I have a selfish interest in living in a well-governed state – my quality of life will be better;

if some interfering economist comes up with an argument that it is irrational to vote, there is a risk that he will be persuasive;

in which case, the people he persuades will be the sort of people who can follow economics arguments;

and they will thus stop voting;

which means that the electorate will become generally stupider;

and are thus less likely to elect good governments;

harming me.

I will therefore support Brian’s argument without even reading it through.

It does, every time they renew the contract[s] for tv coverage.

So the reason I said you were double counting is because you incorrectly characterized your second number, the likelihood that your state influences the outcome of the election. It has nothing to do with swing-statehood and everything to do with how the electoral votes split in absence of your state.

If you take trouble to do things that benefit other people in order to gratify your imagination, is that selfish or altruistic?

Only in a democracy.

You say “I think the probability that your vote matters is the probability that (a) you’ll be important to how your state votes.”

There’s still seems to be a logical problem here, since the the class of how a state votes depends on the the individual votes that comprise the class. The probable value of individual votes cannot directly depend on “how your state votes” since it partially causally determines how the state votes. So we need more sophisticated logical syntax to express what may be, after all, a reasonable point: that if (besides me) a majority of plus two or more will vote one way, there’s no reason for me to add another vote to the pile.

I’m still not finding this satisfying, since it sounds artificial: I don’t really have a way of evaluating the probability that my vote is the majority plus third on the pile. Or, in Conchis’ example: it would be analogous to if I had to decide whether or not to help push, but the number (if any) of other people who are pushing either way won’t be revealed to me until after I’ve decided.

It also occurs to me that one problem with looking at voting as a Prisoner’s Dilemma game is that preferences diverge so much that some people’s payoff matrices aren’t typical of Prisoner’s Dilemmas. A lot of people vote because they see it as a civic duty or as a positive good in its own right. You can defect on these people all day and they’ll happily trudge back to the polls next November – for them, the expected value of voting is almost entirely independent of the likelihood that their vote will make a difference or of what other people do.