Ranked Choice Voting: a new way to vote

Mandi Jo Hanneke


If you’ve followed state, national, or local politics at all during the last year, you may have heard about Ranked Choice Voting and a number of initiatives to implement it. Cambridge, MA uses it. Maine just passed it. Here’s a primer on what it is, why it’s important, and why the Charter Commission put it in the proposed Charter the way it did.

What is Ranked Choice Voting (also known as Instant Run-Off Voting and Single-Transferable Voting)?

Imagine there are 8 people running for the 3 at-large Councilor seats in the new charter. In our current system, called block voting, each voter picks three or fewer. The 3 candidates who have the most votes win, no matter how much support they garnered.

With Ranked Choice Voting, each voter ranks the candidates in order of preference. The votes are tallied, starting with every voter’s first preference. A candidate is elected once she receives ¼ +1 of the votes cast. 1 If 3 candidates don’t meet that threshold after the first count, then the candidate finishing last is eliminated and the votes for that candidate are transferred to the second-ranked candidate. In addition, once a candidate is elected, excess votes for that candidate are transferred to the next-lower-ranked candidate. This process continues until three candidates have received ¼ +1 of the votes cast.2

These two videos describe it much better than I ever could:

Ranked Choice Voting in Single Winner Elections Ranked Choice Voting in Multiple Winner Elections

Why is this important?

You’ve heard of the “spoiler effect” or “throwing your vote away” when there are 3 candidates for 1 seat, right? The thought is, you really support candidate C, but she doesn’t have a chance of winning, and if you actually vote for her, your least preferred candidate, B, might win because enough votes have been pulled away from Candidate A, your second choice. So, what do you do? Vote for C on principle, recognizing that B might win, or vote for A so that B doesn’t win, denying C a chance to show how many people actually support her?

With Ranked Choice Voting, the choice would be easy: Rank Candidate C first, Candidate A second, and Candidate B third (or not at all). That way, you vote your conscience, and ensure your vote isn’t “thrown away.”

But what are the other benefits, you ask? Well, with Ranked Choice Voting, minority candidates are more successful in elections in which there are multiple seats open. Further, those elected tend to better represent the views of the entire electorate (proportional representation) versus in a block voting system, in which voters who support the majority view can control the election of all seats, ensuring that no candidate with minority views gets elected (slate voting).

There’s also some evidence that the campaigns are more positive. After all, candidates are seeking not just your first vote, but your second and third vote too, (or 4th, 5th and 6th), so they don’t want to alienate entire groups of voters by negative campaigns.

There’s a feeling that the campaigns focus more on issues, too, for similar reasons.

This is a huge change from the current block voting system, in which a slate of candidates representing the majority’s views can win all the seats in an election, thereby blocking out the minority political views from seats at the table.

Why is Ranked Choice Voting only in the Charter Commission’s Transition Article?

A charter is meant to outline the broad strokes of a government, not the minutiae of running elections. Adopting Ranked Choice Voting requires setting forth very specific information, including how ties are broken, how excess votes are allocated down-ballot, etc. Such legislation is not meant to be in a Charter, because the laws might need changing as technology changes.

Since a Charter is not meant to be amended frequently (and it’s hard to do because of that), the implementation methods of Ranked Choice Voting are best suited for a bylaw, not a charter.

The Charter Commission unanimously supports the implementation of Ranked Choice Voting, and we made the language as strong as possible to have it implemented as soon as possible. But the specifics are best left to a bylaw and a committee that is tasked with specifically looking at the options and implementation methods of other cities, towns and states.

We’re excited to have Ranked Choice Voting in the Charter.

It will help increase the diversity of elected officials, keep campaigns focused the issues, keep candidates positive, and help ensure that councilors are elected in roughly the same proportion as the factions in town they represent.

It’s the fairest election system out there today. We look forward to implementing in Amherst in future elections.


1The threshold for election is 1 vote greater than the inverse of 1+the number of seats available (1/2 + 1 for 1 open seat; 1/3 +1 for 2 open seats; 1/6 + 1 for 5 open seats; 1/7 + 1 for 6 open seats; etc.).

2Cambridge, MA uses this system to elect their councilors and school committee member. The specifics of the Cambridge council elections in 2017 using this method are located here: http://www.cambridgema.gov/election2017/Council%20Round.htm. In addition, a good explanation of the Cambridge system and its benefits can be found here: https://fairvote.app.box.com/v/cambridge-ballot-image-report


Comments 16

  1. Yes, I agree! This piece is definitely the best part of the planned charter. By far. Especially the part where you write “[candidates] don’t want to alienate entire groups of voters by negative campaigns.”

    And, consequently, I am also glad that town meeting bashing has slowed down recently on this blog, because town meeting bashing does have that effect of alienating, of a large group of the most politically active town residents, you know.

    Ranked choice voting helps ensure that every voter is heard better, and the diversity of candidates get elected (given, of course, that a diversity of candidates exists to begin with.)

    True, advocating groups can still help name-recognition, thus pushing their candidate to the front easily, when the group of candidates are not widely known, but still, if we had this system on the federal level, the election process would be simpler and we would not have to deal with a quite peculiar entity in the highest office now. We will see if, and how it will work out on the local level.

    1. Gabor is right: we should always be deferential to “a large group of the most politically active town residents”, especially since the charter might not pass. Their feelings, at this delicate time, are very important. We should avoid “bashing”.

      On the other hand, if the charter does pass, Gabor, and every other resident in town, will assume a brand new exalted status, hitherto not known in Amherst: that of “constituent”. Voters with power: that will be interesting.

  2. Ah, I almost forgot. When the planning board, and now the ARA is all appointed by plan, and without check-and-balances as the charter is currently written, Ranked Choice Voting will not have as strong a chance to be effective.

    1. The Town Manager supported the switch from Manager-appointed to Council-appointed Planning Board, Gabor, because he wants planning to be supported by the public, chosen by their Representatives. Under the current system, one person appoints the PB, the Manager, with no input from the voters. But under the proposed charter, Precinct Representatives, who are directly accountable to the voters, will approve and appoint them.

      Our current Planning Board has over 100 years of combined experience among them, and the position requires expertise and experience. Not just anyone can collect signatures and run for Planning Board. Therefore, they are subject to the same vetting process as any other volunteer committee or board member in Amherst. And then approved by our Precinct Representatives, who are directly accountable the voters.

      The difference is: a Council Representative, voted on every two years, is more accountable than a Town Meeting Member, voted on every three years. Two years is more accountable than three, it is that simple. Because democracy is one person, one vote. Representative Town Meeting is the illusion of democracy. The “purest form of democracy” is an election by the voters, one person, one vote.

  3. Amherst, and the world, is very different now than it was in 1938 (when our current charter was written). I’m excited about ranked choice voting and many of the other thoughtful aspects of the proposed update to our local government. I believe it will empower the voters and give us thoughtful spending and planning, while keeping what works today. Thank you Nick and Mandi Jo for breaking it down for us.

  4. Thanks, Nick. A very clear and useful explanation of ranked choice voting. To me it appears to have no downside.

  5. It’s interesting to speculate about what last year’s state representative race would have been like if Ranked Choice Voting had been in effect. With six candidates for one seat, it would have given voters more options than just choosing one of them, and would have helped an unsuccessful candidate who was a lot of voters’ second or third choice.
    It’s difficult, however, to see how Ranked Choice Voting would apply in elections under the current system. There are usually too few candidates in a precinct to make it relevant in Town Meeting elections, and too little information about how the candidates differ. And elections for Select Board and School Committee have gotten less competitive than they used to be.

  6. I’m guessing that a change in the system will result in more people taking part.
    Another plus that wasn’t mentioned: I lived for eight years under Ranked Choice system in Ireland, and it created a place for all the math whizzards to shine on election night: there was a running tally from the time the polls closed, at polling stations, on the radio, in the pubs (especially in the pubs) as the return trickled in. Those election specialists kept up with the tally as it changed, building suspense in the process. One more reason to get involved in the election!

