10/16/2007 03:28 pm ET Updated May 25, 2011

Obama Won't Win Mississippi and It Doesn't Matter

With Barack Obama claiming Mississippi for himself in the 2008 general election, John Edwards doing the same for Tennessee and North Carolina, and Joe Biden (!) projecting he'll win Kentucky, you'd think the Democrats have a shot at a clean sweep of the South next year. Listening to the Giuliani campaign (always a mistake), you'd also think that a repeat of Reagan vs. Mondale is in the works if he is the nominee. Not surprisingly, the reality is quite different, according to a series of general election match-ups recently conducted by SurveyUSA in 17 states in the South and elsewhere.

The polls do not back Obama's claim to Southern states with large African-American populations, but they do show great strength for Edwards in the South and pretty much everywhere else. On the Republican side, the numbers are sobering for Rudy Giuliani, mostly favorable for Fred Thompson, and an unmitigated disaster for Mitt Romney, especially in the South. The usual caveats apply to early polling.

The bad news for Obama is that his projected general election performance lags far behind Edwards' and even Hillary Clinton's in the South. The worse news is that Florida is one of the states polled, and Obama does as badly there, compared to Clinton, as in Alabama. One happy glimmer for Obama is that in the only other Southern state that will likely matter, Virginia, he does nearly as well as his Democratic rivals when matched up against Giuliani, Thompson and Romney (all three Democrats beat all three Republicans handily or are tied in a state that hasn't voted for a presidential Democrat in over 30 years). As in primary polling, where his best early state is Iowa, Obama is stronger in general election polling in states with small African-American populations such as Washington or Wisconsin.

Edwards, on the other hand, will credibly argue that he can win anywhere this time around despite the fact that his presence on the 2004 ticket did not help win one single Southern state, including his own, North Carolina; he is also far behind Clinton and Obama in polling for next year's South Carolina primary. On average, Edwards beats the Republican candidate in every state surveyed, except for Alabama and Kansas (he is in an average tie in the latter). He also does better than Clinton and Obama in 13 of the 17 states.

Outside of the South, Clinton blows away her Democratic rivals in general election match-ups in states such as New York, Massachusetts and California, but who cares? None of these will be even close in 2008.

More hope for Obama: in most Midwestern and Western states, he does as well or better than Clinton in general election matchups. He is also as competitive as she is in swing states all over the country, with two glaring exceptions: Florida, as noted above, and, perhaps worse, Ohio, where SurveyUSA shows him losing to every one of the three Republicans, including Romney.

On the Republican side, at this point, the race is more about who will lose by the smallest margin. Here, Giuliani's preposterous claim of competitiveness in Democratic states along the coasts is put to a clear lie (Giuliani loses by double-digits on average in New York, California and Massachusetts). Predictably, Thompson does better than Giuliani in the South, but more importantly their general election performances are tied in the Midwest, and Thompson is only slightly behind in swing states, despite a far lower level of recognition nationally.

As for Romney, his strong performance in polling in early primary states has not translated into general election strength: on average, he loses every single one of the 17 states surveyed except for Alabama (although he'd lose to Edwards there), mostly by double digits.

I've called Democratic voters pragmatic for their strong backing of Clinton in primary polls so far, but in truth, the pragmatic (and conventional) choice should be Edwards. That he trails so far behind his two main rivals in most polls may be a strong demonstration of Democrats' yearning for change, even if paradoxically they may once again not be putting up the stronger potential general election candidate. For once, though, it probably won't matter: in a year in which dissatisfaction with Republicans is likely to peak even higher than in 2006, even Obama is showing early strength in swing states and deep into G.O.P. territory such as Virginia or Kansas, where, for instance, he would easily defeat Romney.