Very early on the night of the 2020 presidential election, it was clear that something unexpected was happening with Hispanic voters.
Florida was one of the first states to report results, and the numbers in heavily Hispanic Miami-Dade County, which Hillary Clinton won by nearly 30 points in 2016, showed Joe Biden leading Donald Trump by a much narrower margin. The shift held, with Biden winning by only about seven points. If Biden had won the county by the same margin as Clinton, he would have lost Florida by 116,000 votes, not 372,000.
That shift was quickly attributed (including by myself) to the density of Cuban immigrants in the area. While Hispanics nationally tend to vote more heavily Democratic than Republican, Cubans in Florida vote more heavily Republican. That Trump had focused heavily on socialism during his campaign, even hosting a Cuban immigrant to give a speech during the Republican convention, suggested that this shift was probably in part a response to Trump’s campaign rhetoric.
It was understood that Biden was underperforming in Florida, but the results were still a surprise. Then, as results continued to trickle in, it became obvious that the shift wasn’t limited to South Florida. Across the country, the higher the density of Hispanics in a county, the more it tended to shift toward Trump in 2020. Mind you, Biden still won a lot of those counties, as he did Miami-Dade, but by a much narrower margin than Clinton did in 2016.
Digging into the numbers a bit, we see a key reason for that. The number of votes Biden received was only slightly higher than the number of votes Clinton had amassed in those densely Hispanic counties. The number of votes Trump received, though, spiked.
If we group counties, the pattern becomes clearer. Counties that are about 30 percent Hispanic, according to Census Bureau data, saw little shift in their overall presidential vote margins relative to 2016, with Biden’s vote percentage increasing over Clinton’s by more than Trump’s improvement. As the density of the Hispanic population increased, though, those metrics shifted dramatically in Trump’s favor.
The question that has lingered since the election is why? What prompted these more heavily Hispanic areas to vote more favorably for Trump?
Just looking at these data, it’s hard to say. It’s hard to know, for example, if the increase in votes for Trump is a function of votes being drawn from Biden — people persuaded to support the incumbent president — or of people coming out to vote who otherwise wouldn’t have, a shift in turnout.
Research published by Equis Labs (and first reported by Axios) offers some insight into that question. The shift for Trump appears to have been a function of both persuasion and increased turnout.
“We know enough to say it’s a mistake to assume the 2020 shift was exclusively about turnout, or about vote-switching by Clinton voters,” a presentation released by the firm argues. “The truth is closer to both.”
To that point, Equis walks through data from several states with large Hispanic populations. It shows, for example, that the decreased attention paid to immigration in the 2020 election likely helped Trump with Latino voters, since it shifted the focus to the economy, where Trump polled far better. This likely helped shift views of Trump to his benefit. (We use the term “Latino” to refer to the Equis data, in keeping with the researchers’ phrasing. The Census Bureau breaks out demographic data for those of Hispanic ethnicity.)
The most intriguing part of the analysis (in my opinion) is how that shift overlapped with lower-propensity voters. In Equis’s data, those Latino voters who were less likely to vote in 2020 showed increased approval of Trump relative to 2019 before the election.
That overlaps with other evidence of an increase in turnout among Latinos. Data from Catalist, for example, show that a third of Latinos who voted in 2016 or 2020 voted only in 2020. (For Whites, by contrast, only about a fifth did.) In Nevada, about 21 percent of Latino Trump voters didn’t vote in 2016 — even though 17 percent of them were eligible then. In the Texas Rio Grande Valley, Equis found that nearly 6 in 10 of Latino swing voters who cast ballots had voted no more than once in the 2014, 2016 and 2018 federal elections.
One of the factors that drove Trump’s success in 2016 was that he managed to motivate low-frequency voters — largely Whites without college educations — to come out and support his candidacy. The Equis data suggest that he did something similar in 2020, spurring much more enthusiasm for voting among conservative Latinos, including both conservative Republicans (about 14 percent of the Latino voting population) and conservative Democrats (11 percent). Those groups were not only more motivated to vote but also became more approving of Trump’s presidency from 2019 to 2020 (though only modestly in the case of those Democrats).
If you can get 25 percent of a population to be more likely to vote and more supportive of your candidate, that’s a good recipe for seeing improvement with that population. Particularly if the rest of the population is either not more motivated or not more negative, which the Equis data suggest was the case with Latinos.
The Equis analysis notes that there are still a number of unanswered questions, but it nonetheless offers an intriguing new lens through which to consider what happened in 2020. The presentation is also pointed about the ways in which mainstream analysis is oversimplified.
“Thinking in terms of ‘turnout vs. persuasion’ misses the larger story about Latinos who feel on the sidelines of the political process,” it reads. That’s unquestionably true — and a better sense of why Americans choose not to vote would certainly give us a better understanding of those occasions when they suddenly do.