Conclusion

With ranked-choice voting expanding in popularity across U.S. cities and states, research has followed. However, the research has yielded few clear conclusions. For the most part, the effects of RCV appear to be generally positive, but marginal in many areas and absent in others.

At this point, we can point to a few widely supported conclusions. First, voters tend to like ranked-choice voting once they experience it, and most report finding it easy to use. Support tends to correspond to age, with younger voters most enthusiastically embracing change while older voters are more resistant. This is hardly surprising, since age corresponds with support for the status quo in many aspects of life. Campaigns also appear to be more civil and positive under RCV. And to the extent that RCV combines the primary and the general election into one, it increases turnout.

However, many of the other hoped-for benefits, such as more diverse candidates (by gender, race, and ideology), higher turnout, and more viable parties are harder to detect. Nor is there any evidence that RCV changes policy outcomes.

At the same time, the research should also allay fears that RCV is too confusing or discriminatory. Voters get RCV, and they like RCV. To the extent there are gaps in understanding, these gaps can be overcome through education and training. Familiarity builds comfort and support. And information and training can overcome obstacles.

Research on RCV’s impacts is limited by several factors. First, and most obviously, reforms take several election cycles to impact candidate and politician behaviors, as political actors and voters learn and adapt over time, and norms change. Political change is a complex process. And it is still possible that as RCV spreads and comes into wider usage, its benefits may come into clearer focus. After all, the experiences of other countries using RCV are mostly positive, and there are good theoretical reasons to believe RCV will have meliorating effects on our politics, especially as voters and candidates adjust to the system.

Second, the fact that cities do not choose to adopt RCV randomly challenges researchers’ ability to draw clear conclusions about how much before and after changes were due to the implementation of RCV as opposed to other trends and factors. Several researchers have attempted to work around this by using matched cities or synthetic controls, but given the small sample sizes and the idiosyncrasies of cities like Berkeley, Oakland, San Francisco, and Minneapolis, such techniques still make clear conclusions elusive.

Third, local politics is very different from national politics. Since most ranked-choice voting implementations have occurred at a local level, we are largely limited in observing the effects in cities that are heavily Democratic, often nonpartisan, and in elections that tend to have very low turnout and citizen engagement. It is possible that effects generalize to a national level. But it is just as possible that the interaction of local politics and ranked-choice voting creates effects that are unique to local politics.

Fourth, ranked-choice voting within the single-winner election contest, one city or state at a time, is a relatively small change. In most elections, the candidate who would have won under plurality voting is also the candidate who won under ranked-choice voting.

Certainly, as more cities and states utilize ranked-choice voting, and over longer periods of time, more research opportunities will emerge, and benefits of RCV as well as any major drawbacks may be easier to observe and measure.