ECONOMY AND FINANCE INEQUALITY IN BRAZIL Income, Wealth and Tax Distribution Research Center on Macroeconomics of Inequalities(Made) June 2024 The authors explore Brazil’s multidimensional social in­ equalities that have strong regional, gender and, most im­ portantly, racial aspects con­ nected to historical social pow­ er relations and exploitation. The Brazilian tax system is markedly regressive. Labour is disproportionately taxed in comparison to capital and the widespread use of consump­ tion taxes disproportionately burdens low-incomes. The au­ thors argue that a progressive tax reform should seek to eliminate such disparities by eliminating privileges granted to the rich by the state. The government of president Lula has promised to reform taxation of the super-rich and is being supported by civil so­ciety in an active campaign. Brazil has put the creation of a minimum global tax rate on billionaires on the G20 agenda. An excise tax of two per cent is estimated to gene­ rate almost R$ 42 billion in taxes in a year.

 Content 2 INCOME AND WEALTH INEQUALITY IN 3 Income distribution 3 The profiles of the rich  4 Typical sources of income  7 Wealth inequality in Brazil  7  OVER LABOUR, WHITE OVER BLACK: THE REGRESSIVE BRAZILIAN TAX 8 Regressive taxes today and 20 years ago 8 Potential impacts of alternative tax reforms  9 Government transfers: progressive overall, but with heterogeneous individual effects  10 Tax evasion  10 POLICY 11 References 12

FRIEDRICH-EBERT-STIFTUNG – INEQUALITY IN BRAZIL INTRODUCTION This paper offers an introductory overview of inequality in the largest national economy of Latin America, focusing on key dimensions such as income, wealth and tax distribution. First, the recent literature on particular aspects of Brazilian inequalities is examined, with a special focus on findings re­ ported in publications by the University of São Paulo’s Re­ search Center on Macroeconomics of Inequalities(MadeFEA/USP). As evidence shows, Brazil’s severe inequalities stubbornly persist over time. The country’s multidimension­ al social inequalities have strong regional, gender and, most importantly, racial aspects that are connected to historical social power relations and exploitation. Brazilian govern­ ments and policymakers have made little attempt over time to reduce these structural inequalities through taxation (Schiozer et al., 2021). This would require extensive tax re­ forms, as the Brazilian tax system is markedly regressive. For instance, labour is disproportionately taxed in comparison to capital(Goto; Pires, 2022) and the widespread use of con­ sumption taxes disproportionately burdens low-income strata(Silveira et al., 2022; Orair, 2022; Cardoso et al., 2022). We argue that a progressive tax reform should seek to elim­ inate such disparities by reversing historical and institutional privileges granted to the rich by the state.

Income and wealth inequality in Brazil INCOME AND WEALTH INEQUALITY IN BRAZIL INCOME DISTRIBUTION Given the sparse – or unreliable – evidence relating to wealth inequality in Brazil, we chose to start our analysis with a fo­ cus on the various sources of information on income ine­ quality. In this regard, some recent studies using administra­ tive data from tax records have shown that household sur­ veys systematically underestimate income concentration at the top of the distribution structure(Medeiros et al., 2015; Gobetti; Orair, 2017). Bottega et al.(2021a) combine data from the Brazilian consumer expenditure survey( Pesquisa de Orçamentos Familiares, POF) for 2017–2018, and person­ al income tax statements(DIRPF), adjusted for total family income obtained from the System of National Accounts (SNA). They find that income appropriation by the 0.5 per cent at the top of the pyramid is 2.47 times higher when in­ come for the top of the distribution pyramid is corrected 1 . 1 Along these lines, De Rosa et al.(2022, p. 32) find that Brazil is one of the Latin American countries for whom“inequality trends during the high-growth years(2003–2013) change after the survey’s reported Using the same methodology, Bottega et al.(2021b) esti­ mate that, for individuals aged 18 and over with positive income, the wealthiest 10 per cent appropriate 53.7 per cent, or more than half, of national personal income, which is 106 times the share appropriated by the poorest 10 per cent(Graph 1). Appropriation by the top 1 per cent and 0.1 per cent is even more startling: 24.6 per cent(almost a quarter) and 12.1 per cent of national personal income, re­ spectively, are concentrated in this group. Thus, even among the richest strata, we can observe a high degree of inequality, since the top 1 per cent captures almost half of income earned by the top 10 per cent while, similarly, the top 0.1 per cent accounts for almost half of the top 1 per cent’s income(Table 1). income is augmented to include ignored top incomes from admin­ istrative data and macroeconomic incomes of the household sector and total economy from the national accounts.(...). More­over, dur­ ing the low-growth years at the end of our period of analysis(post2015), inequality has increased faster in the augmented series than in the raw series”. Graph 1 Appropriation of national income by income deciles and demographic composition 60 Appropriation of total income by group (%) 50 40 30 20 10 0 1 2 3 4 5 6 7 8 9 10 Position in Income Distribution White men White women Black men Black women Source: Made-FEA/USP(Bottega et al., 2021b, Graph 3) 3

