Exit Poll Thoughts
As I explained a couple of posts back, the main reason to think there was fraud in the election was that so many exit polls were discordant with the actual vote counts which favored Bush. One study suggested that the odds of this being by chance are 1:50,000.
Something clearly is wrong.
One possibility is that there was outright fraud that awarded Bush extra votes.
Another possibility is that someone tampered with the exit poll data, just to screw with people. This seems unlikely since there is no clear reason why someone would do this.
The last possibility to explain the discordance is that the exit poll data were analyzed improperly. Of course this was their excuse for what happened in Florida in 2000 with Gore, and I never quite bought it. I'm sure they'll eventually come out and give this excuse this time. The idea is that they use certain models, for how people vote, to plug their exit poll data into, since they really only sample a small percenatge of people voting. For instance, they may have a model for how many rural voters are going to vote versus how many urban voters vote. And let's say that the rural vote came out in much stronger numbers than expected and their weighting of types of voters didn't account for this. This would mean that they would underestimate Bush votes, because rural voters went 65-70% for Bush. (Urban voters went 65-70% Kerry.)
All this would be settled, of course, if the exit poll analyzers simply told us what their numbers were and what sort of models they used to interpret the data. It seems like there is enough importance to this election that they could share some of their (what they probably feel is) proprietary information. Transparency would alleviate a lot of worries here, and when people don't talk about these things-- it makes people worry about a cover-up.
So there COULD be a reasonable explanation for the discordant exit poll data and it may simply mean that the Republicans were better at getting out the vote than expected. We do have to confront this possibility, since it is hard to see multiple counties in a state willfully producing fraudulent votes.
I'm not ruling out some fraud-- clearly there were voters who were discriminated against and e-voting machines simply can't be trusted. It's just not clear how they would have rigged massive fraud.
Something clearly is wrong.
One possibility is that there was outright fraud that awarded Bush extra votes.
Another possibility is that someone tampered with the exit poll data, just to screw with people. This seems unlikely since there is no clear reason why someone would do this.
The last possibility to explain the discordance is that the exit poll data were analyzed improperly. Of course this was their excuse for what happened in Florida in 2000 with Gore, and I never quite bought it. I'm sure they'll eventually come out and give this excuse this time. The idea is that they use certain models, for how people vote, to plug their exit poll data into, since they really only sample a small percenatge of people voting. For instance, they may have a model for how many rural voters are going to vote versus how many urban voters vote. And let's say that the rural vote came out in much stronger numbers than expected and their weighting of types of voters didn't account for this. This would mean that they would underestimate Bush votes, because rural voters went 65-70% for Bush. (Urban voters went 65-70% Kerry.)
All this would be settled, of course, if the exit poll analyzers simply told us what their numbers were and what sort of models they used to interpret the data. It seems like there is enough importance to this election that they could share some of their (what they probably feel is) proprietary information. Transparency would alleviate a lot of worries here, and when people don't talk about these things-- it makes people worry about a cover-up.
So there COULD be a reasonable explanation for the discordant exit poll data and it may simply mean that the Republicans were better at getting out the vote than expected. We do have to confront this possibility, since it is hard to see multiple counties in a state willfully producing fraudulent votes.
I'm not ruling out some fraud-- clearly there were voters who were discriminated against and e-voting machines simply can't be trusted. It's just not clear how they would have rigged massive fraud.
posted by spooked at 11:14 PM
0 Comments:
Post a Comment
<< Home