Oakland’s mayoral election

Concurrent with the statewide general and national midterm elections earlier this month, the city of Oakland, California, elected its mayor. For the first time, the city used the Alternative Vote (instant runoff). The result was a bit of a surprise, in that longtime Democratic Party powerhouse Don Perata was expected to win. However, he lost to Jean Quan, who will be the city’s first woman and first Asian-American mayor.

From the San Francisco Chronicle:

When first-place votes were initially counted after the Nov. 2 election, Quan had just 24 percent, and Perata had 35 percent. But Quan proved to be a more popular second and third choice among supporters of the other eight candidates, and in the end, she had 51 percent to Perata’s 49 percent.

This is, of course, exactly how the system is supposed to work: ensure the election of the majority-supported candidate in the event that the candidate with the most first-preference votes is short of 50%+1 of the total votes cast. But the Perata camp is not amused. His political consultant called ranked-choice voting “an injustice” and the result a “travesty” because his candidate won 78% of the precincts, and led by a margin of 10 percentage points.

“In any other contest, it would be a landslide win, not an election loss.”

Normally, even in the realm of first-past-the-post elections, we do not think of plurality candidates with 35% as landslide winners, but Perata himself said:

“I don’t understand how ranked-choice voting works.”


Just as obviously, Quan did.

Quan had been campaigning for months for people to vote for “anybody but Don.” She had told supporters to list City Councilwoman Rebecca Kaplan as their second choice.

Kaplan, in turn, told her supporters and others to list Quan second or third.

The strategy paid off for Quan when Kaplan, who finished third, was eliminated and her votes redistributed. Quan won 75 percent of them – pushing her from a 10,372-vote deficit to a 2,058-vote victory.

Perata, the Chronicle notes, never told supporters whom they should list second or third. And, apparently, never appealed for any other candidates’ second choices. Stupid strategy, given the electoral system.

8 thoughts on “Oakland’s mayoral election

  1. I’m not surprised that Perata himself didn’t understand the IRV format, but I am shocked that his campaign staff evidently didn’t understand how to mount an effective campaign for an IRV structured system.

    A mayoral race for a city the size of Oakland should draw experienced political operatives to the campaigns of the top tier candidates. Clearly the Perata campaign staffers never took your PMP class or an equivalent!

  2. I get the general impression that one can only rank three candidates.


    If you look on the news report above here.

    Why can’t one just rank all the candidates until they exhausted? Is California so behind when it comes to voter machine technology? That is the flaw of this implementation of RCV.

    I think that voters should be allowed to rank just 1 candidate if they want, but much more than three if they want as well, and ranking all the candidates on the ballot.

    It seems to me that some jurisdictions allow variations of the Alternative Vote.

    1. Preferences are optional, just rank one. (Queensland, New South Wales)
    2. Preferences are limited, you must rank 1, 2, 3, but it is optional to rank more. (Papua New Guinea)
    3. Preferences are mandatory, just rank all the candidates on the ballot. {Even if there are 23 candidates running, thus I will copy the How to Vote Card, my party gave me} (Australia Federal Elections)
    4. Preferences are optional, just rank one, but you can’t ranked beyond preference number 3. You can’t rank a 4th preference and so on. (U.S implementation of RCV at the local level) I get the general impression that this is the Oakland system.

  3. Perata understands the Alternative Vote perfectly well. His statements to the contrary are not made in good faith. Some insight into why he doesn’t like it can be found in this BeyondChron article on his political biography. Short version: he’s gotten used to winning because the opposition splits its votes. In addition, he has a lot of campaign money, making delayed runoffs less of a problem for him than for opponents like Quan. Finally, delayed runoff encourages negative campaigning.

    Of the numerous AV contests in San Francisco, Oakland, Berkeley and San Leandro on November 2, by far the most interesting to election methods geeks was the 10th Board of Supervisors district in San Francisco. Of 21 candidates, the plurality “winner” got barely over 12% of the first choice votes. Another 4 candidates got between 11% and 12% each, and 3 more got between 6% and 8%. The eventual winner started out in third place after the first round. Of the ballots counted in the first round, almost 54% did not name either of the two finalists and ended up in the exhausted pile. Voters can rank at most three candidates, a voting machine limitation which contributed to some unknown extent to creating all those exhausted ballots.

  4. I’ll take Bob’s point to be that what Perata understood was a different game: how to undermine IRV–something DC also alludes to.

    But if his objective was to win this election, he followed a strategy that was doomed to defeat.

    I was not aware that Burlington had abandoned IRV. When did that happen?

  5. Well it wouldn’t be the first time a group of overconfident people with robust campaign coffers tried to play football on a baseball diamond. (Northwestern at Wrigley Field)

  6. #5. The Burlington repeal vote was on March 1 of this year. Those who voted against AV and for repeal appear to have been motivated (at least in part) by opposition to the Progressive Party mayor who was elected twice in AV elections. I gather that he was involved in some kind of corruption scandal, but I don’t recall any details. Armchair social choice theorists make a lot out of the fact that, the second time he was elected, he would not have been the winner of a Condorcet count. But that fact doesn’t appear to have motivated many voters.

    MSS is, of course, right about Perata’s campaign strategy. I suspect that Perata was overconfident about his own chances.

    My main point was really that politicians who claim not to understand AV are — I hate to say it out loud but I have to — lying through their teeth. It’s their not-very-subtle way of claiming that AV is too complicated. Another example is San Francisco mayor and soon-to-be Lieutenant Governor Gavin Newsom, who voted for the same candidate as his first, second and third choices and made sure that this factoid got picked up by reporters. (At least according to San Francisco and Oakland rules, such ballots do count for the one candidate named.) I’m sorry to harp on this, but it makes me boiling mad.

    #3. The only optical scan voting machines currently certified for use in California are limited to three rankings for technical reasons to obscure to go into here. That is a very important limitation (see San Francisco District 10, where over half the first-round ballots became exhausted before the winner emerged). It’s going to be an uphill battle to overcome this limitation — just as it’s an uphill battle to overcome many other limitations of voting machines and election procedures.

  7. > “he would not have been the winner of a Condorcet count”

    Interesting that the voters can use the straight-majoritarian process of a referendum to “correct” what they view is the non-majoritarian result of an election. (Assuming that the electorate’s composition doesn’t radically change between the “wrong” election result and the referendum, so that generally one can talk about the “same” voters).

    Of course, plurality is no less likely to elect the “wrong” candidate. One wonders how FPTP for the Commons would have fared after 1983 or 1987, if British voters had been able to demand a citizen-initiated, majority-nationwide referendum on the electoral system. (Caveat that electing a mayor is by the same polity-wide majority that passes a referendum. Perhaps a better analogy might be: Would the US Electoral College have survived long after 2000 if it could have been replaced by nationwide I&R?)

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.