The Washington PostDemocracy Dies in Darkness

Democrats are coming into November in a familiar position: Urgently needing to push turnout

Enthusiastic supporters cheer Hillary Clinton in Manchester, N.H. on Feb. 3, 2016. (Jabin Botsford/The Washington Post)

Voting takes effort. No matter what, you have to do something to vote: Go to a polling place at a certain time, put a ballot in the mail, whatever. That's assuming you want to vote, that you have a candidate you like or that you've had the time to figure out which candidate is the best choice — and also that you have registered. This is a big reason that turnout correlates to age and income. People with more income tend to work steadier hours; older people tend to be less likely to have moved recently. If it's easier to get to the polls and if you don't have to re-register, you're more likely to cast a vote.

This is a disadvantage to Democrats. More Democratic voters are younger and have less income. More are nonwhite, which also correlates to younger age and lower incomes. So Democrats routinely have to do more work to get voters to the polls.

In November, that challenge may be bigger than usual.

Gallup has tracked the extent to which people plan to vote over the past several cycles. The peak of certainty that people would vote at this point in the election came in September 2008, when at least 80 percent of members of each party said they planned to cast a ballot. In 2012, that fell among both parties and among those younger than 55. This year, it has fallen further still. Less than half of those 35 and younger say they're certain to vote this year; fewer than two-thirds of Democrats say the same.

The most recent Washington Post-ABC poll gives us a closer look at how different voting blocs have shifted since four years ago.

Our numbers are less stark than Gallup's. In September 2012, 84 percent of registered Democrats told us that they were certain to vote, compared with 85 percent this year. (In each year, the figure for Republicans was 90 percent.) Republicans are less likely to say they are following the race closely, but, overall, Americans are paying about as much attention as they did at this point in the Obama-Romney race.

That said, there are some significant changes. Those 65 and older are more likely to say they are paying close attention to the race; black voters are much less likely to say they are. Older voters turn out consistently, so they are not more likely now to say they plan to come out and vote, but among black voters — a critical Democratic constituency — that number is lower than four years ago.

(Notice that the horizontal scale on the two charts below is not the same.)

A few bright spots for the Democrats: Hispanic voters are following the race more closely and report a bigger intent to come out and vote. Historically, Hispanic voter turnout has been far lower than other groups, so there's a lot of room to grow. College-educated white men are less likely to say they plan to vote than four years ago, and they lean to Donald Trump. (That decrease may be a function of relative apathy about the candidates, as might the decrease among black voters.) Those who support Clinton are more likely to say they plan to vote this year than those backing Obama said at this point in 2012.

The overall problem for Democrats is that their strongest supporters are often those least likely to get to the polls. Notice, below, that the blue dots are mostly lower than the red dots, meaning they report less likelihood to go to the polls.

The exceptions are driven mostly by white women with college degrees, a group that backed Mitt Romney in 2012 and this year support Clinton. That's an important exception — and it's also why white college grads are on the left side of the graph.

How do Democrats ensure that their voters go to the polls? They run get-out-the-vote efforts. These can be hard to measure, since campaigns do not release accurate numbers about how many voters they have contacted or persuaded.

Another question for the Democrats: The Post-ABC poll asked Americans if they'd been contacted by either campaign, and the results were about even, with a fifth of respondents saying they'd been contacted by each campaign. This is self-reported, people saying that, yes, one of the campaigns contacted them by phone, in-person or online to seek their support. That means that some people may include things like email fundraising solicitations or may simply misremember. (The figure is about the same in battleground states, which is not what you'd expect to see.) But there's not the obvious gap here between Clinton and Trump that you might expect, given the robust outreach efforts of the former and the ... less robust outreach of the latter.

Turnout efforts ramp up exponentially as the election approaches. That means sooner rather than later in early-voting states, but it's another way in which it can be hard to evaluate how those efforts will affect the outcome. As always, the need for Democrats to get their voters to the polls is bigger than the Republicans' — but this year, that need may be bigger than most.

Scott Clement contributed to this article.