Saturday, November 06, 2004

Some More Questions



I am still digesting the election results, though now the process has shifted largely to the thinking part of me. And what I get is more and more questions.

Consider the information I had right before the elections: The most recent polls were showing a very close presidential race, and many were predicting Kerry slightly ahead of Bush. The news about new voter recruitment quite strongly implied that Democrats were being far more successful in this than Republicans. The exit polls that were leaked during the election day showed a very similar state of affairs.

Then the actual results were quite different. I have been told in various explanations in the media that the Republican effort of getting the vote out just happened below the media's radar. My question: How is this possible? How could the media not pay attention to the Republican vote effort? I am not blaming the media here, by the way, I am only trying to understand what happened.

Next the exit polls failed to reflect the actual results. Traditionally, exit polls have been pretty good at predicting the final results. Why were they so much off this time? Why were they all wrong in the same direction, i.e. showing too many Kerry voters? I would think that if they were off due to the fact that a sample can be unrepresentative by fluke, at least a few sets of results should have been wrong in the other direction, i.e., showing too much support for Bush.

I have also read that the early exit polls were wrong while the later ones were more correct. At the same time, I have read that the later exit polls available on the net were adjusted by the actual vote counts. This would seem to be a very bad thing if it is actually true. What are the facts here? Also, is it true that the exit polls were good predictors of the final results in paper ballot areas but not in the machine voting areas? And if this is true, why?

Finally, if the exit polls are not to be relied on, why are they relied on when it comes to the argument that moral values was the most important reason voters went for Bush? After all, if the exit polls are wrong, they can't be used for some arguments and not for other arguments. If the percentages in the exit polls were unreliable, do we really know what the most frequent argument for voting the Bush-ticket might be?

Questions. So many questions, so little time.