Jan. 27, 2009

Did people 'lie' about race on November election surveys? No evidence, says CBS News panelist

"Race did play a role in the 2008 presidential election," but not the way most people think, said Kathleen Frankovic '68, director of surveys and a producer for CBS News, at a Cornell panel discussion last week.

"[Race] was only an issue among Democrats and independents early in the election season," said Frankovic. Race became much less of an issue toward the end of the election season, she said. "[Political] party trumped" any negative feelings in relation to voting for a black candidate, she said.

Speaking at the Survey Research Institute's Annual Speaker Series at the ILR Conference Center, Jan. 21, Frankovic said that "there was no evidence of a Bradley effect" in the November election. The term was coined after Tom Bradley, the black mayor of Los Angeles, lost to a white Republican in the 1982 California gubernatorial race, despite being ahead in voter polls going into the election. In the national election last November, Frankovic cited a CBS News survey that found no evidence of white Americans saying they were going to vote for Barack Obama and then doing the opposite.

Frankovic noted, however, that when answering surveys, people tend to boast about their behavior. When CBS News asked respondents in surveys if they, as well as the country, were ready for an African-American president, most people said they were more ready than the country was.

"People tend to up their own status and deflate that of others," said panelist Stephen Ceci, Cornell professor of developmental psychology, and that could be one reason people lie on surveys. For example, in surveys of Cornell psychology students in the 1960s, Ceci found that "students systematically put higher SAT scores on their surveys" than those they actually earned.

Another reason why people may lie on surveys, Ceci said, is that survey questions may "engender a conscious protective or defensive response." People may not want others to know something about them or may be embarrassed by the survey questions.

"It is very difficult to know if someone is lying on a survey, and it should not be readily assumed that lying is occurring," said panelist Norbert Schwarz, a psychology professor at the University of Michigan. "People need time to think and respond to questions asked, and if surveys don't allow this, there will inherently be discrepancies in the survey results."

His research, he said, has shown that survey results shift even when people don't lie or don't know how to lie on the surveys.

"Lying can sometimes be confused with being polite," or people just forgetting, said Schwarz. He pointed to a survey that found a 40 percent underreporting of one-day hospital stays because people responding to a survey simply forgot. More careful thought and survey design can improve the accuracy of people's responses, but "lying is not the full story," he said.

Graduate student Marcus Walter is a writer intern at the Cornell Chronicle.