Author’s note: In early 2023, I wrote about Oregon House Bill 2004, which let voters decide whether to adopt ranked choice voting in statewide elections, including those for president, US House and Senate, and governor. If adopted, it would also give cities, school districts, and other local entities in Oregon guidelines to adopt ranked choice voting in their elections.
The legislature passed the bill in June 2023, putting it to voters to approve or reject this fall as Measure 117. Today, by way of offering an explainer on what a switch to ranked choice voting looks like for Oregon voters, I’m re-sharing my 2023 research on the substance of that bill, including four ways evidence shows ranked choice voting gives voters more voice and more choices in elections. I’ve made a few minor updates to reflect the measure on the ballot in Oregon this November.
How ranked choice voting works
Ranked choice voting would make some key changes to Oregon’s current pick-one voting system, particularly in contests where voters have more than two candidates to choose from (as they do in almost every state election).
Using ranked choice voting, Oregon voters would no longer be limited to selecting a single candidate (though they could choose just one if they wanted to). Instead, voters would be free to rank candidates in order of preference.1
If no candidate receives more than 50 percent of first preferences, ballots go to further rounds of counting where last-place candidates are eliminated and their voters’ later choices are considered, continuing until one candidate has over 50 percent of the vote among the remaining candidates.
Here are four reasons voters tend to like the option to rank candidates on their ballots:
1. Winners earn stronger bases of support
Pick-one elections often work fine for elections with just two candidates in the running: voters select their favorite, and the person with more votes wins. But things get tricky when voters have three or more candidates to choose from.2 Maybe only 40 percent of voters cast their ballots for the winner, while two contending candidates each receive 30 percent—so the winner has less than majority support. Or maybe there are ten candidates, and the winner comes out with support from only 15 or 20 percent of voters—quite the minority! When the top candidate has a plurality (the most votes) but not a majority (more than half of the votes), more voters voted against the winner than for them.
Of Oregon’s seven gubernatorial elections since 2000, four saw the winning candidate finish with less than 50 percent of the vote.
This situation happens all the time in Oregon. Of the seven gubernatorial elections since 2000, four saw the winning candidate finish with less than 50 percent of the vote.3 In 2022, with high-profile nonpartisan candidate Betsy Johnson drawing nine percent of the vote, Tina Kotek won the race for Governor with only 47 percent of voters supporting her.
It’s even worse in primary elections. Oregon’s Republican primaries for governor since 2000 usually saw nine or ten candidates,4 and in only one of those races did a winner receive a majority of the vote.5 2022 was a particularly extreme example of this, with 19 candidates running; Christine Drazan came out on top but garnered only 23 percent of the vote. But say those other 77 percent of voters really didn’t like Drazan, and voters would have jointly preferred another candidate even if their first choice lost. Those 18 other candidates split the anti-Drazan vote, and the Drazan supporters, a small minority overall, got to pick their favorite, overriding the majority.6
Ranked choice voting would help that majority coalesce around a single candidate, mitigating vote splitting among multiple similar options. Since Drazan wouldn’t be immediately elected in the first round of counting, later rounds might show that a stronger candidate had the support of more voters. Or they could show that other voters did support Drazan, giving her a clearer base of support heading into the general election.
By counting voters’ later choices once their first choice has been eliminated, ranked choice voting helps identify stronger candidates that are supported by more voters. That means stronger party nominees, happier voters, and officeholders that know they’ve appealed to Oregonians.
2. Minor-party candidates don’t invert results
Pick-one voting is also heavily vulnerable to spoiler candidates, a specific type of vote splitting where a minor candidate siphons enough votes away from the leading major candidate closest to their platform that the major candidate loses the election. This happened quite famously in the 2000 presidential election in Florida when Green Party candidate Ralph Nader, whose voters generally preferred Democrat Al Gore over Republican George W. Bush, received nearly 100,000 votes and Bush beat Gore by just over 500 votes.
