clock menu more-arrow no yes

The hidden politics of New York City’s new ranked-choice voting system

Reformers hope it can transform campaigns. Are they right?

New York City mayoral candidates (left to right) Eric Adams, Scott Stringer, Maya Wiley, Kathryn Garcia, and Andrew Yang.
Getty Images

What if I told you that instead of voting for just one candidate in the next election, you could vote for several — ranking them in order of your preference?

Registered Democratic voters in New York City are getting that opportunity. Their mayoral primary on June 22 will be the city’s first to use ranked-choice voting — and that race will be the biggest spotlight yet for this system in the United States.

Ranked choice is the reform of wonky activists’ dreams. They believe that more traditional elections, where whoever gets the most votes is simply the winner, can go wrong when there are multiple candidates, as someone most voters oppose can win due to the split of the vote. One famous example is Ralph Nader being the third-party “spoiler,” apparently drawing votes away from Al Gore and tipping key states to George W. Bush in 2000.

Ranked choice can, in theory, avert this outcome, because it asks voters to rank candidates in order of their preference. As votes are counted, the lower-performing candidates are gradually eliminated, and votes for them are redirected to those voters’ backup choices. (In the 2000 example, when Nader gets eliminated, ballots that ranked Nader as first choice and Gore as second turn into votes for Gore, increasing his total.)

Ranked-choice voting is still pretty rare in the US, but it’s getting less so. Several cities (most notably San Francisco, Oakland, Minneapolis, and St. Paul) have used it for over a decade. Maine adopted it for federal races in 2018, and Alaska will try its own version of the system next year. And with New York, it’s arrived at the country’s most populous city.

Elections inspector Brian Linden (left) explains ranked-choice voting to Elena Batyuk before she casts her ballot in the New York City mayoral primary election on June 14.
Mary Altaffer/AP

Supporters of the new system hope it will have even more dramatic benefits — indeed, that it will help cure much of what they think ails American politics. They argue that ranked choice’s incentives could lead to less negative campaigning and polarization. And many of them hope it will help new ideas and types of candidates, which had been previously blocked by the two-party system and party establishments, to thrive.

“Fundamentally, what a ranked-choice ballot as we propose it does is give a voter a backup to their first choice,” says Rob Richie, president and CEO of FairVote, the nonprofit group that has done more than any other to evangelize for ranked choice across the country. “From that, it creates really positive incentives for how candidates act and how voters act.”

But not everyone has been thrilled about New York’s new system. Several Black members of New York’s City Council unsuccessfully sued to try and block it last year. “Some progressive white folks got together in a room and thought this would be good, but it’s not good for our community,” Hazel Dukes, the president of the NAACP’s New York state chapter, said at the time.

As ranked choice has slowly caught on in more places across the country, questions about complexity and equity have dogged its proponents. Are its advantages worth the drawbacks? And how does it actually play out with actual voters in the real world?

What is ranked-choice voting, and how does it work?

The “plurality winner” system used in most US elections — where the candidate with the highest number of votes wins, even if they’re short of a majority — can get weird in contests with multiple candidates.

The very conservative Republican Paul LePage won Maine’s governorship in 2010 with 37.6 percent of the vote, while a moderate independent got 35.9 percent and the Democratic nominee got 18.8 percent. More Democrats surely would have preferred the independent to LePage, but they couldn’t coordinate behind him. Another version of this phenomenon is the “spoiler,” where a third-party candidate apparently draws votes mainly from one major contender and tips the outcome to the other, as arguably seen with Nader and Gore in 2000.

These are examples where Democrats felt robbed, but it can happen to any party or faction. Republicans blamed Ross Perot’s candidacy for electing Bill Clinton in 1992. In party primaries, sometimes multiple establishment candidates split the vote and let a reformist or ideologue win, and sometimes it’s the other way around.

The outcome can seem kind of arbitrary: It hinges on just how the vote happens to split, and who happens to end up with the biggest slice of the pie, rather than what a majority of voters want.

Enter ranked-choice voting, which is supposed to let voters more effectively express their “backup” preferences. In theory, this produces something closer to a consensus choice. It differs in key ways from the elections Americans are most familiar with.

