This is an updated version of this brief, taking into account new exit poll results through November 11, 2020.
When Barack Obama won his second presidential term eight years ago, the exit polls clearly showed that he benefitted from what I called the “new American mainstream”—the growing voter blocs of youth, people of color, and the college-educated. That calculus changed in 2016, when Donald Trump shocked the political world with a “demographic blowback” among, largely, older whites without college degrees that took him to the White House.
Now, with Joe Biden declared the winner of the 2020 election, his victory over Trump reflects elements of both of these demographic constituencies. And if exit polls are to be believed, several voting blocs comprised of white and older Americans contributed to his success.
This is not because white voters suddenly flocked to Biden and the Democrats—although in some cases they did. Nor is it the case that Democrats should abandon their growing “new American mainstream,” because they absolutely need them to succeed in the future. Instead, 2020’s exit polls indicate that in key Rust Belt and Sun Belt battlegrounds, Biden benefitted from lower Republican margins among some of the groups that handed Trump his 2016 victory.
National polls show lower white and older support for Trump
Exit polls released by the national election consortium Edison Research (accessed November 11, 2020) allow for national- and state-level comparisons with those from 2016. Figure 1 shows the shifts in Democratic minus Republican (D-R) voter margins for racial groups. (The D-R margin is defined as the percent voting Democratic minus percent voting Republican. A positive value shows a Democratic vote advantage whereas a negative value indicates a Republican vote advantage.)
While whites continued to favor the Republican candidate in 2020—as they have in every presidential election since 1968—it is notable that this margin was reduced from 20% to 17% nationally. At the same time, the Democratic margins for each of the major nonwhite groups was somewhat reduced. The Black Democratic margin—while still high, at 75%—was the lowest in a presidential election since 2004. The Latino or Hispanic and Asian American Democratic margins of 33% and 27% were the lowest since the 2004 and 2008 elections, respectively. These shifts do not apply to all states, and are not applicable to most battleground states where voters of color were crucial to Biden’s win
It is clear that white voting blocs start at different levels of Democratic or Republican support. In fact, there was a modest decline in Republican support in a key Trump base: white men without college educations. This group showed a reduced Republican advantage from 48% to a still sizeable 42% between 2016 and 2020.
Yet among white voters with college educations, there were notable shifts in Biden’s direction. White male college graduates reduced their support for Trump from 14% to 3%. At the same time, white female college graduates boosted their Democratic support from 7% to 9% nationally. Moreover, in key battleground states, white female college graduates generally registered greater support for Biden in 2020 than they did for Hillary Clinton in 2016.
The other noteworthy change that impacted the 2020 election concerns D-R margin shifts for different age groups. It should not be surprising that after a summer of protests and activism among young people, persons age 18 to 29 registered a rise in Democratic support between 2016 and 2020, from 19% to 24%. Some of this is attributable to the increased Democratic-leaning nonwhite presence in this age group, due to the nation’s changing demographic make-up.
There was also less Republican support among older segments of the population: ages 45 to 64 and ages 65 and older. This decrease in Republican support was even more pronounced for 45- to 64-year-old whites: from 28% in 2016 to 19% in 2020 (see downloadable Table A). This is evident in several battleground states.
Lower white support for Trump in Pennsylvania, Michigan, and Wisconsin
The three crucial northern battlegrounds of both 2016 and 2020 are the relatively “white” states of Pennsylvania, Michigan, and Wisconsin. Each one flipped for Biden this election.
Exit polls in all three states indicate that more Democratic-favorable (or less Republican-favorable) margins among different white blocs between 2016 and 2020 contributed to Biden’s wins. See Figure 4 and downloadable Table B.
Beginning with Pennsylvania, it is clear that white college graduates—both male and female—voted more Democratic in 2020 than in 2016. Particularly noteworthy is the change for white male college graduates (from a 17% Republican advantage to just 2%), while white female college graduates increased their Democratic support from 14% to 19%.
Pennsylvania’s seniors, too, registered a smaller Republican margin, while white noncollege men and women remained fairly strongly Republican. But the shift among white college-educated Pennsylvanians toward Democrats, strong nonwhite (especially Black) support, and greater support among young people were enough to seal Biden’s win.
Michigan’s flip to Biden’s column in 2020 is even more dependent on higher D-R margins among each of the white voting blocs. An especially strong shift in Democratic advantage occurred for white female college graduates, from 6% in 2016 to 20% in 2020. Also, large declines in Republican advantages were evident for white college and noncollege men. Among the latter, the Republican advantage fell from 44% to 30%. Moreover, voters in older age groups flipped from a Republican advantage in 2016 to a Democratic advantage in 2020. Along with the strong Democratic support among Michigan’s Black voters, the shifts in the state’s white voting blocs helped Biden considerably.
Wisconsin, the last of the northern battleground trifecta, also showed the same or increased D-R voting margins for all white voting blocs. This was the case for white noncollege men and women. The former group lowered its Republican advantage from 40% in 2016 to 27% in 2020, while the latter lowered from a Republican advantage of 16% in 2016 to 5% in 2020. White female college graduates registered 2020’s highest Democratic margin, 23%, with white male college graduates shifting from an even Republican-Democratic split in 2016 to a 3% Democratic advantage in 2020. Biden also benefitted from higher Democratic margins among voters age 18 to 29 and 40 to 64, as well as from strong support among the state’s Black population. But as with Pennsylvania and Michigan, a more Democratic-leaning white electorate in this whitest of the three states contributed to his win.
Whites made Biden competitive in racially diverse Sun Belt states
As the final votes were being counted, three Sun Belt states remained competitive between Biden and Trump: Arizona, Georgia, and Nevada. While their final outcomes also depended on nonwhite racial groups, white voting blocs in these states shifted since 2016 in ways that benefitted Biden. See Figure 4 and downloadable Table B.
Take Arizona. It is a state that has not voted for a Democratic presidential candidate since 1996. While rapidly diversifying, its older white population has leaned heavily toward Republicans. This time was different; white college graduate women and men flipped sharply toward Democrats, from 2016 Republican advantages of 2% and 12%, respectively, to 2020 Democratic advantages of 15% and 3%. Likewise, white noncollege men reduced their Republican support from 28% to 10%. In addition, Arizona’s senior population flipped from Republican support to even Democratic-Republican support.
These shifts, as well as increased Democratic support among 18- to 29-year-olds and continued Democratic support from the state’s Latino or Hispanic voters, contributed to Biden’s vote gains in Arizona.
Georgia, a longtime deep red Republican state, has been inching toward “battleground” status due to its large and growing Democratic-leaning Black population. Yet its strong white Republican margins have led to GOP presidential wins since 1996. This year, those white Republican margins were reduced enough to make the state competitive.
The biggest swings from Georgia’s 2016 results were reduced Republican support among white college-educated men and women. The former saw its Republican margin shrink from 55% to 12% between 2016 and 2020; the latter’s shrank from 29% to 10%. Reduced Republican margins were also evident for voters age 45 and older. These shifts, along with still solid support from Black voters, were instrumental in Biden’s strong showing in Georgia.
Nevada has voted for Democratic candidates since Obama first ran in 2008. As one of the most rapidly growing and racially diverse battleground states, it was anticipated to be in Biden’s column well before the election. However, the late reporting of votes in Clark County made it competitive to the end. White female college graduates did flip from Republican to Democratic between 2016 and 2020, and white men—both college and noncollege—showed lower Republican margins. These are important for Biden, because Nevada exit polls showed 2016 to 2020 declines in Democratic support from Black, Asian American, and especially Latino or Hispanic voters there.
Several other states were expected to come close to flipping to Biden in 2020, but their final votes favored Trump. Among these are North Carolina, Florida, and Texas. Each showed lower Republican margins or flips to Democratic support for most or all white voting blocs, especially white female college graduates. (See relevant statistics in downloadable Table B.) However, they were not able to overcome shifts in other groups (such as lower Democratic support among Latino or Hispanic voters in Florida and Texas) to bring Biden a victory.
What these shifts mean for future elections
The exit polls and results from this year’s presidential election paint a somewhat different picture than the previous two races. After Obama’s second victory in 2012, Democrats were touting a voter constituency made up of young people, diverse voters, and college-educated whites that they felt would provide them solid support for several elections to come. It even prompted Republicans to issue an “autopsy” urging the inclusion of a wider voter base. Yet after Trump’s 2016 victory with strong support from older, less urban, and noncollege whites, many Republicans stayed onboard their earlier train.
In retrospect, it seems that both the 2012 Obama coalition and the 2016 Trump coalition overperformed in those elections. The 2020 results suggest neither party can rely solely on those particular sets of voters. As I have written, there is no doubt that changing demographics—especially rising diversity—should benefit Democrats in the long run (see also the latest States of Change report).
But in the interim, the results of the 2020 election make plain that both parties need to address the interests of a coalition made up of all of these groups. The Trump presidency did not do this—perhaps a Biden presidency can.
Report Produced by Brookings Metro