Daniel Duque
Jens Matthias Arnold
Expanding Social Protection and Addressing Informality in Latin America
Abstract
Brazil has built a sophisticated social protection system over the years, contributing substantially to reductions in poverty and inequality. At the same time, the fiscal burden of the system is considerable, and the financing of many social protection benefits relies on labour contributions and taxes by those who have a formal job, which is not the case for almost 40% of the workforce. These contributions drive a significant wedge between workers’ take-home pay and the cost of creating these formal jobs for employers, which in turn may be one of the reasons behind widespread labour informality. Reducing these high non-wage labour costs for low-wage earners with salaries close to the minimum wage may create substantial benefits for formalisation. This chapter proposes reforms that could reduce the cost of formal job creation and promote formalisation, provided that alternative financing sources for the labour taxes and contributions of low-wage formal workers can be identified. This could potentially allow many more workers to join the formal sector and benefit from stronger social protection benefits.
Over the last decades, Brazil has built one of the most sophisticated and extensive social protection schemes in Latin America. Benefit coverage and levels exceed that of many countries in the region, including an almost universal old-age pension coverage, a universal public healthcare system and conditional cash transfers that reach almost 20 million households. These social protection benefits cover risks like longevity, health and poverty and are contributing substantially to the reduction of poverty and inequality in Brazil. Spending on social policies amounted to more than 16.5% of GDP in 2023.
The current system has evolved over the years through repeated attempts by policy makers to make improvements and address punctual shortcomings and challenges that became priorities at different points in time. These repeated additions to the system have made the system larger and more complete but have not necessarily made the system as a whole more consistent. From today’s perspective, Brazil’s social protection system as a whole has scope for improving both the efficiency with which it uses public funds to improve social outcomes, and for the incentives it creates.
Social protection in Brazil shares many similarities with other countries in the region: Its origin lies in social security benefits loosely modelled on systems found in advanced economies, mostly financed through contributions in labour receipts, and with benefit eligibility limited to those who pay contributions, which is a common definition of formal sector workers today. These have been complemented by social assistance benefits for those working in the informal sector, largely financed from budget resources, i.e. general taxation revenues. Nowadays, many different, at times overlapping, benefits co-exist and give rise to a social protection system that displays some degree of fragmentation.
One of the consequences of the elaborate social protection system is the sizeable financial burden that its financing imposes on society, which is only set to rise as the population ages, and the distortions resulting from the way that it is financed. With most of formal-sector social security financed through charges on (formal) labour, Brazil is characterised by high non-wage labour costs in the formal labour market, consisting of payroll taxes and social contributions. This high tax burden drives a significant wedge between the cost of creating a formal job and employing a person informally.
In 2023, around 39% of the labour force lacked access to a formal job. Informality has several serious consequences. For one, it precludes workers from better earnings opportunities and job quality. This is consistent with poverty rates that are more than five times higher among those in informal work, whose incomes are more than 45% lower than those of formal workers (IBGE, 2024). In addition, informality deprives a significant share of the population of access to some social security benefits, as will be described in more detail in this chapter.
Furthermore, informality also entails other disadvantages for people’s well-being and productivity, including the exclusion from access to credit, related to the absence of any declared income. This precludes informal workers from using access credit markets to smooth consumption in the case of unexpected spending needs, and it reduces the scope for productivity-enhancing investments into the activities of self-employed workers. Where entire firms are kept outside the formal sector, the difficulties in access to credit are likely to hamper productivity and growth of these firms, making informality a clear obstacle to productivity growth. Finally, informality also reduces tax collection. This is true from a static perspective due to lost tax revenues today. But the losses are probably even larger from a dynamic perspective, as firms and workers could obtain higher incomes over time if they were not held back by the constraints that informality imposes on their development, which would eventually lead them to pay more taxes.
While informality has many causes, improving the incentives resulting from the current financing of social protection can probably lead a long way towards reducing it and promoting formal job creation. The current structure of costs and benefits associated with labour and social security legislation makes formal job creation costly – and informality a common form of tax evasion. Improving incentives for formal job creation is therefore a promising avenue to reduce informality, alongside other policies such as refinements in labour regulations and better enforcement.
This chapter on Brazil tries to map a way forward for strengthening social protection mechanisms in a way that unties both the access and the financing of social protection from formal-sector employment. For old-age pensions, it argues that coverage and benefit levels have advanced well, but the financing of these pensions continues to present challenges for formalisation incentives. With respect to benefits for those in working age, the chapter argues that Brazil’s main conditional cash transfers programme Bolsa Família, which has been substantially upgraded in 2023, can play a key role in this transformation and can be a key tool for breaking the vicious circle between poverty and informality in Brazil.
Some of the proposed improvements in incentives will require finding alternative financing sources to replace current ones. This will necessitate either raising revenues from other sources or finding savings by reconsidering some of the current spending items and raising their efficiency. Brazil has significant scope to improve public spending efficiency, as documented in several editions of the OECD Economic Surveys of Brazil, and some of these financing proposals will be detailed in section 6 of this chapter.
Informality has declined significantly since the 1980, when around 60% of the labour force was working in informal jobs (Figure 5.1). After a period of minor ups and downs hovering around these high levels during the 1980s and 1990s, the Brazilian labour market saw an unprecedented decline in the incidence of informality from the beginning of the 2000s.
This decline in informality occurred in the context of a remarkable transformation of the labour market, both in terms of its sectoral composition and the distribution of income, with marked improvements in the distribution of income towards lower inequality levels. Informality dropped by more than a third between 2000 and 2012, declining from above 60% at the beginning of the millennium to below 40% in 2016. This was even more remarkable given that the minimum wage rose by 61% in real terms over the same period.
Share of informal workers
Note: Years 1991, 2000 and 2001 are interpolated. Self-employed workers without registration are classified as informal.
Source: IBGE, Pesquisa Nacional por Amostra de Domicílios, available at: https://www.ibge.gov.br/en/statistics/multi-domain/science-technology-and-innovation/20620-summary-of-indicators-pnad2.html
A decomposition exercise suggests that large part of the decline in informality between 2003 and 2012 can be traced back to improvements in educational attainments, reflected in a declining participation of less educated and less experienced workers, in addition to sectoral shifts and a direct absorption of unemployed workers into the formal sector (Barbosa Filho & Moura, 2015). Similarly, modelling results lead to the conclusion that compositional changes in the Brazilian workforce are the most important factor behind the reduction in informality that occurred over this period (Haanwinkel & Soares, 2021).
Since 2016, informality has remained fairly stable around, and more recently slightly below, 40%. While this means that the achievements of the last decades in terms of lower informality have been preserved, the secular downward trend has come to an end, and informality has remained persistent, pointing to structural factors that prevent it from declining further.
Similar to other countries in the region, the range of social protection benefits available to formal and informal workers is generally not the same, although it is less dissimilar in Brazil than elsewhere in Latin America. This is because Brazil has been particularly successful in expanding non-contributory benefits available to informal workers, who -as a result- are not left entirely without coverage by social protection.
