Theme 08 of 43

Public Policy, Transfers & Social Protection

The New Bolsa Família pays an average of R$ 677 a month to 18.7 million families — and covers 163 families per thousand inhabitants in Maranhão against 24.5 in Santa Catarina.

Bolsa Família payments, audited and released by the Office of the Comptroller General (CGU), record every transfer made to every beneficiary family in every municipality in the country. The Brazilian Public Sector Accounting and Fiscal Information System (SICONFI), kept by the National Treasury, reveals how municipalities actually execute their budgets. Together, the two panels show a social-protection system with near-perfect territorial coverage. Active benefits from the INSS complete the picture with the BPC, which pays one minimum wage to low-income elderly people and people with disabilities.

New Bolsa Família: 18.7 million families, R$ 677 a month

Monthly payments released by the Office of the Comptroller General (CGU) show a program on a different scale from the 2021 Bolsa Família. In September 2026, 18.7 million families were paid, in every municipality in the country, with an average of R$ 677 — three and a half times the R$ 190 per payment of 2021. In 2025 the program transferred R$ 157.1 billion, five times the 2021 total. The number of families, however, is falling: there were 20.5 million in January 2025, 9% more than today.

New Bolsa Família — indicators (September 2026 unless noted)
IndicatorValue
Families paid in the month18.7 million
Total paid in the monthR$ 12.66 billion
Average per familyR$ 677
Families paid in January 202520.5 million
Total paid in 2025R$ 157.1 billion
Total paid in 2021 (former Bolsa Família)R$ 30.4 billion

The average per family rose from R$ 190 to R$ 677 — the program is no longer token money in household budgets, but it lost 1.8 million families in twenty months.

Where does Bolsa Família go?

The average value barely changes between states — from R$ 665 in Bahia to R$ 735 in Roraima, because the benefit is calculated per person and the North has larger families. What changes is coverage. Dividing families paid by IBGE's 2025 population estimate, Maranhão has 163 beneficiary families per thousand inhabitants; Santa Catarina, 24.5 — almost seven times fewer. All nine Northeastern states and five Northern ones are above 135 per thousand; no state in the South, Southeast or Center-West goes above 80.

New Bolsa Família by state (September 2026)
StateFamilies paidFamilies per thousand inhabitantsAverage value
Maranhão1,144,350163.1R$ 695
Piauí537,549158.8R$ 673
Bahia2,341,654157.5R$ 665
Amazonas601,406139.2R$ 724
Roraima74,679101.1R$ 735
São Paulo2,182,84447.4R$ 671
Paraná537,04445.2R$ 672
Santa Catarina200,35224.5R$ 676
Brazil18,695,57887.6R$ 677

In Maranhão there is one Bolsa Família family for every six inhabitants; in Santa Catarina, one for every forty — the program's map is the poverty map.

Parliamentary earmarks: from R$ 25 to R$ 47 billion

In 2022, committed parliamentary earmarks totaled R$ 25.5 billion, close to the Bolsa Família of the time. Since then earmarks have almost doubled: R$ 35.4 billion in 2023, R$ 44.8 billion in 2024 and R$ 47.1 billion in 2025, according to the Transparency Portal. Because the New Bolsa Família grew even more, the comparison flipped: in 2025 earmarks were equal to 30% of what was paid to families. More than half of the committed value has no individual legislator as author — it comes from the general rapporteur, committees and state caucuses.

Parliamentary earmarks by year (Transparency Portal)
YearEarmarksCommittedPaid in the year
20226,108R$ 25.5 bnR$ 17.0 bn
20236,110R$ 35.4 bnR$ 21.9 bn
20246,990R$ 44.8 bnR$ 31.4 bn
20256,310R$ 47.1 bnR$ 30.6 bn
2026 (through October)5,634R$ 40.1 bnR$ 25.9 bn

Parliamentary earmarks grew 85% between 2022 and 2025 — a R$ 47 billion budget decided by bargaining, beside a R$ 157 billion program decided by income criteria.

