HORSERACE - State and National Poll Update (Updated 9 PM CST)
NOTE: I have adjusted this data to account for the whole slew of Mason-Dixon polls which reader JB was kind enough to bring to my attention.
I decided that since there were so many screwy polls arriving at our electronic doorstep between now and Tuesday, I owe it to you to sort through them. Below are the unweighted averages of all the polls which I consider to be reputable.
Now, since I am exising polls whose methodology I do not like, I think it only appropriate to let you know which polls I am using. One of these polls is Strategic Vision, which does work for Republican politicans. Realclearpolitics will not use their polls to filter their averages. However, they have a good track record and are generally in line with other state polls. Thus, I see no problem using them. I will include equally reputable Democratic polls, should any come to my attention (note that I do not think Democracy Corps is sufficiently reputable).
One technical note, which might surprise you: I have decided to incorporate the Newsweek poll on an ad hoc basis. The reason for this is that Newsweek is, bar none, the best at publishing its internals. They publish everything. So, I have decided to include their polls, pending a review. This one looks pretty kosher, though I think they slightly under-sampled Democrats and over-sampled Independents -- I have included LA Times and ABC News, both of which over-sample Democrats. So everything balances out.
I never had a good a priori reason for excising Newsweek. As my reasons against Newsweek have always been a posteriori, i.e. they just come up with screwy results that nobody is replicating, I have decided that they can be included pending review. Since Newsweek is really, really, really good about releasing its data, and the data seems up-to-snuff this week, this warrants their inclusion. My position on the poll also warmed after I discovered that NBC/WSJ use the same polling outfit, Princeton Associates, for their polls. NBC/WSJ is a fairly decent poll -- so the Newsweek methodology cannot be all that screwed up (inherently at least).
In retrospect, I believe I have been a little too hyperbolic when it comes to Newsweek. My apologies to Eleanor Clift (on this matter, anyway).
Ohio
Bush: 47.93%
Kerry: 46.89%
MOE: 1.4%
(Respondents: 3,511; Polls Used: LA Times, Cleveland Plain-Dealer, Mason-Dixon, Strategic Vision)
Based on these results, we can be 80.51% confident that Bush presently has a lead.
Florida
Bush: 49.13%
Kerry: 45.58%
MOE: 1.4%
(Respondents: 4,683; Polls Used: Insider Advantage, Strategic Vision, LA Times, Gallup, Quinnipiac, NY Times, Mason-Dixon)
Based on these results, we can be 99.88% confident that Bush presently has a lead.
Iowa
Bush: 49.24%
Kerry: 45.98%
MOE: 1.9%
(Respondents: 2,675; Polls Used: Mason-Dixon, Research 2000, Strategic Vision, Gallup)
Based on these results, we can be 99.34% confident that Bush presently has a lead.
Wisconsin
Bush: 47.77%
Kerry: 46.36%
MOE: 1.2%
(Respondents: 1,971; Polls Used: Mason-Dixon, Strategic Vision, Badger Poll)
Based on these results, we can be 81.33% confident that Bush presently has a lead.
Minnesota
Bush: 45.1%
Kerry: 46.5%
MOE: 2.1%
(Respondents: 2,189; Polls Used: Mason-Dixon, Strategic Vision, Humphrey Institute, St. Cloud State University)
Based on these results, we can be 68.79% confident that Kerry presently has a lead.
Pennsylvania
Bush: 47.08%
Kerry: 48.52%
MOE: 1.4%
(Respondents: 4,251; Polls Used: West Chester University Gallup Quinnipiac, LA Times, Temple University)
Based on these results, we can be 92.07% confident that Kerry presently has a lead.
Nationwide
Bush: 49.57%
Kerry: 46.28%
MOE: +/- 0.9%
(Respondents: 8,954; Polls Used: ABC News, Fox News, LA Times, Gallup, Newsweek, Battleground)
Based on these results, we can be 99.9997% confident that Bush presently has a lead.
Present Probability that Bush will win the Electoral College: 96.36% (This is the probability that Bush wins FL and IA and WI or OH. Thus, we can be 96.36% confident that Bush would receive a minimum of either 271 EVs or 281 EVs).
Evaluation: Bush looks to be in very strong shape in IA and FL. At this point, his numbers are inching closer and closer to 50%. His numbers in Ohio are coming back to their levels from mid-October. I believe this is due mostly to bad polls replacing good polls and then being replaced in turn by good polls. He still likely retains a lead in WI, but the Mason-Dixon numbers are somewhat unsettling (though they are mitigated by the MN numbers, which in this post are skewed by an outlying survey from St. Cloud State).
Bush also seems to be inching upward in PA, due to a string of polls calling the race a tie (Quinippiac even has him up), which might result in a big pro-Bush surprise on Election Night. Gallup should release another PA poll sometime before 11-02. That should be the one to look for.
Right now the EV math is looking awfully tough for Kerry. He is definitely behind in FL, IA and, though I do not cover it here, NM. This gives Bush a minimum of 266 EVs. Plus, Bush is likely leading in OH and WI -- and I think Kerry will be unable to hold MN when all is said and done. The word on the ground is that BC04's organization in MN is a sight to behold. The big question on my mind right now is not whether Bush gets to 269, but whether he breaks 300 (which he would do if he carries FL, IA, NM, WI, OH and MN -- that would be 306). |