5.3.1. Pensions have achieved almost universal coverage but create strong distortions
Old-age pensions have achieved almost universal coverage in Brazil through a combination of contributory and non-contributory schemes. Public spending on pension was 11.9% of GDP in 2023, which is high relative to the share of the elderly in the population (Figure 5.2, Panel A). Pension expenditures will continue to rise as Brazil’s population ages faster than most others in the world (UN, 2024[1]). The median age has increased by 6 years between 2010 and 2022, and the population aged 65 and above has risen by 57% over that period. In parallel, life expectancy at birth has increased by almost 6 years since 2000, reaching about 76 years in 2021, 80 years for women and 72 for men (OECD, 2021[2]).
In practice, all those with formal work are affiliated to either the general pension scheme for private-sector employees (RGPS) or to a civil servant regime run by the federal, state or municipal governments (RPPS). While the benefits and fiscal costs are different across the two regimes, with the civil servant regime being somewhat more generous in many instances, both systems guarantee a pension of at least one minimum wage to their beneficiaries. Replacement rates in Brazil’s contributory pension systems are generally higher than in most OECD countries. For those earning an average wage, pension benefits under the RGPS replace 88.4% of that gross income, compared to 50.7 in the OECD average.
The 2019 pension reform marked an important step to contain future increases in pension costs and strengthened the system’s sustainability (Box 5.1). It introduced a general minimum retirement age of 62 years for women and 65 for men, raising the effective retirement age to levels closer to the OECD average (Figure 5.2, Panel B). It is estimated that the 2019 pension reform will generate savings of about 10 percentage points of GDP over 10 years (OECD, 2020; IFI, 2019).
In November 2019, Brazil approved a pension reform containing a wide range of parametric changes. These included the introduction of a general minimum age of 62 years for women and 65 for men, while previously retirements based on years of contribution led to effective retirement ages of 53 years for women and 56 for men. Contribution rates were made progressive, leading to visibly higher contributions for high-wage earners and to lower rates for minimum wage earners. Survivor pensions will no longer be 100% of the deceased spouse's pension but vary according to the number of dependants. Limits on the accumulation of benefits were also introduced.
The reform improves the distributional footprint of the pension system. By establishing a binding retirement age, it ended the current practice where low-income earners with part of their working life in informal employment retire later than others who retire on the basis of years of formal contributions. Moreover, the reform reduces inequalities between the private sector regime and the public sector regime, as the latter was significantly more generous before.
Note: The net replacement rate is the individual net pension entitlement divided by net pre-retirement earnings, taking account of personal income taxes and social security contributions paid by workers and pensioners.
Source: OECD Pensions at a Glance 2023; World Bank; Tribunal de Contas da União; and OECD calculations.
One peculiar feature of pension and other formal-sector benefits, and an important driver of increasing pension costs, has been the current indexation pattern of these benefits. Brazil’s constitution mandates that no old-age pension benefit can be lower than the minimum wage. At the same time, the real value of the minimum wage has seen sharp increases of 33% over the last 15 years. After these increases, the minimum wage is now located between the 60th and the 70th percentile of the income distribution, meaning that 60-70% of Brazilians have monthly incomes of less than one minimum wage (Figure 5.3). This strong effective indexation explains Brazil’s high pension replacement rates (OECD, 2023b) and has mostly benefited households with above-median incomes, exacerbating income inequality.
Note: Extreme poverty is defined as earning less than 2.15 USD 2017 PPP per day. Poverty is defined as earning less than 6.85 USD 2017 PPP per day.
Source: IBGE (PNAD Contínua) and Banco Central do Brasil.
Formal social security pension benefits have been complemented by non-contributory pensions since the early 1990s. The expansion of non-contributory pensions has been instrumental for achieving an almost universal coverage as of the age of 65, with approximately 99% of those aged 65 receiving either a contributory or a non-contributory pension, of which 20% receive a contributory social security pension and 80% a non-contributory pension.
Non-contributory pensions are delivered through two schemes. In rural areas, those who have (presumedly) worked most of their lives in agricultural or forestry activities will receive a rural pension without any requirement of past contributions, with a guaranteed benefit level of one minimum wage, plus a 13th monthly payment per year. In urban areas, those who reach the age of 65 and incomes below 25% of the minimum wage have access to a targeted non-contributory pension, the so-called Benefício de Prestacão Continuada (BPC), which provides a monthly minimum wage to urban low-income workers, but no 13th monthly payment. The amount of these non-contributory, equivalent and automatically indexed to the level of the minimum wage, implies that non-contributory pension benefits are at the same level as social security pensions for most low-income earners, as the minimum wage also acts as a floor for social security pensions. In fact, the vast majority of social security pension benefits granted under the RGPS and the RPPS amount to one minimum wage.
Both these pensions are constitutional rights, so every worker who is eligible to either one of them will have access to it, regardless of any federal budget constraint. This way, pensions among the elderly have become virtually universal. These programmes have essentially eradicated extreme poverty among the elderly, which now affects less than 1% of older age groups (Figure 5.4).
Source: PNADC 2023.
5.3.2. Social assistance benefits to those in working age are effective in reducing poverty
Social assistance benefits for those in working age consist primarily of conditional cash transfers, which achieve a broad, almost universal coverage among the target group. Brazil’s flagship anti-poverty programme, the Bolsa Família programme (BFP), reaches about 27% of households, using a fairly precise targeting to poor households through a large public registry, called the Cadástro Unico. It combines a low eligibility line, i.e., the maximum income that a household can have to become eligible for a benefit, and a benefit level that varies mostly according to the number of children in the household.
Before the pandemic, Bolsa Família’s budget saw a steady 13% decline in real terms from 2014 to 2019 (Federal Government, 2020). Since 2010, readjustments in the underlying Bolsa Família extreme poverty line, which defines eligibility, had been insufficient to offset inflationary losses. The increasing pressures on Bolsa Família spending were partly related to rigid budget rules that include automatic indexation of social security benefits, to which Bolsa Família benefits do not belong, and minimum spending floors in certain spending areas. In the face of rising mandatory spending, continuous cuts in discretionary spending, to which Bolsa Família belongs in budget terms, was the only lever to meet fiscal targets. Public investment, which is also considered discretionary spending, suffered the same fate of being crowded-out by the inability to rein in mandatory spending, most of which is enshrined in Brazil’s constitution and therefore politically more difficult to change (OECD, 2020). In the absence of readjustments during 2014-2016, the average monthly benefit plummeted by 20% over those three years, a time when poverty and inequality increased (Barbosa et al., 2020). New adjustments only came in 2016 and 2018, but they were insufficient to replace the previous losses. At the end of 2019, the average BFP benefit was still 16% below 2014 levels.
In April 2020, the federal government launched the Auxílio Emergencial (Emergency Aid) programme to mitigate the strong impact of the Covid-19 pandemic on informal workers. Its amount was originally proposed at BRL 200 per month by the Executive Branch but then raised to BRL 500 by Congress and finally defined at BRL 600 in an agreement between both. The benefit was intended for informal workers and unemployed people from low-income families. In order to receive it, adults only had to register in a mobile application created for that purpose. Beneficiaries of the Bolsa Família programme became eligible automatically but had to substitute one benefit for the other.
With this broad eligibility, the programme reached around 66 million beneficiaries in August 2020. At that time, half of the population was living in a household where at least one member had received the transfer. The programme was implemented throughout the whole year 2020, with benefit levels cut to half in the outer months of the year. The programme was continued with a slightly different design between April and September of 2021.
