Report card for ranked-choice voting

Steven Hill // Published December 9, 2008
What are you doing today? How would you like to be voting in runoff elections for the Board of Supervisors? That's what many would be doing if San Francisco hadn't voted in 2002 to replace the old December runoff system with an "instant runoff" system known as ranked choice voting.

Whether using ranked choice voting or December runoffs, the goal is the same: to elect officeholders with majority support from the public. But with ranked-choice voting, you accomplish this in one November election.

We now have had five elections since 2004 using ranked-choice voting to elect the mayor, Board of Supervisors and other offices, providing some basis for assessing its impact. One significant difference between ranked choice and the old December runoff has been a dramatic increase in voter turnout. By finishing the election in November when voter turnout tends to be highest (because voters are showing up to vote for president or governor), a lot more San Franciscans are having a say in who represents them on the Board of Supervisors.

For example, this year in the District Three race, 22,407 voters participated in the final round of the instant runoff, with the winner of that race having 13,316 votes. In the December 2000 runoff election to decide the same District Three seat, only 12,414 voters participated, with the winner garnering 7,202 votes. Voter turnout dropped by 40 percent between the November 2000 election and the December runoff, and surely would have done the same this year following a high turnout presidential election.

Instead, in all supervisorial races in 2008 the number of voters participating in the ranked-choice voting races was much higher than in previous December runoff elections, even when accounting for higher turnout in the 2008 presidential election over the 2000 presidential election.

San Francisco taxpayers also are saving millions of dollars by not holding a separate runoff election in December. Based on numbers released in 2003 by the Elections Commission, it costs at least $3 million to administer each citywide election.

With ranked-choice voting, San Francisco avoided a citywide December runoff for assessor-recorder in 2005, as well as for 10 supervisorial races from 2004 thru 2008. That means approximately $6 million in savings on administrative costs, as well as several hundred thousand dollars more saved on the costs of public financing for supervisor runoffs. A couple million dollars have been spent on voting equipment and education to implement ranked choice voting, but that has more than paid for itself. While democracy shouldn't have a price tag, there is no point in spending money needlessly on two elections when you can finish the job in one.

In terms of representation, the Board of Supervisors that was just elected via ranked-choice voting will be the most representative in the history of San Francisco. Seven out of 11 members are racial/ethnic minorities, three are women, the gay community is represented, and there is a range of ideological viewpoints.

Statistical analysis also shows that voters are handling the task of ranking their candidates. In 2008, San Francisco voters on average used 2.3 of their 3 rankings, with voters in the highly competitive races using slightly more, 2.5. That means most voters are using all three of their rankings, while some use only two rankings and a few only one ranking.

The number of overvotes - ballots where voters picked more candidates than they are allowed, rendering them invalid - is a good measure of voter confusion. Overvotes occur in all races, even for president or state Assembly when some voters erroneously vote for more than one candidate. While the rate of overvotes in ranked choice voting races is a bit higher than in non-ranked-choice races, it still has been low, generally less than 1 percent of voters.

Several exit polls have been conducted asking voters their opinions about ranked choice voting. The most thorough of these, conducted by San Francisco State University, found that 87 percent of those polled said they understood ranked choice voting, while 61 percent preferred it over the old runoff system (only 13 percent preferred the December runoffs, while 27 percent said it made no difference).

Certainly there is room for improvement, but by any objective measurement San Francisco has taken to ranked-choice voting elections, and is leading the nation in this important reform.