In ten days, Washington State’s general election will be certified and the results of the qualifying round of this year’s local elections cycle will become final.
With only a hundred ballots listed by King County Elections as awaiting tabulation, and with almost a month having now passed since we announced the electoral findings from our October 2021 survey of the Seattle electorate, it seems like a good time to revisit our poll findings and examine to what extent they anticipated the outcomes in the actual returns tallied by King County Elections.
Our survey, conducted by Change Research from October 12th-15th, 2021, was our second poll of Seattle this cycle and one of just three independent polls conducted in the Emerald City between the certification of the August 2021 Top Two general election and the deadline to return ballots in this general election.
617 likely voters participated, all online.
We released the results of all of the electoral questions on Tuesday, October 19th, via press conference and here on NPI’s Cascadia Advocate.
Our objective in commissioning these surveys was to help everyone interested in Seattle politics get a better sense of the electoral landscape and dynamics.
As I noted last summer, polls can’t predict the future. That is not their purpose. What polling can do, however, is indicate how people (especially likely voters!) may be feeling about proposed legislation, ballot measures, an issue, or a candidate election at a particular juncture. It’s crucial to note that all polls are snapshots in time and that polls vary in terms of their quality and credibility.
The methodology used to conduct a poll is of great importance.
If a poll has a properly representative sample and asks neutral questions, the results can be useful. But if the inputs are bad, the outputs will be bad.
That’s why we and Change Research worked hard to design and field a survey that would be credible. We wanted the data to be useful. We are firm believers in rigorous, high quality research, and we practice what we preach.
Since we’re now at a juncture where we can safely compare our data to the (mostly) complete unofficial results, let’s dive in and see how the returns mesh with our polling from about a month ago.
What the poll found: Of the two finalists for Mayor, the poll found that former Seattle City Councilmember Bruce Harrell had a sixteen point lead over current Council President M. Lorena González, with nearly a majority of respondents expressing a preference for Harrell and just under a third expressing a preference for Gonzalez. 18% said they were not sure and 2% said they would not vote.
What happened in the election: Harrell is decisively defeating González, who conceded the race last week. As of today, Harrell has 58.58% of the vote and González has 41.12%. That’s a margin of victory a little more than seventeen points, which is awfully close to the margin in our polling.
Analysis: Harrell entered the general election as the frontrunner, having won the August Top Two election by a narrow margin. There is no evidence that Harrell ever lost his advantage. While he ultimately prevailed over González in August by just a couple points, he was able to jump out to a much bigger lead before the voting began in this final round, owing in part to the weakness of González’s campaign, which didn’t offer a compelling message to voters.
The other two independent polls I mentioned earlier — conducted by Elway for Crosscut and by Strategies 360 for KOMO — each found Harrell ahead of González by a few points at two different junctures in September. By the time our survey fielded, Harrell had built an even bigger lead… one he would not relinquish.
The remaining undecided voters seemingly went to both candidates in roughly equal proportions, leaving Harrell’s already sizable lead intact. Harrell’s more than two-to-one advantage with voters ages 65 and older was a key factor.
Seattle Times reporter Daniel Beekman asked the González campaign for reaction to our poll finding as part of his story on the survey’s findings.
Campaign manager Alex Koren’s response was to cite the campaign’s own internal polling from a month prior showing González and Harrell tied at 45% apiece, and to question the veracity of our survey’s sample. From Beekman’s article:
Asked about the NPI poll, González campaign manager Alex Koren said the campaign’s own poll in mid-September showed the race tied at 45%.
In a Crosscut/Elway poll conducted in early September, 42% of respondents picked Harrell, 27% picked González and 24% were undecided. In a KOMO/Strategies 360 poll conducted in mid-September, 40% picked Harrell, 33% González and 27% were undecided.
The NPI poll may have underestimated support for progressive candidates like González and City Council Position 8 incumbent Teresa Mosqueda, Koren said.
The mostly complete results clearly demonstrate that our poll’s modeling and sampling was sound, whereas the González campaign’s poll was an outlier.
González’s internal polling was never corroborated by any other data that we saw, whereas our finding in favor of Harrell was, we understand, corroborated by private polling commissioned in October by some of González’s allies.
Their data may not have been publicly available, but ours was. We deliberately chose to release our findings in the hopes of providing a service to the public.
