Author’s note: On March 2, 2021, voters in Burlington overwhelmingly (64 percent to 34 percent) voted to return to ranked choice voting for city council elections beginning in march 2022. The measure must be approved by the state legislature and signed by the governor to be enacted.

If the 2016 US presidential primaries made you think “there must be a better way to vote” and you started poking around to learn more about democracy reforms, you probably came across Ranked Choice Voting (RCV) or Instant Runoff Voting (IRV) as a way to give voters more voice, reduce negative campaigns, and prevent unrepresentative winners. Once you start looking into it, it’s just a matter of time until you hear about Burlington, Vermont. Opponents point to Burlington as “proof” that IRV is doomed to failure and repeal.

In fact, Burlington’s experience proves no such thing, but it does offer three important lessons that reformers should heed. Here’s what happened.

What went down in Burlington, VT?

Prior to 2005, Burlington, Vermont used plurality voting with a 40 percent threshold, meaning that a candidate could become mayor even though 60 percent of voters preferred other candidates, and that in a crowded race where no candidate won at least 40 percent, taxpayers had to fund an additional runoff election. In 2005, voters in Burlington enthusiastically—by 64 percent—approved a ballot measure to switch to using Instant Runoff Voting, a form of Ranked Choice Voting, for mayoral elections. Instant Runoff Voting allows voters to express their preferences for more than one candidate and requires a candidate to win with a majority of active ballots in a single high-turnout election. The city of 39,000 became the seventh American city to use IRV (Australia, Ireland, London, and any organization using Robert’s Rules of Order had already used IRV, a.k.a preferential voting, for decades).

But then things went sideways: after the second mayoral election using IRV, Republicans became angry that the Progressive candidate beat the Republican. Then the Progressive mayor was embroiled in a corruption scandal, fueling a campaign to repeal IRV as a referendum on the mayor. In a low-turnout election in 2010, Burlington voters reversed and voted for a new ballot measure—by 52 percent—to repeal the system and return to plurality voting.

Background: How Instant Runoff Voting works

Instant Runoff Voting, a form of Ranked-Choice Voting for electing a single winner like a president or mayor, allows election administrators to simulate a runoff in a single election to find the first candidate to win support from a majority of active voters. Voters rank the candidates in order of preference. If no candidate wins a majority of first-choice votes, the least popular candidates are eliminated and their votes transferred to the next-ranked candidate until someone wins a majority of active ballots.

Every voter gets one vote per round (just like everyone gets one vote in the primary and one in the general or in the special runoff). If your favorite candidate makes it to the last round your vote counts for her in each round (just like you would vote for her in the primary and again in the general), but if your favorite is eliminated, in the next round your vote will count for your next-ranked candidate who is still in the running (if your primary favorite got eliminated, you would then vote for your preferred of the two candidates in the general). The ranked ballot in essence asks each voter “Of these candidates, which is your favorite? And if she was eliminated, which of the remaining candidates would you vote for?” and then eliminates and counts the rounds automatically, without forcing voters to come back to the ballot.

The Burlington elections

Progressive Bob Kiss won Burlington’s first ranked-choice race for Mayor in 2006. None of the five mayoral candidates won a majority of first choice votes, so candidates with fewer votes were eliminated and their votes automatically transferred until Kiss won with a majority—54 percent—of active ballots.

Kiss also won Burlington’s second IRV mayoral race in 2009. Again, there were five candidates in the race and none won a majority of first-choice votes: Republican Kurt Wright led with 33 percent, Progressive Rob Kiss followed with 29 percent, Democrat Montrell had 23 percent, and an Independent and a Green candidate had the fewest. The candidates with fewer votes were eliminated and their votes automatically transferred until Progressive Kiss beat Republican Wright with 52 percent of active voters. (Active voters were those that marked a rank for either Kiss or Wright. Five percent of Burlington voters ranked neither, effectively choosing not to vote in the automatic runoff—their ballots were “exhausted.”)

Original Sightline Institute graphic, available under our free use policy.

The repeal campaign

After the 2006 election, most Burlington voters were happy. In an exit poll, 63 percent said they like the ranked-choice ballot and 71 percent said it was a better way to express their preferences. There was one exception—Republican voters were not fans of ranked ballots. The 2009 election, in which Republican Wright had the most first-choice votes, further inflamed Republican hostility. The GOP Chairman claimed the IRV election was a “travesty” and the “will of the voters was circumvented.” Republicans started a campaign to repeal ranked voting; Wright was a vocal advocate for repeal.

Most voters in Burlington were not outraged that the Republican lost, so repealers might not have gotten far. But less than a year after the 2009 election, Mayor Bob Kiss was enveloped in a scandal involving the use of taxpayer money to prop up a telecom company. Critics of ranked-choice voting took advantage of the misconduct to turn their referendum on IRV into a referendum on Bob Kiss. Repeal slogans were: “If you don’t like Bob Kiss, vote against IRV” and, sarcastically, “Keep Counting ‘Til Bob win.” When the repeal campaign won, Republican Wright announced that the vote was “a clear rebuke to the Kiss administration.”

The repeal won with 52 percent in a low turnout election and was likely not a broad rejection of IRV, but rather an expression of intense dislike by a slice of the electorate. In fact, repeal might have won with votes only from voters who wished Wright was mayor instead of Kiss. The 2010 repeal won only in the two wards where Wright had won in 2009. Voters in three other wards voted nearly 2 to 1 to keep IRV. The repeal garnered 3,972 votes, compared to 4,059 voters who preferred Wright to Kiss.

Burlington elections returned to plurality voting with a 40 percent threshold. Meanwhile, a dozen or so other American cities and counties continue to use IRV with apparently few problems. So what can we learn from Burlington’s experience?

Lesson 1: Entrenched interests will fiercely fight for the status quo

Entrenched interests that benefit from the status quo will, naturally, obstruct reform.

It happened in Ann Arbor, Michigan, in the 1970s. The city’s left-leaning voters had a history of splitting votes between the Democrat Party and the (far left) Human Rights Party, allowing Republicans to win with mere minority support year after year. Voters adopted IRV and used it in the 1975 mayor’s race, allowing the Democrat, with votes from the eliminated Human Rights candidate, to win with a majority of active ballots. The Democrat was the first African-American mayor of Ann Arbor. Outraged Republicans successfully repealed IRV in a low-turnout election unusually dominated by Republican voters.

  • Our work is made possible by the generosity of people like you!

    Thanks to Marc Daudon & Maud Smith Daudon for supporting a sustainable Cascadia.

  • Democrats don’t like losing either. For example, in Oakland’s 2010 ranked-choice race for mayor, Jean Quan beat heavily-favored Democrat Don Perata to become the first Asian-American woman to be mayor of a major American city. The Perata campaign used the exact words that the Burlington GOP chairman used, calling the election a “travesty” and an “injustice” because Perata would have won in a “landslide”—if he didn’t need to win a majority! Perata’s spokesperson said Oakland would “pay the price.” Turns out, all Californians may have paid the price.

    Ann Arbor, Burlington, and Oakland suggest that IRV might help people of color and minor party candidates win, but when it does, entrenched interests will furiously respond by painting the wins as illegitimate and attempting to regain control.
    Tweet This

    In 2016, the California legislature passed a bill allowing all cities in California to use ranked-choice voting systems like IRV. To legislators’ bewilderment, Democratic Governor Jerry Brown vetoed the bill. Brown had supported fellow Democrat Perata in the Oakland mayor race.

    Ann Arbor, Burlington, and Oakland suggest that IRV might help people of color and minor party candidates win, but when it does, entrenched interests will furiously respond by painting the wins as illegitimate and attempting to regain control.

    Reformers face an uphill battle against inertia and powerful interests. Advocates should plan for education and outreach about the reasons for and benefits of electoral reform not only during the campaign to pass reform but for years after.

    Lesson 2: Advocates should be clear about the solutions reform is offering

    No electoral system can solve every political problem. And single-winner systems like IRV don’t offer all the benefits that proportional representation provides. But advocates should be clear about the problems a proposed reform can solve and nurture support for those solutions well after the campaign is over. Academics have studied IRV in the United States and internationally, so we can safely say it can do the following:

    It defeats “vote splitting”

    In plurality voting, the system used in most American and Canadian elections, two somewhat similar candidates can split the votes of a majority of voters, throwing the election to a different candidate with support from a mere minority. Most people think of the Bush-Gore-Nader problem, where the majority of voters split between the Democrat and the Green Party candidate, allowing the Republican to win with less than half the votes.

    Plurality voting discourages candidates in third or lower place from running and scares voters away from voting for them. IRV solves this problem. Voters can rank their favorite less-popular candidate first, and if she is eliminated their votes still go to their next-ranked candidate who is still in the running, and the winner will have support from a majority of active voters.

    It gives more voters a voice

    In the United States, elections are often decided in the primary. Even in competitive elections, primary voters have the power to narrow the options so that general election voters might not see anyone they like on the ballot. That’s a problem because primary voters are a small, unrepresentative group—older, whiter, and more partisan than general election voters. For example, just 21 percent of voting age Oregonians are 65 or older but 55 percent of primary voters are 65 or older.

    Original Sightline Institute graphic, available under our free use policy.

    It saves money

    If IRV eliminates a primary, that can save taxpayers’ money because cities and counties pay per item on the ballot. For example, every primary race Seattle could eliminate by letting voters express their preferences on a ranked ballot in the general election could save city taxpayers around $450,000, enough to nearly double the City Librarian’s Office funding.

    It tamps down negative campaigns

    Plurality voting fuels negative campaigns. When voters only have one vote and the candidate with the most vote wins even if he didn’t win a majority, making voters hate your opponent is just as good as making them like you.

    Instant Runoff Voting short-circuits this troubling pattern by rewarding candidates for reaching out and winning voters over, as Minneapolis Mayor Betsy Hodges explains. Research shows ranked-choice elections lead to more cooperation and civility among political candidates (who are hoping other candidates’ voters will still rank them second) and reduce negative campaigning.

    Lesson 3: A small group of opponents can repeal reform in a low-turnout election

    In both Burlington and Ann Arbor, the repeal campaign won in a low-turnout election. Even if most voters are happy with IRV (63 percent of Burlington voters liked the ranked ballots), if those voters don’t come out to defend it from repeal in an off year, it could still go away.

    Advocates should be aware they may need to keep supporters fired up and ready to defend electoral reforms from attacks for years. For example, almost 20 years after it passed, dissidents were still trying to repeal New Zealand’s proportional electoral method, but voters turned out to defend it.

    Final advice

    If you want a better way to elect the president or mayor, don’t let overblown warnings about Burlington and IRV scare you away from reform. Even critics of IRV agree that plurality voting (the status quo) is the absolute worst method, but the status quo seems normal and safe whereas any departure from the status quo seems risky. When advocates for other reforms join the entrenched interests to attack IRV, the result, as in Burlington, will be a triumph for plurality voting. Those who want a better way to elect a president or mayor must join together to defeat  the status quo.