Lindsey Graham won the most watched (and expensive) U.S. Senate race of the year. By a lot. He wasn’t supposed to. Not by a lot, anyway. Not if you paid attention to public opinion polls leading up to the election. Most polls gave him the odds, but not big odds.
Some polls even gave his challenger, Democrat Jamie Harrison, a slight chance to unseat him.
And for about an hour after the polls closed on Nov. 3, the early results really did suggest a squeaker was in the works. A certain reporter even tweeted this:
That tweet withered on the vine within another hour. By the time the (still technically unofficial) numbers got tallied, Graham had won the typically 10-point red state of South Carolina by, well, 10 points.
And although Graham in his victory speech chided that pollsters “have no idea what you’re doing,” the long-serving Republican senator had been visibly shaken by what looked to be a very serious challenge to his job. At the hearings of then-Supreme Court candidate Amy Coney Barrett, Graham asked for help to fend off a highly funded opponent. Each candidate threw $100 million at this election. Everything seemed so even, until it wasn’t.
So what happened with the polls? Again? Why did they so underestimate the chances of a hard-right Republican like they did in 2016, when almost every major poll predicted we would have the first female U.S. president?
In the audio story below, two of South Carolina’s – and the country’s – top polling experts, Dr. Scott Huffmon of the Winthrop Poll at Winthrop University and Dr. Steven V. Miller of the Palmetto Poll at Clemson University, discuss how pollsters ran into familiar problems with conservative underrepresentation, misperceptions about polling itself, the role of news outlets regarding public opinion realities, and why the future isn’t 1975 anymore.
Scott Morgan is the Upstate multimedia reporter for South Carolina Public Radio. Follow Scott on Twitter @ByScottMorgan. And follow South Carolina Public Radio @SCPublicRadio and on Facebook.
