Ranked Choice Voting in the Bay Area

by Grace Ramsey, Rob Richie // Published October 16, 2014
oakland skyline bw by geolio d1ubw8s

Interested in this topic? Sign up to receive our newsletter and other updates on elections and electoral reform.

 Ranked choice voting (RCV) ballots are already being cast by some early and absentee voters in the four Bay Area cities holding RCV elections this November: Berkeley, Oakland, San Francisco, and San Leandro. Unlike Bay Area cities and counties that limited the November ballot to just two candidates in the historically low turnout primary last June (with some cities picking their winners then, even when the winner finished with far less than 50%, as in Hayward), RCV cities are presenting voters with a full array of candidates in a number of key races, most notably mayoral elections in Oakland and San Leandro. 

In light of the upcoming elections, we wanted to share a few facts about ranked choice voting:

Since adoption of ranked choice voting, Oakland winners earn more votes: Of the 18 Oakland offices elected by RCV in 2010 and 2012, the winner had more votes than the winner of the previous non-RCV election in 16 of them. This is a common pattern in RCV elections that avoid electing winners in low-turnout June primaries or December runoffs.

Fewer voters now skip city elections: Voters are more engaged in city elections with RCV. The four Bay Area cities with RCV all hold key elections when the top of the ballot is a high-profile race like president or governor. Traditionally, a good number of voters skip their city elections, but that number has declined sharply since adoption of RCV. San Francisco provides a good example. In the 16 RCV elections for Board of Supervisors that required multiple rounds to determine a winner in 2004-2013, the median proportion of voters that skipped this race on their ballots was 9.7%. In the 11 elections for Board of Supervisors that went to runoffs between 2000 and 2003, before the adoption of RCV, the median was 14.4%, nearly 50% higher.

Candidates must treat every voter as a swing voter, and make an affirmative case: With RCV, the strongest candidates will need to treat every voter as a potential swing voter. In a vote-for-one system, candidates write off anyone who is settled on another choice. But with RCV, there is a direct incentive to find common ground with such voters to earn a second or third choice. This argument for RCV has been made eloquently by the winners of the most recent RCV elections for mayor of Minneapolis and Portland, Maine.

Scholars find RCV has a clear impact on civility and substance: FairVote is part of major scholarly study on the effect of ranked choice voting on the civility and substance of campaigns. Initial findings, summarized here, suggest that RCV is having a big impact. Voters from RCV cities were significantly less likely to report that candidates criticized one another “a great deal” than were voters from non-RCV cities (5% to 25%) and were nearly three times as likely to say that candidates had not criticized one another at all (36% to 12%). RCV city voters were also significantly more likely to indicate that they were "very satisfied" with campaigns, and over 60% of respondents in RCV cities supported the system.

Bay Area voters had more trouble with the first Top Two primary ballots than RCV: The huge majority of voters in Bay Area cities handle RCV ballots well. As one measure, less than 0.4% of voters invalidated their ballots in their first contested mayoral elections with RCV in these cities. But it was another story with the first Top Two primary ballot in June 2012. In that election, more than 1.5% of voters cast an invalid ballot in the U.S. Senate primary in Berkeley and San Francisco, and more RCV Famous People Ballot Image
than 4.5% of voters did so in the Senate primary in Oakland and San Leandro. That means that more than 15 times as many Oakland and San Leandro voters invalidated their ballot in their first-ever Top Two primary election for Senate compared to their first-ever mayoral election with RCV.

RCV winners defeat their strongest opponent 1-on-1, as pollsters are now showing: With RCV, no voter ever has their ballot count for more than one candidate at a time. Voters rank the candidates, and all first choices are tallied. Last-place candidates are eliminated, and their ballots are added to the totals of their voters’ next choices until one candidate wins with more than half the votes. Whenever the election comes down to two candidates in the final round, as when Jean Quan defeated Don Perata in the 2010 mayoral race in Oakland, the winner is the candidate who wins 1-on-1. A poll in early September 2010 showed that Quan was already neck-and-neck with Perata in a one-on-one comparison, but the pollster instead focused on less useful aggregate totals of first, second, and third choices. We are encouraged to see that pollsters this year are providing those 1-on-1 comparisons to show just where the state of the race is.

The limit of 3 rankings is due to voting equipment, not the law: Once Alameda County and San Francisco secure voting equipment that allows for more rankings, as is already being done elsewhere, Bay Area cities will allow voters to rank more candidates. This will be helpful in elections with large fields, but note that having three rankings already provides voters with a three-times greater chance to have their vote count than Top Two election systems with large fields.

Voting smart is a part of all Bay Area elections: Given the current limitation of three rankings, voters who want to have their vote count in the final round would be wise to vote for their true favorite with their first choice, and then use their second and third choice for acceptable candidates who seem likely to get to the final round. That is going to happen with the great majority of voters regardless, but RCV gives voters who support lesser known candidates the freedom to vote their conscience without wasting their ballot.

If Oakland still had its old runoff system, it would have eliminated all but two candidates in a low-turnout, unrepresentative June primary. In such a contest, you would have needed to be tactical with your one and only vote and ignored any candidate who didn't seem to have a chance to win. It's quite possible that an absolute majority of voters would have voted for losing candidates, as was the case in last year’s Boston mayoral election, in which the primary "winners" earned just 18% and 16% of the vote (and were two white men, finishing just ahead of several candidates of color).

Electing the most representative candidate: There have been dozens of hotly contested RCV races in the Bay Area, including 31 in which there was an "instant runoff" to determine the winner, and five in which the ultimate winner trailed in first choices. In every single election, the winner has been the candidate who defeats all other candidates when matched against them 1-on-1.

We look forward to this year's elections with anticipation. Most candidates will lose, but there is a winner: voters who have the chance to engage with more campaigns, vote for their true favorite and elect a candidate who has proven their ability to represent voters in city government.