Business Week - November 3, 2003
ECONOMIC VIEWPOINT By Robert J. Barro
What's In A Name For Black Job Seekers?
In some respects, economic conditions for black Americans have improved substantially relative to those for whites over the past four decades. Recent studies by University of Texas economist Finis Welch show that the ratio of black to white wages for the average employed male rose from 43% in 1940 to 80% in 2000 (though the surge in black unemployment over the past couple of years may have eroded some of the gains).
On the negative side, some of that convergence reflects the withdrawal of low-skilled blacks from the labor force, including the large jump over the past 20 years in the number of black prison inmates. This rise in incarceration, by removing many low-wage black earners from the labor force, may account for much of the apparent gains in black wages since 1980. Moreover, gaps between whites and blacks remain large in schooling, out-of-wedlock childbearing, health, and other economic and social indicators.
In a new working paper from the National Bureau of Economic Research, economists Roland G. Fryer Jr. and Steven D. Levitt point out that children's names are another area with a large gap between blacks and whites. This research was inspired by so-called audit studies, begun in Britain in 1970, which found that potential employers systematically discriminated against job seekers with résumés that contained minority-sounding names. But Fryer and Levitt conclude that in the U.S., the black-white differences in naming are a consequence of disparities in socioeconomic conditions of blacks and whites, not an independent cause of disparities.
There's no question that gaps in naming practices are striking. As an example, in California from 1989 to 2000, out of 457 boys named Tyrone, 445 were black. Out of 277 girls named Shanice, 274 were black. In contrast, out of 2,328 girls named Molly, only 6 were black. Using "white" names as the comparison group, the tendency for distinctive naming was much weaker among Hispanics and nonexistent for Asians.
However, the type of name chosen turns out to reflect the parents' socioeconomic status as well as their race. For example, a distinctive "black" name is much more likely to be chosen when the mother is unmarried, when the mother has other children, and when the mother is in poorer health (as signaled by a lower birth weight of the child). In data available only for the 1990s, a more distinctive black name is also associated with lower parental education and lower per capita income in the mother's Zip Code.
Using data on all births in California from 1961 to 2000, Fryer and Levitt find that black names were similar to whites' up to the early '70s. That changed in the early '70s alongside the Black Power movement. By contrast, in the '60s, a black child's first name was not closely related to the mother's socioeconomic situation.
Fryer and Levitt's study shows that the more black-sounding a person's name, the more likely the parents have a lower socioeconomic status. Employers thus might infer that a job seeker with a black-sounding name is more likely to have grown up in a less educated and poorer family. If employers believe, rightly or wrongly, that such a background lowers the chance of job success, this may help explain why audit studies find that employers react negatively to black-sounding names on résumés, which contain current education but little other socioeconomic information. Thus, a recent NBER working paper by Marianne Bertrand and Sendhil Mullainathan found that call-back rates on identical (and made-up) résumés were substantially higher for white-sounding names such as Emily and Greg.
True, employers could be racially discriminating. But once Fryer and Levitt take into account parents' socioeconomic status, the researchers find that a person's name no longer predicts much about later economic outcomes, such as whether he or she winds up living in a rich or poor area. That is, ultimately it matters whether your parents are well-educated or rich but not whether they name you Shanice or Molly. Despite the apparent bias of employers against black names on résumés, the whole hiring process operates efficiently enough not to give much weight to names per se.
Thus, the key issue is whether black-white gaps in income and other economic and social indicators are still narrowing rather than whether employers are discriminating against black names. It's the socioeconomic status of blacks that needs attention, not whether a child is named Shanice or Molly.