Study: ranked-choice elections don’t reduce polarized voting

A voter filling out an election ballot.

Citizens still tend to vote in racial blocs as much as in the old system

Election day is here again, and more than 10 cities around the country will use ranked-choice voting (RCV), a system where voters rank their preferences instead of casting their ballots for one candidate. Proponents of RCV say it’s less racially polarizing than the standard two-round primary because candidates have to appeal to voters from different racial backgrounds to win a majority. But a recent San Francisco State University study suggests that RCV doesn’t live up to its campaign promises.

The research, from San Francisco State Associate Professor of Political Science Jason McDaniel, was published in the California Journal of Politics and Policy last month. The study concludes that RCV does not cause reductions in racially polarized voting.

McDaniel, who has taught political science at SF State since 2009 and previously published research linking RCV to a drop in voter turnout rates, began delving into the effect of RCV on racial voting blocs five years ago. Taking precinct-level estimates of voting data in San Francisco and Oakland mayoral elections from 1995 to 2015, McDaniel calculated voting preferences between black, white, Latino and Asian voters, comparing elections before and after the implementation of RCV in both cities to calculate the voting divide. In the 2015 San Francisco mayoral election, for instance, 77 percent of Asian voters chose the winning candidate, Chinese American Ed Lee, while 48 percent of white voters voted for him, making the vote divide between the two groups 29 percent.

Jason McDaniel

Associate Professor, Jason McDaniel

McDaniel’s initial research suggested that RCV could sometimes reduce the vote divide between racial groups. But in order to get a better grasp of what was going on, McDaniel compared his findings from RCV cities with cities using a traditional primary system. In terms of the racial voting divides, McDaniel saw no difference.

“Any reduction in racially polarized voting that we’ve seen was caused by other factors, not by ranked-choice voting,” he said.

Those factors, McDaniel says, are difficult to pin down, but include the demographic makeup of the cities and candidates. The most important factor is a candidate’s race.

“When there is a same-race group candidate, group cohesion increases and racial polarization increases,” he said.

While last summer’s special mayoral election in San Francisco was not included in McDaniel’s study, he plans to delve into the data and determine how it relates to his current research.

“Sometimes people will say ‘Identity politics is a problem, these things are keeping us from coming together,’ and that’s just the incorrect way of seeing it,” said McDaniel. “Politics is rooted in these identities and these are the identities that matter most to people when voting … So rather than trying to get rid of them, take them seriously.”

But McDaniel says he does not want his work to present the idea that politics is irreparably cleaved along racial lines.

“There are actually multiracial and diverse coalitions on both sides,” said McDaniel. “But it’s not about changing the voting system. It’s about working out compromises and governing, and we need to encourage our politicians to do that.”