You don't. What Guerra did was send a bunch of undergrads to polling places in the November election, who asked voters how they will vote in March. Only problem with that is that turnout in the presidential election was 70%, while turnout in March will be more like 30%.
Brianne Gilbert, one of the researchers, said they did not try to screen for likely March voters. "We did not ask any screeners since this was not a poll dedicated to the mayoral election," Gilbert said via e-mail.
Needless to say, that's important. And that makes the topline results -- in our view -- unreliable. About the best you can say from this is that Garcetti and Greuel are the two front-runners. But we knew that already.
The results are a bit more interesting when you get into the crosstabs. Even if the unlikely voters aren't screened out, you should still be able to get a sense of the candidates' relative strengths among various demographic groups.
It appears that the prototypical Garcetti voter is a young, liberal, agnostic, half-Asian, half-Latino male who just moved to L.A. and rents an apartment somewhere outside the Valley. If, on the other hand, you meet an older, white, Valley-dwelling woman, then you can bet she's voting for Greuel.
Jan Perry, the only African-American in the race, does best with African-Americans. Kevin James, the only Republican, does best with Republicans. But again, we knew that.
One last interesting nugget that might even be true. According to this poll, Mayor Antonio Villaraigosa's endorsement would hurt the candidate who received it. It would help slightly with blacks and Asians and a lot with Latinos. But it would also hurt a lot among whites, and whites will make up about half the electorate. That makes the endorsement a slight net negative overall.
A private poll we posted in October showed Greuel slightly ahead among all voters
, and that's what we're inclined to believe until we see better data. That poll also showed Greuel with a much bigger, 2-to-1 advantage among the "likeliest" voters. We're much more skeptical of that, because it defined "likely voters" as anyone who voted in three of the last six primary elections. But in order to have voted in three primary elections, a voter would have to have been in L.A. and of voting age for at least 12 years. Plenty of 20-somethings and new L.A. residents will vote in March, and according to LMU's poll, they'll tilt toward Garcetti.
At this stage, the hardest thing to do is to predict who will actually show up to the polls. In part that's because a lot of people who will vote don't know it yet. They don't know who's running. They haven't thought about the issues. They haven't seen the ads, and we don't really know who will have the money to pay for them. Once those things start to come into focus, it'll be a little easier to judge.
For now, numbers offer only the illusion of accuracy. We can be safe in saying that Greuel and Garcetti are atop the field, and Perry and James are somewhat behind. We're also pretty confident -- not just from that leaked October poll, but from other sources as well -- that Greuel is slightly but perceptibly ahead of Garcetti. But that doesn't mean too much at this point, since most voters are undecided and the campaigns haven't begun.
Speaking of which: the Democratic Party of the San Fernando Valley is holding its endorsement meeting Monday night. That's Greuel's home turf, but the Garcetti campaign has been contesting it heavily in hopes of denying her the 60% threshold required for the endorsement.
Greuel, it turns out, was a registered Republican during the Reagan years, in her 20s. She re-registered as a Democrat long before she ran for office. This is likely to come up, if it hasn't already, in the party endorsement process as evidence that Greuel is not a lifelong, committed Democrat.
Agi Kessler, the chair of DPSFV, said she had not heard that Greuel was once a Republican. But, she added, "I would find it hard to believe it would make a difference based on her record."
Eric Bauman, chair of the L.A. County Democratic Party, said he was also unaware of Greuel's youthful indiscretion, but added that "I doubt anyone doubts her Democratic bonafides." Just last week, Greuel picked up the endorsement of liberal Democratic Senator Barbara Boxer.
With three Democrats in the race -- Greuel, Garcetti and Perry -- it might be difficult for any one of them to get the delegate supermajority needed for the party's endorsement in the primary.
Update, 9:28 a.m.: Not even the Garcetti campaign thinks it's ahead. In a memo on the LMU poll, Garcetti's campaign calls it a "dead heat race."
The campaign also claims it has "grassroots enthusiasm" because Garcetti has more Facebook friends and Twitter followers, because 200 people showed up to the opening of his campaign headquarters on Saturday, and because he got more signatures on his nominating petition. Straws in the wind.
Update 2, 11 a.m.: The Guerra poll has a sample that is 30% Latino. For reference, when Antonio Villaraigosa ran in the May 2005 runoff, he set a Latino turnout record of 25%. Granted, there are more Latino voters now than there were then. But not even the Garcetti campaign thinks Latino turnout in the March primary will be 30%.
"Minimum 25, possibly higher," was all that Bill Carrick, Garcetti's consultant, would predict in a conference call this morning.
Garcetti's consultants held the conference call to tout the outcome of the Guerra poll, but they were also careful not to embrace his methodology.
"No one is claiming this is the end all, be all of next year's mayor's race," said Fred Yang, one of Garcetti's pollsters. That's for sure.
Yang did say that the campaign conducted its own internal poll over the summer, which showed Garcetti "slightly ahead of Ms. Greuel." From the sound of it, Garcetti's internal poll was more methodologically sound than Guerra's. But we can't scrutinize it because the Garcetti campaign won't release it.
Carrick also said the Garcetti campaign is working to "expand the traditional primary universe," in part by taking advantage of online registration, which suggests that the traditional primary universe isn't that favorable for them.
Update 3: Eric Hacopian, Perry's consultant, dismisses the poll:
"This is at best a poorly executed exit poll with extremely flawed methodology. The results of an exit poll of November voters, which does not screen for voter history can not be taken seriously by any one who has ever done a poll or run a campaign."
But tell us how you really feel.
Update 4, 12:30 p.m.: John Shallman, Greuel's consultant, is similarly dismissive:
"I hope Eric believes the poll. We like running against misinformed opponents. The fact ultimately is that his campaign is grasping and needs to grasp for any good news they can at this point. The best news they can grasp onto is a poll that shows them losing at the end of the day... He'd probably celebrate a survey of 4 out of 5 dentists showing him he's ahead."
For good measure, here's Shallman's prediction for Latino turnout: 21%.
First posted at 10:42 p.m. Sunday.