The voting: In typical US elections, for each office on the ballot, a voter just picks one candidate. But in a ranked-choice election, they get to pick several, marking one candidate as their first choice, another as their second choice, and so on. The number of ranking slots for each race varies by jurisdiction, but in New York City there are five.

An example graphic from the New York City government showing a ranked-choice ballot.
New York City

The counting: The real action happens when the votes are counted. If no candidate wins a majority in the initial tally, the reallocation rounds begin, and lower-performing candidates are eliminated one by one, reality show style:

  • If there are 13 candidates, then after all the first-choice votes are tallied, the candidate in 13th place is out.
  • People who voted for the eliminated candidate will have their ballots reallocated to their second-choice pick.
  • Then the 12th-place candidate will be eliminated, and those ballots get reallocated.
  • And so on and so on, until eventually, one candidate ends up with a majority of the remaining votes in the winnowed field.

The strategy: One key to understanding this system is that you cannot hurt your first-choice candidate by ranking others below them. Your backup rankings will only matter if your first-choice candidate is eliminated from contention (due to getting low support from others). One argument is that you should figure out your true first choice — the person you genuinely want to win, strategic considerations aside — and rank them first. (Though for an example of how even that could go awry, check out this New York Times article on the “Alaska dilemma.”)

From there, it gets even trickier. In New York City, there are 13 candidates on the ballot. Eight of them are generally viewed as credible — Eric Adams, Kathryn Garcia, Andrew Yang, Maya Wiley, Scott Stringer, Dianne Morales, Shaun Donovan, and Ray McGuire. But voters are only permitted to rank five.

This means voters could well use all five of their rankings on candidates who get eliminated. For example, it’s plausible that, after reallocation, the race will come down to Adams versus Yang. If your ranked picks were Garcia, Wiley, Stringer, Morales, and Donovan, your ballot will become “exhausted” and play no role in the final round.

So if you do want to affect the final outcome, you should try to think about who the race is likely to come down to. Perhaps you dislike both Adams and Yang, but if you would prefer one over the other, it might be smart to rank your preference fifth.

Another complication, though, is that the race may not come down to Adams versus Yang — other polls have shown Adams versus Garcia and Adams versus Wiley as the final matchup, and it’s also possible polls will be off and we’ll get a surprise. If you want to maximize your chances of affecting the outcome, you should probably include at least three of the Adams-Wiley-Garcia-Yang quartet in your rankings. But that might come at the cost of leaving off some lower-profile candidates you genuinely like. It’s a trade-off.

New York City mayoral candidate Eric Adams campaigns in Brooklyn on June 15.
Michael M. Santiago/Getty Images

How it really plays out: In recent decades, there have been 236 ranked-choice elections in the US with single winners and at least three candidates running. FairVote has assembled data on how these races turn out. In about 94 percent of these elections, the candidate leading in first-choice preferences ended up winning.

So despite all the drama about backup choices and reallocation, the outcome is usually the same as in a plurality winner system.

Ranked-choice fans, though, would argue the exceptions are the point. Perhaps most of the time, the plurality leader really is acceptable to the majority. But when they’re not, ranked choice lets the plurality leader be unseated — usually by whoever’s in second place during round one.

Jared Golden (D) defeated a Republican incumbent US House member in Maine in 2018 in a ranked-choice election only after voters for independent candidates were reallocated. Jean Quan (D) won Oakland’s mayoral race in 2010 by defeating a controversial incumbent who led in the first round, because the anti-incumbent voters consolidated around her during reallocation.

Is there another agenda behind ranked-choice voting?

All kinds of political players can back ranked choice if they view it as in their interests (like mainstream Maine Democrats, who blamed the plurality system for LePage’s governor win). But it’s an odd combination of moderate centrists and establishment-skeptical progressives who tend to be true believers in the system.

These groups, despite ideological differences, have some things in common. Both share a deep-seated belief that the public is really on their side and would vote accordingly were it not for distorting influences. Both believe that, if people were freed up to vote for their “true” first choice (without worrying about who could win or “wasting” a vote), candidates with their views would thrive.

