If it holds on Election Day, the current 24-point gender gap in House vote preference would be the widest on record for a midterm election in exit polls dating back to 1982 (albeit slightly; it was 21 points in 1994 and 20 in 2014).
Women also now express a disproportionate sense of urgency: Seventy percent say it’s more important to them to vote this year than in past midterms, more than the 62 percent of men who say so. In August it was even, 65 percent in both groups.
While this survey points to continued tailwinds for the Democrats, some results offer hope to the GOP. One is the bump in Trump’s approval; the other is that fewer registered voters now say they’d rather see the Democrats win control of the House, to act as a check on Trump – still a majority, 54 percent, but down from 60 percent in late August.
Perhaps echoing the Kavanaugh controversy, Americans trust the Democrats over the GOP to handle “equal treatment of men and women in U.S. society” by an even wider margin, 55-29 percent – and 76 percent call this especially important in their vote.
The Democrats also lead in trust to change the way Washington works (45-35 percent, also highly important to 76 percent), as well as to handle immigration (50-38 percent, highly important to 70 percent) and Supreme Court nominations (49-38 percent, with 66 percent calling it highly important).
The Republicans are more competitive on two remaining issues – a close 41-45 percent Democratic-Republican split in trust to handle the economy, and 45-42 percent on taxes. Typically, when the Democrats are competitive on taxes, the GOP has a problem.
With voter registration still open, these results are among all adults. But they’re highly similar among registered and likely voters alike. (There are two exceptions: The Democrats’ lead on immigration narrows in some likely voter models, and the importance of Supreme Court nominations gains.)
In a more general measure, the Democratic Party has an 8-point lead over the Republican Party, 45-37 percent, in trust to handle the main problems facing the country. This, too, holds among registered and likely voters alike. And it includes a shift by gender: In 2014, women divided by 41-36 percent, Democratic-Republican, on this question; now it’s 53-32 percent among women. Results among men are unchanged in the same period – 37-42 percent then and now.
Trump, for his part, has the highest approval rating of his career for handling the economy, 49 percent, with 46 percent disapproving. But his overall approval has remained in a 36-42 percent band since he took office, averaging 38 percent – the lowest average approval for a president in his first 20 months in office in polling dating back to Harry S. Truman.
Similarly, at 41 percent today, Trump’s approval rating is the lowest for a president heading into his first midterm election since Truman in 1946. (Truman’s party lost 54 House seats.)
Trump’s low approval is a substantial problem for his party. Among registered voters who disapprove of the president, 64 percent say it’s extremely or very important to them to support a candidate who shares their opinion of him. Among those who approve of Trump, 57 percent say the same. That 7-point gap was not present last month. Moreover, the fact that Trump is 10 points more unpopular than popular among registered voters (53-43 percent) puts his party at a disadvantage.
Vote preference of 53-42 percent, D-R, looks not at all like the 2014 midterms, when ABC/Post polling at this point showed an essentially even 46-44 percent split, or 2010, when it was 47-43 percent. Instead it highly resembles 2006, a 54-41 percent result at about this time.
The Republican Party gained 13 House seats in 2014 and 63 in 2010. In 2006, the Democrats gained 31. They need 23 this year to win control.
Another result adds to this evidence – an evaluation of the leanings of political independents, who have accounted for a plurality of Americans in annual averages in nine of the last 10 years. In ABC/Post polls across 2014, independents on average leaned slightly toward the Republican Party – 37 percent toward the Democrats, 41 percent toward the GOP. In 2010 it was dead even, 42-42 percent. In 2006, by contrast, independents leaned Democratic by 17 points. And this year they’ve done so by a 14-point margin, 46-32 percent.
Turnout and Overvote
The Democrats, to be sure, also can find grounds to worry. Their support is strongest among minorities and young adults, groups that tend to have lower turnout in midterm elections. It’s also focused in urban areas, where the Democrats generally already control House seats; specifically, registered voters favor Democratic candidates by 63-31 percent in cities, vs. an essentially even 49-46 percent in suburbs and 44-53 percent in rural areas.
That suggests a Democratic overvote in districts they already hold, and other evidence backs it up. In districts rated as solid or likely Republican by the ABC News Political Unit, Republican candidates lead by 55-40 percent. In districts rated as solid or likely Democratic, Democrats lead, by a much larger margin – 68-28 percent – more than they need in these locales. And in the 66 House districts rated as only leaning either way, or as tossups, it’s a 46-47 percent D-R race.
That said, this poll suggests a higher than typical midterm turnout, which should advantage the Democrats. Among all registered voters, 76 percent say they’re certain to vote next month, vs. 63 percent at this time in 2014 and 70 percent in 2010. While voting intention is high across the board, that’s especially the case among Democrats and Democratic-leaning independents, nonwhites and 18- to 39-year-olds.
Shifts among groups tell an important part of the story. Conservatives favor Republican House candidates by a wider margin – 58 points – than in ABC/Post polls at this point in any of the past three midterm elections. (It was 49, 46 and 38 points in 2014, 2010 and 2006.) But moderates favor Democrats by 30 points, triple the margin in 2014, double what it was in 2010, and about matching its 2006 level. Liberals, for their part, favor Democrats by a vast 76 points.
Whites favor Republican candidates by 6 points, compared with 18 points in 2014 and 12 in 2010; whites were +4 Democratic at this point in 2006. It’s among nonwhites that Democrats take their advantage, with a 46-point lead in a group that accounts for 32 percent of registered voters and anywhere from 27 to 30 percent of likely voters.
Democratic candidates have a 21-point lead among college graduates, 58-37 percent, again similar to 2006 and far from what essentially were dead heats among graduates in 2014 and 2010. But Democrats have a scant +4 edge among nongraduates, similar in this case to 2014 and 2010 and well off the 2006 polling, +13. Since nongraduates account for nearly two-thirds of the population, the Democratic weakness there represents a threat.
That said, the Democrats have some surprising comparative strengths. They trail the GOP by 9 points in rural areas of the country, as noted; that compares with 26- and 17-point deficits at this point in rural America in 2014 and 2010. By region, the Democrats are especially strong in the comparatively liberal Northeast, a 27-point lead, again far more than in 2014 or 2010 and quite similar to 2006.
But it’s independents, the quintessential swing voters, who may matter most. They currently favor Democratic candidates by 14 points, 52-38 percent. Independents by contrast were +4 and +13 Republican at this stage in 2014 and 2010. Now they look – yet again – as they did in 2006.
This ABC News/Washington Post poll was conducted by telephone Oct. 8-11, 2018 among a random national sample of 1,144 adults, with 65 percent reached on cell phones and 35 percent on landlines. Results have a margin of sampling error of 3.5 points for the full sample, including design effects due to weighting. Partisan divisions are 33-26-35 percent, Democrats-Republicans-independents.
The survey was produced for ABC News by Langer Research Associates of New York, N.Y., with sampling, data collection and tabulation by Abt Associates of Rockville, Md. See details on the survey’s methodology here.