The (Superior Electoral Court) postponed this Tuesday (9) the analysis of the censorship requested by () the disclosure of, amid public messages from ministers about the impacts that the decision will have on this year’s elections.
After the president of the electoral court, , voted to maintain his request by the PL pre-candidate, minister Estela Aranha asked for a review of the case — more time for reflection, but with a deadline still undefined.
In the same session, other magistrates expressed concern about the issue and even considered setting rules for research institutes.
The Atlas/Bloomberg survey had been released on May 19. 5,032 people were interviewed using the Atlas RDR method, an acronym for random digital recruitment, from May 13th to 18th.
The survey showed that Flávio had dropped 6 points in the second round scenario against (PT), after the “Dark Horse” case — with the disclosure of the former banker.
This Tuesday, Estela Aranha did not speak at the session, but Kassio stated that she had anticipated to the presidency her intention to have more time.
The president of the TSE read his previously given decision, in which he stated that there was evidence of the use of the questionnaire as a mechanism to induct the interviewee.
“Although research institutes have technical autonomy to define the methodology used, in surveys that carry out this prerogative it does not remove jurisdictional control in cases where there is evidence of distortion of the research”, he said.
On Monday (8), Kassio issued a preliminary decision and forwarded it to the other ministers for analysis. Until further deliberation, AtlasIntel is prohibited from disclosing, promoting or republishing the survey that showed a drop in Flávio’s voting intentions.
Even after the request for review, some ministers made comments on the topic. Dias Toffoli, who assumed a position as effective minister of the TSE this Tuesday, argued that the ideal is to take advantage of the trial to debate the topic more broadly.
“What we are doing here is defining what parameters this court will have in relation to any and all research and any candidate. In relation to any and all research and any candidate, we cannot decide one way in relation to one candidate and another in relation to another candidate”, he stated.
“What we are going to decide in this case is the future. And what would be this limit of what is inducement and what is not inducement? This limit cannot be subjective,” he said.
According to Toffoli, “in the same way that a research institute can lead to a result X, another research institute can lead certain questions to a result Y.”
Afterwards, Minister André Mendonça, vice-president of the TSE, added that, regardless of the outcome of the specific case, it is part of the court’s task to try to bring fair rules to the electoral process. And, according to him, part of this construction are research institutes.
“It is important that we redouble this awareness of research institutes, of the public role they play before not only , but also the entire Brazilian society. The importance of research institutes also being agents of cooperation for the impartiality and fairness of the electoral process”, he stated.
Minister Floriano de Azevedo Marques shared the concern. “I understand that this moment of definition, not only in the case, but as a setting of a precedent, demands that we already have the instructive completeness to know all the variables and, then, to deliberate together on the decision and the precedent”, he stated.
In the request to suspend the publication of the survey, the team of the PL pre-candidate for the Presidency stated that the arrangement of questions and themes and the “use of associations” between Flávio and the owner of , , “contaminate and induce the answers of the interviewees”.
In a statement, AtlasIntel defended the scientific rigor of the research and said that the collection of voting intentions took place without the audio being played during the questionnaire. According to the company, the material was only presented to users at a later stage — without the possibility of returning to questions or changing answers already recorded.
One of the lawyers signing the request is a former TSE minister. She is part of the Bolsonaro legal team.
According to her, in oral argument at the TSE, controlling research is not new. She also stated that the survey under discussion became known and was circulated for 20 days, unlike others that are prevented from being made public.
According to the questionnaire made available by AtlasIntel to the TSE, the content of was displayed to respondents, but as the last item in the survey. Voters who collaborated in the survey were asked 48 questions, the first of which were about their voting intentions.
In the last question, respondents analyzed a video with audio and could drag to the right when they were “evaluating it more positively” and to the left when they were “evaluating the content more negatively”. The play had images of Flávio and Vorcaro, to illustrate the dialogue.
Kassio stated that, in fact, such circumstances corroborate the arguments that indicate “the possible use of inductive stimuli capable of contaminating subsequent responses regarding image, rejection and voting intention, reinforcing the legal plausibility of the thesis that the research may have gone beyond the limits of regular statistical measurement”.
The minister stated that this was not a methodological disagreement, but a possible induction of the interviewee based on the questionnaire, “especially due to the sequential order of the questions and the use of expressions with a negative evaluative charge”.
Kassio ordered the research institute to present, within two days, additional technical documentation on the survey methodology, especially in relation to the use of Flávio and Vorcaro’s audio, and asked the Public Electoral Ministry to comment on the case within one day.
