Last week, one of the most commented subjects about the elections were the opinion polls. Different institutes conducted surveys in person, with home interviews, between September 9 and 13 and presented diverging results: in the IPEC poll Lula da Silva (Workers' Party) has 46% and Jair Bolsonaro (Liberal Party) has 31%; and in the Quaest, Lula has 42% and Bolsonaro has 34% of voting intentions. But, after all, what justifies this difference between the main candidates?
The answer lies in a fundamental question, but one that is not made clear to readers of the surveys: the methodology. This is because an economic cleavage is made so that the survey contemplates percentages of the different income brackets of Brazilian society.
This is exactly the point of divergence between the institutes. On the one hand, IPEC stipulates that 57% of the voters currently earn up to two minimum wages; for the Quaest, this percentage is 38% - this income bracket is the one which most concentrates votes for Lula. On the other hand, the Quaest considers 22% of the population earning more than five minimum wages, while IPEC considers 15% - this is the income range that most concentrates votes for Bolsonaro. Thus, it was expected that the results would be conflicting.
:: Bolsonaro’s campaign uses presidential e-mail to receive electoral donations ::
So why don't the institutes consult the same percentage of the different income brackets? The answer may concern the decisions of the current Brazilian president. Bolsonaro postponed for two years the census that should have taken place in 2020. As a result, the most updated data we have about the Brazilian population is from 2010. With twelve years of disparity, after years of dismantling public policies to combat poverty and a pandemic, it is not possible to determine the percentage of the population in each income bracket - and, consequently, which institute is closer to reality.
In general, there is a tendency to rely more on the data presented by IPEC, because it uses a higher percentage of people in the range of up to 2 minimum wages in its methodology. It is believed to be the closest to the real thing when considering, for example, the "National Survey on Food Insecurity in the Context of the Covid-19 Pandemic in Brazil", which pointed to the existence of 33 million people going hungry in Brazil. With this, Lula would have an approximate 15 percentage point advantage over Bolsonaro.
A new Datafolha poll is scheduled for this Thursday (15). The institute was the one that came closest to the voting percentages in the troubled 2018 elections.
Edited by: Thales Schmidt
