Friday, December 08, 2006

 

Close elections

In the last few years, it seems as if we've had an unusual number of very close elections. This goes back at least to the 2000 presidential election in Florida, but even in 2006 we've had a number of Senate and House races that were close enough for the outcome to be in doubt days or weeks after election day.

I don't have anything definite to say about this, but I've been wondering whether there's any significance to it, and if so, what the explanation might be. Have there actually been more very close elections lately that should be expected, or have we just noticed them more? If this is statistically significant, what's the explanation? Is somebody cheating just barely enough to win? Or is somebody trying to do so and falling a little short? Or is it a result of gerrymandering, with artificial 50/50 districts being created so that other "safe" districts can have a clear majority? (The latter doesn't apply to Senate or presidential races.) Or are the parties becoming better at calculating just how much money they have to spend on campaigns? After all, if it's going to cost $1 million to win by 50.5%, and $2 million to win by 60%, then that second million is wasted money. Meanwhile, the other side is doing the same thing, which brings game theory into it.

But first, we need to know whether there's anything significant going on in the first place.

Is there a statistician with too much time on his/her hands among my Vast Army of Loyal Readers?

Comments: Post a Comment



<< Home

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