Voting

Is voting cost-effective? (0.00 / 0)

Steven Landsburg has argued that individuals shouldn't vote because the probability of casting a tie-breaking ballot is too small. For instance,

[In 2000], about 6.5 million votes were cast for major party candidates in New York state and 63 percent of them went to Al Gore. Assuming an electorate of similar size with a similar bias, my chance of casting the deciding vote in New York is about one in 10 to the 200,708th power. I have a better chance of winning the Powerball jackpot 7,400 times in a row than of affecting the election's outcome.

I think our subjective odds of casting a tie-breaking vote are not this bad, even in New York. Landsburg's formula assumes that the number of votes for a particular candidate in a two-way race follows a binomial distribution with number of trials equal to the (assumed even) number of voters in the state and success probability p. It's not obvious to me that this is the right model to use, but even if it is, we shouldn't assume a single value for p (e.g., saying p = 0.63 as he did for New York); rather, we should come up with a subjective probability density over p and integrate with respect to that density the probability of changing the election given a particular p over the possible values of p. In the region p ~= 0.5, our probability of impact will be large (for p=0.5 the probability of casting the tie vote is 1 in 3070), and this may make the resulting integral much larger than Lansburg's calculations show.

There appear to be copious publications on this topic (see, e.g., this discussion of various approaches to "pivotal probabilities"), and I'm sure other people who are familiar with the literature could say a lot more.

Even if the probability of a vote influencing the outcome of an election is high, it remains to be seen whether one should actually vote. Perhaps instead of taking a hour to drive to the voting booth, you could work for an hour getting out the vote in areas very likely to vote for your chosen candidates. Or maybe you could spend the hour working and earn money to donate toward similar efforts.

Even if it would be more cost-effective to pursue one of these latter options than to vote yourself, you might still vote personally just to avoid the appearance of hypocrisy. And perhaps if your actions have a large influence on the actions of others (e.g., if you're a rock star), then you would want to vote to encourage others to do the same--since nearly all of those other people wouldn't otherwise be doing something more cost-effective with their time.

---

by: Alan Dawrst @ Tue Feb 26, 2008 at 18:55:42 PM CST

voting (0.00 / 0)

chyeeeeeeeeeeeaaaaaaahhhhhhh Utilitarians FTW!

---

by: hanyo66 @ Tue Mar 04, 2008 at 08:59:48 AM CST