The results of the 2016 election proved to be quite different from what polls predicted leading up to Election Day. WKSU's Kabir Bhatia asked Ryan Claassen, a political scientist at Kent State University, for some of the reasons why the polls were so far off.
One reason is that much of the polling came in before FBI Director James Comey caused the Clinton campaign to relive the e-mail scandal less than two weeks before the election. Other possibilities are linked to social desirability: people told pollsters they weren’t going to vote for Trump, when in fact they planned to.
Claassen has done research into social desirability that supports that view: a number of Republicans either worried about admitting they were going to defect from the party, or worried about admitting they would be voting for Trump.
The entire question of inaccurate polling is linked to the Bradley Effect, according to Claassen, and Hillary Clinton may have been on the receiving end of a gendered version that effect.
It’s named for the 1982 California gubernatorial campaign between Los Angeles Mayor Tom Bradley and Attorney General George Deukmejian. Deukmejian won by about 1.2 percent, despite opinion polls showing a lead for Bradley.
The discrepancy has been dubbed “The Bradley Effect,” referring to a possible tendency by voters to tell pollsters that they are undecided or likely to vote for a black candidate, while actually voting for his opponent.