Here are a few of the questions we hear most often about Ranked Choice Voting:
- Is there legislation in the State House to enact RCV?
- For which elected offices could RCV be used?
- How much will it cost to implement RCV?
- Is RCV the same system Cambridge uses for its city elections?
- How resistant would RCV elections be to fraud and hacking?
- What about traditional runoff and "top two" voting systems?
- What about Approval, Score, or Condorcet Voting Methods?
Yes! In fact, there are three bills in the current session that would advance RCV in the commonwealth. Bill H.377, sponsored by State Representative Jay Kaufman, would enact RCV for all state offices, both primary and general elections. It has 15 co-sponsors. Companion bills H.2897/S.380, sponsored by State Representative Mike Connolly and State Senator Jamie Eldridge, respectively, would create a local option for cities and towns to use RCV for their municipal races by passing a law, without changing their charter as is currently required in most cases. The local option bill has 31 co-sponsors, a new record for RCV legislation in Massachusetts.
With very few exceptions, all elections in Massachusetts could use RCV. We are in favor of using RCV for all elected offices in the Commonwealth, from local offices like mayor and city councilor, to state legislative seats, county offices like district attorney and sheriff, statewide seats like governor, and federal congressional and senate seats. Ideally, it would be used to choose which presidential candidate receives the state's electoral votes, as well.
The only existing elections to which RCV could not apply would be presidential primaries. In Massachusetts, each party allocates its convention delegates proportionally to the presidential primary contenders, so there's no need for RCV to find a winner within the state. Also, should the National Popular Vote interstate compact take effect, using RCV to choose our individual state's electors would no longer make sense, as the president would be chosen by the popular vote in all 50 states and the District of Columbia. However, if we were to have a national popular vote for president, we favor that vote use RCV nationwide.
Many voting machines in Massachusetts will need to be replaced to support RCV, but 1) most of our voting machines are antiquated and in dire need of upgrade anyway; and 2) it will cost the state and taxpayer next to nothing to purchase new machines, thanks to available federal funding for voting system modernization.
Voting machines in Massachusetts are in rough shape. According to a report by the Brennan Center for Justice, we are one of only 14 states using machines that are more than 15 years old. Over 90% of our voting machines are no longer manufactured and fail to meet even the 2005 federal certification guidelines. In 2014, the bipartisan Presidential Commission on Election Administration (PCEA) called the state of voting machines in our state and country an “impending crisis.”
The good news is that money to upgrade them is ready and available by way of the Help America Vote Act (HAVA). According to the 2016 annual expenditure report from the U.S. Electoral Assistance Commission, Massachusetts has over $43.4 million in unused HAVA funds, more than any other state, plus an additional award of $8.3 million in 2018, bringing us to $51.7 million we can use for voting machines upgrades without taxpayer expense.
There are only two modern voting machines authorized for use in Massachusetts — the ImageCast Precinct by Dominion and the DS200 by ES&S — and they both support RCV. Upgrading our more than 2000 outdated voting machines to the DS200 would cost about $12 million, less than a quarter of our available HAVA funds. Leasing them would cost about $1.9 million per year. Upgrading to ImageCast Precinct machines would cost about $16 million.
We anticipate additional voter education costs of about $500,000 for the first election cycle, or about 7 cents per resident, and the education costs could possibly be covered by HAVA funds, as well. New machines would also enable more municipalities to switch to RCV, which could render a net savings to cities by eliminating the need for their costly preliminary elections.
There are two forms of Ranked Choice Voting: one for electing a single candidate, used for offices like governor, senator, or mayor; and another for electing multiple candidates at once, which Cambridge uses to elect its city council and school committee. Since nearly all elected offices in Massachusetts are single-seat, including all state and federal offices, our work has centered on the single-winner form.
Known in Cambridge as "proportional representation," multi-winner RCV offers the same voter experience as single-winner RCV: just rank your candidates in order of preference. As Cambridge voters have seen first hand, using RCV for multi-seat offices provides representation for minority voices while still guaranteeing that a majority of voters elect a majority of seats. Compared to cities of similar demographics, Cambridge boasts higher voter turnout and greater diversity in whom they elect. Under Cambridge’s RCV system, an astonishing 92% of voters see their first or second choice elected, while more than 95% see their first, second, or third choice elected.
Multi-winner RCV is used in in legislative elections around the world, and historically, has been used in many cities around the US, including several in Massachusetts. Today, with vote-counting computerized and urban populations more diverse than ever, US cities should consider multi-winner RCV to provide diverse representation to their city councils, school committees, and other governing bodies.
RCV elections in Massachusetts would be just as resistant to fraud and hacking as existing plurality elections. However, there are some additional safety measures Massachusetts should take to shore up the integrity of all our elections, regardless of voting method.
Massachusetts is doing a couple of key things right with respect to election integrity. First, every polling location uses paper ballots that are either optically-scanned or hand-counted, ensuring every election has a voter-verifiable paper audit trail. DREs (direct-recording electronic machines) are not authorized for use in the state. Second, none of our voting machines are connected to the internet. Neither of these two features would change with a switch to RCV.
Currently, Massachusetts conducts audits for elections that take place during presidential election years. Our audits compare the cast vote record to a manual count in a random subset of precincts, and they are entirely compatible and will continue with ranked choice voting. We agree with the election integrity community that such audits should be conducted after every election, whether or not the election uses ranked choice voting.
Lastly, it is common for U.S. jurisdictions using RCV to post their cast vote record online for public inspection and outside analysis, providing a degree of transparency to RCV elections above and beyond most plurality elections. We expect Massachusetts will do the same, should we adopt RCV.
In a traditional runoff system, if no candidate receives a majority of the vote in the election, a separate runoff election is held at a later date between the top two vote-getters. In Louisiana and elsewhere, a traditional runoff systems is used in conjunction with a non-partisan blanket primary, also known as a "jungle primary," in which all candidates from all parties compete in a single primary election. It is possible for two candidates from the same political party to advance from the jungle primary to the runoff. A variant on the jungle primary system is California's "top-two" system, in which the the two top vote-getters advance to the runoff even when a candidate receives an absolute majority in the primary.
Like ranked choice voting, traditional runoff systems ensure the winner has a majority of the vote. However, they attain that laudable goal in a crude fashion. Unlike RCV, which welcomes new voices and choices to our elections, traditional runoff systems yield majority winners by excluding candidates from the final, decisive election. Top-two systems are the worst offenders in this respect, because the final election is guaranteed to only have two voices in the debate. To make matters worse, those two voices may be from the same party, so the breadth and depth of the electoral debate may be especially limited. While traditional runoff systems are superior to plurality, for those interested in expanding the range of choice on the ballot and deepen our political debate, they fall short.
Traditional runoff systems do not eliminate as many spoilers as RCV does either. Consider a primary election between 6 candidates: 2 candidates from party A, each of whom get 20% of the vote, and 4 candidates from party B, each of whom get 15% of the vote. The two candidates from party A would advance to the final election, despite the fact that 60% of the voters chose a candidate from party B. RCV fixes this problem by simulating a series of runoffs, each of which eliminates a single candidate, but with only one trip to the polls.
With the exception of Cambridge, cities in Massachusetts use a top-two system for their municipal elections. If there are three or more candidates for Boston Mayor, for example, all three complete in "preliminary election" and the top two advance to the general. If our cities adopted RCV, they could eliminate these costly and time-consuming preliminary elections and find a majority winner with a single election.
Yes, there are a seemingly endless number of alternative voting systems in existence. Unfortunately, none of them is perfect. Indeed, for any system one can find a desirable property that the system does not satisfy. However unlike these other systems, ranked choice voting is tried, tested, and works well in practice. Approval, score, and Condorcet methods are not used anywhere in the world for governmental elections, so whatever benefits they may convey are unproven in the context of real, political elections.
On the theoretical front, approval and score voting fail a critical test that voting theorists call the majority criterion. This is the simple property that a candidate who is the first choice of an absolute majority of voters should always win. RCV, Condorcet methods, and even plurality voting satisfy this, but approval and score do not. Under approval or score, a candidate could be the first choice of 99% of voters and still lose in theory.
RCV is also one of the few methods that satisfy a property called later-no-harm, which we believe is necessary in the context of high-stakes, competitive elections. Namely, under RCV a vote for your second choice cannot hurt the chances your first choice will be elected; a vote for your third choice does not hurt the chances of your first or second choice; and so on. This is not true under approval, score, or Condorcet voting — a vote for a later choice works against an earlier choice. When a voting system violates later-no-harm, voters face pressure to bullet vote, meaning to cast a vote for only one's first choice. Case in point: approval voting was used to elect the Dartmouth Board of Trustees, but once the Trustees elections became mildly competitive, they abandoned approval due in part to concerns over bullet voting. The more voters bullet vote, the closer the system will resemble plurality in practice, and then we're back to square one in terms of improving our voting system.
We also consider the Condorcet criterion to be important. This is the property that the candidate that would win a head-to-head race against every other candidate should always win. While RCV, approval, and score voting may fail to elect the Condorcet candidate, in practice RCV has done so in virtually every single election. Due to strategic vulnerabilities of Condorcet methods, including later-no-harm, and the additional complexity Condorcet requires to resolve cycles, we strongly prefer RCV for political elections.
Furthermore, approval, score, and Condorcet were all designed to be used in single-winner elections only. Ranked choice voting works well for both single-winner and multi-winner elections. For elections that involve a mixture of single-winner and multi-winner races, we strongly prefer the simplicity of using a uniform voting method across the board.
These reasons help explain why ranked choice voting is the preferred voting method of the major electoral reform organizations around the world, including FairVote in the United States and the Electoral Reform Society in the UK. Again, no system is perfect, but if there's one thing all advocates of alternative voting systems agree on it's that our current plurality voting method is surely the worst.
Sightline's Guide to Methods for Electing an Executive Officer
Why Approval Voting is Unworkable in Contested Elections
The Troubling Record of Approval Voting at Dartmouth
New Lessons from Problems with Approval Voting in Practice