In the Washington Post, Richard Hanania conducted an experiment to find out how an apology from a politician affects the public’s perception of that politician:
I recruited a sample of 511 individuals and had them read two texts. First, they read about Rand Paul’s 2011 comments suggesting that he disagreed with parts of the Civil Rights Act. Paul had said that while he denounced racist behavior, part of his definition of freedom meant the right to discriminate on private property. About half of the participants read a conclusion to the story that made Paul seem apologetic, while the rest were led to believe that he stuck firm to his comments. (In actuality, Paul never apologized for his statements, but began to deny that he ever questioned the Civil Rights Act.)
Respondents then read about the suggestion by then-Harvard President Larry Summers in 2005 that genetic factors help to explain the lack of high-performing female scientists and engineers at top universities. After reading the comments and hearing about the outcry, half the participants were told that Summers defended himself by saying he believed that “raising questions, discussing multiple factors that may explain a difficult problem, and seeking to understand how they interrelate is vitally important.” The rest learned that he had apologized and read a brief statement Summers made expressing regret for his comments and reflecting on the damage that they had caused.
Contrary to the conventional wisdom, neither Summers nor Paul was helped by an apology. Among respondents who read that Paul was apologetic, 63 percent said that the controversy made them less likely to vote for Paul. Among those who didn’t read about any apology, 61 percent said they were less likely to vote for him — a statistically insignificant difference.
The results for the Summers controversy were even more surprising. Of those who read about his apology, 64 percent said that he “definitely” or “probably” should have faced negative consequences for his statements about women. However, that number dropped to 56 percent when respondents were led to believe that Summers stood firm in his position. Moreover, the surprisingly negative effect of Summers’ apology was even larger among the groups that arguably should have appreciated the apology: women and liberals.
Given these results, why would politicians apologize at all? It may be simply out of habit or because they are following a script that has for the most part gone unquestioned. To be sure, my experiments certainly don’t suggest that it is always inadvisable to apologize. Nor can my findings speak directly to Trump.