The significant income support provided by the federal government during the pandemic mitigated much of the economic impact of the pandemic on low-income households (Duque, 2020). However, with a fiscal cost of almost 8% of GDP, the programme’s price tag was very high in international comparison, putting Brazil in second place among emerging economies with respect to non-health Covid spending, second only to Chile (IMF, 2021). When the Auxílio Emergencial ended in October 2021, the Federal Government modified Bolsa Família and renamed it to Auxílio Brasil. The new programme was similar to the design of Bolsa Família, with an adjustment of the eligibility line from BRL 89 to BRL 100 per person and a continuously strong focus on families with young children.
While the base benefit of BRL 190 per family and month was not too far from pre-pandemic levels, a temporary bonus raised the benefit level to BRL 400 per family (around 25% of median earnings), more than doubling the spending on the programme (IFI, 2021). To finance this considerable extra expense, the federal government obtained several temporary waivers from Congress in 2022. In early 2023, the government reinstated Bolsa Família, discontinuing Auxílio Brasil and permanently increasing the cash transfer value to BRL 600 plus variable benefits, leading to an average benefit level of BRL 680, around 38% of median earnings. The minimum benefit, which had already been temporarily raised during the pandemic to address massive income losses, now represents about 46% of the minimum wage (Figure 5.3). Preliminary evidence suggests that at this level, Bolsa Família benefits have come closer to showing a negative correlation with labour participation (Duque 2023, 2024). This measure raised the total costs of the programme to around 1.5% of GDP in 2024. The original conditionalities related to school attendance and healthcare were fully re-established in the new Bolsa Família design.
5.3.3. Labour market benefits are limited to formal workers
One area covered by contributory social insurance but not by non-contributory benefits is a more comprehensive protection against economic risks related to unemployment, health and maternity, including paid maternity and sick leave. Access to these benefits is probably the most salient difference in benefit coverage between formal and informal workers in Brazil. There are no provisions for insuring informal workers against sudden income losses such as those related to dismissals or business failures.
Income insurance needs of informal workers are generally higher than those of formal workers. Informal workers tend to be affected more frequently by income shocks than formal workers (Gomes et al., 2019). This is related to a higher volatility among formal workers both in terms of effective earnings and hours worked. For example, between the first quarter of 2018 and 2019, informal workers’ hours worked had an average standard deviation of 24% of their average hours, also referred to as coefficient of variation, compared to 18% among formal workers. Similarly, informal workers’ earnings had a coefficient of variation of 37%, above the 29% for formal employees.
Social assistance benefits such as Bolsa Família could in principle cover these insurance needs and help informal workers who face sudden income losses, if they were disbursed rapidly after an income loss occurs. In practice, however, the long delays between applying for Bolsa Família benefits, and their eventual concession is severely limiting the potential of conditional cash transfers for helping a family in a timely manner in moments when a member of the family loses her or his income.
Formal workers have access to two separate programmes that provide benefits in case of dismissal. Although quite different in nature, the two schemes serve essentially the same purpose of providing resources to dismissed formal-sector workers (OECD, 2020).
Seguro Desemprego is a benefit paid for an amount of up to two minimum wages over a period of 3 to 5 months following a dismissal deemed without just cause as regulated by law. This benefit established by the 1946 Constitution and is nowadays regulated by Article 7 of the Brazilian Constitution of 1988. The criteria for accessing the Seguro Desemprego include having received salaries from a legal entity or an individual equivalent to it for periods that can vary from 6 to 12 months, depending on the number of previous benefit request, besides other minor requirements. The benefit is granted for a maximum period varying from three to five months, counting from the date of dismissal. The benefit amount depends on the previous salary. For low-income workers earning up to 1.5 minimum wages, the benefit is set at 80% of the former salary. For workers earning more than 1.5 minimum wages, the replacement rate decreases with the salary, with an overall cap at approximately 2 minimum wages (BRL 2640 in 2023). With a fiscal cost of about 0.4% of GDP in 2023, Seguro Desemprego is financed through earmarked resources from a turnover tax on firms.
The Fundo de Garantia do Tempo de Serviço (FGTS) is an individual unemployment account scheme created in 1966, with the aim of providing better financial stability to formal workers. Contributions to the scheme are made by the employer at a rate of 8% of the salary, or 11.2% for domestic workers, with top-ups by the Federal government that generate a fiscal cost of about 0.3% of GDP. FGTS allows workers to withdraw their individual account balances upon dismissal or for the purpose of a real estate purchase. Account balances have traditionally been remunerated below market rates and even below inflation for all years but two since 1995, giving the FGTS balances a notion of forced savings for which workers could get a better return elsewhere. The balances are used to finance public policy programmes including in the areas of housing and infrastructure. More recently, annual withdrawals even without a dismissal have become an option for FGTS, but workers who exercise this option have no further withdrawal available in case of dismissal, which defies the original purpose of the scheme to insure workers against income losses associated to dismissals.
In addition to these two schemes, Brazil provides an in-work benefit for low-wage formal workers (Abono Salarial), on the condition that they have been registered for at least five years, excluding domestic workers and employees of individuals. It provides an annual benefit that tops up the salaries of formal workers who earned on average up to two minimum wages per month over the year, pro-rated for those who worked less than the full calendar year. Domestic workers and those employed by individuals are not covered. This programme was originally meant to provide in-work benefits to those with low incomes, but given the current parameters and widespread informality, Abono Salarial no longer fulfils this purpose. All formal workers earn at least one minimum wage and after many years of strong real increases, the minimum wage currently marks the 65th percentile of the income distribution. The poorest 30% of Brazilian households received only 17% of the distributed Abono Salarial benefits in 2017 (CMAG, 2020). By contrast, a childless couple of two formal employees earning between 1 and 2 minimum wages each, for instance, will both be eligible for this transfer, despite being among the top 40% in the income distribution, based on calculations using the 2019 PNAD Contínua household data. All in all, the Abono Salarial mostly benefits the middle of the income distribution, while the actual low-income workers are those with informal jobs earning less than a minimum wage. The programme budget was about 0.3% of GDP 2023, financed through earmarked resources from a turnover tax on firms.
5.3.4. The distributional impact of current social spending
On the whole, Brazil’s current set of social protection spending is not characterised by a strong targeting towards low-income households. The public sector spends around 12.9% of GDP on formal-sector pensions and social security, which far exceeds spending on any other social protection programme. This is followed by around 1.7% of GDP spent on conditional cash transfers including the Bolsa Família programme. Non-contributory pensions account for public expenditure of around 0.9% of GDP, while Brazil spends 0.7% of GDP on labour market benefits (Figure 5.5, Panel A).
At the same time, public spending on social security pensions has a larger impact on the disposable incomes of high-income households than on those of low-income households, even relative to disposable incomes, as depicted in the upward sloping green line in Panel B of Figure 5.5. This implies that public pensions exacerbate rather than attenuate income inequalities. The opposite is true for non-contributory pensions and especially conditional cash transfers, which leave a larger footprint on the incomes of low-income earners than on higher income quintiles. Labour market benefits are also slightly progressive, but less so than the latter two categories. The progressivity of labour market benefits comes particularly from their lesser role for the top income quintile, probably as a result of the caps on overall benefits from unemployment insurance schemes. For informal workers, who are more vulnerable than those with formal jobs and largely concentrated in the lower income quintiles, the patterns depicted in Figure 5.5 means that they are eligible for only a minor fraction of the overall social protection spending, as they are excluded from both social security pensions and labour market benefits.
Source: Panel A: Tesouro Nacional, Estatísticas Fiscais do Governo Geral 2023. Panel B: IBGE, POF 2017/18.
Another way of looking at the success of social protection policies in supporting vulnerable households is to look at the percentage of the population participating in social protection, unemployment benefits and active labour market programmes, including both direct and indirect beneficiaries. With respect to this metric, Brazil occupies an intermediate position among countries covered, with around 50% of the population having some kind of access to such benefits (Figure 5.6). When looking at the benefit incidence and narrowing down on the most vulnerable households, i.e. the poorest 20% of the population, Brazil fares less well in international comparison: Only 4% of the total social protection and labour market benefits go to the bottom quintile of the income distribution, compared to around 9% in the median country depicted in Figure 5.6.
Note: Data are based on national representative household surveys. Social protection and labour market programmes include social insurance, social safety nets, unemployment benefits and active labour market programmes.
Source: ASPIRE: The Atlas of Social Protection, World Bank, available at https://www.worldbank.org/en/data/datatopics/aspire
High non-wage labour costs, a large part of which finances formal-sector social protection, drive a significant wedge between the cost of formal and informal work in Brazil (Firpo and Portella, 2021). Taken together, the different contributions paid by workers and employers add more than 35% to the take-home pay of a formal-sector worker remunerated at the minimum wage. A more detailed breakdown of contributions paid by workers and employers is presented in Table 5.1. Pension system contributions of 20% for the employer and 7.5% for the worker represent the lion’s share of non-wage labour costs. In addition, employers pay 8% of the salary of formal workers to the FGTS saving system of individual unemployment accounts, an average of 3.25% for a workplace risk insurance scheme. All these contributions provide some kind of direct benefits to workers, although the FGTS contribution is taxed implicitly by remunerating balances below market rates. No such direct link to worker benefits exists for a series of payroll taxes totalling on average 4.6% of wages and whose revenues are earmarked to finance education and training programmes and a public-sector agency in charge of land reform. By contrast, the unemployment insurance scheme Seguro Desemprego is not financed from contributions or taxes on workers’ salaries, but from a turnover tax on companies. If it were financed through charges on salaries, this would have added another 2.5 percentage points to non-wage labour costs in 2018 (Firpo and Portella, 2021).
Contributions and payroll taxes paid by employers and workers
Note: Earmarked payroll taxes provide no direct benefits to workers that are linked to their contributions. They include taxes used to finance Sistema S, an entity providing worker training, Salário Educação which provides funding for public basic education for all, and the public-sector agency INCRA. The rates applied for workplace risk insurance and Sistema S are an average of their lower and upper bound, respectively 0.5-6% and 0.6-3.1%. The average is calculated using the ‘Relação Anual de Informações Sociais’ (RAIS) in 2019.
Source: Adapted from Firpo and Portella (2021), on the basis of RAIS 2019 data.
These contribution levels are high in international comparison, as they are only exceeded by Japan and European economies, which have more sophisticated welfare systems, like France, Sweden and Belgium (Figure 5.7). These high non-wage labour costs create strong incentives for tax avoidance through informal employment and are one of the explanations behind Brazil’s significant levels of informality, both among workers and among firms. For an establishment employing low-skilled labour that is facing the decision to create a formal or an informal job, the over 40% labour cost difference between an informal and a formal worker will be hard to compensate through other means, especially in markets where firms using informal labour compete with fully compliant firms.
Attacking this root cause of informality would be one promising way of reducing informality in Brazil, even if it is not the only policy lever available. For a low-skilled worker, accepting an informal job is associated with access to less social benefits, but the differential is not that much. Basic benefits like access to an old-age pension or universal public healthcare are available to informal workers on similar terms as to formal workers, including with respect to the level of the pension benefit. The incentives to seek a formal job are diminished by the high level of social security contributions, coupled with the relatively limited differential in benefits.
This situation creates a vicious circle in which the high non-wage labour costs that finance formal-sector social protection effectively preclude many informal workers from getting a foothold in the formal labour market, which in turns cements persistent inequalities in social protection coverage between formal and informal sector workers. Reducing the cost differential between formal and informal jobs would help to break this vicious circle resulting from the interplay of informality and social protection.
Existing evidence based on a natural experiment that reduced non-wage labour costs for some sectors suggests that reducing non-wage labour costs would have strong potential to reduce informality among firms (Rocha et al., 2018). More specifically, industries eligible for the reduction experienced an increase of 4.8% in the number of formal firms as a result of the formalisation of existing informal businesses.
Income tax plus employee and employer social security contributions, 2024
Notes: Data for 2024 except for Brazil (2019). Single individual without children at the income level of the average worker.
Source: OECD Labour taxation – average and marginal tax wedge decompositions database; and OECD Taxing Wages in Selected Partner Economies: Brazil, China, India, Indonesia and South Africa in 2019.
Brazil’s situation is in many ways similar to those of other Latin American countries, but there are also key differences. The almost universal coverage of the population with old-age pensions, for example, is a feature of Brazil’s social protection system that several other countries in the region will likely look at with envy, while the wide reach of conditional cash transfers contributes significantly to poverty alleviation. A universal public health system provides basic health benefits to all residents regardless of their labour market status. At the same time, the wide coverage with non-contributory benefits also implies that the incentives for workers to join the formal sector are lower than in other countries in the region, because the range of benefits they can enjoy by moving from informal to formal work changes less than in other countries.
Informality affects almost 40% of the workforce, holding back social progress, productivity and tax revenues. This is keeping the economy trapped in an equilibrium that delivers far from optimal outcomes. Moving from the current situation into a more desirable equilibrium without informality and better opportunities for professional development and productivity growth will require thinking outside the box and undertaking significant reforms to reduce the current disincentives to formal job creation. Given Brazil’s long history of endemic informality, small changes are unlikely to have any major impact. Neither the strong growth nor the massive improvements in access to education since 2000 have been sufficient to bring the lower end of the income distribution into the formal sector, even if part of those who managed to move up the skills ladder have been able to find formal employment.
Brazil’s starting point for reconciling universal access to social protection with financing mechanisms that avoid unnecessary disincentives for formalisation is in some ways easier than that of other countries in the region, given its elaborate system of non-contributory benefits in the area of pensions, conditional cash transfers and healthcare. While many policies have aimed to tackle Brazil’s widespread informality, one area that has not seen much progress is a systematic focus on reducing non-wage labour costs low-income earners to reduce the cost differential between informal and formal jobs. Brazil’s social security contributions have been at internationally high levels for decades (Table 5.1).