What the poll found: Of the two finalists for Seattle City Attorney, our poll found a big lead for Republican Ann Davison, with abolitionist Nicole Thomas-Kennedy nineteen points behind. 43% of respondents to the poll said they were voting for Davison, while 24% said they were voting for Thomas-Kennedy. A significant number, 30% said they were not sure. 2% said they would not vote.
What happened in the election: Davison prevailed and will go on to win the election, but not by anything resembling the margin in our poll. Davison currently has 51.52% of the vote, while Thomas-Kennedy has 47.7%. On Election Night, Davison had a much, much bigger lead: 58.25% to Thomas-Kennedy’s 40.96%.
Analysis: This is a race in which the dynamics definitely changed during the voting period. The late ballots prove it. The reason Thomas-Kennedy is doing so much better in the election than in the survey is that she was able to reel in a lot of those undecided voters in the final days and weeks of the 2021 campaign.
Those not sure voters had to go somewhere (93% of our survey takers said they were “definitely” planning on voting) and most went to Thomas-Kennedy.
Unlike González, Thomas-Kennedy’s campaign offered a strong closing argument in favor of her candidacy that was widely shared on social media platforms. The election results, compared to our polling, suggest that argument resonated.
Had Thomas-Kennedy campaigned with the branding of a progressive Democrat and not had to contend with former Democratic governors and Supreme Court justices endorsing her opponent (who was rated “not qualified” by a coalition of bar associations) she might have prevailed in the election, even despite her Twitter history, which her opponents seized upon in an attempt to discredit her.
City Council, Position #8 (At-Large)
What the poll found: Of the two finalists for Seattle City Council Position #8, one of the council’s two at large positions, our poll found an eight point lead for Councilmember Teresa Mosqueda. Mosqueda, the only incumbent on the ballot in the four citywide races, got 39% in the survey, while challenger Kenneth Wilson got 31%. 26% said they were undecided and 3% said they would not vote.
What happened in the election: Mosqueda is easily defeating Wilson, with a percentage almost identical to her finish in the August 2021 Top Two election. Mosqueda has 59.36% of the ballots cast, while Wilson has 40.19%. Mosqueda is presently the best performing candidate out of all eight candidates in the four citywide races. On Election Night, however, her lead over Wilson looked a lot more like the margin in our poll. Then, she had 52.4%, while Wilson had 47.1%.
Analysis: When we released our polling, I noted that there were a significant number of undecided voters in the race and added that I saw multiple plausible scenarios for how the contest could play out. One of those was that we’d see a repeat of the Top Two election dynamics, in which Mosqueda did not perform all that impressively in our preelection polling, but went on to do just fine in the actual election, where “not sure” isn’t an option.
And that is the scenario that came to pass.
Wilson has as much support as all of Mosqueda’s challengers collectively had in the summer, which suggests that people who were looking for an alternative to Mosqueda in the general election happily gravitated to his candidacy, whilst most of the “not sure” voters simply came home to Mosqueda, propelling her to a comfortable victory and a second term on the Seattle City Council.
It is not at all uncommon for voters in local races (especially races that don’t have partisan labels on the ballot) to be not sure who they are voting for pretty late in the election cycle, including through a lot of the home stretch. Change Research’s Ben Greenfield and I discussed this phenomenon in our post-results release Q&A.
City Council, Position #9 (At-Large)
What the poll found: Of the two finalists for Seattle City Council Position #9, the council’s other at-large position, our poll found a four point lead for Fremont Brewing cofounder Sara Nelson. Nelson received 41% support in the survey, while her opponent, author and activist Nikkita Oliver, received 37%. 21% said they were undecided and 2% said they would not vote.
What happened in the election: During the voting period, Nelson surged out to a big lead over Oliver, mirroring what happened in the August election. On Election Night, Nelson had 60.31% of the vote and Oliver was under forty percent, with 39.46%. But as ballots were counted, Oliver closed the gap, and the race is now closer… not as close as the City Attorney race, but certainly closer than it was. Nelson currently stands at 53.87%; Oliver has 45.96%.
Analysis: In the August Top Two election, Oliver was eventually able to overcome Nelson’s Election Night lead and claim the first place spot thanks to late ballots swinging several successive counts in their favor. However, they weren’t able to do that in this final round because Nelson jumped out to a much bigger lead.
In August, Nelson’s Election Night lead over Oliver was 7.4%, which was not an insurmountable spread. But in this round, Nelson’s Election Night lead was over twice as big: 20.85%. And so, even though Oliver outperformed Nelson in the late ballots again, it wasn’t enough to change the outcome of the race.
For Oliver to pull off another comeback, they would have had to have been closer to Nelson in the initial returns. But that scenario didn’t come to pass.
Our polling found that the highest geographic concentrations of undecided voters in this race were in neighborhoods like West Seattle, Magnolia, and downtown. The poll found that Nelson was already doing really well in North Seattle neighborhoods like Bitter Lake while Oliver had firm, enthusiastic support in neighborhoods like the Rainier Valley and Beacon Hill.
If you look at the Election Night precinct level data (visualized here thanks to Jason Weill), you can see that Nelson won most of the early vote in those neighborhoods with the most not sure voters — like West Seattle. That accounts, at least in part, for how Nelson pulled off the general election surge.
Seattle School Board races (Districts #4, #5, and #7)
What the poll found: Of the six finalists for Seattle School Board races, our poll found Vivan Song-Maritz, Michelle Sarju, and Brandon Hersey with leads. Sarju and Hersey had large leads of more than twenty-five points over opponents Dan Harder and Genesis Williamson, while Song-Maritz had a ten point lead over her opponent Laura Marie Rivera. Majorities or near majorities of voters said they were undecided. Smaller percentages said they would not vote.
What happened in the election: Song-Maritz, Sarju, and Hersey are all winning, and all are winning by margins exceeding those in the poll, which is to be expected considering how many undecided voters there were. Hersey has the biggest margin of victory because his opponent dropped out months ago, with 92.71% of the vote to Williamson’s 6.63%. Michelle Sarju has 85.14% of the vote to Dan Harder’s 14.63%. And Song-Maritz has 71.97%, while Rivera has 27.67%.
Analysis: Our polling showed that there were three clear frontrunners in these Seattle School Board races, and all of them were well positioned for victory.
Not surprisingly, the undecided voters flocked to the frontrunners and each of them jumped out to big leads in the initial returns that they did not relinquish.
Song-Maritz was initially at 67% and was able to go up to 71% in the late ballots. Sarju and Hersey, meanwhile, started out in the eighties and nineties on Election Night, so their races saw less movement, but even they picked up support as the counting went along, while Harder and Williamson slipped.
Hersey was the only incumbent on the ballot, with Song Martiz and Rivera having ousted appointed boardmember Erin Dury in the August Top Two election.
Although our poll couldn’t predict what would actually happen in the election, it did have the potential to indicate who might win, as explained in the introduction above. And those indications turned out to be right across the board.
All seven of the candidates who placed first in our survey are finishing first in their races, and all seven of the candidates who placed second are finishing second.
Our October 2021 general election survey ended up being the only public poll in Seattle during the final month of the 2021 election cycle. While it was nice to see Elway/Crosscut and Strategies 360/KOMO do polls in September, we would have liked to have had another October survey to compare our data to.
With credible public polling getting harder and harder to find, what we may do next time around, resources permitting, is simply try to conduct a second preelection poll right before the voting ends. That way, we won’t have to hope that another trusted entity will come along and provide fresh data.
Polls cost a lot of money to commission, and a lot of thoughtful work goes into designing them. If you appreciate the research NPI is doing, we invite you to become a member of our organization. You can donate annually or monthly.
If you belong to an organization that would be interested in joining NPI, we offer Associate Memberships and sponsorship opportunities.
- Change Research, a Public Benefit Corporation based in California, surveyed 617 likely November 2021 Top Two election voters in Seattle from Tuesday, October 12th to Friday, October 15th on behalf of the Northwest Progressive Institute. All respondents participated online.
- Change used targeted advertisements on Facebook, targeted advertisements on Instagram, and text messages sent via the echo19 and/or Scale To Win platforms to cell phone numbers listed on the voter file for individuals who qualified for the survey’s sample universe, based on their voter file data.
- Regardless of which of these sources a respondent came from, they were directed to a survey hosted on SurveyMonkey’s website. Ads placed on social media targeted all adults living in Seattle. Those who indicated that they were not registered to vote were terminated.
- As the survey fielded, Change used dynamic online sampling: adjusting ad budgets, lowering budgets for ads targeting groups that were overrepresented and raising budgets for ads targeting groups that were underrepresented, so that the final sample was roughly representative of the population across different groups.
- The survey was conducted in English, and has a modeled margin of error of 4.1% at the 95% confidence interval.
Additional information about the methodology is available here. Finally, if you missed it, be sure to read this Q&A with Change Research’s Ben Greenfield about how to read assess the general election poll findings that we released.