Tentative conclusions on democracy & governance
-
“What-if games” and the VT IRV veto
Last week, Vermont governor Jim Douglas vetoed legislation to elect the state’s congressional delegation by instant runoff voting.
The New Yorker’s Hendrik Hertzberg wrote a good critique of Douglas’ veto message. Hendrik is probably right that partisan calculations decided the matter, but the governor gave other reasons. Some were factually incorrect (e.g. the current system is 200 years old). Others were tired spin (e.g. bastardized understandings of “one person, one vote” that IRV somehow violates).
One novel argument stood out:
Moreover, voters should not be asked to cast their ballots based on a wide range of hypothetical, theoretical or imaginary outcomes.
I take it to mean: IRV raises the information costs of strategic voting. Douglas earlier had fed this argument to the Burlington Free Press, which editorialized against IRV on March 25:
While backers tout instant runoff voting — IRV for short — as a way to increase the voters’ say, what it really does is mess with the process of marking the ballot, forcing people to vote in both a real election and, as the governor said, a hypothetical one [...] Many voters might play the “what if” game in choosing whom to vote for when there is more than two candidates in the race, but that kind of conjecture has no place on the official ballot.
While technically correct under certain conditions, this is a bad argument against reform for at least three reasons. One, those conditions are rare. Two, no voting system is strategy-proof. Three, strategic voting is less important under IRV than under plurality.
Why do governors even bother explaining vetoes of election reform bills?
5 responses to to ““What-if games” and the VT IRV veto”
-
Jack, I’m not sure what you mean by “raises the information costs of strategic voting.” Even if the Governor understood the concept of strategic voting, I don’t think that’s what he’s referring to in his argument. I think he’s suggesting that it’s somehow strange or requires to much effort from voters to ask how they might vote in a future runoff, when that runoff may not happen. I don’t find that strange or too burdensome, let alone an overriding concern, and neither, apparently, do people who actually use IRV.
To be most generous to the Governor’s reasoning, we could see it as an argument for two-round runoff over IRV. As the argument goes, TRR gives voters time to deliberate and re-consider the final two choices, rather than require a complete ranking in advance, which requires extra homework by the voter. It’s a fair point, but not a winning one. However, it’s a completely moot point, considering Vermont doesn’t use TRR for federal offices, nor is he proposing it.
This was a partisan maneuver, so we shouldn’t expect much more than nonsense arguments.
-
Thanks for your thoughts, Greg. You could be right, or we could both be wrong. Maybe Douglas was being deliberately confusing. If that were the case, how do we explain the BFP’s responsiveness to this line of argument?
I was also looking for the strong form of his reasoning. It looked like a repackaged replay of the “anomalous results under IRV” argument. This is an argument we get from opponents who argue IRV is not strategy-proof – that under some conditions, voters do have incentives to misrepresent their preference orderings. Knowing how to do so nonetheless requires an arguably impossible amount of information about the other voters’ preferences. Hence my point about information costs.
-
For what (little) it’s worth, at first I read the “hypothetical” and “imaginary” part of the veto message pretty much the way Greg did. But I see Jack’s point. Rather than both being wrong, you could both be right. Since the test of the message is obviously confused, the writer could have been influenced by partly digested versions of both arguments.
-
I would not be surprised if that were the case given the active (mis)use of social choice theory in lobbying against IRV.
-
I’m no fan of IRV (I like approval and, with proper instructions,
low granularity range) but I agree that the arguments given
against IRV by Douglas are weak.>[Jack]This is an argument we get from opponents who argue IRV is
>not strategy-proof – that under some conditions, voters do
>have incentives to misrepresent their preference orderings.
>Knowing how to do so nonetheless requires an arguably
>impossible amount of information about the other voters’
>preferences
Disagree. ‘Bury the second choice front runner’ is an
simple IRV strategy that does not require very much
information about how other people vote.
Leave a reply
-
Recent Comments