Theme 25 of 43
Federal Budget, Parliamentary Amendments & Budget Execution
The federal government spends twenty times more on debt interest than on Bolsa Família, and R$ 40 billion in earmarks went out with no identified recipient.
Budget execution by managing unit, published by the National Treasury, parliamentary earmarks registered by the Office of the Comptroller General, tax revenue by state from the Federal Revenue Service, and rural credit operations from the Central Bank together form the most complete portrait available of how federal public money is decided, collected and actually spent. That portrait shows a budget where a large share of what's authorized never becomes delivered service, a relevant slice goes with no trace of a recipient at all, and debt service alone consumes twenty times more resources than the country's main cash-transfer program.
Total parliamentary earmarks (2022-2025)
Between 2022 and 2025, Congress committed R$ 152.7 billion across 25,518 parliamentary earmarks, averaging R$ 5.98 million per earmark. That is almost what Bolsa Família paid in a full year (R$ 163.5 billion in 2024) — except decided by lawmakers, committees and state caucuses, without the targeting logic toward those who need it most that shapes the design of a social program. In 2026, through September, another R$ 40.1 billion has been committed.
| Indicator | Value |
|---|---|
| Total earmarks | 25,518 |
| Total value | R$ 152.7 billion |
| Average value per earmark | R$ 5.98 million |
In four years, parliamentary earmarks added up to almost a full year of Bolsa Família — R$ 152.7 billion.
Budget execution: how much becomes real spending?
In none of the years analyzed did the budget execution rate come close to 100%. In 2019, only 44% of what was committed became liquidated spending; even in 2024, the series' best-performing year, 30% still went without full execution. The gap between what's promised and what's delivered isn't a technical detail — it's the distance between the political discourse of the budget and its actual delivery to the population.
| Year | Committed (R$ bi) | Liquidated (R$ bi) | Rate |
|---|---|---|---|
| 2018 | 12.0 | 5.6 | 46% |
| 2019 | 13.9 | 6.1 | 44% |
| 2020 | 37.5 | 18.2 | 49% |
| 2021 | 33.4 | 16.0 | 48% |
| 2022 | 25.5 | 17.2 | 68% |
| 2023 | 35.4 | 22.1 | 62% |
| 2024 | 44.8 | 31.5 | 70% |
Historically, more than half of the authorized budget is never actually executed — the difference disappears from public radar without anyone needing to explain why.
General Rapporteur's earmarks: extreme concentration
In 2022, the last year they existed, the Budget general rapporteur's earmarks reached R$ 8.6 billion committed, directed mostly toward health; between 2020 and 2022 they totaled R$ 46.6 billion. It's a concentration of decision-making power that escapes the usual process of parliamentary dispute spread across hundreds of representatives and senators, concentrating in a single role — the rapporteur — the capacity to decide where billions of reais will be spent.
| Function | Value (R$ bi) |
|---|---|
| Health | R$ 6.36 bi |
| Social Assistance | R$ 0.96 bi |
| Multiple | R$ 0.56 bi |
| Urban planning | R$ 0.26 bi |
| Education | R$ 0.21 bi |
| Sports | R$ 0.19 bi |
A single role — the general rapporteur — came to decide where R$ 8.6 billion went in one year, a concentration of decision-making power incompatible with any notion of a democratically contested budget.
Sectoral concentration of earmarks
More than half of all parliamentary earmark value, 51.8%, goes to the Health function — R$ 79.2 billion. Public Security, with 1.4% of the total, and Social Assistance, with 2.6%, are left with residual fractions against that amount. The sectoral concentration of earmarks reveals where lawmakers concentrate electoral bargaining power, which doesn't always coincide with where social need is greatest.
| Function | % of Total | Value (R$ bi) |
|---|---|---|
| Health | 51.8% | R$ 79.2 bi |
| Special charges | 16.8% | R$ 25.6 bi |
| Urban planning | 7.6% | R$ 11.6 bi |
| Agriculture | 4.4% | R$ 6.7 bi |
| Education | 3.6% | R$ 5.6 bi |
| Social Assistance | 2.6% | R$ 3.9 bi |
| Public Security | 1.4% | R$ 2.1 bi |
More than half of all parliamentary earmarks go to health — a single sector concentrates the entire Congress's electoral bargaining power.
