In the long arc of Brazilian democracy, May's polling numbers tell a story not merely of electoral preference but of institutional trust — a former president who once governed from the left now leads an incumbent whose confrontations with courts, voting systems, and judicial norms have quietly unsettled the voters who decide elections. Luiz Inácio Lula da Silva extended his lead over Jair Bolsonaro to 17 percentage points in one major survey, with a second-round projection of 54 to 34 percent, as moderate Brazilians appear to be rendering a quiet verdict on the costs of democratic erosion. The
Lula extends lead over Bolsonaro as Brazil moderates shift in new polls
Voters believe the president is wrong to confront the courts
So Lula gained ground in these May polls—17 points in one survey, 7 in the other. That's a meaningful shift, but I'm curious what actually changed. Did voters suddenly like Lula more, or did they like Bolsonaro less?
It looks like the latter. Nunes specifically points to three things Bolsonaro did that drove moderates away: attacking the voting system, fighting with the courts, and pardoning Silveira. The pardon seems to have been the crystallizing moment.
But wait—on the voting system, the poll shows distrust actually fell from 27 percent to 22 percent. So Bolsonaro's attacks on electronic voting didn't move opinion his way. That's worth noting.
Right. The attacks didn't work. And among moderates, 54 percent disapproved of the pardon while only 17 approved. That's a huge gap.
How solid is the runoff projection? Lula at 54 percent, Bolsonaro at 34 percent—that's a 20-point gap.
The Quaest poll has a 2-point margin of error, so that's a real gap. But it's a projection, not a measurement. We don't know if moderates will actually break that way in a runoff, or if Bolsonaro could consolidate his base differently.
True. But the first-round numbers are what we can measure now, and Lula is within striking distance of 50 percent. That's the real story—he's close to winning without a runoff.
And these are the first gains Lula's made all year?
Yes. The Genial/Quaest poll explicitly says this is his first rise in 2022. He'd been stuck or losing ground until May.
One thing to keep in mind: these are two different polls with different methodologies. Quaest interviewed 2,000 people in person over three days; PoderData called 3,000 by phone over two days. The Quaest numbers are more dramatic. We should be cautious about treating them as equally reliable without more context.
So the election is in October. That's five months away. Can these leads hold?
That's the question. But Nunes is saying that most committed voters have already decided. The race is really about moderates, and moderates have started moving away from Bolsonaro.
And we don't know what Bolsonaro will do next, or what events might shift the race. These polls are a snapshot from early May.
El Pulso
- Lula's polling lead has widened to 17 points in one survey and 7 in another, placing him within striking distance of an outright first-round victory at the 50 percent threshold.
- Bolsonaro's pardon of a jailed congressman — issued hours after sentencing — has become a flashpoint, with 45 percent of voters disapproving and moderates rejecting it at a rate of 54 to 17 percent.
- The president's sustained campaign against electronic voting machines has backfired: public distrust in the system has actually fallen from 27 percent to 22 percent since September.
- Moderate voters — the undecided middle neither camp has locked in — are drifting away from the incumbent, and analysts say this constituency will determine Brazil's political direction.
- A runoff scenario now projects a decisive Lula victory of 54 to 34 percent, suggesting the race's center of gravity has shifted well beyond the margin of Bolsonaro's base.
In the long arc of Brazilian democracy, May's polling numbers tell a story not merely of electoral preference but of institutional trust — a former president who once governed from the left now leads an incumbent whose confrontations with courts, voting systems, and judicial norms have quietly unsettled the voters who decide elections. Luiz Inácio Lula da Silva extended his lead over Jair Bolsonaro to 17 percentage points in one major survey, with a second-round projection of 54 to 34 percent, as moderate Brazilians appear to be rendering a quiet verdict on the costs of democratic erosion. The October election remains months away, but the center, so often the silent arbiter of political fate, has begun to move.
Two polls released in mid-May showed Lula da Silva extending his lead over President Jair Bolsonaro ahead of Brazil's October elections — the leftist former president's first measurable gains since the year began. The Genial/Quaest survey put Lula at 46 percent to Bolsonaro's 29, a 17-point margin up from 14 the previous month, while PoderData showed a tighter 7-point gap. Both carried a 2-point margin of error.
