To see who is ahead in presidential election polls, the first-round vote and runoff scenarios must be read separately.
Even with the same two candidates, polls presenting the full field and head-to-head polls show entirely different margins.
The key point in these figures is that Lula’s first-round lead and a runoff too close to call within the margin of error appeared together.
3-Line Summary
1. Presidential election polls must be viewed separately for the first round and runoff
2. Lula led the first round in all five polls
3. In the runoff, whether the result falls within the margin of error is key
The Common First-Round Order Across Five Polls
In the five polls compiled by VEJA this week, Luiz Inácio Lula da Silva (president affiliated with PT) was presented as the first-round leader in every poll. His rival, Flávio Bolsonaro (senator affiliated with PL), ranked second in every poll. The fact that the two candidates appeared in the same order is separate from the fact that the figures in each poll were the same.
The first round presents the names of multiple candidates together. Under this condition, the gap between Lula and Flávio was 5–9.7% points. Datafolha showed 38% versus 33%, while BTG/Nexus showed 39% versus 33%. The gaps in Real Time Big Data and Quaest were 8% points each, and AtlasIntel presented 43.4% versus 33.7%.
These numbers do not by themselves represent the final result. Each institution used different polling methods and samples, and the article compares the figures on that basis. Still, the fact that the candidate order in the first round was the same across all five polls is the clearest common denominator in this coverage.
The third position went to Augusto Cury (candidate affiliated with Avante). According to VEJA’s report, he ranked third in all five polls, registering 7.8–11%. At BTG/Nexus, his figure rose from 2% to 11% in one week; at Datafolha, it changed from 2% on August 21 to 8% in the new poll. However, the figures from different institutions cannot be strung together and read as one definitive approval rating.
In Head-to-Head Runoff Polls, 1–2% Points Made the Difference
The picture changes in runoff scenarios with only two candidates remaining. BTG/Nexus recorded Lula at 46% and Flávio at 45%. Real Time Big Data showed the two candidates tied at 44% each. Quaest showed 42% versus 41%, while Datafolha showed 46% versus 44%.
VEJA classified these four polls as technical ties, meaning differences within the sample’s margin of error. The 2% point difference in Datafolha was also described as a tie in that article. Therefore, it is difficult to say that Lula’s lead in the first round would continue at the same margin in the runoff based only on the observation that he was ahead in the first round.
The exception is AtlasIntel. In its runoff scenario, the institution presented Lula at 47.1% and Flávio at 42.6%, and stated that the margin of error was 1% point. VEJA distinguished this result as a Lula lead outside the margin of error. In other words, four of the five polls showed a close race, while one showed a gap outside the margin of error.
Looking only at the numbers, the first-round and runoff questions are different. First-round figures reflect the choice when the full field of candidates is present, while runoff figures reflect the choice between two candidates. When reading these presidential election polls, rather than mixing the two figures and declaring who “will win,” it is important first to see which conditions the figures answer.
The Next Datafolha and Ideia Releases Are Not Results Yet
According to O GLOBO, Datafolha will conduct in-person interviews with 2,004 people from Tuesday through Friday and plans to release the results on Friday, the 11th. The poll’s margin of error is 2% points, and it includes questions presenting a candidate list including Lula and Flávio, spontaneous-response questions, and runoff scenarios.
In the previous Datafolha poll, conducted on September 3, the first round was 38% versus 33%, while the runoff scenario was 46% versus 44%. O GLOBO reported that the new poll also includes a question about the impact of the conflict surrounding the Federal Supreme Court (STF) on the electoral contest. This concerns the poll’s design and release plan, not response results that have already been published.
CartaCapital reported that Ideia plans to release a new poll of 1,500 people by telephone next Wednesday, the 9th, with a margin of error of 2.5% points. This, too, is a plan announced before publication. For understanding this trend, it is important not to give already released poll figures and plans for future polls the same weight.
References
Tags #PresidentialElectionPolls #pesquisaeleitoralpresidente #Lula #FlávioBolsonaro #Datafolha #RunoffElection #FirstRoundElection #OpinionPoll #ApprovalRating #MarginOfError #AugustoCury #Ideia