Welcome to Pollapalooza, our weekly polling roundup. We’re back after a week off (and on a Monday instead of a Friday); we got sidetracked with … other things.
Poll of the week
The Trafalgar Group, a polling firm that tends to work with Republican candidates and groups, put out an Indiana poll that cuts against type: It shows Democratic Sen. Joe Donnelly defeating Republican businessman Mike Braun 51 percent to 39 percent. Democrats shouldn’t go jumping for joy just yet, though. The poll was a clear outlier, as previous polls had shown a tight race. Trafalgar also isn’t a gold-standard pollster.1 But there have been only two previous polls of the race, both of which were done by pollsters who had FiveThirtyEight pollster ratings of C+ or lower, so this is a contest that’s crying out for more and better polls. Given Democrats’ dominance in a set of high-quality polls in three other Midwestern states last month, it’s certainly possible that Donnelly is ahead.
But the really interesting findings from the Trafalgar poll came after they asked who respondents would vote for if Donnelly voted to confirm Brett Kavanaugh, President Trump’s recent Supreme Court nominee. Rather than giving Donnelly a bigger lead, as you might expect in a state as red as Indiana, it actually tightened the race considerably: In that scenario, Donnelly led Braun only 39.4 percent to 38.5 percent. And when voters were asked what they’d do if Donnelly voted against Kavanaugh, the Democrat took a 7-point lead, 45 percent to 38 percent. It’s interesting, though, that Donnelly’s lead shrunk from its original 12 points after voters were asked about Kavanaugh no matter how Donnelly hypothetically voted.2
The poll is a point in favor of the argument that some red-state Senate Democrats have more to lose by voting for Kavanaugh and demoralizing die-hard Democratic voters than by voting against him and risking a loss of Republican support. Notably, Braun’s vote share doesn’t budge from 38-39 percent in any of the three trial heats; his supporters don’t seem to be swayed by what Donnelly does with regard to the Supreme Court. But Donnelly’s voters seem to care a lot. Eleven percentage points’ worth of his initial support winds up in the undecided column when he votes “yes” on Kavanaugh, suggesting plenty of Democrats would be unhappy with that decision. But when he votes “no,” only 5 points’ worth of his supporters switch to undecided — most likely swing voters who support Kavanaugh enough to say they’ll base their November votes on his confirmation.
It’s worth noting that plenty of those Democratic base voters who moved to the undecided column would probably vote for Donnelly in the end anyway. But there is also historical precedent for the somewhat counterintuitive idea that vulnerable senators should not cross party lines in Supreme Court votes. During Robert Bork’s confirmation battle — he was nominated by President Ronald Reagan in 1987 but the Senate ultimately refused to confirm him — several Southern Democrats who relied on black voters to get elected felt that they couldn’t support Bork because of his opposition to the Civil Rights Act. And in 1991, the defeat of Democratic Sen. Wyche Fowler Jr. of Georgia was linked to his vote to confirm President George H.W. Bush’s nominee Clarence Thomas, which reportedly weakened Fowler’s support among women voters because Thomas had been accused of sexual harassment before the vote.
But this poll is hardly the last — or only — word on the subject. Trafalgar said this survey is the second in a series of polls exploring whether voting for Kavanaugh will help or hurt certain senators at the ballot box. The first poll, of West Virginia, had a very different finding: Voting for Kavanaugh gave Democratic Sen. Joe Manchin a commanding lead over Republican Patrick Morrisey, but voting against Trump’s nominee dropped the race into a statistical tie.3 It remains entirely possible that what kind of electoral fallout senators face for this fall’s Supreme Court vote will depend on the state. Indiana supports Trump, but West Virginia really supports Trump. Presumably, senators like Donnelly and Manchin will have to make an educated guess about how their voters feel — and if it turns out they’re out of touch with their constituents, they could be out of a job come January.
Other polling nuggets
- A Monmouth poll of New Jersey’s 3rd district shows a close race between Republican incumbent Tom MacArthur and Democratic challenger Andy Kim. The 3rd district voted for Barack Obama by 5 points in 2012, but went for Donald Trump by 6 points in 2016.
- In Minnesota, which is hosting two senate races in 2018, an Emerson College online poll shows one race much closer than the other. Democratic incumbent Amy Klobuchar leads her Republican challenger Jim Newberger by 24 points. But in the special election for the seat vacated by Democratic Senator Al Franken, who resigned amid allegations of sexual harassment, the Democrat, Tina Smith, leads by only four points against Republican Karin Housley.
- In New York’s 27th district, where Republican Chris Collins has suspended his bid for re-election after being charged with insider trading, a poll by Clout Research looks at the viability of several Republicans against Democrat Nate McMurray. The poll shows every Republican tested beating out the Democrat, with state Senator Pat Gallivan running up the largest margin, at 20 points.
- A Global Strategy Group poll of 400 Latino voters in Florida found that Democratic Sen. Bill Nelson was favored to win re-election over Republican Gov. Rick Scott 44 to 41. Among Cuban Americans however, Scott was favored 57 to 33 over Nelson.
- Younger adults are less likely to believe that Democrats will win control of Congress after this midterm election, according to a CNN poll conducted by SSRS. Fifty-three percent of respondents under age 45 believe that Republicans will retain control of congress, while only 44 percent of those over age 45 feel the same way.
- 47 percent of Democrats have a positive view of capitalism, according to a Gallup Poll, which is down from 56 percent in 2016. Fifty-seven percent of Democrats now say they view socialism positively.
- According to a YouGov poll, 31 percent of Americans say they read between one and five books per year, 16 percent say they read between six and 10 books, 35 percent say they read more than that (with 6 percent reporting that they read over 50 books a year!), and 12 percent say they don’t read books at all.
- 77 percent of Americans say that using genetic engineering to make aquarium fish glow is taking the technology too far, a Pew Research Center survey finds. Sixty-seven percent say the same of manipulating genes to bring back extinct species. Americans were split 55-43 on whether it was acceptable to use genetic engineering to make meat more nutritious, but 70 percent said using it to limit mosquito reproduction was appropriate.
- If you like polling, you’ll be sure to love FiveThirtyEight’s brand-new 2018 House forecast! Check back daily for an updated forecast of the 2018 midterm election for the House of Representatives.
Trump approval
According to FiveThirtyEight’s latest averages, Trump has an approval rating of 42 percent and a disapproval rating of 52.5 percent. That -10.5 point net approval has been remarkably steady for months now. Last week, his approval rating was 41.9 percent and his disapproval rating was 52.5 percent, for a net approval of -10.6. At this time last month, his approval rating was 41.8 percent and his disapproval rating was 52.8 percent, for a net approval of -11.
Generic ballot
In an average of polls of the generic congressional ballot, Democrats currently lead by 7.8 percentage points (47.7 percent to 39.9 percent). That lead has widened a bit from last week, when it was 6.5 points (47.6 percent to 41.1 percent) — but it’s narrower than one month ago, when Democrats’ advantage stood at 9.3 points (48.7 percent to 39.4 percent).
Check out our 2018 House forecast and all the polls we’ve been collecting ahead of the midterms.
CORRECTION (Aug. 20, 2018, 1:10 p.m.): A previous version of this article referred incorrectly to Democratic Sen. Bill Nelson in Florida. He is running for re-election to the U.S. Senate, not for governor.