Thursday, November 09, 2006

A Thumpin'

What constitutes a thumpin'? According to Pres. Bush today:
Q Thank you, Mr. President. You just described the election results as a "thumping."
THE PRESIDENT: I said the cumulative -- make sure -- who do you write for?
Q The New York Times, Mr. President.
THE PRESIDENT: Oh, yes, that's right. (Laughter.) Let's make sure we get it -- the facts. I said that the elections were close; the cumulative effect --
Q Is a thumping.
THE PRESIDENT: -- thumping. (Laughter.)
I will, for the moment, have to respectfully disagree with the President. My dispute with Mr. President lies in my hypothesis that elections tend towards the mean. That is to say that people like close elections, and when polls spread candidates too far apart, you will generally see a correction of sorts. Generally. That being said I do not believe that given a long enough X(time) the curves will intersect. There seem to be districts which poll very close, but have a very large lean towards one party over the other. That is not to say that a heavily Democrat district will ever swing red, but it might correct itself so as not to thump one candidate over the other.

That being said, the President's remarks seem to indicate a different view on election results, some sort of randomness. Let's say, for the moment, that in the contested races the average victory was +/-2% which we can approximate to an equal likelihood for either candidate to win, given enough events. When the President said that, "Look, this was a close election. If you look at race by race, it was close. The cumulative effect, however, was not too close. It was a thumping." What he must have been saying was that every race was close, but given the number of events, which are in themselves 50/50, we just saw a run of one particular outcome. It's like flipping a coin 1000 times and getting 8 heads in a row; it does not mean that your coin or probability is broken.

But the President conceded in his remarks that the electorate was sending a message, which means that the aggregate effect was related to the individual events. It only appeared close because of the hypothesis I stated above. The President could only be correct if he truly believed the Democrat swing was a random series of events.

Groove pointed out to me that the President might have meant that it was only a small portion of the vote that actually swung the election. 45% are going to vote one way or the other every time, so it is the 10% which actually mattered, and they collectively combined to a large aggregate effect, but that it does not necessarily culminate to a democratic "wave." My answer to this is just that every election is defined by the middle. A thumpin' then must be defined by the context of elections which are fought at the margins. Thus for one party to win the majority of the middle in every election ipso facto constitutes a thumpin'.

Thank you President Bush for using such an awesome word and allowing me to write this post about thumpins'.

No comments: