## Brazil’s VAT Reform: Improving Equity

## Source details

**Canonical URL:** [Brazil’s VAT Reform: Improving Equity](https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026132-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2026/english/wpiea2026132-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2026/english/wpiea2026132-source-pdf.pdf.json)

---

### Reform overview and timeline
- The 2023 VAT reform merges five taxes on goods and services (PIS/COFINS, IPI, ICMS, ISS) raised across three levels of government into a dual value added tax and an excise.
- A constitutional amendment was approved at the end of 2023, followed by complementary legislation defining rates and implementation details in 2024 and 2025.
- Implementation will start in 2026 and be completed in 2033; some compensation funds for Brazil’s states will remain in place for almost 50 years.
- Design features:
  - General rate of 28 percent (rounded up from 27.94 percent).
  - Zero rate for most basic goods of the Cesta Basica.
  - Reduced rates on selected goods.
  - A cashback mechanism for the poorest, partially reimbursing VAT paid (with full reimbursement of water and sewage, cooking/natural gas, energy and telecommunications at the federal level).
  - A narrowly defined new excise limited to products harmful to health and environment, including alcohol, sweet beverages, combustion engines, cigarettes, and lottery.
- Primary objective: eliminate distortions and reduce complexity of the current consumption tax system while maintaining revenue neutrality (enshrined in constitutional amendment EC 132/2023).

### Status quo: Brazil’s tax structure and key statistics
- Total tax revenues to GDP in Brazil: 32 percent in 2023.
- OECD average tax-to-GDP: 34 percent in 2023.
- Consumption taxes account for 41 percent of total revenue and 13 percent of GDP in 2023.
- Social security contributions and payroll taxes: 8.6 percent of GDP (OECD average 9.2 percent).
- Subnational revenue collection:
  - Combined state and municipal total tax revenues were 29.6 percent of total revenue in 2022.
  - Most subnational revenues (90 percent) are raised via taxes on goods and services.
- Complexity and fragmentation:
  - Five different taxes across three levels of government generate approximately 10 general tax rates and more than 300 differentiated rates and exemptions.
  - Strong subnational variation across 26 states and 5,568 municipalities.
- SIMPLES regime:
  - Firms with annual turnover of up to BRL 4.8 million per firm (about USD 950,000) eligible for a special tax regime replacing consumption and income taxes with one single payment based on turnover.
- Regressivity indicators (ranges reported in the literature):
  - Poorest decile: consumption tax liability of 20-45 percent of their monetary income.
  - Richest decile: consumption tax liability around 10 percent of monetary income.

### Expected efficiency and macro effects
- Replacing existing consumption taxes with a dual VAT, aligning tax bases, and harmonizing administration is expected to:
  - Reduce complexity and boost efficiency and transparency.
  - Facilitate compliance and reduce administrative costs.
  - Diminish heterogeneity in tax treatment and unleash productivity gains.
  - Boost output by eliminating consumption taxes on intermediate inputs, improving allocation of production factors.

### Data, methodology, and scope of analysis
- Model: World Bank’s SimVat tool calibrated with Brazilian data from the 2018 PNAD household survey and adjusted with latest reform parameters.
- Microsimulations evaluate household impacts, dissecting effects of exemptions, zero/reduced rates, excises, and cashback.
- Key calibration and assumptions:
  - Calibration data: POF 2017/2018 across 185 categories (about 3/4 of total POF expenditure data).
  - Excise assumptions (guided by law PLP68/2024):
    - 46 percent on wine and beer.
    - 62 percent on other alcoholic drinks.
    - 32 percent for sugar-based drinks.
    - 0.25 percent on fuel for vehicles.
    - 250 percent on tobacco.
  - Income definition: disposable, non-adjusted monetary income (includes government transfers, pensions, and wealth variations).
  - Expenditures are non-adjusted and subject to underreporting; total household consumption from POF is about 80 percent of national accounts’ household consumption.
  - Partial equilibrium framework assuming final (pre-tax) consumer prices remain unchanged.
  - Any VAT concession is assumed fully passed on to consumers.
  - Microstatic simulations maintain total revenue collection under the baseline, corresponding to a reference rate of 27.9 percent computed based on approved reform parameters (SERT 2024).

### Main findings — distributional and equity outcomes
- Overall distributional outcome:
  - New VAT results in a relatively equitable distribution of the tax burden for lower- to middle-income households, though the poorest still carry the heaviest load in terms of disposable income.
- Instrument-level impacts:
  - Reduced rates: worsen VAT regressivity and increase the reference rate to 23.9 percent (from baseline full VAT 21.5 percent).
  - Zero rates/exemptions: reduce tax burden for bottom three deciles; raise the reference rate to 24.3 percent under revenue neutrality.
  - Partial cashback: lowers the tax liability of the poorest by 5 percent of income (and by 3 percent of expenditure) relative to a full VAT; keeps the reference rate at 21.8 percent under revenue neutrality.
- Composition effects:
  - Shifting tax burden from manufactured goods to services benefits lower-income households and levies the richest households the most.
  - Pre-reform collection imbalance: 37 percent of consumption taxes collected from manufacturing despite manufacturing’s GDP share of 12 percent; services accounted for 48 percent of collection pre-reform but around 60 percent of GDP.
  - Taxation of manufacturing expected to decline by 1.4 percent of GDP, or about 30 percent.
- Aggregate trade-offs:
  - Equity gains are moderate because concessions require a higher VAT reference rate to maintain revenue neutrality.
- Key quantitative distributional results:
  - Tax liability as a share of income: second to seventh decile average around 14 percent of disposable income; lowest decile 23 percent; richest decile around 9 percent.
  - When measured as share of expenditure: poorest households pay an average tax rate of 16 percent; middle to upper households face 23 percent.

