Datafolha Survey in São Paulo: Covas, 32%; Boulos, 16%; Russomanno, 14%; France, 12% | 2020 elections in São Paulo

• Bruno Covas (PSDB): 32%
• Guilherme Boulos (PSOL): 16%
• Celso Russomanno (Republicans): 14%
• Márcio França (PSB): 12%
• Arthur do Val – Mamãe Fali (Patriot): 4%
• Jilmar Tatto (PT): 4%
• Joice Hasselmann (PSL): 3%
• Andrea Matarazzo (PSD): 2%
• Marina Helou (Chain): 1%
• Vera Lúcia (PSTU): 1%
• None / blank / null: 7%
• I don’t know: 3%

Levy Fidelix (PRTB), Orlando Silva (PCdoB) and Antônio Carlos Silva (PCO) had less than 1%.

Last Wednesday (11), the Electoral Justice authorized the disclosure of this investigation, carried out by the Datafolha institute in partnership with Globo. The prohibition of disclosure had been requested by the coalition of Celso Russomano (Republicans). The coalition questions the sampling criteria of the interviewees and the absence of a runoff simulation without the presence of the candidate Bruno Covas (PSDB), who appears ahead in the polls.

The dissemination of the survey was authorized with the following clarification required by the Court: “This survey is contested in the Electoral Tribunal for the alleged absence, in its results, of considering the economic level of the interviewees, as well as dividing the level of education of these, in the sampling plan, having been in two categories (fundamental and medium level: drop of 67%; higher level: drop of 33%) ”.

Alessandro Janoni, Datafolha research director, said that the institute uses as a reference in the 2020 elections the same variables of sample planning and data weighting that for more than 35 years dictates the monitoring of the elections in the city of São Paulo, with the aim of representing all strata of the São Paulo electorate.

• See the previous Datafolha survey, published on November 5.

In relation to the previous survey:

• Bruno Covas (PSDB) went from 28% to 32%
• Guilherme Boulos (PSOL) went from 14% to 16%
• Celso Russomanno (Republicans) went from 16% to 14%
• Márcio França (PSB) went from 13% to 12%
• Jilmar Tatto (PT) went from 6% to 4%
• Arthur do Val – Mamãe Falei (Patriota) kept 4%
• Joice Hasselmann (PSL) stayed at 3%
• Andrea Matarazzo (PSD) went from 3% to 2%
• Marina Helou (Rede) remained at 1%
• Vera Lucia (PSTU) went from 0% to 1%
• Orlando Silva (PCdoB) went from 1% to 0%
• Levy Fidelix (PRTB) went from 1% to 0%
• Antônio Carlos (PCO) remained at 0%
• None / blank / null was 9% to 7%
• I don’t know if it stayed at 3%

See the previous Datafolha survey, published on October 22.

See the previous Datafolha survey, released October 8.

Compared to the survey conducted between November 3-4, Russomanno had a steeper decline among voters with basic education (from 23% to 15%), among the poorest, with household incomes of up to 2 salaries (from 24%). 15%) and among evangelicals (25% to 19%).

These were also some of the segments in which Bruno Covas has advanced the most since the last week: the least educated nurtures the toucan from 30% to 41%; on the lowest-income platform, from 25% to 34%; and among evangelicals, from 25% to 32%. In the portion with income of 5 to 10 salaries, the toucan grew from 23% to 31%, and among voters aged 60 and over, from 38% to 47%. Covas’ voting intention among the elderly contracted his performance in the age group from 16 to 24, in which he dropped from 19% to 17%, his worst result among the worst sociodemographic cuts in the survey. Boulos and França varied within the margin of error in the main segments of the electorate.

The poll also asked who voters would not vote for at all. The percentages were as follows:

• Celso Russomanno: 49%
• Joice Hasselmann: 32%
• Bruno Covas: 24%
• Jilmar’s Tattoo: 23%
• Guilherme Boulos: 23%
• Levy Fidelix: 22%
• Orlando Silva: 18%
• Márcio França: 17%
• Arthur do Val: 15%
• Vera Lúcia: 13%
• Andrea Matarazzo: 12%
• Antônio Carlos: 11%
• Marina Helou: 11%
• Reject all / would not vote for none: 3%
• Would vote for any / reject none: 2%
• Don’t know: 4%

Respondents could indicate more than one answer, so the sum of the mentioned factors is more than 100%.