BPC: 6.4 million benefits, with disability driving the growth

The Continuous Cash Benefit (BPC) pays one minimum wage to low-income people aged 65 or older and to low-income people with disabilities. INSS data on active benefits show 6.42 million BPCs in January 2026, R$ 9.7 billion a month. Recent growth comes almost entirely from disability: 3.75 million benefits, 34% more than in January 2023, while the elderly BPC grew 13% over the same period.

Active BPC benefits (INSS), January 2023 and January 2026
BenefitJan/2023Jan/2026Change
Person with disability2.80 million3.75 million+34%
Elderly (65+)2.37 million2.67 million+13%
Total5.17 million6.42 million+24%
Average value—R$ 1,518one minimum wage

The disability BPC gained almost a million benefits in three years — it now pays more people than the elderly BPC.

Garantia-Safra: family farmers in the semi-arid region

The program that protects family farmers in the semi-arid region against drought losses reaches only a small fraction of who, in theory, should be covered. Only 15% of family farmers in the region receive the benefit — coverage too low for a program designed specifically for the country's area most vulnerable to drought.

Garantia-Safra: value, coverage, and reach
IndicatorValue
Value per familyR$ 1,200
Families served200 thousand
Coverage (semi-arid region)15%

Only 15% of family farmers in the semi-arid region receive Garantia-Safra — coverage far too low for the policy meant to protect them from drought.

Duration of transfers: how long poverty lasts

Comparing the average duration of different transfer programs, it's clear most function as a temporary bridge, not a permanent solution. Bolsa Família lasts, on average, three years per family; unemployment insurance, a few months; emergency aid, a one-off episode. Most families that pass through these programs later return to poverty.

Average duration of transfer programs
ProgramAverage duration
Bolsa Família3 years
Unemployment insurance3-5 months
Emergency aidOne-off

Social transfers function as a bridge, not a solution — most families return to poverty after the program ends.

Who manages public policy: women in education, men in the mayor's office

IBGE's municipal survey records the profile of whoever heads each policy area across all 5,570 municipalities — and reveals that the sexual division of labour reproduces itself inside the public machine with almost caricatural precision. Women are 69.7% of education managers and 56.5% of health managers, the two portfolios historically associated with care work, but only 12.1% of those holding the mayor's office, the position with budget and political power. This is not a glass ceiling — it is a glass wall: women are not absent from public management, they are confined to a specific set of portfolios. The racial breakdown reveals a parallel and equally sharp pattern: Black people are 2.3% of mayors but 6.5% of culture managers, nearly three times more. The smaller the portfolio's budget and power, the greater the presence of Black managers.

Profile of those heading each municipal policy area (IBGE municipal survey)
Area% women% Black or mixed-race% Black
Mayor12.1%30.6%2.3%
Sport32.4%44.0%6.3%
Culture47.7%41.3%6.5%
Health56.5%37.3%3.5%
Education69.7%40.3%5.1%

Women are 7 in every 10 municipal education managers and barely 1 in 10 mayors — Brazilian public policy is executed by women and decided by men.

Powerful cross-references

Explanatory hypotheses

Until 2022 the Bolsa Família value could be explained by the minimalism hypothesis: small transfers kept families out of the most extreme destitution without the fiscal cost of a transformative transfer — the jump to R$ 677 per family shows that limit was a political choice, not a fiscal one. Clientelism theory helps explain parliamentary earmarks as an instrument of political exchange — resources allocated by bargaining, not by technical need criteria. And the growth of the disability BPC, faster than the elderly one, suggests that the most contested entry point to social protection is now the medical assessment.

Policy implications

With the average Bolsa Família value already above R$ 600, the challenge shifts to coverage: tracking why 1.8 million families left the program since January 2025 tells whether they left through income or through exclusion from the registry. Reducing the volume of discretionary parliamentary earmarks could free up resources for technical anti-poverty policies. Tying earmarks to objective social-vulnerability indicators could correct part of today's distortions. Simplifying CadÚnico, the unified social-registry system, with process digitization, could reduce wrongful denials of benefits like BPC. And long-term conditional transfer programs — not just emergency bridges — have the potential to break the cycle of poverty that today repeats across generations of the same family.