### Box 1 specifics — Subnational revenue and reform sequencing
- Subnational structure:
  - Three-tiered federation: 26 states, a Federal District, and around 5,500 municipalities.
  - Subnational taxes on goods and services provide 90 percent of subnational tax revenue; property taxes provide the other 10 percent.
  - States and municipalities are responsible for education, health, public security and basic sanitation.
- Dual VAT split (reference rate around 28 percent):
  - CBS (federal) covering about 1/3 of the reference rate → CBS would reach 19 percent.
  - IBS (subnational) covering remaining 2/3 → IBS would reach 9 percent.
- New excise (IS) with a maximum rate of 250 percent on specified goods.
- Implementation plan:
  - Phase out federal taxes on goods and services by end-2027.
  - IS starting in January 2027.
  - States and municipal taxes merge into IBS within 2028–2033.
  - Regional compensation funds remain for almost 50 years.
- Harmonized tax base features:
  - Full refund for VAT on intermediate goods.
  - Move from origin to destination principle (zero-rating for exports).
  - Avoids double taxation across supply chains.
- Subnational discretion and anti-incentive rules:
  - States and municipalities have some discretion in setting IBS rate after an initial freeze.
  - IBS and CBS may not be subject to incentives or differentiated regimes except as provided in the Federal Constitution.
- Reduced rates, exemptions, and legacy concessions allowed to continue, including special regimes like free trade zones and SIMPLES.

### Disaggregated policy instruments — simulation highlights
- Baseline full VAT reference rate used in comparisons: 21.5 percent.
- Reduced rates:
  - Worsen regressivity; increase reference rate to 23.9 percent.
- Zero rates/exemptions:
  - Reduce burden for bottom three deciles; increase reference rate to 24.3 percent.
- Partial cashback (reform design):
  - Most effective single tool among assessed instruments for improving equity while containing reference rate increase; lowers poorest tax liability by 5 percent of income and 3 percent of expenditure; reference rate at 21.8 percent under revenue neutrality.

### Pivot to services taxation — implications
- Reform aligns tax collection with services’ contribution to GDP (post-reform: 60 percent vs 48 percent pre-reform).
- Simulated pre-reform revenue contributions:
  - Services: 49.3 percent.
  - Manufacturing, mining, and fuels: 50.7 percent.
  - Agricultural (basic goods): 0.2 percent.
- Expected sectoral change:
  - Manufacturing sector taxation to decline by 1.4 percent of GDP (about 30 percent).
- Distributional effect:
  - Shift benefits lower-income households; higher-income households pay more due to greater services consumption.

### Cashback designs and distributional outcomes
- Full Cashback:
  - Refund of full VAT paid on all goods and services (except excise).
  - Equity effect: disposable income of the poorest up to 25 percent higher.
  - Requires removal of most exemptions/zero/reduced rates to remain revenue neutral.
- Targeted Cashback:
  - Full cashback for VAT on basic goods and services only.
  - Improves equity but leaves poorest with substantial residual liability.
- Reform Cashback (Current Design):
  - Partial cashback on selected goods/services.
  - Provides relief but less effective than full or targeted cashback.
- Operational notes:
  - Real-time compensation via e-invoicing could strengthen compliance and formalization incentives.
  - Full cashback would concentrate refunds in less-informal transactions, potentially improving formalization incentives.

### Relaxing revenue neutrality — scenario and effects
- Comparative benchmarks: Indirect tax collection in Latin America and the OECD averages at 10 ½ percent of GDP (around 2 percent of GDP lower than Brazil’s).
- Reference rate implications:
  - Under reform law revenue neutrality: reference rate up to 28 percent.
  - If aligned with peers: reference rate around 21 percent.
- Budgetary gap: relaxing to peer levels would leave a budgetary gap of around 1-2 percent of GDP.
- Equity impacts measured relative to income:
  - Under reform settings: wedge between lowest and highest decile corresponds to 14 percent of income.
  - Lowering overall collection target by around 20 percent reduces this wedge to 10 percent of income.
  - Poorest households’ tax burden drops to 17 percent of disposable income under lower collection target.
- Equity impacts when measured relative to expenditure:
  - Improvements are less pronounced; wedge narrows less due to limited effect of cashback and zero rates on lower deciles compared with lowering statutory rates at the top.