  7. The following information about Ranked Choice Voting was passed along by UMass Political Science Professor Ray La Raja:

    Election Day 2017 featured several high profile ranked choice voting elections in cities and towns. Altogether, more than 200,000 voters ranked their choices in Minneapolis, St. Paul, Cambridge, Mass. and Takoma Park, Md in cities with a total population of about 844,500. Voters had real, meaningful choices, and could honestly rank those choices without needed to consider strategically how others would vote.

    It is now clear: ranked choice voting can help foster a healthy political environment with competitive elections and high voter turnout.

    Prior to 2017, our research has conclusively shown that RCV improves turnout compared to two-round, primary/runoff systems by eliminating the steep drop in turnout between rounds of voting: one election, not two. However, the data was too limited on the impact of RCV on turnout in the general election alone. A thorough look at election data from across many elections showed that RCV did not impact turnout significantly. Meanwhile, one researcher applied a particularly flawed methodology to a much smaller sample of elections in order to find that RCV hurt turnout. We attempted to demonstrate the flaws of this research, but it nonetheless created a persistent myth that RCV somehow deterred voters from coming to the polls.

    Fortunately, the 2017 election has exposed that myth: voters came out to vote at remarkably high rates in every city with ranked choice voting, flying in the face of a decades-long pattern of turnout dropping in municipal-level elections across the nation.

    In Minneapolis, about 43 percent of registered voters voted in 2017, up from 33 percent in 2013, a very strong showing for a local, off-year election. As Minneapolis City Clerk, Casey Carl put it, “That’s a really good turnout. That’s on par with 2014 midterm elections. … People are turning out everywhere throughout the city.” The election featured a competitive race between the incumbent mayor, Betsy Hodges, and two serious competitors, Jacob Frey and Tom Hoch; it also resulted in the election of Minneapolis’s first transgender city councilmember. More voters cast ballots in the mayoral race than in any prior mayoral race in Minneapolis for at least the last 20 years.

    In St. Paul, turnout was highest in at least two decades. Prior to the election, an opponent of ranked choice voting openly mocked Ramsey County Elections Director Joe Mansky as an “eternal optimist” for expecting 58,000 voters to come out and vote, claiming that high turnout had been abandoned with St. Paul’s prior system, in which 59,235 voters participated in 2001 and 59,154 voted in 2005 before plunging to 34,411 in 2009 and 31,175 in 2013. The number of voters ranking their choices in 2017 in fact exceeds 61,600 — 39 percent of registered voters in the city (the number may increase as provisional and other ballots are counted).

    St. Paul’s first open seat mayoral election saw greater turnout than the contested mayoral elections under the prior system. It appears the election will be decided on the first count, with rising star Melvin Carter being ranked first on a majority of ballots to become St Paul’s first African American mayor. A veteran St. Paul reporter wrote yesterday about how substantive the campaign had been, with each of the 10 candidates bringing something important to the debate.

    In Cambridge, about 32 percent of registered voters voted in 2017, a sizeable increase from its 2015 elections, and affected in part by the number of student voters who are less likely to participate in local elections than national ones. With only six incumbents seeking re-election for the city council’s nine seats, the election demonstrated how RCV elections can accommodate significant competition in a positive, issue-oriented campaign. The odds are that once again, at least 95 percent of voters will have ranked a winner in their top three rankings despite the large field.

    Takoma Park has not published voter turnout statistics at press time, though city clerk Jessie Carpenter did report that in Ward 2, which was the only race with a three-candidate contest, turnout reached a record-high of 40 percent of registered voters. We will update this report with more information as it becomes available.

    Opponents of ranked choice voting sometimes attempt to make ranking candidates seem like a daunting task. It is not — it is empowering and literally as easy as 1-2-3 to cast a smart, effective ballot. Ranking choices is a natural way of understanding our preferences, in and out of politics, and Americans rank things in order all the time. And the system encourages candidates to reach out and engage with more voters to earn a ranking from voters, so both candidates and voters end up learning more about their choices, their government and their community..

