With the 2020 Election well under way–as voters around the country stand on long lines to vote early in person and place completed paper ballots into drop boxes and mail slots — eyes are on Maine, which is poised to become the first state to use ranked choice voting in electing the president.
Maine already broke new ground in 2018, when it became the first state to elect its Congressional representatives and a U.S. Senator–Angus King, an Independent–with ranked choice voting aka RCV, which even altered the outcome in one race. Democrat Jared Golden initially trailed Republican incumbent Bruce Poliquin by about 2,000 votes in District 2, but ultimately beat him by about 3,000, after picking up roughly 90% of the second-choice votes of the two independent candidates in the race. Aided by RCV, Golden became the first challenger to oust an incumbent in that District in a century.
For this year’s presidential race, Maine’s four Electoral College votes are at stake. That is not many, and it could be as few as 3 because Maine is one of two states where those votes can potentially be split between two candidates. Nevertheless, Maine is considered a swing state and the use of ranked choice voting could determine who gets its 3 or 4 Electoral College votes and conceivably determine the outcome.
Not only that. This election stands a real chance of flipping the Senate from Republican to Democratic control and RCV is even more likely to make the difference there. One of the Republican seats most in play is that of Senator Susan Collins of Maine, whose failure in 2018 to vote against the confirmation of Brett Kavanaugh for the Supreme Court exposed the hollowness of her claim to be a pro-choice moderate. Running against Collins are Democrat Sara Gideon and two independents, Lisa Savage a former Green, and Max Linn, who has run in the past as a Republican. Recent polls show Gideon with a slight lead over Collins, close enough that reallocated votes could make the difference. As of Oct. 28, one poll had Savage at 4.7% and Linn at 1.7%. With RCV, those Savage votes are likely to go to Gideon, bolstering her advantage over Collins.
What is RCV?
RCV is a democracy-enhancing approach to the way we conduct elections. It enables the will of the voters to be expressed to a fuller extent than under the standard approach — which the Brits call “first past the post”– where the person who garners the largest number of votes in a multi-candidate field can win with less than half the votes and even if opposed by a majority of voters whose votes are split among multiple candidates.
For something that brings such a significant benefit, RCV works fairly simply. Instead of selecting just a single candidate who is their top choice, voters have the option of also indicating their second choice, their third, and so on, depending on how many candidates are in the race and what the particular RCV law allows.
If one candidate reaches the 50% threshold outright, then they win, just as they would in the absence of RCV. It is when no candidate achieves a majority off the bat that RCV kicks in. The candidate with the least number of votes is eliminated and their votes are reallocated to one or more other candidates who were the second choices of those whose top choice they were. If necessary, that reallocation process is repeated until one candidate surpasses the 50% threshold and is declared the winner. That candidate will be the one who has the support of the broadest majority of voters even if they did not have the most first-choice votes.
That is far better than a traditional election in which the candidate who gets the most votes, say maybe one-third, can win because they have more than anyone else, even if that candidate might be the last choice of a majority of voters. That is how Maine ended up with Governor Paul LePage from 2011 to 2019. He first won in 2010, in a five-candidate field, with a plurality of 37.6%. He was reelected by a higher plurality of 48% in 2014 when he had only two opponents but he was extremely unpopular and in his last year in office, had disapproval rate of 53%, 12 points higher than his approval rating.
Not surprisingly, RCV is sometimes referred to as instant runoff voting, since it works like a runoff election but far more quickly and cheaply.
Perhaps the biggest advantage of RCV is that, for those whose preferred candidate has no chance of winning, their votes are not wasted. They get to vote for who they want but still get to have a say on who actually does win. No one has to feel that their vote was wasted, or didn’t matter, even if their first choice loses.
Which brings us to its other major (and related) advantage: avoiding lesser-evil-ism, the scenario in which a voter –whose preferred candidate cannot possibly win — reluctantly casts their ballot for a candidate who is not their top choice but is more acceptable than a third candidate who might win but is unacceptable. In contrast, the current top-vote-getter-wins system forces the voter to prioritize keeping the bad candidate out over voting for the one they actually want. With RCV, they can do both.
The 2018 U.S. Senate race in New Jersey is a prime example of how this works. Democratic incumbent Robert Menendez was running for reelection against Republican Bob Hugin. Also on the ballot were two third-party candidates, Madelyn Hoffman of the Green Party and Murray Sabrin, a Libertarian.
Normally, a Democratic victory would have been in little doubt given NJ voter registration figures and past voting patterns. But Menendez was a tainted candidate, who had been tried on federal corruption charges the year before, accused of misusing his office to do favors for close friend and major donor, Salomon Melgen, a Florida doctor convicted in 2017 on 67 charges of health care fraud.