### Robustness checks and distributional caveats
- Income adjustment:
  - Adjusting monetary income for the lowest decile reduces measured regressivity; quantitative effect: tax liability of the poorest drops by around 7 percent of adjusted income after equalizing consumption whenever expenditure exceeds household resources.
- Cashback access frictions:
  - Lowering take-up to 50 percent increases the tax burden on the poorest by around 3 percentage points; middle to upper deciles broadly unaffected.
- Informality and SIMPLES considerations:
  - In 2024, around 38 percent of total employment was in the informal sector (IBGE/PNAD, PNADC).
  - Poorer households consume more in the informal sector; this could make consumption taxes de facto more progressive than model estimates, making the baseline an upper bound of tax liabilities.
  - SIMPLES and small retailers may charge different effective taxes; SIMPLES estimated average rate: 3.16 percent.
- Overall robustness statement:
  - Main conclusions—instrument ranking and relative equity impacts—remain robust across tested adjustments (income correction, cashback frictions, informal-sector caveats).

### Policy trade-offs, recommendations, and implementation priorities
- Trade-offs:
  - Combining exemptions, zero and reduced rates, and partial cashback mitigates regressivity but at substantial revenue and efficiency cost, raising the reference rate and complicating administration.
  - Reduced rates perform worst on equity and raise reference rate to 23.9 percent.
  - Zero rates/exemptions improve equity for bottom three deciles but raise reference rate to 24.3 percent.
  - Partial cashback is most cost-effective at improving equity while containing reference rate rise.
  - Full cashback delivers strongest redistributive effect (poorest disposable income up to 25 percent higher) but is costly and operationally demanding.
- Short- to medium-term policy options:
  - Expand the role of the cashback system.
  - Further limit exemptions, zero and reduced rates.
  - In the medium-term, consider shifting away from heavy reliance on indirect taxation toward more progressive direct taxation (personal income taxes and property taxes).
- Implementation and governance priorities (2026-2033 transition):
  - Integrate management of CBS/IBS.
  - Fully refund input tax credits on a timely basis.
  - Harmonize revenue administrations and adopt a single, risk-based compliance strategy.
  - Leverage electronic systems for payments and beneficiary identification to aid targeting and reduce fraud.
- Complementary reforms:
  - Reforms in income and capital gains taxation to move toward a more progressive tax system.
  - Revisit SIMPLES and other special regimes to incentivize firm growth and formalization, while managing perceived regressivity.
  - Adopt a holistic approach integrating tax policy with social assistance and broader economic reforms.

### Key reform policy settings and quantified parameters
- General (reference) rate: 28 percent.
- Differential rates:
  - Zero rate: Essential food basket (cesta básica) and specified items.
  - 40 percent of reference rate: Additional food products, education, health, sport, culture; agricultural inputs, cleaning and health care products, regional air passenger transport (*).
  - 70 percent of reference rate: Professional services (intellectual professions, scientific, literary, or artistic nature).
- Special regimes and rates:
  - SIMPLES (estimated average rate of 3.16 percent).
  - Fuels: levied on production only, credits not allowed in the trade sector.
- Cashback settings:
  - 100 percent CBS and 20 percent IBS cashback (an overall 46.6 percent) on acquisition of cooking gas (canisters up to 13 kg), natural piped gas supplies, domestic electricity, telecommunications, water, and sewage bills.
  - 20 percent IBS and CBS for all other purchases.
- Transition and governance timing:
  - Transition phase explicitly referenced as 2026-2033.
- Progressivity indicators (selected values):
  - Gini Before tax: 0.529
  - Gini Full VAT: 0.401
  - Gini Reform: 0.443
  - Kakwani Reform vs Full VAT: 0.042
  - Kakwani Full Cashback vs Full VAT: 0.099
  - Kakwani Reduced rate vs Full VAT: 0.018
  - Kakwani Exemption vs Full VAT: 0.036
  - Kakwani Targeted Cashback vs Full VAT: 0.042
  - Kakwani Full Cashback vs Reform: 0.057
  - Kakwani Reduced rate vs Reform: -0.024
  - Kakwani Exemption vs Reform: -0.006
  - Kakwani Targeted Cashback vs Reform: 0.001

*Brazil’s VAT Reform: Improving Equity — Working Paper No. WP/2026/132*

### Introduction _________________________________________________________________ 5

### Introduction

### Reform overview and timeline
- The 2023 VAT reform merges five taxes on goods and services (PIS/COFINS, IPI, ICMS, ISS) raised across three levels of government into a dual value added tax and an excise.
- A constitutional amendment was approved at the end of 2023, followed by complementary legislation defining rates and implementation details in 2024 and 2025.
- Implementation will start in 2026 and be completed in 2033; some compensation funds for Brazil’s states will remain in place for almost 50 years.
- Design features:
  - General rate of 28 percent (rounded up from 27.94 percent).
  - Zero rate for most basic goods of the Cesta Basica.
  - Reduced rates on selected goods.
  - A cashback mechanism for the poorest, partially reimbursing VAT paid (with full reimbursement of water and sewage, cooking/natural gas, energy and telecommunications at the federal level).
  - A narrowly defined new excise limited to products harmful to health and environment, including alcohol, sweet beverages, combustion engines, cigarettes, and lottery.
- Primary objective: eliminate distortions and reduce complexity of the current consumption tax system while maintaining revenue neutrality (enshrined in constitutional amendment EC 132/2023).