FRIEDRICH-EBERT-STIFTUNG – INEQUALITY IN BRAZIL Table 1 Appropriation of national income by selected income strata and demographic composition Income strata Demographic group 0–90% 10% 1% 0.1% Black women 10.3% Black men 12.6% White women 11.2% White men 11.7% Source: Made-FEA/USP(Bottega et al., 2021b, Table 1, translated). 4.0% 8.0% 13.9% 27.8% 1.2% 2.2% 5.9% 15.3% 0.5% 0.6% 3.0% 8.0% These results are in line with the evidence provided by the World Inequality Database(WID) for Brazil 2 : between 2001 and 2019, the top 10 per cent share of pre-tax income 3 for individuals aged 20 and over fluctuated between 60.9 per cent to 57.1 per cent, exhibiting remarkable stability over the past two decades. In the same vein, the share of the top 1 per cent rose from 23.7 per cent in 2001 to 25.4 per cent in 2012, then trending downwards in the following years, reaching 20.3 per cent of pre-tax national income in 2019. By contrast, the bottom 50 per cent appropriated an income share of 8.6 per cent, in 2001, 10.8 per cent in 2013, and 10.1 per cent in 2019. corresponds to 42 per cent of the 10 per cent poorest, be­ ing overrepresented, alongside black men, at the bottom of the income distribution. Thus, racism and racial hierarchies manifested in the social division of labour seem to correlate with income inequality more than gender relations, even though there is also evidence that men in general as well as within each racial category outnumber women among the top earners. Still, white women represent a greater portion of the top 10 per cent, 1 per cent and 0.1 per cent, than black men: 27 per cent, 23 per cent and 15 per cent, in contrast to 20 per cent, 13 per cent and 10 per cent, respectively(Graph 2; Graph 3). THE PROFILES OF THE RICH Looking at the profiles of the rich, Bottega et al.(2021b) show that the top of the income distribution structure is characterised by gender and racial inequalities. First, white men make up 42 per cent of the highest income decile, 57 per cent of the top 1 per cent, and 69 per cent of the top 0.1 per cent, while black women only account for 10 per cent, 6 per cent and again 6 per cent of these top income earners, respectively. We should note that black women represent 26.6 per cent and white men 22.6 per cent of the Brazilian population, so it is not surprising that the former Second, and from another related perspective, national per­ sonal income is disproportionately appropriated by a few white men(Bottega et al., 2021b). White men among the top 1 per cent(around 705,000 individuals or 0.57 per cent of the Brazilian population) appropriate a greater share of na­ tional income(15.3 per cent) than all black women together, who number 32.7 million individuals, but appropriate only 14.3 per cent of national income. Similarly, white men among the top 10 per cent, representing 4.2 per cent of the popula­ tion, absorb 27.8 per cent of national income. This figure is higher than the share of both white women and black men, who correspond to 23.3 per cent and 27.6 per cent of the population, and obtain 25.1 per cent and 20.6 per cent of national income, respectively(Graph 4; Table 1). 2 For details on the methodology used by WID researchers to esti­ mate income inequality in a vast sample of countries, see Bajard et al.(2021), Chancel and Piketty(2021), Chancel et al.(2023). For Latin American countries in particular, see also De Rosa et al.(2022). 3“Pre-tax national income is the sum of all pre-tax personal income flows accruing to owners of the production factors labour and cap­ ital before taking into account the operation of the tax/transfer sys­ tem, but after taking into account the impact of the pension system. The key difference between personal factor income and pre-tax in­ come is the treatment of pensions, which are counted on a contribu­ tion basis by factor income and on a distribution basis by pre-tax in­ come. The population is comprised of individuals over age 20. The base unit is the individual(rather than the household) but resources are split equally within couples”. For more details, see Sources& In­ formations in https://wid.world/country/brazil/. Third, among white men, income is more concentrated at the top in comparison to other demographic groups(Botte­ ga et al., 2021b): the income share appropriated by the highest decile within each demographic group is 60 per cent among white men, 53 per cent among white women, 44 per cent among black men and 42 per cent among black women. Similarly, the top 1 per cent within each demo­ graphic group appropriates 29 per cent, 24 per cent and 16 per cent of each group’s total income for white men, white women, and black people, respectively. As should be expected, income inequality among black people is less pro­ nounced, although still high. 4

Income and wealth inequality in Brazil Graph 2 Demographic composition of the total population and of each income decile Demographic composition (%) 100 26 80 60 27 40 23 20 22 25 22 17 14 10 42 36 31 34 32 20 24 29 30 29 33 24 28 27 30 27 28 25 23 24 20 20 26 22 19 34 42 22 24 29 11 14 17 16 17 0 Population 1 2 3 4 5 6 7 8 9 10 Total population and income deciles White men White women Black men Black women Source: Made-FEA/USP(Bottega et al., 2021b, Graph 1) Demographic composition (%) Graph 3 Demographic composition of the total population and of selected strata from the top of the income distribution structure 100 10 6 6 26 13 10 80 20 15 23 60 27 27 40 23 69 57 20 42 22 0 Population 10 1.0 0.1 Income quantiles(%) White men White women Black men Black women Source: Made-FEA/USP(Bottega et al., 2021b, Graph 2) Graph 4 Appropriation of national income by selected income strata and demographic traits 60 Appropriation of total income by the top (%) 50 40 30 20 10 0 0–90 10 1.0 0.1 Quantiles(%) White men White women Black men Black women Source: Made-FEA/USP(Bottega et al., 2021b, Graph 4) 5