The case ended in a mistrial in November 2017 and in early 2018, prosecutors opted not to retry Menendez and drop the charges after the U.S. Supreme Court, in a separate case, narrowed the definition of public corruption. Menendez did not escape entirely unscathed; the Senate Select Committee on Ethics “severely admonished” him in April 2018. He was so tarnished that his sole opponent in the primary, Lisa McCormick, an unknown who did no campaigning, garnered almost 38% of the primary vote from Democrats desperate for an alternative.
Menendez ultimately trounced Hugin in the general election by a 12-point margin (54% to 42%). Despite his eventual margin of victory, however, some polls at some points showed Hugin not so far behind, as close as 2%. As a result, many voters –myself included—would rather have voted for a third-party or independent if doing so did not risk electing Hugin, a Trump supporter who hiked drug prices in his former role as a pharma executive and of course, an addition to a Republican Senate majority that put Mitch McConnell in charge of Supreme Court confirmations. So we held our noses and cast our ballots for Menendez. I never wishe for RCV more than then when it would have allowed so many of us to vote our hopes and values rather than our fears.
Another benefit of RCV – actually just another way of looking at no lesser-evil-ism — is that it prevents the spoiler effect: where a marginal candidate siphons off enough votes from one of the more popular ones to shift the outcome in favor of a candidate held in disfavor by a majority of the electorate. That is what happened in the 2000 presidential election, where Ralph Nader’s votes in Florida exceeded the margin of victory George W. Bush held over Al Gore. Similar accusations have been made that the votes cast for Jill Stein in the 2016 presidential race are what swung several swing states to Trump and garnered him his Electoral College victory though the picture is more complicated and less clear than in the 2000 election.
Experience with RCV indicates that it helps elect more minorities and women to public office. It does so by broadening the pool of potential candidates and making the process more inclusive. RCV brings new people with different ideas into politics by making it more possible for outsiders to win. Those who are new to politics and less widely known can more effectively build support where it is not an either-or choice against more established candidates but a system of tiered approval.
Third-party and independent candidates need no longer be seen as spoilers but as viable candidates whose appeal is based on the quality of their ideas and less defined by the long-shot nature of their campaigns or the lack of a political machine behind them.
In decreasing the structural advantages of mainstream and better known candidates, RCV boosts the odds in favor of challengers, who might lack the funding, support base and other advantages enjoyed by incumbents but whose ideas might appeal to a wide swath of voters. Even if those challengers do not win the election, the increased possibility that they might do so would serve as an incentive to keep incumbents on their toes and make them more responsive to the needs of constituents.
RCV for Primary Voting
RCV would be particularly advantageous for the drawn-out U.S. presidential primary process where a multiplicity of candidates at the start tends to get substantially whittled down by the time the last few states get to weigh in. The process was even more prolonged this year because of COVID-19-related delays, stretching the timeline from the Iowa caucus on Feb. 3, when there were a dozen Democrats competing for votes, until the last primary, Connecticut’s, on August 11 (postponed from the usual April date) which had just two—Joe Biden and Bernie Sanders. In fact, the field had shrunk to those two as early as the Wisconsin primary in April, though some other names still appeared on the ballot.
People in states that tend to have late primaries (like New Jersey!) often complain that they have no say in the primary because the nomination is a done deal by the time we get to vote. That is certainly true for New Jersey which typically holds it primaries in June, but pushed the date back to July this year. At that point, two names appeared on the ballot but Biden had clearly wrapped up the nomination.
Voters in the early primary states also have cause to complain when they vote for candidates who subsequently drop out. Supporters of Elizabeth Warren, Pete Buttigieg, Amy Klobuchar, Kamala Harris, et al. might have liked the chance to have a say in the final showdown between Biden and Bernie.
We could eliminate that problem by holding all the primaries on the same day. Another approach would be to use RCV to conduct them. That way, some voters might still be unhappy that their first-choice candidate dropped out of the race but their votes could still have mattered if they had marked Biden or Sanders as an alternate choice. That would not only have made the process more democratic but could have potentially swung the outcome.
Actually, RCV was used in this year’s primary by the Democratic parties in four states–Alaska, Hawaii, Kansas and Wyoming. FairVote, a nonpartisan organization that advocates for RCV, tracked how it went. In Alaska, for example, no candidate got more than 50% on the first tally but after the reallocation of votes, Biden prevailed over Sanders, 55-45. Nearly 11% of primary voters had indicated as their first choice a candidate who had already dropped out. Thanks to RCV, more than 92% of them had their ballot count for either Biden or Sanders.
Better Campaigns, Higher Turnout
RCV doesn’t just lead to more democratic outcomes. It also elevates and improves the way campaigns are conducted because it deters negative, scorched earth electioneering. A candidate is far less likely to run attack ads against an opponent if she wants that candidate’s supporters to mark her as their second or third choice.
