What our poll can teach us after election outcomes

We were told this couldn’t possibly be accurate because the primary results were so different from our poll results. In August, in a field of 15 candidates, Harrell captured 34% of the vote, while Gonzalez earned 32% — a much closer contest.

I answered the criticism on Twitter at the time, in this way:

It’s illogical to compare outcomes in the primary with a poll of people who are likely to vote in the general election and expect the numbers to be identical. First, more people vote in the general election than in the primary. And second, they are basically different groups of people with different contests in front of them.

The primary results, the poll results and the general election illustrate how three different groups of people answer similar — but different — questions. We didn’t claim to predict the future or the outcome of the election. In the end, the election is the election, and the poll is just a snapshot in time.

But I had a feeling this poll would prove prescient. And not just because I knew the science behind it was solid or that I trusted our pollster, Stuart Elway, to do his best to interview the people who would most closely represent the electorate.

Other polls in the weeks following ours showed similar results, but with a shrinking group of undecideds. Polling gave the campaigns some clear hints about what people were thinking and why.

“It seems in our poll that the outlines of the campaign were pretty clear,” Elway told me a few days after the Nov. 2 election. The longtime Washington pollster added, however, “Recognizing a challenge and meeting it are quite different things.”

In the end, our poll and the other independent polls that followed did offer a powerful glimpse into the election’s eventual outcome. As the mayoral campaigns continued, the margin grew and Harrell won with nearly 58% of the vote, compared with Gonzalez’s 41%.

Elway said we had some help getting this right, in addition to carefully designing the poll and balancing the participants to make sure they mirrored the group expected to vote in November.

Both Harrell and González were both fairly well known in advance, so that eliminated some unknown factors. In one particular race, our poll was not predictive. Before the primary, candidate Kenneth Wilson, running for Seattle City Council Position 8 against incumbent Teresa Mosqueda, had very little name recognition. About 33% of survey respondents told our pollsters they planned to vote for Mosqueda, with just 17% supporting Wilson and 39% undecided. Mosqueda did end up winning, but by a smaller margin, 59% of the vote to Wilson’s 40%.

Elway found it interesting that in the mayoral election, the undecided voters eventually mirrored those who had already made up their minds in September. He credits the campaigns for delivering on that prophecy by staying on one basic path.

“The campaigns really didn’t alter that fundamental current,” Elway said. People weren’t happy with their city government, as previous Crosscut/Elway polls had shown. Harrell, who served on the city council from 2008-2019,  campaigned as an outsider coming in to change things, even though he had been off the council only for a few years himself. That approach appealed to voters and he won.

This story was first published in Crosscut's Weekly newsletter and written on Friday. It has been updated with the near final election results, as of Monday. Want to hear more from journalists like Donna Gordon Blankinship? Sign up for the newsletter, below.