Race and gender effects on employer interest in job applicants: new evidence from a resume field experiment

ABSTRACT

We sent nearly 9000 fictitious resumes to advertisements for job openings in seven major cities in the United States across six occupational categories. We randomly assigned names to the resumes that convey race and gender but for which a strong socio-economic connotation is not implicated. We find little evidence of systematic employer preferences for applicants from particular race and gender groups.

KEYWORDS:

• Discrimination
• field experiment
• resume audit study

JEL CLASSIFICATION:

• J71
• J64

Acknowledgements

The authors thank Scott Delhommer, Jared Dey, Lucas Singer, Trey Sprick and David Vaughn for research assistance.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ For ease of exposition, throughout the note we use the term ‘race’ to indicate race/ethnicity.

² Per 2000 Census data, see https://www.census.gov/genealogy/www/data/2000surnames/.

³ The X-vector includes controls for whether the resume is the first one to be sent to the employer, whether it was accompanied by a more-enthusiastic greeting, city and occupation indicators, a flexible time trend to account for seasonality in employer responses, information on work-history gaps and education credentials. The findings we present below are not substantively sensitive to which controls we include in the models, which is not surprising given the random assignment of names to resumes.

⁴ We also estimate models where the dependent variable is an indicator for an explicit interview request and obtain similar results to what we report below. These confirmatory results are omitted for brevity but available from the authors upon request.

⁵ The lower bound of the 95% confidence interval for the African American effect in columns 1 and 2 of Table 1 rules out an effect as large as the one found in Bertrand and Mullainathan (2004).

⁶ Qualitatively similar results are obtained using the interview-request models (results are omitted for brevity).

Additional information

Funding

This work was supported by the Spencer Foundation, the Economic and Policy Analysis Research Center at the University of Missouri, and CALDER.