Gallup releases the demographics of this year's likely voters, and there's reason for both Chicago and Boston to cheer.
On balance, though, there's one number that looks particularly good for Romney.
In 2008's final Gallup poll, 39% self-identified as Democrats, while 29% self-identified as Republicans. That's a +10% advantage for Democrats.
In this month's Gallup poll, 36% of self-identified as Republicans, while 35% of likely voters self-identified as Democrats. That's a +1% GOP advantage.
So, as Ari Fleischer points out, the likely voter electorate has shifted 11% in the GOP's directions since 2008.
That's very good news for Romney.
The good news for Obama is that the '12 electorate, otherwise, looks very similar to the '08 electorate with similar numbers of minorities, young people, and women (i.e. the Obama coalition) expected to turn out.
Thus, demographically, things look good for Obama.
But... the reason why Romney's good news might offset Obama's is that whites seem to be trending toward Romney much more significantly than they did toward McCain.
That means that Romney can get big gains even if the demographic composition of the overall electorate remains the same -- and that, indeed, seems to be what's happening.
.... given the relatively similar demographic composition of the 2012 and 2008 electorates, the election's outcome may hinge more on how groups vote rather than to what extent they will vote. And most groups are currently less likely to support Obama now than they were in 2008. However, Obama's seven-point margin of victory in the 2008 election leaves him considerable breathing room to lose electoral support yet still win the election.
At this point, though, Gallup Daily tracking of likely voter preferences suggests Obama has lost more support than he could afford to, given his current 50% to 47% deficit to his Republican challenger, Mitt Romney. To close that gap in the final weeks of the campaign, Obama would need to have subgroups favorable to him, such as blacks or young adults, turn out at rates that match or exceed those of groups less favorable to him, or to increase his support among key subgroups even if their turnout remains the same.