Brazil's Big Bad Biden Bet
Lula makes his biggest political calculation yet
Back in January 2024, Joe Biden posted a video saying “I’m still the only person to ever beat Donald Trump—and I’m looking forward to doing it again.” Fateful words from a forgettable president. Few remember, but this position was gospel among those who insisted the former president was “far beyond cogent”.
To be fair, Biden never got the chance to prove himself right. And if you look at polls from the time—as well the election result—his claim wasn’t entirely off.
Fast forward to the Brazil-averse in 2025, President Lula, just under a fortnight ago, sent a not-so-subtle message to his 2022 election opponent, Jair Bolsonaro: “if it’s against me [he’ll] lose again”. Sound familiar? Like Biden, Lula is the only one to have beaten his outsider opponent—and has, therefore, decided he’s the only one who can do it again.
If Lula runs in 2026, he’ll be 81 at the height of the campaign—the same age as Biden when he imploded on the debate stage. Unlike Biden, though, Lula is showing less signs of slowing down. After undergoing brain surgery late last year, the president was quick to return to work as a show of strength. Yet, while age is worth mentioning, it's a clickbait aside to the thesis of this piece.
Lula’s big bad Biden bet isn’t on being Biden’s age, but believing the more extreme opponent will be the easiest to beat.
Back in the 2024 Republican primaries, there was a feeling among Democrat strategists that a Trump nomination would be favorable to Biden’s re-election cause. Even more moderate Republicans were convinced that a Ron DeSantis candidacy would force the Democrats to swap Biden out for a younger opponent.
The fallout from Trump’s final days in office gave many pundits the view that the Donald was a spent force for the average American voter. The base might have been alive and kicking, but the normies would never go back. They were wrong.
Lula’s return to openly criticizing the ineligible Bolsonaro suggests he’s of a similar mindset. Recent polling from Atlas shows he may be onto something. A runoff against Bolsonaro shows him with a comfortable lead, while against São Paulo governor, Tarcísio de Freitas, it’s a technical tie.
It’s probably worth digressing here and unpacking these polls. At first glance, the Workers’ Party (PT) would probably be popping champagne corks. Lula beats every single opponent—even regular Fox News contributor, Eduardo Bolsonaro. Winning bigly, right? Well, not quite.
In none of these scenarios—even against memelord Pablo Marçal—does Lula cross the 50% mark in the second round. This is a direct reflection of, what I would argue, is the most important polling number: a president’s disapproval rating. Lula’s disapproval rating, according to Atlas, sits at 51.4%—an almost exact mirror of his vote share against Pablo Marçal.
Before that fateful debate night, the Democrats were adamant that despite the Biden Administration’s poor approval ratings, the presence of Trump on the ticket would see them over the line. They were wrong, again.
While the PT base seems pleased that Bolsonaro remains ineligible for 2026, my guess is that the vibe is shifting in Lula’s campaign war room. The fact that Lula is even publicly mentioning Bolsonaro is an indication of as much. If you’re so convinced, and happy, that Bolsonaro won’t be running in ‘26, why even evoke his name?
In The Big Short, Selena Gomez, and that other guy, break down the ‘Hot Hand Fallacy’, which in the other guy’s words is when “people think whatever’s happening now is gonna continue to happen into the future.”
For PT, that means something like this: ‘Better to have the crazy guy running again in ‘26 so we can build a big beautiful coalition like we did in ‘22’. The people loved that then, and they’ll definitely love it now.’ The problem with this assumption is that, like Selena’s synthetic CDOs, PT believe they’re working with a 4D-chess mathematical formula that will deliver them victory—when, in reality, it’s just vibes, man.
Since Brexit, Trump, and Bolsonaro, the Centrist-Dad’s Army has taken a greater interest in the “dark arts” of political polling. This was a good start, but it’s led to some pretty misguided conclusions. The consensus among the army’s foot soldiers is that the right is much better at using polling to run hyper-targeted, dark-magic campaigns. It’s a greatly exaggerated interpretation.
What actually happened in Brexit and Trump 1.0 was that people like Dominic Cummings picked up on a 2012 US election study, where researchers analyzed a large poll on run on Xbox using Bayesian statistics. Long story short, campaigners identified pitfalls in conventional polling techniques; for example, not weighting for education among younger white male voters.
The math looks kind of complicated, but ultimately, it’s picking up on vibes. Back in 2022, a week before the second round of the Brazilian election, I ran a much simpler, and wonkier, analysis to predict the outcome.
My outcome: 50.35% (Lula); 49.65% (Bolsonaro).
The actual outcome: 50.9% (Lula); 49.1% (Bolsonaro).
Not too shabby. But it doesn’t make some mathematical wizard of the political darks arts.
What I actually understood was the vibe. Lula’s 2022 win wasn’t about coalition building and bringing Alckmin into the tent; it was about Bolsonaro’s disapproval rating. As much as I tweaked the predictors for Ciro’s first-round vote distribution, it was honestly meaningless in the context of a sitting president with an >50% disapproval rating.
The last days of Bolsonaro’s government were chaos; but so were Trump’s in 2020 and PT’s in 2016. The point? Even bad governments are re-electable with the passage of time. Lula has no direct control over Bolsonaro’s eligibility in 2026, but wishful thinking can be a dangerous pastime. If Trump 2.0’s election has taught us anything, it’s that electorates will vote you out if the price of eggs is too high.
One of the main takeaways people have from The Big Short is that the smartest guy in the room always bets on the worst possible outcome. Everyone walks away from the movie wanting to be Michael Burry (Christian Bale) or Mark Baum (Steve Carrell). I call this ‘The Big Short Trap’. I’d break it down now, but I’d rather wait until Podgecaste is successful enough to get a Hollywood actor to do an explainer video for you.
My takeaway, however, is a little different: know when to cut your losses, and you’d much rather be Jared Vennett (Ryan Gosling) than Bury or Baum.
If Biden had been more Vennett than Baum, he may not have loyal Democrats writing op-eds calling for his legacy to be thrown under the bus. Yet, here we are, and his big bad bet serves as a lesson for the history books. Although there are politicians in the more near present who could heed the warnings.
To paraphrase Vennett’s wise words, if someone’s offering you fire insurance on a burning house, you take it.