The growing number of Democrats running for president in 2020 are a diverse group of candidates. But they won’t be as diverse as the people voting for them.
According to projections from the Pew Research Center, a higher number of those who will cast votes for president next year will be younger than their counterparts in past presidential election years. There will be more eligible Latinx than African-American voters. And because of an increased number of naturalized citizens, one in 10 voters will have been born outside the United States.
Nonwhites will account for a third of eligible voters — their largest share ever — driven by long-term increases among certain groups, especially Hispanics. At the same time, one-in-ten eligible voters will be members of Generation Z, the Americans who will be between the ages 18 and 23 next year. That will occur as Millennials and all other older generations account for a smaller share of eligible voters than they did in 2016.
These projections from Pew—long considered the gold standard of polling research—are just that, projections based on demographic trends. But if the 2018 midterm election taught us anything, it’s that conventional wisdom about certain voting habits and voting trends from past elections don’t necessarily apply anymore. The old polling models of who votes and in what numbers need a major overhaul.
Too bad pundits—and some candidates—haven’t learned that lesson.
It’s perfectly fine and even logical to look at traditional factors when trying to decide what voters are looking for in a 2020 presidential candidate. Although no one can predict who will show up on Election Day in 2020 (and who will cast ballots in the growing number of states with early and mail-in voting), exactly who those voters will be seems to be the most important factor of all.
The 50.3 percent voter turnout rate in 2018 hit a 50-year high for a midterm election, with 118 million certified votes. In 2014, turnout was 36.7 percent, the lowest turnout in 72 years. The 50-plus percent number is still lower than the average turnout in a modern presidential election, which usually ranges somewhere in the high 50th percentile—from 55.7 percent in 2004 to 61.6 percent in 2008.
Nevertheless, voter turnout in 2018 delivered some clues on new voting habits, and the voting shifts are worth noting. Voter registration surged in several states with key races, especially in the number of new young voters up to age 40. Much of that upsurge came after the mass shooting at Marjory Stoneman Douglas High School in Parkland, Florida, one year ago.
These factoids are from a compilation story by Bloomberg on voter numbers:
- Voter turnout increased in House races across the country from 2014 to 2018. The vast majority shifted left.
- Turnout from 2014 to 2018 increased in every district except two.
- State election laws can influence turnout (see election results in Georgia, North Dakota, etc.).
Lots of factors can weigh on whether someone votes, but the biggest reason for high turnout in 2018 likely has more to do with the national political climate than local races and candidates.
“The obvious explanation is Donald Trump,” said Michael McDonald, associate professor of political science at the University of Florida who also runs the United States Election Project. “That really spurred an interest in politics—and whether you love him or hate him, you're showing up to vote because you want to have your say.”
While it’s true that members of Trump’s base wanted to have their say, an even greater number of those who disapprove of Trump wanted to have their voices heard. We’ve seen these voting results reported many times since November 2018, but if you voted then, you probably will vote in 2020. According to another set of data from Pew Research about 2018:
- The gender gap in voting preference is not new, but it is at least as wide as at any point over the past two decades.
- Women college graduates stand out for their strong preference for the Democratic candidate.
- Whites with less education — particularly men — supported the Republican.
- Majorities of voters ages 18 to 29 (67 percent) and 30 to 44 (58 percent) favored the Democratic candidate.
- Voters ages 45 and older were divided (50 percent Republican, 49 percent Democrat).
Probably the most important fact of all from Pew, because first-time voters are often given less weight in polling: “Among voters who said this was the first midterm in which they voted, 62% favored the Democrat and just 36% supported the Republican.”
So who will the voters be in 2020, and where will they come from? Here are more observations from Pew Research on the likely 2020 electorate and how they might vote:
- Younger generations differ notably from older generations in their views on key social and political issues.
- Nearly a quarter of the electorate (23 percent) will be 65 and older, reflecting not only the maturation of Baby Boomers but also increased life expectancy.
- Baby Boomers and older generations, who will be 56 and older next year, are expected to account for fewer than four-in-ten eligible voters in 2020.
- The changing racial and ethnic composition of the electorate likely has political implications in part because nonwhites have long been significantly more likely than whites to back Democratic candidates.
Probably the most interesting—and potentially most significant—projection from Pew:
Post-Millennials are on track to be more racially and ethnically diverse than their predecessors: In 2020, Gen Z eligible voters are expected to be 55% white and 45% nonwhite, including 21% Hispanic, 14% black, and 4% Asian or Pacific Islander. By comparison, the Boomer and older electorate is projected to be about three-quarters white (74%).
Nate Silver of FiveThirtyEight.com recently joined former Obama economic adviser Austan Goolsbee at a forum at the University of Chicago Institute of Politics. Goolsbee, who teaches at U of C, and Silver, an alumnus, discussed predictions about 2016 and 2020. Silver still defends his 2016 work, since he said he gave Donald Trump a better chance of winning than other outlets. Yet for his 2020 predictions, he sees no reason to revamp his analytical model. “Not only do I think that adjustments are unnecessary, I think it’s the wrong thing to do,” Silver said. His current predictions are “even money” for a Trump re-election with a slight edge to California Sen. Kamala Harris to be the Democratic nominee. (Way to go out on a limb, there, Nate!) He further predicted a “messy election” and “trench warfare” in the Democratic primaries, despite the fact that most of the declared candidates enjoy collegial working relationships.
But after the 2018 election, how can anyone say that analytical models don’t need adjusting? The sheer number of women candidates and voters made up a very different electorate than past numbers would suggest. Almost every subgroup of women in CNN's national exit polls moved toward Democrats, CNN reported days after the midterms. Those groups included white women, Latinas, white college-educated women, white non-college-educated women, Democratic women, and independent women. The percentage of black women voters—traditionally the most reliable Democratic voters of all—stayed about the same.
Maybe the Nate Silver of February 2019 should remember what the Nate Silver of November 2018 tweeted 10 days after the midterms (via a story from Vox):
As FiveThirtyEight’s Nate Silver pointed out, some 60 million people voted for Democrats in the House this year. That’s a big number, considering about 63 million people voted for Trump in 2016. ...
Voters do appear to have been extra-engaged in 2018, and how that will translate over the next couple of years remains to be seen.
“Trump’s not going away in 2020,” McDonald said. “I wouldn’t be surprised to see record turnout.”
Any candidate who takes any voter for granted as we approach 2020 might find himself or herself on the losing end of an election—if he or she is on the ballot at all. Pundits who predict that “this is what voters are looking for” (as Politico did in a “How to Choose the Most Electable Democrat in 2020” piece) might as well turn in their laptops.
When “crazy socialist” ideas like Medicare for all, greatly increased taxes on the wealthy, an ambitious climate action plan like the Green New Deal, and common-sense gun laws including universal background checks are getting high support from voters, all bets could be off.