FRIEDRICH-EBERT-STIFTUNG – INEQUALITY IN BRAZIL Graph 5 Income composition of the top centile according to occupational status 100 Proportion (%) 80 60 40 20 0 Black men White men Demographic group White women Black women Domestic worker Military Private sector employee Public servant Employer Self-employed worker Non-remunerated worker helping another family/household member Source: Made-FEA/USP(Gomes et al., 2022, Graph 3) Graph 6 Average Composition of Wealth 100 Proportion (%) 80 60 40 20 0 75–80 80–85 85–90 Financial Assets Source: Made-FEA/USP(Martins, Arthen and Gomes, 2024, Graph 2) 90–95 96 97 98 Position in Income Distribution(%) 99 99.01– 99.8 Real Estate Mobile Assets Other assets and rights 99.81– 100 6

Income and wealth inequality in Brazil TYPICAL SOURCES OF INCOME To investigate the typical sources of income for the rich, we build on the findings of Gomes et al.(2022). It should be pointed out that, in order to evaluate the heterogeneous in­ come compositions of distinct demographic groups, espe­ cially with regard to gender and racial hierarchies, the au­ thors opted not to adjust data obtained from POF 2017– 2018 with the DIRPF administrative data. This allowed them to be able not only to look at the different sources of in­ come for the populations of interest, but also to calculate the tax rate for different demographic groups according to their relative position in the income distribution structure. For both purposes, however, it is important to remember that household surveys such as POF, as mentioned, underes­ timate the true level of income inequality. Notwithstanding these limitations, one of the main findings of Gomes et al.(2022) is that among the top 1 per cent there are significant differences between white men and other demographic groups with respect to the main income sources(Graph 5). More specifically, while white men in the top percentile earn almost half of their income as self-de­ clared employers, black men in the same income group earn only 7 per cent of their income as self-declared employers. It is mostly employment in the formal and public sectors that generate black people’s and white women’s income 4 . In contrast, the income share of public servants for white men in the top percentile is 19 per cent. Moreover, formal work­ ers appropriate much less: 36 per cent of total income. This has important consequences for taxation. WEALTH INEQUALITY IN BRAZIL and are primarily derived from an extrapolation of income inequality data rather than on direct sources of information regarding asset distribution. Given the difficulty of finding adequate data sources on the distribution of wealth in Bra­ zil, most of the estimates for wealth inequality are unrelia­ ble or at least subject to methodological limitations. As Me­ deiros(2015) emphasises, evidence based on DIRPF(per­ sonal income tax statements) usually underestimates the wealth of the rich, given the undervaluation of self-de­ clared assets(usually declared by their original purchase val­ ue and not their true market value), tax evasion, patrimoni­ al confusion with legal entities and fraud. Also, the impos­ sibility of individualising the wealth of married couples inev­ itably distorts the results, regardless of the methodology used to separate it. Despite these difficulties, by combining PNADc(research based on a sample of Brazilian households) with DIRPF, one can corroborate and gather important information regard­ ing income and wealth profiles. From these combined data, Martins, Arthen and Gomes(2024) found that only 27 per cent of Brazilian individuals possess some wealth. However, this wealth is strongly concentrated at the top of the distri­ bution structure, with 0.2 per cent of wealthiest individuals having an average wealth of R$ 13 million, whereas the av­ erage wealth of the population as a whole is R$ 97.208. Fur­ thermore, in addition to the unequal distribution of Brazilian wealth among individuals, the composition of wealth is striking, i. e. how it is distributed according to each quantile of income share. One may see, in Graph 6, that as we pro­ gress in income distribution towards 0.2 per cent, the pro­ portion of physical assets decreases, whereas financial as­ sets become the most important form of wealth. According to the World Inequality Database(WID), in 2001, 73.7 per cent and 41.1 per cent of Brazilian national wealth was concentrated in the hands of the top 10 per cent and top 1 per cent, respectively. The share of the bottom 50 per cent, on the other hand, was only 1.2 per cent. Almost two decades later, in 2019, the top 10 per cent and the top 1 per cent appropriated 79.6 per cent and 48.3 per cent of na­ tional wealth, respectively, while the bottom 50 per cent registered negative participation in national wealth at –0.4 per cent, which means that a remarkable wealth ine­ quality is both cause and effect of a profound income ine­ quality in Brazil. These results should be interpreted with caution since they are based on a series of imputations(Bajard et al., 2021) 4 It is worth noting that black women stand out as the only demo­ graphic group with a reduced participation of self-employed work­ ers among the top 1 per cent earners: less than 10 per cent of their total income is appropriated by these workers, while this share rises to around 20 per cent for white people and black men. At the same time, the share of income related to self-declared employ­ ers is higher for black women(almost 20 per cent) in comparison to black men and white women. Since black women represent a small fraction of top earners, these differences might be related to eco­ nomic phenomena. 7

