What's in a Name? Reconciling Conflicting Evidence on Ethnic Names

January 01, 2006

Bureau of Labor Statistics data show that median weekly earnings for full-time workers in 2004 were 28.6 percent higher for white men than for African-American men and 15.6 percent higher for white women than for African-American women. Differences in education, experience, occupation and industry may explain parts of these gaps.1 Several years ago, economists Joseph Altonji and Rebecca Blank showed that after accounting for these worker characteristics, the gap in hourly wages between blacks and whites who worked full-time for all of 1995 was still 7 percent.

What portion of this gap—if any—can be attributed to discrimination rather than productivity-related effects? Several recent studies examined one signal of race—distinctively ethnic names—and asked whether those names are disadvantageous to African-American workers. Two studies offer seemingly opposite viewpoints; one focuses on opportunity and the other focuses on outcomes. The studies also discuss a variety of explanations, only some of which may indicate direct racial animus.


Economists Marianne Bertrand and Sendhil Mullainathan investigated the impact of having an ethnic name on the initial interview process. For their experiment, they created fictitious résumés of high and low quality, based on important characteristics, such as experience, career profile and various skills. They selected employment ads for sales, administrative support, clerical and customer service jobs in Boston and Chicago newspapers and sent four résumés—two of each quality level—in response to each ad. As a control for nonracial differences across résumés, African-American names were assigned to one résumé of each quality, and white names to the remaining two. Distinctively African-American names were chosen based on the ratios of black babies assigned that name to white babies assigned that name. For example, the authors used Kenya and Hakim for black résumés and Allison and Brad for white résumés.

Bertrand and Mullainathan found that résumés with white names received 50 percent more calls for interviews than résumés with black names. In addition, 8.4 percent of the employers contacted at least one more white applicant than black applicant, while only 3.5 percent of employers contacted at least one more black applicant than white applicant.2 The value of a high-quality résumé also varied between the two races. White résumés of high quality received 27 percent more calls than those of low quality, but for black résumés, the difference was only 8 percent in favor of the high-quality ones.

Do these results necessarily suggest discrimination? Bertrand and Mullainathan sought alternative explanations but failed to find a hiring rule consistent with their findings. For example, employers might seek to interview a certain number of African-American applicants, perhaps to match their proportion in the population. A disproportionate number of black applicants might then result in high-quality black applicants not receiving interviews. However, if firms were setting such hiring rules, one would expect the call rate for black applicants to even out when aggregated across all industries. In fact, it does not. The pervasiveness of a racial gap across all occupations and industries sheds doubt on the existence of such a hiring rule.

Bertrand and Mullainathan argued that their results suggest differential treatment based on name—and, hence, race—at least in the job search process. Explaining the difference in response rates using existing economic theories other than pure employer discrimination proved difficult since the only substantive difference was in the assigned names. Consequently, Bertrand and Mullainathan could not rule out employer discrimination.


Another study on the role of names in the labor market found results that seem to contradict Bertrand and Mullainathan’s conclusions. Economists Roland Fryer and Steven Levitt used information collected on non-Hispanic black and non-Hispanic white babies born in California between 1961 and 2000.3 The authors measured how distinct an African-American name is by calculating a Black Name Index (BNI), which measures the percentage of babies with a given name who are black.4

Fryer and Levitt found the BNI to be related to a number of variables associated with socioeconomic status. For example, single black mothers, as well as younger and less-educated black parents, are more likely to give their children distinctively ethnic names. Additionally, lower birth weight is correlated with a higher occurrence of ethnic names. Fryer and Levitt also found that the local socioeconomic environment can spill over to the likelihood of receiving an ethnic name. For instance, increasing per capita income in the residential ZIP code decreases the incidence of ethnic names. Moreover, children born in hospitals with lower percentages of black births—an indicator of the degree of neighborhood segregation—and children whose births are paid for by private insurance are, on average, less likely to be given ethnic names.

If employers believe both that low social background hinders human capital accumulation and that an ethnic name is a signal of low socioeconomic status at birth, then they may infer that an ethnic name signals low productivity. In this case, employers might forgo interviewing a person with an ethnic name on the basis of inferred productivity rather than animus. However, if employers use names to facilitate racial animus instead of as a signal of productivity, then one would expect to find variations in the effect on economic outcomes, and a black adult with an ethnic name would be worse off economically than an otherwise similar black adult with a race-neutral name, on average.

Do Fryer and Levitt’s data indicate that having an ethnic name leads to worse adulthood outcomes after controlling for background characteristics? The authors used data on women who were born in California in 1973 or 1974 and later gave birth there by 2000. The authors compared information on the woman’s own birth certificate, which provides information about her socioeconomic conditions at birth, with information about her that is available on her child’s birth certificate, which provides information on the woman’s adulthood economic outcomes.

Fryer and Levitt did not find any strong relationship between these measures. They found weak statistical evidence in favor of a relationship between BNI and life outcomes for only a few of the outcomes.5 The authors estimated that an increase in the woman’s BNI from 50 to 100 (i.e., from a race-neutral to a black-only name) is associated with a 0.9-percentage-point increase in the percent of African-American babies in the hospital, a 0.02-percentage-point increase in the probability that the woman was unmarried at the time of her baby’s birth, just under a $100 decrease in the per capita income among African-Americans in 1989 in the woman’s ZIP code and an increase in her child’s BNI by three points.

In light of their results, Fryer and Levitt concluded that having a distinctively African-American name will not directly cause worse economic outcomes in adulthood. Rather, they argue that such a name typically goes hand-in-hand with a worse socioeconomic background and, hence, lower productivity on average. After the authors controlled for negative economic conditions at the time of birth, they found that name alone has virtually no impact. They argue that this evidence supports the notion that employers may be inferring productivity from an ethnic name.


Fryer and Levitt suggested three theories consistent with the findings of both studies. First, an ethnic name could provide a signal of race to an employer. While some discriminatory employers might interview fewer minority workers, the overall effect on a black worker’s life outcomes might be mitigated by other fair-minded employers. Second, a distinctively black name might signal low social background and, hence, potentially lower levels of human capital.6 Thus, when corrected for social background, the effect of name alone on outcomes disappears. Third, ethnic names might have a direct impact on calls for interviews and length of unemployment duration but not have a strong influence on the outcomes that Fryer and Levitt studied.

Despite the variety of interpretations that can reconcile the findings from the two studies, some caveats are necessary. Foremost, neither study directly measures outcomes of people with ethnic names. Bertrand and Mullainathan’s use of fictitious résumés does not determine actual job market outcomes, whereas Fryer and Levitt used indicators but not direct measures of economic outcomes like personal income or wealth.7

To sum up, Bertrand and Mullainathan suggested that racial discrimination may affect the likelihood of being interviewed by some companies. However, it is unclear whether discrimination in some interviews leads to worse economic outcomes overall. Fryer and Levitt asserted that outcomes, as the authors define them, do not appear to be worse for those with ethnic names after controlling for social background. Only a small percentage of employers in the Bertrand and Mullainathan study seemed to discriminate based on name. Thus, that number of discriminatory employers may not be sufficiently large to affect job market outcomes across the board. Additionally, some employers may be attempting to infer underlying productivity from ethnic names.

In the end, ethnic names appear to serve as a hindrance in the labor market, but the exact extent has yet to be conclusively determined.


  1. According to the Census Bureau, among those aged 25 and older in 2004, 28.2 percent of whites had a bachelor’s degree or more, compared to 17.6 percent of African-Americans. Meanwhile, 19.4 percent of African-Americans had less than a high school degree, compared to 14.2 percent of whites. [back to text]
  2. Eighty-eight percent of the employers gave equal treatment, with the majority calling none of them. [back to text]
  3. The available data—obtained from birth certificates—include the baby’s first name, race, sex, date of birth, hospital of birth, birth weight, mother’s maiden name, parental ages and marital status. Beginning in 1989, parental education, residential ZIP code, form of payment, and the mother’s first name and date of birth are available. [back to text]
  4. Fryer and Levitt found that use of ethnic names increased over time, beginning with the Black Power movement in the 1960s and 1970s. The mean BNI for African-American babies increased from 60.9 during 1961-1967 to 71 during 1989-2000, where 50 represents an even split in the race of children with that name. Additionally, the percentage of African-American babies with a BNI of at least 80 increased from 20 percent in the 1960s to 45 percent in 2000. [back to text]
  5. The woman’s education and that of her baby’s father, her age at her first birth, her baby’s birth weight, whether or not she has private insurance, and her total number of children were not affected by her BNI. [back to text]
  6. Bertrand and Mullainathan, on the other hand, found no evidence suggesting that social background, as proxied by the mother’s education, affected call rates. [back to text]
  7. Fryer and Levitt warned that their results concerning economic outcomes may not generalize because their sample included only women who remained in California and gave birth by age 27. [back to text]


Altonji, Joseph G. and Blank, Rebecca M. “Race and Gender in the Labor Market,” in Orley Ashenfelter and David Card, eds., Handbook of Labor Economics, Vol. 3, pp. 3,143-259, New York: Elsevier Science, 1999.

Bertrand, Marianne and Mullainathan, Sendhil. “Are Emily and Greg More Employable Than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination.” American Economic Review, September 2004, Vol. 94, No. 4, pp. 991-1,013.

Fryer, Roland G., Jr. and Levitt, Steven D. “The Causes and Consequences of Distinctively Black Names.” Quarterly Journal of Economics, August 2004, Vol. 119, No. 3, pp. 767-805.

About the Authors
Kristie Engemann
Kristie M. Engemann

Kristie Engemann is a senior coordinator in the St. Louis Fed External Engagement and Corporate Communications Division. 

Kristie Engemann
Kristie M. Engemann

Kristie Engemann is a senior coordinator in the St. Louis Fed External Engagement and Corporate Communications Division. 

Mike Owyang
Michael T. Owyang

Michael T. Owyang is an economist and senior economic policy advisor at the Federal Reserve Bank of St. Louis. His research focuses on business cycles and time series econometrics. He joined the St. Louis Fed in 2000. Read more about the author and his research.

Mike Owyang
Michael T. Owyang

Michael T. Owyang is an economist and senior economic policy advisor at the Federal Reserve Bank of St. Louis. His research focuses on business cycles and time series econometrics. He joined the St. Louis Fed in 2000. Read more about the author and his research.

Views expressed in Regional Economist are not necessarily those of the St. Louis Fed or Federal Reserve System.

For the latest insights from our economists and other St. Louis Fed experts, visit On the Economy and subscribe.

Email Us

Media questions

Back to Top