In recent years, hundreds of U.S. cities have adopted ranked choice voting (RCV). Unlike traditional elections, in which voters select a single winner, RCV allows you to rank candidates in order of preference. Proponents argue this helps voters express their true preferences and ensures the winner better represents the community.

But does it work? SPA Assistant Professor of Government Jonathan Colner and coauthors Cheryl Boudreau and Scott A. MacKenzie (both of University of California – Davis) take up this question in “Start Spreading the Views: Improving Political Expression and Representation in Local Elections Using Ranked Choice Voting,” published in Political Research Quarterly this summer.

“Like any voting system, RCV has an array of advantages and disadvantages [across] different electoral environments,” Colner said. “Our project, by focusing on comparability across office level and prestige, helps us understand what types of electoral environments are best suited to RCV, and what kinds of interventions work . . . to encourage voter knowledge and political expression.”

While people usually have plenty of information about mayoral candidates, they often know very little about "down-ballot" offices like city comptroller or city council.

“Lower prestige offices receive less attention from newspapers and have less money for voter outreach, leaving voters unsure of the policy positions of the candidates,” said Colner. “With less information, voters will often rank fewer candidates or be less efficient at ranking those most aligned with their policy preferences.”

To test this, the authors conducted a study during the 2021 New York City Democratic primary, the largest municipal RCV implementation in U.S. history. They surveyed over 900 residents, measured their policy views on 20 local issues (e.g., police funding, charter schools), and asked participants to rank candidates for mayor, comptroller, and city council. Random voters were shown “endorsements” from an ideological group or local labor unions.

Their findings confirmed that voters are much better at using RCV for "high-prestige" offices. For mayor, voters were highly likely to rank multiple candidates and successfully choose those who matched their policy views. For city council, voters ranked fewer people and often chose candidates whose views were opposite to their own.

“To rank multiple candidates, voters must not only know their most preferred candidate, but also how they’d comparatively rank up to four more,” Colner explained. “Doing this effectively requires a significant amount of political knowledge.”

Finally, the authors found that information from outside groups can help "fix" RCV’s issues in lower-level races. Ideological endorsements helped voters identify which candidates actually shared their policy preferences, significantly improving the "match" between voter and candidate. Union endorsements encouraged people to rank more candidates and helped them pick winners, though not necessarily a perfect policy match.

Ranked choice voting is not a "magic wand" for democracy, the article maintains. Its success depends entirely on the information environment. For RCV to improve representation, the authors suggest that cities must do more than just change the ballot—they must help "spread the views" by making candidate positions and endorsements more visible to the public.

Colner and coauthors have several related research projects upcoming, including one examining the impact of personalized voter guides on the 2020 RCV elections in San Francisco, and another measuring the effect of endorsements covering multiple candidates.