What If It Wasn’t A Binary Choice?
Chris Stirewalt, Charlotte Lawson, Sarah Isgur :: The Dispatch
When New York City voters take to the polls in June’s mayoral primaries, they’ll be presented with the option of ranking their top five candidates to fill their party’s ticket. Ranked-choice voting, though more difficult to conceptualize than traditional plurality voting systems, has taken hold in municipalities across the country for its ability to—in theory—select the candidate backed by the broadest consensus.
So how do different voting methods yield unique electoral outcomes? While the United States had traditionally relied on a winner-take-all vote, election experts have put forth a variety of strategies designed to mitigate unwanted election outcomes. Many have already been incorporated into our voting systems.
The likelihood of a bad outcome varies under different systems, but the inadvertent selection of an unpopular candidate by voters hoping to improve their preferred candidate’s chance is most likely to occur in voting systems with limited choices.
Let’s start with the norm. Nearly every state, with a few exceptions , selects the candidate with the greatest number of votes for most offices. Under the plurality vote, a voter’s second choice is taken into as much consideration as their tenth choice—which is to say, not at all. These races are the easiest for pollsters to gauge, but they’re also the most susceptible to tactical voting, which ups the likelihood of paradoxical outcomes.
In the case of party primaries utilizing a plurality system, the “winner” in “winner-take-all” often skews far to the left or right as a result. Larry Diamond, former director of the Center on Democracy, Development, and the Rule of Law at Stanford University, explained to The Dispatch that low voter turnout combined with a “first-past-the-post” voting method creates a “perfect storm for polarization” as state primary outcomes are dictated by as little as 8 percent of the electorate and moderate voters are discouraged from voting third-party.
“The method doesn’t respect the voters’ second, third, and fourth-ranked candidates,” said Don Saari, author of Decisions and Elections and mathematics professor at the University of California, Irvine. Drawing an analogy to the academic world, Saari explained that the plurality system is akin to a situation in which universities quantified student success only by As received, not accounting for Bs, Cs, Ds, and Fs. The failure of this algorithm is that it allows students with both several As and several Fs to outflank students with more consistent scholastic achievement.
The same comparison applies to ranked-choice, in which plurality is the driving metric despite the voting system’s deference to preferential ballots. In the case of New York City, votes are first tallied by first choice. If no candidate wins a simple majority (highly likely in a field of at least a dozen serious Democratic contenders), the candidate with the fewest votes is eliminated. If your first choice is eliminated, your votes are then transferred to your second choice. The process continues until a candidate surpasses 50 percent of the vote.
The method is sometimes also referred to as “instant-runoff voting,” because like the two-round system, the election requires a certain threshold of votes to trigger a victory. Unlike the runoff system, the criteria necessary to get a candidate past that threshold are built into the first ballot voters complete.
“The problem with a runoff election is it asks a lot of voters to come back later and vote all over again. Ranked-choice voting enables all votes to be counted right there at the same time,” Diamond explained. “You can program a voting machine very simply to do these ballot transfers if no one gets a majority.”
According to the New York Campaign Finance Board , “cities that have implemented Ranked Choice Voting have elected more women and more women of color, making their elected officials more representative of their communities.” In addition to promoting diversity in elected office, ranked-choice also seeks to minimize negative campaigning, lessen the spoiler effect, and broaden public approval of the elected candidate.
Primarily for the latter two reasons, Maine became the first state to implement ranked choice in a statewide election after Republican Gov. Paul LePage twice won office with less than a majority of the vote. Alaska also adopted a form of preferential voting in a November voter initiative, eliminating the party primary in favor of a blanket primary that advances the top four finishers to the general election, which then uses ranked-choice voting to select a candidate.
“There could be two Republicans, a Democrat, a Green Party candidate, a Libertarian, somebody from Mars, and somebody from Venus,” said Diamond, adding that the system afforded Sen. Lisa Murkowski the political flexibility to convict former President Trump because she no longer needs to beat out a Trump-endorsed candidate in the party primary (and is likely to have broader appeal in the ranked-choice general election).
Another alternative is to create multi-member districts. Proportional ranked-choice voting, introduced under the recent Fair Representation Act, is designed to lessen the effects of gerrymandering and diversify the geographical spread of political party representation. But by selectively upping the number of congressional seats in districts across the country, this system results in a very low percentage threshold necessary to be elected to federal office.
“If you have a district size of one person, then you only need to win 50 percent plus one. If you have a district size of three members, then you only need logically to win 25 percent of the vote plus one,” Diamond said. So to be elected a five seat district, candidates only need to pull one-sixth of the vote plus one. While this could help moderate Republicans win office in deeply blue states and vice versa, it may also have the adverse effect of advancing radical candidates.
The method most effective at eliminating negative outcomes, Saari argues, is the largely untried Borda count. As with ranked-choice and proportional ranked-choice voting, voters list their candidates in order of preference. But instead of relying solely on plurality, the Borda method assigns value to candidates according to their places on each ballot.
The clearest parallel is to how colleges and universities calculate GPA. Candidates get 1 point for last place, 2 for next-to-last, and additional points in ascending order. The substantive outcome: voters select the least divisive and most widely-accepted candidate, allaying partisanship and infusing nuance into the picking process. “It captures the information from paired comparisons with less work and fewer things that can go wrong. It can be proved that it is the method that is the least likely to give you paradoxical outcomes,” said Saari, who worked on the mathematical proofs to back the method.
“I think our worst villain is the voting system, because it throws away valued information about who the voters want,” Saari added. “I strongly believe in a ranked voting system to try to recapture what it is that the voters want, but if you’re using plurality at each stage in a ranked-choice voting system you’re throwing away valuable information.”
Do you like this page?