In recent Aussie elections, we have been seeing increased inaccuracies and the subsequent irrelevance of professional polls. It seems the pollsters are getting it right less often. At least the weather bureau seems to be improving their forecasts!
Some analysis of the inconsistent political forecasting is offerred by Andrew Leigh who suggests that the pundits are at least as good, if not a better guide, than the pollsters.
The looming USA Presidential election is set to follow the precedent. The varying polls around the nation and from each state offer no consistency and have Bush and Kerry at varying levels of win or lose.
Take a look at the three-way tussle between Bush-Kerry-Nader since the beginning of September, we see Bush consistently leading by as much as 16 points. He only loses to Kerry on four ocassions, and only by a maximum of 4 points with eight of the 62 polls suggesting a tie.
If we forget Nader (and 99% of USers do) the head-to-head counts still suggest a similar story. Of the 43 polls representing the same time frame, 4 suggest a tie and only 4 pick Kerry to win by 3 points or less. Bush is a clear leader with a lead of up to 13 points.
The conclusion is clear:
(by a narrow margin or a landslide!)
To pick the winner might be easy enough, but to ascertain the margin we may as well throw darts. My cynicism gives the odds to Bush with a significant 4-5% majority. I say this because a close election could see a repeat of 2000. The USA cannot afford to have any doubt as to the winner of this election and the GOP is not likely to leave the way open for another legal challenge.
[+/-] show/hide this post
No comments:
Post a Comment