Run This Off
By Mitch Berg
One of the biggest defeats in Saint Paul Tuesday night was the passage of Instant Runoff Voting.
“IRV” will bring endless resolutions, opaque unaccountable recounts, and – in the worst cases – elections where “Majorities” can only be reached by tossing ballots who don’t vote for any of the “finalists” in the byzantine counting processes – meaning winners will get “majorities” of ballots with preferences, but minorities of actual votes cast.
On the upside, “victory” parties will go on forever; Minneapolis likely won’t see official results from its various races until Christmas.
Christmas.
Jeff Rosenberg at MNPublius tries to explain last night’s Minneapolis election results. And I do mean “tries”. Jeff is a capable enough writer, and explains statistics as well as anyone in the Twin Cities’ ‘sphere. (They’re often as not wrong and out of context, but they’re explained well!).
To wit:
There were no major surprises, and few chances to really see ranked-choice voting come into play in Minneapolis last night. Joe Bodell has the complete results at MN Progressive Project [Sometimes called “Minnesota “Tragedy of Spyrochaetal Paresis” “Progressive” Project”]. There were a few races were 2nd-choice votes will be needed to officially put one candidate over 50 percent, but in all but one case the results are clear, with the leading candidate receiving 47 or 48 percent of the 1st-choice vote and the nearest challenger with under 40 percent. These races would all take miracles for the 2nd-place finisher on the 1st-choice votes to come from behind.
Which is a good thing, because recounts will be just plain hell.
There were two cases, though, where we’ll see ranked choice voting (i.e. Instant Runoff Voting) come into play, both involving the Park Board. In District 5, Carol Kummer and Jason Stone are nail-bitingly close. This one will be determined by 2nd-choice votes — in fact, it may even need to be decided by the 3rd-choice votes of those whose first choices were Barland and Looney.
Because Minneapolis voting machines could not be calibrated to tabulate anything other than the 1st-choice votes, the ballots will all need to be tabulated by hand, which means it will be some time before we know the winner of the Park Board race.
Is this reminding anyone of rotisserie league baseball yet?
I mean, that seems to be IRV’s main feature; it gives wonks stuff to chatter about. Forever.
Congrats, Saint Paul. You turned your electoral system into a friggin’ wonk’s parlor game.
November 5th, 2009 at 7:10 am
The “more democracy equals better government” theory. Unless you want to vote on abortion rights or gay marriage, I suppose.
November 5th, 2009 at 8:23 am
I thought it worked just fine. Alas, not anywhere close to enough folks in Minneapolis voted “Anybody But Rybak,” which is what I did.
November 5th, 2009 at 9:10 am
I compared the election results for Minneapolis and Saint Paul and found they were anything but fine. I looked at the results for Saint Paul and I knew the results instantly. The Minneapolis results took a few seconds to decipher. It was interesting, but making election results anything other than crystal clear is not a bebefit.
In my view, IRV is a useless complication. How much did it cost (money, time, effort on the part of all involved) to avoid the possible expense of a recount or run-off in case of a tie?
November 5th, 2009 at 10:06 am
If the only benefit of IRV were to avoid the runoff problem, it wouldn’t be worth the trouble; ties are rare.
What IRV can do, among other things, is allow those disatisfied with one-party dominance to let their vote give expression — and effect — to a consensus that the single-party choice is the wrong one. Demonstrably, when the consensus is, as it was in Minneapolis, in favor of the one-party nominee, IRV doesn’t (and, shouldn’t) prevent his election.
But, yes, it does take a few seconds to make sense of the results, at first; a full explanation of the algorithm takes a few minutes; and manifestly absurd results can happen, just as they can in the electoral vote system, or the plurality sweeps system.
November 5th, 2009 at 11:38 am
“Because Minneapolis voting machines could not be calibrated to tabulate anything other than the 1st-choice votes, the ballots will all need to be tabulated by hand…”
Congratulations Minneapolis and now St. Paul, you’ve returned to the stone age in ballot counting. And how will anybody know if the ballot counting machines get it wrong in the future since the complication makes it less transparent? Also, I’d like to know whether the machines were “calibrated” enough to check for invalid ballots (voters voting the same person multiple times, people who thought the ranking was vertical rather than horizontal, etc.) I can’t wait for the next general election.
November 5th, 2009 at 11:43 am
Margaret,
I need to go and take a look at the actual ballot; I haven’t yet.
I’d love to find out if anyone actually ran a usability test on the ballot (or, for that matter, if anyone in Henco government knows what a usability test is).
I may go dig for one and run some basic metrics and practices on it just for fun.
November 5th, 2009 at 11:52 am
I’d like to know whether the machines were “calibrated” enough to check for invalid ballots (voters voting the same person multiple times, people who thought the ranking was vertical rather than horizontal, etc.)
Yup. I asked one of the poll workers about that; there’d been a few folks who had initially made at least the first error, and had to (well, got to) redo their ballots.
I’ll be interested in Mitch’s professional evaluation of the ballot. I thought it was easy to understand, both graphically and in terms of the short instructions, but I’m probably not a good test of that, as I was pretty familiar with IRV, already.
November 5th, 2009 at 1:28 pm
I’d like to know if the machines were calibrated (or have ever been calibrated) enough to recognize candidates without a D or G after their names.
As for the usability study, I think they used the same company that tested those butterfly ballots in Florida. It’s not as if it matters, or anything. Complex balloting and counting only means we need more government assistance to straighten things out. Wait for the announcement of a new Department of Counting Ballots, staffed by bureaucrats ready to selflessly count until they get the desired answer.