There is an organized effort in Minnesota — and currently and particularly in Rochester — to spread the use of ranked choice voting. RCV, a form of Instant Runoff voting, is a system in which each voter can vote for more than one candidate per office, ranked by their personal preference. See the nearby sample ballot. The voter marks a ‘1’ or fills a #1 bubble related to the most preferred candidate, a ‘2’ beside the second-most preferred, and so forth. The number of allowed choices can vary. Minneapolis allows up to three choices, St. Paul allows up to six.
Tallying rules can vary but generally, RCV ballots are first counted with first-preference votes. If a candidate receives a majority of the votes, s/he wins. If there is no winner, the candidate receiving the least number of votes is disqualified and the ballots cast for that candidate are replaced with those voters’ next ranked choices. The process continues until a candidate receives a majority of the votes.
Chair Aaron Miller had an editorial on this subject published in the January 31, 2018 Post-Bulletin.
Generally, proponents claim the following advantages for RCV:
- No need for primary or runoff elections
- More civil tone in campaigns
- Allow voters to vote for their favorite candidate no matter the party
- Majority wins
We’ll address #1 and #4 in some detail, below. As you’ll see #4 is incorrect. It may be true that RCV promotes greater campaign civility: candidates may want to stay on voters’ good side in order to retain higher rankings even when not the #1 choice. But there is no guarantee of better tone just because of the system. As for #3, there’s nothing in the typical voting system that prevents people from voting for their favorite candidate.
We’ll look at cons as part of the following section and in the conclusion.
Fair Vote Minnesota (FVM)
Fair Vote Minnesota is an organization working to convince people that RCV is a great thing. How could anyone be against a fair vote? Following, are some of their claims, taken from their web site, and some thoughts about them.
FVM: Does RCV violate the “one person, one vote” principle?
Yes! FairVote says no. By RCV definition, that’s incorrect. Voters whose higher ranked candidates are eliminated get subsequent votes that other voters do not get. It’s true that in any given round of tallying, no voter has more than one vote in the round. But some voters get the opportunity to keep throwing different votes into the mix. Voters who are least representative of their communities get more votes counted than everyone else. On the other hand, some voters may have no vote in the later rounds (if they did not rank any of the remaining candidates). Both ways, RCV is not “one person, one vote”. FairVote notes that the Minnesota Supreme Court ruled that RCV complies with “one person, one vote”. Are judges infallible now?
FVM: Fixing Problems?
FairVote says RCV eliminates “wasted” votes. No votes are wasted in our current system; they’re all counted. On the contrary, RCV can cause votes to go to waste. When there are more candidates than voters are allowed to (or have chosen to) rank, voters who have not ranked any still-standing candidates in late tally rounds have no vote in the race. This isn’t theoretical. It happens. If a voter casts a vote in a race but doesn’t have a vote in the decisive round, how is that not a wasted vote?
FairVote says RCV gives voters more choice. RCV doesn’t provide more choice. It changes where the choices occur. Instead of using primaries or run-offs, it puts all candidates on a single ballot. Perhaps they’re assuming more people will run. A balloting system doesn’t prevent people from running.
FairVote says RCV solves the “spoiler” problem. We’re guessing they’re referring to a candidate (A) with philosophies close to another candidate (B) thereby making it harder for candidate B to win and easier for candidate C to win. If that’s their definition of spoiler we still don’t know what the perceived problem is. We’ll presume, again, the problem is the hurt put on major party candidates by spoilers. If so, this is a curious thing to consider a problem as they already praised RCV for providing more choice. In any case, there’s no way of knowing if RCV is better or worse regarding the “spoiler problem”. In a three-way race, if a spoiler does poorly but well enough that no other candidate gets 50% of the vote, the spoiler is out and the problem is solved. Otherwise, the spoiler may have still spoiled. After all is said and done, is it spoiling, or just the will of the voters?
FairVote says RCV increases voter participation. But RCV is more complicated than checking a single box per office. Voters need to be armed with more information about more candidates and properly reflect their preferences on a more complicated ballot. A study of Minneapolis RCV voting showed a significant rise in spoiled ballots after introducing RCV. Having your ballot thrown out isn’t increasing participation even if it felt like you participated. Plus, the study revealed that those whose ballots were thrown out tended to be those of less affluent people and people of color (based on the neighborhoods where the ballots were discarded). Is that something FairVote considers fair?
FairVote says RCV reduces the cost of running campaigns. Balderdash. Candidates run as hard as they can as long as they’re running. On the one hand, RCV pats itself on the back for avoiding primaries. But with primaries, some candidates stop running when they lose in the primary. No primary? Everyone’s running to the very end.
FairVote says RCV reduces the cost of elections. Saving primaries and run-offs saves money. But running RCV elections costs more than running current elections because the ballots are more complicated and requires most jurisdictions to get new voting machines. Not all races require primaries but in an RCV world, we’ll pay more for every election where it’s used.
FairVote says RCV ensures majority outcomes. It absolutely doesn’t. As mentioned above, voters who do not rank last-standing candidates have no vote in the final result. It’s perfectly plausible – and has happened in actual RCV elections – that victors win with tallies less than 50% of the number of voters and a total number of votes counted than the number of people who cast votes for the office. This is well understood. How can FairVote still claim to ensure a majority outcome?
FairVote claims RCV yields office-holders who more fully represent the views and desires of voters. Actually, results of RCV balloting can yield surprising results compared to expectations. Here’s an example with 100 voters:
Group 1 is 35 voters. In rank order, they vote for Bill , George, and Ross.
Group 2 is 31 voters. They vote for George, Bill, and Ross.
Group 3, the 34 remaining voters, vote for Ross, George, and Bill.
Tally round 1: No one gets a majority. George is eliminated with only 31 votes.
Tally round 2: George’s votes from group 2 now go to Bill so Bill wins 66 to 34.
What’s interesting about this result? Almost two-thirds of voters – groups 2 and 3 – preferred George to Bill. But according to the final tally, Bill won with two-thirds of the votes! It seems reasonable to expect that with Ross out of the picture, and George preferred by a super-majority of the voters among the two remaining candidates, that George should win.
FVM: Where it’s used
FairVote mentions some places where ranked choice voting is in use but doesn’t mention something else: places that have avoided it or tried it and got rid of it. Is that fair? Fort Collins, Colorado defeated it almost 2-1. In the UK it was defeated 2-1. Duluth decided against it 3-1. Burlington, Vermont tried it and removed it. FairVote looks like an accident looking for a new place to happen.
FVM: Does RCV disenfranchises communities of color and less affluent voters?
FairVote says no and touts a university poll as proving it. Well, how about studying actual voter results from the ballot box instead of college kids on street corners? Such studies have shown ballot spoilage significantly higher than with non-RCV ballots, particularly in precincts largely populated by less affluent voters and voters of color. FairVote goes to the trouble of saying that there is “no evidence to support the claim that minority voters felt [emphasis added] disenfranchised” by RCV. Hmmm. Maybe they just don’t know their ballots were thrown out. Is it OK to throw out votes when the voters don’t know it?
FVM: Will new equipment be needed to tabulate RCV elections?
FairVote never really answers the question. For most jurisdictions, including Rochester, the answer is yes. Not only that, there are questions about mixing RCV ballots with “normal” ballots. It’s thought the two cannot be used together in the same election without a lot of trouble. Usually, where RCV is used, those elections are held in odd-numbered years with “normal” ballot elections retained in even-numbered years.
FVM: What about other alternative voting systems, like approval voting?
FairVote says they prefer RCV because the League of Women Voters of Minnesota liked it based on a study started in 2003. FairVote doesn’t like approval voting because “a candidate can lose despite being the first choice of more than half of voters.” This is a curious statement. Approval voting doesn’t use ranked choices, just yes or no as options for each candidate for an office. It’s true that a candidate granted a “yes” vote from more than 50% of voters could lose an election but that would only be because another candidate got still more, right on up to 100%. Is it fair to elect a candidate marked “yes” on 100% of the ballots (or even just 52%) instead of a candidate with 51%? Yep. Hey, FairVote, nice play with the facts. Not.
FairVote is in business to sell RCV. Like a used car salesman in a horror movie. Only with better teeth and clothes. And a web site. Perhaps the FairVote folks believe in what they’re selling but that doesn’t mean they’re making any sense or seeing or telling the whole story.
To add some of the cons not mentioned in detail:
- Complicated ballot
- Complicated ballot counting
- Voters need to know more (than just their #1 candidate)
- Voter suppression (based on higher rate of ballot spoilage)
- Lack of transparency
(some jurisdictions have refused to share raw ballot data)
- Possible manipulations
- All votes have to be in before counting can begin
- Disputed ballots, late ballots, or vote integrity problems require tally restart
- Manual recounts extremely complicated
- Likely to require local RCV elections to be held in odd years only
- May increase overall election costs
For some of us, the breakage of “one vote per voter” is enough to discard RCV. But as you can see, there are plenty of other problems, too.
Ranked Choice Voting is a bad idea.
updated 15May18: fixed a typo in penultimate paragraph
updated 19Feb18: added One Person One Vote graphic and link to P-B editorial