FRIEDRICH-EBERT-STIFTUNG – INEQUALITY IN BRAZIL CAPITAL OVER LABOUR, WHITE OVER BLACK: THE REGRESSIVE BRAZILIAN TAX SYSTEM REGRESSIVE TAXES TODAY AND 20 YEARS AGO The Brazilian tax system is regressive: 20 years ago, the poorest decile was subject to a tax rate of 32 per cent (28.3 per cent indirect and 3.7 per cent direct taxes), while the richest decile paid a tax rate of 22 per cent(10 per cent indirect and 12 per cent direct taxes)(POF 2002–2003, Sil­ veira, 2010). The top 10 per cent at that time had an aver­ age per capita income almost 90 times higher than the av­ erage per capita income of the poorest decile. Given the extent of inequality before taxation, we can on­ ly conclude that the progressivity of direct taxes is, at best, modest, especially if we consider their reduced share of overall tax revenues: 22.5 per cent, in contrast to a 27.7 per cent share in the case of payroll taxes and a 44 per cent share for indirect taxes in 2019(Cardoso et al., 2022). A tax reform confronting this is thus extremely urgent, especially with regard to personal income tax(IRPF) and its unex­ plored potential, since Silveira(2010) estimates that taxes on motor vehicles(IPVA) and on the ownership of urban real estate properties(IPTU) are neutral or even regressive. There has been little change in the regressivity of the Bra­ zilian tax system in the last decades(Silveira et al., 2022). In 2017–2018, the tax rate applied to the poorest decile was 26.4 per cent(23.4 per cent indirect taxes and 3.1 per cent direct taxes), in comparison to a tax rate of 19.2 per cent(8.6 per cent indirect taxes and 10.6 per cent direct taxes) for the top 10 per cent. Again, it is worth noting that the average per capita income of the richest decile is 36.4 times higher than the average per capita income of the poorest one, which underscores the limited progres­ sivity of direct taxes. Still, in the case of personal income tax( Imposto sobre a Renda da Pessoa Física or IRPF), 80 per cent of it comes from the top 10 per cent, and 97 per cent from the top 30 per cent, enhancing its strate­ gic importance in any future tax reform. The challenge when it comes to the redistributive capacity of the IRPF, however, is its regressive effect on the top 1 per cent who are subject to the tax, in comparison to the richest decile in general. Taking a closer look at the top 1 per cent, Gomes et al. (2022) find that black men at the top have a higher tax bur den than white men at the top 5 . Moreover, personal income tax in Brazil is regressive for white men at the top and pro­ gressive for black men at the top 6 . Looking at the profiles of a“rich white person” and a“rich black person” in Brazil, it is easy to understand this discrepancy. Black men at the top are usually public servants, and therefore pay the tax rate as­ sessed on labourers. White men at the top, on the other hand, are more frequently in the position of employers in social relations of production, which means they are usually remunerated and taxed as capital(Graph 7). In fact, Silveira(2010) had previously shown that direct tax­ es placed a disproportionate burden on wage workers, in comparison to employers and own-account workers. Goto and Pires(2022) also present evidence in this regard: if we add payroll taxes up to the highest marginal tax rate for per­ sonal income tax(IRPF), the tax rate rises from 27.5 per cent to 55.5 per cent 7 . This approach allows the authors to simu­ late the tax rate of individuals with the same monthly in­ come(R$ 10,000.00 8 ), but working in different occupa­ tions, and they find that a formal employee and a public servant bear a tax burden of 42.3 per cent and 38.1 per cent, respectively, while a partner who provides services for a company operating under the presumptive income tax re­ gime faces a tax burden of 16.3 per cent. This huge discrep­ ancy has its roots not only in the exemption of profits and dividends distributed to individuals(Law nº 9.249/1995), but also in the“poorly calibrated parameters of simplified taxa­ tion regimes for small and micro-enterprises” 9 (Orair, 2022, 5 Still, white people are responsible for most of the burden in connec­ tion with direct taxes(Gomes et al., 2022), which is simply a conse­ quence of the overrepresentation of white people at the top of dis­ tribution structure. 6 White men among the top percentile have a tax rate of 8.8 per cent with regard to direct taxes, which is less than white men in the next lower top 9 per cent, who face a tax rate of 10.6 per cent, whereas black men among the top percentile bear a tax rate of 13.1 per cent, more than black men at the next lower top 9 per cent, who pay a tax rate of 10.7 per cent. This also seems to be the case for women, but the reduced size of the sample for black women does not allow one to draw any conclusions in this regard(Gomes et al., 2022). 7 It is worth noting that, while the highest IRPF marginal tax rate (27.5 per cent) is considerably lower than the OECD average(39.7 per cent), the tax rate including payroll taxes(55.5 per cent) is significantly higher than the OECD average(44.5 per cent)(Goto; Pires, 2022). 8 Around US$ 2,000.00 in September 2023. 9 Our translation. 8

Capital over labour, white over black: the regressive Brazilian tax system p. 77). As a consequence, we can observe in Brazil the phe­ nomenon of pejotização, or the migration from employ­ er-employee relationships to an independent-contractor-­ only hiring policy(Goto; Pires, 2022; Orair, 2022). POTENTIAL IMPACTS OF ALTERNATIVE TAX REFORMS In this scenario, Bottega et al.(2021a) and Bottega et al. (2021b) simulate the impacts of an eventual elimination of the unjustified privilege concerning distributed profits and dividends. Considering alternative tax rates of 15 per cent and 20 per cent, Bottega et al.(2021a) find that a profit and dividends tax would not have a substantial effect on Brazil’s Gini coefficient. Even so, according to Bottega et al.