### Status quo: Brazil’s current tax system (key statistics and features)
- Total tax revenues to GDP in Brazil: 32 percent in 2023.
- OECD average tax-to-GDP: 34 percent in 2023.
- Consumption taxes account for 41 percent of total revenue and 13 percent of GDP in 2023.
- Social security contributions and payroll taxes: 8.6 percent of GDP (OECD average 9.2 percent).
- Brazil’s indirect taxation complexity:
  - Five different taxes across three levels of government generate approximately 10 general tax rates and more than 300 differentiated rates and exemptions.
  - Cascading taxes arise from overlapping bases and incomplete credits for intermediate inputs, with strong subnational variation (26 states and 5,568 municipalities).
- Subnational revenue collection:
  - Combined state and municipal total tax revenues were 29.6 percent of total revenue in 2022.
  - Most subnational revenues (90 percent) are raised via taxes on goods and services.
- SIMPLES regime: firms with annual turnover of up to BRL 4.8 million per firm (about USD 950,000) are eligible for a special tax regime that replaces consumption and income taxes with one single payment based on turnover.
- Regressivity indicators (ranges reported in the literature):
  - Poorest decile: consumption tax liability of 20-45 percent of their monetary income.
  - Richest decile: consumption tax liability around 10 percent of monetary income.

### Expected efficiency and macro effects
- By replacing existing consumption taxes with a dual VAT, aligning tax bases across municipalities, states, and the federal level, and harmonizing administration, the reform is expected to:
  - Reduce complexity and boost efficiency and transparency.
  - Facilitate compliance and reduce administrative costs.
  - Diminish heterogeneity in tax treatment and unleash productivity gains (references: Cebreiro et al. 2025; Cavalcanti et al 2024).
  - Boost output by eliminating consumption taxes on intermediate inputs, improving allocation of production factors (Domingues and Freire Cardoso 2020; Badel and Kolerus 2025).

### Data, methodology, and scope of analysis
- The paper uses the World Bank’s SimVat tool calibrated with Brazilian data from the 2018 PNAD household survey and adjusted with the latest reform parameters for tax rates and cashback.
- Microstatic simulations evaluate the impact of reform policies on household income and consumption.
- The analysis dissects the relative impact of exemptions, zero and reduced rates, excises, and the cashback mechanism.
- The paper assesses distributional implications of the shift from goods to services taxation and simulates alternative policy settings including relaxing revenue neutrality.

### Main findings (distributional and equity outcomes)
- Overall distributional outcome:
  - The new VAT would result in a relatively equitable distribution of the tax burden for lower- to middle-income households, though the poorest still carry the heaviest load in terms of disposable income.
- Instrument-level impacts:
  - Reduced rates: reform’s policy settings for reduced rates worsen the VAT’s regressivity.
  - Zero rates: mainly applied to Cesta Basica products, reduce the tax burden of the bottom three deciles by 7 percent of their income compared with a full VAT without concessions.
  - Cashback mechanism: an important relief for poorer households, lowering their tax liability by almost 10 percent of their income relative to a full VAT.
- Composition effects:
  - Shifting the tax burden from manufactured goods to services benefits lower-income households given consumption patterns; this shift levies the richest households the most.
- Aggregate equity trade-offs:
  - Equity gains remain moderate because a higher VAT reference rate is needed to compensate for revenue loss from concessions to maintain revenue neutrality.
- Policy-simulation results:
  - Targeting the poorest with an expanded cashback while removing other concessions would improve their household income by 25 percent (via tax reduction) and allow for a lower reference rate.
  - Relaxing revenue neutrality and aligning revenue collection closer with peers improves regressivity relative to income but not relative to consumption.

### Literature and interpretation notes
- Measurement matters: VAT regressivity depends critically on metric choices (income vs. expenditure), treatment of savings and informality, and assumptions about pass-through to consumer prices.
- Evidence summarized:
  - Single-year VAT burden measured as percent of income often shows high regressivity; including savings and informality can reduce measured regressivity.
  - When measured relative to expenditure, VAT is often proportional or progressive in some studies.
  - Reduced and zero rates are common but theoretical and empirical literature finds limited support for multi-rate VATs as the best redistribution tool when targeted cash transfers are available.
  - Targeted cash transfers, especially if delivered in real time at point of purchase (e.g., digital cashback), are an efficient instrument to compensate low-income households and can enable a progressive VAT design.

*Source: IMF Working Paper — Introduction section of "Brazil’s VAT Reform: Improving Equity" (content unit from provided PDF).*

### Box 1. Brazil’s Subnational Revenue Collection

### Box 1. Brazil’s Subnational Revenue Collection

### Subnational structure and revenue shares
- Brazil is a three-tiered federation of 26 states, a Federal District, and around 5500 municipalities.
- The Constitution of 1988 granted broad powers to states and municipalities in tax and expenditure functions.
- Subnational taxes on goods and services provide 90 percent of subnational tax revenue; the other 10 percent comes from property taxes.
- In 2022, combined shares of state and municipal tax revenues amount to 29.6 percent of total revenue (comparators: Germany 33.6, Australia 18.4, Canada 47.5).
- States and municipalities are responsible for education, health, public security and basic sanitation.

