Monday, October 27, 2008

Why The Race is Tighter Than You Think

Iowahawk, premier blogger satirist, sent me over to Tidestemmer, a worthwhile trip. In an article titled "Random Error Doesn't Look Like This," Gene Fama argues what I have noticed in over three decades of watching elections; Republicans almost always do better than the polls indicate. People forget how close the 1976 election turned out to be, 50.1% to 48.0%. That was an incredible result given the albatross of Watergate, the Nixon pardon, "There is no Soviet domination of Eastern Europe" statement, and the MSM portrayal of Ford as a bumbler. (Hey, is there some sort of theme here?)

Look at this graphic on exit polling discrepancies from 2004 and you will notice the consistent anti-Republican bias in result. Why? (It is not that Republicans cheated in every state.) My theory is that pollsters for the MSM are more comfortable polling in locations where there are like minded people and they somehow are self selecting fellow liberals to do exit polling to a small but measurable degree. This is basic to anyone who has studied the need for double blind testing in psychology, the area of science most closely linked to polling.

Gene also discusses the so-called Bradley effect, but turns it on its head. Perhaps, in an atmosphere supercharged by the constant playing of the race card, people who disagree with Obama on the issues, but who don't want to be called racist are saying they will vote for him as well. Because Obama did not seem to have a Bradley effect going in the primaries (he closed well), I tend to buy Gene's explanation. In the primaries, the Democrats all agreed with one and other on the policy issues, hence no effect.

To re-iterate my point from yesterday, just more reason to keep voting and get your fellow freedom-lovers to vote as well.


1 comment:

K T Cat said...

I just saw that Mississippi now has something like 123% of its population registered to vote. The polls mean nothing if they only call each person once.