In a 1996 “Simpsons” episode, evil aliens Kodos and Kang secretly replace and impersonate Bill Clinton and Bob Dole during that year’s election.
When their plot is revealed, an angry crowd of voters vows to vote third party. Kodos scoffs at their empty threat, saying, “What, and throw away your vote?” The voters realize they are beaten. Looking on, ‘90’s third-party candidate Ross Perot punches through his straw hat in frustration.
It’s the age-old “spoiler” problem: Voters who prefer third-party candidates won’t pull the lever for them, for fear that doing so will waste their vote on a candidate who has no chance of winning – or even worse, throw the election to their least-preferred candidate. The spoiler issue has special resonance this year, with large numbers of voters disliking both Donald Trump and Hillary Clinton.
This November, Maine may lead the way in resolving this issue, with a referendum for “ranked-choice voting” for state elections. Under RCV, also called “instant runoff voting,” voters indicate their first choice, second choice, third choice, etc. for all the candidates who are running.
If one candidate wins a majority of first-place votes, she is declared the winner. However, if no one candidate gets a majority of first-place votes, the candidate with the fewest first-place votes is eliminated, and his votes are redistributed to the other candidates based on who those voters chose as their second choice. This process continues until one candidate has a majority of votes and is declared the winner.
I’ve studied and written about RCV systems for decades, first as a lawyer for the U.S. Justice Department then as a law professor teaching election law. In 2008, I led a successful referendum effort to pass RCV for Memphis City Council elections. Implementation in Memphis has been delayed because of outdated voting equipment and lack of political will, but I still believe the system has much to commend it.
A proven system
RCV is used currently in municipal elections in 11 U.S. cities, including San Francisco, Minneapolis and Oakland. Five states use it for military and overseas ballots. It’s used at the national level in a number of countries, including Ireland and Australia. The venerable Robert’s Rules of Order recommends it for some situations. Even the Oscars use it.
2016 might be a banner year for RCV. On the November ballot in Maine is a referendum which would institute RCV for state elected offices. The measure was largely inspired by the success of controversial Maine Republican Governor Paul LePage, who has described himself as “Donald Trump before Donald Trump became popular.”
LePage has been elected twice with less than a majority of the vote – the first time with only 37 percent support. In both cases, the presence of an independent candidate on the ballot split the anti-LePage vote.
In fact, in the last four Maine gubernatorial elections, the winner did not have majority support. And there’s no way under our current plurality system to avoid situations where the winner ends up being someone who is the least-preferred candidate of a majority of voters.
The traditional response to that problem is to have a separate “runoff” election between the top two vote-getters some weeks after the initial election. These runoffs occur in many jurisdictions around the country whenever no one candidate receives a majority of all votes cast in the first election. Not only are these separate elections time-consuming and expensive, but experience shows that turnout drops precipitously in the second round, with only a tiny fraction of voters deciding the final outcome. For example, in Memphis, average turnout in city council races drops from about 38 percent to about 5 percent between the first and second rounds of voting.
It is because of cost, time and turnout that many states have been abandoning their runoff elections. North Carolina and Florida are the latest states to abandon runoff elections for just these reasons.
RCV solves this problem because only one election is needed. As a result, recent scholarship has shown, turnout rates in RCV elections exceed those in systems with runoffs.
The turnout advantage is intuitive even as compared to plurality elections without runoffs. By solving the “spoiler problem,” RCV provides opportunities to third-party candidates and other lesser-funded, lesser-known candidates. With the outcome less predictable, elections become more competitive, which has the natural result of boosting turnout.
California gets innovative
Another alternative to business as usual is the “top two” system used in California, which abolishes individual party primaries and has all candidates of all parties run together in one primary election, with the top two vote-getters appearing in the general election. While admirable, the “top two” system still has a spoiler problem, since voters can’t rank, and often results in just two Democrats or two Republicans as the final-round choice. Indeed, some in deep-blue California have complained that it tends toward one-party rule.
RCV prevents this outcome, ensuring that the winner has majority support. It provides more opportunity for third parties, and is more reflective of the popular will.
Rank choice voting also decreases the effectiveness of negative campaigning. In an RCV election, you want to get your base to vote you as #1, but you also want to appeal to the base of like-minded candidates, hoping you can become their #2 choice. For that reason, it makes less sense to wage a scorched-earth, trash-your-opponent campaign.
While used primarily at the local level, there’s no reason RCV couldn’t be used at the presidential level too. As the Supreme Court said in Bush v. Gore, under the Constitution, state legislatures have full power to decide how they will allocate their Electoral College votes. Variations aren’t unknown. For example, Maine allots electors by congressional district for fear that, in a three-way race, one candidate could get all the electors with just 34 percent of the vote.
All those frustrated by their choices this election season, and those interested in fair, representative outcomes, should remember that the “spoiler problem” is an eminently fixable problem.
Kodos and Kang may seem scary, but they’re not invincible.