### Tax assignments and implications for VAT reform
- In LAC, typical subnational taxes: property taxes, motor-vehicle licenses, taxes on specific services, municipal fees.
- In OECD countries, subnational revenue often comes from taxes on income and profits; in Brazil, consumption taxes constitute the main source of revenue at the subnational level.
- High reliance on consumption taxes at the subnational level amplifies the challenges the VAT reform faces.

### The 2023 VAT Reform: structure and timeline
- The reform will gradually replace existing consumption taxes with a dual VAT by 2033.
- Dual VAT design:
  - Federal component (CBS) covering about 1/3 of the reference rate.
  - Subnational component (IBS) covering the remaining 2/3 of the reference rate.
- With a reference rate around 28 percent (as currently assumed), the CBS and IBS would reach 19 and 9 percent respectively.
- A new excise (IS) with a maximum rate of 250 percent would focus on goods harmful to the environment and health.
- Implementation plan:
  - Phase out federal taxes on goods and services by end-2027.
  - IS starting in January 2027.
  - Taxes of states and municipalities will gradually merge into the IBS within 2028–2033.
  - Regional compensation funds will stay in place for a transition period of almost 50 years.
- Political economy shaped compromises: a single-rate structure was originally considered (PEC 45/2019) but not adopted.
- A subnational piggy-back VAT is characterized as particularly efficient to raise revenue relative to practicable alternatives.

### Harmonized tax base, subnational discretion, and anti-incentive rules
- One major achievement: a harmonized tax base for both federal and subnational VAT components, anchored in the constitution since December 2023.
- Features of the unified base:
  - Full refund for VAT on intermediate goods.
  - Move from origin to destination principle (implies zero-rating for exports).
  - Avoids double taxation of consumption across supply chains and increases transparency.
- Subnational fiscal sovereignty:
  - States and municipalities have some discretion in setting their IBS rate after an initial freeze period.
  - Legislation explicitly establishes that IBS and CBS may not be subject to incentives and financial or tax benefits or specific, differentiated, or favored tax regimes, except as provided in the Federal Constitution.

### Reduced rates, exemptions, and legacy concessions
- The reform allows reduced (including zero) rates and exemptions and maintains certain concessions and specific regimes.
- Rate structure beyond the reference rate:
  - Reference rate: 28 percent.
  - Additional applicable rates: zero, 40 and 70 percent of the reference rate, and exemptions.
- Zero rates applied to selected “essential foods” from the Cesta Basica.
- Reduced rates levied on additional food items, goods and services related to public administration, local transportation, books/magazines/newspapers, and health and education services provided by non-profit organizations.
- Special regimes maintained, including free trade zones (e.g. Manaus) and the tax regime for small enterprises (SIMPLES).
  - SIMPLES businesses can opt-in to VAT taxation benefiting from tax credits while maintaining preferred rates for other taxes.
- Many concessions reflect preferential treatments inherited from the various taxes the dual VAT replaces; political economy and acquired rights explain continuation of many concessions.

### Digital cashback mechanism for lower-income households
- A partial cashback system targets vulnerable families registered in Brazil’s Cadastro Unico.
- Eligibility: households with income below half of the minimum wage per capita qualify.
- Cashback rates:
  - 100 percent CBS and 20 percent IBS cashback (an overall 46.6 percent) on acquisition of cooking gas (canisters up to 13 kg), natural piped gas supplies, domestic electricity, telecommunications, water, and sewage bills.
  - General 20 percent IBS and CBS cashback on other purchases.
- The cashback takes the form of a (partial) refund (or welfare transfer) linked to e-invoices.
- Complementary Law No. 214/2025 allows all levels of government to establish (by law) higher refund percentages and simplified refund calculation mechanisms in exceptional cases.
- Compatibility checks with available family income and other mechanisms will help avoid fraud; precise anti-fraud mechanisms remain to be determined.

### Equity analysis — data, methodology, and key assumptions
- VAT liabilities estimated by income and consumption deciles using World Bank’s SimVAT model adjusted with approved reform policies.
- Calibration data: national budget survey (Pesquisa de Orçamentos Familiares, POF) 2017/2018 across 185 categories (about 3/4 of total POF expenditure data).
- Excise assumptions (guided by law PLP68/2024):
  - 46 percent on wine and beer.
  - 62 percent on other alcoholic drinks.
  - 32 percent for sugar-based drinks.
  - 0.25 percent on fuel for vehicles.
  - 250 percent on tobacco.
- Income definition: disposable, non-adjusted monetary income (excludes non-monetary components like in-kind transfers or imputed rents but includes government transfers, pensions, and wealth variations).
- Expenditures are non-adjusted and subject to underreporting; total household consumption from POF is about 80 percent of national accounts’ household consumption.
- Model assumptions and limitations:
  - Partial equilibrium framework assuming final (pre-tax) consumer prices remain unchanged.
  - Any VAT concession is assumed fully passed on to consumers.
  - Exemptions are treated as zero rates; any non-recoverable input VAT is assumed fully borne by the producer with no impact on final consumer price.
  - Microstatic simulations maintain total revenue collection under the baseline, corresponding to a reference rate of 27.9 percent computed based on approved reform parameters (SERT 2024).

### Distributional outcomes and findings
- Under approved reform policies:
  - Tax liability as a share of income is relatively similar for the second to seventh decile, with consumption taxes averaging around 14 percent of disposable income.
  - Lowest income decile faces a higher liability amounting to 23 percent of their income.
  - Richest decile faces around 9 percent of income in tax liability.
- Overall assessment:
  - The new VAT reflects some degree of regressivity but performs relatively well compared with other VAT systems despite a high reference rate.
  - When measured as share of expenditure, the new VAT shows a relatively progressive tax structure:
    - Poorest households pay an average tax rate of 16 percent on their expenditure on goods and services.
    - Middle to upper income households face an average rate of 23 percent.
  - These average tax rates are noted as among the highest worldwide, but the relative progressivity in expenditure is highlighted as a positive outcome.

### Box 2 — Scenario Map (Reform Scenario / Baseline)
- Description: The approved VAT reform merges five existing taxes into a dual VAT and an excise, with a VAT reference rate of 28 percent, zero rates for basic goods, reduced rates for selected items, and a cashback mechanism for the poorest households.
- Equity Impact: The reform achieves a relatively equitable distribution of the tax burden for lower- to middle-income households, but the poorest still bear a heavier load relative to their disposable income. The combination of zero rates, reduced rates, and cashback mitigates regressivity but requires a higher VAT reference rate to maintain revenue neutrality.

*Source: Box 1. Brazil’s Subnational Revenue Collection (excerpt) from the IMF working paper "Brazil’s VAT Reform: Improving Equity."*

### 2. Disaggregated Policy Instruments

### 2. Disaggregated Policy Instruments

### Overview of individual instrument simulations
- The paper simulates the equity impact of each policy instrument introduced in the reform individually, compared to a hypothetical full VAT (single rate, no concessions).
- Baseline full VAT reference rate used in comparisons: 21.5 percent.

### Reduced rates
- Finding: Reduced rates worsen regressivity, as benefits are spread across all income groups.
- Mechanism: Reduced rates are relatively expensive, pushing up the VAT reference rate under revenue neutrality.
- Quantitative effect cited: reference rate increases to 23.9 percent from 21.5 percent under the full VAT, penalizing lower income households more in relative terms.

### Zero rates and exemptions
- Finding: Zero rates and exemptions, focused on basic goods (Cesta Basica), reduce the tax burden for the bottom three deciles but are costly because all households benefit.
- Mechanism: Require a higher VAT reference rate under revenue neutrality.
- Quantitative effect cited: reference rate reaches 24.3 percent under revenue neutrality when zero rates/exemptions are applied.

### Partial cashback (reform design)
- Finding: Partial cashback targeted at vulnerable households (those registered in Cadastro unico) is the most effective single tool for improving equity among the assessed instruments, but does not fully eliminate VAT regressivity.
- Quantitative effects versus full VAT baseline:
  - Lowers the tax liability of the poorest by 5 percent of income.
  - Lowers the tax liability of the poorest by 3 percent of expenditure.
- Mechanism: Tailoring refunds to vulnerable households keeps the instrument more affordable and the reference rate at 21.8 percent to maintain revenue neutrality.

---

### 3. Pivot to Services Taxation

### Description
- Reform shifts tax burden from goods (specially manufactured goods) to services, aligning tax collection with services’ contribution to GDP.
- Post-reform share of services in GDP: 60 percent vs 48 percent pre-reform (statement: "aligning tax collection more closely with the actual contribution of services to GDP (60 percent vs 48 percent pre-reform)").

### Distributional effect
- Finding: Shift benefits lower-income households (whose consumption baskets are more goods-heavy) while higher-income households pay more due to greater consumption of services.
- Sectoral collection imbalances pre-reform:
  - 37 percent of all consumption taxes collected from the manufacturing sector despite this sector’s contribution to GDP of 12 percent.
  - Services, including utilities, accounted for 48 percent of collection pre-reform but around 60 percent of GDP.
- Reform effects:
  - Harmonization of bases and rates, full input tax crediting, and a cap of the reference rate at 28 would shift tax collection from manufacturing to services.
  - Taxation of the manufacturing sector is expected to decline by 1.4 percent of GDP, or about 30 percent (Cebreiro et al, 2025).
- Simulated revenue contribution assumptions for pre-reform scenario used in distributional simulation:
  - Services revenue contribution: 49.3 percent.
  - Contribution by manufacturing, mining, and fuels: 50.7 percent.
  - Agricultural (basic goods): 0.2 percent.

---

### 4. Expanding the Cashback Mechanism

### Three cashback designs simulated (Full VAT as baseline)
- Full Cashback:
  - Eligible households receive a refund of the full VAT paid on all goods and services (except excise).
  - Equity effect: Greatest equity gains, raising the disposable income of the poorest by up to 25 percent.
  - Design note: Under this scenario the VAT neither features reduced rates nor zero rates nor exemptions; refunds exclude excise (IS).
- Targeted Cashback:
  - Full cashback for VAT paid on basic goods and services only (including Cesta Basica items, water, energy, health, education, transportation).
  - Equity effect: Improves equity but leaves poorest with substantial residual tax liability.
- Reform Cashback (Current Design):
  - Partial cashback on selected goods/services as approved in the reform.
  - Equity effect: Provides relief but is less effective than full or targeted cashback options.

