Sunday, November 4, 2012

The Polls Are Now Unequivocal

There is just no way out of it now.

Either a lot of pollsters are going to have a lot of egg on their face or Barack Obama will be re-elected.

Obama's lead in the national polls ( model) is now at its largest margin (Obama +1.2) since October 4, the day after the first debate. The effect of the first debate has generally been overstated as Obama's lead had started to decline before that and he had his largest margin in the Pollster model on September 21 (Obama +4.3). But the margin Obama has now is slim but real.

10 different national polls were released on Sunday and there is almost no variance. 3 have the candidates tied, 7 have Obama with a tiny lead from 1 to 3 points. Rasmussen now has the candidates tied nationally. Only Gallup has not released their poll yet though I'm not sure anything they release has much credibility at this point.

In the all-important state of Ohio, Obama's lead is now 3.2 points in the model. This is his largest lead since (say it all together now) October 4. 11 polls have been released in Ohio in the last few days. Only Rasmussen has it tied. 10 others have Obama winning by between 2 and 8 points. Tonight's PPP poll has Obama up by 5 points.

How about Romney's last-minute Pennsylvania gamble? Pollster has Obama up by 5.6 points. But Pennsylvania has not been polled as extensively as some other (actual) swing states. What if Romney pulls off a miracle there. Can he win?

Probably not. It has flown under the radar a bit but Virginia has been trending to Obama recently. His lead in the Pollster model is now 1.1 points and this on the heels of 6 polls released there in the last few days. Those polls ALL have Obama up by 1 to 6 points. If Obama wins Ohio and Virginia, he can lose Florida, Pennsylvania, and hell, let's give Romney Colorado too. It is still an Obama victory.

Bottom line: Romney needs the polls to be wrong. Could they be wrong? Yes. All of them? Not likely but yeah. It is possible. But pollsters are in the business of being right about this stuff. A Romney win now would represent polling failure of a widespread nature we've never seen. Pollsters have gotten individual states wrong at times in the past. But a whole bunch of states? It hasn't happened in the modern era.

No comments: