Skip to Content

Support MinnPost

This coverage is made possible by a grant from The Joyce Foundation.

Confused about ranked-choice voting? Let’s sort it out

A sample ballot from the website FairVote Minnesota.
FairVote Minnesota
A sample ballot from the website FairVote Minnesota.

First in an occasional series on ranked-choice voting.

Are you ready for the first really big test of ranked-choice voting (RCV) in Minnesota?

Minneapolis and St. Paul have both moved into the vanguard of cities experimenting with ranked-choice voting, which in the past was often referred to as “instant-runoff voting” or IRV. (I liked IRV better because Irv was my dad’s name and because, for me, the idea that this method creates a series of instant runoffs until there are only two, and finally, one candidate remaining also seems like a good start to explaining how it works. But pretty much everyone seems to have agreed to call it “ranked-choice voting.”)

Being MinnPost readers, and therefore among the info-elite, you very likely have heard about this and probably have some idea how it works. Perhaps you have dipped a toe into the arguments for and against ranked-choice voting.

Today (in fact, with this very piece you are now reading) MinnPost kicks off an occasional series exploring RCV including its history, how it works, arguments for and against, how it is affecting the campaigns in the Minneapolis mayoral race, and some speculation on how it might affect the outcome.

In November, Minneapolis and St. Paul will both employ RCV. Because of the open mayoral seat in Minneapolis (the first no-incumbent mayoral race in Minnesota’s biggest city in 20 years) and because of the huge field of candidates in that race, the success or failure of RCV in Minneapolis in November will be the biggest trial yet for RCV in Minnesota.

If you don’t live and vote in either of the Minnesota’s big cities, you might still want to pay some attention to this new voting system because the effort to adopt RCV for statewide elections will continue, and the performance of RCV this November will surely be a talking point both for and against the expansion of the system.

But if you live in Minneapolis or St. Paul, if you plan to vote in November and if you want to maximize the impact that your support might have, you need to figure out not only whom you support for mayor or city council but who are your second and third choices. You don’t have to rank three choices, but you really should.

If you think — as many people to whom I have spoken in working on this issue seem to — that you can help your favorite candidate more by listing only him or her, and leaving the second and third choice options blank, you are wrong. But more on that later.

'First past the post'

In the democratic world, many systems of government and election procedures thrive. By far the most common system in the United States is the one sometimes called “first past the post,” in which each voter votes for one candidate and the candidate who gets a plurality wins, even if — in the case of a multiple candidate field — the winner gets less than a majority of the votes.

There are exceptions, other than RCV. For example, in Georgia they use a double runoff (but not instant runoff) system. If there is a contested party primary with more than two candidates, in either party, and no one gets a majority of the vote, they hold another election a couple of weeks later between the top two finishers, so each party gets a nominee who has won a majority of the intra-party vote (although, very often, the runoff has a much lower turnout than the first). Then, in the November general election, if there are more than two parties and no candidates gets a majority of the total vote on that round, they have another runoff to guarantee a majority winner. So it is possible — and not just possible, this happened in the 2008 U.S. Senate race — for the election to take four rounds. In the aforementioned 2008 Senate race, the final round had to be held in December.

Effective DemocracyIn Minnesota, our current governor, Mark Dayton, was elected in 2010 with less than 44 percent of the vote in a three-way race. In fact, no Minnesota governor has been elected with a majority vote since Arne Carlson in 1994, five gubernatorial cycles back. Minnesota is unusual in this regard (and many others) because over the past 20 years, it has featured three substantial (legally “major”) parties—all three of which have fielded credible candidates for governor in all of those races, which naturally makes it harder for any candidate to win an outright majority.

For some people, this is a problem that muddies up the plurality winner’s mandate to govern, at least symbolically, and it leaves supporters of the runner-up thinking that their candidate might have won if their candidate had been able to go one-on-one against the winner.

One of the arguments for ranked-choice voting — although the argument is not perfect — is that it greatly increases the likelihood that the winner of an election will have majority support and whatever additional democratic legitimacy comes with that fact. That is one of the chief arguments in favor of ranked-choice voting. There are others, and arguments against RCV (including the argument that RCV does not actually guarantee a majority winner) that will be covered in subsequent installments of this occasional series.

For today, to keep this short, let’s just settle for a quick overview of the history of RVC and of where it is has been used.

RCV history

The system that became ranked-choice voting was devised by Massachusetts-based architect William Robert Ware in 1871. RCV (although not in its current form) was first used in Australia in 1893 in local elections, and first used nationally, also in Australia, in 1918. RCV is still used there in elections for the lower house of the national legislature. In several other nations (including Britain) various forms of RCV are used by parties in choosing their nominees.  

In the United States, in addition to Minneapolis and St. Paul, RCV is used in municipal elections in two other big cities —  San Francisco and Oakland. It was adopted by referendum in Burlington, Vt., in 2005, used in a couple of mayoral elections, but then repealed (also by referendum) in 2010 after an outcome in which the winner did not receive a plurality on the first round. RCV is not currently used statewide in any U.S. state.

