On a November ballot initiative, Massachusetts voters will be asked if they would like to change our election system to ranked-choice voting. Instead of casting a ballot for a single candidate, voters will be asked to rank all the candidates in order of preference. If no candidate gets an outright majority, then through a series of additional tabulations, voters’ second, third, fourth, and lower preferences are added to the vote until one candidate has accumulated more than 50 percent and is declared the winner.
The main rationale for adopting RCV is that it prevents election results like the one in the Sept. 1 primary in the Fourth Congressional District. With nine candidates on the ballot, Jake Auchincloss won the primary election with only 22 percent of the vote. Many people think it is unfair that a candidate can win with only 22 percent of the vote. They may also think it’s unfair when two or more candidates who offer similar platforms split the vote, and thereby give the election to a candidate with different policy priorities. RCV is meant to address these problems. With voters able to rank candidates, the eventual winner will have a majority of support once voters’ lower choice preferences are taken into account.
But as with any reform, ranked-choice voting solves one set of problems while creating new ones. Here are some issues worth considering.
First, there is a basic principle in psychology called loss aversion: People tend to feel losses more strongly than they feel gains. Nobody likes it when their candidate loses an election. But people will feel burned if they think their candidate won and then the victory was taken away from them, especially when this happens to candidates from demographic groups that have been historically under-represented in political office.
Consistent with this concern is evidence from Jesse Clark, a PhD student at MIT who is writing his doctoral dissertation on RCV. Using experiments in which some participants use RCV ballots and some use traditional ballots, Clark finds that RCV makes voters feel less confident in the system and that the RCV rules “are stacked against the voter and their party.”
Second, RCV makes voting harder. Clark’s evidence shows voters find RCV ballots harder to fill out. We also know from studying the results of RCV in other communities that a substantial number of voters, about 10 percent, but as many as 27 percent, do not rank enough candidates for their votes to count in the final tally. Experts call this ballot exhaustion or ballot truncation. The more candidates on the ballot, the more widespread the problem.
Many scholars argue that ballot truncation affects who wins and renders RCV problematic. A key question is which voters typically do not fill their ballots out completely in an RCV election, because groups of voters lose voting power when they don’t fill out their ballots fully. It’s a hard question to answer because the secret ballot prevents researchers from determining the demographics of voters who cast truncated ballots.
Nevertheless, we do know that a basic pattern tends to emerge when voting is made more complicated. The demographic groups who are disproportionately affected when voting is made harder are low-income, less-educated, and racial-minority voters. In Maine, which recently adopted RCV, truncated ballots were significantly more commonplace in communities with lower rates of college education.
Proponents of ranked-choice voting argue that voters are free to rank just one candidate. That is true, but it means that those ballots will count less than the ballots of voters who rank all candidates. By adopting RCV, we would abandon a system that guarantees ballot equality in favor of a system where people like me — a PhD who loves politics and can easily rank nine candidates in an election — have yet another leg up on voters who don’t have the luxury to learn about so many candidates. If my first-choice candidate is eliminated, my second-choice candidate gets a vote. If you didn’t rank all the candidates, your vote counts less than mine.
Third, RCV proponents believe that the current system discourages citizens from participating because they may feel their ballots are wasted or that the candidates whom they like don’t stand a chance. Perhaps adopting RCV will increase turnout to such a degree that it will offset issues like truncated ballots, particularly among marginalized communities. Unfortunately, evidence suggests that turnout is unaffected or negatively affected by the implementation of ranked-choice voting.
Finally, RCV is relatively new, and there is much we do not know about its effects on elections. It takes longer to fill out an RCV ballot. Will that contribute to long lines on Election Day? Will RCV favor well-financed candidates more or less than the current system? Will RCV harm or help the prospects of candidates of color? The evidence is mixed, and comes almost exclusively from a few communities in the Bay Area of California, where the racial dynamics are different from Massachusetts. Voters who are considering the ballot initiative will have to weigh what kind of appetite they have for these unknowns. One thing is almost certain: The state will be subjected to lawsuits contesting the legality of the measure.
Proponents of RCV believe that avoiding elections won with small pluralities gives the democratic process more integrity, a more important value than any potential downsides, including an extra burden on voters. Maybe they think RCV doesn’t make voting meaningfully harder or more complicated. Research suggests otherwise. There are plenty of ways to make our election system better. Let’s advocate for ones that lower barriers to participation rather than raise them.
Eitan Hersh, a political scientist at Tufts University, is author of “Politics Is for Power.”