FairVote has built deep ties with progressive reformers, but for over a decade, its board chair was John Anderson, a former Republican member of Congress who ran for president in 1980 as an independent candidate challenging Jimmy Carter and Ronald Reagan; a more recent leading donor to FairVote and to ranked-choice campaigns across the country has been John Arnold, an idiosyncratic Texas billionaire who has voiced frustration with the two parties and said both of them should moderate. FairVote and Arnold backed a particular ranked-choice ballot initiative in Alaska that passed and could help Sen. Lisa Murkowski (R) defeat a conservative challenger next year (though there are questions about whether it would, in practice).

But it’s not just moderates. Green Party leaders Howie Hawkins and Jill Stein are also fans, as are “Squad” members Reps. Alexandria Ocasio-Cortez (D-NY) and Ayanna Pressley (D-MA). Progressives with an interest in government reform also like it — Rep. Jamie Raskin (D-MD) is a former board member of FairVote, and the liberal nonprofit stalwart Common Cause has been a key supporter.

Revealingly, in cities where ranked choice has been implemented, the most enthusiastic backers have often been progressives who have tense relations with the Democratic establishment or “machine.” Many of these cities already had a way of dealing with the plurality winner issue for a crowded contest: a runoff, where the top two candidates from the initial round face off at a later date. The winner of a two-candidate race by definition has a majority of the vote.

Yet ranked choice’s most enthusiastic backers in cities hated those runoffs. They pointed out that turnout usually dropped between the initial election and the runoff, claiming that made the winner less representative. They also complained holding another election was expensive and inconvenient for voters who had to turn out at the polls again.

For many critics, though, the real issue may have been that they didn’t like the runoffs’ typical outcomes: Establishment candidates tended to beat reformers or outsiders. “Big money would really spend bigger in runoffs. It’s easier to negatively blast one of the two candidates in runoffs,” Richie said. “It’s easier to control.” Supporters hope ranked choice will be tougher for these interests to game. (The counterargument is that these establishment forces are only getting their way by, well, persuading more people to vote for them.)

Reformers also hope ranked choice will make campaigns more pleasant overall. The idea here is that every candidate is going to want to be listed as the second or third choice from voters who don’t already support them, so they’ll be less likely to alienate those voters. This, reformers hope, will disincentivize negative campaigning (though it won’t eliminate it entirely).

Who are the critics, and what are the criticisms, of ranked-choice voting?

Many objections to ranked-choice voting are situational. Republicans generally dislike the system, due to either instinctive conservatism or a belief that general election vote splits generally tended to help them. (Still, the Virginia GOP implemented the system at its convention this year in an effort to prevent an extreme Trump superfan from winning the governor’s nomination.) Democratic establishment forces in cities feared ranked choice would disadvantage them compared to the runoff system: Gavin Newsom voted against ranked-choice voting while he was on the San Francisco Board of Supervisors, and as governor he vetoed a bill that would have allowed it in more California cities.

But there are more principled objections as well. I’m a weirdo who loves complexity and making ordered lists of things. Not everyone is like me, though, and the task of ranking multiple candidates in a crowded field may make voting seem more daunting. The added element of gaming out who the final two candidates are likely to be, to ensure you use one of your limited ranking slots on at least one of them, can also be challenging in a complex race.

Relatedly, many fear that less privileged voters — voters who don’t speak English, who are lower-income, or who are less educated — will have more difficulty with the new system, if they haven’t been sufficiently informed about how to use it. Perhaps they may be more likely to have their ballots thrown out due to improper rankings. Or perhaps they may be less likely to use all their ranking slots, making their ballots disproportionately likely to be discarded in a later round.

“We’re asking people to fundamentally shift and change the way they’ve voted their entire lives, in a very short period of time, with very little education, in the wake of a global pandemic,” says Christina Greer, a political science professor at Fordham University.

Mayoral candidate Eric Adams had his own self-interested reasons to oppose ranked choice — he feared it would hurt him — but he made this argument last year, saying, ”Voting can’t be for the astute, the technically astute people. It must be for everyone, and we’re not ready right now.”

Before the second televised debate for the New York City mayoral race, supporters gather outside the CBS studios to cheer for their choice on June 10.
Andrew Lichtenstein/Corbis/Getty Images

Another critique holds that ranked choice can fall short of its promise of producing a winner with majority support among voters, because of “ballot exhaustion.” During the various reallocations and rounds of a ranked-choice count, there’s a drop-off in the overall number of ballots still being counted. Partly this is because voters have limited ranking slots, and all their ranked picks could get eliminated. Other voters don’t even use all their slots in the first place, listing no backup candidates, so their ballot is exhausted when their first choice is out.

