In this week’s midterm elections, Maine will break new ground by using ranked-choice voting to select who to send to Washington DC. But another kind of ranked-choice voting has also gotten lots of attention — the single transferable vote, a type of proportional representation. Jack Santucci gives an overview of how this form of ranked-choice voting works, its history, and what its wider adoption might mean for US politics. He argues that the system helps to build parties – not kill them – and ensures that those in the majority remain responsive to the needs of voters who feel unrepresented.
Not in about 70 years have we seen so much interest in proportional voting. Otherwise known as multi-winner ranked-choice, its basic idea is to shake up the two party system. A majority of votes would mean a majority of seats — unlike what we may get in this week’s election. Meanwhile, parties would be more diverse on the inside, instead of being so ideologically rigid.
Lost in the story is how all this works. The key is a constructive tension. Parties still will want to win their majorities. The difference is that, with multi-winner ranked-choice, they’ll have to cater to protest voters. I’ll explain what I mean based on recent research.
Multi-winner ranked-choice voting: what it is and how it works
In this type of voting, the voter ranks candidates in order of preference. Winning a seat requires a quota. If your vote doesn’t help some candidate get a quota — either because that candidate is hopeless or because they have enough votes already — your ballot flows to your next-ranked pick.
The size of the quota is the critical difference between two kinds of ranked-choice voting. In a one-seat district, it’s a majority, and that’s the system in San Francisco or Maine.
In an earlier time, though, many US cities had multi-seat districts — Cleveland, New York City, Cincinnati, Worcester (Mass), and 20 more. The larger the district, the lower the quota — typically about 10 percent of the vote.
Contrast ranked voting with the current system, where the most votes in a district determine the winner. That makes it rational for all sorts of people — money, the media, and party hacks — to winnow a candidate field. In the words of the late Kenneth Arrow, the “plurality system chokes off free entry.”
Ranked-choice voting doesn’t kill parties
When any politician can win a seat with 10 percent of the vote, you might expect parties to evaporate. Or you might expect legislatures to be filled with third parties. Not so.
In any democracy, there needs to be a legislative majority. That can be issue-by-issue, or the same group of people can just stick together. Humans are lazy, so they prefer the latter. This is why, even in countries with more than two parties, we tend to see the same parties working together. Sometimes they even announce such deals before an election begins.
“Big Top” by Gordon McKinlay is licensed under CC BY NC SA 2.0
One irony in this country’s history with multi-winner ranked voting was a tendency toward two-party competition. In Cincinnati, for example, the Republican Party squared off against the Charter Party in 16 elections, from 1925 to 1957. Charter was a coalition of Progressive Republicans and institutional Democrats. Other parties came and went — Progressive Democrats for two elections, Roosevelt Democrats in one of them. But of 144 winners over 32 years, only four weren’t from one of the two major parties — three in the late 1930s, at the height of the Great Depression.
And in Massachusetts cities, seven of which had multi-winner ranked systems, parties like the Cambridge Civic Association or Citizens’ Plan E Association were common fare. In 1932, in fact, the national group behind this reform also recommended forming “good government” parties.
The ranked-choice system itself gave people reasons to form political parties. Say you’ve decided to form a majority. The next step is to get it elected. Parties would do this by asking voters to use their rankings just for its candidates. That way, ballots wouldn’t flow “accidentally” to the opposing party or independents.
If parties can “game” ranked-choice voting, how is it better than what we have now? Weren’t they just choking off free entry?
Remember that parties want to win a majority. Sometimes, those coalition deals would close out key constituencies. Government might ignore important issues like civil-service desegregation or what ethnic neighborhood a new highway would run through. Property taxes were another concern.
Sometimes, issues unrelated to current policies would make their way into public debate. Examples included air pollution, a woman’s role in politics, and what to do about the Depression.
Independent candidates were common in periods when people didn’t feel that government was responsive. This is when the magic of ranked-choice kicked in.
The minority party in city council would see the ferment in election returns. To it, independent candidates would represent an untapped market for votes. In the extreme, independents might get a seat of their own, depriving some party of a majority. So, when the next election came around, the minority party would recruit independents, putting them onto a major-party slate. Getting back to a majority meant being responsive.
Responsiveness was visible in council business. One way to see this is to track party unity on city council votes. As newcomers challenged status quo policies, fights would ensue between them and the old guard. These fights turn up in data as party-unity dips, typically after some party or faction had returned to majority status. Then, as the party worked out its new disagreements, party-line voting would recover.
Over the long haul, the interplay of party strategy, independent challengers, and issue-based politics was constantly remaking parties. In the language of business, this is constructive tension.
Beyond extreme claims
I don’t want to oversell ranked-choice voting and the constructive tension it may lead to. I have no measure of whether such dynamics exist in just-pick-one voting systems. They sound like they should, but do they?
Multi-winner ranked voting does not blow up the two-party system. It does not open the third-party floodgates, nor lead automatically to policy change. On all of these metrics, it may disappoint.
But, over several elections, constructive tension with independents makes parties work harder to win their majorities.
- This article is based on the paper, “Evidence of a winning-cohesion tradeoff under multi-winner ranked-choice voting” in Electoral Studies.
Please read our comments policy before commenting.
Note: This article gives the views of the author, and not the position of USAPP– American Politics and Policy, nor of the London School of Economics.
Shortened URL for this post: http://bit.ly/2F7KEav
About the author
Jack Santucci – Drexel University
Jack Santucci is Assistant Teaching Professor at Drexel University, Adjunct Professor at James Madison University, and a 2017 graduate of Georgetown University’s doctoral program in Government. He is writing a book on the rise and fall of multi-winner ranked voting in American cities, 1893-1962.
For multi-winner RCV, the quota for N places to fill is ( TotalNumberOfVotes / (N + 1) ) + 1. For example, if there are three seats to fill and three candidates each have a little more than a quarter of the votes each, there are not enough votes left for anyone else to pass them.
Fairvote.org has a good example: https://www.fairvote.org/multi_winner_rcv_example My example uses the same example at 1/10 scale, but with the same results.
Above is the Droop quota. That site also gives the Hare quata ( TotalNumberOfVotes / N ) as an option.