Federal Judge Upholds Maine's Ranked-Choice Voting

A Republican representative lost his seat in the new instant runoff system, so he sued.


Ranked voting
Peter Mautsch / Maranso Gmbh /

A federal judge has pushed away a legal challenge to Maine's new ranked-choice voting system.

In the November midterms, incumbent Republican Rep. Bruce Poliquin won the first round of votes against Democratic challenger Jared Golden. But he did not get more than 50 percent of the votes, since there were two other independent candidates in the race. Under Maine's new election rules, put into place by the voters, a candidate for Congress must get a majority of the votes, not just a plurality.

To resolve this problem, Maine voters are asked to rank the candidates in order of preference. If no candidate gets a majority vote, the candidate who received the least votes is eliminated from the race. The ballots are tallied again, but for those who voted for the eliminated candidate, their second choice is counted as their vote instead. And so it goes until a candidate gets the majority vote. It's essentially an instant runoff system.

In Maine, more of those independent voters selected Golden than Poliquin as their second choice, and that pushed Golden ahead to narrowly win with 50.6 percent of the vote. Poliquin sued to try to stop the vote count, challenging the constitutionality of ranked-choice voting.

U.S. District Court Judge Lance Walker, appointed by President Donald Trump, roundly rejected Poliquin's suit in a 30-page ruling. Poliquin and the other plaintiffs—a couple of Poliquin voters—had argued that Article 1, Section 2 of the U.S. Constitution calls for plurality voting. But it does not. The text says members of the House of Representatives will be elected by popular vote, but it doesn't state whether the winner should be determined by a plurality or majority.

Walker ruled that this is intentional. States have historically been free to decide the threshold for victory, and—as Walker noted—some have separate run-off elections to determine a majority vote winner.

Walker also rejected an argument by the Poliquin-voting plaintiffs that the system violated the Equal Protection Clause of the 14th Amendment by not treating their votes with the same "weight" as others. The plaintiffs only voted for Poliquin and declined to rank the other candidates, which is permitted. They argued that this meant that their votes had less "weight" then those who ranked their candidates. This made little sense and the judge rejected it. They chose not to rank the other candidates, but the option was presented. And during each round, each person's vote counted only once.

Now, there is an issue where a person's ballot is "exhausted" if he or she voted only for a candidate who is eliminated from subsequent rounfs. In that case, that person's vote is arguably wasted. But that's no different from what happens in conventional election system—and it isn't what happened to Poliquin voters anyway. Their votes were equally counted in the second round of votes.

So it looks like Maine's ranked-choice voting system is here to stay. This shouldn't be a surprise, as a handful of cities have already been using such systems. But Maine is the first to use it for broader elections.

Maine voters actually wanted to use it for more than just congressional elections. They authorized its use for statewide races and state lawmaker races too. But Maine's constitution, unlike the United States Constitution, specifically orders that election winners for those races be determined by a plurality vote. In order to comply with the voter-approved ballot initiative, lawmakers need to amend the state's Constitution. Right now state Republicans have been resistant. We'll see if this ruling gets them to accept the public's will.

Read Walker's order here.

NEXT: California Regulators Want to Tax Texts You Sent 5 Years Ago

Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

  1. Can you vote for the same person as all choices?
    Say there are multiple candidates from “mainstream” parties, but only one Libertarian, can you put the Libertarian down as all your preferences? Sort of a modified ‘none of the above’?

    1. Looking at the sample ballot, you can only vote once for a given candidate and do not need to vote for any of them. So in a four person race you could just check the box for the Libertarian. Alternatively, you could vote Libertarian first, somebody else second and leave the rest blank, or vote for all 4 candidates in the order of your preference.

    2. It might be possible structurally, but it’s completely irrelevant, given how ballots are tabulated.

      In ranked choice voting, you declare your order of preference among the candidates. Then votes are tabulated in a series of “instant runoffs” known as rounds. In each round, your vote is assigned to your most preferred candidate that is still in the running.

      So, if you make the Libertarian your first choice, and she is still on the ballot in round 2, your vote will ALWAYS go to the Libertarian. If she is eliminated in round 2, your vote will be discarded, because NONE of your preferred candidates are still in the running. So, putting your preferred candidate in first place is sufficient. Putting them in second place is irrelevant.

    3. That doesn’t make any sense.

    4. According to the instructions on the ballot:

      Fill in no more than one oval for each candidate or column.

      So no, marking the same candidate across for all choices would likely get the ballot tossed.

      One of the downfalls of this system is that theoretically a candidate ranked last in the first round could have enough votes to break the 50% threshold by the “third” round (largely the second choice of the second candidate eliminated) but having already been eliminated can’t possibly win.
      Simple example: Given candidates A, B, C, & D in the first round get 31%, 26%, 23%, & 20% respectively then D is eliminated. The second choices for 20% of D voters are split 8% to B and 12% to C making the tally 31% A, 34% B, & 35% C thus A is eliminated. Now suppose that all 31% of A voters put D as their second choice. Had D not been eliminated in the first round D would now have a 51% majority but because D was already eliminated it shifts down to the third choice of A voters.

      1. While theoretically true, there is no alternative runoff system that would succeed in getting D elected.

        1. Perhaps, but I’m sure there’s a mathematician somewhere who could construct something obscure enough where D could win. Imagine further that the second choice of all B & C voters was also D such that by adding all the first and second choices for a total of 200%, D would get 100% while A, B, & C would get 31%, 34%, & 35% respectively. But yes, it is more than a bit contrived at that point.

        2. You are absolutely correct. And, Instant Runoff Voting (IRV) used in Maine is unconstitutional – the Plaintiffs used bad arguments. IRV discards votes, which is the winning argument. The discard algorithm violates 14th Amendment Section 1 equal protection and Section 2 voting rights. That’s because ranking information is the same thing as a vote. Condorcet Voting – a round-robin ranked choice system – uses ALL votes/ranks and matches candidates head-to-head. It is quite possible for “D” to be the “Condorcet Winner” or round-robin winner. I’ve been advocating voting reform (with zero success) for over 20 years because, without it, forget about viable 3rd parties.

          I’d recommend looking at what actually happened in the Burlington Vermont mayoral election that used IRV in 2009 for some of the doleful effects of IRV’s discard algorithm.

          1. The vote is not discarded – it was used in the first round. The voter decided to not participate in the subsequent rounds.

  2. Can’t guess how many voters don’t understands how it works.

    1. I think something like 99% voted for the DeRps anyway – so either that’s the number that don’t understand the point of this voting system – or the new system serves no purpose at all.

      1. The other choice is that some of them understand the point, and don’t think about it the way you want them to.

        1. No. If that 99% is actually happy with the DeRps, then this new voting method serves no purpose.

      2. In the election in question, there were ~22k out of ~200k votes for independent candidates, so more than 10%.

        Prior to RCV, it’s hard to say how many people are really independent, since many Independents would vote “strategically” to avoid the spoiler effect. With RCV, we can finally see people’s TRUE preferences (in the first round.) A good showing may enable Independents to build a constituency.

        RCV is very good for the viability of Independents in the long run!

        1. In NM, they somehow manage to tabulate who voted for whom, versus registered party. I guess it’s still anonymous.

          But in the 2008 Pres race, for example, “Independents” voted like, 51% D and 49% R, and .002% actually for a third party. It was pretty sad.

          1. If RCV were available on a national level, those numbers might change. But otherwise, most people know you gotta vote for a potential winner or you are just “throwing your vote away”.

      3. Actually, the system did serve it’s purpose, eliminate the “split vote” or “spoiler” problem in plurality races when there are more than two candidates.

        In the first round of the Maine 2nd District race, Polquin was ahead, but without a majority. When the two independent candidates were eliminated, and voters who had them as their first choice had their second choices tabulated, the Democrat had a majority, and won.

  3. Good on them. Now they just need to change their state constitution so they can use it on all elections as intended.

    1. The legislature has already referred it to the Committee of Not Until Hell Freezes Over.

  4. Does this new system affect ballot access? And if so how?

    1. It has an indirect, positive impact on access.

      RCV simulates a series of runoffs. If runoffs were held on a seperate days, that would burden the voter to find time to vote in each runoff. RCV makes it easy for a voter to participate in every “runoff” with just one trip to the polls. So in that way, it improves access and participation.

  5. This new system of voting is nonsense. We should be able to “send a message” to either — or both — of the two, major parties, by refusing (with our 3rd party vote) to give either of them a majority of the votes. This new system negates that because it results in always giving someone a 50% plus total.

    1. A comment from Australia on your “new system of voting”. Ranked choice or preferential voting as it is known in Australia was first used at our federal level 100 years ago on 14 December 1918. It had been used by some states even before Federation.

      Maybe Americans are just a little bit slower in picking up in what is happening in the rest of the world.

      1. I lived in Tasmania from about one year old to fifteen from the late forties till the early sixties

        Tasmania used Hare-Clark or the Single Tranferable Vote system of Proportional Representation for elections for its lower house while for the Federal lower house (House of Representatives) RCV was used

        There seem to be many reasons that these systems have never caught on in the USA and Canada (which could really use it) but the one that I seem to hear most often is that voters won’t understand it

        Which is sort of strange since I’ve never really felt that Australian’s were exceptionally smart compared to Americans or Canadians

        1. Also, note that RCV and STV are not really all that common in the “rest of the world”

          Canada and the UK both use “first past the post” while many Europian countries use “party list” proportional representation which is pretty much the system that gave PR its bad name

        2. Maybe Americans are just a little bit slower in picking up in what is happening in the rest of the world.

          They copied the secret ballot from us and it was originally called by them the Australian ballot. That shows that they are capable of adopting good ideas but why the American political system ? like the British ? uses first past the post requires a great deal more explanation.

    2. One property of a decent voting system is it comes to an answer. RCV does this by stimulating runoffs. That’s not a bad thing.

      You can still “send a message” by voting independent in the first round, forcing the election to later rounds. And because RCV eliminated the spoiler effect, we should see more people “sending a message”. When enough of them do so, the power will shift.

