How are Lula da Silva and Flávio Bolsonaro heading into the second round? The latest poll puts them at 47% and 46%, two days before the election

A second-round election, or runoff, is a later vote between the leading candidates when the first round does not produce a clear winner under the election rules. It narrows the contest so voters choose directly between the finalists. This matters because a first-round lead does not guarantee the presidency.

Brazil’s first round produced two finalists: Flávio Bolsonaro received 47 percent and Lula received 45 percent, according to the article. Their close result led to another vote. The second round therefore asks voters to select one of these two candidates rather than choosing among a larger field.

The runoff was scheduled for two days after the reported survey, while both candidates prepared for a final televised debate. Datafolha placed Flávio at 47 percent and Lula at 46 percent, with both reaching 50 percent among valid votes. The result remained highly competitive, so the second-round choice was decisive.

The poll percentages measure respondents’ stated intention to vote for each presidential candidate. They are not final election results. Datafolha reported 47 percent for Flávio Bolsonaro and 46 percent for Lula da Silva, based on interviews conducted before the second round.

The difference is only one percentage point. The survey’s margin of error is two points, meaning each estimate can plausibly vary around its reported figure because the poll questioned a sample rather than every voter. That makes the candidates’ positions a technical tie, as the article explains.

The numbers show a very close race, not a guaranteed Flávio victory. Depending on sampling variation, the actual level of support could be higher or lower for either candidate. Among valid votes, the poll gives both candidates 50 percent, reinforcing the uncertainty before the final debate and election.

Datafolha conducted 2,520 face-to-face interviews for the survey. The poll was commissioned by TV Globo and Folha de S. Paulo and registered with Brazil’s Electoral Court under code BR-09709/2026. These details describe the survey’s sample and formal registration.

A 95 percent confidence level is a statistical measure of how often a sampling method would capture the true population result within its stated uncertainty across repeated surveys. It does not mean there is a 95 percent chance that one exact percentage is correct. The article also reports a two-point margin of error.

That combination makes the poll informative but not definitive. It can identify broad patterns and a likely close contest, yet small differences should be treated cautiously. Here, the one-point separation between Flávio and Lula falls inside the two-point margin, producing a technical tie rather than a conclusive lead.

Valid votes are the votes counted for the candidates after blank votes, null votes, and undecided responses are excluded from the calculation. This measure focuses only on respondents who express a valid choice between the two finalists. It can therefore differ from percentages calculated across all respondents.

In this survey, the valid-vote figures are evenly divided: Lula da Silva has 50 percent and Flávio Bolsonaro also has 50 percent. The equality does not mean every respondent chose a candidate. It means the candidates have the same share after the excluded categories are removed.

This distinction matters because the broader figures are 47 percent for Flávio and 46 percent for Lula. Those percentages include the survey’s other responses, while the valid-vote calculation redistributes the comparison among the two candidates. The result reinforces the article’s description of a technically tied contest.

Lula’s support is strongest among voters earning up to two minimum wages, where he leads 54 percent to 39 percent. He also leads in the Northeast, 65 percent to 30 percent, among Black voters, 57 percent to 37 percent, among people with primary education, 53 percent to 40 percent, and among Catholics, 55 percent to 39 percent.

Flávio Bolsonaro has his clearest advantages among evangelicals, 69 percent to 26 percent, and in the South, 58 percent to 33 percent. These figures show that the candidates draw support from different parts of Brazil’s electorate. Religion, region, income, race, and education all correlate with different voting preferences in this poll.

The divide helps explain why the national race is so close. Lula’s Northeast and lower-income advantage offsets Flávio’s strength in the South and among evangelicals. Neither candidate dominates every group, making turnout and final decisions within these blocs especially important.

The final televised debate gives Lula da Silva and Flávio Bolsonaro one last major opportunity to present their arguments before voting. The article says they are using it to try to convince undecided voters. In a contest this close, even a small shift could matter.

The poll places Flávio at 47 percent and Lula at 46 percent, but the two-point margin of error makes that difference inconclusive. Among valid votes, both stand at 50 percent. These figures show why debate performance, campaign messages, and voters’ final decisions could affect the result.

The debate cannot guarantee a change in voting intentions. Its effect depends on whether viewers are persuaded and whether those intentions become actual votes. The article does not quantify the undecided group, so its precise impact cannot be calculated. Still, the narrow polling gap makes the final stage politically significant.

Brazil’s presidential election system generally uses two rounds when no candidate secures a majority of valid votes in the first round. The leading candidates then face each other in a runoff, and the candidate with more valid votes in that final contest wins. This explanation uses established background beyond the article’s specific description.

The article gives a clear example. Flávio Bolsonaro led the first round with 47 percent, while Lula received 45 percent. Despite that advantage, Flávio still had to compete again against Lula. In the reported second-round poll, the candidates were virtually even, showing how the later contest can differ from the first.

A first-round leader can lose because supporters of eliminated candidates may choose the other finalist, abstain, or change their preference. Regional and social differences can also matter. Here, the second-round poll’s 47-to-46 split and 50-50 valid-vote result show why the first-round lead was not decisive.