### Distributional and implementation observations
- Maximum benefit for the poor achieved under full cashback, particularly if refunds apply to the lower two (and a half) deciles (excluding excises).
- Targeted cashback leaves poorest with highest tax liability among deciles despite benefits.
- Partial cashback extends to a limited number of eligible goods and allows partial refunds only, leaving substantial residual tax liability for poorer households.
- Operational considerations:
  - Real-time compensation via e-invoicing and digital payments could strengthen compliance and formalization incentives.
  - Full cashback refunds likely to be concentrated in manufactured goods and services with lower informality, potentially improving formalization incentives for buyers and sellers.

---

### 5. Relaxing Revenue Neutrality

### Scenario description
- The paper simulates lowering the overall indirect tax burden to levels closer to Latin American and OECD averages by relaxing revenue neutrality.
- Comparative benchmarks: Indirect tax collection in Latin America and the OECD averages at 10 ½ percent of GDP, around 2 percent of GDP lower than Brazil’s.

### Quantitative and distributional effects
- Reference rate implications:
  - Under reform law revenue neutrality: reference rate up to 28 percent (reform cap mentioned).
  - If revenue neutrality relaxed and collection aligned with peers: reference rate around 21 percent (compared with 28 percent under the reform).
- Budgetary gap: relaxing to peer levels would leave a budgetary gap of around 1-2 percent of GDP.
- Potential policy responses to gap: alternative, potentially more progressive fiscal policies (e.g., rebalancing from indirect to direct taxation such as personal income taxes and property taxes).
- Equity impacts when measured relative to income:
  - Under reform settings: the wedge between the tax liability of the lowest and the highest decile corresponds to 14 percent of income.
  - Lowering overall collection target by around 20 percent reduces this wedge to 10 percent of income.
  - Poorest households’ tax burden drops to 17 percent of disposable income under lower collection target.
- Equity impacts when measured relative to expenditure:
  - Improvements are less pronounced; the wedge between lowest and highest decile becomes smaller when measured as a share of expenditure, reflecting weaker effects from cashback and zero rates on lower deciles compared to lowering statutory rates at the top.

---

### 6. Robustness Checks and Distributional Caveats

### Income adjustment
- Adjusting monetary income for the lowest decile (to account for underreporting) improves the equity pattern but does not eliminate VAT regressivity.
- Quantitative effect: Tax liability of the poorest drops by around 7 percent of adjusted income after equalizing consumption whenever expenditure exceeds household resources; overall VAT pattern remains regressive.

### Cashback access frictions
- Introducing access frictions to cashback reduces the equity gains for the lower three deciles compared to the reform outcome.
- Overall conclusions about instrument ranking remain robust to these access-friction adjustments.

### Informality and SIMPLES considerations
- Informality:
  - High informality in Brazil (note provided: "In 2024, around 38 percent of total employment was in the informal sector (IBGE/PNAD, PNADC)").
  - Poor households consume more in the informal sector; this could make consumption taxes de facto more progressive than model estimates, implying the baseline is an upper bound of tax liabilities.
  - Counterpoint: Lara Ibara et al (2021) argue informality would not materially change distributional results given food sales concentration in supermarkets and most retail food taxed at zero or low rates.
- SIMPLES:
  - Poorer households purchase more from small retailers and service providers operating under SIMPLES subject to lower (fixed) taxation, not modeled here.
  - Taxes charged under SIMPLES do not account for zero rates and other concessions.

### Overall robustness statement
- The main conclusions—instrument ranking and relative equity impacts—remain robust across the tested adjustments (income correction, cashback frictions, informal-sector caveats).

---

### 7. Policy trade-offs and summary conclusions

- Combining exemptions, zero and reduced rates, and partial cashback mitigates regressivity relative to a full VAT without concessions and relative to pre-reform patterns, but at substantial cost in revenue and efficiency.
- Concessions increase the reference rate needed under revenue neutrality, complicate compliance and administration, raise opportunities for avoidance and fraud, and can create incentives for rent-seeking and further base-narrowing.
- Among instruments assessed:
  - Partial cashback (targeted to vulnerable households) is the most cost-effective at improving equity while containing the reference rate rise (reference rate at 21.8 percent under revenue neutrality with partial cashback).
  - Full cashback delivers the strongest redistributive effect (poorest disposable income up to 25 percent higher under full cashback scenario) but is more costly and operationally demanding.
  - Reduced rates have the weakest equity performance and increase regressivity by raising the reference rate to 23.9 percent.
  - Zero rates/exemptions improve equity for the bottom three deciles but are costly and raise the reference rate to 24.3 percent.

