I’m posting at the Economist’s Democracy In America website today, discussing the post-debate fallout that has seen President Obama’s lead in national tracking polls completely dissipate. Today’s Gallup poll of likely voters shows the race tied at 48% a piece. Rasmussen has Romney up by 1%, 48%-47%, as does the RCP composite poll, which has had Romney ahead the last two days for the first time since early summer. In my Economist blog post, I argue that we shouldn’t be surprised by this, given that the aggregate forecast from 13 political science forecast models essentially call this race a dead heat. Three of the forecasts have Obama winning the popular vote, five more have him ahead, but with the race too close to call, and five have Romney winning outright. The mean prediction of the two-party vote is 50.3% for Obama, and the median is 50.6% – both well within the polls’ margin of error. So, given the fundamentals, the question was always why Obama had managed to outperform the fundamentals for as long as he did. I try to provide an answer in the blog post, but the short response is that he was more effective at framing the narrative in a way that negated Mitt’s strengths and masked his own weaknesses. However, there is some evidence that the race was tightening before the debate, and that Obama’s relatively poor debate performance coupled with Mitt’s strong one served primarily as a focal point that accelerated a process that was already underway.
Note also, consistent with what I have been arguing for some time, that the battleground states also tightened considerably in tandem with the national polls. This is to be expected and it is a reminder that these states do not operate in a national vacuum, but instead are influenced by broader factors affecting the race. Put another way, I have never bought into the argument that the President is protected by an Electoral College firewall.
My other point, however, is that this race is far from over; Romney has gained ground, but I think it fair to say that he has not pulled into a lead. Indeed, I expect the race, now that it has recalibrated to a point much more consistent with our forecast models, to stay quite close for the remaining 28 days. This is also a good time to remind everyone not to overreact to any single poll. Looked at in the aggregate, however, they will become an increasingly accurate barometer of the state of the race as we head to Election Day. In the meantime, strap yourself in – it’s likely to be one helluva finish!