allowance can be received by all citizens regardless of their income or material status but upon the result of the so-called “social assessment” by the social workers. Amount in 2018 (BGN):  Varies, but cannot exceed 5 times the GMI › Disability Benefits Benefit:  Targeted financial support for buying and repairing supportive means, devices, equipment and medical products Income test:  Non-contributory, Non-means tested: Medical referral 64 Eligibility conditions:  These are in-kind benefit provided to persons with permanent disabilities. The support is granted upon a medical referral and without means-test. The support is provided in form of reimbursement of the expenses actually made by the beneficiary, but not more than a fixed maximum Appendix level. The maximum level of the possible financial support is determined per each item (mean, device, equipment, or other product) by an Ordinance of the Minister of Labour and Social Policy. The allowance was paid until 31 December 2018 as the Integration of People with Disabilities Act. Amount in 2018 (BGN):  Variable Benefit:  Monthly allowance for social integration Income test:  Non-contributory, Non-means tested Eligibility conditions:  These are benefits paid to permanently disabled people. There are 7 types of social integration allowances (for transport services; for ICT services; for education; for medical bath treatment and rehabilitation; for dietary purposes and medicines; for accessible information; for paying the rent of a municipal dwelling). Their amounts are calculated as a percentage of GMI and vary from 15%, for example—monthly allowance for transport services, to 300% of GMI, for example the allowance for medical bath treatment and rehabilitation. As of 1 January 2017, only permanently disabled people above 18 years of age are entitled to this allowance. The allowance was paid until 31 December 2018 as the Integration of People with Disabilities Act. Amount in 2018 (BGN):  Variable Benefit:  Supplement for care by assistant Income test:  Non-contributory, Non-means tested Eligibility conditions:  This benefit is granted to mothers of many children, persons with reduced working capacity over 71%, and children up to age 16 with permanent disabilities. Mothers of many children are entitled to free travel once per year—1 return ticket by the railway or bus transport anywhere within the country. To facilitate the social adaptation individuals with reduced working capacity over 71%, children up to age 16 with a specific type and degree of disability of 50% or more, and war-disabled individuals have the right to free travel twice a year—2 return tickets by rail or bus transport anywhere within the country. Amount in 2018 (BGN):  75% of the social-old age pension (BGN 123.28, US$145.8 2011 revised PPP per month) › Pension Benefits Benefit:  Pension for insurance and old age Income test:  Contributory, Non-means tested THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Eligibility conditions:  This is a benefit it is a contributory benefit granted to persons who have reached 65 the standard retirement age and have a fixed number of years of insurance. For men, the condition in 2018 was 64 years of age and 1 month and 38 years and 6 months of insurance and for women, the Appendix condition in 2018 was 61 years of age and 2 months and 35 years and 6 months of insurance. Since 1 January 2016 individuals who have the necessary length of contributory service can get an insurance and old-age pension maximum 12 months prior reaching the statutory retirement age, i.e. to retire up to one year earlier. In that case the amount of their pension shall be reduced by 0.4% per each insufficient month. The minimum amount of the insurance and old-age pension is determined by the Public Social Insurance Budget Act. Amount in 2018 (BGN):  The amount of the pension depends on: (1) the amount of the average contributory income for the country for the last 12 months preceding the date of retirement; (2) the monthly ratios between the past earnings of the claimant and the average contributory income for the country for the assessment period; (3) the duration of the pension insurance (number of years). The minimum amount of the benefit is BGN 200. The maximum amount of the pension is equal to 35% of the maximum amount of the contributory income for the country Benefit:  Pension for invalidity due to general sickness Income test:  Contributory, Non-means tested: Medical referral Eligibility conditions:  This is a benefit granted to persons with permanent disability upon medical referral. The minimum required period of insurance varies according to the age of the claimant—it is between 0 for persons up to age 20 and 5 years for persons of age 30 or above. Amount in 2018 (BGN): The calculation formula is the same as for the pension for insurance and old age. However, the number of years