(2021b), a 15 per cent tax rate on profits and dividends plus an addi­ tional tax rate of 35 per cent with respect to IRPF for the top 1 per cent 10 could generate R$ 46 billion(around US$ 9.2 billion in September 2023) in tax revenues. In other words, even if a progressive tax reform has limited capacity to directly reduce inequality, it may introduce a new source of revenue to defray government expenses, and in particular redistributive policies, including monetary transfers such as Bolsa Família, known for their success in mitigating extreme poverty and reducing inequality(Soares et al., 2007; 2009; 2010; Barros et al., 2007; Silveira, 2010; Hoffmann, 2013; Silveira et al.; 2020; 2022). Silveira et al. 10 People with monthly income above R$ 26,857.00 (around US$ 5,371.00 in September 2023). (2022) show how the Brazilian government expanded its ca­ pacity to redistribute income through social expenses during the first decade of the 2000s: transfers and taxes reduced the Gini coefficient of market income inequality by 15 per cent(POF 2002–2003) and 21.8 per cent, respectively(POF 2008–2009). Moreover, government spending on public services such as health and education has a significant im­ pact when it comes to reducing inequalities, as these ex­ penditures benefit low-income strata disproportionately(Sil­ veira; Palomo, 2023). But even if we consider a fiscally neutral tax reform, the redis­ tributive potential offered by a profit and dividends tax(15 per cent or 20 per cent) is substantial: such an increase in tax rev­ enues could secure a corresponding reduction in consump­ tion taxes. Cardoso et al.(2022) simulate these effects, show­ ing that such a reform would have a net expansionary effect on GDP(1.2 per cent in the case of a 15 per cent tax rate, and 1.9 per cent with a 20 per cent tax rate), since it stimulates consumption. Consequently, even if the reform were fiscally neutral, it would have a positive effect on tax revenues. More­ over, it would promote an expansion of employment(2.14 per cent and 3.25, respectively) and increase the disposable in­ come of all income brackets, except the highest one(above 30 minimum wages), in which 82 per cent of income derived from profits and dividends is concentrated. Another tax reform that has begun to gain traction during Brazil’s temporary presidency of the G20 is the creation of a minimum global tax rate on billionaires. This proposition originates with Chancel et al.(2022) idea of implementing a minimum global tax of 2 per cent on the wealth possessed by this same group. Along the same lines, Martins, Arthen Graph 7 Effective Excise Income Tax – Current and After Wealth Tax on the top 0.2 per cent 15 Effective Excise Tax (%) 10 5 0 75–80 80–85 85–90 90–95 96 97 98 Position in Income Distribution(%) Current 2 per cent Wealth Tax 2.5 per cent Wealth Tax Source: Made-FEA/USP(Martins, Arthen and Gomes, 2024, Graph 5) 9 99 99.01 99.81 –99.8 –100 3 per cent Wealth Tax

FRIEDRICH-EBERT-STIFTUNG – INEQUALITY IN BRAZIL and Gomes(2024) have estimated potential tax revenues that would accrue to the Brazilian government, as well as potential distributional effects in the event this proposal is implemented for the top 0.2 per cent of the income distri­ bution pyramid. They modeled various scenarios with differ­ ent excise taxes, i. e. 2 per cent, 2.5 per cent and 3 per cent on wealth, with the revenue obtained corresponding to the difference between the excise tax and how much the quan­ tile already pays in income taxes as a proportion of their wealth. The smaller amount would generate almost R$ 42 billion in taxes in a year, whereas the other two would pro­ vide approximately R$ 60 billion and R$ 78 billion, respec­ tively, in the same period. When one considers the strong regressivity of Brazil’s effective tax incidence on income when we look beyond the top 1 per cent in the distribution structure, it is apparent that the imposition of an excise tax greater than 2.5 per cent could reverse this pattern and re­ turn to the progressive pattern observed up until the richest 1 per cent of individuals. GOVERNMENT TRANSFERS: PROGRESSIVE OVERALL, BUT WITH HETEROGENEOUS INDIVIDUAL EFFECTS Regarding the effects of government transfers on income inequality, Gomes et al.(2022) present some interesting ev­ idence. Echoing an extensive literature(Soares et al., 2007; 2009; 2010; Barros et al., 2007; Silveira, 2010; Hoffmann, 2013; Silveira et al.; 2020; 2022), they show that, taken to­ gether, government transfers are progressive, that is, they benefit the lower strata disproportionately, therefore reduc­ ing inequality. For instance, they represent around 25 per cent of the total income of the lowest decile, less than 20 per cent of the income appropriated by the highest de­ cile, and around 15 per cent of earnings by the top 1 per cent. Three types of monetary transfers are responsible for this result: the world-famous Bolsa Família, the Benefício de Prestação Continuada or BPC, which guarantees a monthly payment equivalent to a minimum wage for the elderly and disabled; and the public pension regime for retired formal workers, RGPS( Regime Geral de Previdência Social). TAX EVASION According to estimates by the National Union of National Treasury Attorneys(SINPROFAZ) 11 , between 1 January and 31 December, Brazil lost R$ 626.8 billion 12 in 2022 due to tax evasion, which amounts to 6.3 per cent of GDP in the same year. Thus, a significant share of tax revenues is lost to the Brazilian government budget every year, reducing its ca­ pacity to provide public goods and services for the popula­ tion. Tax evasion is related to the size of the“shadow econo­ my” 13 . Medina and Schneider(2018) estimate that the Bra­ zilian shadow economy amounts to 37.6 per cent of GDP (the average over 1991–2015). In comparative terms, the largest shadow economies are located in Zimbabwe (60.6 per cent) and Bolivia(62.3 per cent), while the small­ est ones account for 8.9 per cent and 7.2 per cent of Aus­ tria’s and Switzerland’s GDP, respectively. The average size for all 158 countries included in the sample is 31.9 per cent, which means that Brazil assumes an intermediate position on a global scale. In turn, Clemente et al.