Lula leads Bolsonaro by 39 points among Auxílio Brasil recipients

Cash transfers alone cannot overcome deeper currents in voter sentiment
Bolsonaro's flagship welfare program failed to build electoral support among its intended beneficiaries.

Bolsonaro's flagship Auxílio Brasil program failed to build electoral support, with beneficiaries favoring Lula by a 39-point margin in May 2022 polling. The 21-point national lead for Lula (48%) over Bolsonaro (27%) reflects broader voter preference, with income-based divides showing Lula dominates lower-income voters.

  • 59% of Auxílio Brasil recipients said they would vote for Lula versus 20% for Bolsonaro
  • Lula led nationally by 21 points (48% to 27%) in the May 2022 Datafolha poll
  • Among voters earning below two minimum wages, Lula won 56% to Bolsonaro's 20%
  • Datafolha surveyed 2,556 voters across 181 municipalities May 25-26, 2022

A Datafolha poll shows 59% of Auxílio Brasil beneficiaries plan to vote for Lula versus 20% for Bolsonaro, undermining the social program's intended electoral strategy for the incumbent president.

In May 2022, with Brazil's presidential election five months away, a Datafolha poll delivered unwelcome news to Jair Bolsonaro's campaign. Among the 18 million families receiving Auxílio Brasil—the social welfare program Bolsonaro had created to replace Bolsa Família and, strategically, to build electoral support—59 percent said they would vote for Luiz Inácio Lula da Silva. Only 20 percent backed the sitting president. The 39-point gap was a stark repudiation of what had been designed as a political asset.

The Auxílio Brasil program represented a deliberate pivot in Bolsonaro's governance. It succeeded the long-established Bolsa Família, which had been created in 2003 and become synonymous with the left. By replacing it with his own initiative, Bolsonaro hoped to claim credit for social spending and lock in electoral loyalty among the poorest voters. Inside the government, officials had told reporters in April that they expected the program's launch—along with a companion gas voucher—to produce a measurable bump in the president's polling numbers. The Datafolha results, conducted May 25-26 among 2,556 voters across 181 municipalities, suggested that calculation had failed entirely.

The broader picture was no more encouraging for Bolsonaro. Nationally, Lula commanded 48 percent of voting intentions, a 21-point lead over the incumbent's 27 percent. Ciro Gomes, the third-place candidate, drew only 7 percent. The remaining field—including André Janones, Simone Tebet, and a handful of other contenders—collectively accounted for less than the margin separating the top two. When voters were asked unprompted, without candidate names presented to them, Lula's advantage widened further: 38 percent named him as their choice, compared to 22 percent for Bolsonaro. This represented a significant shift from March, when the unprompted figures had been 30 percent and 23 percent respectively.

Income proved a decisive dividing line. Among voters with monthly household earnings below two minimum wages, Lula won decisively with 56 percent support against Bolsonaro's 20 percent. In the middle tier—families earning five to ten minimum wages—the two candidates tied at 37 percent each. But among higher-income households, those earning more than ten minimum wages monthly, Bolsonaro reversed the dynamic, claiming 42 percent to Lula's 31 percent. The pattern suggested that Bolsonaro's base had consolidated among the wealthy while his attempt to build support among the poor through direct cash transfers had largely backfired.

The poll's methodology was rigorous. It was registered with Brazil's Superior Electoral Court under number BR-05166/2022 and carried a margin of error of two percentage points. Datafolha noted that the survey was not directly comparable to its March measurement because the field had shifted—some candidates had dropped out, others had entered—making apples-to-apples comparison difficult. But the institute also noted that Lula's 38 percent in the unprompted measure represented the strongest performance Datafolha had recorded since beginning its tracking in May 2021.

Indecision had also declined. In the unprompted format, 29 percent of voters remained undecided in late May, down from 32 percent two months earlier. The electorate appeared to be crystallizing, and the direction of that movement was unmistakable. Bolsonaro's signature social program, designed to win hearts and votes among the poorest Brazilians, had instead become a reminder of his unpopularity with that very constituency. With five months until the election, the data suggested that no amount of cash transfers could overcome deeper currents in voter sentiment.

The government had expected a polling bounce after the program's launch, but instead Lula's lead grew
— Government sources cited in reporting
La Conversación del Hearth Otra perspectiva de la historia
Inventor

Why did Bolsonaro think replacing Bolsa Família with Auxílio Brasil would help him politically?

Model

It was a straightforward calculation: take credit for a popular program by putting his name on it, rebrand it, and hope the beneficiaries would remember him at the ballot box. The government even timed the rollout to coincide with the campaign season.

Inventor

But it didn't work. Why would people who receive money from a president vote against him?

Model

Because cash transfers, however welcome, don't erase everything else. Bolsonaro's approval was already underwater. The program was a gesture, but it couldn't overcome his broader unpopularity or the fact that Lula had created Bolsa Família in the first place—people remembered that.

Inventor

The poll shows Lula winning among the poor but Bolsonaro winning among the rich. Isn't that a normal political split?

Model

It is, but the magnitude matters. Bolsonaro was supposed to be the outsider, the anti-establishment figure. Instead, he'd become the candidate of the wealthy. That's a fundamental realignment.

Inventor

What does this tell us about the power of social spending in elections?

Model

That it's not magic. Money helps, but it doesn't buy loyalty if voters don't trust you or if they see you as fundamentally opposed to their interests. Bolsonaro was trying to have it both ways—cut spending elsewhere while buying votes with this one program—and voters saw through it.

Inventor

Did anything surprise the government about these numbers?

Model

Almost certainly. They'd been telling reporters in April that they expected a polling bounce. Instead, they got the opposite. The program launched, money flowed, and Lula's lead actually grew.

Contáctanos FAQ