of insurance is corrected with a coefficient based on the percentage of reduced working capacity. There are certain minimums (which varies according to the degree of reduced working capacity) and the maximum amount is 35% of the contributory income for the country Benefit:  Pension for invalidity due to employment injury or occupational disease Income test:  Non-contributory, Non-means tested: Medical referral Eligibility conditions:  This is a benefit granted to persons with permanent disability upon medical referral. A minimum qualifying period is not required. The calculation formula is the same as for the pension for insurance and old age Amount in 2018 (BGN): The calculation formula is the same as for the pension for insurance and old age. However, the number of years of insurance is corrected with a coefficient based on the percentage of reduced working capacity. There are certain minimums (which varies according to the degree of reduced working capacity) and the maximum amount is 35% of the contributory income for the country 66 Benefit:  Social Old-Age Pension Income test:  Non-contributory, Means-tested: Appendix Gross annual income per family member for the last 12 months < GMI Eligibility conditions:  This benefit is granted to people over 70 years old. In order to receive the benefit, the person should not be in receipt of any other pension Amount in 2018 (BGN):  BGN 123.28 (US$145.8 2011 revised PPP) per month Benefit:  Social Invalidity Pension Income test:  Non-contributory, Non-means tested Eligibility conditions:  This benefit is granted to persons 16 years of age with reduced working capacity of over 71%. Amount in 2018 (BGN):  The amount of the social invalidity pension varies according to the degree of reduced working capacity and should be determined as a percentage of the social old-age pension. Benefit:  Pension supplement in case of deceased spouse Income test:  Non-contributory, Non-means tested  his is a benefit granted to a pensioner whose spouse has deceased. It is granted Eligibility conditions: T for life but should be terminated in case of re-marriage. The supplement cannot be combined with a survivor’s pension, but full accumulation with other pensions is possible. Amount in 2018 (BGN):  26.5% of the pension (or the sum of pensions) of the deceased spouse. Benefit:  Pension from supplementary mandatory pension insurance Income test:  Contributory, Non-means tested Eligibility conditions:  These are benefits paid to persons having been insured in a private fund for supplementary mandatory pension insurance. Pension contributions are accumulated in the so-called “individual accounts”. There are two types of such funds: Universal Pension Funds—the insurance is mandatory only for people born after 31 December (1)  1959) The insurance in the Universal Pension Funds entitles the insured person to: (i)  a supplementary lifelong old-age pension after acquisition of an insurance and old age pension; a lump-sum payment of up to 50 % of the resources accrued on the individual account in (ii)  case of lifelong, permanently reduced working capacity exceeding 89.99% (70.99% until 31 December 2015); (iii.) a lump-sum payment or payment by installments of amounts to the survivors of a deceased person or of a pensioner. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES (2)  Professional Pension Funds)—the insurance is mandatory for persons working under the 1st 67 and 2nd category of labour, i.e. in risky and unhealthy working conditions. The insurance in the Professional Pension Funds entitle the insured person to: Appendix (i) fixed-period early-retirement occupational pension; lump-sum payment of up to 50 % of the resources accrued on the individual account in (ii)  case of lifelong, permanently reduced working capacity exceeding 89.99% (70.99% until 31 December 2015); lump-sum payment or payment by installments of amounts to the survivors of a deceased (iii)  insured person or of a pensioner. Since 1 January 2016, people insured in both Universal Pension Funds and Professional Pension Funds have been given the opportunity to decide whether to remain insured only in the State Public Social Insurance with an increased pension contribution. Under certain conditions, individuals who have decided to be insured only in State Public Social Insurance can restore their insurance in the Universal Pension Funds. Individuals previously insured in Professional Pension Funds who have decided to be insured only in the State Public Social Insurance cannot reverse their choice. Amount in 2018 (BGN):  Varies Benefit:  Pension from voluntary pension insurance Income test:  Contributory, Non-means tested  his benefit is paid to persons having been insured in a private fund for Eligibility conditions: T supplementary voluntary pension insurance. Pension contributions are accumulated in the so-called “individual