Services
Commentary
Instant runoff voting and buyer's remorse
by Chuck RepkeThe advocates of ranked-choice, or instant runoff voting (IRV), make a lot of false or misleading claims. St. Paul voters need to know the truth before Election Day, when they will decide whether to adopt the system.
Under IRV, voters would rank, in order, their preferences for mayor or City Council. If no candidate won a simple majority when the votes were counted, the candidate with the lowest votes would be eliminated and his supporters' second choices distributed to the remaining candidates.
This would continue until one candidate achieved a majority of the votes still being counted, or until only two candidates were left standing.
Advocates claim IRV would save money by eliminating primary elections. This is not true. In St. Paul, the School Board will still hold a costly primary election, though the turnout for that primary will be hurt by the absence of mayoral and council races.
IRV advocates claim the system is "democracy of the future," "as simple as one, two, three." In the last two years, seven cities or counties in the United States have used it, and three of them have either repealed IRV or have IRV up for repeal this year. Cities are suffering buyer's remorse:
Tacoma, Wash. Pierce County spent $1.6 million on a voter education campaign for IRV in 2008, yet 66 percent of 90,000 voters polled said it was confusing, frustrating and a waste of time. It is on the ballot this year for repeal.
Cary, N.C. After the IRV election in 2007, 30 percent of voters polled found it confusing and 22 percent said they "did not understand IRV at all." Cary has stopped using IRV.
Aspen, Colo. The results of an IRV election revealed that one losing candidate for City Council would have won the election if 75 of his supporters had voted for him as their second instead of as their first choice. The complicated IRV counting system cost him the election. The repeal of IRV is on the ballot this year.
Advocates claim that IRV will increase turnout. This is not true. In San Francisco's first mayor's race with IRV, voter turnout was 40 percent less than in its last traditional election.
This year the IRV-format election in Minneapolis may hit record low turnouts. With 11 candidates running, there has been almost no media coverage of the mayor's race, and the incumbent has agreed to debate only one of his opponents.
In St. Paul, meanwhile, Eva Ng came in second in the primary and has moved on to the general election. Her primary showing enhanced her credibility and has given her access to media coverage as the alternative candidate for mayor.
Advocates claim that IRV makes sure that the winner "always has a majority vote." This is not true.
In Burlington, Vt., the mayor was re-elected with only a majority of those ballots still being counted in the final round. Seventeen percent of ballots cast for candidates in the first round were eliminated during the IRV counting.
In San Francisco, in 10 of 11 elections for which IRV has been used, the winner received less than the 50 percent plus one that would constitute a majority.
Advocates claim that IRV increases minority participation and thereby helps minority candidates. This is untrue. Using IRV, Takoma Park, Md., with close to 40 percent minority population, elected an all-white City Council.
There is no evidence that IRV helps minority candidates. In the highest-minority and lowest-income precincts of San Francisco, 20 percent of the voters do not fill in a second choice on their ballot.
Free, fair elections are the hallmark of any democracy, and every voter deserves to be treated equally. Voters need to know that their votes count, and that voting for the candidates they want can only help them, not hurt them. Instant runoff voting is far too likely to confuse, frustrate and inhibit voters.
-----
Chuck Repke, St. Paul, works in neighborhood development and government relations. He is co-chair of the No Bad Ballots Committee, which is organized to oppose IRV.
Comments (23)
Chuck - you are correct when you write that IRV does not ensure a winner always has a majority of the vote. It all depends on how you define "majority". I define majority as 50% plus one vote of the total number of votes counted in the first round. In Cary NC in 2007, 3022 votes were cast in the first round. No one had 1512 votes, so they went to IRV. The eventual "winner" was awarded the office with 1401 votes - 111 votes short of the threshold. In all but a very few instances, all IRV ever does is give you larger plurality wins - but at what cost? IRV confuses voters, is not easy to count (unless you get conned into using an Excel spreadsheet where you can't actually see the votes being counted), and costs extra each and every time you have to rank candidates whether or not you have a winner in the first round.
Chuck,
Unfortunately, you have relied on information from Joyce McCloy of NC Voter, a zealous opponent of IRV who has distorted the reality of much of what you discuss in your piece.
For example:
* Cary, NC:
Cary didn't repeal IRV. Rather, the city council didn't affirmatively vote to try it again for a second oone-time use in 2009. The key council meeting on this issue took place when it was unclear how ballots would be counted.
Your highly selective use of the polls from Cary distorts the reality of what that city's voters think about IRV.A politically diverse city (majority registered Republicans, but enough Democrats to have a mayor endorsed by Democrats), Cary voters overwhelmingly prefer IRV to traditional runoffs.
More than 70% of voters preferred IRV to their former runoff system in a North Carolina State exit poll in 2007, and then a full poll conducted by Cary in 2008 affirmed an overwhelming preference for using IRV again -- indeed, on a scale of 1 to 9, with 1 being most opposed to IRV being used and 9 being most in favor, 67% indicated a 7 or higher (including 51% indicating the highest level of 9) while only 7% indicated 3 or less. That's nearly a ten to one margin for IRV.
Pierce County, WA:
We'll see what happens n the vote on repealing IRV on November 3rd, but both daily papers serving the county have endorsed keeping IRV, as has the League of Women Voters.
Aspen, Colo.
Aspen had its highest turnout election ever this year and hotly contested races. You express concern about the nonmonotonicity effect, yet fail to acknowledge that EXACTLY the same thing could happen in St. Paul's current two-round runoff system. Furthermore, the winning candidate would have defeated the candidate in a one-on-one race -- hard to say that's not a fair result.
Turnout, campaign spending, fairness and majority rule:
IRV elects the candidate who has a majority of the vote in voter's preferences between the top two -- just like runoffs, not everyone will have a preference between the final two.
It also is absolutely true that turnout is almost always higher in decisive elections with IRV. St. Paul's low turnout primaries are decisive for the candidates failing to advance to the top two, and the San Francisco runoffs that would have taken place in December which IRV has avoided would have had far lower turnout than the November elections held with IRV. It's also true that these low turnout first rounds or runoffs almost always have whiter, wealthier, older electorates than higher turnout elections.
As a corrective to your statements about racial minorities and IRV, see a paper on how minorities and others handle ranked ballots:
http://www.fairvote.org/?page=2290
Note that IRV doesn't change winner-take-all dynamics. The Takoma Park city council was all-white before it had IRV and is so now. But note that San Francisco after several IRV elections has its most diverse Board of Superivisors in its history.
One reason some establishment folks like two-round systems is it helps candidates who can get more campaign cash. Candidates are forced to run two elections, and one-on-one races are great for negative attacks as. In the last three mayoral and county executive races with IRV, the big spenders have lost. In Pierce County, the losing Republican spent nearly four times as much as the winning Democrat. In Burlington, the losing Republican heavily outspent the winning Progressive candidate (who spent the 4th most in the race). In Aspen, the losing candidate raised more money than ever raised in a mayoral race in that city.
Unsurprisingly, it is those big spenders that in each case are the ones leading efforts to repeal IRV.
Wow, it is interesting that the first person that commmented on my commentary is the top employee of the National (Fairvote)organization funding these ballot efforts around the country.
I got most of my information from Mr Richies own web site. I didn't say Cary repealed it, I said they stopped using it. When you say 70% prefered it and I say 30% didn't understand it, both things are true.
First let me be clear if you are reading this or you listen to the discussion on almanac tonight or you read an editorial, you will be able to understand IRV. You will be fine.
My point in trying to stand up to this national campaign is to remind my friends on the left that just because they understand something doesn't mean that everyone does. I got into politics almost 40 years ago because of the civil rights movement and one thing I know as a fact is that it is important to make the voting franchise as simple as possible and when report after report on Fairvote's own web site says 85% of voters understood what they did when they were in the voting booth, I hear 15% of the people may have lost their franchise because they didn't understand it.
And when you look at the San Francisco report again on Richie's own web site, you will see that the communities that have issues with this systems are those with language issues, seniors and low income.
So, it will be a great thing for my college educated friends but for once in your lives think about the 15% who won't get it.
Chuck Repke
Your sources of information have an anti-ranked-voting bias. There are numerous factual errors and distortions in what you pressent. As one example, you state that in Burlington 17% of the ballots cast for mayor in the first round were excluded by the end of the tally. Wrong. A bit over 6% of the ballots did not rank either of the two finalists, and were thus "exhausted" (not 17%). But compare that to the drop-off in voter participation typical in a separate runoff election, which in Burlington have been around 50% in recent runoffs.
Terry editing is tough what I was trying to say was that 17% of the people who voted for the two candidates dropped before the final round ended up with no vote in the final election. (That is a lot of words and I don't know if it sounds any better.) But it is a normal drop off in US IRV races that the top of the ticket has about 15% that made no second choice. In your case 17.5% of the people who's first choice didn't make it had no vote in the final round and that creates the 6.8 percent of the public that didn't know that the actual choice would be between the Green and the GOP and thereby didn't have a vote.
My point is that they may have liked to chose between the Green and the GOP but because they didn't imagine it would happen, your system didn't allow them to vote in the final round.
Do you realize the level of sophistication that one has to be at to assume all of the run offs that might happen and continue to make choices? They may have seen the Green and the GOP as a choice between the devil and the deep blue sea, but most people if they knew that choice would be there would choose the deep blue sea... and because they don't imagine it happening they are excluded.
This is what I like - vigorous debate. But I don't see that happening in this town. A few watered down "forums" where each candidate gets a total of 10 minutes because there are so many.
I like the two person debate, just like here. It's fascinating and informative. Like Rob Rich pouncing to spin with "didn't affirmatively vote" or "second oone-time use" (sic), gotta love this. And then his big words, "nonmonotonicity" what the heck is that? Who does he think he is? And how do elections have it? It's an insiders game to be sure.
I miss being informed about the candidates positions on the issues and how they differ from each other, having the cream rise to the top and really debating the issues rather than just a beauty contest. My guess is we were sold something we truly didn't know much about.
Readers can decide for themselves.
I am writing from North Carolina, a state where IRV has been piloted in 2 towns and now only 1 town is now participating.
Some people think it is ok that a significant number wof voters who don't understand IRV. Personally, I thought we had worked hard to eliminate barriers to voters. Didn't we get rid of the poll tax, the literacy test, and hopefully the butterfly ballot in order to protect the vulnerable segments of our population?
If you would like to learn why voter advocates are concerned about the problems with IRV, please visit these sites below:
Instant Runoff Voting in the United States
www.instantrunoffvoting.us (website)
http://instantrunoff.blogspot.com/ (blog on IRV in US)
Protect North Carolina Elections - Stop Instant Runoff Voting - http://irvbad4nc.blogspot.com/
(blog about IRV in North Carolina)
NC Voter website
www.ncvoter.net about voter verified paper ballots, improved voter access, and increased transparency in elections.
And here is a just released report on IRV by a professor employed by the Vermont Legislative Research Division.
Report on IRV by Anthony Gierzynski, Ph.D. University of Vermont
http://www.uvm.edu/~vlrs/IRVassessment.pdf
About us: The North Carolina Coalition for Verified Voting is a grassroots non-partisan organization fighting for clean and verified elections. We study and research the issue of voting to ensure the dignity and integrity of the intention of each voting citizen. We passed a law requiring paper ballots, audits and standards for voting machines, we increased access to voter registration by eliminating the "no match no vote" rule. We improved post election audits.
McCloy went to court against Diebold Election Systems to protect our newly passed law. The NC Voter Verified Coalition has consistently fought for increasing access, participation and ensuring the voter franchise. Contact Joyce McCloy, Director, N.C. Coalition for Verifiable Voting - phone 336-794-1240
Chuck -- I'm glad you've been perusing our website. As to San Francisco, the evidence refutes your claims that people of color and less well-educated people have difficulty with instant runoff voting. That is also the case in Burlington, where this was also studied. As mentioned above, take a look at the information in this report:
http://www.fairvote.org/?page=2290
Note that not understanding something doesn't mean that people will not vote according to their interests. I'm not great with fixing car engines, but can drive my kid to school as effectively as the mechanic who lives down the street. In presidential races,many people might tell a pollster they don't understand the rules of the Electoral College, but still cast a smart vote for president. Similarly, a voter who fully understands IRV and a voter who isn't sure about it will both do the same thing if they read the instructions on the ballot: rank your favorite candidate first, your next favorite second and your third favorite third.
Rob, I was referring to the actual San Francisco report on voter turn out and which precincts had what numbers and not the "study" that was done by "Fairvote" after the fact. One looks at raw data from turn out and gets a different impression than what "Fairvote" gets when it interviews people.
Reading the reports from the leading national advocate for a particular outcome and the only group in the country I know of that has funds to "study" the issue. Makes me reflect back to those great days when we would read "studies" paid for by Philip Morris promoting the benefit of smoking because you exercised your lunges by taking deep drags. In "Fairvote" after the case interviews, one is supposed to believe that whites filled out three choices at an extremely lower rate than persons of color, despite the fact that the precinct data doesn"t reflect that from the percentages of those dropped.
It is like the actual data that Cary, NC had compared to the "exit polls" conducted by the "independent expert." "Fairvote."
My point, in this entire effort is to try to get people to use basic common sense and test with people they know if these assertions, about how simple this is, make sense and if they can"t see where people will be cut out by the process.
Chuck -- The numbers don't confirm your analysis, and other independent studies like San Francisco State University have spent serious resources studying instant runoff voting with conclusions similar to ours.
Chuck -- In discussing Burlington you say that a certain number of voters didn't expect the final choice to be the "Green" and the Republican.
The "Green" in fact was the Progressive Party incumbent mayor, and the Republican was the biggest spender and one of the clear frontrunners. It was indeed clear to voters that they were strong candidates. It's just that a certain number of voters for the two eliminated candidates didn't have a preference between them.
IRV allows people to choose a "lesser of two evils," but they don't have ot -- voters have the right to abstain, and a certain number made that choice. Given that 99.99% of voters cast a valid first choice, it's clear they knew what they were doing.
Rob you do sound like your minions here in Minnesota. You believe that people are choosing not cast a vote in the General Election/Final Runoff and I believe that given the opportunity to vote for one of the final two they would.
What drives me crazy with the IRV advocates is in even number years/Governor elections where everyone is on the ballot, they say that we should change to IRV because "everyone" wants to be able to make that second or third choice to vote their heart and the head. Those voters want to cast a vote Green and Democrat or Independence and Republican. In those elections most of the time we are talking about numbers in the 5-10% range that they are so concerned about, but when 15% of voters don"t cast second choice ballots in their IRV municipal elections, well that is too bad or it was just their choice.
In my heart and from closing in on 40 years of working on political campaigns I know who those voters are that aren"t getting a second bite of the apple and Rob, you do too. Honestly, how many people when asked the question, "Did you understand what you did in the voting booth?" are going to answer that they didn"t know what they were doing. That 15% say it is amazing, 22% in Cary, NC stunning.
We had a recount here in Minnesota and there were a lot of absentee ballots with problems. If surveyed every person who sent in an absentee ballot would say they did it right. They would be wrong about it, but they wouldn"t have sent it in If they thought they messed it up, but those are the surveying tactics of "Fairvote."
I just think this is the most disgustingly selfish thing that I have ever seen supported by my brothers and sisters on the left wing of my party and I think it sets voting rights back years.
One last quick question to the head of "Fairvote", because I went and pulled of the San Francisco data.
The real executive summary that San Francisco did says that they polled the voters in 2005 and 54% of them knew that they would be using IRV when they went to vote. Now, that was the second year of using the system but 46% had no clue that they were going to be asked to rank candidates. Yup, I am sure that a great deal of study was able to be done in that ranking process, Rob.
So, lets see 54% of voters knew when they went to the polls what they were doing and then magically when "Fairvote" talks to them later 85% knew what they were doing...
Yup there's a guy at R.J.Reynalds that got a job open for these guys.
Chuck -- I'm sorry you feel so hostile to this idea and to me, but perhaps in this case you might apologize.
There were exit polls by SFSU done after BOTH the 2004 and 2005 elections in San Francisco. Our data accurately reflects the 2004 one here:
http://www.sfgov.org/site/uploadedfiles/election/Elections_Pages/SFSU-PRI_RCV_final_report_June_30.pdf
Your polling stat from the November 2005 election (the first citywide one) is important. 54% knew about IRV coming in, but 99.6% of voters cast a valid ballot.
Note that there are a lot of people in San Francisco who work with low-income people and people of color (the city is majority non-white), and the system is overwhelmingly popular among such community leaders. One example is the Asian Law Caucus, which did its own extensive exit poll in 2006 and found very strong support among Asian American voters.
This is the case in other places. Memphis, an African American majority city, voted 71% for instant runoff voting last year, with the strong backing of groups like the NAACP. In 2006, Oakland voted 69% for IRV, with strong backing from community leaders of color.
FairVote doesn't have "minions." We have partners in communities around the country who are independent thinkers working hard to do what's best for their community.
Onward.
I find Rob's comment interesting, in that people prefered it to a "traditional runoff".. Of course, people only want to go to the polls once. This is, IMO, misleading.
The question then becomes, how often does any election actually go to a runoff election? Cannot be very often.
Chuck -- I'm sorry you feel hostile to this idea and to me, but note that here were exit polls by SFSU done after BOTH the 2004 and 2005 elections in San Francisco. Our data accurately reflects the 2004 one here:
http://www.sfgov.org/site/uploadedfiles/election/Elections_Pages/SFSU-PRI_RCV_final_report_June_30.pdf
Your polling stat from the November 2005 election (the first citywide one) is important. 54% knew about IRV coming in, but 99.6% of voters cast a valid ballot.
Note that there are a lot of people in San Francisco who work with low-income people and people of color (the city is majority non-white), and the system is overwhelmingly popular among such community leaders. One example is the Asian Law Caucus, which did its own extensive exit poll in 2006 and found very strong support among Asian American voters.
This is the case in other places. Memphis, an African American majority city, voted 71% for instant runoff voting last year, with the strong backing of groups like the NAACP. In 2006, Oakland voted 69% for IRV, with strong backing from community leaders of color.
FairVote doesn't have "minions." We have partners in communities around the country who are independent thinkers working hard to do what's best for their community.
Onward.
Wayne -- runoffs are quite common in Cary, NC. In fact, there is one right now because the first-round leader had 49.97% rather than a majority - meaning the city will spend tens of thousands of dollars, the candidates will need to raise big bucks quickly and quite possibly there will be a lower turnout in the final vote. This election is sure to keep instant runoff voting on the frontburner there.
For some good numbers on how turnout almost always drops significantly in congressional primary runoffs (there have been 116 such runoffs in 1994 to 2008), see:
http://www.fairvote.org/?page=1489
In St. Paul, of course, the low turnout is in the first round in September, as with the miniscule turnout this year. Yet that election eliminates most voter choice in the election cycle.
About Rob Richie's comment regarding Cary, North Carolina. IN spite of the fact that FairVote sent out a national alert to have IRV advocates pummel Cary Town Council members with emails and phone calls, the Town Council said No More to IRV. When Cary tried IRV, they didn't ask the voters first. When Cary was asked to do it again, they had several meetings and also heard from the Cary voters. The decision was that Cary was better off with traditional runoff elections where the winner really does get 50%+1 of the vote. Don Frantz, who won by IRV in 2007 was asked by a IRV supporter if he still preferred traditional runoffs now (in Oct 2009):
"Why ask me?
Ask all of council as we are the ones who decided to not utilize IRV in Cary elections.But my answer is "no". I do not like instant runoff voting and have given my reasons as to why many times. I'll take in elections over funny math and 30% voter confusion any day. You also assume the initial results would be the same if IRV had been utilized. I don't believe that would be the case. __________________Don"
http://irvbad4nc.blogspot.com/2009/10/rebutting-fairvote-misinfo-on-cary-nc.html
With traditional runoff elections, EVERYONE gets a say in the final election.
With IRV, only some voters get a say in the "runoff".
Notice that Rob Richie and Terry Bouricius, the two persons defending IRV in this forum, are both PAID representatives of the well-funded "Fair" (better called Fairytale) Vote, not objective observers.
IRV threatens the integrity, fairness, accuracy, transparency, and verifiability of elections.
For more information on IRV see http://electionmathematics.org IRV page or see Joyce McCloy's web sites on IRV. Information from Fair Vote is deliberately misleading and much of the original false claims about the IRV/STV methods solving the spoiler problem and finding majority winners (it does neither) was spread from there.
Actually, the situation in Burlington was even worse than Chuck Repke said. (I'm speaking as author of study of the Burlington 2009 IRV election, read it at http://rangevoting.org/Burlington.html .)
It wasn't just that IRV failed to produce majority support for the winner (Bob Kiss). In fact, a majority of voters preferred Montroll over Kiss and said so in their ballots. (IRV ignored many of those
preferences and eliminated Montroll; a problem with IRV is it counts some stuff voters say, but not other stuff, and which gets ignored and which gets counted can seem pretty whimsical.)
Another majority preferred Montroll over Wright. In fact, majority preferred Montroll over X for every single X, including the winner Kiss. So it would seem Montroll was the "majority winner" but IRV elected Kiss. Also, the so-called-majority that IRV advocates use to justify Kiss's victory, namely the IRV "last round" where Kiss beat Wright, was a smaller majority than every one of those Montroll-over-X majorities!
So it is pretty clear IRV thwarted the will of the majority in this election, plus exhibited a lot of other crazy pathologies. I personally recommend "range voting" and its simplified form "approval voting."
See
http://rangevoting.org
http://RangeVoting.org/Approval.html
It is a pity that a rather silly and bad system (IRV) which arose as a historical accident in about 1890, is
getting the attention that it is.
I happened upon your commentary because I was looking for a different news item about Takoma Park, where I live.
Using Takoma Park to support your argument is, in a word, ridiculous, because so few races are even contested. Rob Richie (whom I've had the painful pleasure of playing basketball with, but otherwise don't know) would be more familiar than I w/ the history of IRV in the city, but I can tell you that in the 2007 elections for city council and mayor, there was only one contested race and there were no minority candidates.
In the elections about to take place Nov. 3, there is one Latino candidate and one candidate who might qualify as a "minority" simply by virtue of his name. The Latino candidate is running against a white incumbent who received vocal support from African-American residents at our recent nominating caucus. (In fact, he regained his seat a few elections back when the voters of his ward, which is predominantly made up of African-Americans and other minorities, cast enough write-in ballots to defeat the African-American incumbent).
I might add that our mayor is openly gay, as are two council members, not that the issue ever comes up. Does that count toward diversity? (Probably not in our city.)
The fact remains, however, that w/o competition, IRV isn't an issue.
Great job on Almanac Mr. Repke! It's definitely an uphill battle to promote change, but honest caution is necessary too. Thanks!
Interesting most of the comments here are from out of state. And it always fun to see the pro-plurality (status quo) people with the radical range vote supporters against IRV, as if they're in united agreement in any way!
My own argument isn't against IRV, but against eliminating the primary. Minneapolis should have retained a primary to reduce the field to "viable candidates" - those that can get 15% or higher in the primary. I mean we could have an IRV primary and IRV general election.
Minneapolis will also be using the STV variation of IRV for multiple seat elections like Park board. That's more progressive use of ranked ballots. Potentially 90% of voters can count towards winners in STV, versus 50% or less of voters in the old plurality-at-large could control all the seats, so that might bring some interesting diversity of representation. School boards could also benefit by this method.
Doesn't media coverage and public debate narrow the field of candidates?
It's weird that proponents of the status quo pretend to be so upset that IRV can't always award the vote to a candidate with support from a true majority of voters. Our current system awards a candidate the winner if they have a plurality. Attacking IRV for operating only as well as our current system of voting seems less like debate and more like verbal trickery--you are saying it is the same, but trying to make it sound like it is worse by how you say it.
Post a comment
Please be civil, brief and relevant.
E-mail addresses are never displayed but they are required to confirm your comments. All comments are moderated. MPR reserves the right to edit any comments on this site and to read them on the air with attribution. Please read the Comment Guidelines before posting.
You must be 13 or over to submit information to Minnesota Public Radio. The information entered into this form will not be used to send unsolicited e-mail and will not be sold to a third party. For more information see Terms and Conditions and Privacy Policy.
Latest News & Features
News Cut
with Bob CollinsIn The Spotlight
-
Sound Learning
Looking for ways to ensure your students are doing quality research from credible sources? Sound Learning is a launching point to Minnesota Public Radio's content on the Web. -
The Current Music Blog
Your daily note for good music, news and pop culture. With attempted jokes.
- Become a Member
- Member Connect Accounts
- Leadership Circle
- Sustaining Members
- Member Discounts
- Become a Sponsor
- Tax Information
- Planned Giving
- Vehicle Donation
- Donate Stock
