Voices & Choices

Ranked choice voting’s midterm report

Ranked choice voting’s midterm report

The following blog is based on a new FairVote report on ranked choice voting elections in 2018

In the first half of 2018, nearly half a million voters ranked their choices in elections for the most important offices in their communities. Voters in Santa Fe, New Mexico elected their first full-time mayor, voters in San Francisco elected their mayor in a hotly-contested special, and voters in Maine ranked their choices in state and congressional primary elections – and then convincingly upheld ranked choice voting (RCV) in a statewide referendum.

We’ve now seen RCV in action enough to know it works. Looking at the 2018 results shows:

  • Voter turnout surpassed expectations.
  • Implementation of RCV was smooth and inexpensive.
  • Voters used the ballot well, ranking their choices and making few errors.
  • Outcomes were fair, with winners earn both core and broad support.

These benefits come from an easy change: allowing voters in rank their preferences among the candidates: first, second and so on. The tally then simulates a series of runoff in which the last-place candidate is defeated, and ballots for that candidates go to their next choice until someone wins with a majority of the vote.

Santa Fe, New Mexico: Santa Fe first adopted a charter amendment for RCV in 2008. Because Santa Fe historically relies on the state for voter equipment, implementation was delayed. However, after New Mexico added new software to its machines, RCV could finally be implemented for its elections in March, although it took great community organizing and a lawsuit to ensure it happened. Five strong mayoral candidates ran. Alan Webber won with 66 percent of the vote after leading in the first round with 39 percent. One city council race was also decided decisively in an instant runoff.

CA_results.pngSan Francisco, California: San Francisco has used RCV since 2004, after a 2002 ballot measure win. Following the death of Mayor Ed Lee, the city called a special election that would take place during the state primary elections on June 5. Eight candidates ran, with three frontrunners: Board of Supervisors President London Breed, former state Senator Mark Leno, and Supervisor Jane Kim. Both Jun and Breed had campaigned well under RCV in upset wins for the Board of Supervisors, and Leno had helped the city adopt RCV in 2002. Sharing many policy positions, Leno and Kim each asked their supporters to rank themselves first and the other second. Breed led in first choices by 12% and won the final instant runoff by 1 percent.

Maine congressional and state primaries: In 2016, Mainers voted to become the first state to adopt RCV for all of their state and congressional elections in a great citizen-led campaign. The new law then faced a gauntlet of legislative and legal challenges from incumbents, but ultimately the will of the people prevailed. RCV was used for the first time in Maine’s partisan primary elections on June 12, with seven Democrats and four Republicans running for governor. Four Democrats were also on the ballot for the 2nd Congressional District primary. Janet Mills won the Democratic nomination for governor and Jared Golden won the Democratic nomination for the CD-2, with both candidates securing decisive majorities after an instant runoff. Shawn Moody on the Republican nomination for governor on the first round, and the one state legislative primary with more than two candidates was won on the first count as well.

In the same election, Maine voters voted to keep using RCV in its November general elections for congressional offices (including this November, when there will be at least three candidates in elections in Maine for U.S. senate and its two congressional districts) and in future congressional and state primary elections. RCV doubled its 2016 victory margin. Because of an adverse ruling by Maine’s highest court, RCV will not be used for general elections for governor and state legislature without a change to the state constitution, but Mainers seem determined win RCV for all their elections.

Voter Turnout In San Francisco, turnout was nearly 53 percent, far higher than the 29.7 percent percent turnout in the June 2014 primary without an RCV races and the state primary average this year of 38 percent. A total of 250,868 voters cast a vote, which was far higher than votes for governor and U.S. Senate. The Maine primary drew the most Democrats in Maine’s primary history. Republican turnout was higher than in all but one of their gubernatorial primaries since 1998 In Santa Fe, 20,604 voters cast a valid vote for mayor, topping the highest turnout in any recent mayoral election, including the 17,022 votes cast in a comparably contested mayoral race without RCV in 2014. Local news emphasized the high number of candidate debates and overflowing attendance -- with many apparently desiring to learn more about the full field.

We’re seeing more candidates running with a good understanding of how RCV is grounded in reaching out directly to as many voters as possible. High voter turnout continued the 2017 trend, where all four cities with RCV contests – Minneapolis, St. Paul, Cambridge, Mass. and Takoma Park, Md. – had surges in turnout.

Voter Experience and Use of RCV Ballots: The evidence from RCV in practice plainly shows that voters are comfortable ranking their choices This year had extremely low invalidating overvote rates across RCV elections with different ballot designs and systems, including only 0.13 percent in Santa Fe’s five-candidate mayoral election, 0.25 percent in San Francisco's eight-candidate mayoral race, 0.24 percent in Maine Democrats’ second congressional district primary with four candidates, and 0.34 percent in Maine Democrats’ seven-candidate gubernatorial primary. Voter error that invalidated ballots was far higher in the non-RCV race for governor in San Francisco.

In Santa Fe, a large and representative exit poll found that 94 percent reported being satisfied with their voting experience, and that their level of confidence in the process was higher than that of New Mexico voters statewide in the 2016 presidential election. Of those who voted in the mayoral contest, 65 percent ranked all five candidates, and 88 percent ranked at least two. In San Francisco, 85 percent of voters ranked at least two. In Maine Democratic primaries for Congress and Governor, nearly nine-in-ten voters ranked two. All of Maine’s RCV winners earned absolute majorities of the first round vote. In contrast, runoff winners in more than half of the 23 congressional primary winners this year actually earned fewer runoff votes than in the first round.

Implementation details varied significantly across the three jurisdictions. Both Santa Fe and San Francisco used equipment that allowed the round-by-round tallies to be released immediately once ballots were processed, meaning that they ran RCV tallies on election night and RCV results were final as soon as ballots were processed. Maine’s election took longer to report final RCV results, because it took longer for the ballot data to arrive in the state capital.

Fairness of Outcomes: How RCV Candidates Seek Broader Support: RCV has many benefits, but probably the most intuitive is ensuring that winners earn more support. RCV routinely outperforms either single-choice plurality or two-round runoff election systems by this measure, serving to avoid situations where the winner earned only low plurality or were elected in a low-turnout runoff election. An additional way of measuring success at voter engagement is to consider what proportion of the voters highly ranked the winner. This measure does not affect the outcome, of course, but expands the result beyond the relative support of the winner compared to their rivals.

Every RCV race in 2018 has been won by a candidate who was ranked in their top three by at least 60 percent of voters. Such levels of support demonstrate that even many of those backing the strongest challenger are comfortable with the winner.  Despite a hard fought and extremely close mayoral election in San Francisco, for example, 47 percent of 2nd place finisher Mark Leno’s voters and 37 percent of third-place finisher Jane Kim’s voters ranked the winner London Breed among their top three candidates. Winners in these races are earning a more convincing mandate and have compelling reasons to govern in ways that satisfy more voters.

Cities and states with RCV are experiencing healthy, positive campaigns that are drawing relatively high turnout from voters. Voters seem to appreciate the opportunity to rank their choices, and they do so without making serious errors when compared to non-ranked contests. Winners emerge with greater consensus support in their communities and real mandates. It’s time to make this a national norm.

Rob Richie is president and CEO of FairVote.

Join Us Today to Help Create a More Perfect Union