Sunday, November 7, 2010
Quan Probably Was Leading the Entire Time, And She Has To Be the Odds-On Favorite
Politics / People Robert Gammon — Sun, Nov 7, 2010 at 9:30 AM
Statistically, Councilwoman Jean Quan will be tough to beat in the Oakland mayor’s race. For either Don Perata or Rebecca Kaplan to move past her in the ranked-choice tabulations, then the ballots yet to be counted will have to be much different from those that have already been recorded. Such an occurrence may be possible, particularly because we don’t know what parts of the city the late ballots came from. But it’s not likely.
Here’s why:
Let’s start with Perata. Looking at the numbers, the ex-senator is currently 2.18% behind Quan in ranked-choice tabulations, a difference of 1,876 votes. That may not sound like a lot, but there are probably only about 10,000 votes left to count in Oakland. The county registrar said Friday night that there were about 15,000 ballots total to count from Oakland, Berkeley, and San Leandro. And so Oakland’s share of those is probably about 10,000.
That means Perata would need to beat Quan by at least 19 percentage points on the remaining ranked-choice ballots. It’s a long-shot. The reason is that Quan beats him by a wide margin on second- and third-place votes. So far, she has picked up 20,051 of them, compared to his 9,176 (most of hers come from Kaplan supporters, as we noted Friday night). In other words, Perata will have to pick up a large percentage of the remaining first-place votes, while hoping that Quan won’t continue to pummel him on second- and third-place selections on the uncounted ballots.
Again, that would require the remaining ballots to be much different than the ones already counted. They’ll have to favor him by a huge margin. It’s possible, but unlikely, because in the post-election counting, Perata’s numbers have been shrinking. As ballots have been counted since Election Night, the ex-senator’s percentage of first-place votes has dropped from 35.20% to 33.96%. Quan’s, by contrast, have inched up from 24.30% to 24.64%, a swing of 1.58 percentage points in her favor. So not only does Perata need that trend to reverse itself, but he needs it to shift back in his favor in a big way.
Indeed, Kaplan may have a better shot of overcoming Quan than Perata does. Here’s why: Before the ranked-choice tabulations show her being eliminated in the penultimate round, Kaplan is only 2.26% behind Quan, a difference of 2,066 votes. That means Kaplan needs to beat Quan by about 23 21 percentage points on the remaining 10,000 ballots.
It’s a lot, but Kaplan possesses at least one advantage that Perata doesn’t: Joe Tuman. The results show that when Tuman is eliminated in the ranked-choice balloting, Kaplan picks up far more votes than Quan: 4,361 to 2,818. By contrast, Quan and Perata basically split Tuman votes 2,818 to 2,724.
But even with the Tuman boost, Kaplan still needs to beat Quan by a wide margin on the remaining first-place votes. And again, that means the uncounted ballots will be have to be much different from the ones already tabulated. The good news for Kaplan is that her first-place numbers have moved up since Election Day, rising from 20.82% to 21.48%. But, remember, so have Quan’s. That will have to change significantly for Kaplan to pull it out.
In the end, Kaplan and Perata’s only real hope is that the remaining ballots come from their electoral strongholds in the city. For Kaplan, that likely means Temescal and uptown/downtown. For Perata, it would likely have to be East Oakland and Fruitvale. Although he does well in the hills, Quan does, too, because it’s her council district.
Finally, let’s address the issue of whether Friday night’s preliminary ranked-choice results really did show a “turnaround” as it’s been portrayed. It sure seemed that way, because until then, we only knew that Perata was beating Quan among first-place votes by a significant margin. But looking at the ranked-choice numbers, specifically how the second- and third-places have broken in a big way for Quan, it now seems likely that she would have been labeled the frontrunner from the beginning if the county registrar had decided to run the tabulations on Election Night. By waiting until Friday, it made it seem that Perata was in the lead, when in fact, Quan probably was.
Having said that, Friday night’s results were surprising (as they would have been if known on Election Night). It’s hard to believe that anyone could have credibly predicted that Quan would trounce Perata so thoroughly on ballots that had Kaplan as the first-place choice. It even shocked Quan.
The reason is that a September poll commissioned by the chamber of commerce appeared to show that Perata would get a boost if Kaplan were out of the race. In short, it indicated that more Kaplan supporters would pick Perata than Quan as their second choice. But either the poll was flat-out wrong, or Kaplan supporters shifted sharply away from Perata in the final six weeks of the election and went heavily for Quan instead.
Comments (8)
Showing 1-8 of 8
Good point, the "Quan overtakes Perata" coverage seems to have missed the fact that "Anybody but Don" was alive and well. We'll know for sure today, so that's something. Good article Robert.
like
dislike
Robert:
My maybe inartfully expressed point is that the composition of the Kaplan vote at 12% is very different than the composition at 21%. And very different from the composition of the Kaplan vote at 29%, after the Tuman dump. I think the dynamics of the race changed over the 6 week period for certain, as the "anybody but Don" vote congealed. But I also think the 17 point difference between 12% and 29% were voters mostly voting for Kaplan, Quan, Tuman in various rankings. It's not surprising that they would have overwhelmingly gone to Quan.
I will be very interested in the post-campaign discussion of mistakes, given how razor close this has become.
For one thing, has anyone discussed or noticed how Perata's push for people to only list him and no one else might have actually hurt him with 2nd and 3rd place voters? If you were considering Perata, but say you wanted Tuman or Kaplan first, if you bought the logic of the Perata campaign, wouldn't you just have ONLY voted for Tuman or Kaplan and no one else (since you liked them best), and excluded PERATA when otherwise he would have received your 3rd place vote (which of course would be a full vote for him at this final stage of the process!)?
That's another reason I think this campaign was handled so poorly by the Perata folks. Given the money he raised, I'd give them an "F" on execution. If he was going to get a 1st place vote from the voter anyway, what freakin difference should Perata care who that person puts 2nd or 3rd, even if it's Tuman or Kaplan or even Quan? The fact they apparently did very much care only convinces me more of what numbnuts they are, as they should have been doing what Quan did, and courting Tuman and Kaplan voters hard instead of spending precious 'persuasion capital' convincing their own #1 placement voters to not list anyone else. Perata's argument implicitly said "put me first or don't vote for me." Well, he should have been more careful of what he asked for; indeed, I think his own moronic 'strategy' is what killed him if indeed he loses. I'm convinced that if he had done more outreach to Tuman and Kaplan voters, given more reasons for his positions and explanations for why he'd be a good choice in his ads (instead of just 'Believe in Oakland' repeated ad nausium), and participated a little more in forums and various campaign events, even with all the problems that have been highlighted, I think he'd have picked up enough 2nd's and 3rd's to have clinched it. It almost seems his own people really fundamentally misunderstand ranked choice voting and its implications, which is odd if this is their profession, because it really just isn't that conceptually challenging an idea and it's been done before. mfraser
Parke,
I'm not sure that's correct. The chamber poll showed for first-place votes:
Perata 26%
Quan 22%
Kaplan 12%
Undecided 25%
If the poll was correct, then that would indicate Kaplan picked up the most undecideds, since she now has 21.5% first-place votes. Perata picks up almost as much, since he now sits at 34% first-place votes, with Quan only getting about 2.5% of undecideds.
That means Quan voters didn't swing to Kaplan, listing Kaplan first and Quan second -- at least in not any significant way. And even if most of the undecideds that Kaplan picked up listed Quan second on their ballots, it's not enough to explain the huge swing to Quan from Kaplan voters. It still looks like either the poll was wrong, or Kaplan people turned away from Perata and went to Quan in the final six weeks.
Daniel,
I'm sure you're right about Tuman votes going to Kaplan and then to Quan. But remember, there were only a total of 4,361 votes that could have done so (Kaplan's total take from Tuman). But Quan picks up more than 15,000 votes from Kaplan if she's out. That means more than 10,000 Kaplan supporters picked Quan over Perata. At minimum, it's a 2-1 margin, since Perata only gets about 5,000 from Kaplan.
I think a lot confusion stems from sloppy reporting on IRV. It is not just the people who put Kaplan first that transferred to Quan, but probably also those who ordered their votes Tuman then Kaplan then Quan.
Re: your last sentence. The other, and likely correct, explanation is that the Kaplan electorate 6 weeks ago was smaller than the Kaplan electorate on Election Day. Kaplan surged late (with a late large IE from the Nurses and some good endorsements), moving some Quan #1, Kaplan #2 voters to Kaplan #1, Quan #2. But, thanks to RCV, they all came back.
Author Archives
-
Jerry Brown’s Optimism Has Been Costly
- May 18, 2012
-
A’s Planned Move to San Jose Looks Dead
- May 17, 2012
- More»