This happens frequently. Of the 128 US ranked-choice elections that have proceeded to at least a second round of counting, only 64 ended up producing a winner with a majority of the voters who turned out, per FairVote.

In the San Francisco mayoral election of 2018, for instance, there were 251,032 votes counted in the first round; in the final and decisive ninth round of counting, only 229,408 votes remained. This was a drop-off of 21,624 votes, or 8.6 percent of the total, and because of it, the narrow winner, London Breed, did not end up getting a majority of votes cast.

Ranked choice is more successful at producing majority winners than the simple “plurality winner” system. But ballot exhaustion rates raise questions about whether it’s really better at it than a runoff. Supporters argue that the problem was that, in many of these elections, voters only had three ranking slots, so ballots were exhausted more easily as all three were eliminated. But adding more ranking slots means foisting more complexity on voters.

Some real voting systems nerds have a different critique. They say that, compared to alternatives such as “Condorcet voting,” ranked choice doesn’t do enough to help the consensus “backup” choice. The complaint here is basically that the one-by-one candidate elimination in the count can end up eliminating most voters’ second choice, if that candidate doesn’t have enough first-choice votes. (This doesn’t seem to happen very often, but it did happen in Burlington Vermont’s 2009 mayoral race, and it could happen in Alaska’s Senate race next year.)

Finally, some simply don’t agree with the reformers’ view that “party establishment” and “negative campaigning” should be dirty words. Reformers “who believe politics is the problem often don’t end up thinking about how parties help voters,” says Jason McDaniel, a political science professor at San Francisco State University.

There’s a school of thought among political scientists that negative campaigning can actually be useful and helpful to voters, because it at least can be more substantive than hazy, image-burnishing positive ads. Ranked choice, in contrast, might incentivize a sort of inoffensive lowest common denominator campaign, rather than a bold one. (Though you do have to stand out to some extent to avoid getting eliminated.)

Party endorsements and campaigns can also help simplify choices for voters who trust the party. And runoffs have their virtues — they concentrate the mind, bringing clarity to a crowded field, by offering a binary choice, rather than tasking voters with researching and forming opinions on multiple little-known candidates, as they will in a ranked-choice election.

When the rules change, people adapt

With a decade or more of experience with ranked choice in some jurisdictions, the sky generally has not fallen, but politics hasn’t been fundamentally transformed either. Voters adapted — and so did the parties.

The theory that ranked choice gives progressives a marginal boost in primaries in liberal cities seems plausible, but the system certainly hasn’t shattered establishment power. The task of negative campaigning may have been effectively outsourced to independent groups rather than the candidates themselves.

Some studies say ranked choice may have resulted in somewhat more minorities and women winning, versus comparable cities that didn’t use ranked choice. But it’s too early to say for sure: Ultimately, we’re mainly talking about a handful of high-profile mayoral races and dozens of lower-profile city races so far. Evidence on turnout is mixed, but it doesn’t seem that ranked choice makes turnout fall off a cliff.

Overall, despite ranked choice’s impressive spread in recent years, Richie, FairVote’s president, is still looking down the road. The group was initially called Citizens for Proportional Representation (CPR, for “resuscitating democracy”), in the hopes that proportional election results would lead to a multi-party system. But Richie told me that moving to a proportional system was too heavy a lift for the time being.

“I and the people that got the organization going were visionary but pragmatic at the same time,” Richie said. “The country’s ready for ranked-choice voting. It is, we believe, an important step toward a conversation about proportional voting.”

The real reform he’s long thought the US needs, though, is multi-member districts. This would allow for proportional ranked-choice voting — where seats are split up according to how the vote breaks down.

But that’s a reform for another day.

Recode

No, the supply chain mess is not a war on Christmas

Explainers

The $5 trillion insurance industry faces a reckoning. Blame climate change.

Politics & Policy

Can the US cut drug prices without sacrificing new cures?

View all stories in Explainers

Sign up for the newsletter The Weeds

Understand how policy impacts people. Delivered Fridays.