Those 30% that Bolsonaro

There's a magic number that keeps popping up in national political debates: 30%. It's the level of support Bolsonaro supposedly has among the population, which many people treat as a proven fact and a kind of natural phenomenon.   

This 30% discussion begins with a statement by Duda Mendonça on the eve of the 2002 election. He said that, regarding the PT (Workers' Party), there were three types of people in Brazil, each corresponding to a third of the population: those who liked the PT, those who were against it, and those who were in the middle, neither liking nor disliking it. It seemed to be a great truth, stated in a simple way. 

But that was never truly the truth. Only exceptionally, in the hundreds of state and local elections for the Legislative and Executive branches that we have held since redemocratization, did the rule prevail: there was never a "floor" of 30% for votes for PT candidates. Nor in the presidential elections after that. Even in the PT's biggest victory, Lula's reelection in 2006, his opponent obtained almost 40% of the valid votes. Where would they come from, if anti-PT sentiment was only a third? Duda's formulation was good, but inaccurate. 

Even less accurate is the interpretation of popularity indicators, including presidential ones. According to Datafolha, Lula was considered an "excellent" or "good" president by 83% of people and "bad" or "terrible" by 4%. Where are the thirds? They didn't appear for Dilma either, neither in good times nor bad. For PT governors and mayors, this law never applied. 

Reasoning with the anti-PT "third," the same applies. There are dozens of examples of governors from other parties elected or approved by more than 66% of the electorate, that is, also supported by PT sympathizers.

It's not because, in relation to the party, there would be this division that every political-ideological cleavage in Brazil would have to follow the "thirds" pattern. When imagining it, Duda Mendonça was only thinking about the PT (Workers' Party), and extending the idea to everyone would be absurd. 

Bolsonaro did indeed achieve 30% approval during the second half of 2019, but this was merely a point in a downward trajectory that began immediately after the election, when he was still popular. According to Datafolha, in December 2018, 65% had favorable expectations and only 12% unfavorable expectations regarding what the government would be like. 

But from January 2019 onwards, everything went downhill. Less than 30 days into his term, according to an Ipespe/XP poll, the captain's approval rating plummeted from 60% to 40%, while disapproval doubled, reaching 20%. Twelve months later, at the end of his first year in office, in the same series of polls, the negative rating had doubled, reaching 39%, and the positive rating had fallen to 32%.

In the first five months of 2020, the trend accelerated: according to Ipespe/XP surveys, disapproval reached 50%, while approval fell from 32% to 26%. By the end of May, the proportion of those who disapproved of the captain was almost double that of those who approved.  

The shift is easy to explain: due to his stupid and irresponsible response to the Covid-19 epidemic, Bolsonaro's approval rating dropped among people with higher levels of information and education. In the overall public opinion, the decline wasn't more pronounced only because there was a slight improvement among lower-income and lower-education segments, perhaps driven by the Emergency Aid program. 

A series of other surveys conducted online in April and May point in the same direction, all showing a significant drop in approval ratings (Atlas: 58% disapproval/23% approval; Fórum: 39% disapproval/26% approval; Quaest: 49% disapproval/19% approval). These results are not directly comparable, but they help to visualize a trend. 

Bolsonaro doesn't "have" 30%, nor has he ever had, if we understand that to mean a proportion of the population that identifies with him, follows him, and respects him in a stable manner. He started with more than double that, but he lost support as his president failed and he became incapable of representing them. 

Research suggests that his effective size is somewhere between 8% and 12% of the population—those who admire him, share his ideas (?) and defend him. From there to 30%, the distance is large, and there are several reasons why people who don't identify with him might support him circumstantially, the main one being time. Many approve of him simply because they think it's too early to disapprove of him.

Time, however, is also his main adversary, as his incompetence and flaws become more evident each day. Now, with the epidemic, everything has accelerated. His performance was quickly seen as catastrophic and a disgrace to Brazil.

Just as an exercise, let's reason that he still "has" 30%. As the captain himself would say, "so what?", what to do with that support if the remaining 70% don't follow or respect him? Because if there's one thing all the polls show, it's that the "regular" rating tends to fall and what increases is disapproval. 

Bolsonaro might be able to maintain a hypothetical 30% approval rating, but he wouldn't win a majority election.