Excerpts from a Nate Cohen article today on The New Republic.
....Obama campaign officials said they expect the non-white share of the electorate to increase by 2 percentage points to 28 percent of the electorate. Many Republicans are skeptical that the electorate will be as diverse as it was four years ago, let alone more so. After all, Obama’s wide margin of victory was dependent on a historic turnout from young and African American voters, raising the question of where these new minority voters might come from, especially with resurgent GOP enthusiasm.
One obvious source of additional minority voters is the new cohort of 18-22 year olds who will be voting for the first time in 2012. Less than 60 percent of 18-22 year olds are white and in 2008, approximately 62 percent of 18-24 year old voters were white. Even if young voters turnout at a lower rate than they did four years ago, the new wave of 18-22 year old voters should reduce the white share of the electorate by more than one percent even before accounting for—and I apologize if this is crass—the departure of a disproportionately white wave of elderly voters.
Getting additional minority voters beyond the new wave of young voters is more difficult. Low-turnout Latino voters are the obvious source, but most polls suggest that they aren’t as enthusiastic as they were four years ago. And just for good measure, Obama would need to prevent a decline in turnout among those who voted in ’08, which could undo the gains from young voters. After all, Obama could lose plenty of this year's 23-32 year old voters who turned out in large numbers for Obama four years ago. Resurgent Republican enthusiasm can also play a role, since the 2008 exit polls suggest that about two million (predominantly white) Bush voters may have dropped out of the electorate between 2004 and 2008. Their return could cancel-out many of Obama's gains.
But voter registration data suggests that the Obama campaign might just be pulling it off. Only a few states provide a racial breakdown of their registered voters, but several southern states do and the most diverse southern states—Georgia, Florida, and North Carolina—each show the non-white share of registered voters increasing by 1.6 to 2.7 percentage points, with much of the increase coming from Latino voters. Whether these voters turnout on Election Day is hard to say, but it certainly seems that the pool of available voters has become more diverse to an extent roughly in-line with the Obama campaign's expectations.
Even though minority turnout is a crucial question for the election, it’s relatively peripheral to the outcome of Ohio. In 2008, minorities represented 17 percent of Ohio’s electorate compared to 24 percent nationally and 15 percent in 2004. Most of the state’s minorities are African Americans and overall, Ohio’s turnout only increased by 96,000 votes. Of course, Obama would certainly benefit from a higher turnout among minority voters, but this just isn’t a state where Obama’s ability to further increase the non-white share of the electorate is essential to his chances. Realistically, there are more opportunities to increase the non-white share of the Sun Belt states like Florida, North Carolina, or Virginia.
A wave of new young and perhaps latent Latino voters could increase the non-white share of the electorate, but it's hardly assured with polls showing diminished enthusiasm among those two groups. Nonetheless, the pool of registered voters is more diverse than it was four years ago, so a strong turnout could potentially help the president in several key Southeastern states. But at the moment, the Sun Belt states are somewhat peripheral to the Electoral College calculus, mainly because Obama holds a clear lead in Ohio, a state where subtle differences in non-white turnout aren't likely to prove decisive. The fact that a serious substantive disagreement about the race could influence a few key battlegrounds while leaving others unscathed highlights the volatility created by a diverse set of close states.