Refutations of false or misleading claims by Howard Dean about Instant Runoff Voting (IRV)

Howard Dean [Vermont Governor 1991-2003, briefly the frontrunning USA presidential candidate in 2004, chair of Democratic Party 2005-2009, founder of Democracy For America] published a pro-instant-runoff-voting op-ed "How to move beyond the two-party system" in the New York Times Friday 7 October 2016. As far as I currently am aware, Dean is the most prominent US politician yet to have written an op-ed advocating better voting methods. It is unfortunate that he knew so little about them – we therefore must applaud Dean's vague instincts, but cannot support his specifics. You can read our verbatim copy of Dean's op-ed on the other page [and see also our page giving NY Times readers' letters responding to Dean]. Unfortunately it contained numerous errors, lies, and/or misleading claims, as we will now show.

First of all, Dean's title "How to move beyond the two-party system" is very misleading. As Dean correctly notes, the country that has used IRV more than every other combined has been Australia, whose lower House has been elected via IRV for nearly 100 years. (But note: Australia's senate is not elected via IRV.) The net result of that was this: In the 3 consecutive House elections of 2001, 2004, and 2007, electing 150+150+150=450 seats in all, a grand total of zero third-party House members were elected, despite a goodly number of third-party Senators (e.g. as of 2016 their Senate contains 9 Greens, 4 "One Nation," 1 "Family First," 1 "Liberal Democrat," and 3 "Xenophon team" senators, in addition to the 26 Labor and 30 LibNat major-party senators). IRV evidently yields 2-party dominance at a comparably massive level to the USA even in a country whose third-parties are much better off than in the USA. So a more accurate title would have been "How to stay stuck in the two-party system."

Second, Dean's terminology "ranked-choice voting" instead of the correct name "instant runoff voting" also is very misleading. In fact, there are an infinite number of voting systems based on rank-ordering style ballots, in principle – and over 100 have been seriously proposed for actual use. Instant runoff is just one of them, and not a terribly good one either. By the misuse of the name "ranked-choice voting" to mean IRV, Dean fosters the utterly false impressions that there is only one rank-order ballot voting system, or that there is some sort of consensus for IRV as "the" ranked system.

Dean misleadingly conflates several "ranked choice" voting systems into one. (E.g. London and Australia do not employ same system.) Dean's later description shows the system he wants is instant runoff voting (IRV), although Dean himself never uses that name.

Dean then says "the fundamental issue is majority rule" and claims IRV shows "how best to uphold majority rule." Unfortunately IRV can and has opposed majority rule, a counterexample being the IRV mayoral election in Dean's own city of Burlington in 2009! Montroll was preferred by a voter majority versus X, for every rival X, according to the official ballots – but elected Kiss, despite those official ballots preferring Montroll by 4067-3477 majority over Kiss. Also, IRV can elect X even though Y is preferred to X by a landslide supermajority (indeed even if Y is preferred versus every rival by a landslide supermajority) and Y has more top rankings (indeed even if Y has more rankings above a threshold, for every possible threshold simultaneously, than every rival). And IRV's so-called "majority winner" can be the same person as its "majority loser" (elected by the reversed ballots) – oops.

But Dean at the same time makes an even more "fundamental" (to use his word) error than that. And that is: "majority rule" is, as a fundamental principle, not even a good idea: If anybody contends it is ("naive majoritarianism"), then we can force that person to logically contradict himself. To avoid that fate, Dean needs to restart from a better underlying philosophy, free of internal contradictions.

Then Dean claims that with IRV, "Civility is substantially improved. Needing to reach out to more voters leads candidates to reduce personal attacks and govern more inclusively."

While that is a touching sentiment, unfortunately Dean gives no evidence for this speculation. What we genuinely do know is that IRV is biased in favor of "extremists" and against "centrists." So if, to win, it is better to seem extreme, is that going to "improve civility" and yield "more inclusive" governance? This is, to say the least, not obvious.

Dean then falsely claims with instant runoff voting (IRV), "major parties don't have to fear being 'spoiled'." But in Burlington 2009, Montroll was the candidate of Dean's own Democratic Party (major) and lost because Wright was a "spoiler" – had Wright dropped out of the race, Montroll would have won. (Exactly like Nader "spoiled" it for Gore in 2000.) Again, Burlington's official IRV ballots say so. In Florida 2000, in contrast, it is not possible to prove Nader was a "spoiler" from the ballots alone – poll data showing Nader voters preferred Gore over Bush are also needed.

Soon after this fiasco, Burlington repealed IRV.

Dean contends IRV is "as simple as 1-2-3" but yet, as we have just shown, he, who considers himself an expert, was unable to understand it, and did not realize what happened with IRV in his very own home city!

A better and simpler (and more popular, say surveys) single-winner voting system is "score voting": voters score each candidate on an 0-to-9 scale (or, if they do not want to express an opinion about candidate X, they may leave him unscored); highest average score wins. Even simpler, although probably not as good, is "approval voting."

As Dean remarked near the start of his op-ed "We keep repeating this cycle." But the cycle is: yet another uninformed so-called expert who never educated himself about voting theory produces yet another op-ed full of lies, errors and/or distortions; it sometimes encourages voters to enact IRV; and later the discouraged/conned voters, after yet another disaster, repeal it. Just like happened in Dean's own city of Burlington. That is the actual cycle that keeps happening, in some cases right in front of Dean's nose.

To break that cycle, people like Dean need to actually learn about voting methods, then recommend ones which are going to be less disappointing. I suggest he read this website, and/or get in touch with us.

Postscript: Newsweek then published an article ("How Election 2016 Would Be Different With Ranked-Choice Voting," 17 Oct. 2016, by Paul Raeburn) based on the Dean op-ed. But contrary to what Newsweek claimed, Nobelist Ken Arrow has endorsed "range voting", a voting system not mentioned in the Newsweek piece, and which is not subject to his impossibility theorem, despite that article's false claim that Arrow and/or "game theorists" have shown there is no voting system that escapes that theorem.

Return to main page