That doesn’t mean there is no room for legitimate criticism of an opponent’s positions with RCV, just not for mudslinging and distortions, which can alienate voters entirely and keep them from voting at all. Thus, a decrease in negative campaigning because of RCV could have the added benefit of increasing voter turnout.
The same is true for the broader and more diverse array of candidates you are likely to get with RCV. Voters who see candidates who look more like them or come closer to sharing their views are more likely to show up to vote.
History of RCV
Invented in Europe in the 1850s for use in multi-winner elections, RCV was adapted for the single-winner kind by an MIT professor and began to make inroads in the U.S. during World War I and in the decades after. By the early 1940’s, about two dozen U.S. cities were using it for municipal elections but it was repealed in all of them except Cambridge Massachusetts, which continues to use it.
Use of RCV remained more constant in other countries, in such places as Australia, Malta and Ireland, and spread to other places, including New Zealand, Scotland and Fiji. In the last two decades it has started to make a comeback in the U.S., chiefly on a municipal level, starting with its adoption by San Francisco in 2002. Currently, it is used in about 20 U.S. cities including Cambridge, Oakland, Berkeley, St. Paul, Minneapolis, Santa Fe, New Mexico and Portland, Maine.
Utah in 2018 and Virginia in 2020 enacted legislation authorizing RCV in local elections. Five states—Alabama, Arkansas, Louisiana, Mississippi, and South Carolina—use a ranked-choice system for military and overseas voters.
In 2019, New York City became the largest political entity in the U.S. to adopt RCV, approving it by a nearly 3-to-1 margin. When it takes effect next year, NYC voters will be able to rank up to five choices in primary and special elections for mayor, city council, comptroller and public advocate. The City’s Independent Budget Office estimates it could save $20 million for each election cycle that might otherwise require a runoff election (required when no candidate reaches 40%).
From 2006 to 2013, North Carolina had a law that created an RCV pilot program used to fill certain judicial vacancies. It was repealed, however, as part of a bill that imposed a voter ID requirement and other voting restrictions that were struck down by a federal appeals court in 2016.
It is not just governments that have turned to RCV for fairer tallies. Dozens of colleges have done so, including Auburn University, CalTech, Carnegie Mellon, Cornell, Duke, GWU, Georgetown, Harvard, MIT, Northeastern, Oberlin, Rice, Stanford, Temple, Tufts, and six University of California campuses—Berkeley, Davis, Los Angeles, San Diego, Santa Barbara and Santa Cruz.
Other entities have also adopted it, including the American Association of University Women, American Mensa, National Organization for Women and associations for chemists, geologists, actuaries, philosophers, psychologists, bicyclists and law and medical students. Since 2009, it has been used in voting for the Academy Awards.
RCV on the Ballot
RCV is on the ballot this fall in a number of locations, including two states. Alaskans will decide whether to institute ranked-choice voting for general elections, including the vote for president. In Massachusetts, the ballot proposal would apply RCV to state offices and congressional races.
In 2020 alone, legislators in 20 plus states and Washington DC have introduced more than 65 bills to institute ranked-choice voting at various levels, in addition to the referenda in Alaska and Massachusetts.
Pending in the New Jersey Legislature, is A1200/S1820, which would establish RCV for races for governor, state legislature, U.S. Senate and House and for presidential primaries and general elections. Voters would get to rank at least six choices deep, or more if the voting machines (and the size of the field) allowed. The law would take effect immediately but RCV would not take place until a year or more after certification by the Secretary of State that all voting machines in the state have the capability to handle RCV. (Another reason to switch to hand-marked paper ballots.)
Introduced in January 2020, the bill has not been given a committee hearing in either house. That is also what happened with A5205, an identical bill in the prior legislative session (2018-2019).
On the federal level, several pending bills would further RCV: The For the People Act of 2019, H.R. 1, passed by the House of Representatives on March 8, 2019, would, among many other things, promote the purchase of voting systems capable of RCV. The Fair Representation Act, H.R. 4000, supports amending several laws including the Help America Vote Act and the Reapportionment Act of 1929 to mandate the conversion of all congressional districts from single- to multi-member districts elected by ranked-choice voting as well as the creation of state-level nonpartisan redistricting commissions for congressional redistricting. Both bills would need to be reintroduced when the new session begins in January 2021.
Previous legislation, the Voter Choice Act of 2005, H.R. 2690, sought to require the use of RCV for general elections for federal office. It was never voted on.
The Fight for RCV in Maine
It’s worth taking a look at the many obstacles that had to be overcome in Maine to enable RCV. There were multiple referenda, repeated legislative action and lawsuits in both state and federal court, leaving the question of whether RCV would be used in this fall’s presidential election unresolved until as late as October 1.
Particular language in Maine’s state constitution was part of the problem and not likely to be common elsewhere but the political battles and challenges are likely to be present in other states and could prove instructive.
In October 2015, a grassroots group called the Committee for Ranked Choice Voting delivered more than 70,000 signatures they had gathered in support of RCV to Maine’s Secretary of State, who certified the issue for a vote on November 8, 2016.
Maine Question No. 5 on the 2016 ballot, formally An Act to Establish Ranked-Choice Voting, asked voters: “Do you want to allow voters to rank their choices of candidates in elections for U.S. Senate, Congress, Governor, State Senate, and State Representative, and to have ballots counted at the state level in multiple rounds in which last-place candidates are eliminated until a candidate wins by majority?” (Note that the question did not authorize RCV in a presidential race.) It won approval by a vote of 52% in favor, 48% opposed.
Subsequently, Maine’s Attorney General questioned whether RCV was inconsistent with language in the state constitution saying candidates for governor and legislature would be seated if they won a plurality of votes. In contrast, RCV is predicated on determining which candidate has the support of the majority, rather than a mere plurality. As a result, the Maine Senate asked the state Supreme Court for its opinion on whether RCV was consistent with the constitution. In May 2017, the court issued an advisory opinion that found a conflict did exist and that RCV could not be used for the gubernatorial and state legislative races. The court found no issue, however, with RCV in primaries or federal elections.
Later that year, the Maine Legislature passed LD 1646, a law stating that RCV would not be used in 2018 or in 2020, and that the Ranked Choice Voting Act would be automatically repealed in its entirety unless the Maine Constitution was amended prior to December 1, 2021.
The Committee for RCV got back into action to try to save the law through another referendum petition in support of a people’s veto. In Maine, any state law can be blocked through that mechanism. If enough signatures are gathered, the law—in this case the one repealing RCV—is put on hold and the people get to vote on the issue at the next election. That is what happened: more than 80,000 people signed the petition, suspending the repeal and allowing RCV to be used in the 2018 primary (though not for governor or state legislature) and for the people to weigh in on the repeal at the same time.
Before that occurred, Maine’s Secretary of State went back to the Supreme Court for an advisory opinion on whether the stay of the repeal meant RCV had to be used for the June primaries and the court said that it did. Then the state Senate jumped into the lawsuit to argue against RCV, a position the court unanimously rejected.
A month before the 2018 primary, the Maine Republican party adopted a rule that its party nominations were to be decided by a plurality vote. That same day, May 4, it sued in federal court for a declaration that, in light of the new rule, the First Amendment required that Maine not apply RCV to the GOP primary. The court refused, holding Maine had a legitimate interest in applying RCV to the primary.
RCV was used for the first time in the 2018 primary for both state and federal office and the accompanying referendum in support of RCV passed with 54% in favor, meaning RCV would be used going forward for all primaries and for general elections for federal office. As discussed earlier, in the fall 2018 election, RCV determined the outcome of one of the state’s Congressional races, resulting in the victory of the Democrat Jared Golden.
Before the second round of counting, however, Republican Bruce Poliquin who had the most first-round votes but not a majority, asked a federal court to stop the process. The court refused and when the second round of counting showed that Poliquin had lost, he requested a recount, then called it off before it was done but continued with the lawsuit, appealing his loss in the lower court to a federal appellate court. After losing the appeal, he finally conceded the election.
In August 2019, legislation expanded RCV to the presidential election and primary but the Governor delayed action on the bill, LD 1083, so it would not go into effect until after the 2020 primary.
As Election Day approached, several lawsuits over RCV were playing out.
Hagopian v. Dunlap, was filed in federal court this past July by 3 Republicans and an Independent, who contended that RCV unconstitutionally disenfranchises voters whose ballots are eliminated before the final round of tabulation, including those who do not fully indicate their preferences. The court ruled against them on August 14.
Payne v. Secretary of State was filed in state court in April by the Committee for RCV who sought to block a People’s Veto effort aimed at preventing the use of RCV in presidential races by repealing LD 1083. The Committee argued that the veto petition was filed too late or otherwise improperly and the issue got kicked up to the Maine Supreme Court for a decision, which was issued on August 13. The Supreme Court found the petition was not late or improper and sent the case back to the lower court, which found that the number of valid signatures fell short of the required 63,000 plus.
Another state court lawsuit was not finally wrapped up until October 1. Jones v. Secretary of State was filed by Maine Republicans to block use of RCV in voting for president. After the Maine Supreme Court held in favor of RCV on Sept. 22, the Republicans asked that the decision be put on hold while they appealed to the U.S. Supreme Court. On October 1, Maine’s top court denied the request, saying that it was not likely plaintiffs would prevail in the appeal and also that ballots had been printed and some had even already been returned.