Home | Opinion | Columnists | Mark Mellman

The politics of identity

While politics, like all social activities, often implicates questions of identity, this primary season, chock-full of historic candidacies, affords many more opportunities than usual for identity politics, and for its evil twin — outgroup prejudice.

In casting our votes, we, in part, express our identities — communicating who we are and what we hold dear. More prosaically, this often manifests itself in folks voting for people like them.

Republicans are no strangers to identity politics. Mike Huckabee, a Baptist minister, enjoys the fervent support of fellow Evangelicals, but gets votes from few others.

On the Democratic side, African-Americans have united behind Obama, giving him 82 percent across the Super Tuesday states, though that total masks some variation. Blacks in Massachusetts gave him just 66 percent, compared to 88 percent in Georgia.

White women are also voting their identities. On Super Tuesday, 60 percent voted for Clinton, compared with 36 percent for Obama. Again, states varied, with Clinton winning this segment by a 52-point margin in Tennessee but actually losing white women narrowly in New Mexico.

The dark underside of identity politics also appears to be on display among white men, particularly in the South. Drawn neither to Clinton by gender nor to Obama by race, white men have emerged as a swing bloc in Democratic primaries.

However, there is evidence that race may be keeping at least some out of the Obama camp. Consider this: Obama won white men in Connecticut, capturing 57 percent.

Among that same segment in Alabama he garnered just 27 percent. Obama received 32 percent from Tennessee’s white males, but 55 percent from those in California.

It is easy to see race — out-group identity politics — as the culprit in these gaps. Southern white men seem largely unwilling to join the Obama camp, while their brethren outside the South give him majorities.

But the story may be a bit more complex than simple racial animosity.

First, there is a glaring exception. In Georgia, Obama actually won white men, albeit by a slim margin.

Second, socioeconomic status has an important impact on the choices of white men. In Connecticut, for instance, college-educated white men supported Obama by almost 2 to 1, whereas non-college white men gave Clinton a 7-point margin. In Massachusetts, 58 percent of white college men went for Obama, but 60 percent of non-college white men voted for Clinton.

The same class patterns held in much of the South as well, with 59 percent of Tennessee’s college-educated white men supporting Obama and 72 percent of non-college white men voting Clinton. In Georgia, too, Obama won college-educated white men by 21 points, while losing those without a college degree by 31 points.

Part of the difference in the voting behavior of white men in the North and South may result from differing socioeconomic status. In Connecticut, where Obama fared quite well among white men, 62 percent have a college degree, whereas in Alabama, where white men did not support him, only 35 percent are college graduates.

Similarly, in Tennessee, just 34 percent of white male Democrats sport college diplomas. However, Georgia, the anomalous case, looks more like Connecticut in this respect — 60 percent have a college education.

Identity politics may also be evident in the generational divide among white voters. Obama’s daughters still don Halloween costumes, while Clinton’s holds a high-powered job. Older citizens more readily identify with Clinton, younger voters with Obama — and that is how they vote. Among whites under 30, Clinton garnered just 39 percent, while among those over 60 she won 60 percent.

With so many voters expressing their identities through their ballots, the nominee will need to forge a broader identification with larger swaths of the electorate.

Mellman is president of The Mellman Group and has worked for Democratic candidates and causes since 1982. Current clients include the majority leaders of both the House and Senate.