(2021) find that Brazil is one of the Latin American countries with the lowest probability of tax evasion: 18 per cent, in comparison to 68 per cent, 67 per cent, 48 per cent and 28 per cent for Argentina, Uruguay, Bolivia and Peru, respectively. Still, the likelihood of tax eva­ sion in the case of Brazil is higher than for European coun­ tries such as Germany(10 per cent), Denmark(11 per cent), France(11.7 per cent) and Finland(13 per cent). The authors argue that Latin American countries have a tax structure that favours tax evasion. Indeed, the Brazilian tax system is complex and not very transparent, so“the difficulty in mon­ itoring and auditing heightens the opportunity for tax eva­ sion”(Clemente et al., 2021, p. 11). Thus, a tax reform that simplifies the tax structure, such as a reform of indirect tax­ es currently under discussion in the National Congress, could help mitigate this situation. On the other end of the spectrum, the public pension re­ gime for retired public service workers( Regime Próprio de Previdência Social or RPPS) is regressive: it is a source of in­ come overrepresented among the top 10 per cent, and es­ pecially the top 1 per cent of the income distribution struc­ ture. This regressivity is more pronounced within the demo­ graphic groups of white women and black men, since top earners within these groups are usually public service work­ ers. Last but not least, one important feature of government transfers is its pro-female nature across all income deciles, but especially for the top 9 per cent(excluding the top 1 per cent) of earners in each demographic group: these transfers (mainly because of the RPPS and the RGPS) account for 24 per cent and 22 per cent of the total income of black and white women, respectively, in comparison to 18 per cent of the total income of black and white men among the top 9 per cent(Gomes et al., 2022). 11 https://www.sinprofaz.org.br/tag/sonegometro/(Retrieved Septem­ ber 30, 2023). 12 Around US$ 125.4 billion in September 2023. 13“[S]hadow economic activities may be defined as those economic activities and income earned that circumvent government regulation, taxation or observation(...); hence all productive economic activities that would generally be taxable were they reported to the state(tax) authorities”(Medina; Schneider, 2018, p. 5). 10

Policy recommendations POLICY RECOMMENDATIONS The fragility of information concerning wealth inequality in Brazil translates into a need for official data producers, such as the Brazilian Institute of Geography and Statistics( Instituto Brasileiro de Geografia e Estatística , or IBGE), in coopera­ tion with other government institutions, to provide statistics on the distribution of wealth in the country. Such an en­ deavour is a crucial step in the further development and re­ finement of future research, and essential if better-informed public policies are to be implemented. Still, there is plenty of evidence that the Brazilian tax system reinforces existing structural inequalities. As is known, tax revenues as a percentage of GDP(the tax-to-GDP ratio) are comparable to the OECD countries’ average: 33.1 per cent and 34.1 per cent, respectively, in 2021(Tesouro Nacional, 2022; OECD, 2022). However, the composition of Brazilian taxes is extremely regressive, since the system is heavily de­ pendent on indirect taxes. The latter account for 45 per cent of tax revenues, in contrast to a 32 per cent share for OECD countries(Silveira et al., 2022). Hence the importance of a tax reform that increases the share of direct taxes. A pro­ gressive tax reform should guarantee that people with the same capacity to pay are taxed to the same extent. As we have seen, the inability to observe such a principle of hori­ zontal equity reproduces racial inequality and white privi­ lege, since black men at the top end up being overtaxed rel­ ative to white men. Besides, a progressive tax reform is sup­ posed to increase the tax burden on the rich, reducing the weight at the bottom of the distribution structure in terms of the tax burden and, therefore, respecting the principle of vertical equity. In this respect, the small role played by direct taxes in tax revenues is an indication that there is considera­ ble latitude for such an increase. – an increase in the maximum marginal rate for the inher­ itance and gift tax(ITCMD) from 8 per cent to 30 per cent, with incremental progressive rates of 5 percent­ age points 14 ; – an overall change in the composition of tax revenues, reducing the weight of taxes on consumption and in­ creasing the weight of direct taxes. – the adoption of a 2 per cent tax rate on wealth for the super-rich, i. e. the 0.2 per cent at the top of Brazilian income distribution pyramid. It is worth noting that the last recommendation implies a comprehensive understanding of the tax system, and points to the importance of a reform that simultaneously considers modifications to income and consumption taxes. A reform of indirect taxes is currently taking place in Brazil, but the political debate is unfortunately not expanding into a broad­ er discussion, which is unfortunate considering the urgent need for a reform of direct taxes. Moreover, the potential positive effects of such a reform in terms of economic growth and a possible expansion of tax revenues might be hindered by the recent approval of a new fiscal framework imposing severe constraints on government spending and, hence, the public provision of goods and services. This rare opportunity to transform Brazilian historical and structural inequalities should be seized. Thus, in view of the foregoing, we recommend: – the adoption of a new marginal tax rate of 35 per cent, levied on the top 1 per cent of the income distribution structure; – the adoption of a tax rate on distributed profits and dividends between 15 per cent and 20 per cent; – the broadening of the corporate taxation base, includ­ ing a review of the simplified taxation regimes for small and medium-sized enterprises; 14 As Pires(2022) has suggested. 11

