A few days after I posted my critique of Jim Harper's post advocating voting, Bob Murphy posted an excellent critique that affirms my point and goes further.
Recall Jim Harper's statement that even though your vote is highly unlikely to affect who's elected in an election with lots of voters, it still matters for the margin of victory or defeat. I pointed out that your vote has no noticeable effect on the margin either. I wrote:
If you find yourself persuaded by Jim's point about margin of victory, ask yourself this: Think of a candidate, candidate A, whose views you liked a lot more than those of his opponent, candidate B. To make it real think about an actual candidate A whom you voted for. Now ask yourself, without checking data, by what number of votes did he/she win?
Bob puts it this way:
Do you really think anybody cares whether a sitting president won by a margin of 2,321,210 as opposed to 2,321,211? Since that type of miniscule difference will affect the behavior of precisely ZERO people-including the policymakers in DC whom Harper said respond to margins of victory-it follows that casting a vote in a presidential election confers basically no benefits, on the margin, if we are thinking in purely instrumental terms.
But he has a further critique. Here's how he starts his hypothetical situation:
I have been trying for a while to get even my economist friends to "think like an economist on the margin" when it comes to casting votes for president. I think I've finally got it! Suppose the rules are tweaked so that (for some reason) the government approaches Jim Smith and tells him, "Every other eligible voter still gets one vote, just like before, but we've got a deal for you, Jim. You get your first vote for free, like before, but after that, you can cast as many additional votes for president as you want, but you have to pay $100 per additional vote. So how many do you want to buy?"