The results have startled him. As COVID-19 spread in March and April, did signers give less of their capital to shareholders (via dividends and stock buybacks)? No. On average, signers actually paid out 20 percent more of their capital than similar companies that did not sign the statement. Then, as the coronavirus swept the country, did they lay off fewer workers? On the contrary, in the first four weeks of the crisis, Wry found, signers were almost 20 percent more prone to announce layoffs or furloughs. Signers were less likely to donate to relief efforts, less likely to offer customer discounts, and less likely to shift production to pandemic-related goods. “Signing this statement had zero positive effect,” said Wry. Why, though, would it produce a negative effect?
Wry told me he has yet to nail down a causal explanation. (His first theory—that signing the statement drew counterpressure from institutional investors—found no supporting evidence.) But he said his findings are not inconsistent with psychological explanations. Behavioral psychologists have observed an effect they call “moral self-licensing”: If people are allowed to make a token gesture of moral behavior—or simply imagine they’ve done something good—they then feel freer to do something morally dubious, because they’ve reassured themselves that they’re on the side of the angels.
Read: How capitalism drives cancel culture
One of the starker examples happens to involve police and racial prejudice. In a 2001 study, the Princeton and Stanford psychologists Benoît Monin and Dale Miller asked subjects to imagine themselves as the police chief of a small town that has historically been exclusively white. Police officers in the department harbor racist attitudes and—a few years earlier—an African American officer had quit, citing the hostile environment. Now they need to hire a new officer. Should ethnicity be a factor? Is the job better suited for a Black candidate? Or a white one?
It didn’t matter, participants said.
But a second group was allowed to rubber-stamp the hiring of a Black candidate for an unrelated consulting job prior to being presented with this scenario. It was the most perfunctory of decisions—the three white consulting candidates were less qualified—but apparently enough to establish the participants’ moral bona fides such that they could then comfortably veer into prejudice. This second group was measurably more likely than the first to say that the policing job was better suited for a white person.
Was this all about avoiding the appearance of racism? Interestingly, no. The effect persisted in a similar experiment, even when no one else could see the subjects’ choices. So they weren’t protecting solely their reputation. It was also, at least partially, about their self-image. Maintaining a consistently good view of one’s self is very important to people—and very easily accomplished. In a 2008 version of the police-chief study, merely indicating that they would vote for Barack Obama in the upcoming election licensed participants to favor a white applicant for the position. In another setup in the 2001 study, the chance to disagree with brazenly sexist statements enabled people to favor male candidates over identically qualified women.