Friday, September 08, 2006

Voting and Violence

Prof.Levy's History of Economics classes have been fun. I have had other history classes before but this is the best history class I have ever had. We are doing the economics and public choice of Homer's Iliad right now. Yesterday, we talked about how the pre-history of the gods was violent and bloody, and how Earth and Olympus were common to all gods and the rest of it was chosen by lot. The distribution was unequal so that it remains 'fun' for the gods. What I came back home with was the following idea, which is so obvious now in retrospect. If every vote counts and if there is a bimodal distribution then the population size matters a lot and both groups know that by killing one or a few people in the opposite faction they can win. Thus there is inherent instability, violence and lots of killing.

I can take this idea home to the small town and village politics in India. The way the political system works (to my limited knowledge) is that there are local representatives in the state assembly and parliament. In many cases there are two strong contenders and others vying for the same constituency. The contenders can try to persuade through fiery speeches or bribe the population to vote in their favour. However, there is no assurance in the absence of threats and coercion that individuals will vote the way they said, since ballots are secret. So kill a bunch of people in the opposing camp, that increases your chance of winning in two ways, one through the number of people capable of voting for you, and the second the killing sends a signal to the rest of the population that they should vote for you if they value their life.

Hmm! That does not seem like a bad idea at all. Especially since violence does escalate during elections in India. By the same logic booth capturing also makes sense. If you know a certain locality is the strong hold of the oppositon and all those votes matter then just capture the booth and ensure that your candidate wins. It would be a cool statistic to find the correlation between number of election violence deaths and margin of victory.

No comments: