The 2020 presidential contest between Donald Trump and his Democratic challenger Joe Biden will be remembered as one of the most momentous elections in the nation’s history. 

Held in the middle of the COVID-19 pandemic and a deep economic crisis, it nonetheless produced a record turnout, thanks in part to the decision by many states to expand mail-in ballots and early voting, and to nationwide polarization that sparked both supporters and opponents of Trump to flock to the polls. 

More than 159 million people cast ballots, a dramatic 15 percent increase over the 2016 election, with Biden eventually securing a margin of seven million in the popular vote, and a clear majority of the all-important electoral college vote.

On Election night, however, media accounts claimed that Latino voters – and to some extent, African American men – had made a significant and unexpected shift toward Trump. 

According to those accounts, Biden captured “only” 65% of Latino votes to Trump’s 32%, a smaller margin than Hillary Clinton had achieved four years earlier. This was proof that Latinos are more conservative than many Americans realized, the media claimed, and some stated  there is no such thing as a cohesive “Latino vote.”                                                                                                       

This narrative, in our opinion, is a false one.  It lacks both a deep analysis of this year’s overall Latino vote and any context of historical trends within the Latinx community. It was largely based on flawed exit poll data and on preliminary vote returns from two small geographical areas of the country, Miami-Dade County and the Rio GrandeValley of Texas. 

Most importantly, the narrative completely ignored the main story of this election – an unprecedented increase in turnout by Latinos – and it obscured the pivotal role those voters played in delivering key battleground states for Biden, specifically Arizona, Wisconsin, and Pennsylvania. 

Furthermore, it failed to recognize a demographic trend that has become more pronounced with each passing year: the emergence of a new Latino “Brown Belt” in the United States, one that is centered in California and the Southwestern states, and which is on the cusp of transforming national politics in the coming decades. 

The much-cited 65%-32% split among Latino voters for president came from one place – the National Exit Poll that the Edison Research firm has conducted for the major broadcast networks every election since 2004. That survey has been repeatedly criticized in the past because it undersamples Spanish-speaking voters, those with less education, and those who live in majority Latino precincts and because it oversamples precincts in Cuban American communities, which historically have tended to vote more for Republicans. 

“It is an example of the worst possible poll that has somehow become believed as ‘official’ or ‘fact’ by folks in the media,” UCLA political scientist Matt Barreto says. “Within political science, it will be hard to find a single Ph.D. who believes anything in the [Edison] exit poll as representative.”

 Barreto, a co-director of the UCLA Latino Politics and Policy Initiative, also worked as a consultant for the Biden campaign this year. He insists a more accurate measure of voting patterns among Latinos, African Americans, Asian and Native Americans is the American Election Eve Poll (AEEP), which interviewed more than 15,000 voters nationwide, including more than 5,000 Hispanics. The latter poll found the Biden-Trump split among Latinos to be 70%-to-27%, not 65-to-32.

 Yes, Barreto acknowledges, there was a small percentage increase for Trump compared to his 2016 race against Hillary Clinton, but that ignores the far more important result of this election – a huge increase in Latino turnout that provided a much bigger margin in actual votes for Biden among Latinos.

More than a month after the election, no one knows for sure the increase in Latino voters in 2020. We probably won’t have precise numbers until next year, when the Census Bureau releases the survey of voters it conducts following each major election. But we have a pretty good idea of the range.

The Edison poll, the one used by all the major media firms, reported that Latinos comprised 13% and African Americans 12% of the 159 million Americans who voted. If true, that would mean 20.6 million Latinos cast ballots  – an astonishing increase of 8 million over the 12.6 million the Census Bureau says voted in 2016 – a 63% jump. But like everything else about the Edison poll, those turnout numbers are highly suspect. “Random guesses,” Barreto calls them.

He and the other UCLA experts believe that Latinos more likely represented 10.5% of the electorate this year. That would translate into 16.7 million ballots cast by Latinos, an increase of nearly 33%, or 4.1 million voters compared to four years ago. 

Barreto’s figure is much lower than what Edison estimates, but it still reflects an extraordinary increase. Remember, turnout among all Americans this year jumped by 15% – and that was considered historic – so the increase in Latino turnout was truly off the charts, whether you go by Edison’s calculation or by that of UCLA’s Latino Initiative.

