Ranked Choice Voting
RCV in NYC: Who Didn’t Rank?
An examination of NYC voters that chose not to rank candidates in the June 2021 primary election.
Written by: Jay Wendland and Erin Carman
On June 22, 2021 New York City became the largest population in the US to use ranked choice voting (RCV). It is not the first municipality in the U.S to use it, as San Francisco has since 2004 and has been joined by several other cities including Santa Fe and Minneapolis. Additionally, Maine adopted RCV in 2018 and has used it in local, state, and federal elections; Alaska followed Maine’s example by adopting a version of RCV by ballot initiative in 2020. While RCV has expanded its usage (it is projected to be used in 74 jurisdictions in upcoming elections), its impacts are still not fully understood.
Proponents of RCV point to various benefits it brings with it. It provides an avenue for voters to better express their preferences, which should lead to increased feelings of political efficacy, and thus perhaps increased turnout. There was indeed increased turnout in the NYC primary in June, with 23 percent of voters showing up to vote across party identification. 2013 was the last year to have both Democratic and Republican mayoral primaries and the 2021 turnout increased by almost 250,000 more voters. There was a slight decrease (1 percent) in Republican turnout, but an increase (3 percent) in Democratic turnout. This was only one election, so we should not treat this singular election as a sufficient example of RCV driving up turnout, as others have pointed out, changes liker higher public matching dollars were also different in 2021. But it may lend credence to the idea that RCV increases turnout — at least for some voters.
RCV allowed voters the ability to express their preferences through ranking the candidates running for office. In New York’s primary, voters could rank up to five candidates, and in so doing we saw 85 percent of all ballots still active in the last round of tallying. In the mayoral race, 88.5 percent of voters ranked more than one candidate, with 48 percent ranking on all five lines. Most voters in the Edison Research Exit Poll that opted to rank did so for reasons related to feelings of political efficacy. In the city council races, we saw the percentage of voters that opted to rank candidates drop to 69 percent. Similar to the mayoral race, we saw similar reasoning as to why these voters opted to rank.
However, a topic that has been underexplored is why voters have opted against ranking. By choosing to only rank one candidate and not fully engage with the RCV process, these voters are opting to treat the election as though it were a plurality election. These voters chose to only rank one candidate, despite the outreach efforts by candidates, parties, nonprofits, and other organizations to help educate voters about the RCV process and encourage them to engage with it. So, why did New York voters opt not to rank? Based on the Exit Poll data collected by Edison Research, which polled over 4,000 voters, many of these voters made a conscious choice not to rank. In fact, the top reason for not ranking more than one candidate was that the voter only liked one candidate (80 percent). Others chose not to rank because they did not know enough about the other candidates, were afraid that ranking candidates would hurt their preferred candidates, did not know they could rank candidates, or did not understand how to fill out their ballot. Regardless of why these voters opted not to rank, it seems clear that a vast majority made a conscious choice to not rank.
However, it is important to dig a bit deeper into who chose to not rank to determine if there are systematic problems that need to be addressed. First, when we examine racial patterns among those that did not rank candidates, we see that minorities were slightly less likely to rank candidates. About six percent of white voters ranked only one candidates, while 16% of Black voters, 17% of Hispanic voters, and 10% of Asian voters opted to rank only one. When we examine gender differences, we see there is no difference in the percentage of men and women who opted to only rank one candidate (11%).
Analyzing age-related differences highlights an important set of gaps that exist. Only eight percent voters under 50 years old ranked only one candidate, while 14% of those over 50 ranked only one candidate. Finally, we see a pretty large discrepancy in non-ranking participants when we look at education.
Nearly 1 in 4 (23%) of those without any college education ranked only one candidate, while only 6% of those with an advanced degree chose to rank only one candidate.
Thus, while we saw NYC primary voters embrace ranked choice voting in its inaugural use— there were some patterns that emerged that warrant further exploration and caution before this new electoral mechanism is more fully embraced. While most of those who opted against ranking candidates made an informed choice not to do so (80%), it is still important to note that these voters were mostly minorities, older voters, and those without a college education. So, as NYC prepares for future elections in which RCV will continue to be used, these groups of voters may need further outreach and education regarding the RCV process.
