Advertisement
Essay
Ranked choice voting in NYC showed us what elections could look like
On June 23, the “Late Show with Stephen Colbert” had an unusual pair of guests, New York City Comptroller Brad Lander and New York State Assemblyman Zohran Mamdani. They took turns praising each other. Lander said he found Mamdani inspiring, lauding his effort to lead a debt relief program for taxi drivers. Mamdani said that Lander had been “a principled and progressive leader” who is responsible for many of the laws that are celebrated in the city.
The catch: They were running against each other in New York City's mayoral Democratic primary.
If you are used to politics as usual, this might seem like a cruel joke. We are so accustomed to malicious mudslinging that it is hard to even imagine civility and decorum in our elections. But this deficit is not hardwired into the system; it is a by-product of the mechanisms that run our democracy, mechanisms that we have the power to change.
One of those is plurality voting (also known as relative majority or first-past-the-post), which we’re all familiar with. Plurality voting is used to choose all but a handful of the approximately 520,000 elected officials in the U.S., including our president. Among its many flaws, plurality encourages negative campaigning because it’s a winner-take-all system.
In the usual plurality elections, a voter selects one candidate, and the candidate with the most votes wins. Seems simple and intuitive, and it is. It is also deficient because a candidate can win with less than 50% of the vote. Plurality can lead to vote-splitting (when similar candidates divide a majority of the votes, allowing a less popular candidate to bubble up to the top) and spoilers (a candidate who has no chance of winning siphons away enough votes from a major candidate, enabling a third candidate to win), which means that a candidate who isn’t actually the number-one choice of the people comes out on top.
I write this from a Massachusetts district represented in Congress by Jake Auchincloss, who won his Democratic primary in 2020 with only 22.4% of the votes. Several progressive candidates split the vote, and the rest will be history: Because of the political infrastructure that protects incumbency, Auchincloss, an undeserving winner, will likely sit in Congress for another 30 years.
But Massachusetts isn’t alone. Many of our elections, from school boards to the U.S. Senate, produce winners supported by fewer than 50% of the voters. Whenever this happens, there is a the danger that vote-splitting or a spoiler candidate was involved.
Advertisement
In 2021, New York City started using a different system — ranked choice voting (RCV) — which takes the mud out of the equation. RCV is better at capturing people’s collective preferences. In an RCV election, a voter ranks the candidates by order of preference. The voter does not have to rank all of the candidates; ranking fewer – even only one – is fine.
Ranked ballots are tallied like this: If a first-choice candidate wins over 50% of the votes, they are the winner. If no candidate crosses that threshold, the candidate with the fewest number of first-place rankings is eliminated. The votes for that candidate are then transferred down to the second-choice candidate on those ballots. In effect, whoever had the eliminated candidate at the top of their ballot now gives their vote to their second-choice candidates. This process continues until someone crosses the 50% threshold.
RCV eliminates vote-splitting and spoilers by virtue of voters being able to indicate their choices by order of preference. It also encourages voters to cast their first-place preference honestly — without any worry about “wasting” their vote. (If you vote for a less popular candidate in a plurality election, your vote is irrelevant, and, what’s worse, you run the risk of contributing to a spoiler). From a mathematical point of view, RCV asks for more information from the voter so that the algorithm can do more with it. Because it has more data to work with, RCV is much better at representing the will of the voters than plurality voting.
Plurality is a zero-sum game – fewer votes for everyone else means more votes for me – a game that breeds cutthroat ruthlessness. Being nice to a competitor is not a winning strategy. But for RCV, it is. This method rewards positive campaigning since the iterative process rewards candidates ranked near the top. This is precisely why Lander and Mamdani decided to play nice and cross-endorse each other. With a smile, they each told the electorate, “I hope you rank me first, but if you rank the other guy first, then please rank me second.”
The strategy worked. No candidate won over 50% of the vote outright. (Mamdani came the closest with 43.5% of the votes). The RCV machine kicked in and Mamdani received almost 100,000 additional votes from those who ranked him second. We still don’t have the details about which candidates those second-place votes came from, but it is highly likely they were largely from voters who had ranked Lander first. The Mamdani-Lander cross-endorsement, combined with the “Don’t rank Cuomo” movement, did the job. After the first RCV round, Mamdani won with 56% of the votes.
Cuomo did not play to the strengths of RCV; for example, he did not reciprocate the endorsement of two other candidates (although that may not have made a difference since those two only received about 12,000 votes). After the first RCV round, Cuomo only garnered about 53,000 second-place votes, not nearly enough to overcome Mamdani’s gain. Ranked choice underscored that Cuomo is not a casualty of vote-splitting or spoilers: he was simply truly disliked by voters.
But the real story from this primary election is how the voters felt about RCV. (Mamdani still would have won if this had been a plurality election, so RCV didn’t deliver a plot twist.) Turnout was historic — the highest in decades — which supports the research surrounding responsiveness to RCV. Voters typically find RCV easy to use and engaging. This election was no different.
Bostonians should be excited about all this since an effort to bring RCV here is underway – and support for it is growing. In a major milestone from last May, Boston City Council voted 8-4 to advance a Home Rule Petition to implement RCV in Boston’s elections. RCV’s future in Boston looks good. As it should.
Follow Cognoscenti on Facebook and Instagram. And sign up for our weekly newsletter.
