A new Washington Post/ABC News poll released over the weekend pointed toward a scenario in which President Trump could win reelection despite his deep and enduring unpopularity.

The poll showed that all of the top-tier Democratic candidates defeat Trump in head-to-head matchups by sizable margins among all U.S. adults. But among registered voters, those matchups suddenly become mostly dead heats.

Specifically, among all adults, Joe Biden leads Trump by 55 percent to 41 percent; Bernie Sanders leads by 51 percent to 45 percent; Elizabeth Warren leads by 51 percent to 44 percent; Kamala D. Harris leads by 51 percent to 43 percent; and Pete Buttigieg leads by 48 percent to 44 percent.

But among registered voters, Biden leads Trump by 53 percent to 43 percent, while all the others are locked in a dead heat with the president.

To be sure, polling right now isn’t at all predictive. But this does offer an opportunity to ask whether Democrats are reckoning with the possibility that a voter turnout failure could allow Trump to squeak through to a second term. Making this more likely, Trump could again prevail in the electoral college while losing the national popular vote.

Obviously, the matchups among registered voters don’t by themselves tell us much about what turnout might look like, since registering people is half the battle, while getting them to actually vote is a separate matter.

But still, the disparity in the matchups among all adults on one hand and among registered voters on the other is a reminder that as a general matter, Democrats will reduce the chances of another fiasco if they do a good job at tapping into the pool of eligible voters. That will, of course, entail registering them and turning them out.

Right now, most signs are that senior Democrats and Democratic-aligned data crunchers expect that interest in the 2020 election will be something approaching thermonuclear, which in theory should make it possible to do that successfully.

Writing at the Atlantic, Ronald Brownstein recently offered a definitive analysis of 2020 turnout expectations. As Brownstein reported, figures collected by Michael McDonald of the University of Florida, an expert in demographics and voting behavior, suggest that as many as two-thirds of eligible voters may cast ballots in 2020 — which would be the highest percentage in over a century.

The Democratic-aligned firm Catalist recently estimated that turnout could climb as high as 160 million — dwarfing the 138 million who turned out to vote in 2016. We’ve already seen a hint of what this might look like with the massive turnout in the 2018 midterm elections, which drove the large national popular vote win that delivered control of the House to Democrats.

But what remains unclear is whether this expected expansion of the electorate in 2020 will work to Democrats’ advantage a second time. While turnout is expected to grow among minorities, young people and college-educated whites, just as it did in 2018, it’s also possible that turnout could surge among non-college-educated whites — Trump’s base — a demographic that did not turn out at high rates in 2018, relatively to 2016, but could do so again once Trump is back on the ballot.

The wild card in all this is continuing demographic change — and here is where the pool of eligible voters again comes into play. As Brownstein summarizes:

The nature of the population eligible to vote is evolving in a way that should indeed help Democrats. McDonald estimates that the number of eligible voters increases by about 5 million each year, or about 20 million from one presidential election to the next. That increase predominantly flows from two sources: young people who turn 18 and immigrants who become citizens. Since people of color are now approaching a majority of the under-18 population -- and also constitute most immigrants -- McDonald and other experts believe it’s likely that minorities represent a majority of the people who have become eligible to vote since 2016.
The generational contrast in the eligible voting pool is also stark. States of Change, a nonpartisan project studying shifts in the electorate, projects that Millennials (born, according to the organization’s definition, from 1981 to 2000) will constitute 34.2 percent of eligible voters next year. Post-Millennials (born after 2000) will make up another 3.4 percent. That means those two groups combined will virtually equal the share of eligible voters composed of Baby Boomers (28.4 percent) and the Silent and Greatest Generations (another 9.4 percent).

Given the depth of alienation from Trump among minorities and young voters, those trends should favor Democrats. To put it simply, younger, racially diversifying and socially tolerant America is growing, and Trump appears determined to do all he can to drive it away.

But there are other wild cards as well: The percentages of eligible voters who actually show up also matter. In 2016, that lagged among nonwhites and young voters, helping enable Trump’s victory. That turned around in 2018, boosting Democrats, but what happens in 2020 obviously remains to be seen.

Senior Democrats appear wary of this: Last year, the Democratic-aligned super PAC Priorities USA released an analysis showing that alarmingly high percentages of unregistered African American millennials don’t feel that either party represents them. The super PAC noted that registering and involving them in politics must be an urgent Democratic priority going forward.

In the end, one big question is whether eligible nonwhite and millennial voters will increase their turnout rate more in 2020 (relative to 2016) than older and white voters do. That could happen, and it would benefit Democrats. But it does remain very possible that Trump could pull off another miracle with working-class white turnout.

All of this should serve as a reminder that Trump himself won’t be enough to produce the turnout that Democrats need to beat him. A great deal will turn on their own ability to register — and engage — these voters.

Read more: