How White Will the Electorate Be?

Probably not quite this white.
Probably not quite this white. Photo: Christopher Anderson/Christopher Anderson/ Magnum for New York Magazine

Last month, Emory University political scientist Alan Abramowitz noticed something unusual. Most of the polling about the presidential race found President Obama leading Mitt Romney, by mid-to-high single digits. But the Gallup poll had Romney ahead. Polls, like everything else remotely affecting the presidential race, have come to be the objects of intense debate, each studied and scrutinized for its differences, but Gallup’s holds particular weight among horse race watchers: It runs a tracking survey that produces new findings every day. Also, because it has such a long history, many of the many articles comparing Obama’s poll standing with a predecessor — is he like Reagan? Carter? — rely on it.

Abramowitz found that Gallup’s results were mainly attributable to different assumptions about the demographic composition of the electorate. Among the stuff that white people like, of course, is the Republican Party, and so the share of total voters who are white people matters a great deal. Gallup was projecting that whites would account for about 77 percent of ballots cast, a higher figure than other surveys, which pegged the number somewhere in the lower seventies. Obama’s political adviser David Axelrod went so far as to accuse Gallup of “methodological problems.” Them’s fightin’ words, at least within the small world of polling obsessives.

Who’s right? That’s impossible to say, though we can guess. But what makes the Gallup mini-controversy important is not its value in the daily political spin wars. It’s important because the same reason Gallup’s surveys may be missing nonwhite voters (which will have no bearing on the actual outcome) is the same reason Obama’s field workers face a challenge in getting them to the polls (something that does matter, potentially a great deal).

The economic dislocations of the Great Recession have undone much of the organizing work that Democrats performed in 2008. Not long ago I spoke with a union leader who told me that a huge share of her members were no longer registered at their previous address.  The bulk of the union’s political work was simply finding them. Many had moved or were living in somebody else’s home. Obama needs those voters, without whom a vital and favorable dynamic may tilt against him. The white share of the electorate has been falling steadily for two decades, from 87 percent in 1992 to 83 percent in 1996 to 81 percent in 2000, 77 percent in 2004, and 74 percent four years ago. Most of that decline came from the growth of Latino and Asian-American voters, though Obama’s 2008 election also benefited from an unusual burst of African-American voting. Given that Obama enjoyed a seven percentage point cushion in 2008, he could again lose a large share of the white vote and still eke out a win. The number-crunching blog electionate calculates that Obama, who won 43 percent of the white vote last time, could still win with 38 percent of the white vote — as long as the white share of the electorate does not increase.

It’s not hard to imagine, though, how just that might happen. Republicans are more enthusiastic this time around, and Democrats less so — and then there are the aforementioned dislocated Democrats who might be MIA. This would explain why Obama’s campaign has devoted the vast bulk of its resources to turning out its base. It would also explain why polls have produced such divergent shares of the nonwhite vote. The pollsters’ challenge is to guess how many nonwhite voters will show up. Obama’s is twofold: First to make sure it has found them, and then get them to do it.