Wednesday, September 10, 2008

Presidential Polls Are Wrong, Wrong, Wrong

November 1, 2008 Update
Note: to blog surfers and others who recently found this post. This post was written on September 10, 2008, when the polls showed Senator McCain taking over the lead from Senator Obama. The purpose of the post was to point out that the sampling procedures had been changed to favor a more Republican oriented response. This is a tremendous no-no in any poll or sampling procedure. If you are a Republican hoping to read that the current polls which show Obama with anywhere from a three to eight point lead are wrong, I have to disappoint you. While there is a significant difference among the polls, it is likely Obama still has a four to five point lead. There are significant flaws in some of the polls, such as one that assumes that 42% of the voters will be evangelicals, which is more than twice the percentage of evangelicals that voted in the 2004 presidential election. Besides, the election is decided on a state-by-state basis were Obama has more than enough states to give him the necessary 270 electoral votes. I'm not saying that McCain can't win, but rather that it is extremely unlikely that he will do so.


According to the most recent polls (Gallup, CBS, USA Today etc.) Senator McCain has pulled out in front of Senator Obama. But as everything else in this campaign nothing is what it appears to be. Alan Abramowitz, cited in the Huffington post, said, "One reason for the dramatic difference between the two recent CBS polls is that the two samples differed fairly dramatically in terms of partisan composition. The first sample was 35.2% Democratic, 26.2 percent Republicans, and 38.6 percent independent. The second sample was 34.9% Democratic, 31.1% Republican, and 34.0% independent. That's a change from a 9 point Democratic advantage to a 3.8 point Democratic advantage. That alone would probably explain about half of the difference in candidate preferences between the two [CBS] polls."

Simply put, you cannot change the sampling populations from one poll to the next and have any consistency (reliability). Mr.Abramowitz doesn't think the overall spike in Republican sampling among all three polling firms is an accurate reflection of the electorate. He goes on to say, "Given how important polls can be in the coverage of the race, even a slight assist to McCain during a period in which he is exciting the Republican base could help him solidify a new narrative in the race, regardless of the partisan facts on the ground."

The people of this country need accurate information in order to make an informed decision about whom to vote for in the presidential election. If the polls are inaccurate, and the McCain campaign is certainly not accurate, and Obama is spending most of his time responding to "misinformation" from the McCain campaign, then the American electorate falls back to choosing someone they are most comfortable with. That's the McCain strategy. It was the Bush strategy and it got us eight years under what will go down in history as one of the worst presidential administrations. Do we want that again?

No comments: