HORSERACE - Poll Update 10/31/04
NOTE: Several of you have written to point out that one cannot really reconcile these polls in a statistically perfect way. This is due to the different methodologies and weights that each poll uses. I agree with you, and actually have mentioned these problems before. When I started this project, I mentioned that this presumes that there are no such problems with this analysis. As this is not the case, I concluded that these figures should only be used to give the reader a sense of the state of the race. This method is better, much better, than the Realclearpolitics method, but it is not perfect.
With that said, there were a few new reputable polls that came out today. Here are the new averages.
Florida
Bush: 48.23%
Kerry: 46.70%
MOE: +/- 1.3%
(Respondents: 5,220; Polls Used: Gallup, Quinnipiac, LA Times, NY Times, Mason Dixon, Strategic Advantage, Insider Advantage)
Given these results, we can be 94.18% confident that Bush presently has a lead.
Ohio
Bush: 48.44%
Kerry: 48.55%
MOE: +/- 1.1%
(Respondents: 7,502; Polls Used: Gallup, Columbus Dispatch, Cleveland Plain-Dealer, LA Times, Strategic Vision, Mason Dixon)
Given these results, we can be 55.17% confident that Kerry presently has a lead.
Iowa
Bush: 47.90%
Kerry: 46.40%
MOE: +/- 1.5%
(Respondents: 3,481; Polls Used: Mason Dixon, Des Moines Register, Strategic Vision, Gallup, Research 2000)
Given these results, we can be 90.99% confident that Bush presently has a lead.
Minnesota
Bush: 44.79%
Kerry: 48.15%
MOE: +/- 1.6%
(Respondents: 3,850; Polls Used: Minneapolis Star-Tribune; St. Paul Pioneer Press; St. Cloud State University; Humphrey Institute; Strategic Vision)
Given these results, we can be 99.88% confident that Kerry presently has a lead.
Michigan
Bush: 44.51%
Kerry: 47.49%
MOE: +/- 1.9%
(Respondents: 2,425; Polls Used: Mason Dixon, Detroit News, Research 2000, Strategic Vision)
Based on these results, we can be 98.12% confident that Kerry presently has a lead.
Pennsylvania
Bush: 47.68%
Kerry: 47.85%
MOE: +/- 1.4%
(Respondents: 5,373; Polls Used: West Chester University; Gallup; Quinnipiac; LA Times; Temple University; Mason Dixon)
Based on these results, we can be 57.53% confident that Kerry presently has a lead.
Wisconsin
Bush: 49.30%
Kerry: 45.50%
MOE: +/- 1.8%
(Respondents: 3,090; Polls Used: Strategic Vision, Badger Poll, Mason Dixon, Gallup)
Based on these results, we can be 99.86% confident that Bush presently has a lead.
New Mexico
Bush: 49.17%
Kerry: 47.00%
MOE: +/- 2.4%
(Respondents: 1,625; Polls Used: Albuquerque Journal, Mason Dixon)
Based on these results, we can be 97.32% confident that Bush has a lead.
Nationwide
Bush: 49.17%
Kerry: 47.00%
MOE: +/- 0.9%
(Respondents: 11,532; Polls Used: Gallup, NBC News, CBS News, Battleground, Fox News, ABC News, Pew, Newsweek, LA Times)
Based on these results, we can be 99.9997% confident that Bush presently has a lead.
Analysis: I know that many of you were undoubtedly distressed by Gallup's numbers tonight. However, it is important to view them in context. Even with Gallup's results, Bush still definitely has a lead nationwide, according to my method. His average is also at the point at which it would be impossible for Kerry to beat him in the popular vote.
As for the states, Mason Dixon and Gallup gave us almost unanimously different results (they only agreed on IA). I am honestly quite surprised by this. As Gallup has larger samples, each statistic skews toward Gallup. What we can say with at least 90% confidence is that Bush is winning in Florida, New Mexico, Iowa and Wisconsin. Winning those states and losing OH and NH would put him at 276 EVs.
It is also important to note how close Pennsylvania has become. Kerry's numbers have definitely trended down in the last seven days.
As for OH, let me just say that it is a virtual tie right now. I would be more than willing to take a virtual tie into Ohio -- I like BC04's odds in that situation. It is a solidly Republican state, and BC04 is definitely ready to do battle on the ground.
Generally, given Bush's strong position in IA, WI and NM -- Kerry would have to run the gamut of the big states. He'd have to win OH, PA and FL. If Bush wins any of these, the game is over. Given these polls, Kerry has a 1.85% chance of doing that -- which is to say, he has virtually no chance.
All of this strikes me as very consistent with a difference between the candidates of 2%-3%. If Bush wins the election by 2.17%, which is what my average currently shows, we would expect him to get between 276-300 Electoral Votes. In other words, we would expect Bush to win 2 of the 3 "Big 3" states. Right now, that looks likely to happen.
Beware the Gallup Poll
That 49-49 result is due to Gallup's decision to allocate the remaining undecided voters based upon a tried-and-true formula -- 9 to 1 for the challenger.
Pew, on the other hand, decided to allocate the undecideds they found equally. This is what Pew had to say: "Pew's final survey suggests that the remaining undecided vote may break only slightly in Kerry's favor."
So, it appears that Pew's data told them something that the CW of undecided voters, i.e. their breaking toward the incumbent, is off. Gallup just did what my 9th grade geometry called a "plug and chug." As Polipundit notes, The New York Times survey indicates the same result as Pew. They found that undecideds do not seem to be breaking for Kerry.
Who is right? I do not know. I am going to cut my loses and put their non-leaned numbers into my poll averages. This puts us at Bush 48.71% and Kerry 46.2% (also factoring in this evening's Battleground Poll update).
The fellas at Emerging Democratic Majority, incidentally, smells a rat with Pew. They are all atwitter that it is (gasp!) "over-sampling" Republicans by 3.6%. "The horrah!" Maybe, but they wound up with the same result as The NY Times poll, which over-samples Democrats.
And if they are over-sampling Republicans, it simply means that they are not weighting by party ID, which in turn means they found more Republicans out there than Democrats. You want to dismiss that out of hand, guys? Do ya? They also note that Democrats have out performed Republicans at the polls for 14 years by 3-5%. Fellas, it ain't gonna happen this year. It just ain't.
They also note with interest that Pew predicted a Bush win in 2000. True true. I, in turn, will note with interest that CBS News/NY Times predicted a Gore win in 2000. |
