At 6:11 p.m. yesterday, before the polls closed, I wrote that our exit polling suggested Toni Preckwinkle had the Cook County Board president's race locked down.
And I was right. Our survey honed in on Preckwinkle's strong performance early in the day, and continued to highlight her lead as the election progressed.
And yet ... our poll was wrong. I predicted Preckwinkle would snag 69% of the vote, and noted that the poll had an 8% margin of error. Preckwinkle ended the day with 49% of the vote -- well outside that margin.
Such are the joys and pains of exit polling.
In many ways, we succeeded on a grand scale. We illuminated the key trends that shaped the election and provided some excellent guidance on which candidates Cook County voters would go for. And we did it starting just after lunch, even though the polls closed at 7. We provided some seriously useful, interesting and timely information.
We predicted Pat Quinn would win Cook County, and he did, 54% to 46%. We said David Hoffman was running strong in the county, and that Giannoulias was in for a tougher race than he'd bargained for. Hoffman nabbed 34% of the Cook vote, compared to 36% for Giannoulias, the odds-on favorite. And ultimately we said the gap between the two of them would be narrower than our margin of error, which it was.
It's a damn good performance, considering that we're a three-month-old news organization that's never run an Illinois exit poll. Though networks often pay tens of thousands of dollars for these things, our expenses totaled $200 -- most of which went for a $100 rental car that we needed to send our political reporter to polling places on the South Side.
We also brought a new level of transparency and currency to election coverage. We designed the survey results so they'd be a valid representation of the election at each moment of the day. That allowed us to post updates minute by minute as the survey results rolled in. We treated the vote as a football game in progress, with score updates every few minutes. All in all, it amounted to a new kind of political coverage.
On the other hand, we don't get great marks when you look carefully at a few of the numbers. I was off on the Preckwinkle forecast, albeit in the right direction. And I pegged Hoffman's Cook County tally at 51%. As noted, he finished with 34% -- a value that's way outside our margin of error.
Those results suggest some room for improvement. As I said early yesterday, we had originally planned to hit 25 polling places, but staffing considerations and logistics got in the way. We ended up with a dozen.
Not enough, when you consider that our results were undone by unpopular candidates who performed really well in a few precincts. In the County Board race, the poll showed Todd Stroger and Dorothy Brown in single digits, but together they accounted for 37% of the vote. Our sample didn't include any polling places near Stroger's 8th Ward base. A bigger sample likely would have netted more Stroger voters and a better overall result.
Similarly, our results were dead on when it comes to Giannoulias. But when you consider those who voted against him, problems arise. Hoffman polled well in some areas of Chicago and the Cook suburbs. But 23% of voters gave the nod to Cheryle Jackson, and we missed many of them, probably because of the same geographical gap that led us astray with Stroger.
We'll plan on sampling many more precincts in November, and may tweak the process to ensure that we get a representative group of polling places in minority neighborhoods. And hopefully we'll be able to broaden the poll to include precincts across Illinois.
Meanwhile, I want to highlight the contributions of staff writers Adrian Uribarri and Alex Parker, correspondent Ben Meyerson, and three Northwestern University j-schoolers who staffed polling places: Sarah Chacko, Andrew Paley and Alice Trong. They all spent many hours standing out in the cold trying to convince recalcitrant voters to take our survey.
Phil Meyer, who pioneered the use of social science techniques in journalism back in the 1960s and has led the field ever since, advised on the design of our survey, and recommended we collect data at 25 precincts. Next time I'll do what he says.
Windy Citizen 