FRIEDRICH-EBERT-STIFTUNG – INEQUALITY IN BRAZIL REFERENCES Bajard, F.; Chancel, L.; Moshrif, R.; Piketty, T.(2021): Global wealth inequality on WID.world: estimates and imputations. World Inequality Lab – Technical Note N° 2021/16. 10p. Barros, R. P.; Carvalho, M.; Franco, S. O(2007): Papel das transferên­ cias públicas na queda recente da desigualdade de renda brasileira. In: Barros, R. P., Foguel, M. N.; Ulyssea, G.(Ed.). Desigualdade de renda no Brasil: uma análise da queda recente. Brasília: Ipea. v. 2. pp. 41–86. Bottega, A.; Cardomingo, M.; Carvalho, L.; Fernandes, R. C.; Orair, R.; Palomo, T. R.; Silveira, F. G.(2021a): A proposta de reforma tributária para o Imposto de Renda de Pessoa Física e seus efeitos sobre a desigualdade. Nota de Política Econômica nº 15, Made USP. Bottega, A.; Bouza, I.; Cardomingo, M.; Pereira, F. P.; Pires, L. N. (2021b): Quanto fica com as mulheres negras? Uma análise da dis­ tribuição de renda no Brasil. Nota de Política Econômica nº 18, Made USP. Bottega, A.; Pires, L. N.; Forquesato, P.(2023): Considerações sobre o efeito distributivo e a perda arrecadatória de um reajuste da tabela do IRPF. Nota de Política Econômica nº 34, Made USP. Cardoso, D. F.; Cardoso, G. S.; Ventura, T.(2022): Modificações na composição da carga tributária brasileira: efeitos econômicos e distribu­ tivos. In: Pires, Manoel(Org.). Progressividade tributária e crescimento ­econômico. Rio de Janeiro: FGV IBRE. pp. 243–284. Chancel, L.; Piketty, T.(2021): Global Income Inequality, 1820–2020: the persistence and mutation of extreme inequality. Journal of the European Economic Association , 19(6):3025–3062. Chancel, L.; Piketty, T.; Saez, E.; Zucman, G.(Eds.)(2022): World ­Inequality Report 2022. Harvard University Press. Chancel, L.; Moshrif, R.; Piketty, T.; Xuereb, S.(2023): Historical ­Inequality Series on WID.world – Update. World Inequality Lab – Techni­ cal Note N° 2023/01. Clemente, F.; Lírio, V. S.; Aransiola, T. J.(2021): Brazilian Evidence on Tax Evasion and Enforcement: A Case Study of Global North–South Com­ parison. Criminal Justice Review, XX(X), 16p. De Rosa, M.; Flores, I.; Morgan, M.(2022): More Unequal or Not as Rich? Revisiting the Latin American Exception. World Inequality Lab – Working Paper N° 2022/13. 79p. Gobetti, S.W.; Orair, R. O.(2017): Taxation and distribution of income in Brazil: new evidence from personal income tax data. Revista de Economia Política , 37(2), pp. 267–286. Gomes, J. P. F.; di Rada, R. P.; Cardomingo, M. R.; Pires, L. N.(2022): Privilégio branco na estrutura tributária brasileira: uma análise intersec­ cional de impostos diretos e transferências. Nota de Política Econômica nº 27, Made USP. Goto, F.; Pires, M.(2022): A proposta de reforma do imposto de renda: diagnóstico e análise. In: PIRES, Manoel(Org.). Progressividade tributária e crescimento econômico. Rio de Janeiro: FGV IBRE. pp. 119–138. Hoffmann, R.(2013): Transferências de renda e desigualdade no Bra­ sil(1995–2011). In: Campello, T.; Neri, M.(Org.). Programa Bolsa Família: uma década de inclusão e cidadania. Brasília: Ipea. pp. 207–216. Martins, G. K.; Arthen, G.; Gomes, J. P. F.(2024): Taxação dos Su­ per-Ricos no Brasil: Efeitos sobre Arrecadação e Distribuição de Renda. Nota de Política Econômica, n º 50, Made USP. Medeiros, M.(2015): Distribuição Da Riqueza No Brasil: Limitações a Uma Estimativa Precisa a Partir Dos Dados Tabulados Do IRPF Disponíveis (The Distribution of Wealth in Brazil: Limitations for a Precise Estimate from the Existing Tabulated Data from IRPF). Available at SSRN: http://ssrn.com/abstract=2641192. Medeiros, M.; Souza, P. H.; Castro, F. A.(2015): O Topo da Distribui­ ção de Renda no Brasil: primeiras estimativas com dados tributários e comparação com pesquisas domiciliares(2006–2012). Revista de Ciências Sociais, Vol. 58, No. 1: 7–36. Medina, L.; Schneider, F.(2018): Shadow economies around the world: What did we learn over the last 20 years? IMF Working Paper nº 2018/017. 76p. OECD(2022): Revenue Statistics 2022: The Impact of COVID-19 on OECD Tax Revenues. OECD Publishing, Paris. Available at: https://doi. org/10.1787/8a691b03-en. Orair, R.(2022): Imposto sobre a renda corporativa: tendências interna­ cionais e análise comparada do modelo brasileiro. In: Pires, Manoel(Org.), Progressividade tributária e crescimento econômico. Rio de Janeiro: FGV IBRE. pp. 51–78. Pires, M.(2022): Tributação de heranças e doações. In: PIRES, Manoel (Org.). Progressividade tributária e crescimento econômico. Rio de Ja­ neiro: FGV IBRE. pp. 181–198. Schiozer, N.; Marcolin, J. V. S.; Bouza, I. C.; Leme, J. P. V. M.(2021). O Imposto de Renda da Pessoa Física no tempo: um breve histórico do discurso político brasileiro nas reformas tributárias. Nota de Política Econômica nº 17, Made USP. Silveira, F. G.(2010): Tributação, previdência e assistência sociais: impac­ tos distributivos. In: Castro, J. A. D. O.; Santos, C. H. M. D. O.; Ribeiro, J. A. C. O.(Org.), Tributação e equidade no Brasil: um registro da reflexão do Ipea no biênio 2008–2009. Brasília: Ipea, 1ª ed., pp. 67–124. Silveira, F. G.; Palomo, T. R.(2023): The Brazilian State’s redistributive role: changes and persistence at the beginning of the 21 st century. Discussion Paper nº 275, Ipea. Silveira, F. G.; Palomo, T. R.; Cornelio, F. M.; Tonon, M. R.(2022): Previdência e assistências sociais, auxílios laborais e tributos: caracterís­ ticas redistributivas do Estado brasileiro no século XXI( Working Paper nº 7). Made – FEA/USP. Silveira, F. G.; Passos, L.; da Silva, E. G.; Palomo, T. R.(2020): Impac­ tos redistributivos das transferências públicas monetárias e da tributação direta: evidências com a POF 2017–2018. Nota Técnica nº 89, Ipea. Soares, F. V.; Soares, S. S. D.; Medeiros, M.; Osório, R. G.(2007): Programas de transferência de renda no Brasil: impactos sobre a desi­ gualdade. In: Barros, R. P.; Foguel, M. N.; Ulyssea, G.(Eds.), Desigualdade de renda no Brasil: uma análise da queda recente. Brasília: Ipea. v. 2. pp. 87–130. Soares, S. S. D.; Osório, R. G.; Soares, F. V.; Medeiros, M.; Zepeda, E. (2009): Conditional cash transfers in Brazil, Chile and Mexico: impacts upon inequality. Estudios Económicos, número extraordinario, pp. 207–224. Soares, S. S. D.; Souza, P. H. G. F.; Osório, R. G.; Silveira, F. G. (2010): Os impactos do benefício do Programa Bolsa Família sobre a de­ sigualdade e a pobreza. In: Castro, J. A.; Modesto, L.(Ed.), Bolsa Família 2003–2010: avanços e desafios. Brasília: Ipea. v. 2. pp. 25–52. Tesouro Nacional(2023): Estimativa da Carga Tributária Bruta do Governo Geral. Boletim da Secretaria do Tesouro Nacional. Minis­ tério da Fazenda. Available at: https://sisweb.tesouro.gov.br/apex/ f?p=2501:9::::9:P9_ID_PUBLICACAO:46589. 12

imprint ABOUT ThE AUThORS IMPRINT Research Center on Macroeconomics of Inequalities (Made): The Research Center on Macroeconomics of Inequalities(Made), based at the University of São Paulo (FEA/USP), is dedicated to understanding and discussing the economic causes and consequences of the multiple facets of inequalities, focusing mainly on the Brazilian reality. Made actively contributes to academic debate and public policy proposals, fostering dialogues between researchers, policymakers, and representatives of organized civil society. In doing so, it seeks to expand access to economic knowledge from a critical perspective, theoretically and empirically. Moreover, Made collaborates with other national and international research centers, working to establish the fight against inequalities as a cornerstone of state policies in Brazil and around the world. Published by: Friedrich-Ebert-Stiftung e. V. Godesberger Allee 149| 53175 Bonn| Germany Email: info@fes.de Issuing Department: Division for International Cooperation/ Global and European Policy https://www.fes.de/referat-globale-undeuropaeische-politik Responsible: Sarah Ganter| International Financial Policy sarah.ganter@fes.de Editing: Dr. James A. Turner Design/Typesetting: pertext, Berlin| www.pertext.de Illustrations: edeos – digital communication GmbH Orders/Contact: Katrien Klüver, katrien.kluever@fes.de Made researchers involved in this publication: Amanda Martinho Resende holds a Master’s degree in Economics from the University of São Paulo(USP). She is a researcher at the Center for Research in the Macroeconomics of Inequalities(Made/FEA-USP). She has a Bachelor’s degree in Economics from the Federal University of Rio de Janeiro(UFRJ) and a Bachelor’s degree in Law, with an emphasis on State and Society, from the Pontifical Catholic University of Rio de Janeiro(PUC-Rio). The views expressed in this publication are not necessarily those of the Friedrich-Ebert-Stiftung(FES). Commercial use of media published by the FES is not permitted without the written consent of the FES. Publications by the FES may not be used for electioneering purposes. ISBN 978-3-98628-561-6 © 2024 Guilherme Klein Martins is a professor at the University of Leeds(UK), holds a Ph.D. in Economics from the University of Massachusetts Amherst, and a Master’s and a Bachelor’s degree in Economics from the University of São Paulo(FEA/USP). Guilherme Arthen is a Ph.D. student in Economics at the University of São Paulo(FEA/USP). He holds a Master’s degree from the same institution. João Pedro Freitas is an undergraduate student in Economics at the University of São Paulo(FEA/USP). www.fes.de/bibliothek/fes-publikationen

INEQUALITY IN BRAZIL Income, Wealth and Tax Distribution The paper offers an introductory over­ view of inequality in the largest national economy of Latin America, focusing on key dimensions such as income, wealth and tax distribution. The authors ex­ plore Brazil’s multidimensional social in­ equalities that have strong regional, gender and, most importantly, racial as­ pects connected to historical social power relations and exploitation. The Brazilian tax system is markedly re­ gressive. For instance, labour is dispro­ portionately taxed in comparison to capital and the widespread use of con­ sumption taxes disproportionately bur­ dens low-income earners. The authors argue that a progressive tax reform should seek to eliminate such dispari­ ties by reversing historical and institu­ tional privileges granted to the rich by the state. They also point out the lack of data on wealth inequality and call for more transparency. The government of president Lula has promised to reform taxation of the su­ per-rich and is being supported by civil society in an active campaign. A tax re­ form that has begun to gain traction dur­ ing Brazil’s temporary presidency of the G20 is the creation of a minimum global tax rate on billionaires. An excise tax of two per cent is estimated to generate al­ most R$ 42 billion in taxes in a year. The authors point out that an excise tax greater than 2.5 per cent could even re­ verse the pattern of strong regressivity of Brazil’s effective tax incidence on income beyond the top 1 per cent in the distribu­ tion structure and return to the progres­ sive pattern observed up until the richest 1 per cent of individuals. Further information on the topic can be found here: https://www.fes.de/themenportal-die-welt-gerecht-gestalten/ weltwirtschaft-und-unternehmensverantwortung/steuergerechtigkeit