A new poll of Washington's 8th Congressional District race foreshadows a hard year for Republicans, but New York Times polling guru Nate Cohn — who grew up in the district — says it may still be a harder lift for Democrats than some believe.
The real-time poll, watched by many local political junkies, showed its results on a map and chart as they came in between Sept. 24 and 26, with the early responses favoring Schrier, then bouncing toward Rossi before settling into a dead heat.
Cohn has more than a passing interest in Washington state politics, as someone who grew up in Auburn and graduated from Whitman College. On Twitter, he has called the 8th “America’s greatest district.” He agreed to an email Q & A Thursday with The Seattle Times to break down the poll results.
Q: It’s been addictive for a lot of political junkies to watch your live polling experiment. What was your goal in trying this out? What have you learned?
Most Read Local Stories
- How December weather is shaping up in the Seattle area after that warm, wet November
- Inslee, Washington state Democrats discuss delaying WA Cares long-term care payroll tax
- Six Seattle shootings in a little over 24 hours leave 3 injured
- How trellises are transforming Washington's apple orchards
- Washington Supreme Court won't redraw political maps, will accept redistricting commission's work
A: After 2016, we wanted to demystify polling. We wanted people to see what polling really is, how hard it is, and why polls are so much fuzzier and uncertain than the seemingly precise numbers make it seem.
If you watched the WA-8 poll finish live, you saw Rossi and Schrier trade the lead back and forth over the final respondents. And if we had kept polling, Rossi could have easily retaken the lead. That’s what we mean when we say there’s a margin of error of 4.6 points, but I think it feels a lot more real if you watch it live.
Q: One obvious lesson for those of us watching the poll play out: People don’t like to answer the phone. You made 40,568 calls to reach 505 people. Is that a pretty common ratio? Is it getting more difficult to conduct phone polls? How do you deal with that in your modeling?
A: It’s a little worse than average, perhaps especially for a district that’s not particularly urban, but it’s not atypical. It’s getting more difficult and more expensive to conduct phone polls every year, and we have to deal with it in a lot of ways. We try and call more people from groups who we think are less likely to respond. We try and adjust our sample to give more weight to respondents from groups who didn’t respond in the numbers we hoped.
Q: Your WA-08 poll came out basically even, Kim Schrier 46, Dino Rossi 45. This in a district that Rossi carried easily in his previous statewide runs for office. How bad a result is that for Republicans? How does it compare with your polls nationally?
A: It’s a tough result for the Republicans. Rossi is all but an incumbent in terms of his name recognition, and voters in our poll had a favorable impression of him by a 6-point margin. But this is a tough national environment for the Republicans and it’s a district that voted for Hillary Clinton.
If you dig into our results, Schrier had a 20-point lead in the north King County part of the district. It’s competitive in statewide elections and it’s voted for Rossi in the past, but voted big for Hillary Clinton. It doesn’t seem like voters there are going to entertain voting for a Republican in this political environment, and that’s a trend we’re seeing in well-educated areas across the country.
That said, this is a stronger result for Republicans than a lot of other contests. Rossi is still highly competitive, after all, and that’s more than can be said for some other Republicans in districts that lean Democratic in presidential races.
Q: What are the caveats about the results you’d offer in terms of who you reached vs. those who didn’t answer the phone? I noticed you didn’t reach your target among 18-29-year-olds, for example. Do you have a sense the poll may overstate or understate the supposed Democratic wave brewing?
A: I wouldn’t dig through the detailed results to try and identify the places where it looked like one side was being underestimated or overestimated. I would mainly just emphasize that it’s just one poll, and the “true” state of play could easily lean either way.
Q: As a WA-08 native, what else jumps out to you about the 8th District and its voters in these midterms? (For example, you mentioned on Twitter misconceptions people have about the district.)
A: I think the typical national political observer tends to erroneously caricature WA-8 as a well-educated, high-tech suburban district, like those outside of other fairly similar major cities like Denver or Washington, D.C., or Minneapolis. But Eastside, north King County suburbs are only about 25 percent of the district, and the rest is decidedly middle to working class. And this is a district that often splits its ticket and votes for Republicans for state offices, including for Rossi in the past, even as it votes for Democratic presidential candidates. Put it together, and I think this is a tougher lift for Democrats than some people assume.