Why Jamal Can’t Get a Job
That’s the title of a story in the current issue of the Chicago Graduate School of Business magazine. The article summarizes the results of a novel study by the auspiciously named Prof. Marianne Bertrand of U Chicago and Senhil Mullainathan of MIT’s Economics department. Here’s their experiment:
Between June 2001 and May 2002, they sent out about 5,000 resumes in response to 1,300 help-wanted ads in Boston and Chicago. They found that resumes with white-sounding names received 50 percent more calls than those with African American-sounding names, despite identical qualifications.
A few methodological details: the resumes were taken from resumes careerbuilder.com and americasjobbank.com, but the names were changed to represent either white or black names. To determine which names went with which races, they culled names from birth records between 1974 and 1979 to find the most common names for each race, and followed that up with “man on the street” interviews to verify that perceptions matched the data.
In practice, some high schools are better than others, so identical numerical GPAs might not necessarily mean the same thing on two resumes. In other words, it’s possible that the call-backs were skewed white not due to racial discrimination per se but rather because whites disproportionately come from good high schools. To control for this, they used addresses from mixed-race neighborhoods and high schools in Boston and Chicago. In a nutshell, the applicants truly are identical on paper, except for their names. The result: black names get 1/3 less call backs.
In the second phase of the study, they also boosted the resumes (adding experience, honors, …) of matched pairs of resumes and found that a better resume increased the odds of a call back for a white name, but not for a black name.
But at least there’s some bad news for everyone who’s poor:
“Whoever you are, whatever your resume is, we find people who live in ‘worse’ neighborhoods — meaning low income, black, low education – those people had lower response rates … Living on the South Side of Chicago hurts you whether you’re white or black.
Speaks for itself. The full paper is available here.
UPDATE: Reader Bo emails an alternative hypothesis, that’s definitely not without merit:
It could be that people discriminate not against black names but against unusual names. They should have sent out resumes with typical black names as well as unusual Jewish, Indian, etc. names to get a better idea of whether they were seeing true racial discrimination or wierd-name bias.
Here’s my response:
That’s a decent point, and the authors mentioned that they’d like to replicate it with other races. So it was probably a matter of time and money and there’s likely more to come.
Interestingly, to my mind, the “black last names” are actually not particularly unusual. They are: “Jackson, Jones, Robinson, Washington, and Williams”. So any non-black minority is bound to have even more unusual last names.
I suppose that to confirm the racism hypothesis, they’d need to find little effect from, say, clearly Jewish names. So if Shlomo and Ehud get the same number of callbacks, while Jamal and Tyrone can’t, then you have an issue. Stay tuned.
In the interim, I’ll apply Occam’s Razor and accept the simplest explanation.