FairVote unabashedly believes that the second IRV election for mayor in Burlington this month was a great success. Of course in a partisan election some people will be disappointed by the result, but objectively IRV: 1) successfully prevented the election of the candidate who would likely have won under plurality rules, but would have lost to either of the other top finishers in a runoff; 2) saw voters handle the system remarkably well, with 99.99% valid ballots and equally effective use of rankings in low-income wards and high-income wards despite a minimal voter education effort by the city; and 3) contributed to producing a campaign among four serious candidates that was widely praised for its substantive nature.

However, there are opponents of IRV who have taken to the Internet to tout their "analysis" of the Burlington election in hopes of derailing the expanding use of IRV. Interestingly these IRV opponents can unite in criticizing IRV from opposite perspectives. In other words, some opponents oppose IRV because they feel IRV's accommodation of multiple parties undercuts the two-party system, while others feel IRV doesn't do enough to break the two party "duopoly." Some feel IRV places too much value on voters' first choices, while others feel IRV places too little value on first choices. And so on.

In this post I will briefly address one faulty analysis that is posted on the web site of Warren Smith who advocates a novel voting system called "range voting" that is not used in any public elections. This site devotes a lot of space to tearing down other election reform proposals (and thereby, effectively propping up the status quo).

The analysis opens with a personal attack against me as an individual, which is rather odd. I am used to personal attacks having served a decade on the city council and another decade as a state legislator as a member of the Progressive Party. I have also worked as an election administrator for non-profit organizations when not working as an analyst for FairVote.

Now turning to the defective analysis...

The irony is, that this anti-IRV-biased analysis is co-signed by a Vermont professor named Anthony Gierzynski. Gierzynski is a strong defender of plurality voting for statewide elections in Vermont and a two-party system where third parties don't run candidates. In 2002 he in fact ran for state legislature as a Democrat against incumbent state legislator, the Progressive Party's Bob Kiss, who later went on to be the winner in Burlington's two IRV elections for mayor. Professor Gierzynski does not disclose this obvious source of potential bias when offering analysis of Bob Kiss' victories in the 2006 and 2009 Burlington elections.

The new analysis also is being cited by other IRV opponents who defend current plurality elections (vote-for-one-top-vote-getter) and two-party system, even though it states: "IRV still seems to have performed better in this election than plain plurality voting, which (based on top-preference votes) would have elected Wright. That would have been even worse, since Wright actually was a "lose-to-all loser" among the Big Three, i.e. would have lost head-to-head races versus either Kiss or Montroll."

Here are just a few misconceptions in the analysis:

1. The author, incorrectly states that Burlington's IRV election suffered from a monotonicity failure. You can read about the monotonicity criterion  here. In fact, no such failure occurred. To understand the context, here is the situation...Republican Wright was ahead in first choices in the initial tally with 2951 votes, next came Progressive Kiss at 2,585, and the Democrat, Montroll, was in third with 2063. There were also two other candidates with 1,306 and 35 votes respectively (and some write-ins). In the final runoff tally, Republican Wright had 48.5% and Progressive Kiss won with 51.5% (with some voters sitting out the runoff by not ranking either of the finalists).

What his analysis actually shows is that non-monotonicity could have affected the election, but did not. ...If just over 25% of the supporters of Republican Wright had abandoned their true first choice and instead voted for any other candidate (although to meet the non-monotonicity definition they would need to switch to Kiss), they could have kept their favorite candidate, Wright, from making it into the runoff, and allowed the Democrat to face off against Kiss in the final runoff, where the Democrat would beat Kiss. But this did not happen, and there is nobody who thinks it was a sensible strategy for any voters or candidates to advocate. Certainly it had no impact on how candidates campaigned nor ever would have.

Another way that non-monotonicity could have occurred would be if Wright or Montroll lost the election because they got too many first choices that might have gone to some other candidate instead. Since Montroll didn't even make it into the final runoff there is no way this could apply to him (any fewer first choices would just confirm his elimination). That leaves Wright. Again, even if some of his supporters had voted for any other candidate first, Wright would still lose the runoff between either Kiss or Montroll. Smith would need  1,279 first choices  that actually went to Wright, to get Smith into the final runoff. But  at that point Wright would be in fourth place and have no chance of advancing to the final runoff.  So it is mathematically impossible for a switch of first choices away from Wright to have made him a winner. Thus, despite the Smith-Gierzynsnki analysis, there was, in fact, no non-monotonicity event in the Burlington election.

It is also worth noting that this same kind of non-monotonic strategy (Republicans conspiring to help elect the Democrat to block the Progressive) could be pursued under the old separate runoff system just as well. In fact, non-monotonicity is a much bigger risk under two-election runoffs because with a separate runoff, strategic manipulators can change their first choice on their ballot between rounds, which can't be done with IRV.

2. In another point in the analysis, Smith and Gierzynski attack IRV for failing to elect the apparent compromise Condorcet-winner. This is disingenuous because Warren Smith himself dismisses the Condorcet-criterion, since his favored method also fails to meet this criterion -- in fact, range voting can elect the Condorcet-loser, which IRV never can, and could quite possibly allow the defeat of a candidate who won an absolutely majority of 51% or more of voters' first choices.

They observe that in the Burlington election, the candidate who came in third in the initial tally was a compromise choice who could have beat either Wright or Kiss in a head-to-head election. This is called a Condorcet winner, and is a mathematically valid calculation, though little regarded in American politics. The point is that this compromise candidate would come in third and lose badly under the plurality system most IRV opponents support, and would also be eliminated in a traditional two-round runoff election system. In fact, such compromise candidates have a better chance under IRV than any other voting system used by any government anywhere in the world. But some anti-IRV, pro-plurality activists illogically use IRV's failure to elect the third-place candidate in this election as an excuse for attacking IRV and supporting plurality elections.

This criticism of IRV is legitimate when coming from Condorcet advocates, (which Warren Smith is not, since his favored system also fails the Condorcet criterion). It comes down to a matter of what values one feels are most important in an election process - -both in who should win and what kind of campaign you want to see run. I and many other experts feel the "mutual-majority" and "later-no-harm" criteria are far more important than the Condorcet criterion, for example.

Condorcet voting methods discount the relative importance of first choices, to the extent that a candidate who came in last place in terms of first choices, or even wins no first choices at all, but who is a broadly acceptable second choice, can win a Condorcet election. One concern is how that might affect candidate policy discussion, where the avoidance of alienating any voters becomes more important than the earning of any first choices. But this, at least is an area of legitimate disagreement over what values to reward in an election process, where reasonable and honest people can hope to resolve their differences through open discussion.

So the Smith-Gierzysnki analysis shows that IRV is better than plurality (and in fact better than two-election runoffs), yet plurality voting advocates twist the story to claim it shows how bad IRV is. The lead author of the report favors other un-tested voting methods that he naively thinks are better than any other system, and constantly attacks IRV in hopes of winning support for his favorite theoretical system. The net result is to help maintain the status quo.

For more defense of IRV in Burlington, including my critique of Gierzsynki's unfounded claim that IRV creates a bias against less well-educated voters, see my post at Vermont Daily Briefing.