Beneath a Close Election Contest Lie Deep Rifts Among Groups (POLL)

Profound rifts among groups lie beneath the close presidential contest.

— -- Profound rifts among groups lie beneath the close presidential contest, underscoring the country’s fundamental political divisions not only by gender, education, race and ethnicity but also by factors ranging from religious belief to residential area.

See PDF with full results here.

There’s a 22 point gender gap in the contest, nearly double the norm in elections since 1976. And that pales compared with other gaps — 44 points between college- and non-college-educated whites, 65 points between whites and nonwhites, 66 points between rural and urban residents and 97 points between white evangelicals and likely voters who don’t profess a particular religion.

Overall, the race has gone from up 1 point for Trump to even to up 2 points for Clinton in four-night averages in this poll, produced for ABC by Langer Research Associates. These are not significant changes, given the survey’s sample size, about 1,100 likely voters. That said, Trump has had a nonsignificant but numerical advantage just twice in ABC/Post polls, up 2 points in a two-way test May 19 and up 1 point Oct. 30.

Clinton has held significant and even substantial leads, but these have been sensitive to events and propensity to vote, given the lukewarm nature of her support (and Trump’s).

This is reflected in levels of strong enthusiasm for the candidates — tepid overall, with patterns among groups that help identify their strengths and weaknesses. Trump has a 7 point edge in strong enthusiasm, with challenges in enthusiasm for Clinton among independents.

Another factor, early voting, has accelerated, up from 21 percent of likely voters Oct. 30 to 27 percent now. After starting better for Clinton, vote preferences in this group have tightened to 50 percent for Clinton, 45 percent for Trump in the latest four-night average.

Primary Support

Also, unhelpfully for Clinton, there’s a hint of some slippage in turnout. While the results don’t reach statistical significance, former Sanders supporters appear a bit less likely to vote.

Trump is doing less well with Republicans and GOP leaners who wanted someone else to win the nomination; he has 75 percent support in this group, with 15 percent going to Clinton, 7 percent to Johnson or Stein. Two factors mitigate the damage: leaning Republicans have grown more apt to say they supported Trump in the first place (47 percent, up from 40 percent in August), and there’s no slippage evident in turnout among leaning Republicans who preferred someone else for the nomination.

Groups

Differences among groups abound, with one of the largest between evangelical white Protestants — a core Republican constituency — and those who profess no religion. The former back Trump by 77 to 19 percent; the latter, Clinton by 64 to 25 percent. The groups are identical in size, each making up 17 percent of likely voters. The gap between them is typical.

Also consistent with the 2012 election, Clinton leads by 58 to 33 percent among urban voters, while Trump has an even larger lead in rural areas, 68 to 27 percent; they’re running even in the suburbs. A difference here is that rural dwellers account for just 18 percent of likely voters while urban residents make up 33 percent.

Those gaps reflect partisanship: Most urban voters, 61 percent, are Democrats or Democratic-leaning independents, while 57 percent of rural voters are Republicans or GOP leaners.

The racial gap, while wide, is narrower than typical. Whites support Trump by 53 to 39 percent; nonwhites favor Clinton, 72 to 21 percent. But that 65 point divide is a bit smaller than the average in exit polls since 1976, 73 points, as well as 26 points smaller than the record racial gap, 91 points, in 1984. That’s mostly because of lower support for Clinton among nonwhites who are not black or Hispanic; 48 percent back her, versus 66 percent for Barack Obama in 2012.

In a particularly prominent result in this election, gender and education continue to spell broad differences among whites. College-educated white women, consistently strong for Clinton all year, now back her by 59 to 32 percent. Non-college-educated white men, steadily one of Trump’s best groups, support him 64 to 26 percent.

White men with a college degree continue to divide closely, making them a battleground group. Non-college-educated white women — the largest of these groups, at 24 percent of likely voters — have wavered in their support, generally for Trump but by varying margins. Lately, they look more settled on Trump, now by 61 to 33 percent.

Other notable gaps are between white evangelical Protestants and their non-evangelical counterparts (38 points), college graduates and those without degrees overall (33 points), college- and non-college-educated whites in particular (44 points) and, as noted, men and women (22 points). The gender gap this year is larger than the average of 13 points in exit polls since 1976. The previous peak was 22 points, same as now, in 2000.

Enthusiasm

Just 46 percent of Clinton’s supporters and 53 percent of Trump’s express strong enthusiasm for their candidate, including fairly narrow majorities even in their own parties. It’s similar for Trump among independents who lean Republican, 45 percent, but drops to 35 percent for Clinton among independents who lean Democratic.

Among pure independents, 28 percent are strongly enthusiastic about Trump, compared with only 10 percent who are similarly excited about Clinton.

As with vote preferences, variation in strong enthusiasm for the candidates illustrates the cultural bases of the two parties, most starkly by residential area and religion. In rural areas, 42 percent are strongly enthusiastic about Trump, versus 16 percent for Clinton. In urban areas, by contrast, just 17 percent express strong enthusiasm for Trump, versus 33 percent for Clinton. Suburbanites divide evenly.

Nearly half of white evangelical Protestants are strongly enthusiastic about Trump. Only 10 percent of those who do not identify with a religion feel the same way, with 36 percent in this group very enthusiastic about Clinton.

By race and ethnicity, strong enthusiasm for Clinton peaks at 58 percent among black likely voters and 38 percent among Hispanics, dropping to 21 percent among whites. By contrast, strong enthusiasm for Trump reaches 30 percent of whites; it’s just 12 percent among nonwhites.

Education and, to a slightly lesser extent, gender are key factors. Strong enthusiasm for Trump reaches 32 percent among likely voters who haven’t gone beyond high school, and it peaks at 37 percent for Clinton among postgraduates. By gender, 3 in 10 women are strongly enthusiastic about Clinton; 3 in 10 men feel the same about Trump.

Again, there’s great variation among whites by education and gender. Strong enthusiasm for Trump peaks at 40 percent of white men without a college degree and 34 percent of white women without a degree. This flips among white women who have a college degree; 34 percent are very enthusiastic about Clinton. College-educated white men aren’t particularly enthusiastic about either candidate.

Enthusiasm among age groups underlines turnout challenges for Clinton. While Trump’s support is low among 18-to-29-year-olds, only 21 percent of young likely voters are very enthusiastic about Clinton; an additional 32 percent support her but less enthusiastically. By contrast, strong enthusiasm for Trump jumps among those 40 or older, peaking at 33 percent among seniors, a more reliable voting group.

Methodology

This ABC News/Washington Post poll was conducted by landline and cellular telephone Oct. 29 to Nov. 1, 2016, in English and Spanish, among a random national sample of 1,167 likely voters. Results have a margin of sampling error of 3 points, including the design effect. Partisan divisions are 37-30-29 percent, Democrats-Republicans-independents.

The survey was produced for ABC News by Langer Research Associates of New York City, with sampling, data collection and tabulation by Abt-SRBI of New York City. See details on the survey’s methodology here.