## Introduction

Ethnic and racial differences in human capital and earnings are one of the key determinants of inequality in Latin America (De Ferranti et al. 2004; Hall and Patrinos 2006; Ñopo 2012). Although the specific social dynamics faced by Afro-descendants and indigenous peoples differ, both of these populations have systematically higher poverty rates and are disproportionately represented among the poor.1 Fiscal redistribution is one of the policy levers the state has to address ethnoracial inequalities. As shown in a series of existing studies that apply the common methodological framework developed at the Commitment to Equity Institute (Tulane University), fiscal policy unambiguously reduces income inequality, albeit to different degrees.2

Does fiscal policy also reduce income inequality between ethnic and racial groups?3 This article analyzes the effects of fiscal policy on inequality along ethnic and racial lines for Bolivia, Guatemala, and Brazil in 2009, the year for which comparable analysis for these three countries is available.4 These three countries were selected because they have large Afro-descendant (Brazil) and indigenous (Bolivia and Guatemala) populations, and welfare gaps between whites and nonwhites are quite high. Using the self-identification method to classify individuals,5 the indigenous populations in Bolivia and Guatemala represent 54.2 and 40.7 percent of the total populations, respectively, the largest shares in the region. In the last census round in Brazil, 50.8 percent of the national population self-identified as Afro-descendant, the largest both in absolute and relative terms in Latin America.6

A salient indicator of ethnoracial inequality is the extent to which the indigenous and Afro-descendant populations are disproportionately underrepresented among the higher income groups and disproportionately overrepresented among the poor.7 If whites and nonwhites were similarly affected by social circumstances and market forces, one would expect the share of the nonwhite population at every income stratum to be roughly equal to its share in the total population. However, this is far from true. In Bolivia, Brazil, and Guatemala, the share of the nonwhite population is roughly 54.2, 50.8, and 40.7 percent, respectively. And, yet, before fiscal redistribution, nonwhites represent 71.7 percent in Bolivia, 74.2 percent in Brazil, and 60.8 percent in Guatemala. The overrepresentation of nonwhites among the poor is just the other side of the coin of how poverty affects the different ethnoracial groups. The indigenous and Afro-descendant peoples face a considerably higher probability of being poor than does the white population. The probability of being poor is measured by the head count ratio.8 In Bolivia and Guatemala, the head count ratio (with the national poverty line) is 14.7 and 20.6 percent, respectively, for the nonindigenous population, but 31.5 and 46.6 percent, respectively, for the indigenous population. In Brazil, the head count ratio is 5.2 percent for the white population but 14.6 percent for Afro-descendants.

To what extent does fiscal policy reduce the gap in the probability of being poor for different ethnic and racial groups? I will answer this question using fiscal incidence analysis, which essentially allows us to trace the changes in people’s incomes from incomes before taxes and transfers to incomes after taxes and transfers. The specific components of fiscal policy looked at here are direct taxes and transfers (such as personal income tax and cash transfers), indirect taxes (such as value-added taxes or VAT and excise taxes), and subsidies (such as energy and food price subsidies). The article draws from the following Commitment to Equity Institute studies: Bolivia: Paz Arauco et al. (2013, 2014); Brazil: Higgins and Pereira (2013 and 2014) and Pereira (2017); and Guatemala: Cabrera, Lustig, and Moran (2015).9 The results can be compared across countries because the studies use the same fiscal incidence methodological framework, described in Lustig and Higgins (2013) and Lustig (2017a). The household surveys used for the analyses are, in Bolivia, Encuesta de Hogares, 2009; in Brazil, Pesquisa de Orçamentos Familiares, 2009; and in Guatemala, Encuesta Nacional de Ingresos y Gastos de las Familias, 2009–2010.

In the three countries, although fiscal policy reduces ethnoracial gaps, the change is negligibly small and in some instances does not exist. While, as I will show below, cash transfer programs targeted to the poor tend to redistribute more resources to Afro-descendants and indigenous groups, they are too small to make a significant difference in terms of ethnoracial differentials in poverty rates.

## Fiscal Incidence Analysis: A Brief Overview

Fiscal incidence analysis is used to assess the distributional impacts of a country’s taxes and transfers.10 Essentially, it consists of allocating taxes and public spending (particularly social spending) to households or individuals so that one can compare incomes before with incomes after net taxes. The fiscal incidence analysis used here—known as the “accounting approach”— is a point-in-time analysis and does not incorporate behavioral or general equilibrium effects. That is, no claim is made that the original or market income equals the true counterfactual income in the absence of taxes and transfers. It is a first-order approximation that measures the average incidence of fiscal interventions.

However, the analysis is not a mechanically applied accounting exercise. The incidence of taxes is the economic rather than statutory incidence. Consistent with other conventional tax incidence analyses, here we assume that the economic burden of direct personal income taxes is borne by the recipient of income. The burden of payroll and social security taxes is assumed to fall entirely on workers. It is assumed that individual income taxes and contributions by both employees and employers, for instance, are borne by labor in the formal sector. Individuals who are not contributing to social security are assumed to pay neither direct taxes nor contributions. Consumption taxes are assumed to be shifted forward to consumers. In the case of consumption taxes, the analyses take into account the lower incidence associated with own consumption, rural markets, and informality. These assumptions are strong because, in essence, they imply that labor supply is perfectly inelastic and that consumers have perfectly inelastic demands for goods and services. In practice they provide a reasonable approximation.11

A fiscal incidence study must start by defining the basic income concepts that are used to assess the effects of fiscal policy on people’s incomes and their distribution. In this article, there are three basic income concepts.

Market income, also called primary or original income, is total current income before direct taxes and transfers.12 It equals the sum of gross (pretax) wages and salaries in the formal and informal sectors (also known as earned income); income from capital (dividends, interest, profits, rents, and so on) in the formal and informal sectors (excluding capital gains and gifts); consumption of own production; imputed rent for owner-occupied housing; and private transfers (remittances, pensions from private schemes, and other private transfers such as alimony).

Disposable income is market income minus direct personal income taxes on all income sources (included in market income) that are subject to taxation plus direct government transfers (mainly cash transfers but also including near-cash transfers such as food transfers, free textbooks, and school uniforms).

Consumable income is disposable income plus indirect subsidies (such as food and energy price subsidies) minus indirect taxes (such as value-added taxes, excise taxes, and sales taxes).

These concepts are summarized in Figure 1.

Figure 1

Fiscal redistribution and income concepts. Source: Lustig and Higgins (2013).

## Taxes, Transfers, and Subsidies

### Bolivia

#### Taxes

In 2009, total tax revenues amounted to 26.9 percent of GDP. In Bolivia, income tax is negligible and was not included in the analysis.13 There are four indirect taxes applied to consumption were included in the analysis: a value-added tax (VAT), a transaction tax, a special tax on hydrocarbons, and a specific consumption tax (excise taxes), all of which account for 41 percent of total tax revenues in 2009 (or 11 percent of GDP).

#### Transfers

In 2009, the year of the survey used for the fiscal incidence analysis, spending on direct transfers equaled 2.0 percent of GDP. With spending at 1.4 percent of GDP in 2009, the largest direct cash transfer program was the noncontributory pension program Renta Dignidad. Implemented in 2008, the program built on an earlier transfer program created in 1994 (Bono Solidario, known as Bonosol). Beneficiaries are citizens aged sixty or older. In 2009, the program benefited close to 800,000 people. Direct transfers also included two flagship conditional cash transfer programs: Bono Juancito Pinto and Bono Juana Azurduy.

The Bono Juancito Pinto was launched to promote primary school attendance. All children between six and nineteen years of age attending public schools are eligible for the program. The transfer consists of a yearly payment equal to 200 bolivianos, approximately $0.18/day (in 2005 purchasing power parity, or PPP) paid once a year, conditional on proven attendance during the school year.14 According to the program’s roster, 1.7 million children benefited from the transfer in 2009, with public expenditures equaling 0.3 percent of GDP. The Bono Juana Azurduy was created in 2009 with the purpose of promoting prenatal health, infant checkups, and an increase in the rate of hospital-attended births. Only mothers and children without access to health insurance are eligible. The program consists of a maximum transfer of 1,820 bolivianos over a maximum period of thirty-three months (equivalent to an average of$0.58 PPP/day). According to the program registry, 776,045 women and children benefited from the program in 2009. Public expenditure on the program reached 0.02 percent of GDP.

Desayuno Escolar (school breakfast program) provides breakfast to children between the ages of four and nineteen who attend school. The per capita average cost of the program is 9 bolivianos per month, about $0.1 PPP/day. In 2008, the program benefited 1,985,158 people. Resources spent on the program reached 0.2 percent of GDP in 2009. The war veterans’ transfer program Beneméritos del Chaco consists of an average monthly payment of 1,254 bolivianos, equivalent to$13.2 PPP/day, paid to veterans of the Chaco War (1932–1935). In 2009, the payment benefited more than one thousand veterans and resources spent on the program amounted to 0.14 percent of GDP.

#### Subsidies

Indirect subsidies include the subsidized fraction of liquid gas and the subsidized fraction of gasoline consumed by households. In 2009 subsidies were equivalent to 0.6 percent of GDP.

### Brazil

#### Taxes

Total fiscal revenues at the federal, state, and municipal levels were about 35 percent of GDP in 2009, much higher than the Latin American average.15 Direct taxes (personal and corporate) represented 45 percent of the taxes levied by the government, and indirect taxes represented 55 percent. The Brazilian tax system is exceedingly complex. The taxes included in the incidence analysis are direct personal income taxes and consumption taxes. The most important indirect tax is the ICMS, a state tax levied on the sale or physical movement of goods, freight, transportation, communications services, and electricity. ICMS accounted for 21 percent of fiscal revenues in 2009. Other important indirect taxes are the COFINS (federal tax on goods and services to finance the social security deficit), ISS (municipal tax on services), PIS (federal tax on goods and services to finance social services for workers), and IPI (federal tax on industrial products). They correspond to 10.8, 4.1, 2.9, and 2.8 percent of fiscal revenues, respectively. Many indirect taxes operate with their own administering department, which may be at the federal, state, or municipal level, compounding on each other, something known as the “cascading effect.” These effects are especially important due to their impact on consumer purchasing power. Exemptions on consumption taxes are almost nonexistent in Brazil and, hence, the effective rates paid on basic food products in Brazil can be especially deleterious for the poor.

#### Transfers

In 2009, direct transfers included in the incidence analysis were roughly equal to 4 percent of GDP. Special Circumstances Pensions, the largest program in terms of spending (2.3 percent of GDP), are designed to smooth the impact of idiosyncratic shocks or are means tested. They are paid in the case of an accident at work, sickness, or related idiosyncratic shock. Although these pensions are funded by the contributory pension system, they are considered noncontributory because they have low or no requirements in terms of length of contribution period. In 2009, there were about 2.9 million beneficiaries (INSS 2010) and the average benefit per person was $5.22 PPP per day. Bolsa Família, Brazil’s flagship conditional cash transfer program, transfers cash to eligible families in exchange for complying with certain conditions. Eligible families are poor families with children less than eighteen years of age or with pregnant women, and all extreme poor (the latter group is regardless of having children). Eligibility is determined through partially verified means testing; households with income below the cutoffs are incorporated into the program. The conditions are prenatal and postnatal care sessions for pregnant women, adherence to a calendar of vaccinations for children up to age five, and a minimum level of school attendance for children ages six to seventeen. There are no conditions for the “fixed benefit” given to extremely poor households. In 2009, the government spent 0.4 percent of GDP on 41.2 million individuals living in beneficiary families, and the average benefit per person living in a beneficiary household was US$0.35 PPP per day.

Benefício de Prestação Continuada (Continued Payment Benefits, BPC) is a noncontributory pension program that provides a monthly monetary transfer of one minimum salary (465 reais per month or US$8.83 PPP per day in 2009) to elderly poor or incapacitated poor. Elderly means sixty-five years old and older, and incapacitated is determined by doctors based on ability to work. In 2009, the government spent 0.5 percent of GDP on this program and there were 3.2 million beneficiaries and the average benefit per person living in a beneficiary household was US$2.18 PPP per day.

Other social transfers such as Food Transfer Programs (Programa de Aquisição de Alimentos, PAA), Unemployment Insurance, and other smaller programs represented 0.9 percent of GDP.

## Concluding Remarks

The importance of reducing ethnoracial inequalities arises both from what this divide means ethically as well as its causes and consequences. Today’s ethnic and racial inequalities are often the product of morally condemnable societal actions such as discrimination in the present and subjugation of indigenous groups and slavery in the past. Due to the discrimination that indigenous peoples and Afro-descendants face in Latin America and the resulting gaps, these ethnic and racial groups are what philosophers would call “morally relevant” groups and, as a result, policies developed to correct these gaps would be deemed ethically acceptable (Nickel 2002). In addition, ethnic and racial inequalities are found to be associated with lower overall development and growth (Alesina et al. 2012; Easterly and Levine 1997). Thus, addressing ethnoracial inequalities may have the additional benefit of generating higher welfare levels for everyone.

While fiscal redistribution can play a role in reducing income inequality in the ethnoracial space, our results show that the impact is very limited. Using several indicators, I showed that taxes and transfers reduce ethnoracial inequality very little or not at all. This is due in large part to the small magnitude of government spending on targeted cash transfers programs. However, the difference in, for instance, the number of poor people by ethnicity and race is so large, that it is unlikely it can be significantly reduced through just income transfers. Nonetheless, at the minimum, the design of transfer programs should avoid exacerbating ethnoracial inequalities. As discussed above, this is the case with the Special Circumstances Pensions in Brazil, which disproportionately benefit the white population because Afro-Brazilians are less likely to be enrolled.