Datafolha also addressed spontaneous voting intentions, where the voter says who they will vote for without the names of the candidates being presented. See the results:

• Bruno Covas (PSDB): 21%
• Guilherme Boulos (PSOL): 14%
• Celso Russomanno (Republicans): 7%
• Márcio França (PSB): 7%
• Arthur do Val (Patriot): 3%
• Jilmar Tatto (PT): 2%
• Andrea Matarazzo (PSD): 1%
• Joice Hasselmann (PSL): 1%
• PT candidate: 1%
• Others: 5%
• White / null / none: 9%
• Don’t know: 28%

Degree of knowledge of the interviewee about the candidate’s number

• Bruno Covas (PSDB): 49% know (correct mention) and 51% do not know (7% incorrect mention and 44% do not know the number).
• Celso Russomanno (Republicans): 29% know (correct mention) and 70% do not know (6% incorrect mention and 64% do not know the number).
• Guilherme Boulos (PSOL): 59% know (correct mention) and 41% do not know (4% incorrect mention and 37% do not know the number).
• Márcio França (PSB): 43% know (correct mention) and 57% do not know (10% incorrect mention and 47% do not know the number).
• Jilmar Tatto (PT): 72% know (correct mention) and 28% do not know (3% incorrect mention and 25% do not know the number).

The survey also asked the degree of decision of the vote. 67% of voters are fully determined and 32% say it can still change. See the degree of decision in relation to each candidate:

• Bruno Covas: 69% are fully determined; 30% can still change the vote.
• Guilherme Boulos: 81% are fully determined; 19% can still change the vote.
• Celso Russomanno: 56% are fully determined; 44% can still change the vote.
• Márcio França: 59% are fully determined; 40% can still change the vote.
• Jilmar Tatto: 82% are fully determined; 18% can still change the vote.

Among those who stated that they can still change their vote, Datafolha also asked, if interviewees do not vote for the first option, which other candidate is more likely to receive their vote.

• Bruno Covas: 19%
• Celso Russomanno: 15%
• Márcio França: 14%
• Jilmar’s Tattoo: 7%
• Guilherme Boulos: 6%

See the distribution of these votes:

31% of the voters who vote for Russomano would vote for Bruno Covas; 42% among Boulos voters; 23% among French voters; 8% among Tatto voters.

29% of the voters who declared their vote in Covas would vote for Márcio França; 24% among Russomanno voters; 30% among Boulos voters; 27% among the voters in Tatto.

24% of the voters who declare their vote in Covas would vote for Celso Russomanno; 2% among Boulos voters; 17% among French voters; 17% among the voters in Tatto.

Guilherme Boulos would vote for 10% of the voters who declared their vote in Covas; 5% among Russomanno voters; 15% among French voters.

5% of the voters who vote in Covas would vote for Jilmar Tatto; 5% among Russomanno voters; 14% among Boulos voters; 14% among French voters.

Second shift simulations

Datafolha also questioned, if the second round of the mayoral elections were today, who would voters vote for:

• Bruno Covas 59% x 25% Celso Russomanno (white / null: 15%; I don’t know: 1%)
• Bruno Covas 56% x 30% Guilherme Boulos (white / null: 12%; don’t know: 2%)
• Bruno Covas 53% x 34% Márcio França (blank / null: 11%; don’t know: 3%)

• The survey was commissioned by TV Globo and the newspaper “Folha de S. Paulo”.
• Margin of error: 3 percentage points more or less
• Who was heard: 1,512 voters in the city of São Paulo aged 16 years or older.
• When the survey was conducted: November 9-10, 2020.
• Identification number in the Electoral Tribunal: SP-05584/2020
• The confidence level used is 95%. This means that there is a 95% probability that the results reflect the current electoral moment, considering the margin of error.

Electoral surveys: what is a sample, margin of error and level of confidence.

VIDEOS: Meet the candidates of the SP City Council