Sen. Kamala Harris (D-Calif.) addresses the audience at the Politics and Eggs meeting held at the New Hampshire Institute of Politics on the campus of Saint Anselm College in Manchester, N.H., on Feb. 19. (CJ Gunther/EPA-EFE/Shutterstock)
Data analyst and political columnist

By now, the demographic story of the 2016 election is familiar to the point of exhaustion. Donald Trump, through a combination of populist economics, cultural conservatism, “law and order” and a hard-line immigration policy, lost some traditionally Republican suburbanites but won enough formerly Democratic blue-collar white voters to get an electoral college majority and become president. Trump’s blue-collar strength shocked plenty of people who collectively repented by reading “Hillbilly Elegy," sending boatloads of reporters on expeditions to Youngstown, Ohio, and trying to see literally every political story from the vantage point of a white, middle-aged central Pennsylvania man who voted for Barack Obama twice but now wears a MAGA hat because he lost his job in the steel plant.

People tend to view stories about electability through that exact lens, and often assume that what “electability” means is the ability to win over blue-collar white Midwesterners. And while they’re an extremely important group — whites without a college degree made up about 45 percent of the national electorate in 2016 — they aren’t the whole ballgame. There are also votes to be gained among Asians, Hispanics, blacks and college-educated whites across the country. If Democrats want to win in 2020, they should pick a truly demographically electable candidate — one who can appeal to all these groups simultaneously.

Appealing to these different constituencies is a tightrope walk — the most electable Democrat (and the most electable version of Trump) will be the one who can project different images and messages to different audiences without seeming “inauthentic” to any of them. But stepping out onto that tightrope is worth it: The right candidate might be able to put together a bigger coalition, put new competitive states on the map and substantially increase his or her party’s chances of winning (and keeping) the White House.

This installment of our electability series will focus on exactly which voters are up for grabs and what sort of candidate might be able to win them over in 2020. We’ll start by talking about the voters that Democrats left on the table in 2016, look briefly into Trump’s possible growth areas and then use all of that data to describe the sort of flexibility and multifaceted appeal a candidate would need to be demographically “electable.”

Democrats are leaving lots of votes on the table — and many aren’t working-class whites

The Democrats have become increasingly racially diverse over time, but they’re still leaving Hispanic and Asian votes on the table.

According to demographers Ruy Teixeira and Robert Griffin, voters who are Asian or some other race (not Latino, African American or white) made up about 5.5 percent of the electorate in 2016 and voted for Hillary Clinton over Donald Trump, 55.9 percent to 35.7 percent. That’s a solidly Democratic margin, but these numbers conceal room where Democrats could grow — only 49 percent of eligible Asian and other voters turned out to vote, which is almost 20 points lower than the turnout rate for black or white, blue-collar voters.

Democrats would probably benefit from increasing turnout among Asian voters. Despite the internal diversity of the group (Pew found that 24 percent of Asian Americans were of Chinese origin, 20 percent were of Indian origin, 19 percent were Filipino and the rest come from 16 other groups), Asians are generally Democratic. Polls show that they’re less conservative than the rest of the country and more supportive of a larger government that provides more services. Preelection polls show that the baseline partisanship of different groups varies somewhat (e.g. Filipino Americans are less Democratic than Chinese Americans), but as a whole they voted for Clinton by a roughly 20-point margin.

Not all of these voters would help Democrats win the electoral college — many live in deeply blue states, so they likely won’t be a big factor in 2020 — but they could help Democrats pad their popular-vote margin. Moreover, Asians are a growing group, and it’s never a bad thing to try to get increasingly numerous voter groups on your side.

Hispanic/Latino voters, like Asian voters, don’t turn out at high rates and mostly don’t live in swing states. But there are a lot of Latinos throughout the country, and mobilizing them could open up new possibilities for Democrats.

(David Byler/Washington Post)

Most states where Hispanics make up a large portion of the population are blue (e.g. California, New Mexico, Illinois, New York) or have recently trended left (Colorado, Nevada). But Hispanic turnout is still low in Arizona, a red state that Clinton lost by only 3.5 points due partially to her strength in the suburbs. Florida, the quadrennial swing state, also has a solid number of Hispanic/Latino voters, but many of them are Cuban American (who have been historically more Republican than some other Hispanic subgroups). And Texas, the Democratic white whale, has a large nonvoting Hispanic population.

Nationally, only 46.1 percent of voting-eligible Latinos turned out in 2016, so pushing that number up would likely help Democrats both in the popular vote and in key states. Democrats need only a small boost to win Florida again, and Arizona isn’t out of reach if Latinos start to vote more. Increasing turnout isn’t easy. Many Latino voters are young, and nobody has cracked the nut on how to consistently increase turnout among young voters.

A liberal immigration position isn’t a silver bullet, either — Latinos care about a wide range of policy issues, and Trump, despite putting a punitive approach to immigration at the center of his campaign, did about as well as Mitt Romney with Hispanics.

But increments matter, and Democrats don’t need a silver bullet to make improvements. A Democratic Party that focuses more on millennial issues, works on Hispanic representation and continues mobilization efforts could push Florida toward Democrats and nudge Arizona toward some level of purpleness.

Democrats could also work on their turnout among black voters. In 2008 and 2012, black turnout spiked, but it leveled off in 2016. Democrats also increased their (already massive) edge while Obama was on the ballot.

(David Byler/The Washington Post)

Teixeira and Griffin show that if Clinton had replicated Obama’s performance among black voters in 2016, she likely would have barely won Michigan, Wisconsin, Pennsylvania, Florida and North Carolina — which would have given her the presidency. A group of researchers also at The Post’s Monkey Cage blog also showed that if Clinton had replicated Obama’s 2012 turnout numbers among all demographic groups, she would have only lost Georgia by 1.3 points.

If Democrats can turn out more black voters — whether that’s by nominating another black candidate or coming up with a platform that’s more exciting to the group than what Clinton had — they can make some serious inroads. Getting Obama’s numbers back with black voters wouldn’t solve all every Democratic problem, but it’d help a lot.

Democrats probably can’t do much on turnout with white college-educated voters — more than 80 percent of that population voted in 2016. But the 2018 election showed that there might be room for Democrats to make some converts. According to Catalist, a Democratic data firm, white college graduates went from favoring Trump by a 4-point margin to favoring House Democrats by 5 points in 2018.

White, college-educated voters aren’t distributed evenly across the states — many of the most suburban-ish states are safely red or blue. But Trump’s 2016 electoral college victory was built on thin margins in Midwestern swing states. Many of those same states voted for Democratic Senate or gubernatorial candidates in the great suburban revolt of 2018. A candidate who carefully capitalized on some Trump-era cultural fights, passed on others (i.e. skip most of the tweets, pick events like Charlottesville) and projected managerial competence might be able to get even more white-collar votes than Clinton. If someone could do that while increasing black, Latino and Asian turnout, he or she might be able to win without running up the score among Obama-to-Trump blue-collar voters.

But every “electable” candidate should have some sort of plan to gain at least some votes from the white working class. These famous Midwestern Obama-to-Trump voters delivered Trump the presidency, but they’re not down-the-line conservatives. They tend to be to the right on cultural/racial issues, in the middle on moral issues and on the center-left on economic issues. Blue-collar whites made up 45 percent of the electorate in 2016, and the segment that flipped from Obama to Trump is a real part of the electorate in key swing states. A Democratic candidate who managed to put economics at the center, hit Trump for passing traditional Republican laws such as the tax bill and cast themselves as The Real Populist might be able to win back some of these voters.

Obviously, it’s hard to do all of these things — win back Obama-to-Trump voters, push further with white-collar suburbanites and increase turnout among multiple nonwhite groups — at the same time. But doing some of them is helpful. Some back-of-the-envelope math (using Center for American Progress estimates) suggests that combining Clinton’s strength with college-educated whites and Obama’s numbers with blue-collar whites and black voters would (depending on your other assumptions) get Democrats to a mid- to high-single-digit win in the popular vote.

Trump is leaving votes on the table, too

Trump is also leaving votes on the table, but, demographically speaking, he’s less complicated than the Democratic nominee or the next Republican nominee. By nominating Trump, Republicans put reformicon dreams of a multiethnic populist front at least on hold. Trump has never seemed interested in the multiethnic part of that plan. So I would guess (though it’s always important to be cautious with guesses about what Trump will or won’t do) that he’s not going to try make big inroads with nonwhite voters.

But his clearer route to more votes lies with Romney-to-Clinton voters — many of whom were college-educated suburbanites. Republican pollster Patrick Ruffini has calculated that if, on a county-by-county basis, a Republican was able to combine Romney’s strongest areas with Trump’s strongest areas, that candidate would win the popular vote by 1.3 points and carry states such as Minnesota and Virginia to his map.

So how would an ‘electable’ candidate pick up these voters?

I don’t think there’s one single strategy that would allow Democrats to dramatically increase Hispanic and Asian turnout, get Obama’s numbers back with black voters, maintain (or maybe increase) Clinton’s margin with upscale whites while reversing Trump’s gains with the white working class. A Democrat who went hard on economic populism might go too far and give up the suburbs. A Democrat who drove up college-educated white and nonwhite margins through a platform of racial justice and social all-out progressivism might alienate culturally/racially conservative white voters. That’s a lot to ask of a platform.

But demographic electability isn’t about a platform. It’s a balancing act where a skilled candidate can project different images to different constituencies simultaneously. Obama carefully balanced his status as a historic racial symbol with traditional Democratic economic messages, which helped him simultaneously hold onto Northern blue-collar whites and drive up black turnout. George W. Bush was primarily an evangelical to some and a businessman to others. Bill Clinton was a Rhodes scholar who felt at home in rural Arkansas. Electability, especially in the modern iteration of the Democratic Party, means running different, sometimes conflicting, identities and messages in parallel.

The campaign trail will test which candidate is best able to achieve this sort of multidimensionality. Kamala Harris may have an edge on this front. She’s only been in the Senate for two years, so she has a lot of latitude to set her own policy positions. She’s a black and Asian woman, which fits with a lot of the constituencies Democrats want to mobilize. But it’s unclear to me if she’ll want to pursue this sort of political vagueness. She might end up trying to run as a down-the-line progressive and project that image to everyone everywhere.

Joe Biden might also be able to run multiple identities. He could be “Scranton Joe” to some and the vice president to a popular liberal president to others. But he also has a metric ton of baggage and might flop before Iowa votes. At this point, Democrats don’t need to crown one candidate as the most demographically electable. But they should be on the lookout for candidates who can strategically balance their identities, platform, eventual running mate and other spinning plates without getting tagged as disingenuous by their adversaries.

And Republicans would benefit from getting Trump to do more than one thing at a time. Trump has spent the past two years passing traditionally Republican laws, nominating conservative justices and trying by any means necessary to build a border wall. That made parts of his base happy, but it has also kept his approval rating in the low-to-mid-40 percent range.

If Trump were able to simultaneously project a few different images — the competent businessman who managed economic growth, the populist who is delivering for the middle class, the outsider who hates politicians as much as voters do, etc. — he might have a better chance of building a bigger coalition.

There’s nothing wrong with candidates going hunting for voters in Appalachian hamlets and depressed former steel towns. Trump’s 2016 win showed that those voters are real and have a lot of sway. But there are plenty of other voters out there who have yet to have their American heydays. The key to winning the 2020 election for both parties isn’t getting stuck mulling “hillbilly elegies”; it’s figuring out how to convince Americans of all backgrounds that they can move into the future together.

Read more:

David Byler: Want to know which Democrats can actually beat Trump? We don’t have to guess.

David Byler: Think you know which platform will beat Trump? Think again.

Paul Waldman: Will the 2020 Democrat be progressive, or electable against Trump? Probably both!

Paul Waldman: The worst way for Democrats to judge their 2020 presidential contenders

Jennifer Rubin: Here’s the thing about electability