Brazil’s Gangs Move 2.4 Times What the State Spends on Security
Public Security
Key Facts
—The audit. Brazil’s federal accounts court, the TCU, reported on the National Public Security Plan on 8 July.
—The federal share. Union transfers came to 1.19% of state security spending in 2023, 1.17% in 2024 and 1.31% in 2025.
—The asymmetry. Organised crime moves about R$348bn ($67.6bn) a year, some 2.4 times the R$142.81bn ($27.7bn) all three tiers of government spent on security in 2024.
—The unused lever. Between 2023 and 2025 no state ever had a transfer withheld for ignoring the plan.
—The targets. Two of the plan’s thirteen original goals had already been met before it took effect in 2021.
—The orders. The court gave the Justice Ministry 45 days to fix the targets and 60 to produce the evaluation it has never written.
The states pay for Brazil public security, almost entirely. Federal transfers amounted to just over one percent of what state governments spent on security in each of the last three years, a federal audit published this week shows.
That leaves the states self-funding close to ninety-nine percent of the bill. They also run the uniformed police who patrol the streets and conduct most criminal investigations.
Every foreign account of Brazil public security asks whether the army should be deployed or whether Washington should help. The audit points at a duller question, which is who actually pays for the police.
What the Brazil public security audit actually found
The report comes from the Tribunal de Contas da União, the federal accounts court that scrutinises how Brasília spends money. Its subject is the National Public Security Plan, a ten-year framework running to 2030.
The audit is blunt about the consequence. The gulf between federal and state money, it says, sharply limits the Union’s ability to persuade states to follow the plan at all.
Then comes the number that reframes everything. Organised crime in Brazil moves roughly three hundred and forty-eight billion reais a year, close to sixty-eight billion dollars.
Set that against what the federal government, the states and the municipalities together spent on security in 2024, about one hundred and forty-three billion reais or some twenty-eight billion dollars. The criminal economy is nearly two and a half times larger than the public one arrayed against it.
A plan whose goals were met before it began
The auditors then examined the plan’s thirteen original targets. Two of them had already been achieved before the plan came into force.
The goal for vehicle theft, set for 2030, had been cleared in 2020. The goal for robbery-homicide, the crime Brazilians call latrocínio, was set at exactly the rate the country had already recorded that year.
A proposed revision was supposed to correct this. The auditors found it repeats the error, pitching a robbery-homicide target for 2030 that national data have shown was already surpassed in 2023.
The revision also expands the target list from thirteen to twenty-two. Eleven of the new twenty-two are written as percentage changes without saying from what starting point, which makes them impossible to measure.
Nor has the plan ever been evaluated. The law requires an annual assessment, and none has been produced since 2021, while four of the nine documents its governance committee was meant to publish were never written and five stopped appearing after 2023.
There is a reason for the silence. The technical commission responsible for producing those assessments had its members formally appointed only in October last year, and it held its first meeting in January.
The plan had by then been running for four years. Nobody, in that time, had been designated to measure whether it worked.
Why this matters more than the sovereignty argument
The states were not consulted when the plan was designed. Its intervention logic was built in workshops attended by forty-two representatives, every one of them from a federal body.
In April this year the court convened the national councils of state security secretaries, military-police commanders and forensic directors. Their verdict on whether the states had been involved was unanimous, and it was no.
The plan’s one enforcement mechanism is the power to withhold federal money from a state that ignores it. Between 2023 and 2025 that power was never once used, and the auditors note it would barely bite anyway, because the money involved is so small.
Read that alongside the week’s other news and the picture sharpens. The Senate has just voted to route a slice of Brazil’s betting tax into a fund for the Federal Police, and Washington has offered to help Brazil confront its criminal gangs.
Both gestures point at the federal forces. The audit says the forces that do the policing are state forces, and Brasília pays for barely one percent of them.
For an investor pricing Brazilian country risk, that is the binding constraint. It is not sovereignty, not doctrine, and not who deploys which soldier, but a federation in which the government that writes the national plan funds almost none of it.
Who pays for Brazil public security?
The states, overwhelmingly. Federal transfers covered just over one percent of state security spending in each year from 2023 to 2025.
How big is Brazil’s criminal economy?
The audit cites annual financial flows of about sixty-eight billion dollars, nearly two and a half times what all three tiers of government spent on security in 2024.
What must the government do now?
The court gave the Justice Ministry forty-five days to revise the flawed targets and sixty days to deliver the annual evaluation the law has required since 2021.
In depth
Read More from The Rio Times