Poll director Felipe Nunes pointed to three presidential actions driving the shift: Bolsonaro's attacks on Brazil's electronic voting system, his escalating clashes with the Supreme Court and Electoral Tribunal, and his decision to pardon Congressman Daniel Silveira just hours after the court sentenced him for threatening democratic institutions. The pardon proved particularly damaging — 45 percent of voters disapproved, and among moderates the gap was stark: 54 percent opposed versus only 17 percent in favor.
Bolsonaro's voting-system narrative also appeared to be losing its grip. Public distrust in electronic machines had actually declined since September, from 27 to 22 percent — suggesting the president's repeated challenges had persuaded few and perhaps reassured many.
At 46 percent, Lula sits within the margin of error of the 50 percent threshold needed for a first-round win. Should the race go to a runoff, Quaest projects him winning 54 to 34 percent. Analysts note that committed supporters on both sides are largely fixed; the election will be decided by the undecided middle — and that middle, crystallized by the Silveira pardon, has begun to move.
Two new surveys released in mid-May showed Luiz Inácio Lula da Silva pulling further ahead of Jair Bolsonaro in Brazil's race toward October elections, marking the leftist former president's first polling gains since the year began. The Genial/Quaest poll put Lula at 46 percent support against Bolsonaro's 29 percent—a 17-point margin, up from 14 points the previous month. A second survey by PoderData found Lula leading by 7 points, a gain of 2 points over the prior two weeks. Both polls carried a margin of error of 2 percentage points.
The shift reflected a measurable erosion of Bolsonaro's standing among moderate voters, the constituency most likely to decide a close election. Felipe Nunes, director of the Quaest poll, attributed the movement to three specific presidential actions: Bolsonaro's sustained attacks on Brazil's electronic voting system, his escalating confrontations with the Supreme Court and the Supreme Electoral Tribunal, and his decision to pardon Congressman Daniel Silveira just hours after the court sentenced him to prison for threatening the judiciary and democratic institutions. The pardon proved particularly divisive—45 percent of voters surveyed disapproved of it, while only 30 percent approved. Among moderates specifically, the disapproval gap widened sharply: 54 percent opposed the pardon versus just 17 percent in favor.
Interestingly, Bolsonaro's core argument about the voting system appeared to be losing ground. When the Quaest poll asked Brazilians about their trust in electronic voting machines, only 22 percent expressed distrust—down from 27 percent in September. This decline suggested that the president's repeated questioning of the system's integrity had not persuaded the electorate; if anything, confidence had grown. Nunes observed that voters had reached a clear judgment: they believed the president was wrong to challenge the courts, wrong to undermine the voting apparatus, and wrong to shield an ally from judicial consequences.
The Quaest survey, which interviewed 2,000 respondents between May 5 and May 8, showed Lula approaching a threshold that would allow him to win outright in the first round. At 46 percent, he sat within the poll's margin of error of reaching 50 percent—the level needed to avoid a runoff. The PoderData poll, which conducted telephone interviews with 3,000 people from May 8 to May 10, showed a tighter first-round race: Lula at 42 percent, Bolsonaro at 35 percent, a difference of 7 points that fell within the margin of error and represented only minor movement from a fortnight earlier.
Both polls painted a starkly different picture if the race proceeded to a second-round runoff. The Quaest survey projected Lula winning decisively, 54 percent to 34 percent. This gap suggested that moderate voters—those not yet locked into either camp—held the key to the election's outcome. Nunes emphasized that most committed supporters of either candidate had already made their choice; the race would be decided by the undecided middle, and that middle had begun to move away from the incumbent. The pardon of Silveira appeared to have crystallized their doubts, transforming abstract concerns about institutional conflict into a concrete act of presidential power that a clear majority found troubling.
Citas Notables
Voters believe the president is wrong to confront the Supreme Court, to question the credibility of electronic voting machines and to pardon a congressman convicted by the court.— Felipe Nunes, Quaest poll director