BLOG | A week later: who’s our mayor?

November 9, 2010

Oakland residents are still waiting to hear who won the mayoral election of 2010. The outcome will be particularly¬†titillating as Jean Quan has gained a slight lead over Don Perata due to Oakland’s new ranked-choice voting system.

Here’s how it works:

Oakland voters were asked to choose their top three candidates for mayor. If no one candidate receives 50% or more of the vote, the last place candidates are dropped from the running, and their votes are distributed among the last two candidates standing. This is to say, those who voted for a losing candidate as their first choice will have their second and third choice votes will be given to those candidates.

It’s confusing, I know. But it explains why Quan has been able to overcome Perata’s previously strong lead.

Once Rebecca Kaplan was dropped from the running, the second place rankings on her first place votes were distributed. Most people who voted for Kaplan marked Quan as their second choice, explaining why once Kaplan dropped out, Quan received many more votes in her favor.

This system poses an entirely different strategy for politics. Quan and Kaplan succeeded in working this system by somewhat working together with their “anyone but Perata” campaign. By teaming together, they ensured that no matter which on of them made it to the final round would have the backing of the other’s second place votes.

And yet, we still don’t have a mayor a week after the election.

What were your experiences with the ranked-choice voting system in Oakland (or in other elections)? Is this system worth the lengthy wait for a concrete decision? Is Quan’s current lead surprising to you? Let us know what you think by leaving a comment below or contacting

BLOG | A week later: who’s our mayor? was published on November 9, 2010 in Blogs

Print this page Print this page

  • Rob Richie

    The delay is entirely due to completing a tally of the absentee and provisional ballots. Ranked choice voting ballots are scanned just as quickly as non-RCV ballots. The RCV tally is basically about pushing a button. If all the ballots were scanned, the final RCV tally could be run. An update on the RCV tally could be run at any time. Note that the statewide AG’s race is not decided for the same reason.

    Of Oakland voters indicating a preference in the mayor’s race, 99.8% (meaning 998 out of every 1000 voters) cast a valid ballot that counted in the first round.

  • DaveMM

    The delay is DIRECTLY related to RCV.

    Before running a RCV program, ALL VOTES must be in hand, and certified. If not, the results can change due to the ENRON style of vote manipulation.

    An example of how this happens can be seen on this youtube video: “The Case of the Missing Ballots”

    Unlike a normal elections, results can be released early and a winner determined, ie: Candidate A and B are in the lead. The difference is 6,000 votes between them. ROV says there are 4,000 provisional/absentee votes that still need to be counted before he certifies the results. Nevertheless, It is clear candidate A will win. Everyone understands that. That’s the beauty of the traditional elections. Some close races, where the difference is smaller tan the uncounted, then obviously the results will be uncertain. This is very rare, though when it happens, is much talked about.

    In fact, the ROV broke the memorandum of understanding with the cities. It says “the ROV shall not, under any circumstances, provide and RCV election results until all of the ballots have been tallied, including, but not limited to vote by mail, provisional, and early voting ballots. The ROV will provide only final elections results once all of the ballots have been tallied”

    Had they followed the agreement, everyone would still be waiting for ANY information on the RCV results.

    Concerning if Oaklanders understood the system, which is different than indicating a preference, an article with full data on how people voted can seen at this web site.

    Seems to me like people have a half full/half empty argument depending upon if you like RCV. RCV supporters are rather emotional about their system and aggressively defend it. In regular runoff elections, error rates are more like 0.01%, 9999 out of ever 10,000. RCV traditionally shows 10X to 100X errors.