Looking at the nation’s most intriguing experiments in local policy.
The policy: Ranked-choice voting
Where: San Francisco
In place since: 2002
In March 2002, San Franciscans were ready for some new voting rules. The city had long used a two-round runoff system for elections, which usually meant a second round in December to get a majority winner. Voters were tired by then. Taxpayers complained of the cost.
San Franciscans had experienced runoffs in 2000 and 2001, with turnout declines of 51 percent and 44 percent, respectively. Six of the past eight city elections had asked voters to come back to the polls a month later to ensure all city office holders won a majority of votes.
So voters approved Proposition A and became the first US city to adopt instant-runoff voting in the modern era. A coalition of good-government reformers and progressive politicians got behind the voting system, a variation on a system that 24 US cities (including New York City and nearby Sacramento) had adopted between 1915 and 1948 in an earlier era of municipal reform.
These days, IRV goes by ranked-choice voting. Same thing, different feature highlighted.
As the name suggests, ranked-choice voting lets voters mark their first-choice candidate first, their second-choice candidate second, their third-choice candidate third, and so on. Easy as 1-2-3. Each voter has only one vote but can indicate their backup choices: If one candidate has an outright majority of first-place rankings, that candidate wins, just like a traditional election. But if no candidate has a majority in the first round, the candidate in last place is eliminated. Voters who had ranked that candidate first have their votes transferred to their backup — that is, the candidate they ranked second. This process continues until a single candidate gathers a majority.
In San Francisco, there was another selling point besides eliminating costly low-turnout runoffs: Voters could vote with their heart first and their head second, picking their favorite candidate first without feeling they had “wasted” their vote, then picking the lesser of two evils second or third, knowing their vote would still count.
But more than anything, the hope was that ranked-choice voting would encourage more civil campaigning, more engagement with voters, and better coalition-building. Candidates would now be angling for second- and third-choice preferences. They’d be nicer to each other as a result. They might even campaign together.
How it worked:
As San Francisco goes, so goes everyone else. In 2004, Berkeley followed suit; Oakland adopted the practice in 2006. So did Minnesota cities Minneapolis in 2006 and St. Paul in 2009. Unlike San Francisco, Minneapolis and St. Paul didn’t have two-round elections that spent taxpayer dollars on low-turnout December runoffs. They spent it on low-turnout primaries where fewer than one in 10 residents turned out. So why not just have one election, primary and general, rolled into one, where a ranked ballot did the whittling? Most there thought it was a good idea.
Minnesotans were also familiar with the shortcomings of simple plurality voting. In 1998, Jesse Ventura became the state’s governor with just 37 percent of the vote in a three-way race. In 2002, Tim Pawlenty won with just 44.4 percent, also in a three-way race, with an independent getting 16.2 percent of the vote. (Pawlenty won again with a mere plurality in 2006 — 46 percent, with an independent getting 6.4 percent.)
Against the backdrop of plurality elections producing mere plurality winners, the case for ranked-choice voting got stronger: Shouldn’t the winning candidate have to earn a true majority? Ranked-choice voting meant that once second- and third-choice preferences were counted, the winning candidate was indeed the most broadly preferred — a strong selling point.
Twenty cities in the United States have now adopted ranked-choice voting. But the big news was that in 2018, Maine became the first state to adopt it for federal elections. This has catapulted the reform to a national spotlight, with presidential candidates like Sens. Elizabeth Warren and Michael Bennet now championing it.
As promised, it has made politics a little less nasty. In cities with ranked-choice voting, candidates spent less time attacking each other, as compared to similar cities that didn’t adopt ranked-choice voting. Voters in these cities reported being more satisfied with local campaigns as a result (again, as compared to similar cities).
Ranked-choice voting has also increased the share of racial minority candidates, female candidates, and female minority candidates running compared to similar cities. The scholars who have studied this most closely believe more minority candidates ran because under ranked choice, such candidates could reach out to other communities where they might not be the natural first choice and ask for second-choice votes. They believe women were more likely to run because under traditional winner-take-all elections, “women were deterred from running for office by … negative campaigning.” But with less negative campaigning and more cooperative campaigning, women are more likely to run. They’re also more likely to win, scholar Sarah John and her colleagues concluded.
No voting system is perfect. Critics of ranked-choice voting say the task of having to rank multiple candidates unfairly overwhelms low-information voters, and that the added complexity hits poor minorities hardest. The editorial board of the San Francisco Chronicle last year called the process — in place for more than 15 years — “a baffling experience for many voters,” and decried a plan by two San Francisco mayoral candidates to defeat another, London Breed, by “gaming” the ranked-choice system.
However, surveys have shown no difference in understanding of ranked-choice voting between white and nonwhite voters, and more than 90 percent of voters who have used the system describe it as simple. Some studies show declines among minority voters. But most likely, when properly accounting for other factors, turnout has probably remained stable, with no declines in poorer precincts.
But cities are one thing. The RCV cities are solidly on the political left, and some are technically nonpartisan. How it works at the state level, or even the national level (as many, including me, have proposed) with contested partisan elections, is another.
But the national stage is where RCV has its greatest promise. The zero-sum, binary nature of our two-party elections rewards negative campaigning, where winning comes from disqualifying the opponent. It pushes politicians into us-against-them rhetoric so many now decry. Ranked-choice voting would make space for political alternatives to emerge without being spoilers, potentially reorienting our stuck partisan division.
Here, we might look not to the handful of reform-oriented cities but to Australia, which has used ranked-choice voting nationally for 101 years, and with a strong track record of moderate and stable politics. Most likely, the next step for ranked-choice voting is for more states to follow Maine’s lead, adopting it statewide. Maine’s first ranked-choice voting run in November went smoothly, though we’re most likely to see effects kick in with the next election, as candidates adjust to the new system.
But Maine’s adoption did provoke some backlash from Republicans, including Gov. Paul LePage, who was twice elected with a mere plurality. LePage called it “the most horrific thing in the world.”
As for San Francisco, it had a somewhat wild eight-round ranked-choice voting special mayoral election in June 2018. In the end, Breed became the city’s first female African American mayor, with 50.6 percent of the vote after all transfers, even though her two leading opponents joined forces. With 12 years of experience, San Franciscans have learned to use the system. They will use it again this November, for the city’s normally scheduled mayoral election.
Lee Drutman is a senior fellow in the political reform program at New America.