PolySciFi Blog

Sunday, November 05, 2006


My 2 cents on voter rationality

I often seen it asserted that it is irrational to vote as the odds of your one vote swaying an election is approximately infintessimal. For example, that argument is repeated on volokh today (though as part of arguing that it is often rational to vote).

I however find the argument unpersuasive as I commented there:

Giving (denying) your candidate (opponent) a mandate is also a reason to vote
even if the actual outcome is a foregone conclusion. By increasing (decreasing)
the winning difference you enhance (limit) the effectiveness of your candidate
(opponent) even if your vote isn't the deciding vote. Sure it's another epsilon
in the calculation, but it's an epsilon that occurs with certainty.
Basically, in modern democracies the effectiveness of an office holder is positively correlated with their level of democratic support which is most accurately measured by an election. Thus if you support what you believe to be a candidate's policies then you enhance the chances of those policies being implemented by giving them the largest possible election mandate. Conversely, if you oppose a candidate's policies you can diminish the chances of those policies being implemented by voting for their opponent.

If you know how much 15 minutes are worth to you, how much the difference in the candidates' policies are worth to you, and have a handle on how much you'll contribute to a mandate effect, then it's pretty straightforward calcuation for homo economicus. The only difficulty lies in estimating your contribution to (or against) a mandate which alas is not exactly a rigorous process.

Side notes:
As I believe myself to be rational, I cast an absentee ballot before this trip to India. If anyone is curious how I voted, this catallarchy post about sums it up except I'm more hawkish than the poster as I prefer to nip problems in the bud and would've voted for the Libertarian candidate if one had been running. The only time when that hasn't happened was 2004 when a) Badnarik was a nut b) I really really wanted to cast a ballot against John (the Winter Soldier) Kerry c) I had high hopes of Bush pushing through personal accounts for social security.

I also exchange emails with my Congressman & Senators to tell them what I like and don't like (important to do both) in the belief that this is even more effective than voting. I discourage the rest of you from doing likewise so I can maximize effectiveness for my emails. [ed - But by saying this aren't this aren't you really encouraging what polyscifi readers may remain to correspond with their representatives thereby diluting the effectiveness or your remails? You're right! That sneaky little superego has gone and snuck another one by my ego...]

However, I do agree with Mankiw that if you don't think you know enough about the candidates, then you probably shouldn't vote. In light of the preceding discussion, if you have no basis to predict the likely policies, then your voting is little better than noise. Some would say patriotic noise, but noise nonetheless


This page is powered by Blogger. Isn't yours?