After what may have been Romney’s best polling day of the year, Obama fared much better over the weekend.
Heading into the weekend, only YouGov/Economist and RAND showed Obama leading among likely voters. By this morning, those two methodologically distinct surveys were joined by four more traditional polls, including two surveys that showed Romney leading late last week. In both instances, Obama's gains exceeded a net-4 points over the last few days. Perhaps most surprisingly, the ABC/Washington Post poll showed Obama gaining a point compared to their pre-debate survey, now holding a 49-46 lead. On the other hand, Obama did not make any gains in Gallup, even though the survey chopped off the initial few days following the first presidential debate. For Democrats hoping for signs that Romney's bounce was fading, the Gallup numbers are a troubling sign. Nonetheless, Obama's leads in national surveys represented an improvement over last week.
But the state polls continued to indicate that Obama was faring relatively well in the battleground states. PPP showed Obama maintaining a 5-point lead in the critical state of Ohio, adding its name to the list of pollsters to show Romney making negligible gains in the Buckeye State. Interestingly, none of the Ohio polls show Romney gaining more than 2 points compared to a post-DNC counterpart, raising the possibility that the tight race shown by the averages might be the product of house effects more than a large shift. Gravis usually provides Romney friendly results, but even Gravis found Obama leading by 2 points in Colorado, a state where Romney has performed relatively well in post-debate polls.
Compared to previous results pointing toward a larger Romney lead, Romney's one point lead in a Florida PPP poll wasn't necessarily good news for Boston. But it continued a trend of polls showing Romney performing well among Florida's Latino population. PPP and a recent poll of Florida Hispanics showed Obama with only a very modest lead among Florida Latinos, while recent Mason Dixon and Marist polls actually showed Romney in the lead. On this basis, the Guardian's Harry Enten suggested that Obama might have a Cuban problem. Since Obama appears to be performing well among Latino's nationally, Romney's unique strength among Florida's Latinos might be explained by Florida's unique, large, and relatively conservative cohort of Cubans. As far as I can recall, Obama tended to fare better among Florida's Latino population and these Latino subsamples have relatively high margins of error, so it's worth waiting on more data before making firm assessments. Even so, it's clearly a pattern worth following.
In an informative change of pace, much of this weekend's polling action occurred outside of the mutually contested battlegrounds in states. In states generally presumed to tilt toward Romney, like Georgia, Arizona, and North Carolina, Obama held within three points of his ’08 tallies. Most surprising was an lead in Arizona, where most believe that Romney is safe. From a demographic perspective, Obama’s resilience in these states is not surprising. Obama has largely maintained an overwhelming share of non-white voters, and Democrats are extremely dependent on the support of minorities in these three states, especially the two in the southeast. As mentioned here for months, a state like North Carolina figures more prominently in a close national election than many assume. Similarly, a poll showed Romney only down by four points in Minnesota--a predictable result if Romney's ahead nationally or making gains among white voters.