NC and IN results will depend on turnout of key groups

In this trench war, Obama and Clinton have both greatly benefited from the slightest inroad in the other candidate's voting base; in fact, an argument can be made that Obama became the overwhelming favorite simply because of a few percentage up-tick among blue-collar white voters in Wisconsin and in the Potomac primaries. Yet, there has been enough stability in a group' voting behavior that the primary factor has been what proportion of the electorate each group represents. Both candidates have been focused on soring up turnout among their own base or making inroads small improvements in groups that they know are not favorable to them; gone is the obsessive focus on the undecided voter. It might be a shame that the Democratic primaries have become so dependent on identity politics, but seemingly every contest since January has confirmed that fact.

In particular, the proportion of black voters and the proportion of blue-collar voters have been two crucial factors; in Pennsylvania, the main question heading into the Tuesday vote was whether the blue-collar and Catholic vote would turn out, not whether Clinton would hold to her high margin within that group. Tomorrow, the very same question will matter a great deal in Indiana. In North Carolina, black turnout will be a huge factor. Consider how remarkably stable the African-American has been since the South Carolina primary: It would be a huge surprise if Clinton gets anything lower than 8% and anything higher than 12%. Knowing that it becomes crucial to know just what proportion of the electorate black voters will make up in North Carolina.

The average number pollsters and pundits seem to be betting on is 35%. That number would make it quasi-impossible for Clinton to keep it close -- and it is not even on the high-end of hte spectrum of the possibilities. African-Americans make up 41% of the early vote in North Carolina -- though they tend to make up a bigger share of the early vote than of the day-of vote (for instance in 2004). This spreadsheet makes it obvious how difficult a 65-35 breakdown makes things for Clinton: If she manages to get 13% of the black vote, she would need 70% of the white vote to prevail. Unlike Alabama and Mississippi, North Carolina is not racially polarized enough for this to happen.

To make the race close, Clinton will need (1) heavy white turnout or depressed black turnout while (2) holding her usual share of the white vote. But even then the path to a victory is difficult: If blacks make up 32% of the electorate as they did in 2004 and as SUSA's latest poll projects, Clinton would need more than 68% of the white vote to prevail! To pull off an upset, Clinton thus needs to do what she has not done a single time this year: improve her share of the black vote. Even her hopes of keeping the race within single digit depends on the electorate's racial breakdown: If black African-Americans make up 40% of the electorate, Obama would be sure of pulling a double-digit victory, as Clinton would need about 67% of the white vote.

It is different to make similarly precise calculations in Indiana as the white vote tends to fluctuate a bit more based on educational and class breakdown. But the basic idea remains the same: just as in Pennsylvania, the question will be whether white blue-collar voters go the polls or whether they stay at home, perhaps demoralized by talk of Clinton's low chances to win the nomination or convinced by Obama's ads that Clinton's talk of the gas tax holiday is simply a pandering position.

That Clinton's main focus in Indiana has been this group of voters is hardly debatable. The gas tax holiday proposal is clearly aimed at the middle-class and both Hillary and her husband have employed increasingly populist rhetoric on the trail. Check Bill's latest rant against a New Yorker article he accused of elitism and note the use of pronouns: "That's the kind of thing those people that aren't for us say. You know, they think we're dumber'n we are. I know, cuz I grew up in a place like this, and I figured out that people are just as smart here as anywhere else. But they ain't figured it out yet." Clinton does not need to convince white blue-collar voters to migrate to her side; they have been heavily voting for her for months.

That she is putting such emphasis on their vote confirms that both campaigns are focusing on turning out their base, not on convincing any new voter.

Another key question in Indiana concerns the partisan breakdown: Indiana hosts an open primary and polls confirm that, as in many other states, Clinton is strong among registered Democrats while Obama's hopes rest on independents and registered Republicans. How much these two latter groups turn out tomorrow will go a long way towards determining Obama's potential.