Closer to home, the 1990 Oregon gubernatorial election saw anti-abortion independent Al Mobley draw 13 percent of the vote, likely from voters who otherwise would have supported Republican Dave Frohnmayer, leading to the election of Democrat Barbara Roberts with the support of only 46 percent of voters. And in Maine, far-right Governor Paul LePage won his two elections when voters split between a left-leaning independent and a Democrat. This was one major impetus behind Maine’s ballot measure moving the state to ranked choice elections.
Oregonians deserve the ability to vote for candidates who share their values and represent their communities, but the state’s current election system too often discourages voters from doing that. Almost every election comes with a discussion of voting for somebody who’s “more electable,” being sure not to “waste your vote,” or even “a vote for Candidate X is a vote for Candidate Y.” Voters are wary of voting for a less popular candidate they might prefer, worrying that doing so may just help their least favorite major-party candidate win.
But just like vote splitting more generally, ranked choice voting helps deal with spoiler candidates sensibly. The minor candidates are eliminated in early rounds because they received the fewest votes, and their voters’ later choices are counted to see which remaining candidate is preferred by a majority of voters. A minor candidate with only a few points of the vote won’t throw the election to a different major candidate.
3. Voters can choose their honest favorite
Due to a (very sensible) fear of slim plurality winners and spoiler candidates, the pick-one system incentivizes many voters to cast their ballots strategically, picking a candidate they see as electable rather than the candidate they like the most—somebody they think can win, not somebody they think should win. And campaigns take full advantage of this, spending chunks of their time and money on arguments about feasibility instead of making the substantive case for their candidate. I can’t even count the number of pro-Tina Kotek ads I saw in the fall of 2022 that said “I like Betsy Johnson, but she can’t win” or “a vote for Betsy Johnson is a vote for Christine Drazan.”7
Ranked choice voting changes this dynamic entirely. Since their second choice will count if their first choice is eliminated, voters can support a less popular candidate without “taking votes away” from a more popular candidate. Maybe the first round of that Florida election looks the same, with Nader receiving two percent of the vote, but the second round of counting, when Nader voters’ second-choice votes are tallied, puts Gore over the top. Or you can honestly vote for Betsy Johnson first and still know that if she loses in the first round, your vote will count for Tina Kotek in later rounds.
4. More representative candidates can run and win
Thanks in large part to vote splitting, voters and candidates in a pick-one system know they must act strategically, angling to end up with a least-bad outcome. Voters often judge women as “less electable” than men, pushing women candidates more often into the trap of feasibility and strategic voting. And candidates often must decide even before entering a race whether their presence would lead to vote splitting and help elect a minority-preferred contender. This frequently applies to less established third-party candidates, women, and people of color. In Chicago’s 2023 mayoral election, the chair of the City Council’s Black Caucus publicly called for some candidates to drop out to avoid splitting the vote among Black candidates.
When communities can better coalesce around their preferred candidates under ranked choice voting, those candidates can worry less about electability or vote splitting. That likely means opportunities for more women, people of color, and minor candidates to run and earn votes with lower likelihood of throwing the election to a less-preferred winner. One study found that in the Bay Area, women were more likely to win elections in ranked choice cities than in cities that had not adopted ranked choice voting.
In 2006, San Francisco used ranked choice voting to elect a new Supervisor from the majority-Asian District 4, and four Asian candidates ran against two white candidates. While no candidate had more than about a quarter of the vote in the first round of counting, large numbers of second- and third-preference votes from supporters of other Asian candidates in the race helped put Ed Jew over 50 percent in the final round. Even if votes were split among those Asian candidates in round one, ranked choice voting helped those voters to join forces with their second- and third-choice votes behind a candidate of their choice, even with multiple candidates running.
What Oregonians can expect with ranked choice voting
Ranked choice voting would help to mitigate the current pick-one system’s high potential for vote splitting without placing heavy limitation on voter choice and candidate entry. Candidates win by running campaigns more focused on the issues, dodging negative rhetoric and arguments about electability. With ranked choice voting, Oregonians can expect winners who’ve received votes from a majority of voters, a greater range of choices on their ballot, and officeholders more likely to represent their preferences.
Thanks to former Research Assistant Nakeshia Diop for contributing research to this article.