Formal job creation is clearly related to non-wage labour costs. One piece of evidence for this was the introduction of a micro-entrepreneur scheme Microemprendedor Individual (MEI) in 2009. This scheme allowed micro enterprises with up to 1 employee (besides the entrepreneur) to join the formal sector at much lower registration costs. As of 2011, the scheme also included a substantial reduction in non-wage labour costs in eligible sectors. The 2011 reform triggered an increase in the number of formal firms of more than 4% in eligible sectors, as a result of informal firms becoming formal (Rocha, Ulyssea and Rachter, 2018). By contrast, an earlier measure that included only a reduction in registration costs for informal companies had no discernible impact.
The level of non-wage labour costs may not matter in the same way in all circumstances. In fact, informal employment is most prevalent among those with relatively low skills and low wages, while for many high-skilled jobs informality is not a relevant consideration. This suggests focusing the efforts for reducing social security contributions on the lower range of the wage distribution. One potential approach could be to aim for substantive reductions in non-wage labour costs for those earning up to 1.25 or 1.5 minimum wages, in combination with additional contributory benefits, and social contributions levied on wages, for those earning more. These higher contributions would, of course, need to rise gradually with incomes to avoid strong discretionary increases in non-wage labour costs at any given threshold, as this would create incentives to keep wages just below such thresholds.
As argued in Chapter 2 of this volume, one promising way to reduce the formalisation disincentives from social security contributions and payroll taxes would be to exempt low-income ranges from these contributions as much as possible and replacing these resources with general tax revenues. This would shift some of the tax burden from social security contributions to other taxes and contributions, without raising the overall tax burden. Its success and feasibility, however, will ultimately hinge on the ability to create ways to finance the revenue losses from lowering social security contributions for those low-wage workers that are currently formal and paying contributions, even if this is by far the minority of workers in this income range.
5.5.1. Pension contributions
The largest element of non-wage labour costs in formal jobs are the contributions of both workers and employers to the pension system, which sum up to 27.5%-29% of the wage bill for salaries in the salary range between 1-2 minimum wages. Reforming these contributions for lower wage ranges would have the largest potential for reducing non-wage labour costs and improving formalisation incentives. This is particularly relevant in the lower wage ranges because the added value of a contributory pension may not be straightforward to see for many low-income workers, given that they can enjoy similar pension benefits through non-contributory pensions.
Reducing pension contributions for low-income earners could be achieved by merging the current non-contributory pensions and a part of the contributory pension system into a universal basic old-age pension. This basic pension would be paid to all those above the retirement age of 65, regardless of their work history.
Such a pension could be financed from general tax resources, as is the case for current non-contributory pensions, as opposed to from specific social security contributions levied only on formal-sector salaries. This would allow reductions in non-wage labour costs for low-wage earners, which are most affected by informality, and hence significantly improve the incentives for formalisation, while continuing to ensure universal access to basic pensions for all. The basic universal pension would be effective in eliminating old-age poverty, while at the same time reducing one of the major distortions that give rise to Brazil’s high levels of informality.
This could be done by applying a zero rate to pension contributions for those earning up to a certain threshold, while gradually phasing in pension contributions at initially low but increasing rates as wages rise. The exact zero-rate threshold where the contribution rate could switch from zero to a low positive entry rate could be subject to discussion. But given the bunching of the wage distribution right at the minimum wage, this threshold should probably be slightly above but in the vicinity of one minimum wage. One possible option could be to set it at 1.25 or 1.5 minimum wages, but other options in that vicinity may be equally suited.
For those with higher incomes, there would still be a case for a mandatory contributory second pillar of the pension system with benefits that would top up the basic universal pension for those with higher incomes. The main justification for a contributory pillar for higher incomes is the typical short-sightedness of people’s financial planning. Besides preventing old-age poverty, which would be addressed with a universal basic pension from the first pillar, pension systems provide consumption smoothing between the working and the retirement periods, i.e., ensure that retired workers can enjoy pension benefits that deliver a reasonable replacement for their working-age incomes. The basic universal pension provided by the first pillar would not be able to provide adequate replacement ratios for those earning several multiples of the minimum wage.
A contributory second pillar of the pension system should be actuarially fair in broad terms and provide neither subsidies to higher-income pensioners, as the current pension system implicitly does, nor should it expropriate part of their savings from contributions made over the years. There is no reason either why a top-up contributory pension scheme should make any difference between private-sector workers or civil servants. Integrating newly appointed civil servants into the general pension scheme with its basic non-contributory and its complementary contributory pillars would allow substantial savings over time and starkly improve the distributional impact of the pension system, even if introduced only for newly appointed civil servants while preserving the current pension rights for all those already in civil service.
The level of the basic pension benefit could be determined according to several considerations. In principle, a universal basic pension would have the main task of eliminating old-age poverty, suggesting that the poverty line would be a lower bound for such a benefit. This reasoning would likely resonate in a number of Latin American countries where many of the elderly currently go without any pension at all, and old-age poverty levels are high. In Brazil, however, the starting point and the political economy of creating a universal basic pension are quite different, given the current almost universal coverage of the elderly with pensions amounting to one minimum wage. Against this backdrop, a universal basic pension below the level of one minimum wage would likely be politically unrealistic, as it would be perceived by beneficiaries as a step backwards. Setting the basic pension benefit at the level of the current minimum wage would probably be the only politically feasible way forward and ensure continuity with the status quo for most current and future pensioners.
Setting the basic pension at the level of one minimum wage today, however, does not preclude future discussions about the most appropriate indexation mechanism, which will affect the real value of the pension benefits in the future. Over many years, current indexation mechanisms have pushed up the real value of minimum pension benefits, while using these resources for other social benefits that reach those with lower incomes would have allowed significantly stronger progress in reducing inequality (Arnold and Bueno, 2021). Looking ahead, there may be a case for preserving the real purchasing power of future minimum pension benefits without strictly following movements in the minimum wage, which have often outpaced productivity growth over the last decades and may do so again. Many countries automatically index pensions in a way that preserves their real value and passes part of the wage increases of current workers on to retirees, but not necessarily all of those gains.
The proposed two-pillar pension system would strongly improve formalisation incentives and could make real inroads into eventually eliminating the inequalities and the productivity drag resulting from high informality. For a worker in the lower range of the salary distribution, there would essentially be no major cost difference between formal and informal work, except maybe a small contribution to workplace accident insurance. Workplace risk insurance is intimately linked to the employer-employee relationship and Brazil’s workplace accident insurance varies both across sectors and the according to the accident history of individual firms, which is in principle a desirable feature that creates incentives for firms to minimise these risks (Levy and Cruces, 2021). At the same time, current contribution rates of 1% for low-risk or 2% for medium-risk activities are small and unlikely to make a real difference when it comes to formalisation, as long as people perceive that the insurance will protect them in case of a work accident.
Moving towards a tax-financed basic pension would lead to revenue losses as social security contributions would no longer be due on salaries up to the zero-rate threshold. For those with incomes above that threshold, the question would be whether their incomes up to that threshold should also be paying contributions at a zero rate or not. While exempting them as well may have regressive distribution effects when looked at in isolation, it would essentially be the price to pay for maintaining compliance incentives and avoiding large threshold effects. If those earning only slightly above the threshold immediately had to pay contributions on their entire salary, including the amount that would be exempt if their salary was below the threshold, then this would lead to extremely high marginal contribution rates at that threshold. Workers and employers would have strong incentives to set or declare salaries just below the threshold instead. At the same time, there are other ways to introduce more progressivity into the financing of social protection, as discussed in the next section.
5.5.2. Payroll taxes
Another place to start reducing the high non-wage labour costs would be the payroll taxes that finance non-work-related policy programmes including training, the public-sector agency INCRA and public basic education for all. These add 4.6% of extra costs to the cost of formal job creation and have no obvious link to the salary or the employment contract on which their economic burden falls. Given Brazil’s high minimum wage, it is likely that many employers will be unable to pass on these payroll taxes to workers through lower wages, so that these payroll taxes are likely to have significant effects on formalisation (Firpo and Portella, 2021).
These payroll taxes are used to further broad policy objectives. In the case of Sistema S and Salário Educação, both worker training and better financing for public education are desirable objectives for a society, not least due to their positive external effects for the economy as a whole. But they have no particular relationship with formal-sector employment and there is no reason why the tax base that finances them should be only for the receipts from formal labour. By defining the tax base so narrowly, these taxes create distortive incentives to substitute formal labour for informal work. Instead, broad policy objectives should be financed from a broad range of taxes, which could include taxes on all kinds of income and taxes on goods and services. Against the background of Brazil’s long history of earmarking tax revenues to specific expenditure items, which have created challenging rigidities in the budget that make it hard to adjust the budget to a changing world, the creation of these specific payroll taxes for financing these unrelated policy objectives should be seen like an accident of history rather than a feature of a desirable tax structure (OECD, 2020).
Financing policies currently funded through payroll taxes from alternative tax revenue sources would eliminate this part of the distortion towards informality, and the next section will elaborate a number of proposals of where additional tax revenues could come from.
Estimates suggest that pension contributions and payroll taxes from the income range up to one minimum wage currently amount to 1.35% of GDP. This provides a ballpark estimate of the amount of funding would need to be identified from other revenue sources or by reducing current expenditures.
5.5.3. Unemployment benefits
An additional reform that could make a larger difference in terms of reducing the cost wedge between formal and informal employment would be a reform of the unemployment benefit system. Employers’ contributions to the FGTS scheme amount to 8% of wages, or roughly one fourth of the current non-wage labour costs at the level of the minimum wage. There are several reasons why these contributions could be reconsidered. One is the broader question about the relative costs and benefits of an explicit unemployment benefit for low-wage workers in a context where well-developed conditional cash transfers exist to ensure a basic lifeline for those unable to generate sufficient income, including following a dismissal.
That perspective would imply that for the large number of workers earning around a minimum wage, the flagship cash transfer programme Bolsa Família could be tasked with ensuring a basic transfer to low-wage workers who lose their job. Following the recent increase in benefit levels, the resulting replacement rate would not be too distant from current practice in other countries, for those earning around the minimum wage. The basic Bolsa Família benefit of BRL 600 is currently around 46% of the minimum wage, while a single parent with two dependent children would get BRL 900 from Bolsa Família, about 69% of the minimum wage. These sizeable potential replacement rates raise the question of whether an additional separate unemployment benefit scheme is necessary for workers at these wage levels.
A key challenge for Bolsa Família to support families after dismissals will be to make the programme more responsive to changes in beneficiaries’ personal situation by speeding up processing of benefit claims. Programme enrolment for new beneficiaries can sometimes take months or more, but dismissed workers need immediate income support.
Workers who are already programme beneficiaries will also find it easier to seize employment opportunities if they know that the benefit will resume rapidly should they become unemployed again. This area has seen important progress recently, with a gradual rather than sudden withdrawal of Bolsa Família benefits for those whose household income exceeds the programme’s eligibility threshold, for example because a household member found employment. This gradual withdrawal allows households to keep receiving 50% of the benefit for a transition period of 12 months. The benefit will only be cancelled if household income exceeds the eligibility line during the transition period and gets reinstated otherwise. This transition period encourages people on Bolsa Família benefits to take up job offers, without fear of being worse off if they fail to maintain their employment.
Making the programme even more agile than at present would imply more frequent visits to beneficiary households by local social workers. If conditional cash transfers could react and pay out pro-rated monthly benefits within a few days after a worker declares her or his dismissal and hence becomes eligible for the transfer, they could become one reasonable way to insure the temporary income losses associated with dismissals. For a larger part of low-income workers, namely those currently in informal employment and with no access to unemployment benefits at all, such a change would clearly constitute an improvement over the status quo.
For those on the lower end of the income scale, a reform that would ensure swift coverage with Bolsa Família benefits upon dismissal while eliminating the current 8% contribution to the FGTS scheme would combine basic insurance against dismissals with a substantial reduction in non-wage labour costs. As a result, more of those currently in informal employment would likely be able to join the formal sector as employers would be less reluctant to create formal jobs.
For higher income ranges, Bolsa Família would not be able to generate satisfactory replacement rates, and income smoothing considerations would be poorly serviced by Bolsa Família alone. But that does not mean that the current co-existence of two unemployment benefit schemes that -despite their substantial differences in design- essentially aim to ensure workers’ losses in the case of dismissal, would be the optimal setup for higher income ranges either. One of them, the Seguro Desemprego, is financed from tax revenues that go beyond formal payrolls and will be part of the currently implemented reform of indirect taxes that aims to convert several currently distortive indirect taxes into a dual value-added tax (VAT) levied by the federal and subnational governments. In this context, the future design of the current financing source of Seguro Desemprego is expected to be a well-designed VAT, while FGTS is solely financed from contributions levied on formal wage bills. These differences in financing sources give Seguro Desemprego a clear edge over FGTS in the role to insure temporary income losses resulting from dismissals.
Moreover, the benefits that workers receive in return for the 8% contribution that their employers make for FGTS are likely to be valued at less than the contributions, given the history and trend of remuneration below market-rates. In fact, Brazilian workers have revealed a preference for withdrawing their funds from FGTS whenever they had a chance to do so, including when recent reforms allowed yearly withdrawals at the expense of losing access to FGTS funds in the event of a dismissal. A broader reform of the unemployment benefit system could therefore consider abolishing the mandatory returning current balances to FGTS account holders while shifting the role of unemployment benefits towards Seguro Desemprego and Bolsa Família. Such a reform would also allow savings on the current FGTS contributions by the Federal government, whose fiscal costs are about 0.3% of GDP.
Establishing a transfer policy to which in principle all can have access upon losing their livelihood would also introduce an opportunity to build a unified registry of employment (both formal and informal) and transfers. Currently, formal employment is registered in the ‘Relação Anual de Informações Sociais’ (RAIS), while potential social assistance beneficiaries are registered in the ‘Cadastro Único’. The lack of a broader database was one of the main challenges of the implementation of emergency benefits during the pandemic, prompting the Government to use a mobile application where adults could self-declare their employment situation.
Zero-rating mandatory contributions and payroll taxes on low salaries would eliminate much of the current disincentives to formalisation. But such a reform will hinge on finding alternative funding sources for the social security contributions and payroll taxes currently paid by low-income formal workers. These are currently estimated at 1.5% of GDP for incomes up to 1.25 minimum wages and 1.8% of GDP for those earning up to 1.5 minimum wages. Identifying fiscal space to compensating the lost contributions and tax revenues will be challenging, but not impossible, given significant scope for improving spending efficiency
In the same way as Brazil features a more sophisticated social protection system than some other countries in Latin America, it is also characterised by much larger public revenues and spending. With public spending of more than 45% of GDP, the size of Brazil’s public sector exceeds the 40% OECD average and by far exceeds the 28% average of Latin American peers (Figure 5.8). Taxes are characterised by a strong reliance on goods and services taxes, while social benefits, interest payments, education and health are the largest spending items. Social benefits are not only the largest spending category but also the one that has recorded the strongest increase in recent years.
Source: IMF World Economic Outlook database: April 2025; OECD Comparative tables of countries in the global database; and Tesouro Transparente.
While a complete analysis of Brazil’s fiscal accounts is beyond the scope of this chapter -and has been undertaken by several national and international institutions including OECD Economic Surveys of Brazil (OECD, 2020; OECD, 2023) - the purpose of this section is to give a few examples where reorganising public expenditures and taxes could provide the funding for the social protection reforms proposed in the previous section.
Brazil’s relatively high spending and revenue levels in a regional context suggests that looking for additional funding source to compensate for lower social security contributions should start with reviewing some of the current public expenditures, where significant inefficiencies suggest for substantial savings (World Bank, 2018; OECD, 2020; OECD, 2023; Izquierdo, Pessino and Vuletin, 2018).
Improving the targeting and effectiveness of the social protection system as a whole presents room for reducing spending, which could in turn finance lower contributions for those with formal job earning low wages. While large parts of Brazil’s social benefits are well targeted and effective in reducing inequality and poverty, this is not the case for all benefits. Formal-sector benefits like the Abono Salarial and Salário Família, with a budget of about 0.25% of GDP in the last years, could be reconsidered. As discussed in section 3, Abono Salarial is a wage subsidy granted to all formal workers earning between one and two minimum wages per month. The Salário Família benefit, in turn, is a wage subsidy to formal workers with children below the age of 14 and earning up to 1.35 minimum wages. Given that at least 60% of Brazilians have per-capita incomes below one minimum wage, the beneficiary income ranges of these programmes are considerably above the median income. As a result, these benefits are not well targeted and could be reconsidered. Such a change could be carried out gradually, to ease the transition for those formal workers who would lose access to the benefit.
Pension expenditures themselves probably do not present much scope for immediate reductions, not least given the political economy and the recent 2019 pension reform. One further reform option for the future would be to reconsider the current indexation of pension benefits. Currently no old-age pension benefit can be lower than a full minimum wage, whose real value has increased by 33% over the last 15 years and is now between the 60th and the 70th percentile of the income distribution. This strong effective indexation explains Brazil’s high pension replacement rates (OECD, 2021b) and has mostly benefited households with above-median incomes, exacerbating income inequality. Adjusting social security benefits, including pensions, in line with inflation could free up resources to finance reductions in non-wage labour costs for low-wage earners. Removing the automatic link between pensions and the minimum wage would generate increasing annual savings, estimated to reach about 6.1 percentage points of GDP by 2050 (Cuevas et al., 2019).
Beyond the realm of social policies, public-sector wage costs may also present room for reducing public expenditures. Brazil’s public sector wage bill for current and retired civil servants is high in international comparison, including when compared with emerging economies and Latin America (Figure 5.9). Most of public employment is concentrated in subnational governments, where the wage bill is now well above half of the primary expenditure. Containing the growth of the wage bill, by strengthening links between pay and performance instead of seniority, will allow creating fiscal space for alternative financing models of social protection. A reform aimed at modernising public administration management could introduce regular and systematic performance evaluations of civil servants, particularly during the probationary period. By some estimated, such a reform could bring fiscal savings of up to 8% of GDP in ten years (IPEA, 2020).
Source: OECD Annual government expenditure by function (COFOG) database; OECD Annual GDP and components - expenditure approach database; and Tesouro transparente.
On the revenue side, rather than raising tax rates or adding new taxes, the focus could be on reducing tax expenditures such as special regimes and exemptions or reduced rates. Brazil has significant tax expenditures that often fail to serve to purpose they were designed for, or at least fail to do so in an effective way (World Bank, 2018; OECD, 2020; OECD, 2023; Ministério da Economia, 2022). Reducing some of these tax expenditures would lead to higher tax revenues and a larger public sector, but in a way that would broaden tax bases, potentially reducing distortions, and not at the expense of raising statutory tax rates. Taken together, subsidies and tax expenditures amounted to 4.8% of GDP in 2023 at the federal government level, or 7.2% of GDP when including states (de Renzio et al., 2024).
One of the largest tax expenditures is the targeted corporate tax regime for small and medium enterprises called Simples Nacional. This regime generates revenue shortfalls of around 0.8% of GDP alone (DEAP, 2022). It combines a lighter tax burden with a simplified calculation of tax liabilities based on turnover, replacing up to 9 separate taxes. The eligibility condition for the simplified regime is to have revenues below approximately USD 1 million, which is high in international comparison and discourages firms to grow beyond the threshold. Being based on revenues rather than income, it also discourages sourcing intermediate inputs from potentially more efficient external providers. For very small firms, the easier compliance may outweigh these considerations. However, with its high participation ceiling, the regime is currently used by 74% of Brazilian firms and could be much better targeted, including by reducing the participation threshold.
Simples Nacional has the merit of lowering tax compliance costs for small and medium enterprises, which is a considerable argument in light of Brazil’s exceptionally complex system of indirect taxes. However, in 2023 Congress approved a major reform of the current indirect tax system. The reform is based on simple rules harmonised across states, a broad base, full tax credit of the value-added tax on all inputs and zero rating for exports. The proposal unifies three federal taxes (PIS, COFINS, IPI), as well as one state and one municipal tax (ICMS and ISS, respectively), into a dual value added tax (VATs) with one part administered by the federal and another part by subnational governments. The dual VATs will have a common tax base, but states can apply different rates, while taxing on the basis of the destination of goods and services. The reform will take effect gradually to reduce transition costs, beginning in 2026 and ending in 2033 for taxpayers.
The adoption of a unified value-added tax instead of the current multiple consumption taxes will align Brazil with OECD best practices and reduce distortions. The benefits of a well-designed VAT include its neutrality for production and sourcing decisions as long as the VAT is guided by the destination principle, and the expected reduction in compliance costs for firms can hardly be underestimated. In the context of the ongoing complete overhaul of Brazil’s and fragmented system of indirect taxes, the case for maintaining this special regime in its current size will become much weaker. As compliance costs for small and medium enterprises fall in the context of the reform, the eligibility ceiling for participation in Simples Nacional could be reduced gradually until it only covers the smallest firms.
Another reason for reducing the eligibility ceiling of Simples Nacional is related to tax arbitration opportunities for professionals such as lawyers, accountants, architects or consultants (Dutz et al 2018). Filing as a corporate entity under Simples Nacional has become a preferred choice for many high-income professionals, as this can reduce the effective tax burden from almost 50% to as little as 11.5% for those with few deductible expenses (Appy, 2017). By contrast, the smaller Microempreendedor Individual (MEI) programme has a more moderate ceiling of USD 16,000 in turnover, which reduces the scope for abuse. At significantly lower fiscal cost, this programme has contributed to lower informality among low-income entrepreneurs, especially women (OECD, 2023).
