Are Political Polls An Exact Science?
Posted November 5, 2004
RALEIGH, N.C. — For months, the media used polls to predict how candidates would fare Election Day, so how accurate were they?
A WRAL poll, conducted by Mason-Dixon, gave Gov. Mike Easley a 16-point lead. The Voters' Voice poll put together for WRAL,
The News & Observer
and UNC Radio found Easley 13 points ahead. Ballantine picked up a lot of the undecided votes Election Day, but Easley still won by 12 points.
Both polls had the U.S. Senate race neck-and-neck. In the Voters' Voice poll, Richard Burr went ahead by a point with a handful of voters still undecided. They mostly went to Burr for a victory on Tuesday.
In the race for president, the polls had the Bush/Cheney ticket up by nine and six points. On Election Day, the Republican ticket won by 13 points in North Carolina.
"What it says about your polls and polls in general is that it's become a pretty exact science," Republican strategist Tom Fetzer said.
However, others disagree with that assessment.
"Gov. Easley and others don't think they are reliable at all," said Democratic consultant Mac McCorkle, who helped with Easley's campaign.
McCorkle likes to watch trends instead. He tries not to mistake a snapshot for reality on Election Day.
"You can't use it as if I'm going to win. We've seen people make up 15-point deficits in two weeks," he said.
Fetzer and McCorkle both said the pre-election polls should not be confused with exit polls. They said exit polls have lost credibility in the last few elections.
Overall, Mason-Dixon polls did well nationwide. Mason-Dixon conducted polls for the presidential race in 24 states. They were correct in every state except one. Minnesota broke from the prediction because the undecideds went for John Kerry.