retirement accounts”. There are two types of pensions—old-age pension (paid if the person has gained a right to a pension from public pension fund) and invalidity pension (paid upon a medical referral). The pension can be paid as a periodic or lump-sum benefit depending on the choice of the insured person. Amount in 2018 (BGN):  Varies › Social contributions Statutory contributions are compulsory for all individuals having earnings from at least one source of taxable income ( EUROMOD, 2021). Social Insurance Contributions (SIC) are usually split between employers and the employees (EUROMOD, 2021) and borne entirely by the self-employed in the case of self-employment. There is a minimum threshold for the monthly contributory base which varies by the occupation and economic activity of the company where the employee works. There is a maximum threshold for the monthly contributory income when the compulsory SIC are assessed—i.e. they are not due on the fraction of income beyond that threshold. In 2018, the upper threshold was BGN 2600 (US$3075.7, 2011 revised PPP) per month for employees. For the self-employed, in 2018, the lower threshold was BGN 510 (US$603.2 2011 revised PPP) per month and the upper threshold was BGN 2600 (US$3075.7 2011 revised PPP) per month. For registered farmers and tobacco producers in 2018, the thresholds were BGN 350 (US$414.0 2011 revised PPP) per month and BGN 2600 (US$3075.7, 2011 revised PPP) per month. 68 Employees › The employee social contributions for Pension fund are compulsory paid by all employees. Appendix › The employee social contributions for General Sickness and Maternity Fund and Unemployment Fund are compulsory paid by all employees hired on a labour contract. › For civil servants (state administration, military, judicial system, etc.) the total amounts of compulsory SIC are paid by the State. › Employees receiving a pension and at the same time are hired on a labour contract are compulsory insured against all social and health risks. If they perform work under a civil contract and their remuneration is below the amount of the minimum wage (after normative allowed deductions) they do not have the obligation to pay health insurance contributions as they health insurance contributions are paid by the State in their capacity of pensioners. If the remuneration is above the minimum wage after the deductions, they are obliged to pay health insurance contributions. Regarding social insurance, pensioners performing work under civil contract pay pension contributions to the Pension Fund on a voluntary basis. Self-employed › Self-employed individuals have a legal duty to contribute to social insurance funds at a minimum level which covers only pension and health insurance. › A self-employed individual can choose a particular monthly value for her contributory base which could be higher than the minimum threshold and up to the ceiling. › Compulsory SIC are due by entrepreneurs and other self-employed only for those months during which they have conducted business activity (i.e. have maintain records for business expenditures and revenues). › Self-employed individuals can decide to pay contributions to General Sickness and Maternity Fund of the State Public Social Insurance on a voluntary basis. › Fund: Pension Conditions:  People contributing to the pension fund are insured against old-age, invalidity and survivors’ risks. Insurance in the Pensions Fund is mandatory for all categories of insured, regardless of their income, occupation or economic status. Contribution rates: Appendix table 7. Contribution rates by pension fund Fund Employee Employer Self-employed 1st category of labour: Pension Fund 1st category-14.02% PF rate for persons born before 1960 8.78% 2nd category-14.02% 19.8% 3rd category-11.02% THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Appendix table 7. Contribution rates by pension fund (continued) 69 Fund Employee Employer Self-employed Appendix 1st category-11.22% PF rate for persons born before 1959 6.58% 2nd category-11.22% 14.8% 3rd category-8.22% Supplementary Mandatory Pension Insurance Fund (Universal Pension Funds) SMPIF rate for persons born before 1960 5% 1st category -12.0% PPF (Professional Pension Fund) 2nd category-7.0% 3rd category-N/A 1st category -N/A UPF (Universal Pension Fund) 2nd category-N/A 3rd category-N/A SMPIF for persons born after 1959 5% 1st category -12.0% PPF (Professional Pension Fund) 2nd category-7.0% 3rd category-N/A 1st category -2.8% UPF (Universal Pension Fund) 2.20% 2nd category-2.8% 3rd category-2.8% Source: Bulgaria EUROMOD Country report. › Fund: General Sickness and Maternity Fund Conditions: This SIC is used to provide eligibility of the employee to benefits in cases of general sickness or maternity. Contribution to the fund is mandatory for all employees hired on a labour contract. For civil servants, (state