HOME > The ability to

Support for Lula reaches its highest level after legal back-and-forth, according to XP/Ipespe data.

In the heat of the legal imbroglio that almost culminated in his release last Sunday (8), former president Lula saw the level of support for his candidacy reach its highest point in a month. According to a survey conducted by Ipespe between July 9 and 11, the PT leader has 30% of voting intentions in the only first-round simulation that considers his candidacy, a 2 percentage point increase compared to the survey from the previous week.

Support for Lula reaches its highest level after legal back-and-forth, shows XP/Ipespe (Photo: Stuckert)

SÃO PAULO - In the heat of the legal imbroglio that almost culminated in his release last Sunday, former president Luiz Inácio Lula da Silva saw his support for his candidacy reach its highest level in a month. According to a survey conducted by Ipespe (Institute of Social, Political and Economic Research) between July 9 and 11, the eighth commissioned by XP Investimentos, the Workers' Party leader has 30% of the voting intentions in the only first-round simulation that considers his candidacy. Convicted in the second instance for the crimes of passive corruption and money laundering and imprisoned for more than three months, Lula risks having his candidacy barred by the Clean Record Law.

Lula's performance represents a positive fluctuation of 2 percentage points compared to the previous week's survey, within the maximum margin of error of 3,2 percentage points up or down. This was the third time Lula reached 30%, his highest level in the XP/Ipespe series, which began in May. Right behind him in this scenario is federal deputy Jair Bolsonaro (PSL), with 20% of voting intentions, the same level as the previous week. Former senator Marina Silva (Rede) has 10%, while the former governors of São Paulo, Geraldo Alckmin (PSDB), and Ceará, Ciro Gomes (PDT), each have 7%. Blank, null, and undecided votes total 15%.

Since Lula's movement occurred within the margin of error, it is necessary to monitor the Workers' Party candidate's performance in upcoming polls to determine if he has actually gained ground. Even if his candidacy is considered unlikely in political circles, the former president's performance could be an important indicator of his ability to transfer votes in the race.

Bolsonaro leads in all other first-round scenarios tested, with a 9 to 10 percentage point advantage over the second-place candidate. With these results, if the election were held today, it would be impossible to predict who his opponent would be. In the most uncertain scenario, which possibly best projects the start of the presidential race, four candidates appear technically tied: Marina Silva, Ciro Gomes, Geraldo Alckmin, and Fernando Haddad (PT), with Lula's support. The former mayor of São Paulo jumps from 2% of voting intentions to 12% simply by including the information that he would be Lula's chosen candidate, in an exercise testing the former president's ability to transfer votes. This performance shows a 1 percentage point increase compared to the previous week. Other fluctuations are within the margin of error when compared to the results of the last four weeks.

In simulations without Lula, the "non-vote" group reaches 35%. This performance reinforces the degree of uncertainty in this election less than three months before the first round. The XP/Ipespe survey also showed that, in the last presidential election, when the scenario was much less unclear regarding the candidates, 51% of respondents decided on their presidential vote after the start of the radio and television campaign period. Only 39% of voters say they chose a candidate before this stage. Expectations are that this behavior will be repeated in this election, which tends to indicate greater fluctuations in the coming months.

This assessment is corroborated by the fact that, in the spontaneous survey, when no candidate names are presented to respondents, blank, null, and undecided votes total 64% of the electorate. In this scenario, Bolsonaro appears with 15% of voting intentions, technically tied with Lula, who has 13%. Ciro Gomes and Geraldo Alckmin each have 2%, while other candidates do not exceed 1%.

To access the breakdown of votes for the main candidates in each of the scenarios tested, click here.

Second round

As in previous editions of the survey, six second-round scenarios were tested. In a hypothetical runoff between Alckmin and Haddad, the PSDB candidate would win by 30% to 20%, with 44% of votes being blank, null, or undecided. In the third week of May, they appeared to be technically tied, but the difference began to widen in subsequent polls.

In a simulated contest between Lula and Bolsonaro, the Workers' Party candidate now appears ahead, by 40% to 33%, above the maximum margin of error, with 27% of blank, null, and undecided votes. A week ago, the advantage was 6 points, which constituted a technical tie. At the beginning of the historical series, the congressman appeared 2 points ahead, also in a situation of a technical tie.

If Bolsonaro and Alckmin were to face each other, the situation would be a technical tie, with the congressman numerically ahead by a score of 34% to 32%. Blank, null, and undecided votes total 35%. The result is the same as a week ago. The difference between the candidates reached 7 percentage points in the fourth week of May. At no point so far has the PSDB candidate led the race.

In a potential contest between Marina Silva and Bolsonaro, the scenario is also a technical tie, with the former senator numerically ahead by 37% to 33%. Blank, null, and undecided votes total 29%. The congressman was numerically ahead in the first two surveys of the series, conducted in the third and fourth weeks of May, when the difference reached 6 percentage points in favor of the parliamentarian, also within the limit of the sum of the margins of error for both candidates.

A technical tie is also observed in the simulated contest between Alckmin and Ciro, with the PSDB candidate numerically ahead by 32% to 30%, the same difference recorded a week earlier. Blank, null, and undecided votes total 38%. In the last week of June, both appeared with 32% of the voting intentions each.

If Bolsonaro and Ciro were to face each other, the scenario would also be a technical tie (as in the last five weeks), with the congressman numerically ahead, with 33% of voting intentions against 31%. Blank, null, and undecided votes total 36%. In the first two surveys, the congressman won the race by a margin greater than the sum of the candidates' margin of error.

Rejection of candidates

The survey also asked respondents about the candidates they would not vote for under any circumstances. Lula remains the leader in rejection, now at 62%, but technically tied with four other potential candidates: Ciro Gomes (59%), Geraldo Alckmin (59%), Marina Silva (58%), and Fernando Haddad (57%). The trajectory of the main candidates is shown in the table below:

The voter has not yet woken up.

Not even the Brazilian national team's exit from the World Cup a week before the end of the competition, with the defeat to Belgium, was able to generate an increase in the respondents' interest in the upcoming elections. With less than three months until the first round, 36% say they are not interested in the process, while 19% say they are somewhat interested. Only 25% responded that they are paying closer attention to the election. Almost all the variations presented throughout the eight surveys already conducted are within the maximum margin of error of 3,2 percentage points up or down.

Election "pool"

The survey also asked voters who they believe will be elected president in October. In the betting pool, Bolsonaro maintained his lead, with 28% of the indications. Last week, the congressman was the name pointed out by 30% of those interviewed, four percentage points below his peak, recorded in the fourth week of May. Lula, on the other hand, saw his indications fluctuate from 24% to 27%, within the margin of error. This is the highest level for the former president in the historical series of the "electoral pool" evaluated by the survey.

To access the full text of this and all other research, click here

Exclusive

The XP/Ipespe survey was conducted by telephone between July 2nd and 4th, interviewing 1.000 respondents across all regions of the country. Questionnaires were administered "live" by interviewers (with random selection of candidate names in prompted questions) and subject to subsequent auditing in 20% of cases to verify responses. The sample represents all Brazilian voters with access to a landline (at home or work) and a mobile phone, stratified by sex, age, education level, family income, etc.

The confidence interval is 95,45%, meaning that if the questionnaire were applied more than once in the same period and under the same conditions, this would be the chance of the result being repeated within the maximum margin of error, established at 3,2 percentage points. The survey is registered with the TSE (Superior Electoral Court) under code BR-09898/2018 and cost R$ 30.000,00.

Ipespe has been conducting telephone surveys since 1993 and was the first institute in Brazil to do so. tracking telephone polls in election campaigns. The institute's scientific council is chaired by sociologist Antonio Lavareda, and its executive director is Marcela Montenegro.