When the New York City Board of Elections reported the first batch of Democratic mayoral primary results last week, Eric Adams seemed to have a commanding lead. But as election workers continued to count ranked-choice ballots, Adams’s lead eroded. As of this writing, roughly 15,000 votes separate him from next-place finisher Kathryn Garcia.
The Adams campaign has made known its displeasure with ranked-choice voting, which New York City used this year for the first time since 1947. City Councilman I. Daneek Miller, an Adams surrogate, told the New York Times that the system was enabling “an attempt to eliminate the candidate of moderate working people and traditionally marginalized communities.”
Is this characterization fair?
In certain circumstances, it’s possible that ranked-choice voting would amplify the voices of privileged voters. But the evidence that it does so is mixed. And ranked-choice voting could also push politics in more egalitarian directions.
If ranked-choice voting did deepen inequities, here’s how it might happen
The Adams campaign has not issued a long-form critique of ranked-choice voting. But the most convincing version of its concerns might go as follows.
On a New York City ranked-choice ballot, instead of choosing just one candidate, voters can rank up to five candidates in order of preference. If your favorite candidate or candidates do not have enough support to win, your vote is redistributed to the candidates you’ve listed further down.
The backup choices on the ballot are invitations to exercise power. If you rank a candidate as your fifth choice and your top four choices are eliminated, you still have a say in who wins. Of course, ranking so many candidates can be hard; many voters do not fill out complete ranked-choice ballots. The savvier voters, those who enjoy deciding on fifth choices in a mayoral primary, might be the same people who already have a firm grip on politics: White, wealthy, well-educated voters. If that’s the case, ranked-choice voting would deepen the influence of already powerful groups.
Imagine two New York City voters who each ranked Andrew Yang first and Raymond McGuire second. One of them marked a third preference for Garcia. The other thought Adams was her third choice, but wasn’t sure and left the rest of her ballot blank.
As ballots were tabulated, the count revealed that neither Yang nor McGuire had enough support to win. So the ranked-choice voting algorithm set out to reassign their supporters’ ballots. The first voter’s ballot was transferred to Garcia. But our second voter’s influence was over. Since she gave no preference after Yang and McGuire, her ballot was exhausted.
Maybe this voter left her third-choice blank because she was genuinely indifferent. Or perhaps she had the beginnings of a preference for Adams and would have expressed it but for the burden of collecting the necessary information, something political scientists call “information costs.” In this scenario, Garcia would gain on Adams by one vote.
Suppose that, all across the city, the “information costs” of ranked-choice voting fell hardest on marginalized communities, and that these communities would have preferred Adams to Garcia. In such a scenario, ranked-choice voting would indeed deepen inequalities.
As Georgia’s new voting law shows, when Black people gain local power, states strip that power away
But is that how it actually works?
The first piece of this argument is hard to deny. Using ranked-choice voting requires voters to gather and consider a lot of information. Could you name a fourth choice in your jurisdiction’s most recent mayoral primary? The research confirms that ranking several candidates is challenging.
Scholars have also found that well-off White voters have more information about electoral politics, at least when asked trivia-style questions about how the government works and who holds office. But despite these background factors, no clear evidence suggests that the complexity of ranked-choice voting burdens disadvantaged voters.
In fact, one study found that San Francisco neighborhoods with more Black voters were more likely to complete full ranked-choice ballots. And when political scientist Joseph Coll asked 2020 Democratic primary voters to rank their top five candidates for president, he found no association between race and the number of candidates a respondent listed.
There is some evidence that ranked-choice voting’s complexity hurts voters with less education and income. When political scientist Nolan McCarty produced a report on Maine’s ranked-choice voting, he found that precincts with more voters who did not complete college had higher rates of incomplete ballots. And Coll’s survey of Democratic primary voters found that wealthier respondents ranked more choices.
It’s possible that ranked-choice voting reinforces divisions along class but not racial lines. But the studies paint a somewhat inconsistent picture — and they rely on data of limited use. The data tell us how voters have used ranked-choice voting, not how they would use it in a world where the press, political parties and civic organizations fully adapt to the new ballot. Groups might learn to issue slates of endorsements that help voters connect their first preferences to down-ballot choices. We might ask whether voters could equitably use ranked choice under these circumstances, not only what voters have done in the past.
Let’s look at the big picture
Ranked-choice voting might be so complex that it reinforces inequality. Ranked-choice voting could also make our politics more egalitarian precisely because it asks for more information from voters.
For instance, as the Republican Party has tacked to the right on such issues as voting rights and immigration, it hasn’t faced a credible challenge from a new center-right party. Ranked-choice voting could make it easier for moderates to mount such an effort. In a general election that used ranked choice, if you wanted to vote for a new center-right party but still preferred Republicans to Democrats, you could rank a center-right candidate first and the Republican candidate second. Knowing that voters had more choices, the GOP might feel pressure to adopt more moderate stances.
Ranked-choice voting has pros and cons, as any election system does. As Americans decide whether it belongs in our push for a functioning democracy, we may wish to consider all its elements in concert.
Don’t miss any of TMC’s smart analysis! Sign up for our newsletter.
Benjamin P. Lempert (@Ben_Lempert) is a PhD candidate in the University of Michigan’s department of political science, and a recent graduate of the University of Michigan Law School.