administration, military, judicial system etc.) the total amounts of compulsory SIC are paid by the state. The following categories of insured are not obliged to pay contributions to this fund: (1) free-lancers and craftsmen; (2) persons performing work as sole traders, owners or partners in commercial corporations and natural persons who are members of unincorporated associations; (3) PhD candidates unless insured on other grounds; (4) registered farmers and tobacco producers; (5) people who perform work without entering into an employment relationships (so-called “civil contracts”) and who receive a monthly remuneration equal to or exceeding one minimum wage less the expenses for the activity for standard deduction—unless insured on other grounds during the respective month; (6) people who perform work without entering into an employment relationship (so-called “civil contracts”) and who are insured on other grounds—regardless of the amount of the remuneration received. Insured mentioned in points (1), (2) and (4) can pay contributions to the Fund on a voluntary basis. 70 Contribution rates: Employee 1.4% Appendix Employer 2.1% Self-employed (voluntary) 3.5% › Fund: Unemployment Fund Conditions:  The contribution to the Unemployment Fund is used to provide eligibility of the employee to cash benefits in cases of unemployment spells. Contribution to the fund is mandatory for all employees hired on a labour contract. For civil servants, (state administration, military, judicial system etc.) the total amounts of compulsory SIC are paid by the state. The following categories of insured are not obliged to pay contributions to this fund: (1) free-lancers and craftsmen; (2) persons performing work as sole traders, owners or partners in commercial corporations and natural persons who are members of unincorporated associations; (3) PhD candidates unless insured on other grounds; (4) registered farmers and tobacco producers; (5) people who perform work without entering into an employment relationships (so-called “civil contracts”) and who receive a monthly remuneration equal to or exceeding one minimum wage less the expenses for the activity for standard deduction—unless insured on other grounds during the respective month; (6) people who perform work without entering into an employment relationship (so-called “civil contracts”) and who are insured on other grounds— regardless of the amount of the remuneration received. Insured mentioned in points (1), (2) and (4) can pay contributions to the Fund on a voluntary basis. Contribution rates: Employee 0.4% Employer 0.6% › Fund: Work Injury and Occupational Disease Conditions:  The contribution to this fund is between 0.4 and 1.1% and is paid solely by the employer. The contribution rate is determined by economic sector in the Public Social Insurance Budget Act. Contribution rates:  Paid solely by employer-0.7% › Fund: Health insurance Fund Conditions: The compulsory health insurance is the main source of funding for the public health system. Employees and employers pay different rates of contributions. Health insurance contributions are due by entrepreneurs and other self-employed for all months of the year. Only the recognized contributors are eligible to use the services of the state health system. The state budget contributes to the health fund for the pensioners, registered unemployed, individuals under 18 years, students in the tertiary education and public employees. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Contribution rates:  Varies 71 Appendix › Fund: Teachers’ Pension Fund Conditions:  The teachers insure themselves for old age in the Teachers’ Pensions Fund with a separate insurance contribution. The contribution is 4.3% and is paid solely by the employer. Contribution rates:  Paid solely by employer › Fund: Pensions for the individuals under art. 69 Conditions: T he fund has been established since 1 January 2016. The insurance in that fund is obligatory for military servicemen and several categories of civil servants performing their duties within the systems of Ministry of Inferior, Ministry of Defense and Ministry of Justice.  he contribution rate to the Fund depends on whether the insured have decided Contribution rates: T not to insure themselves in the Professional Pension Funds. › Fund: Guaranteed claims of workers and employees Conditions: T he insurance to this fund provides protection to workers and employees in case of insolvency of their employer. The contribution is paid by employers only. Since 2011 the contribution rate to this Fund has been fixed at 0.0% as the accumulated funds are sufficient to cover all arising expenditures on benefits. Contribution rates:  0.0% › Taxes  Personal Income Taxes (PIT) ›  Personal income tax /PIT/ is levied on the earnings of the individual defined in the system as “physical” (or “natural”) person. ›  PIT rates are governed by the Personal Incomes Taxes Act (PITA). ›  The period of taxation with PIT is one fiscal year and every individual with taxable income is obliged to submit a Tax Declaration about her income with the exception of the case that the individual has received income only from employment relations under a Labour Contract or as a kind of public servant—in this case it is assumed that the employer has calculated, withheld, and transferred the due PIT liability on behalf of the employee and submission of a Declaration is not necessary. 