---

### Progressivity indicators (selected values from Table 1)
- Gini Before tax: 0.529
- Gini Full VAT: 0.401
- Gini Reform: 0.443
- Kakwani Reform vs Full VAT: 0.042
- Kakwani Full Cashback vs Full VAT: 0.099
- Kakwani Reduced rate vs Full VAT: 0.018
- Kakwani Exemption vs Full VAT: 0.036
- Kakwani Targeted Cashback vs Full VAT: 0.042
- Kakwani Full Cashback vs Reform: 0.057
- Kakwani Reduced rate vs Reform: -0.024
- Kakwani Exemption vs Reform: -0.006
- Kakwani Targeted Cashback vs Reform: 0.001

*Source: wpiea2026132-source-pdf*

### conclusions remain in place. Moreover, while this adjustment takes care of potential biases for lower deciles,

### Brazil’s VAT Reform: Improving Equity

### Robustness of Distributional Findings
- Adjustment to account for potential biases in lower deciles preserves main conclusions; income of upper deciles is likely underestimated given lack of information on savings to compute lifetime income. The depicted patterns could therefore provide an upper and lower bound of regressivity.
- Lowering take-up of cashback increases regressivity with little effect on middle to higher income households:
  - Baseline assumes eligible households receive full benefits without take-up or access frictions (100 percent take-up).
  - Lowering take-up to 50 percent – holding everything else constant under the reform scenario – increases the tax burden on the poorest by around 3 percentage points.
  - Tax liabilities for middle to upper deciles remain broadly unaffected under reduced cashback take-up.
- Incomplete cashback realization risks:
  - Areas with lower administrative capacity (typically correlated with lower per capita incomes) may see materially reduced realized benefits.
  - Experiences with incomplete cashback could incentivize households to purchase informally, lowering both their tax liability and the equity impact of the reform.
- Incentives for firms:
  - There remains a strong incentive for companies to remain VAT (and thus refund) eligible and therefore process VAT electronically, increasing the likelihood that registered transfer recipients are correctly identified.

### Main Conclusions and Equity Assessment
- The approval of the dual VAT in Brazil is an example of a growth-enhancing structural reform that incorporates distributional consequences.
- Expected macro and distributional benefits:
  - Materially enhance revenue collection and reduce compliance gaps by addressing longstanding inefficiencies, thereby boosting productivity and formal sector growth.
  - Tackles equity considerations by introducing mechanisms—such as a cashback system and zero rates—that help mitigate regressive impacts and reduce the disproportionate tax burden on lower-income households.
  - The new VAT’s distributional pattern is described as relatively equitable and an important step toward addressing regressivity in Brazil’s tax system and helping alleviate Brazil’s high inequality.
- Limitations to full progressivity:
  - Proposed reduced rates, intended to support equity, inadvertently undermine it.
  - The combination of instruments is relatively expensive, pushing up the reference rate levied on all other goods.
  - The cashback is partial in nature under current settings.

### Reform Design and Policy Options to Strengthen Equity
- Short- to medium-term policy options:
  - Expand the role of the cashback system.
  - Further limit exemptions, zero and reduced rates.
  - In the medium-term, consider a strategic shift away from heavy reliance on indirect taxation to align Brazil's tax system with international peers and reduce the overall burden on the most vulnerable.
- Trade-offs and feasibility:
  - Full cashback is both desirable and technically feasible but would require the removal of most exemptions, zero and reduced rates to maintain revenue neutrality – a politically highly challenging step.
  - Broadening the scope of the reform beyond consumption taxation in 2024 would have been very challenging; rebalancing could be feasible once the dual VAT is successfully under implementation.

### Implementation, Governance, and Operational Priorities
- To realize the reform’s full potential during the transition phase (2026-2033), address governance and operational challenges, including:
  - Integrate the management of the CBS/IBS.
  - Fully refund input tax credits on a timely basis.
  - Harmonize revenue administrations to leverage synergies across levels of government and enable implementation of a single, risk-based compliance strategy.
  - Leverage electronic systems for payments and beneficiary identification to aid targeting support and reduce fraud (Cebreiro et al. 2025).
- Complementary reforms needed for lasting equity improvements:
  - Reforms in income and capital gains taxation to move toward a more progressive overall tax system.
  - Revisiting SIMPLES and other special regimes could be perceived as regressive, but may set incentives for smaller companies to grow, formalize, and become more productive, thereby lowering inequality.
  - A holistic approach integrating tax policy with social assistance and broader economic reforms is essential to ensure growth and equity objectives are realized for all segments of society.

### Key Reform Policy Settings and Quantified Parameters
- General (reference) rate: 28 percent.
- Differential rates:
  - Zero rate: Essential food basket (cesta básica) and specified items.
  - 40 percent of reference rate: Additional food products, education, health, sport, culture; agricultural inputs, cleaning and health care products, regional air passenger transport (*).
  - 70 percent of reference rate: Professional services (intellectual professions, scientific, literary, or artistic nature).
- Special regimes and rates:
  - SIMPLES (estimated average rate of 3.16 percent).
  - Fuels: levied on production only, credits not allowed in the trade sector.
- Cashback settings:
  - 100 percent CBS and 20 percent IBS for Acquisition of cooking gas (canisters weighing up to 13 kg), natural piped gas supplies, domestic electricity, telecommunications, water, and sewage bills.
  - 20 percent IBS and CBS for all other purchases.
- Transition and governance timing:
  - Transition phase explicitly referenced as 2026-2033.

*Italic source attribution:*  
*Brazil’s VAT Reform: Improving Equity — Working Paper No. WP/2026/132*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026132-source-pdf.pdf_