As results rolled in from Florida on election night, one thing became clear almost immediately: Joe Biden was underperforming in Miami-Dade County, home to most of Florida’s Cuban community. Later in the evening, a similar story emerged from a handful of heavily Mexican American counties along Texas’s southern border.
Biden ultimately won most of these counties in both Florida and Texas, but by much slimmer margins than Hillary Clinton had four years earlier. However, he did end up bringing Arizona — another state with many Latino voters — into the Democratic column for the first time in a presidential election since 1996.
What divided the electorate in 2020 | FiveThirtyEight Politics Podcast
At the same time — although the eventual results were not at all apparent on election night — Biden won three traditionally blue states that Clinton had lost in 2016, Michigan, Pennsylvania and Wisconsin, in part by improving on her performance among white voters with a four-year college degree.
Of course, the story of why Biden won back the “blue wall” but didn’t win Florida or Texas is much more complex than his gains among white voters with a college degree and his misses with some Latino voters. As we’ve written before, no one group “swings” an election. Nevertheless, it’s worth unpacking how these two demographic groups voted in 2020, as they underscore larger geographical and educational divides in our country.
For one thing, the political gap between rural and urban America appears to have been even more pronounced this year than it was in 2016, which may explain some of Trump’s success among Latino voters. Education was once again a clear dividing line, especially among white voters, with white voters with a college degree moving toward Biden while white voters without a college degree remained largely in Trump’s column.
This is just the beginning of our analysis of how Americans voted in 2020 — over the coming weeks and months, we’ll have more deep dives into various corners of the American electorate as more data becomes available. In this story, we’re zooming in on county-level results in a handful of key swing states to help us understand how demographic divisions in pivotal parts of the country helped — and hurt — the two major candidates.1 So let’s start with what we know about the voting patterns in heavily Latino counties in Arizona, Florida and Texas, and what that tells us about the country’s urban, suburban and rural divide.
What Latino voters tell us about America’s urban-rural divide
In the lead-up to the election, there were plenty of signs that Biden’s support among Latino voters in key swing states might be weaker than Clinton’s in 2016, but some of the shifts wound up being very large. In Florida’s Miami-Dade County, for instance, which is 68 percent Hispanic, Trump narrowed his deficit by 22 percentage points between 2016 and 2020; in Texas’s Starr County, which is 99 percent Hispanic, Trump improved by a stunning 55 percentage points.
However, as the chart below shows, Trump’s gains among Latino voters were hardly universal. In fact, the places where Trump appears to have gained the most support were largely in rural areas or among more conservative Latino voters like Cuban Americans. In suburban and urban areas, the story was much more mixed. (And, to be clear, Biden still won the overwhelming majority of Latino votes.)
One important factor to keep in mind here — which is partially why some of these shifts toward Trump seem so pronounced — is that Trump did really poorly with Latino voters in 2016. According to pre-election surveys, he won just 18 percent of Latino voters in 2016 but 27 percent this year, putting him back in the territory of other recent Republican presidential nominees.
Additionally, part of what we’re seeing here isn’t necessarily something unique to Latino voters at all, but an extension of America’s growing urban-rural divide. As Bernard Fraga, a political science professor at Emory University who studies racial and ethnic politics, told us, “Biden was never expected to make gains in rural populations and he didn’t really try. … So if Trump is the candidate that people in rural America identify with more, that’s going to work with Latinos just like it works with other groups.”
One notable exception to this is Miami-Dade, which is a largely urban area. But Sergio Garcia-Rios, a professor of political science at Cornell University who studies Latino identity, pointed out that Trump may have helped improve his standing among Latinos there by recognizing their diversity and focusing on a few groups where he had the potential to make inroads — like Cubans in South Florida. The Trump campaign made outreach to Cuban voters a priority early on, emphasizing Trump’s tough record on Cuba while also drumming up the (inaccurate) idea that Biden was a “Trojan horse” bringing socialism to the U.S. It’s a message designed especially for voters who trace their roots to Cuba and other countries with histories of socialist dictatorships. And that tack seemed to pay off, especially in Miami-Dade County, where more than half of the Hispanic population is of Cuban origin. Trump still lost the county, but this time just by 7 points after losing it by 29 points in 2016. He made substantial gains across the whole county, but especially in areas heavily populated by Cuban Americans.
Notably, though, Trump didn’t make quite as dramatic inroads in Osceola County, Florida — part of the Orlando metro area — which is heavily Puerto Rican, though he still gained there, losing it by 14 points compared to 25 points in 2016. Dario Moreno, a politics professor at Florida International University, told us that Trump’s improved margin there may have been due to economic anxiety and fear that another wave of COVID-19 restrictions on businesses would hurt an area heavily dependent on tourism.
In fact, in an election where the focus was more on COVID-19 and the economy, it’s not that surprising that Trump returned to a similar level of support among Latinos as other recent Republican presidential candidates, according to Garcia-Rios, as concern about jobs and business lockdowns in particular may have nudged some conservative voters toward Trump.
“We might … disagree about whether Trump’s economy has really helped minorities, but some of them do believe that for their jobs, for their families, Trump is the answer,” Garcia-Rios said, pointing to survey after survey that showed Latinos’ top election issues were COVID-19, health care, the economy and jobs. Other issues like immigration and discrimination, where Trump has a long history of race-baiting and trying to clamp down on migration, were a lower priority this year. “We do a disservice when we paint the Latino vote with a broad brush,” Garcia-Rios added.
There is also the question of whether Biden and the Democrats’ Latino outreach was sufficient. In the lead-up to the election, the campaign’s lack of a cohesive strategy for reaching Latino voters was heavily criticized, and it may have even contributed to some smaller losses among urban and suburban Latinos. For instance, whereas Biden improved over Clinton in the whiter areas on the periphery of Harris County, Texas (the home of Houston), Trump notably decreased his deficits in heavily Hispanic precincts toward the more central areas of the county, which has a large Mexican-American population. Similarly, Trump did somewhat worse this time around in Maricopa County, which is home to the city of Phoenix and about 60 percent of Arizona’s votes, but he did perform slightly better in many Latino-heavy precincts, including the 7th Congressional District that is almost 60 percent Mexican American.
Polls were off for the 2020 election, but not by much | FiveThirtyEight
It’s worth reiterating that Biden still won far more votes in Latino sections of Houston and Phoenix than the president did. But that success may have little to do with Biden’s campaign and more to do with local organizers. This, experts told us, underscores that Democrats really can’t take Latino voters for granted — and a lack of investment in Latino outreach and organizing could hurt them in the long term.
Lisa García Bedolla, a political scientist at the University of California, Berkeley, pointed out that Latino community organizers have been working for more than a decade in Arizona to mobilize Latino voters politically, which almost certainly benefited Biden and Democrats this year. “Demography isn’t destiny when it comes to Latino voters,” she said. “The explanation for why Latinos voted the way they did this year is about history, it’s about geography and it’s about where coalition-building and community organizing has been happening for years — not just what a presidential candidate did or didn’t do this cycle.”
Education remained a dividing line, driving white voters further apart
In the three “blue wall” states that Biden recaptured in 2020, meanwhile, voters were deeply divided by education. The education split has been especially significant among white voters, and this rift appears to have widened as Trump lost ground in Michigan, Pennsylvania and Wisconsin, especially in areas where many white voters have four-year college degrees.
As the chart below shows, the larger a county’s share of non-Hispanic white voters with a college degree, the worse Trump tended to perform compared to 2016. Conversely, even though pre-election polls suggested that Biden might be poised to make some gains among white voters without a college degree, that didn’t happen. Trump held his ground and even made further inroads in predominantly white counties with lower levels of education — but not enough for him to win.
And across the most highly-educated counties in these states — which include some populous suburban areas — Biden improved substantially on Clinton’s margins. In Wisconsin, for instance, Biden won by 0.6 points statewide after Clinton lost by a similarly slim margin in 2016. And part of that came down to Trump’s deterioration in the traditionally Republican Milwaukee suburbs. Trump still won Waukesha County, by far the largest of the GOP-leaning Milwaukee suburban counties, by 21 points, but that marked a 6-point drop from his 2016 advantage there. And that was in part because of how white voters with a four-year college degree broke (around 40 percent of Waukesha’s population is white and college educated). In other words, areas that were once pretty Republican, like Waukesha, still moved away from Trump because of how white voters with a college education trended.