72 › T  here are a wide range of allowances (See Appendix table 7) ›  There is a specific tax allowance for young families with mortgage. The amount of the mortgage interest paid during the year is subtracted from the annual consolidated tax base. The tax Appendix allowance for mortgage interest can be applied, if the family fulfill the following conditions simultaneously: 1. The mortgage contract is signed by a person with a registry marriage. 2. At least one of the married partners is below 35 years old. 3. The home with the mortgage is the only dwelling owned by the family. ›  The mortgage interest tax allowance is valid only for interests paid on the first BGN 100 000 (approximately EUR 50 000) of the principal amount. ›  The mortgage interest tax allowance can be taken only by one of the partners. ›  Since 1 January 2017 a new specific tax allowance is introduced. It is connected with the non- cash transactions. The allowance is 1% from the tax duty, but no more than BGN 500. The tax allowance can be applied if: 1. All incomes are received by bank transactions. 2. The non-cash transactions are 80% or more from all incomes. › There are a wide range of sources exempt from taxation as defined by the Personal Income Tax Act: (a) all social insurance benefits and social assistance allowances (b) income from private pension insurance (after regularly obtaining the right for pension); (c) since the beginning of 2013 incomes from interest in local commercial banks on deposits are taxed automatically and paid directly from the bank account (there is no information about incomes from interest on deposits in the tax declaration) (d) income from interest on treasury bills and bonds issued by the central or local governments; (e) nsurance indemnities and compensations received; (f) income obtained by registered agricultural or tobacco producers from selling unprocessed agricultural goods; (g) as of 1 January 2017—prizes given in the form of a supplementary game or an insignifican amount of prize money from slot machines within the meaning of the Gambling Act, as well as the cash and the object profits from gambling games organized under a license issued under the Gambling Act. › After deriving the amounts of tax bases in each section, the total annual tax base is obtained as a sum of the tax bases from the sections. Tax base for different categories of workers are summarized in *. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES Appendix table 8. Tax allowances, 2018 73 Allowances (BGN Allowances (annual amounts) annual amounts) Appendix Standard deduction for permanently disabled persons, BGN (i.e. with 50% or higher level up to 7920 of disability) Deduction for voluntary social, unemployment, health and life insurance (max.% from the 10% consolidated tax base) Standard child deduction 1 child 200 2 children 400 3+ children 600 Permanently disabled child 2000 Deduction of bequests (max.% from the consolidated tax base) For social and health institutions, Red Cross, NGOs 5% For sponsoring cultural events or products 15% For the National Fund “Children’s Health” 50% Source: Bulgaria EUROMOD Country report. Appendix table 9. Personal income tax schedule, 2018 Income Type Rate All other types of income 10% Income from business or self-employed activities 15% Dividends 5% Interests from special savings accounts 8% Interests from deposits 8% Source: https://ec.europa.eu/taxation_customs/tedb/taxDetails.html?id=60/1514764800 Appendix table 10. Itemized tax base from main income sources Income source Contribution rate Earnings for Labour Contract Gross earnings (GE) SIC 8.78% HIC 3.20% Tax base=GE-SIC-HIC Earnings from Civil Contract Gross earnings (GE) Normatively recognized expenditure (NRE) 25% Taxable income (TI) = GE-NRE SIC 8.78% HIC 3.20% Tax base=TI-SIC-HIC 74 Appendix table 10. Itemized tax base from main income sources (continued) Income source Contribution rate Appendix Income from self-employment as sole entrepreneur Gross self-employment earnings SIC 19.8% HIC 8% Tax base=GE-SIC-HIC Other sources of self-employment income Gross self-employment earnings from each source (GE) NRE (separate by income source) Freelance professions 25% Agricultural activity (processed products) 40-60% Forestry and fishery 40% Income from crafts, trade with decorative plants 40% Artistic/performance activities 