    When it comes to voting and elections, we deserve to have choices, and we deserve better than the artificial limitation of only a single choice in our political world of complex overlapping ideas and coalitions. It should therefore be no surprise that when voters have the opportunity to rank their choices, they turn out and make themselves heard.

  8. Couldn’t the Charter Commission have recommended that we use rank choice voting for electing Select Board members and Town Meeting members as a Charter reform?

    1. Ranked Choice Voting would have little impact on the current political system because the number of candidates for Select Board and Town Meeting have been too low relative to the number of seats available.

      1. There have been a number of contested Town Meeting elections in recent years, e.g. last year in Precinct 9 where 24 people ran for 8 seats. So RCV could also be used for Town Meeting reps and for Select Board. And just as for a city council, it could help raise interest in participation both as a candidate and as a voter.

        So the answer to your question, Janet, is that the Charter Commission felt that the city council was a better form of government for Amherst. And because any kind of change like this requires voter approval, it simplified things to include RCV as part of the charter proposal rather than as a separate proposal.

        1. The increase in the number of candidates for Town Meeting in the 2017 election, especially in Precinct 9, was attributable mostly to outrage over Town Meeting’s rejection of $34 million of state money for new elementary schools. Since 2008, 70 percent of Town Meeting members have been elected from precincts in which there have been 10 or fewer candidates for eight three-year seats, which can hardly be called competitive. Half the precincts have had eight or fewer candidates for eight seats, meaning that these Town Meeting members were self-appointed. The average number of candidates per precinct over that period was 9.2.

          1. I think the increase in the number of candidates is also a response to our current nationwide political climate, and it may therefore be more long-lasting.

            In any case, this is not an argument against using ranked choice voting for town meeting, because whenever there are more than 8 candidates running in a precinct, even if it’s 9 or 10, RCV would be a fairer way to elect representatives than the current system of at-large block voting within each precinct.

            It is, however, an argument for reducing the number of elected representatives so that there are more competitive races more often — which may lead to more substantial discussion of town issues.

  9. So Ranked Choice Voting wouldn’t work in Northampton for their City Councilor races since they have had no or few competitive races, right? So with 13 Council seats up all at once every other year (not even considering the fact that all School Committee and Jones Library Trustees seats are up at the same time), you would need at enough candidates per seat to make Ranked Choice voting operable, right? I guess I am asking– how many candidates do you need for each Council seat to make Ranked Choice voting actually take effect? And why does the Charter winnow the number of candidates in November by forcing a primary (and thereby making it hard for people to fund a run) when there are a lot of early candidates?

    1. Post

      Janet–Yes, if the races are not contested, then having ranked choice voting likely will not change the outcome of an election. Similarly, if only 1 seat is up for election (Oliver Smith Will Elector, for example), and only two people are running, then using ranked choice voting will likely not change the outcome. But, when there are multiple seats up for election (3 at large councilors or 5 school committee members, for example), then having just 1 or 2 people running over the number of seats being elected could make a difference from a traditional block voting scenario, potentially resulting in a different candidate winning that 3rd or 5th seat.

      As for the charter, the transition article might be confusing you. The only preliminary election being held if the charter passes is the preliminary in 2018, per the transition article. Once that election is held, no other preliminaries will be held again. The reason the commission opted to do a preliminary for the very first council election is that the Commission expects that the first council election will be widely contested, and it was not possible to institute ranked choice voting in time for it. Therefore, by holding a preliminary, we avoid a situation where 20 people are running for 3 seats and the winners get just 15% of the vote. By narrowing it down to 6 candidates for 3 seats through a preliminary (which will only be held if there are more candidates than double the number of seats), we are giving the electorate more opportunity to learn about the issues and the candidates and have the winning candidates elected with a larger percentage of the vote. The goal is to have ranked choice voting operative by the regular town election in 2021.

Leave a Reply

Your email address will not be published. Required fields are marked *