Election Day

First of all, I did err in my explanation of the Redskins factor yesterday. It's not a Democrat/Republican thing - it's an incumbent/challenger thing. My mistake.

The weather has turned foul again, lots of cold rain. In fact, the other day when it was sunny and nice out looking, it actually turned out to be decently cold. I never did get to the park.

Yesterday I had class. It was very long an tiring. Or, I was just very tired from the weekend. So I didn't do much last night and went to bed early. They're doing construction in the house here, so apparently 9 AM is the time to sand and otherwise make a ruckuss. Fortunately, I had been sleeping for a long time. I feel quite rested today, which is good. I have class today and I'm going to school early to get my study on. Tomorrow I go back to work.

Today is election day. I encourage all those who are able to vote to get to the polls. This looks to be the tightest election in history (including Florida 00). I feel confident that Kerry can pull off the win, but it's close. I read some interesting things about the polls earlier today. Two major polls (Gallup and one other) have disclosed how they see the "undecided" break as well as what percentage of new voters they think will actually vote. See, the polling system is not an art, nor a science. Someone has to make decisions about how many people actually vote - broken down by category (new voters, young people, old people, etc). Both of these major polls has the "new voter" turn out at 7%, which is about 6% too low. In fact, they have new voter turnout trending downward from 2000, something that just flatly ignores the trend in this election to get out the vote. Why does this matter? Because both polls concluded that new voters are splitting for Kerry between 2:1 and 3:1.

The other factor, the "undecideds", was broken by the most unscientific means possible - they split the 4-6% evenly between the two candidates. The historical record (referred to as the incumbent effect or 'incumbent disadvantage') suggests that undecideds break for the challenger at about 3:1, so I don't know how the people at Gallup or elsewhere can justify their numbers.

The point is, of course, that polling in the US is incredibly inaccurate (Gallup predicted that Bush would beat Gore in the popular vote 52-48% in 2000) and that there is great need for reform. My question is, how much does polling effect the electorate? There were arguments made in 00 that announcing results in Florida early depressed turnout in the panhandle because they thought the race was won by Gore. That may have cost Gore the election. (He lost by 527ish votes, the GOP dominated panhandle turned out, the Democrats there didn't - he pretty much lost the Presidency because of media reports saying he won Florida.)

At any rate, today looks to be a fascinating day in American politics, one that I can only hope results in John Kerry seizing the day. (Of course, when Bush consolidates his power with the military, rounds up all those uppity democrats and declares a police state to preserve his hold on power - well, I'll be laughing at you all from across the pond!)