Brazil’s Congress has approved a plan to drastically cut former president Jair Bolsonaro’s 27-year prison sentence, the punishment handed down in 2025 after the Supreme Federal Tribunal convicted him of plotting a coup to remain in office following his 2022 election loss. The legislation, drafted with the support of the centrão — Brasília’s transactional centre-right bloc — does not nullify the conviction but lowers the maximum applicable term, alters parole calculations and is widely understood to clear a path for Bolsonaro to leave prison years earlier than the original sentence allowed. President Lula da Silva is reportedly weighing a veto. The vote was the most consequential test of the post-Lava Jato consensus that Brazil’s institutions had finally learned how to jail their own former presidents — and is being read across Latin America as a signal of how durable that consensus actually is.
The received wisdom
The mainstream framing, both inside Brazil and in the international press that covers it, has hardened quickly into something close to alarm. The Supreme Federal Tribunal’s conviction of Bolsonaro was, on the standard reading, the moment Brazilian democracy proved itself capable of disciplining an ex-president who, by his own admission, had urged his military aides to consider extra-constitutional steps in late 2022 and early 2023. To take that conviction and, two years later, water down its sentence — at the prompting of legislators many of whom were themselves under investigation for the events of 8 January — looks to liberal commentators like the textbook playbook of democratic backsliding: institutions hold the line, the political class then quietly reroutes around them. Lula’s allies in the PT and the international centre-left have framed the bill as an “amnesty in installments.” Brazilian editorialists have invoked the precedent of Argentina’s repeated military amnesties and the post-Pinochet Chilean compromises. The argument has weight; it is not paranoid.
A different read
But the picture is more complicated, and conservatives — especially conservatives sceptical of the way technocratic anti-corruption drives have been weaponised across the hemisphere — have reasons to take the more uncomfortable view seriously. Begin with the conviction itself. Bolsonaro’s 27-year sentence was the longest handed down to a Brazilian ex-president in modern memory, longer than those imposed on military officers convicted in connection with the 1964–85 dictatorship. It was issued by a Supreme Federal Tribunal whose conduct over the past five years — broad gag orders against journalists, expansive use of inquiry powers led by Justice Alexandre de Moraes, the prosecution of social-media platforms — has drawn criticism even from observers who have no sympathy at all for Bolsonaro himself. The pattern of punishing your defeated opponent with a sentence calibrated to outlive him is one Latin America has seen many times, and conservative commentators were not wrong, in the early 2010s, to worry that Brazil’s anti-corruption institutions were beginning to substitute prosecutorial moral certainty for the slower work of political contestation. The centrão’s bill, ugly as its motives are, also speaks to that real grievance.
The historical parallel worth holding in mind is not Pinochet but Italy. The Mani Pulite investigations of the early 1990s wiped out an entire generation of Christian Democrat and Socialist leaders, and they did so on findings that, in many cases, were factually correct. The longer-term effect, however, was not a cleaner Italian politics but a vacuum that Silvio Berlusconi proceeded to fill — and a population that grew systematically more cynical about the impartiality of magistrates the more the magistrates seemed to make political weather. Peggy Noonan once observed that the most dangerous thing for the rule of law is not the politician who breaks it but the judge whose sentence the public concludes was political. Brazilian voters, on every available poll, are roughly evenly split on whether Bolsonaro’s prosecution was justice, persecution, or both. That split is itself a fact, and it is the fact that the centrão’s bill is exploiting. The honest right-of-centre answer is not to deny the legitimacy of the conviction; it is to insist that the legitimacy of any sentence is partly a function of the public’s belief that the same court would have acted the same way against the other side. Lula, who himself spent 580 days in jail on charges later annulled, knows the costs of that perception better than anyone. His eventual decision on the veto will be read not just on its merits but as a signal of whether his second presidency intends to govern as a partisan actor or as a constitutional one.
The right-of-centre warning is therefore double. There is no defending Bolsonaro’s conduct in the days around 8 January, and the centrão’s bill is plainly self-interested. But the broader Latin American pattern in which presidents punish their predecessors and are then punished by their successors is corrosive of exactly the institutional trust that liberal democracy depends on. Brazil has now done both halves of that pattern within a single decade. What it does next — whether the bill is vetoed, whether the veto is sustained, whether Bolsonaro accepts or contests early release, whether Lula’s party looks for a similar accommodation when its own turn at the bar of justice comes — will determine whether the country looks more like post-Mani Pulite Italy or post-junta Spain.
What to watch
First, whether Lula vetoes, partially vetoes, or signs the bill: a partial veto is the most likely and the most revealing. Second, the response of the Supreme Federal Tribunal, which retains the power to review the constitutionality of any sentence-reduction scheme. Third, whether the recurring narrative of US visa cancellations against Bolsonarista journalists in Latin America becomes a Trump-administration lever in the dispute, dragging the question into the larger US–Brazil quarrel over tariffs and information policy. Fourth, the price action in Brazilian sovereign bonds: investors have a longer memory than editorialists, and they are watching whether Brazil is becoming more or less predictable.
— J