The basic contours of these turnout figures are inescapable. After decades of political experts talking about the growing Latino vote, this year it actually happened. Hispanic voters felt compelled like never before to go to the polls, because COVID-19 felled them more than other groups, because of the constant demonization of Latinos by President Trump, because of his barbaric family separation policies, or because of the threats to health care. One thing is sure, neither the Democratic nor Republican Parties will underestimate nor ignore Latino voters from now on.

Much of the attention on the supposed “shift” toward Trump among Latino voters has focused on the rural Rio Grande Valley of Texas and South Florida, where Trump did indeed make real inroads in actual votes from largely Latino districts. But such shifts were hardly a nationwide trend.

In their post-election analysis, however, Barreto and his UCLA colleagues did not simply rely on the results of the AEEP or a few select geographical areas, they also studied actual raw vote results from election precincts around the country, using a statistical method called “ecological inference” to calculate how both presidential candidates fared with Latinos.

“The real point is that Latinos voted about 70-30 in favor of Biden,” Barreto said, “and with 4 million new additional Latino votes, they provided a critical increase in net votes for Biden in key states.”

In battleground states like Pennsylvania, Wisconsin, and Arizona, Biden’s support among Latinos was actually far higher than the national average. Here are some examples of the percentages Biden received in largely Latino precincts of some key counties in those states in the UCLA analysis:

In Arizona, where Biden prevailed by just 12,000 votes, UCLA estimates Latino voters delivered a net margin to him (after subtracting pro-Trump Latino votes) of 312,000.

In Pennsylvania, where Biden won by 82,000, the net Latino vote for him was 116,000.

In Wisconsin, where he won by 20,500, the net Latino vote for him was 44,300.

Barreto estimates the net vote margin Biden received nationwide from Latinx voters was nearly 7.3 million, slightly more than his overall total popular vote margin of 7 million.

It is becoming increasingly clear that a huge swath of the West and the Southwest of the United States keeps shifting more Democratic in its political leanings, and much of that is due to the growing concentration of Latinos throughout the region, which has already led to a modern-day “Brown Belt,” one that echoes the historic Black Belt of African-Americans in the post-Civil War South. No state better represents this change than Arizona, which now has two Democrats in the U.S. Senate, and which, before delivering a majority for Biden this year, had voted for a Democrat for president just once since 1952.

The Brown Belt stretches from California through New Mexico, Nevada, Arizona, and into Texas, roughly the old Mexican territories the U.S. annexed after the U.S.-Mexico War. In those five states, Latinos already comprise more than 35% of all residents. Two of them, California and Texas, happen to be the nation’s biggest in total population. The 31 million Latinos who inhabit these five states already amount to half of the nation’s total estimated Hispanic population of 60.6 million. 

This false narrative of a sudden Latino shift for Trump also ignores the rich and complex history of the Latinx community and profound trends laid bare by this election.

The fact is that Latino support for Trump was about par for Republican presidential candidates in elections over the past thirty years. The Republican share of votes has varied from a low of 27% for Mitt Romney in 2012, to a high of 44% for George W. Bush in 2004. Even John McCain, when he ran against Obama in 2008, got a similar percentage of the Latino vote as Trump did this year ­– and that’s if you rely on the Edison Poll. The highpoint for Republicans, the Bush victory in 2004, is actually the last time a candidate of their party got a majority of the popular vote.

Finally, some observers are even questioning whether a Latino community actually exists. In our view, Latinx identity in the U.S. is a social construct, created both by the dominant society that needed to define and categorize an “other,” and by this marginalized group itself, organically from the ground by disparate Latin American migrant groups who were forced to unite in order survive and whose children gradually intermarried to create a new social construct – the Latino in America.

Latinos have been the Third Force in American politics for decades now. They largely vote Democratic, but a significant portion is susceptible to shifts if Republicans or other political leaders address even a few of their concerns. A small minority of Latinos have always been drawn to right-wing populism, to national chauvinism, and even to racist views.

The fixation on what Latinos did in this election, however, ignores the biggest question of all, which very few political observers have dared to tackle: why did 58% of white Americans, including 55% of white women, vote for a man for president who was so patently unqualified to perform the basic duties of that office?  Rather than incorrectly claiming that people of color showed unprecedented support for Trump, that is the real story we should be examining.