The same was true in Michigan and Pennsylvania. For instance, in Oakland County, a populous Detroit suburb with a hefty share of educated white voters, Trump’s deficit increased from 8 points in 2016 to 14 points in 2020. Similarly, Trump lost Chester County, a suburb of Philadelphia, where nearly half of the population is made up of white voters with a college degree, by 9 points in 2016 and 17 points this year.
Ashley Jardina, a political science professor at Duke University who studies white identity politics and the education divide, said that white college-educated voters might be especially likely to have been alienated this year by Trump’s handling of the pandemic — potentially leading them to abandon him in even greater numbers than in 2016. “A lot of college-educated whites were not that turned off by Trump’s rhetoric about women or people of color,” she said. “But we’re now in the middle of a pandemic and if you’re a college-educated person who cares about good government, Trump’s handling of the response might really matter to you.”
On the other hand, counties with larger shares of less educated white voters largely stuck with Trump. But most of these counties lie, not coincidentally, in rural areas, where Trump’s electoral strength largely held or improved regardless of its racial or ethnic makeup. Take Clark County in central Wisconsin, where the vast majority of the population is white and doesn’t have a college degree. Trump won the county by 32 points in 2016, but expanded his edge to 37 points in 2020. The same pattern popped up in places similar to Clark, such as Clare County in central Michigan and Mifflin County in central Pennsylvania.
Part of what is happening, according to Arlie Hochschild, a sociologist at the University of California who has written extensively about conservative voters, is that many less educated white voters have come to see Trump as their champion. “They feel that Trump is making them great again — their social class and their identity as whites,” she said. “Many of them feel that as white [people], they’re discriminated against.” She added that even if Biden might have personally appealed to those voters, it might not have been enough to overcome their suspicion that the Democratic Party as a whole was hostile to their worldview.
Importantly, Trump’s gains among white voters without a college degree were less substantial than his losses among educated white voters, and that appears to have cost him in these three states. This was most stark in Wisconsin, where Trump’s margin improved in 39 of the state’s 72 counties, but fell in 31 and didn’t change in two. The counties where he lost ground tended to be bigger and more well-educated, while the ones where he gained were generally smaller and less well-educated. In aggregate, these shifts added up to a narrow loss in Wisconsin for Trump in 2020 instead of the close win he achieved in 2016.
And in general, that pattern illustrates a broader challenge for Republicans: White voters without a college degree are shrinking as a proportion of the broader population, while white voters with a college degree are growing as a share of the electorate.
But Biden’s dramatic gains with white voters with a college degree may not represent a permanent realignment, according to Jardina. “I don’t think [these voters’] opposition to the Republican Party is crystallized, depending on what happens after Trump,” she said. More than anything, she added, the education gulf among white voters this year may illustrate the futility of Democrats’ attempts to bring the white working class back into their coalition. “Democrats have focused on winning back white working-class voters and sort of taken whites with a college degree as a given, which strikes me as a strange, Pollyannaish strategy,” she said. “[White working-class voters] are not their people now.”
As we said at the outset, this is just the beginning of our analysis of how Americans voted in 2020 but it’s worth underscoring what’s surprising about these initial findings.
Even though Trump’s overperformance among specific groups of Latino voters was one of the big stories coming out of election night, what we can see in the data isn’t actually all that unexpected. There were clear signs all along that Biden might be weaker than Clinton among Latinos voters, and the places where Trump appears to have gained support — in rural areas and among more conservative Latino voters like Cuban Americans — just speaks to the diversity of a group that should never have been regarded as politically homogenous to begin with.
On the other hand, the widening of the education gap — and the emergence of geography as another important political divide — might be bigger takeaways, given where the polling stood going into the election. Biden’s gains among white voters without a college degree simply didn’t materialize, while Trump continued to make inroads with less-educated white voters — although it wasn’t enough to win him the election. But importantly, both groups offer a first look into what might be one of the defining trends of the 2020 election: the widening political chasm between urban, suburban and rural America.