October 21, 2008

University of Alabama at Birmingham (UAB) Professor Larry Powell, Ph.D., discusses a phenomenon known as the Bradley Effect and why predicting this year's presidential race may be harder than ever:

"The Bradley Effect is a theory or phenomenon that says that African-American candidates tend to poll higher than they will actually get in voter support," Powell said.

The term Bradley Effect comes from an incident involving former mayor of Los Angeles Tom Bradley, an African-American, who was defeated by a white Republican, George Deukmejian, for governor in 1982 even though the polls had shown that Bradley was leading by a strong margin.

It was attributed to the fact that people lied to the pollsters, saying that they were going to vote for the black candidate but didn't, Powell said. "Since then, some studies have said the [Bradley Effect] theory is not holding up anymore," Powell said. "Some are still questioning it. One of the problems is that we don't have enough data to make a decision on it. Some polls are just wrong."

"Another phenomenon that has been occurring over the last 10 years is that Democrats have been polling less than they have been actually getting [in votes]," Powell said.

"In Alabama in the 2002 election between Bob Riley and Gov. Don Siegelman, Riley went into the last stage of that election with a six- to eight-point lead in every poll, but he won by less than 3,000 votes. What happened was that the Democrat ended up doing better than the polls indicated. So in the presidential election we don't know which phenomenon is going to occur."

Powell is a professor in the UAB Department of Communication Studies. He is an author and a renowned expert on political communication. His books include "Political Campaign Communication: Inside and Out."