MPR/Humphrey Ranked Choice Poll

Minnesota Public Radio has worked with Larry Jacobs from the Humphrey Institue on a post-election poll asking voters (and non-voters) what they thought of ranked choice voting. I've had the opportunity to review the report that was written based on this survey.


(There are two things I should state for the record: first, I am a strong ranked choice voting advocate and serve on the board of FairVote Minnesota; second, Professor Jacobs applied for the contract to conduct the City's post-election poll, which was awarded instead to Saint Cloud State University.)


Here are some of the key take-aways:

- 56% of voters prefer ranked choice voting.
- More than 90% of voters understood how to effectively cast their ballots.
- 68% of voters think we should use ranked choice voting for Gubernatorial races.
- Ranked choice voting saw a significant bump in popularity (13%) among people who voted on November 3.
- More voters ranked a second choice than didn't (52% to 47% for Mayor)
- 10-18% of people state that ranked choice voting will make them more likely to vote, while only 5% state it will make them less likely.

But for some reason, the Humphrey report that lays out the details of the survey has chosen to focus on a different statistic: whether or not people who declined to vote this year - people who haven't used the system - prefer it or not. It found that 54% of these non-voters didn't like ranked choice voting, and would prefer to not participate using the old method. This was a main point that Professor Jacobs chose to make in the MPR story. I find this particular piece of information about as interesting as someone's bad review of a movie they haven't seen.

Fascinatingly, the report takes the overwhelmingly positive data collected by the survey and manages to spin it as "ranked choice voting falling short of expectations." This is clearly a case, in my opinion, of attempting to make the facts fit a preconceived narrative.

For instance, the survey cites that only 52% of voters in the Mayoral race ranked a second choice, while 47% selected only one candidate. He spins this as a rebuttal to the idea that a second ranking is helpful to voters.

This narrative might make sense if clear frontrunner Mayor Rybak hadn't cruised to reelection with 74% of the vote. To put that in context, it means that a large chunk of people who ranked RT first used a second ranking that they were fairly sure they didn't need.


I have a good personal example of this dynamic. In the Ward 2 Council race, unsurprisingly, I ranked Cam Gordon first. There were only two candidates on the ballot, and I was fairly sure that my first choice would win (as he did, with 84% of the vote). I chose not to use my second ranking on that particular race, not because I reject the idea of ranking second choices in general, but because I was confident that I did not need it. To prove this point, I used all three rankings for Board of Estimate and Park Board at-large.


Would more voters have ranked second and third-place choices in the hotly-contested Mayoral race between Rybak, McLaughlin and Hakeem in 2005? I suspect so. It will be very interesting to compare Council races this year where there was a clear frontrunner who received over 60% of the vote (2, 7, 11, etc) to more competitive races where the winner received 55% or less (1, 4, 5, 6, etc). My hypothesis is that voters in those more competitive races will have significantly higher rates of using second- and third-place rankings.


Here's another example: the survey concluded that 47% of voters thought that ranked choice voting would make their vote count for more, while 41% thought it wouldn't make a difference and only 12% thought it would make their vote count for less. The report's spin? That "many do not believe RCV makes their vote count for more."


That's one way of interpreting those data. Another way would be to say that more people believe that the method helps them than believe it doesn't make a difference. Another way would be to say that 88% of Minneapolis voters believe the system either helps or at least doesn't hurt them.


The most interesting information from the survey were the partisan crosstabs. There's a glaring difference between support for ranked choice voting among Democrats, Republicans and Independents. Republicans reject the method 67%-31%, while Democrats embrace it 61%-33%. Most interesting of all, Independents are the group most in favor of ranked choice voting by 63% to 30%. This is one of the best proofs of the idea that Independents aren't necessarily in "the middle" of the political spectrum that I've seen. It's also interesting to put these figures in context: perhaps Democrats would be less excited about the idea and Republicans would oppose it less if it was a Democratic Governor who had won twice with less than a majority of the vote.


One major caveat to all of this analysis, both the report's and mine: this survey had a large sampling error of 5.7% for voters and 6.9% for non-voters, based on a low sample size of only 504 people. I can't remember the last time I saw a poll with a margin of error that high, and it undermines the credibility of the whole undertaking.


On the whole, this survey provides some good information. It's mostly consistent with the informal exit polling that was done on Election Day that shows Minneapolis voters made a fairly easy transition to RCV, and preferred it to the previous method. I look forward to the City-funded survey and evaluation report by Saint Cloud State - and commend interim Elections Director Pat O'Connor for choosing to contract with them - which will go into more depth, have a smaller sampling error, and hopefully present the facts in a more even-handed, neutral, and informative way.