Instant Run-off Voting (IRV for short) has made a big difference in this year’s Oakland mayoral election, and it’s not over yet. Under Oakland’s traditional primary -> run-off system, there would have been a primary election in June, with lots of money spent. The two top candidates (in this case, Don Perata and Jean Quan) would then have faced off in a November run-off, with a lot more money being spent in that election. Essentially, the election could well have boiled down to who was better at raising the huge amounts of money needed to fight two back-to-back electoral battles. In the past, that’s usually meant that the candidate with better connections to big-money special interests wins.
(Parenthetically, in very old-style elections, there was no run-off. The candidate with the plurality of votes cast won outright. This led to people putting up phony candidates whose main purpose was to draw votes away from the prime opponent. Having a run-off at least allowed a clean match-up of the top candidates.)
With IRV, only one election is held, but the voters get to pick more than one candidate. (IRV is also called ranked choice voting.) If there first choice gets knocked out of the running, their vote transfers to their second choice, etc. Thus, unless all the voter’s choices are eliminated, the voter’s voice still makes a difference. Here’s how it played out in the Oakland mayoral election this year (the first time IRV was used in Oakland).
There were ten – yes, ten, count them – candidates in the elections. Some of them were pretty minor and ran only token campaigns; others tried to run low-budget grassroots campaign — a very hard thing to do in a city of over 400,000 people. Most observers acknowledged there were four “major” candidates with sufficient funds and supporters to run credible citywide campaigns — former state legislator Don Perata, city council members Jean Quan and Rebecca Kaplan, and professor and media pundit Joe Tuman. Shown below are the preliminary results of the IRV calculus. (The results are preliminary because several thousand absentee and provisional ballots remain to be tallied.)
As you can see, the IRV procedure shifted votes from lower ranking to higher ranked candidates. Significantly, however, the shift was nonrandom. About two-thirds of Rebecca Kaplan’s votes shifted to Jean Quan when Kaplan was eliminated, while only about one-third went to Perata. Even more striking (although smaller) was the shift from Green Party candidate Macleay. Quan got almost half of his 1500 or so votes, while Perata only got a little more than fifty. The unequal distribution of second and third choice votes allowed Quan to close the gap against Perata and eventually surpass his total. One potentially important factor in the election was candidates’ recommendations about whom to vote for other than themselves. Several candidates, notably Quan and Kaplan, urged their supporters to choose any other candidate except Perata as a second or third choice. It appears many voters followed that advice.
Back when ranked choice voting was being debated, Perata, and several city council members who supported him, came out in opposition to IRV. At this point, it’s pretty clear it was in his self-interest to do so. If not for IRV elections, it’d almost certainly now be Mayor Perata, rather than Mayor Quan.