Are national polls understating President Obama's prospects for victory?

President Obama is performing ahead of 2008 margins in Jan Brewer's Arizona (Pete Souza/White House)

Early in the week, NBC News (and the Wall Street Journal) released their first national poll in a month. The topline numbers moved very little'whereas Barack Obama's lead over Mitt Romney was four points in May, it was three points in June.

NBC's Mark Murray, however, noted some movement in the nearly dozen states identified by the media outlet as "swing states":

Another place where Obama is running ahead: the swing states.
Among swing-state respondents in the poll ' those living in Colorado, Florida, Iowa, Michigan, Nevada, New Hampshire, New Mexico, North Carolina, Ohio, Pennsylvania, Virginia, and Wisconsin ' Obama leads Romney, 50 to 42 percent.

Also in these swing states, Romney's favorability numbers have dropped, possibly reflecting the toll the negative Obama TV advertisements are having on the former Massachusetts governor in these battlegrounds.

A month ago, Romney's favorable/unfavorable score stood at 34-38 percent nationally and 36-36 percent in the 12 swing states.

But in this latest survey, his national fav/unfav score is 33-39 percent and 30-41 percent in the swing states.

As a matter of statistical relevance, in all candor, this poll did not tell us a whole hell of a lot. The margin of error on a subsample of 12 states is pretty stout, so saying that this eight-point edge for Obama is a BFD (so to speak) would lead the math types in the political pundit class to rise in vociferous protest. And, ultimately, they are right about that.

But the poll does underscore a long-held theory about this election, which I have written on in the past as well as others (I have seen PPP's Tom Jensen note it, as well). The theory is that the national polling numbers, which have been in tossup territory for much of the Spring, actually are overstating the case for Mitt Romney's electability.

This theory is based on a few assumptions:

• Because of minimal campaigning by either candidate, and the general voter malaise apparent in much of the polling to date, the president will still easily carry the blue states he won in 2008, but perhaps by slightly depressed margins.
• Because of minimal campaigning by either candidate, general voter malaise, and terrain that was already hostile to begin with, the president is going to lose the red states that weren't particularly competitive in 2008, and quite possibly by more substantial margins.
• The swing states, meanwhile, and notably, will behave much as they did in 2008. There may be some slight depression in the numbers (again, that voter malaise thing), but because the president and his campaign will put maximum effort into these states, both in defending his record and assaulting Mitt Romney's record, the needle will not move as dramatically here.

If these three premises come to pass, then the net result would be a potentially notable reduction in the performance for the president in the national popular vote, but a considerably smaller impact in his performance in the electoral college. Therefore, while the race appears to be a tossup in the national polling, when the votes are actually tallied, the electoral college will still go decidedly in favor of Barack Obama.

That's the theory. How legitimate is it? Go past the jump in order to find out.