So, Bill Rappeleye, Bob Plain, and Justin Katz have a discussion about about instant-runoff voting (IRV) and none of them can explain it right.
Plain starts off the best, by describing what IRV might’ve looked like with the Democratic primary for governor of RI. Voters go in, they rank the candidates in order of their preference for each one (1 for their top priority, 2 for their second choice, and so on), and then the ballots are counted. Raimondo and Taveras have the most votes, and then Plain gets off track, suggesting that it then goes to who’s got more number 1 votes. While I think that might be a system of elections (I’m not sure which), it’s not how IRV works. Plain then gets precariously close by being back on track by talking about runoff elections (the “R” in IRV).
Moderator Rappeleye starts off strong talking about rating candidates in order, and then immediately goes wrong by talking about weighting votes – a feature which doesn’t appear in IRV. Rappeleye’s summary of IRV as a way to cast a vote for someone you believe reflects your values while still being able to throw a vote to someone you think will win has a grain of correctness about it, but is also wrong, wrong, wrong.
Katz at least makes it clear he really doesn’t understand it, but not before he talks about tactical voting by suggesting voters might rank their actual preference low for some reason. That’s wrong. Katz also tries to make the case that because he’s flustered by it, the average person will get flustered by it.
Let’s explain it by having a hypothetical election for best fruit between Apple, Orange, and Banana. Under the current system, known as plurality voting or first-past-the-post the candidate with the most votes wins. That seems natural, but Apple has been winning with 45% of the vote for the past three elections while Orange has gotten 40% and Banana has gotten 15%. This has lead to Orange complaining that Banana is acting as a spoiler. In each successive election, there have been fewer and fewer votes for Banana because Banana’s voters are getting sick of Apple rule and want to vote for a candidate that will “actually win this time.” However, Apple keeps winning, despite not voted against by the majority of voters.
However, the Board of Elections has decided that rather than let things stand, they want to encourage voters to have as many choices as possible, and has instead instituted instant-runoff voting. They looked at doing a more traditional runoff vote method, where there would be another election held between the two candidates with the most votes if no one had a majority, but they decided this would be too expensive and also lead to lower turnout. Here’s a quick flowchart of how it works:
After months of campaigning, three debates, and numerous attack ads, voters go to the polls. They get a similar ballot to the one they had last time, asking which of the three fruits is best. But this time, instead having to fill in an arrow for their number one candidate, voters are asked to assign each candidate a different number from 1 to 3.
When the ballots are counted, something shocking happens. It turns out, that many Orange voters prefer Banana for her oblong shape and ease of eating, but have not been voting for Banana because they thought she couldn’t win. 45% of voters marked their first choice for Apple, but this time, Orange wins only 25% of the vote while Banana wins 30%.
With the least votes, Orange is eliminated as a choice, and Orange’s votes are divided up between Apple and Banana based on Orange voters’ second choice. It turns out that a sizable portion of Orange voters like Apple for his round shape and red hue. But many voted for for Banana as well. Apple wins with 55% of the vote, while Banana comes in second with 45% of the vote.
Okay, so what’s the advantage here? Well, first, it usually leads to a majority winner. This means Chafee wouldn’t win with a third of the vote in 2010, and Raimondo wouldn’t be the primary winner with 2/5ths of the vote this year. If you believe in voter mandates, the candidate that wins tends to have the most votes. It eliminates the need for runoff elections which can be costly and can also have lower turnouts depending on the date; essentially, IRV holds as many runoff elections as needed on the same day.
Katz’s suggestion about voting tactically is nonsensical. Tactical voting is a major feature of the current system. IRV’s biggest strength is its difficulty in making tactical voting successful. Without knowing how other people are going vote, it’s hard to rank choices in order of preference. Theoretically, the tactical way to vote is to raise your second choice vote to a first choice, but it’s only possible to vote that way once you know the election results.
As for the argument that it’s too confusing, well it’s no more confusing than a pick X number of candidates from a list of many candidates. The voter just is asked to rank candidates in order of preference. Super simple.