In both Minneapolis and St. Paul, the question of whether to adopt RCV was submitted to the electorate as a city charter amendment. Minneapolis said yes in 2006 with 65 percent of voters in favor. St. Paul said yes in 2009 by a much closer 52-47.

Let’s leave it at that until the next installment.

Effective Democracy is a year-long series of occasional reports supported by the Chicago-based Joyce Foundation, as part of a grant made to MinnPost and the Wisconsin Center for Investigative Journalism.

Get MinnPost's top stories in your inbox

Related Tags:

About the Author:

Comments (11)

Analyse chap 167 carefully

A careful reading of 167.20 definition has two threshold definitions - why, was it to provide wiggle room to lower the election bar. In 167.60 there is a line that allows a blank 1st choice ballot to be discarded and to defeat write-in votes. What if a voter intended none of the above and either left the choices blank or put in write-ins. In (c) of 167.60 for 'rounds' it seems threshold and not maximum potential threshold sets the 50% bar - is this the wiggle room? We can be assured the election official will have or are in thr process of writing a tabulating rule book implementing the 167 ordinance - such a book would cover handling case situations. The elections office does not seem transparent.

Ranked Choice Voting

Sir:

I appreciate your article on Ranked-Choice Voting. Please allow me to share observations that add to your story.

I also prefer IRV (instant runoff voting). The variant for multiple-winner races, MIRV (multiple instant runoff voting) gives us, as you observed, two men's names, IRV and MIRV.

An English gentleman named Thomas Hare developed MIRV, and thus IRV, in 1859, before the gentleman from Massachusetts in 1871.

Hopkins, MN had MIRV for its City Council elections from 1948 - 1960, when it repealed MIRV by City Charter referendum.

For those who consider ranked-choice voting to be confusing, consider the story of Atri, Italy in 1944. The U.S. Army occupied the town. The military mayor was on Major R. E. Garrigan, a Cincinnati native who was neither a lawyer nor an election clerk, nor one with any political science education except his own first-hand knowledge as a voter in Cincinnati's then-MIRV City Council/Mayoral elections. He did not know Italian; he relied on translators' assistance in his duties. He held the first post-Fascist election in Atri using MIRV. It was also the first time women voted in Atri. Of 1,472 votes cast, only 22 went down as spoiled for failing to follow ballot instructions, a spoilage rate under 1.5% Major Garrigan noted that MIRV empowered more citizens and enabled occupying forces to understand accurately the political sentiments of the townfolk.

Please see my comments in the following link for more.

http://www.minnpost.com/glean/2009/06/first-minneapolis-later-state-cour...

Respectfully,

Peter J. Nickitas

This should be

…an interesting experiment. I'm looking forward to it, and to future installments in this series. I'll be especially interested in how this style/type of voting might influence the positions and behavior of candidates, according to relatively knowledgeable observers. It'll be interesting to see what kinds of speculations turn out to be close to what actually happens, and which ones turn out to be way off the mark.

RCV Doesn't Necessarily Deliver a Majority

The major argument for RCV is that it will deliver a consensus majority. But, that might not happen this time around. With lots of candidates but only three choices, that means that a lot of ballots will be "exhausted" before someone gets a majority. So, when the final count is finished, the top vote-getter may only have 45% of the total vote. That's the same result as the way elections are decided now.

Once that happens, maybe people will understand that RCV sounded good but didn't deliver on its promise. It was a bad idea to start and it's still a bad idea.

RCV does deliver a majority winner

As Steve Titterud discussed below, your understanding of the number of valid votes cast in rounds after the first is wrong. Ballots without a viable candidate in second and later rounds have no vote to count (and thus are exhausted). In mathematical terms, the denominator of the fraction of total votes for each candidate is the total number of ballots with a candidate counted in that round.

RCV is a series of runoff elections, with voters having made a single visit to the polling place. The winner of a runoff must have a majority of the votes counted in that runoff. But the reasoning that says RCV doesn't deliver a majority must hold that in a runoff election the "winner" has to get 50%+1 of the votes in the (earlier) general election. That's not an election rule anywhere.

The problem in Minneapolis isn't RCV. It's the ballot, which the Council refused to change from just three choices to rank.

RCV

Kudos for acknolwedging that RCV does not guarantee a majority winner - an erroneous claim that RCV advocates used to frequently make. Now that the argument has morphed into "greatly increases the likelihood that the winner of an election will have majority support" I am curious to see what, if any, evidence there is to support that claim. RCV did not produce a majority winner in either the Oakland or San Francisco mayoral races, nor in the Burlington, Vermont race that led to the repeal.

Sounds good in theory but. . .

if it's so good then why have so many of the polities which have used it in the past dropped it?

Maybe I need to have RCV explained to me again. Does RCV make any sense at all with a two party election system? If you have more than three parties participating in an election, doesn't RCV make whoever wins a winner by a majority by the way RCV works? In other words, isn't the final result the total first and successive ballots cast for the ultimate winner? Making every winner a majority winner seems to do more than just "increase the chance" of a majority winner, it seems to me.