40% Income from authorship rights (scientific/art products) 40% Taxable income (TI) = GE-NRE SIC 19.8% HIC 8% Tax base=TI-SIC-HIC Gross income NRE 10% Taxable income (TI) = GE-NRE SIC N/A HIC 8% Tax base=TI-SIC-HIC Source: Bulgaria EUROMOD Country report. The tax allowances are subtracted from the consolidated tax base before the final calculation of PIT duty. The final tax base is obtained by reduction of the consolidated tax base by the amount of allowances. › Value added tax VAT is levied at a rate of 20 percent for most goods and services including on the importation of goods into Bulgaria and intra-European acquisitions. Accommodation provided at hotels and similar establishments including the provision of vacation accommodation and rental of places for camping or caravan sites are levied at a rate of 9 percent. › Zero-rated goods  include supplies of goods dispatched or transported to destinations outside the territory of the EU; international transport of goods and passengers; supplies linked to international transport and goods traffic; supplies for handling of goods; supplies involving handling of goods— treatment, processing or repair; supplies of gold for central banks; supplies linked to duty-free THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES trade; supplies of services provided by agents, brokers and other intermediaries linked to the 75 supply of the aforementioned zero-rated supplies; supplies by virtue of international treaties, agreements, accords, conventions; supplies of goods and services where recipients are armed Appendix forces of other countries which are parties to the North Atlantic Treaty; supplies of services linked to importation such as commissions, packing, transport and insurance; intra-European Union supplies; and intra-European supplies of new means of transport. › Exempt goods include supplies linked to healthcare; welfare and social security work; education, sport and physical education; culture; religious organizations; non-profit organizations; land and buildings; financial services, insurance and reinsurance services and gambling; supplies of postage stamps and postal services; and supplies of goods and services to which no input tax credit was applied. › Excise duties Excise duties are levied on various products (so called “excise goods”), e.g. alcohol and alcohol beverages, coffee, tobacco products, fuels and energy products (mineral oils, gas, etc.), electricity, and gambling. The amount of each excise duty is determined per unit of the naturally measured quantity of the respective excise good (e.g. per kilogram, liter, cubic meter, etc.). Appendix table 11. Excise duty schedule. 2018 Good Excise duty Petrol-Leaded BGN 830 per 1000 litres Petrol-Unleaded BGN 710 per 1000 litres Gas oil BGN 646 per 1000 litres Diesel BGN 646 per 1000 litres Heavy fuel oil BGN 400 per 1000 litres Liquid Petroleum Gas (LPG)-propellant BGN 340 per 1000 kg Liquid Petroleum Gas (LPG)-heating fuel BGN 0 per gigajoule Kerosene BGN 646 per 1000 litres Natural Gas-motor fuel BGN 0.85 per gigajoule Natural gas-businesses BGN 0.60 per gigajoule Natural gas-households BGN 0 per gigajoule Coal and Coke BGN 0.6 BGN per gigajoule Electricity-businesses BGN 2 per MWh Electricity-households BGN 0 per MWh Beer BGN 1.50/hl/°Plato Beer-Independent small breweries BGN 0.75/hl/°Plato Wine BGN 0/hl/°Plato Ethyl alcohol BGN 1,100 per hectolitre Cigarettes BGN 109 per 1000 pieces plus 25 percent ad valorem 76 Appendix table 11. Excise duty schedule. 2018 (continued) Good Excise duty Appendix Cigars and Cigarillos BGN 270 per 1000 pieces Fine cut smoking tobacco BGN 152 Other smoking tobacco BGN 152 Source: Excise Duty tables 2018, European Commission, https://ec.europa.eu/taxation_customs/tedb/taxDetails. html?id=4051/1514764800 (http://ec.europa.eu/taxation_customs/index_en.htm) › Other taxes Property (real estate) tax Property tax is a local tax levied on the so called “tax estimate” of the value of any real estate object (building, land spot, etc.). Its revenues are allocated directly to the respective local (or municipal) budget. The tax on buildings is determined along with the sewage fee using differentiated rates on the basis of the tax estimate. The local governments (Municipal Councils) are allowed to set the local tax rates by their own legal act in established by the National Parliament limits (minimum and maximum thresholds). Inheritance tax Inheritance tax is also a local tax and is a version of a tax on capital transfer. It is a liability of both the heritor and heritage-giver and emerges as payable at the time of enacting the inheritance. Any kind of assets (both non-financial and financial) can be objects of the tax and its rate is determined differentially for each inheritor depending on her relation to the heritage-giver and the legal evaluation of the amount inherited. If the recipients