Cornell Belcher. (Photo: Benjamin C Tankersley/For The Washington Post via Getty Images)
In the final days leading up to Election Day, new polls are being released with almost frightening regularity. But no matter how frequently Quinnipiac, Public Policy Polling, Gallup and the others ask voters if they support President Obama or back Mitt Romney, which man cares more about the middle class and whether they're better off now than they were four years ago, one thing almost never changes: the extreme tightness of the race.
That simple fact has cast an aura of suspense over the campaign, leading pundits and other political observers to speculate endlessly about which candidate should be most worried. It also begs the question: in the end, do polls really count?
David Bositis, senior political analyst at the Joint Center for Political and Economic Studies, believes polls matter because they are measurements of what's going on. But Democratic strategist Cornell Belcher, president of Brilliant Corners Research & Strategies, warns against voters reading too much into them. The conventional wisdom throughout the campaign was that the contest between Obama and Romney would be close.
"Polling is both a science and an art. The science of it is frankly the easiest part; the art is a little bit more difficult because you're constructing models based on a number of things that you have to sort of use your best guesstimate at," said Belcher, who conducted a summer poll of African-American voters in key states commissioned by BET Networks.
Pointing back to 2008, Belcher said, poll numbers weren't entirely reflective of the race, especially during the early primary process, in part because in the end 11 percent of the electorate were new voters who'd not been counted among likely voters because they'd never cast ballots before.
"So if you're constructing a typical likely voting model, it would have been off because it didn't take into account the dynamics of the 2008 election. That's the art of it," he said.
Complicating matters is that polling houses use different methodologies, making it difficult to compare who's right or wrong. In addition, they try do it "quick and on the cheap," which means they're "probably also going to be lean on younger votes, particularly those who are cell phone only," Belcher explained. African-Americans and other minorities also tend to be underrepresented.
"If I were to sort of encapsulate all of them, regardless of their models, they're all showing a fairly tight race with Obama up two or three points, which is basically a toss-up race," he said.
But Belcher does not think the election will be the nail biter that many are anticipating. Early voting raw numbers so far are exceeding 2008, which Belcher believes casts aspersions on the idea — and polls — that say Democrats are less enthusiastic in 2012.
"We're lucky to get half of the eligible voters to turn out and if only the most enthusiastic did, we'd be lucky to get a third of eligible voters to turn out," Belcher said.
Bositis says that poll numbers won't influence the way people vote and Belcher tends to agree.
"In the end, it's the turnout and ground operations game that sort of breaks this thing open," he said.
BET Politics - Your source for the latest news, photos and videos illuminating key issues and personalities in African-American political life, plus commentary from some of our liveliest voices. Click here to subscribe to our newsletter.