Tax expenditures in indirect taxes also include the exemption of basic food items, many of which will carry over to the new value added tax. While basic food items are consumed by low-income households and represent a larger share of their consumption basket that for high-income households, these items are also consumed by the better off. In absolute numbers, households with higher incomes tend to spend more on these items than low-income households, meaning that a large share of the lost tax revenues are appropriated by households that are relatively well off. Usually, targeted benefits provide a more effective way to transfer resources to low-income households than such tax exemptions. These exemptions could be phased out gradually, starting with the less obvious “basic products” such as some types of fish and meat including salmon, beef fillet, turkey, duck; dairy products including specialty chesses and whipped cream; highly processed items such as soluble cappuccino powder. Estimates based on the current system suggest that reducing these exemptions, referred to as “cesta básica” in the policy discussion, could lead to as much as 0.33% of GDP in additional tax revenues (de Renzio et al., 2024).
Beyond consumption taxes, there is also scope for reducing tax expenditures in personal income taxes. One of these relates to the high level of the basic deduction, or the income level as of which individuals become liable for paying personal income taxes. In 2023, this income threshold was about 1.6 minimum wages, which is more than what 90% of Brazilians earn. In international comparison, this level of the basic deduction is exceptionally high (OECD, 2018). Reducing this basic deduction in a measured and gradual way presents potential for raising additional revenue from the upper and middle classes without affecting the tax burden on lower and lower-middle class income earners.
Moreover, a current personal income tax deductibility of expenditures for private health and education expenses has regressive distributional effects, given that it benefits only those top 10% income earners with incomes above the basic deduction. Moreover, only 25% of Brazilians are subscribed to private health plans, while most of the population relies on the public health system. Phasing out these deductions could save 0.23% of GDP.
Considering all of the points discussed above, finding the fiscal space to zero-rate social contributions and payroll taxes for low-income earners seems clearly possible, at least from an economic point of view. Whether it will be possible to create a political consensus for this remains to be seen, but the prospects of achieving substantial reductions in informal employment, an area where progress has been limited in recent years, should provide a strong argument for doing so.
Brazil has built an impressive social protection system over the years, with an almost universal pension coverage and anti-poverty cash transfers that have become a model for many other countries. This sophisticated system is contributing substantially to reductions in poverty and inequality. At the same time, as the system has evolved over time and different benefits have been added, the overall consistency of the current system could be improved and the efficiency of public spending on social protection enhanced. While some benefits are highly effective, others may warrant a second thought or refinements. The financing of many social protection benefits relies on labour contributions and taxes by those who have a formal job, which is not the case for almost 40% of the workforce. These contributions drive a significant wedge between workers’ take-home pay and the cost of creating these formal jobs for employers, which in turn may be one of the reasons behind widespread labour informality. Reducing these high non-wage labour costs for low-wage earners with salaries close to the minimum wage may create substantial benefits for formalisation, and is possible, as some of the analysis contained in this chapter is meant to illustrate.
References
[17] Appy, B. (2017), “Tributação e produtividade no Brasil”, in Bonelli R, Veloso F and Castelar Pinheiro A (eds.), Anatomia da produtividade, FGV IBRE, Rio de Janeiro.
[5] Arnold, J. and M. Bueno (2021), “How effective are different social policies in Brazil? A simulation experiment”, OECD Economics Department Working Papers, No. 1662, OECD Publishing, Paris, https://doi.org/10.1787/47087376-en.
[10] Cuevas, A. et al. (2019), Chapter 11 Fiscal Challenges of Population Aging in Brazil, International Monetary Fund, https://www.elibrary.imf.org/display/book/9781484339749/ch011.xml (accessed on 30 August 2023).
[15] de Renzio, P. et al. (2024), RELATÓRIO NACIONAL SOBRE GASTOS TRIBUTÁRIOS Brasil (VERSÃO PRELIMINAR).
[3] Firpo, S. and A. Portella (2021), “Informal and small: How Labor Market Institutions Affect Inequality, Income Volatility and Labor Productivity in Brazil”, UNDP RBLAC WORKING PAPER SERIES, United Nations Development Programme, New York, https://www.latinamerica.undp.org/content/rblac/en/home/library/human_development/informal-and-small--how-labor-market-institutions-affect-inequal.html (accessed on 30 April 2022).
[12] IFI (2019), Relatório de Acompanhamento Fiscal 12/2019, Instituição Fiscal Independente, Brasília, https://www12.senado.leg.br/ifi/relatorio-de-acompanhamento-fiscal.
[8] Izquierdo, A., C. Pessino and G. Vuletin (2018), Better Spending for Better Lives How Latin America and the Caribbean Can Do More with Less, Interamerican Development Bank, Washington, DC, http://www.iadb.org/DIA2018spending (accessed on 25 January 2019).
[6] Levy, S. and G. Cruces (2021), “Time for a New Course: An Essay on Social Protection and Growth in Latin America”, UNDP Latin America and the Caribbean Working Paper Series, No. 24, United Nations Development Programme, https://www.latinamerica.undp.org/content/rblac/en/home/library/human_development/time-for-a-new-course--an-essay-on-social-protection-and-growth-.html (accessed on 21 September 2021).
[9] Ministério da Economia (2022), Orçamento de Subsídios da União (OSU), https://www.gov.br/economia/pt-br/assuntos/avaliacao-de-politicas-publicas/subsidios/transparencia-de-subsidios-da-uniao (accessed on 30 August 2023).
[13] MTP (2022), Projeções financeiras e atuariais para o regime geral de previdência social RGPS, Secretaria de Previdência (SPREV), Ministério do Trabalho e Previdência (MTP).
[16] OECD (2023), OECD Economic Surveys: Brazil 2023, OECD Publishing, Paris, https://doi.org/10.1787/a2d6acac-en.
[14] OECD (2023), Pensions at a Glance 2021: OECD and G20 Indicators, OECD Publishing, Paris, https://doi.org/10.1787/ca401ebd-en.
[2] OECD (2021), OECD Reviews of Health Systems: Brazil 2021, OECD Reviews of Health Systems, OECD Publishing, Paris, https://doi.org/10.1787/146d0dea-en.
[11] OECD (2020), OECD Economic Surveys: Brazil 2020, OECD Publishing, Paris, https://doi.org/10.1787/250240ad-en.
[4] Rocha, R., G. Ulyssea and L. Rachter (2018), “Do lower taxes reduce informality? Evidence from Brazil”, Journal of Development Economics, Vol. 134, pp. 28-49, https://doi.org/10.1016/J.JDEVECO.2018.04.003.
[1] UN (2024), “Population size and growth”, in World Population Prospects, The 2024 Revision, United Nations, New York, https://doi.org/10.18356/0f5a4c11-en.
[7] World Bank (2018), A fair adjustment : efficiency and equity of public spending in Brazil : Volume 1 - Overview, World Bank, Washington, DC, http://documentos.bancomundial.org/curated/pt/643471520429223428/Volume-1-Overview (accessed on 12 December 2019).