are the partner and/or/ the children there is no liability for inheritance tax. In other cases the rates vary between 0.4% and 6.6% for amounts higher than BGN 250,000. Motor vehicle tax Motor vehicle tax is also a local tax and is determined by the Municipal Councils. The local government can set the unit tax as much as three times its minimum threshold. The tax is due to the owners of motor vehicles and is determined according to the power of the engine corrected by a coefficient depending on the year of its first registration. Corporate profit ta Corporate tax is due by all resident companies—Bulgarian or foreign controlled—that are registered as legal entities under the Commercial Law. Also, non-profit organizations or public bodies that have own divisions operating as business units are also subject to the corporate tax. For the period 2011–2020 the tax rate is 10% and is levied on the corporate profit value specifically assessed by a financial accounting procedure under the regulations of the Accounting Law and the Law on Corporate Taxation. THE DISTRIBUTIONAL IMPACT OF TAXES AND SOCIAL SPENDING IN BULGARIA WITH AN APPLICATION TO GREEN FISCAL POLICIES References 77 Appendix Cazcarro, I., Amores, A., Arto, I. & Kratena, K. 2022. “Linking multisectoral economic models and consumption surveys for the European Union”. Economic Systems Research, 34:1, 22-40, DOI: 10.1080/09535314.2020.1856044. European Commission. 2018, “Excise duty tables: Excise duty - Tobacco products (EU harmonised)”, updated on 2019/05/14. Ivanova, D., & Wood, R. 2020. “The unequal distribution of household carbon footprints in Europe and its link to sustainability”. Global Sustainability, 3, E18. doi:10.1017/sus.2020.12. Jellema, J., and Inchauste, G.. 2018. “Constructing Consumable Income: Including the Direct and Indirect Effects of Indirect Taxes and Subsidies.” Chapter 7 in Commitment to Equity Handbook: Estimating the Impact of Fiscal Policy on Inequality and Poverty, edited by Nora Lustig. Brookings Institution Press and CEQ Institute, Tulane University. www.commitmentoequity.org. Lustig, N. 2018. “Commitment to Equity Handbook. Estimating the Impact of Fiscal Policy on Inequality and Poverty”. Brookings Institution Press and CEQ Institute, Tulane University. OECD. 2021. Trade in embodied CO2 (TeCO2) Database. Steen-Olsen, K., Wood, R. and Hertwich, E.G., 2016. The Carbon Footprint of Norwegian Household Consumption 1999-2012. Journal of Industrial Ecology, 20(3), pp.582–592. Vernon, N., Black, S., Parry, H. 2021. “Still Not Getting Energy Prices Right: A Global and Country Update of Fossil Fuel Subsidies,” IMF Working Papers 2021/236, International Monetary Fund World Bank. 2015. Fossil Fuel Subsidies: Approaches and Valuation. World Bank, Washington D.C. © World Bank. World Bank. 2021. Bulgaria Systematic Country Diagnostic. World Bank, Washington, DC. © World Bank. https://openknowledge.worldbank.org/handle/10986/36842 License: CC By 3.0 IGO. Green Transition. Washington, DC. © World Bank. https://openknowledge.worldbank.org/ handle/10986/37308 License: CC By 3.0 IGO. World Bank. 2022. Green Fiscal Reforms: Part Two of Strengthening Inclusion and Facilitating the 78 Appendix CEQ: Argentina: Rossignolo, Dario. 2017. “CEQ Master Workbook: Argentina. Version: May 19, 2017,” CEQ Data Center on Fiscal Redistribution, CEQ Institute, Tulane University, New Orleans, LA. Armenia: Younger, Stephen D., and Artsvi Khachatryan. 2014. “CEQ Master Workbook: Armenia. Version: May 31, 2014.” CEQ Data Center on Fiscal Redistribution, CEQ Institute, Tulane University, New Orleans, LAm and the World Bank, Washington, DC. Belarus: Bornukova, Kateryna, Gleb Shymanovich, and Alexander Chubrik. 2017. “Fiscal Incidence in Belarus: A Commitment to Equity Analysis.” Policy Research Working Paper 8216, World Bank, Washington, DC. Croatia: Inchauste, Gabriela and Ivica Rubil. 2017. “The Distributional Impact of Taxes and Social Spending in Croatia.” World Bank, Washington, DC. Mexico: Scott, John. 2013. “CEQ Master Workbook: Mexico. Version: September 2, 2013.” CEQ Data Center on Fiscal Redistribution, CEQ Institute, Tulane University, New Orleans, LA. Poland: Goraus, Karolina, and Gabriela Inchauste. 2016. “The Distributional Impact of Taxes and Transfers in Poland”. Policy Research Working Paper 7787. World Bank, Washington D.C. Russia: Lopez-Calva, Luis F., Nora Lustig, Mikhail Matytsin, and Daria Popova. 2017. “Who Benefits from Fiscal Redistribution in the Russian Federation?” Working Paper 39, CEQ Institute, Tulane University, New Orleans, LA. Romania: Inchauste, Gabriela, and Eva Militaru. 2018. “The Distributional Impact of Taxes and Social Spending in Romania.” World Bank, Washington, DC. Country Studies: EUROMOD: Bulgaria: 2017-2020 Ekaterina Tosheva, Iva Tasseva, Desislava Dimitrova, Dragomir Draganov, Venelin Boshnakov December 2020 EUROMOD version I3.0