We know that President Biden won the 2020 election (regardless of what former President Donald Trump and his allies say). We just haven't had a great picture of how Biden won.
That is until Wednesday, when we got the clearest data yet on how different groups voted, and crucially, how those votes shifted from 2016. The Pew Research Center just released its validated voters' report, considered a more accurate measure of the electorate than exit polls, which have the potential for significant inaccuracies.
The new Pew data shows that shifts among suburban voters, white men and independents helped Biden win in November, even while white women and Hispanics swung toward Trump from 2016 to 2020.
To compile the data, Pew matches up survey respondents with state voter records. Those voter files do not say how a person voted, but they do allow researchers to be sure that a person voted, period. That helps with accuracy, eliminating the possibility of survey respondents overreporting their voting activity. In addition, the Pew study uses large samples of Americans — more than 11,000 people in 2020.
It's a numbers-packed report, but there are some big takeaways about what happened in 2020 (and what it might tell us about 2022 and beyond):
Suburban voters (especially white suburban voters) swung toward Biden
Suburban voters appear to have been a major factor helping Biden win. While Pew found Trump winning the suburbs by 2 points in 2016, Biden won them by 11 points in 2020, a 13-point overall swing. Considering that the suburbs accounted for just over half of all voters, it was a big demographic win for Biden.
That said, Trump gained in both rural and urban areas. He won 65% of rural voters, a 6-point jump from 2016. And while cities were still majority-Democratic, his support there jumped by 9 points, to 33%.
Men (especially white men) swung toward Biden
In 2020, men were nearly evenly split, with 48% choosing Biden to Trump's 50%. That gap shrank considerably from 2016, when Trump won men by 11 points. In addition, this group that swung away from Trump grew as a share of the electorate from 2016 — signaling that in a year with high turnout, men's turnout grew more.
White men were a big part of the swing toward Biden. In 2016, Trump won white men by 30 points. In 2020, he won them again, but by a substantially slimmer 17 points.
In addition, Biden made significant gains among married men and college-educated men. All of these groups overlap, but they help paint a more detailed portrait of the type of men who might have shifted or newly participated in 2020.
However, we can't know from this data what exactly was behind these shifts among men — for example, exactly what share of men might have sat on the sidelines in 2016, as opposed to 2020.
Women (especially white women) swung toward Trump
The idea that a majority of white women voted for Trump quickly became one of the 2016 election's most-cited statistics, as many Hillary Clinton supporters — particularly women — were outraged to see other women support Trump.
While that statistic was repeated over and over, Pew's data ultimately said this wasn't true — they found that in 2016, white women were split 47% to 45%, slightly in Trump's favor but not a majority.
This year, however, it appears that Trump did win a majority of white women. Pew found that 53% of white women chose Trump this year, up by 6 points from 2016.
This support contributed to an overall shift in women's numbers — while Clinton won women of all races by 15 points in 2016, Biden won them by 11 points in 2020. Combined with men's shifts described above, it shrank 2016's historic gender gap.
Notably, the swing in white women's margin (5 points altogether) was significantly smaller than white men's swing toward Biden (13 points altogether).
Hispanic voters swung toward Trump
Trump won 38% of Hispanic voters in 2020, according to Pew, up from 28% in 2016.
That 38% would put Trump near George W. Bush's 40% from 2004 — a recent high-water mark for Republicans with Hispanic voters. That share fell off substantially after 2004, leading some Republican pollsters and strategists to wonder how the party could regain that ground. Trump in 2016 intensified those fears, with his nativist rhetoric and hard-line immigration policies.
There are some important nuances to these Hispanic numbers. Perhaps most notably, there is a sizable education gap. Biden won college-educated Hispanic voters by 39 points, but the Democrat won those with some college education or less by 14 points.
That gap mirrors the education gap regularly seen in the broader voting population.
Unfortunately, Pew's sample sizes from 2016 weren't big enough to break down Hispanic voters by gender that year, so it's impossible to see if this group's gender gap widened.
Nonwhite voters leaned heavily toward Biden
Unlike white and Hispanic voters, Black voters didn't shift significantly from 2016. They remained Democratic stalwarts, with 92% choosing Biden — barely changed from four years earlier.
Nearly three-quarters of Asian voters also voted for Biden, along with 6 in 10 Hispanic voters and 56% of voters who chose "other" as their race. (Those groups' sample sizes also weren't big enough in 2016 to draw a comparison over time.)
2018 trends stuck around ... but diminished
In many of these cases where there were substantial shifts in how different groups voted, they weren't surprising, given how voters in the last midterms voted. For example, white men voted more for Democrats in 2018 than they did in 2016, as did suburban voters.
A small bump for Biden with white, non-college voters
In 2016, Clinton won 28% of this group that is a key part of Trump's base (and that makes up more than 40% of the electorate). Biden improved slightly upon Clinton's performance here, winning 33% of this group.
However, Trump did not lose ground with this group, holding steady with roughly two-thirds of these voters. That's because the share of this group that may have voted third party appears to have shrunk.
Biden made gains among white Catholics
Pew's data shows that Trump won a majority of this group, as he did in 2016. However, Biden — himself a Catholic — gained considerably among them, compared with Clinton's 2016 performance. In 2016, she won 31% of this group, according to Pew. Biden, meanwhile, won 42% of white Catholics in 2020.
Meanwhile, white evangelicals — a cornerstone of the GOP's base for decades — seem to have voted even more strongly Republican. Pew's data shows 84% of this group voting for Trump in 2020, up from 77% in 2016.
A small gain among younger voters for Trump
The basic pattern of older voters tending more Republican and younger voters tending more Democratic held in 2020. In addition, the silent generation, boomers and Generation X all had roughly the same partisan voting patterns as they did in 2016. However, millennials showed a slightly bigger lean toward Trump than they did in 2016, with his share among that generation moving from 31% to 39%.
Also notably, for the first time, the silent and boomer generations together constituted a minority of voters, with Gen X, millennials and Generation Z accounting for a majority.
What it means for 2022
The data signals that Democrats' strength with Hispanic voters has eroded, but that the party succeeded in making further inroads in the suburbs, including among suburban whites.
It suggests that these groups, already major focuses for both parties, will continue to be so in 2022, with Republicans trying to cement their gains among Hispanics (and regain suburban voters), while Democrats do Hispanic outreach and try to hold onto the suburbs.
However, it's hard to project much into the future about what voters will do based on the past two elections because of their unique turnout numbers.
"It's hard to interpret here, because 2018 was such a high turnout midterm election, and then our last data point, 2014, was a historically low turnout midterm election," said Ruth Igielnik, senior researcher at Pew Research Center.
AILSA CHANG, HOST:
Nearly eight months after the election, we are still learning more about why Joe Biden won the presidency, how Donald Trump lost, and what it might all mean heading into next year's intense battle for control of Congress. There is new data out today explaining how different demographic groups in the country voted, data that is considered more solid than last year's exit polls and their notorious flaws. NPR political correspondent Danielle Kurtzleben has been digging into the new numbers and joins us now.
DANIELLE KURTZLEBEN, BYLINE: Hey, Ailsa.
CHANG: So this new information comes from the Pew Research Center. What are we learning here that we didn't necessarily know back in November?
KURTZLEBEN: We're learning more exact numbers and magnitudes of how things changed. One big thing is how much men were a big part of swinging the election to Biden and especially white men. Trump, you'll remember won white men by a lot, by 30 points in 2016. And he still won them last year, but now we can pretty definitively say that his margin shrank nearly in half to 17 points. And what's really important here is that men's turnout grew quite a bit in an already high turnout year.
But important things we can't tell from this data are why these shifts happen. Was this largely men, for example, who sat out 2016 or people who changed their votes from then? And also, why that happened - we still don't know that. One more thing I want to get out here, though, is suburban voters. We talked so much about them ahead of November of last year. Trump won them by 2 points in 2016. Biden won them by 11 points last year. That's more than half of voters that were in the suburbs. And so they were - they ended up being very crucial to Biden's win.
CHANG: Well, we said that this data is considered more solid than other information we had, like exit polls. Can you just briefly explain why?
KURTZLEBEN: Yes. So Pew does something special here. They match up their survey respondents with state voter records. That's public data. That data doesn't say how a person voted, but it says that a person voted. So that helps with accuracy. It eliminates the possibility of a lot of survey respondents overreporting their voting activity, which does happen in some surveys. In addition, the Pew study has really big samples. More than 11,000 people were a part of this in 2020, so that shrinks down the margins for error.
CHANG: And I understand that there are some key areas where Trump improved from 2016 to 2020, even though - to remind everyone - he did lose the Electoral College, and he did lose the popular vote by a wider margin. What were those areas of improvement?
KURTZLEBEN: You know, one is white women. Originally in 2016, it looked like Trump won a majority of white women. Now, to be clear, he did not. He won 47%, which was still a plurality, slightly more than Hillary Clinton. But he did better in 2020, according to Pew. He won just over half. He did win a majority, 53% of white women voters last year. Now, overall, Biden won women of all races, but by a few points less than Clinton did. With Biden's gains among men, that means the historic gender gap - excuse me - in 2016, it shrank a little bit last year.
But one other group I want to get to is Hispanic voters. Trump won 38% of Hispanic voters according to Pew, and that's really close to the recent high water mark that George W. Bush hit in 2004. He got 40% of Hispanic voters. So while Biden did still win Hispanics overall, this data shows that Trump made some gains and also a big education gap, that Hispanic voters without a college degree voted more for Trump than Hispanic voters with a college degree.
CHANG: And real quick, Danielle, what does this data tell us about the next election, which will be all about the control of Congress?
KURTZLEBEN: I mean, it really signals that these groups that we've been paying so close attention to, Hispanic voters and suburban voters, they have been important. They continue to be important. They continue to be swingy. They continue to be groups that parties are going to go after.
CHANG: That is NPR's Danielle Kurtzleben.
Thank you, Danielle.
KURTZLEBEN: Thank you. Transcript provided by NPR, Copyright NPR.