Making winners majority winners

In 2011 in San Francisco, 194,418 voters cast valid ballots. The winner, Ed Lee, was named as a first, second, or third choice on 84,457 ballots, or 43.4 percent. The second place candidate, John Avalos, was named as a first, second or third choice on 57,160 ballots. 52,424 ballots did not name either Lee or Avalos for any of the three choices.

The RCV reallocation process did not determine a winner until only Lee and Avalos were left and the votes of all the other candidates had been reallocated. If you count only the votes that went to Lee and Avalos, and discard the valid ballots that did not name either of them, then Lee was a "majority" winner.

http://en.wikipedia.org/wiki/San_Francisco_mayoral_election,_2011

So ultimately, it depends on how you define majority winners. If you define majority by receiving 50 percent plus one of all the valid ballots, RCV does not guarantee majority winners. If you are willing discard all the valid "exhausted" votes - votes that did not name the remaining candidates when the RCV process determines a winner - and tell those voters (in excess of 25 percent in SF) that their votes didn't count, then RCV does guarantee a "majority" winner.

Ultimately, the reason you want majority winners is to provide a mandate or "democratic legitimacy," as Eric puts it. But if you have to discard valid votes and reconfigure the definition of majority, you aren't going to achieve that anyway. The real trouble comes up where the candidate who gets the most first place votes doesn't win. That happened in Burlington, Vermont, and led directly to the repeal of RCV there, and also in Oakland, where there is an active repeal movement. What happens is that a candidate is announced as having the most first place votes on election night, and then several weeks later, another candidate is announced as the winner after the reallocation process occurs. That is supposed to produce legitimacy and a mandate?

",,,a winner by a majority by the way RCV works?"

By the rules of RCV, the total votes available are re-calculated after every round of tabulation, due to the influence of "exhausted ballots" and dropped candidates. This is the "by the way RCV works" in your question - i.e., its rules as they are applied, round by round.

Here is where a particular kind of confusion is created by the opponents... They refuse to accept this re-calculation of total votes available, and cling to the ORIGINAL vote total in their complaints that an RCV winner does not necessarily gather a majority total votes cast. They continue to view the Round 1 "total" as the only valid "total" throughout all successive rounds, regardless of the rules of the RCV system as a whole, which might be looking, in Round 6, at a Round 6 total, a different number. It drives the opponents crazy that the Round 6 total of votes available may be less than the Round 1 votes available, and cry foul.

So when answering the question whether a winner has achieved a "majority", they feel the RCV rules of re-calculation should be ignored. I don't know if this explanation clears up any of your confusion or not - I'm just taking a stab at clarification here.

There is a pretty thorough tabular presentation of the round-by-round totals of the San Francisco 2004 mayoral election at http://en.wikipedia.org/wiki/Instant-runoff_voting_in_the_United_States. If you walk through each round, you can see clearly how it works, and how it might well work here. Both proponents and opponents cite this election's data, but they see different things in it.

Those of us who are willing to give this new system an open-minded chance see the point of the opponents.

But the current system in Minneapolis has produced a set of public officials - mayor and city council - who worked and VOTED to invalidate the provisions of the city charter which gave the citizens a right to vote down a massive stadium handout to a billionaire. These public servants went along with invalidating the city charter for the EXPRESS REASON that they knew the citizens would vote the stadium scheme down in a referendum !! That's why they had to cut the taxpayers out of the picture.

And look at the mess this kind of government has gotten us into NOW, considering the NJ trial !!

A system which can produce these kind of public servants is absolutely broken, in my view.

I'm willing to try just about ANYTHING to see if it can put better public servants in office - public servants with more respect for the citizen.

You wondered whether RCV makes sense at all with a two party election system.

Here's the way it could make sense: disrupt the duopoly's tried and true levers of control - the mechanism that has given us the city government we have right now. It makes it nigh unto impossible to elect third party candidates. Give third parties a better opportunity to win office through changing the culture of our elections.

Yes, the opponents have a point. And it could be a little messy at first. But it's worth a try.

Thanks for the clarification

I get how this works. And I tend to agree with Mr. Titterud that if RCV could open a crack in the two party duopoly, that would be worth it to not have a clear majority winner. I'm thinking of how this might have worked in a Presidential election like 2000 where many people would have liked to vote for Ralph Nader but didn't want to throw the election to George W. Bush or the people who did and have been since reviled as having done so.

You've brought up a key point of RCV,...

...which is the "don't waste your vote" marketing angle of the two major parties. This is a rather corrupt argument candidates have advanced in many elections past, and it is due to the election system.

In RCV, the major party candidates cannot afford to ridicule the lesser candidates - because the ISSUES those lesser candidates raise might very well influence ranking by voters as they fill out their ballots.

Therefore, issues which have in the past been ignored or dismissed by the major parties might influence the various rounds of vote calculations and thereby come back to bite the major party candidate in the behind.

I'm also hoping that the public will no longer accept the two party "dog and pony" debate presentation, as though no one else matters. In RCV, it's conceivable that MANY candidates will "matter" - i.e., influence the outcome. And so, for example, when the League of Women Voters puts on one of its "pick one of our two major candidates" performances, it might come to look ridiculous.

We don't know if it's going to work, really. We DO know the current system doesn't work for US.

It works for Zygi Wilf, though.