Lula hit his poll number almost to the decimal. Datafolha’s final survey gave him 45% of valid votes, and on Sunday he took 45.16%. The error was entirely on the other side of the ledger. The final polls we tracked had Flávio Bolsonaro between 42% and 45%. He finished at 47.03%: 56,104,422 votes to Lula’s 53,878,385, a gap of 2.2 million with 99.99% of sections counted.
If your Brazil file for the October 25 runoff still says “Lula favored, tight at the margin,” it describes an election that ended on Sunday night.
What used to work
Through September, the Brazil model was tidy. Lula led nearly every first-round poll by one to six points. The runoff looked like a close contest the incumbent’s machine was favored to win. The operator’s job was to hold anything election-sensitive until a winner emerged, and the whole file hung on one question: who wins on October 25?
That framing is now too small, and it was too small before Sunday too. It treated the presidency as the only thing on the ballot.
What changed on October 4
Three things moved at once.
The order moved. The first-round leader has won every Brazilian presidential runoff since the end of the military dictatorship, according to AFP, and that leader is now Flávio. Analysts quoted by Reuters read part of his surge as protest votes against Lula that pollsters expected to wait for round two.
The Congress is settled. Brazil’s legislature was elected the same night, and AFP describes landslide wins for the right. The Partido Liberal (PL) is projected to hold around 29 Senate seats in 2027, against roughly nine for Lula’s Partido dos Trabalhadores (PT, Workers’ Party). The Atlantic Council’s Ricardo Sennes and Valentina Sader make the point that matters for planning: whoever wins will face a fragmented Congress with no clean majority, and will nominate at least three justices to the Supremo Tribunal Federal (STF, Supreme Federal Court), each needing Senate confirmation.
The pool shrank. Four minor candidates took 7.58% of valid votes between them: Augusto Cury 2.89%, Renan Santos 2.24%, Ronaldo Caiado 2.18%, Romeu Zema 0.27%. By our arithmetic, Lula needs about 4.8 points of that pool to reach 50%, roughly two-thirds of it. Flávio needs about 3, two-fifths. Abstention reached 33.5 million voters, 21.08%, the highest share since 1998. Zema said before the vote he would back any right-wing finalist against Lula, while Santos spent the campaign attacking Flávio. The pool is not a bloc, but the math is not symmetric.
Lula’s own read, after the count, was candid: he admitted he had been convinced he would win in the first round, and said the game was about to start.
Where our September 14 read stands
On September 14 we called the runoff a toss-up: 47% government retention, low confidence, with the Cury bloc as the deciding variable and a 60% break toward Flávio as the tipping point. Sunday dated that call. Cury took 2.89%, so he is not a kingmaker, just one slice of a 7.58% pool. And with Flávio already ahead, he needs only about 39% of that pool, not a supermajority of one candidate’s voters. The 47% figure no longer holds. The next honest update belongs to the runoff polls.
What works now
Sort your Brazil decisions into three bins.
Settled. Concessions, permits and sector rules that depend on the legislature, not the Planalto. The Congress result is in, and BNamericas notes the concessions agenda has carried on across successive governments. Proceed with what doesn’t hinge on who sits in the presidential palace.
Contingent on October 25, and dated. Start with the U.S. channel. The 25% Section 301 tariff on most Brazilian imports took effect July 22, and Lula blamed Flávio’s Washington visit the week before. Exemptions cover about $11 billion in trade, so exposure varies by product. A Flávio win likely opens a negotiation, and a Lula win likely keeps the confrontation. Anything that prices in a tariff assumption, from sourcing shifts to tariff-exposed capex, gets a hold with an October 26 trigger and two finished scenario papers behind it.
Fiscal, with a tripwire. Decree 13.116 takes R$0.63 a litre off federal gasoline taxes and expires October 9, mid-campaign. Renewal tells you the framework bends to the runoff. Lapse puts a visible price rise into the final stretch. On Flávio’s side, Daniella Marques, formerly CEO of the state-owned lender Caixa Econômica Federal, leads an economic team promising a fiscal-framework overhaul and limits on subsidized credit. As of mid-August there was no named finance minister. That name, when it comes, is the signal.
The case for caution
Flávio is the leader, and history favors leaders. But the gap is 1.87 points, Lula’s coalition has the state machinery, and 33.5 million people who stayed home are still up for grabs. Jair Bolsonaro outperformed his polls by almost 10 points in the 2022 first round, according to Chatham House’s Christopher Sabatini, and still lost the runoff.
So don’t swap one favorite for another. A model that mirrors the old error in the other direction repeats it. Write both papers with equal effort, and give each one a day-one calendar.
What to watch
Three dates sort the next three weeks. October 9: whether Decree 13.116 is renewed. The week after: where Caiado, Zema, Cury and Santos land, and what the abstention count looks like in the polls. October 25: the runoff itself, with inauguration set for January 5, 2027. After that comes the February leadership vote in both chambers, which is the real test of whether either president can govern.
Brazil’s Congress already voted. On October 25, the country picks who has to live with it.
Further reading: