The New York City mayor’s race briefly descended into chaos this week when an employee at the city’s Board of Elections mistakenly tabulated 135,000 test ballots — adding 135,000 phantom votes to the total. It was a human screwup, at an institution with a long history of incompetence. If anything, the lesson is that electoral administration is important infrastructure in need of investment. It would be a mistake to attribute any of this confusion to the introduction of ranked-choice voting to New York City.
By any measure, the ranked-choice voting part of the election was a resounding success. Despite fears it would be too complicated, New Yorkers overwhelmingly found it easy and straightforward to rank. Primary turnout hit its highest levels in more than 20 years. And as the counting continues, New Yorkers find themselves watching an unexpectedly tight race unfold between Eric Adams and Kathryn Garcia — who only a week ago seemed out of the running.
How much of this can we attribute directly to ranked-choice voting? Let’s look at what happened. The mayoral race attracted a wide range of high-quality candidates, and quite a lot of media attention. The race was interesting because it was uncertain and competitive — and it was uncertain and competitive in large part because of ranked-choice voting. Meanwhile, Garcia’s late surge appears to be driven by the fact that she was the second choice of many New Yorkers, the exact sort of result ranked-choice voting is supposed to facilitate.
For those who still need a refresher on how ranked-choice voting works in a single-winner election: Voters rank candidates in order of preference. If one candidate has a majority of first-choice rankings, that candidate is declared the winner. If there’s no majority winner, the candidate with fewest votes gets eliminated. Voters who selected that candidate as their first choice then have their votes transferred to their second choice. Votes get re-tabulated. If there’s a majority winner, the counting ends. If there’s no majority winner, the process continues, with candidates getting eliminated until one candidate gets a majority.
What this means in practice is that more candidates become relevant. In a simple plurality election, there’s a strong tendency to focus on the top two or three candidates, and to consolidate around a clear front-runner. Once a race dynamic gets set early, candidates outside the top two or three tend to struggle to gain media attention, fundraising, and activist support because they are not seen as being relevant.
But under a ranked system, where second, third, fourth, and fifth preferences are also relevant, more candidates see a potential shot at winning. More candidates become relevant. Voters want to know about more candidates. Candidates campaign more broadly to score second- and third-place rankings. Candidates form alliances — the most notable in New York being the alliance between Garcia and Andrew Yang, whose supporters were asked to rank Garcia second. The race becomes more dynamic, and more interesting.
Dynamic, competitive races lead to higher turnout. More voters are engaged, and more voters matter. Longer term, this can change both how mayors and city council members might govern as they think ahead to their next election. It can change the types of candidates who might enter.
More immediately, it gets voters accustomed to the idea of a “second-choice” candidate winning office — someone who is broadly acceptable to most voters, a baseline consensus. It also changes how much voters know about various issues, because they spend more time learning about candidates and their positions when they get to rank five instead of pick just one. A more engaged citizenry is essential for a more representative and more responsive system.
Perhaps the biggest criticism of ranked-choice voting leading up the election was that it would be too difficult for voters, because it involves them not only having to consider more candidates but also to rank them accordingly. And New York City’s Democratic primary in particular was teeming with candidates, many of whom were unfamiliar to voters.
To be sure, ranking involves more mental energy than a simple one-and-done vote. And roughly a quarter of voters still only ranked one candidate.
But the vast majority took advantage of ranking, and numerous reporters discovered that actual voters found ranking simpler and more straightforward than critics had asserted. “I brought a bunch of flyers and I Googled everybody online,” said one voter. “It’s real easy if people just learn how to read,” said another. “It demands more of voters — it requires you to be more informed,” said a third. “I think people are going to be less afraid of throwing away their votes.” Polls confirmed the reporting: 95 percent of voters considered ranked-choice voting to be simple.
Though plurality voting is often cast as being simpler, in crowded races like a primary, plurality voting can actually be more complicated because voters have to guess whether or not they might be wasting their vote by supporting a candidate who is not polling close to the top. The advantage of ranked-choice voting is that voters can sincerely rank their preferences “I didn’t have to be so strategic,” said one voter. “I was willing to vote for the person I actually wanted instead of anticipating who was ahead in the polls.”
Though some have complained about the delays in tabulating results, these are primarily due to New York City’s decision to allow extra time for all absentee ballots to arrive, ensuring all voters get counted. By contrast, under the old system of runoff elections, it would take longer to find out the winner. Runoff elections also allow candidates to recalibrate their message to highlight new issues or appeal to certain constituencies — since the polls closed, for example, surrogates for Eric Adams have suggested that ranked-choice itself was being used “to eliminate the candidate of moderate working people and traditionally marginalized communities.” In truth, ranked-choice has ensured the election results remain a reflection of how voters felt about the issues on Election Day.
Furthermore, a traditional runoff election would have cost taxpayers at least another $15 million — with much lower turnout. As Maya Wiley noted: “Here in New York City, ranked-choice voting will prevent a runoff election, which would see significant voter drop-off, especially in communities of color.” Notably, with Adams and Garcia neck and neck in the latest tabulations, this year New York City will either elect its first female mayor or its second-ever Black mayor. And for the first time ever, women will hold a majority of city council seats.
With 125,000 absentee votes still to be tabulated, Garcia may well overtake Adams. If so, it will be because Garcia worked hard to build her appeal to supporters of other candidates, and campaigned with Andrew Yang — exactly the kind of coalition-building ranked-choice voting rewards and encourages.
Primary elections are a particularly amenable to ranked-choice voting because primaries often attract crowded fields, and crowded fields mean that candidates can win with mere plurality support. (Four states adopted ranked-choice voting for presidential primaries in 2020. This worked especially well with mailed ballots, since voters didn’t have to worry as much about their first choice dropping out. If that happened, they could simply move on to their second choice.)
More broadly, many (including me) have touted ranked-choice voting as a potential mitigating force to hyperpartisan polarization, because it encourages more inclusive and compromise-oriented campaigning. In one-party towns like New York (and many other liberal cities), ranked-choice voting tends to generate more civil and less negative campaigning. The logic is simple: In an election in which second-, third-, and sometimes even fourth- and fifth-choice preferences can matter, candidates win by building alliances and drawing broad support. Negative campaigning is especially likely to backfire, and may have hurt Adams in particular, who called Yang a “liar” and a “fraud” in the waning days of the race and so far hasn’t benefited from the ranked-choice sorting.
However, as a nationwide fix to hyperpartisan polarization, the single-winner form of ranked-choice voting used in the New York City primary runs up against a significant obstacle: urban-rural polarization. Because Democrats tend to live in the cities, and Republicans outside the cities, most districts (drawn to keep communities together) wind up being so solidly Democratic or solidly Republican that the general election winner is never in doubt. Call it natural gerrymandering.
This is where the multi-winner form of ranked-choice voting, in which more than one winner prevails, becomes crucial. By expanding the number of winners (ideally to five members), this proportional form of ranked-choice voting allows a more diverse range of candidates to win, ending the winner-take-all dynamic inherent to any single-winner election (even those using ranked-choice voting).
This is hardly a new innovation. It’s a system Ireland has used for a century. And it’s a voting system that New York City actually helped pioneer in the early 20th century. But in the mid-20th century, proportional ranked-choice voting fell into disfavor precisely because of the diversity of candidates it helped elect. It may have been a voting system ahead of its time. But maybe that time is now here, finally.