Tax structure: companies vs. workers
Corporate income tax (IRPJ), levied on company profits, raised R$ 315.2 billion in 2022, against R$ 57.9 billion for personal income tax (IRPF), levied on individual income — a ratio of more than five to one. That doesn't by itself mean the Brazilian tax system is progressive: a relevant part of total revenue still comes from taxes on consumption, which weigh proportionally more on those who earn less, and don't show up in this specific breakdown.
| Year | IRPF (R$ bi) | IRPJ (R$ bi) | IPI (R$ bi) |
|---|---|---|---|
| 2020 | 41.4 | 173.9 | 33.3 |
| 2021 | 56.2 | 248.3 | 41.9 |
| 2022 | 57.9 | 315.2 | 36.3 |
| 2023 | 58.6 | 300.3 | 32.2 |
| 2024 | 33.8 | 153.0 | 17.1 |
IRPJ raises three to five times more than IRPF, but the system as a whole continues to weigh disproportionately on those who earn less, via indirect taxation.
Unpaid carryovers: the government's hidden debt
In 2020, the government registered R$ 290 billion in unpaid carryovers ("restos a pagar") — resources committed but not liquidated within the fiscal year, which remain available for use in future years without new budget approval. Every year, a fraction of this stock — between R$ 8 billion and R$ 15 billion — is simply canceled, money that had already been committed on paper but never turned into service delivered to the population.
| Year | Amount Registered (R$ bi) | Amount Canceled (R$ bi) |
|---|---|---|
| 2019 | R$ 86 bi | R$ 12 bi |
| 2020 | R$ 290 bi | R$ 8 bi |
| 2021 | R$ 180 bi | R$ 15 bi |
| 2022 | R$ 120 bi | R$ 10 bi |
Between R$ 12 billion and R$ 15 billion in unpaid carryovers are canceled every year — money that was promised on paper and never became delivered service.
Discretionary spending: what gets cut
Education and Health, politically protected functions, kept execution rates of 85% and 90% respectively of their allocations. Science, with just 65% executed, and Environment, with 55%, are systematically the first targets for spending freezes when the government needs to cut spending during the fiscal year — a pattern that penalizes precisely the areas of investment with slower, less visible short-term returns.
| Function | Allocation | Executed |
|---|---|---|
| Education | R$ 120 bi | 85% |
| Health | R$ 140 bi | 90% |
| Science | R$ 15 bi | 65% |
| Environment | R$ 8 bi | 55% |
Education and health are protected from spending freezes; science and the environment are always the first to be cut when the budget tightens.
Public debt: debt service
The federal government spends R$ 700 billion a year just on public-debt interest and charges — twenty times more than the R$ 35 billion a year earmarked for Bolsa Família. This imbalance between the cost of public debt and investment in cash transfers shows how much debt service dominates the fiscal space available for social policy.
| Indicator | Value/year |
|---|---|
| Interest and charges | R$ 700 bi |
| Bolsa Família | R$ 35 bi |
| Debt service | 20x BF |
Brazil pays twenty times more in debt interest than it spends on Bolsa Família — a budget where the cost of indebtedness consumes the space that could go to social policy.
Tax burden: composition
Indirect taxes — such as ICMS and IPI, charged on consumption — account for 55% of the entire Brazilian tax burden. Direct taxes, like income tax and CSLL, hold 35%, and wealth taxes, which could directly target large fortunes, hold less than 5% of total revenue. Since consumption tax applies equally to rich and poor in nominal terms, it weighs proportionally more on those who earn less.
| Type | % Total |
|---|---|
| Indirect (ICMS, IPI) | 55% |
| Direct (IR, CSLL) | 35% |
| Wealth taxes | <5% |
| Fees | 5% |
More than half of all Brazilian tax revenue comes from indirect taxes — paid proportionally more by those with lower income.
Secret budget: spending with no trace
Between 2020 and 2022, R$ 40 billion in earmarks were executed with no clear identification of recipient — the so-called "secret budget." Without knowing who received what, any real oversight becomes impossible, whether by the press, control bodies, or civil society itself. It's the most opaque point in the entire federal budget cycle.
| Indicator | Value |
|---|---|
| Secret earmarks (2020-2022) | R$ 40 bi |
| Recipients | Undisclosed |
| Oversight | Impossible |
R$ 40 billion in earmarks with no recipient identification — total opacity at a moment when the budget should be the most public document of all.
After the secret budget, committee earmarks
The Federal Supreme Court declared rapporteur earmarks unconstitutional at the end of 2022, and they disappear from the records from 2023 on. The money, however, did not return to the Executive: total committed earmarks rose from R$ 25.5 billion in 2022 to R$ 47.1 billion in 2025, the highest in the series. The rapporteur's space was taken over by committee earmarks — residual until 2022 and R$ 11.7 billion in 2024 — and by individual earmarks, which more than doubled.
| Year | Individual, defined purpose | Individual "Pix" | State caucus | Committee | Rapporteur | Total |
|---|---|---|---|---|---|---|
| 2019 | 8.5 | — | 4.7 | 0.2 | 0.4 | 13.9 |
| 2020 | 8.6 | 0.6 | 6.5 | 0.6 | 21.2 | 37.5 |
| 2021 | 7.4 | 2.0 | 7.2 | — | 16.7 | 33.4 |
| 2022 | 7.4 | 3.3 | 5.8 | 0.3 | 8.6 | 25.5 |
| 2023 | 13.8 | 7.1 | 7.6 | 6.9 | — | 35.4 |
| 2024 | 17.0 | 7.7 | 8.4 | 11.7 | — | 44.8 |
| 2025 | 17.3 | 7.0 | 11.6 | 11.2 | — | 47.1 |
| 2026* | 16.8 | 6.1 | 7.5 | 9.7 | — | 40.1 |
Special transfers — the so-called "Pix earmarks", which land directly in state and municipal coffers with no defined purpose — went from R$ 0.6 billion in 2020 to about R$ 7 billion a year since 2023. With them and with committee earmarks, whose individual sponsor does not appear in the record, a growing share of the money remains hard to trace back to whoever directed it.
The end of the secret budget did not shrink earmarks: they nearly doubled between 2022 and 2025, and the rapporteur's space passed to the committees.
Powerful cross-references
- Execution × Earmarks: half of the authorized budget doesn't become real spending.
- Rapporteur × Concentration: R$ 46.6 billion in rapporteur earmarks between 2020 and 2022, R$ 8.6 billion in the last year alone.
- Taxation × Inequality: companies are taxed three to five times more than workers via IRPJ vs. IRPF, but the system remains regressive overall.
- Carryovers × Cancellation: R$ 12-15 billion canceled per year — money promised and never delivered.
- Discretionary × Cuts: science executes 65% of its allocation, environment only 55%.
- Interest × Bolsa Família: R$ 700 billion in interest against R$ 35 billion in Bolsa Família.
- Taxation × Regressivity: 55% of revenue comes from indirect taxes.
- Secret budget × Opacity: R$ 40 billion with no recipient identification.
- Supreme Court × Committees: with no rapporteur earmarks since 2023, committee earmarks reached R$ 11.7 billion in 2024 and the total hit R$ 47.1 billion in 2025.
Explanatory hypotheses
The low execution rate can be explained by the hypothesis of the budget as a political bargaining chip: officials commit resources to show action to voters, without any real commitment to short-term execution. The concentration of decision-making power in few hands — general rapporteur, specific committees — reveals a captured Legislature, in which a reduced number of actors control a disproportionate share of budget allocation. The regressive tax structure overall reflects a captured State, in which capital influences the rules to reduce its effective burden. And the secret budget represents the most extreme form of that capture: public money moved with no accountability whatsoever.
Policy implications
Expanding proactive transparency about budget execution allows direct scrutiny by civil society and the press. Tying earmark value to actual execution — treating unpaid carryovers as a control indicator, not a discretionary reserve — tends to improve public-policy delivery. Advancing tax progressivity, with greater taxation on wealth and large fortunes, could correct part of the distortion that today makes the system weigh more on those who earn less. Fully abolishing any remnant of budget with no recipient identification would restore a minimum of democratic accountability. Reducing the cost of public debt would free up a significant portion of the R$ 700 billion currently allocated to interest for social policy. And reinvesting in science and the environment — areas among the first to be cut today — would help restore long-term state capacities that the country continues to neglect.