Let’s play a guessing game. I’m thinking of a person who was first elected to the lower chamber of Congress in 1991. He made a name for himself over the following decades as one of the most ideologically extreme members of Congress, and was known for making favorable statements about authoritarian regimes. He never joined a major party, and did not pass much legislation. When he first decided to run for president several years ago, no one expected him to succeed, but an extremely online following of devoted fans defied all expectations.

Did you guess Jair Bolsonaro or Bernie Sanders?

I was talking about both of them. In an ideological and partisan sense, we could call Bernie Sanders the Jair Bolsonaro of the American left. That said, a Bernie Sanders presidency could play out very differently from that of Jair Bolsonaro.

The current presidential election season in the United States is shaping up to be highly unusual – and not only because the current front-runner, Bernie Sanders, calls himself a socialist. We might look back on 2020 as the year of the Brazilianization of American politics.

As a professor of comparative politics, I usually teach my American students about the politics of countries such as Brazil by pointing out how our systems are similar or different. Today, though, I’ll flip the chalkboard over to explain the US in comparison to Brazil.

Brazilians are used to the idea that politicians switch in and out of parties, and that voters don’t care very much which party a candidate runs on, with the partial exception of the PT. But not in the United States. American presidential candidates are usually people who have gradually built careers inside the Democratic or Republican Party over decades. It is a sign of the weirdness of the present moment that three of our top presidential contenders – Donald Trump, Bernie Sanders, and the billionaire Michael Bloomberg – all joined their parties recently, and only to run for office.

To understand why parties and elections historically worked so differently in the US and Brazil, let’s talk about electoral rules for a moment. We’ll start with the legislature. In the US, we elect legislative candidates in single member districts, meaning that just one person can win the election in each district. The candidate who gets more votes than anyone else wins – even if he or she didn’t get a majority.

One of the very few “laws” known to political science is “Duverger’s Law.” Back in the 1950s, the French sociologist Maurice Duverger noticed something: countries that use electoral rules like those in the United States tend to have just two large parties in Congress. Duverger argued that both “mechanical” and “psychological” reasons explain why plurality elections in single member districts tend to reduce the number of parties. A party has to take close to half the vote in a district to get a seat in Congress. Knowing this, voters avoid minor parties they think have no chance of winning. Compare that to Brazil’s “open list proportional representation” rules, which allow a party to win seats even if it takes only a small percentage of the vote.

The electoral system for the presidency in the US reinforces the tendency toward two parties. Presidential elections here require tremendous resources and organization – and many citizens feel that it’s a waste to vote for a candidate from a party that doesn’t have a strong shot at getting approximately half the vote. As a consequence, our two big parties from Congress naturally fall into the role of running presidential elections as well. This feedback effect further strengthens the leadership of the Republican and Democratic Parties.

Reforms in the 1970s were supposed to change that. In the presidential campaign of 1968, many Democrats revolted when their party refused to give its nomination to the popular anti-war candidate Eugene McCarty. The party instead nominated Hubert Humphrey, who had failed to win a single state primary – and Humphrey lost. Following the 1968 debacle, both parties began a process of internal reform to democratize the nomination process. Primary elections were supposed to move control out of the hands of elites and give more power to ordinary people to choose which candidates their parties would nominate.

However, party leaders didn’t just lie down and give up control. A decade ago, a team of four prominent political scientists published an important book called The Party Decides. Marty Cohen, David Karol, Marty Cohen, and John Zaller showed how party insiders still found myriad ways to shape the outcome through an “invisible primary.” Before ordinary voters ever cast their first vote, party leaders had already sifted through the candidate pool and chosen the likely winner through campaign donations and endorsements.

But in the last two elections, it doesn’t seem like party leaders are deciding. Suddenly, the process for nominating presidential candidates seems more chaotic and less controllable.

Here’s where the Brazilianization of American politics begins.

David Samuels from the University of Minnesota and Cesar Zucco from the Fundação Getúlio Vargas argue that over the past two decades, antipetismo has become a social identity that affects everything about Brazilian politics. Researchers can’t predict elections just by asking which party Brazilians support – now, we also need to know which ones they loathe.

The strange thing is that what scholars call “affective polarization” between parties has also skyrocketed in the US in the past two decades. By Barack Obama’s second term (2012-2016), many Republicans detested him. In a way, it was not too different from how some Brazilians felt about the PT in those years. This is just one of the eerie parallels between our two countries.

It might seem logical that growing animosity towards opponents would make Americans more willing to follow leaders from their own parties. Paradoxically, though, it seems to be eroding leaders’ control instead. As the scholar Julia Azari commented days before the 2016 election, “The defining characteristic of our moment is that parties are weak while partisanship is strong.”

Motivated by their strong opposition to Obama and the Democratic Party, a very high number of Republicans ran for president in 2016. Among them was the outsider businessman Donald Trump. Republican Party leaders were initially not thrilled with Trump’s candidacy. The question was, whom would they support instead?

If 2016 had been a typical election, Republican Party leaders might have piled on to the campaign of someone like Jeb Bush – former governor of Florida and family member of two former presidents. Yet in the unusually large field, Republican leaders couldn’t coordinate.

What leaders on the right cared most about, though, was defeating the left. When they couldn’t unite support behind one of the establishment rightists, they threw their support behind the populist outsider.

Does any of this sound familiar? There are striking parallels between the US 2016 election and Brazil’s 2018 election – even the sorry performance of the favored insiders, Jeb Bush and Geraldo Alckmin. (Incidentally, translating “picolé de chuchu” into English is always fun, as few Americans have had the honor of encountering the noble chuchu.)

Which brings us to 2020. It looks like what happened in 2016 among Republicans might now be repeating itself among Democrats. Just like Republican voters moved rightward as a reaction against Obama, Democratic voters have moved leftward under Trump. In addition, a huge field of Democratic candidates has materialized, all eager to defeat Trump. But party leaders can’t agree on which candidate is best suited to beat him, and they haven’t been able to coordinate very well in the invisible primary. Lacking coordination among insiders, outsiders such as Bloomberg and Sanders who have fractious relationships with party leaders are prospering.

In principle, a solid majority of Democratic voters would surely support whichever insider candidate was the best bet to beat Donald Trump. However, the visible primary isn’t well suited to help them figure that out. The end result may well be that the party nominates Bernie Sanders, who has a devoted base.

So, when I say that Bernie is the Jair Bolsonaro of the American left, I mean that he is more ideologically extreme than many of his potential voters, and that he has weak, instrumental, and fractious ties to his party.

By now, you have probably noticed that Donald Trump also shares those traits. But here’s something interesting: Donald Trump’s relationship to his party has gotten better. Three years into his term, Trump has gained some control over the party, in part by pushing out many of his opponents. By contrast, Bolsonaro…well, he will probably have a better relationship with his new party than his old one, but he may never have an effective party coalition in Congress.

Why has Trump managed to exert partisan control, yet O Mito struggles? A lot of it is about the pre-existing party and media systems. In the extremely polarized, two-party system of the US, moderate Republicans recognize that opposing Trump is the same as helping the Democratic Party. As a result, they bite their tongues. By contrast, Rodrigo Maia appears convinced that he can oppose Bolsonaro without helping the PT.

Would a possible President Bernie Sanders end up like Bolsonaro or like Trump? Partisan polarization in the US could encourage Democrats in Congress to go along with Sanders, just as Republicans have gone along with Trump – but I’m far from certain the current cast of Democratic characters would do so.

What seems like a more solid bet is that a President Bernie Sanders would further polarize the electorate, as Republicans rejected the presidency of a self-described socialist. As Harvard professors Steve Levitsky and Daniel Ziblatt explain in their book How Democracies Die, there are two big ways democracies break down: when insiders consolidate power, or when outsiders overthrow them. And growing polarization raises the risks of both.

Amy Erica Smith: Associate Professor of Political Science, Iowa State University