New Pew Poll Gives Obama a Three Point Lead

In my last post, I said that Obama should be favored to win the election, and suggested he has roughly a 65-70 percent chance of winning. My estimate of Romney’s chances was based in large part on his relatively strong showing in national polls, which still had him even with Obama late last week. Today, however, the Pew Research Center came out with a new poll giving Obama 3 point lead nationally, among likely voters.

Obama’s lead in this poll is actually slightly greater than the survey’s margin of error (2.2 percent). And Pew is one of the best polling firms in the business. In combination with the other evidence, such as the battleground state polls, I think this gives Obama an even higher probability of winning than I suggested yesterday, perhaps 80% or even more.

This is still going to be a close election, and will still be one of the rare instances where the outcome is not a foregone conclusion well before election day. But at this point, it’s looking more like 2004 (a close election where one side nonetheless has a clear edge) than 2000 (close to dead even). The polls are still close enough that Romney has a shot, especially if you buy claims that the pollsters’ turnout models are overestimating the number of Democrats who will vote. But his chances are much weaker than it seemed as recently as a week ago. A Romney victory is possible, but at this point would be a pretty substantial upset.

If Obama does win a narrow victory, it’s possible that the effect of the hurricane will be responsible for pushing him over the edge. But I think it’s at least equally likely that things have just reverted to the outcome that could be predicted based on the underlying condition of the economy, which has long pointed to a narrow Obama victory, as I explained in this September post. Romney might have been able to overcome these odds had he been an unusually appealing candidate, or if the Obama administration suffered a major scandal or some other clear setback. But neither has happened, at least not to any great extent.

UPDATE: It’s worth noting that the hurricane and reversion to the historical mean scenarios are not mutually exclusive. It could be that Obama’s numbers have improved recently for a combination of both reasons.

UPDATE #2: Some other polls released today show a dead heat. On balance, a set of national polls where some give Obama the lead and others point to a tie is not good news for Romney, especially when combined with the state polls. That said, it’s possible he has a better chance than I just suggested above. I still think Obama is very much the favorite, however, and would certainly give him at least the 65-70% chance I suggested yesterday. Whether the probability of his winning is 80% or higher is much harder to say. In any event, we’ll soon see.