Tuesday, January 14, 2003

previous entry | main | next entry | TrackBack (1)


THE RESIDUE OF RACISM: One mantra that persisted throughout the Lottroversy was that racism remains a problem in American society. That's an easy thing to say, but what exactly does it mean?

This story does a nice job describing the nature of the problem:

"CHICAGO, Illinois (AP) -- It helps to have a white-sounding first name when looking for work, a new study has found.

Resumes with white-sounding first names elicited 50 percent more responses than ones with black-sounding names, according to a study by professors at the University of Chicago Graduate School of Business and the Massachusetts Institute of Technology.

The professors sent about 5,000 resumes in response to want ads in the Boston Globe and Chicago Tribune. They found that the 'white applicants they created received one response -- a call, letter or e-mail -- for every 10 resumes mailed, while black' applicants with equal credentials received one response for every 15 resumes sent."

Click here for the actual study. If you read the story, it's clear that the researchers controlled for other explanatory factors. [But c'mon, don't researchers who engage in these studies mine the data for results that favor their pre-existing beliefs?--ed. One of the researchers, the University of Chicago's Marianne Bertrand, has conducted other studies about economic discrimination. This abstract of a co-authored NBER paper suggests that she does not manipulate her data].

UPDATE: OK, I was apparently way behind the curve on this study, which Alan Krueger discussed in his column last month in the New York Times. Brad DeLong, Kevin Drum, and Thomas Maguire posted on this more than a month ago. The criticism of the study is that the name selection could merely indicate a bias against "outside-the-mainstream" names, and not necessarily racism. The authors do seem to have covered this with survey research on attributing names to racial backgrounds -- and they're also quite forthcoming about the drawbacks of their testing approach in the paper. Steven Postrel e-mails to raise a better criticism, which is that the African-American names were the most "countercultural" while the white names were as WASPish as you could get. Point taken.

One possible problem that occurred to me was that the experiment was carried out "between July 2001 and January 2002 in Boston and between July 2001 and May 2002 in Chicago." Since several of the African-American names have Islamic-sounding names, I wondered if those names combined with 9/11 were responsible for the result. Surprisingly, the results (Table 2 in the paper) don't suggest that either.

posted by Dan on 01.14.03 at 03:06 PM