    3. Except in the few states that require a run-off election, the major parties don’t even notice that there wasn’t a majority. For example, see all the idiots screaming that Hillary won and the electoral college “stole the election”. They don’t notice that she did not have a majority, and don’t know that the Constitution not only mandated the EC, but also required a majority of the EC votes to win, with the runoff in Congress. If the popular vote counted and the rules were otherwise unchanged, Trump would still have won, unless Hillary managed to publicly buy several Republican Congressmen and Senators. (IIRC, two Presidential elections were decided by Congress, John Adams and Hayes, and both times they were won by what many considered a corrupt bargain.)

      With an instant runoff, the candidates notice that they weren’t the first choice of half the voters. When whining about this system gets Poliquin nowhere, the next Republican candidate is going to think about that ten percent of the voters that didn’t pick him even as their second choice – and becoming acceptable to those who won’t vote for the major party candidates as a first choice is going to become the way to win close elections.

  6. these election rules being imposed now are totally stupid. if you win an election over all your opposition, whether by one vote or 51% you STILL win. It does not matter how many candidates or parties were represented, YOU STILL WON!!! The only place this is different is the electoral college vote for president. And that should NOT be winner take all. It should be based on the votes each candidate received in each state. If each candidate gets 40% vs 60% then he gets 40% or 60% of the states electoral votes. Had that been in place in 2008 obama would have lost by the largest percentage of the electoral vote in history even give the huge majority in 4 of the most populous states. do the math, its time consuming but simple adding and subtracting. even democrats should be able to do it. Well maybe not Kalifornexicos.

  7. Allow me to explain the ACTUAL constitutional issue that Plaintiffs failed to make. The essence of our democratic voting system is (1) majority decides and (2) every vote must be counted. Maine’s IRV system fails on both counts. Taking (2) first, with a ranked ballot and the IRV algorithm, 1st place ranks/votes are initially counted. If no candidate has a majority, the candidate with the lowest number of 1st place ranks/votes is eliminated and ALL of the lowest candidates rankings/votes on other ballots are discarded. (At this point, Section 2 of the 14th Amendment has been violated because voting rights are being infringed.) The algorithm now decrees that every person’s ballot that had the eliminated candidate ranked first has their 2nd ranking/vote promoted to first and added in to the remaining candidates. This process of discard and rank promotion continues until someone has a “majority”. Clearly a ranking is a vote. I’ve had people argue that a rank is somehow a vote and not a vote in order to avoid the constitutional objection I make. This is pure sophistry. As for (1), IRV’s “majority”, will often NOT be the true majority winner. In social choice theory, where there are 3 or more candidates, the true majority winner – the candidate who beats all others head-to-head – is known as the “Condorcet Winner”. The Condorcet Voting system, which discards no votes, and matches all candidates head-to-head, is the only system that will find the majority winner.

    1. “The essence of our democratic voting system is (1) majority decides”

      So according to you, most states aren’t holding elections properly because they don’t have runoffs when there isn’t a majority. Or perhaps you don’t understand the difference between “majority” and “plurality”?

      As for “every vote must be counted”, does the vote of a Republican in San Francisco count when there is a nearly 0% chance of any candidate he ever votes for winning? Did my vote count when every candidate on the ballot was an effing statist, and they only disagreed about which direction to expand the government and which rights to violate first? In IVR, every voter’s choice is counted. It might not be the first choice, but in first past the post voting, often the _majority_ voted against the “winner” – and a lot of the votes for the “winner” came from voters who considered him far from their first choice, because third party votes are “throwing away your vote”.

      In this election in Maine, over 10% of the voters did not pick either the D or the R as their first choice. I don’t know about Maine politics or this particular race, but that’s two or three times the votes for all third parties in most election. It allowed these disaffected voters a way to show a preference other than D or R without “throwing away their votes”, and puts the major parties on notice that they need to pick up these votes if they want to win.

  8. One other point. Without democratic voting reform, all discussion of a viable Libertarian Party is moot. I assume every libertarian posting here wants to be able for vote for Libertarian candidates without it being a “wasted vote”. I quit the Libertarian Party over its failure to advance voting reform to a major issue. Initially, I pushed Approval Voting (AV) because it’s simple and would be cheap to implement. New optical scanning systems have made ranked ballot voting cheap for states to implement and the Condorcet system is objectively the best system available. I see IRV as a potentially dangerous distraction for libertarians because it WILL NOT help 3rd parties. The proportional representation variant of IRV, the Single Transferable Vote (STV) has been used in Australia for over 100 years. I may help to create some tiny 3rd parties but that’s about it. Also, because of IRV’s “preference truncation” problem, Australia requires ALL candidates to be ranked, leading to “donkey voting”. It’s a mess and should never be adopted anywhere. I could also talk about the non-monotonicity problem of IRV (thanks to discard) but the mere fact that IRV objectively doesn’t make the environment significantly more favorable for 3rd parties should be enough for libertarians to reject it.

Please to post comments

Comments are closed.