CA GOP On Knife Edge; USC Pollsters, LAT Weasel
California Republicans stand at a crossroads: Will they join the ranks of voters in Kentucky, Georgia, Oregon and other states who have chosen Establishment conservatives over Tea Party knuckledraggers? Or will they add California to the roster of states like Nevada, Missouri and Illinois who chose teabaggers who later got skunked by Democrats.
Until this weekend, polls were suggesting that Assemblyman Tim Donnelly, the Tea Party “patriot” was on his way to becoming the GOP standard bearer by defeating former U.S. Treasury operative Neel Kashkari, the Establishment’s choice, in the open primary.
Neither of them has a prayer against Gov. Jerry Brown in November. No non-incumbent Republican has been elected governor of California after a primary election since Pete Wilson won against token opposition nearly a quarter century ago.
But the USC Dornsife/LA Times poll University of released over the weekend found Kashkari slightly ahead of Donnelly in the race for second place and the right to lose to Brown in November.
“Among likely voters in the primary election, Democratic incumbent Brown has 50 percent of the vote, compared to 18 percent for Kashkari and 13 percent for Donnelly, with 10 percent of likely voters still undecided,” USC/LAT reported.
Fun with numbers: So at odds with other recent polls was the result, that the survey’s directors and the LA Times itself
weasley weasely called the race for second place “a statistical tie” and a “dead heat” – which they did because their finding was so close to the margin of error for likely voters in their poll.
In other words, the pollsters weren’t confident enough in their own results (or their model of likely voters*) to assert that Kashkari has surged ahead of Donnelly. “It’s too close to call, but Kashkari has some momentum going into the final stretch,” Dave Kanevsky of American Viewpoint, the Republican firm that conducted the poll along with the Democratic firm Greenberg Quinlan Rosner Research, told the LA Times.
And our friends Seema Mehta and Michael Finnegan (or their editors) at the Times chose not to buy into the survey’s actual findings, hedging their bets, writing: “The difference between the two vying for the second slot in the general election was within the poll’s margin of error.”
But according to the results released by USC, Kashkari’s 5-point lead among likely voters is actually just outside of the poll’s margin of error of +/- 4.4 percent for likely voters.
The race could be called a dead heat if you were looking at all the registered voters in the survey, where the results had Kashkari at 13 percent and Donnelly at 12 percent with a margin of error of +/- 2.9 percent. But then, you wouldn’t be relying on the voters you expect to be a part of the final tally. Moreover, among those who told the pollsters they’d already voted, Kashkari led 15-12 percent.
If you’re going to reject your own poll’s findings, you really ought to explain why.
Bottom line: Most Establishment Republicans are hoping and praying for Kashkari to win second place on Tuesday so they can avoid the inevitable investigative story that would follow a Donnelly victory – into whether their candidate actually has an opposable thumb.
*P.S. On Monday, after this post went online, our pal Timm Herdt at the Ventura Star, posted an item that came out of a media call the pollsters did with reporters that may shed some light on why the Times was hinkey about standing by their poll findings:
…predicting voter turnout in an election that most analysts believe will approach or exceed record-low turnout is a fairly dicey proposition. In a conference call with USC’s pollsters today (Scott Tiell of the Democratic firm Greenberg Quinlan Rosner Research and David Kanevsky of the Republican firm American Viewpoint), they noted the poll’s “likely voter” universe represented 41 percent of their sample of 1,511 registered voters. Given the rate of mail-in-ballot returns so far, no serious political observer expects actual turnout to even approach 41 percent of registered voters. The consensus is 30 percent, plus-or-minus a couple percentage points. The pollsters did the best they could, as they counted as “likely voters” only respondents who had voted in at least one of the two most recent statewide primaries and who said they had either already voted or were almost certain that they would vote.
“Still, it seems likely the poll counted a fair amount of people who won’t actually vote. Perhaps the most instructive number came from the response of those who said they had already voted, and it showed a virtual dead heat: 15 percent for Kashkari, 12 percent for Donnelly.
I predict a Tom Bradley outcome for Kashkari. I think he will win the most votes tomorrow, but because he got out of the gates so late, the absentee ballots will undo him. If Donnelly wins, candidates in about 10 races will have to think long and hard about how they approach Latino voters. They can break hard with Donnelly and risk losing Republican votes or ignore him (and the taunts from their Democratic opponents) and risk losing the Latino vote. Tough situation.
Or you could point out that one of the “independent” analysts USC used for this poll has been anti-Donnelly from the get-go:
“If Tim Donnelly does get the nomination, it could wipe out all of the gains that the establishment Republicans have made nationwide, because the fringe candidates undoubtedly color the discourse,” Mike Madrid, a Sacramento-based Republican strategist, told The Daily Beast. “We know it’s not going to be helpful [if he gets elected]. The question is, how harmful [would it be]?”
Is it evil for Democrats to vote for Donnelly?
Not evil per se, just impious.
Call me jaded, but if Kashkari acutally wins, how long will it take him to pander to the right and say something that undermines any credibility to the claim that “mainstream” Republican policies are somehow moderate?
For God’s sake, the spokesperson for the “Donnelly is dangerous” campaign is his father-in-kind Pete Wilson.
@Andy Since Kashkari invested $2+ millions of his own cash in this election, his “mainstream policies” are those of his own making, as is obvious on inspection, and not of those of the Party.
Doesn’t a MOE of 4.4 mean Tim could be as high as 17.4 and Neel could be as low as 13.6? (Does MOE apply to each candidates’ # or the margin between them?)
Eli, you’re right that MOE applies to each candidate’s percentage. But you’re digging into the statistical weeds here. The +/- 4.4% MOE actually applies to Brown’s 50%. The MOE for Kashkari’s 18% and for Donnelly’s 13% would actually be +/- 3%. The finding of a 5% difference between them would be statistically significant at the 95% confidence level.