Young Voter Turnout Under — or Over — Estimated in California Polls?

They also have exit polling in France. This man did not appear to wish to answer. (Stephen Rees: Flickr)

They even have exit polling in France. This man did not appear to wish to answer. (Stephen Rees: Flickr)

This election, KQED has focused in part on young voters and their views on different issues.

So how much of a role did young voters play in Tuesday’s electoral outcomes? It’s hard to say exactly or — as it turns out — even approximately.

The respected Sacramento political newsletter The Nooner Wednesday pointed to an exit poll showing 18-to-29-year-old voters made up 27 percent of California voters in yesterday’s election. That’s compared to 22 percent in 2008.

Yet the highly regarded Field Poll had predicted on Monday that a maximum of 15 percent of these young voters would turn out.

So what gives? I called Field Poll Director Mark DiCamillo, thinking that he would explain how pre-election predictions can run awry. Instead, he walked me through the exit poll industry in California — and what an interesting trip that was.

He explained that there’s one exit poll in California, and it’s used by all the major networks. DiCamillo, too, has used it widely in past years. But he told me had just gotten through reviewing the data set, and he “doesn’t have confidence” in the poll’s findings this year.

Perhaps the easiest discrepancy to grasp is this one: Proposition 30, the governor’s education tax initiative, passed in Los Angeles County by 20 points, but DiCamillo says the exit poll’s final estimates showed Prop. 30 winning LA County by only 6 points.

In an email to KQED Election Editor Tyche Hendricks, DiCamillo added the following:

Another finding that I find implausible is their estimate that just 55% of the state’s voters in this election were white non-Hispanic. This would be a steep decline from previous CA presidential elections, and from what most pre-election polls were showing. They also show Latino voters supporting Prop. 30 by just 2 points (51% to 49%), which is less than its statewide margin. Every pre-election poll that I’ve seen had Latinos on the yes side by margins of two to one or greater.

DiCamillo stressed to me that his preference is “not to pick fights with his fellow pollsters.” But he’s looked carefully at the exit poll data in many ways, he told me, and concluded, “This is not California.”

Mid-afternoon Wednesday, I called Edison Media Research, which conducted the exit poll. They’re in New Jersey and I was told the main contact for the California exit poll was not in. As if Hurricane Sandy wasn’t bad enough — now it’s snowing there.

DiCamillo was up front about the possible flaws of the Field Poll’s pre-election predictions on young voter turnout. “I don’t want to say ours is definitive,” he said. Online registration was implemented in September, pollsters are not able to sample the newly registered. Almost half of the more than 700,000 people who registered online were young voters.

It’s “reasonable to conclude,” DiCamillo promptly said, that the Field Poll underestimated the young vote, and those young voters may have contributed to the passage of Proposition 30.

So when can we expect definitive information? Not anytime soon.

All that California voters have to provide when they register is name, address, date of birth and political party (or “no party preference”). “Those four things are probably retrievable,” DeCamillo says, “but they won’t be for months.”