A team of economists has uncovered persuasive evidence that local government officials throughout the United States are less responsive to African-Americans than they are to whites.
The researchers sent roughly 20,000 emails to local government employees in nearly every county. The emails posed commonplace questions, like “Could you please tell me what your opening hours are?”
The emails were identical except that half appeared to come from a DeShawn Jackson or a Tyrone Washington, names that have been shown to be associated with African-Americans. The other half used names that have been shown to be associated with whites: Greg Walsh and Jake Mueller. The email sent to each local officeholder was determined by chance.
Most inquiries yielded a timely and polite response. But emails with black-sounding names were 13 percent more likely to go unanswered than those with white-sounding names. This difference, which appeared in all regions of the country, was large enough that it was statistically unlikely to have been a matter of mere chance.
These troubling results were documented in the paper, “Racial Discrimination in Local Public Services: A Field Experiment in the US,” by Corrado Giulietti of the University of Southampton in Britain, Mirco Tonin of the Free University of Bozen-Bolzano in Italy, and Michael Vlassopoulos, also of the University of Southampton. The study is to be published in the Journal of the European Economic Association.
The findings appeared to be a striking indication of racial discrimination in seemingly benign and mundane interactions. The tendency to ignore emails sent by African-Americans was particularly pronounced in sheriffs’ offices, but it was also evident in school districts and libraries.
In a clever twist, the authors analyzed whether the replies were polite, counting responses that included either the sender’s name or words like “hi,” “Mr.,” “dear,” “good” (which captures “good morning,” “good afternoon” and “have a good day”) or “thank” (which captures both “thanks” and “thank you”). By this measure, those with apparently African-American names received 8 percent fewer polite responses than those with white names.
While many studies have found differences in treatment for African-Americans and whites in employment, housing and the criminal justice system, it hasn’t always been clear whether these differences reflect discrimination or other factors.
The usual difficulty is that it’s impossible to find, say, job seekers who are absolutely identical in every respect except race. As a result, it is difficult to conclude whether a white job seeker succeeded — and a black one didn’t — because of discrimination. While statistical techniques can adjust for some of these factors — education, geography and the like — no analysis can account for all of them.
But the new research allows for a clearer conclusion: It appears to have documented straightforward discrimination.
As a real-world experiment, it built on earlier “audit experiments,” as they are known in social science. Perhaps the most famous is a study by Marianne Bertrand of the University of Chicago and Sendhil Mullainathanof Harvard (who is a regular contributor to this column). In that earlier experiment, Ms. Bertrand and Mr. Mullainathan sent fictitious résumés to employers, finding that people with white-sounding names were more likely to receive a positive response than those with black-sounding names.
Economists tend to group explanations of discriminatory behavior into two buckets: taste-based and statistical. If a librarian chooses not to respond because a person is black, that’s taste-based discrimination. In common speech, there’s a simpler label: racism.
Statistical discrimination, on the other hand, occurs when a librarian uses a person’s name or race as a marker for other characteristics. Perhaps an African-American-sounding name signals that a person is more likely to be poor. The librarian happens to be biased against poor people. In this case, race is being used as a statistic for inferring poverty, and it’s the perception of poverty that causes the discriminatory behavior.
But two pieces of suggestive evidence in this study point to the problem here as being straightforward, taste-based discrimination.
First, the authors repeated the exercise — sending an additional 20,000 emails to the same recipients — although this time with a twist. They added a signature line, identifying the sender as a real estate agent. This extra information made the sender’s name — whether it seemed to be African-American or white — less relevant for inferring income or socioeconomic status. If statistical discrimination had driven behavior in the first round, this extra information should have led to less discrimination in the follow-up. It did not.
Second, the pattern of evidence was consistent with taste-based discrimination. While the researchers didn’t determine the race of the people who responded to their emails, they did have data on the racial breakdown of the municipal work forces. The racial gap in email response rates was greater in counties where the proportion of whites was higher.
Taste-based discrimination — basically, racism — isn’t necessarily the result of conscious thought. In an email, Mr. Tonin, one of the study’s authors, said that it’s possible “this behavior is due to some sort of unconscious bias” and, therefore, that “making people aware of the problem may contribute to the solution.”
If awareness really is the first step toward a fix, then the study may be helpful in refining our understanding of racial discrimination in America. It occurs not only in the labor market and the criminal justice system, but also in countless small frictions every day.
The culprit may not be a hate-spewing white nationalist, but rather a librarian or a school administrator or a county clerk, unaware that she’s helping some clients more than others.