Editor’s note: This opinion piece is by Chris Roy, a Republican candidate for Secretary of State.
On Town Meeting Day, Burlington voters ended their experiment with instant runoff voting (IRV). Efforts to apply IRV to statewide elections should likewise be abandoned.
For those unfamiliar with IRV, instead of selecting a single candidate, IRV voters rank multiple candidates in preferred order. If no candidate receives a majority after counting first choices, the second choice of those who supported candidates with fewer votes are added to their totals. If no candidate reaches a majority after this round, additional candidates with lower vote totals are knocked out, with their supporters’ second, third, and fourth choices being applied until one emerges with “majority” support.
The theory is that this process of conducting hypothetical “instant runoffs” best reflects the intentions of the voters. In practice, IRV fails.
After approving IRV, Burlington voters have twice elected Progressive Bob Kiss mayor. In 2006, he received 39% support after tallying first-choice votes, and eventually won after two more rounds of IRV.
In 2009, Kiss was again elected, this time after four rounds of IRV. Democrat Andy Montroll received more combined first- and second-choice votes than any other candidate. Republican Kurt Wright had the most votes through the first and second rounds. Mayor Kiss only received 29% first-choice votes as an incumbent and yet, after four rounds, he emerged the winner.
Besides this unexpected outcome, more problems arise when IRV is employed in the real world. First, candidates are competing not only for first-choice votes, but also to be the second choice of voters who may prefer other candidates. Thus, Burlington ’s mayoral “debates” in 2009 focused on ways candidates agreed with each other. Voters were asked to rank multiple candidates while being denied the opportunity to explore distinctions between the candidates as is the case in traditional campaigns. Traditional campaigning would have been more likely to raise tough questions about the status of Burlington Telecom that might have avoided the crisis that arose after the mayoral election.
Second, IRV’s counting of votes involves computers, proprietary algorithms, and a lack of transparency. I observed the IRV “tally” in Burlington and watched as officials from various wards entered carrying black suitcases like government agents in a spy movie. The suitcases were plugged into a central computer, and each “round” of voting was displayed on a screen with the press of a key.
Third, allowing supporters of less popular candidates to have two, three or four votes skews the outcome in favor of their preferences. In contrast, a true runoff between Candidate A and Candidate B allows voters to focus on a clear choice.
In the wake of their real-world experience, Burlington voters rejected IRV and returned to a traditional method of electing mayors. Burlington ’s adoption and rejection of IRV parallels similar voter rejection in cities that previously approved IRV, including Aspen, Colo., and Ann Arbor, Mich.
Second, one can only imagine the time and expense involved if a statewide recount of IRV is ever required – an outcome that is easily foreseen given IRV’s lack of transparency.
Meanwhile, the Vermont House has referred H.396 to the Government Operations Committee. This bill would apply IRV to the election of Vermont ’s governor. Statewide use of IRV would implicate the preceding problems experienced in Burlington , and three more concerns.
First, it would be a substantial burden to ask Vermont ’s municipalities, from Bennington to Canaan and all towns in between, to administer a computer-based voting system that imposes logistic and technological burdens on local election officials. Not surprisingly, most town clerks have opposed IRV.
Second, one can only imagine the time and expense involved if a statewide recount of IRV is ever required – an outcome that is easily foreseen given IRV’s lack of transparency.
Third, one of the primary arguments in favor of IRV is that it avoids the cost of a runoff election. Under Vermont ’s Constitution, however, any gubernatorial election not resulting in a majority winner goes to the Legislature where, traditionally, the voters’ plurality choice has been respected.
Vermonters should take the lesson learned by Burlington voters to heart, and consign IRV to the ash heap of electoral history.
