On Sunday, Brazil’s Electoral Superior Court announced that no candidate had crossed the 50% threshold in the country’s presidential election, sending voting to a second round on October 30 between incumbent President Jair Bolsonaro, a staunch conservative, and former president Luiz Inácio “Lula” da Silva, a socialist.
Brazil is the largest country in Latin America, and with more than 210 million people, is the most populous in the Western Hemisphere after the United States. Bolsonaro, a populist who was elected in an upset in 2018, has feuded not just with the left-wing opposition, but also with much of the traditional establishment elite on a range of issues from COVID-19, where he opposed lockdowns, to crime and foreign policy.
Bolsonaro forged close ties with Donald Trump and his sons, active political influencers in their own right, and made appearances at CPAC and other U.S. conservative gatherings. In turn, the new Biden administration has made its attitude toward Bolsonaro clear, barely disguising a preference that he be defeated by his left-wing rival, Lula of the socialist Workers Party. This made the first round of Brazil’s election a proxy battle between left and right, and populism and elitism, not just in Brazil, but globally.
That more was at stake than simple partisanship was evident in the results. Despite polls showing a double-digit lead for Lula, who was expected to potentially reach 50% in the first round, thereby avoiding a runoff, Bolsonaro held him to a 5% lead, 48.42% to 43.21%, ensuring a second round. With striking parallels to the U.S. 2020 election, polls showing a large lead for the left-wing candidate, Biden in the case of the U.S., Lula in Brazil, more or less correctly measured the support for that candidate but badly underestimated support for their opponent.
The Economist’s final polling average had Lula leading by 49% to 37%, compared to the final result of 48.42% to 43.21%. Lula was within 0.6% of his polling average but Bolsonaro outperformed him by more than 6%. This is already causing large-scale questioning of polling for the second round showing large, double-digit leads for Lula. The Economist average has one poll showing Lula up 14%.
What happened? And why does it keep happening? One suggestion is that Bolsonaro supporters, like Trump supporters, were much less likely to answer polls. They are apt to view the media suspiciously.
This theory is reinforced by the results. Traditionally, high-income voters cast their ballots for right-wing parties that favor low taxes, while poorer voters cast their ballots for left-wing parties which support higher social spending. But over the last decade, the major dividing line has become education and culture. With the rise of woke corporations and ESG investing, the financial elite identify success as much with the far-left values of their milieu as they do with their take-home pay, perhaps an unintentional result of tax systems so complex that most of the truly wealthy avoid paying most top rates altogether no matter who is in office.
High-income urban and suburban voters are far more likely to answer polls and speak to newspapers, as they tend to be the primary remaining consumers of traditional media. The result is that coverage of politics tends to overstate their importance. This was evident in Donald Trump’s presidency when the major media narrative was how Trump was losing the support of well-off voters who had supported Mitt Romney, John McCain, or George W Bush. Rather than fake news, this was happening, and it manifested further in 2020 when Biden actually exceeded Barack Obama’s support among white voters as a whole and did better than Obama in 2008 among those with college degrees. However, this analysis missed the other part of the story. Traditional Democratic voting constituencies such as working-class Latinos, white-collar whites, and former union voters all were moving toward Donald Trump. By measuring only those population segments moving away from Trump, the media and pollsters missed a critical part of the story.
The same thing appears to have occurred in Brazil. In much of the traditionally right-wing south, Bolsonaro did historically bad for a right-wing candidate. In 2018, in the Federal District representing the capital city, Bolsonaro won 58.37% of the vote in the first round and 70% in the second. This time he won a mere 51.7%. In Rio de Janeiro in 2018, he won 59.79% in the first round and 67.95% in the second. On Sunday, he won only 51.1%. In Sao Paulo, the largest state, he went from 53% in round 1 and 68% in round 2 in 2018 to 47.7% this year. If these swings had been representative nationally, Bolsonaro would have fallen from his 46% first round result in 2018 to around 39%, more or less in line with polls.
But that is not what happened. As Bolsonaro fell in traditional right-wing strongholds, he made inroads into poorer, previously left-leaning areas. In Bahia, the largest left-wing state, and Lula’s stronghold, Bolsonaro won 23.42% in round one of 2018, while this year he won 24.37%. In Ceara, he received 21.74% in round one of 2018, but this year won 25.4%. The pattern nationally, therefore, was for Bolsonaro to lose 6-7% support in his strongest areas while gaining 2-4% support in his weakest. The net result was that Bolsonaro fell from his 46.03% first-round result in 2018 to 43.2% in 2022.
Much as in the United States, it seems that many of the arguments used against Bolsonaro, especially regarding insufficient vigilance in imposing lockdowns, or uncouth language, were either ineffective or alienated working-class voters, who were more concerned with issues like crime and the economy than breathless warnings from elites in both Brazil and the United States that “democracy” and “the constitution” were under threat.
This produced another result reminiscent of the 2020 U.S. election. The simultaneous elections for Congress and state governments were a defeat for the left and a catastrophe for the “centrist” parties which had attempted to position themselves as opponents of left-wing policies, but enemies of Bolsonaro’s government. Those parties dominated the last congress but appear to have been reduced to only around a quarter of the seats in the new one. Bolsonaro’s allies by contrast have won an absolute majority in the House, 258 of 513 seats, and 19 of 27 seats up for election in the Senate. The wealthy voters Bolsonaro alienated nevertheless voted for right-wing parties, whereas those his enemies turned away did not return to the left down-ballot.
It is worth taking a moment to note that despite widespread predictions in the media and alarmist warnings from Washington that Bolsonaro might “resort to violence” or engage in a coup, the elections went off largely peacefully and without incident. While the runoff is likely to be fiercely contested, there is little evidence that it will turn physically violent.
That does not mean it will be entirely clean. Reporting on “fraud” has been highly simplistic in the American press when it comes to Brazil, with a tendency to portray Bolsonaro’s comments in conspiratorial ways which are easy to dismiss. Bolsonaro has been reported as warning against fraud, and suggesting that if he loses, it will be due to fraud. Combined with his fierce rhetoric against the left, this has led to a tendency to conflate the charge with the claim that if he loses it will be because the Left, which is out of power, and Lula, who was in prison until three years ago, somehow influenced the courts and local governments they do not control to throw the election.
Bolsonaro’s point is in fact more nuanced. Brazil has made large strides toward election security, but elections are still heavily influenced by local political machines which, regardless of ideology, are self-interested. The runoff will take place in a context in which the right has won unassailable majorities in congress. This creates an interesting dynamic for some voters, who perhaps want a check on the expanded right-wing majority in the form of a president from the other side of the political spectrum.
But the more relevant concern for Bolsonaro is that the current situation creates a choice for right-wing parties and machine politicians. If Bolsonaro wins, the congressmen and senators elected on his coattails will be obligated to support his government, pass his policies, and confirm his ministers. By contrast, if Lula wins, Lula will only be able to appoint ministers who conservatives in the Brazilian Congress approve of, and implement policies they pass. Given their record of trying to impeach every Brazilian President, it will also likely be much easier to get rid of Lula if needed than Bolsonaro.
That Lula, who will be 77 when he takes office, chose as his vice presidential running-mate Geraldo Alckmin, who was his conservative opponent in the 2006 election, cannot be lost upon establishment members of congress. It is as if Joe Biden had selected Mitt Romney or John Kasich as his running mate in 2020. Could Donald Trump have trusted in the unqualified and enthusiastic support of every Republican governor and senator in a hypothetical 2020 runoff against such a ticket?
Sometimes leaders accused of paranoia are right that people are out to get them. Bolsonaro has faced unprecedented opposition from much of the centrist establishment for his entire tenure, as well as the diehard opposition of the left. What he has tried to warn his supporters about in the runoff is treachery by members of the right, especially in local political machines.
If Lula wins, it is by no means explainable only by fraud. He won twice in the past, and he also won 48.4% in the first round. But the runoff is going to be conducted in a slanted manner, in which Lula can count on not just the support of his own supporters, but tacit assistance from many nominal supporters of Bolsonaro. That is likely to produce tensions and suggestions that Lula’s win, while perhaps reflective of votes cast, is nonetheless in effect a coup by congress and governors, both left and right, against a president who is too strong for their comfort.
This nuance is of course lost on the Biden administration, which sees “democracy” as synonymous with the elites who dominate the traditional institutions of any system. Therefore, the removal of someone who challenges that dominance is hailed as a victory over a “threat to democracy” by Biden and his ilk. Hence actions by center-right officials to aid Lula and undermine Bolsonaro are defined as actions in defense of democracy, while Bolsonaro’s efforts to block them are attacks on it. In reality, neither outcome on October 30 is likely to be fully democratic to American sensibilities. But one of them will involve a reversion to the system of the mid-2010s when the Brazilian Congress dominated and removed powerless puppet presidents at will. Lula’s win at this point would not be a victory for democracy, liberalism, or the left, so much as for machine control of Brazilian politics.
Daniel Berman is a frequent commentator and lecturer on foreign policy and political affairs, both nationally and internationally. He holds a Ph.D. in International Relations from the London School of Economics. He also writes as Daniel Roman.
Was Lula not in charge before and was it not disastrous?