Targeted Tax Relief in Action: How Rio Grande do Sul’s VAT Cashback Aids the Poor

The Devolve-ICMS program in Rio Grande do Sul, studied by the World Bank and SEFAZ-RS, refunds VAT to low-income households through a mix of flat transfers and consumption-based rebates. It shows that while digital systems make targeted tax relief feasible, design trade-offs, gender dynamics, and complementary incentives like lotteries shape its impact on equity and formalization.

The World Bank Group, in collaboration with the Secretary of Finance of Rio Grande do Sul (SEFAZ-RS), has turned a spotlight on an innovative experiment that is reshaping how social protection and fiscal policy intersect in Brazil. The Devolve-ICMS program, introduced in 2021, channels part of the value-added tax (VAT) back to low-income families, combining social relief with incentives for formal economic participation. Already, more than 700,000 households in Rio Grande do Sul have received around US$100 million in transfers. The scheme is particularly significant because Brazil’s sweeping 2023 tax reform includes provisions for targeted cashback at both federal and state levels, making this state-level initiative a test case with national and even international resonance.

How the Program Works

Devolve-ICMS uses the federal government’s social registry, Cadastro Único, to automatically identify low-income households. Eligibility extends to families earning below three minimum wages or less than half a minimum wage per capita, and to those enrolled in Bolsa Família or with children in public high schools. Once selected, beneficiaries receive a prepaid “Citizen Card” issued by Banrisul, the state bank, which is credited every quarter. The payment has two components: a flat R$100 quarterly transfer, similar to an unconditional cash benefit, and a variable rebate calculated from formal purchases documented with electronic invoices. Brazil’s universal system of e-invoicing makes this possible by linking household consumption to tax IDs provided at checkout. In practice, this means families can see part of the VAT they pay returned to them, while the government gains insight into consumption patterns among the poor.

What the Data Reveals

Administrative records from 2023 paint a nuanced picture of how families engage with the program. Perhaps the most striking finding is that over 90 percent of beneficiaries had at least one documented purchase in the formal sector during the year. This challenges assumptions that low-income Brazilians rely almost entirely on informal markets. Yet anomalies also surface: nearly one-fifth of households reported formal spending that exceeded their declared income. This may reflect underreported earnings, reliance on credit and installment purchases, common in Brazil, or the use of tax IDs to register purchases beyond household consumption.

Another key observation is that poorer families spend proportionally more of their income in formal markets than slightly wealthier households. At the lowest levels of income, formal spending averages nearly 60 percent, compared with less than 40 percent among those closer to the upper limit of eligibility. However, averages disguise wide disparities. The median poor household formalizes only about 10 to 20 percent of its income, while some families report consumption nearly equal to or greater than their entire declared earnings. This mix of strong engagement by some and minimal participation by others underscores the uneven reach of the incentive structure.

Trade-Offs in Design

A crucial feature of Devolve-ICMS is the balance between generosity and incentives. Roughly 60 percent of families receive only the flat transfer, because their reported income and spending levels fall below thresholds that would entitle them to variable rebates. For these households, the R$100 quarterly payment exceeds what their purchases would generate, leaving no reason to request invoices or alter spending habits. This design choice ensures predictable fiscal costs for the state, while guaranteeing that even the poorest families receive support. Yet it blunts the program’s role as a tool for encouraging formalization. Policymakers are thus confronted with a delicate trade-off: prioritize equity through unconditional transfers or focus on reshaping behavior through stronger consumption-based incentives. The Rio Grande do Sul experience demonstrates that striking the right balance is far from straightforward.

Women and Lotteries Shape Outcomes

Gender dynamics emerge as another defining feature. Around 85 percent of beneficiaries are women, reflecting policies that encourage female registration in Brazil’s social programs. But women are not just numerically dominant; they are also more engaged. Female-headed households are more likely to collect their debit cards, more likely to register purchases, and more likely to enroll in complementary initiatives. Even after adjusting for income, geography, and household size, women report significantly higher levels of formal consumption. This suggests that VAT cashback schemes may inadvertently empower women by strengthening their role in managing household finances and linking them more closely with formal institutions.

The interaction with Rio Grande do Sul’s Nota Fiscal Gaúcha lottery adds another layer. Families who participate in this prize-based system, which rewards consumers for registering invoices, consistently show much higher levels of formal spending than those who do not. Across income groups, they report formal purchases amounting to 20 to 40 percentage points more of their income. While part of this may reflect self-selection, families already inclined toward formal shopping are more likely to participate; the complementarities between lotteries and cashback are evident. Together, they create overlapping incentives that deepen engagement with the tax system.

A Blueprint for Broader Reform

The lessons from Rio Grande do Sul carry weight far beyond state borders. With Brazil embarking on a dual VAT system that embeds targeted cashback at national and subnational levels, the program offers a living laboratory of what works and what does not. It shows that digital IDs, e-invoicing, and social registries make such schemes feasible on a large scale, but it also warns that if flat transfers dominate, the incentive to formalize spending may vanish. For other developing economies, the experiment is instructive. Consumption taxes are often criticized as regressive, yet cashback mechanisms show promise in softening their impact while potentially nudging families toward formal markets. Success, however, depends on more than technical capacity; trust in government systems, clarity of rules, and ease of access all shape outcomes.

Ultimately, the Devolve-ICMS experiment demonstrates that tax systems can be designed with both equity and efficiency in mind. It does not provide a perfect solution, but it offers a blueprint, one that blends poverty relief with fiscal innovation, exposes the trade-offs between redistribution and incentives, and underscores the importance of tailoring programs to the lived realities of low-income families.