Audit studies have been used for decades to show discrimination by race and sex. In these studies, two otherwise equal candidates are separated by one feature, perhaps race, perhaps first name, perhaps gender. Then the researcher looks at the differing response rates from mortgage companies or landlords or businesses. For example, see this notable 2003 study that showed that whites with a criminal record had better job prospects than blacks with no criminal records. This mode of analysis was recently utilized in an article in the latest issue of the American Journal of Sociology to show differing response rates for gay men looking for a job versus those applicants who were not openly gay:
The study, which is the largest of its kind to look at job discrimination against gay men, found that employers in the South and Midwest were much less likely to offer an interview if an applicant’s resume indicates that he is openly gay. Overall, the study found that gay applicants were 40 percent less likely to be granted an interview than their heterosexual counterparts…
For the study, Tilcsik sent two fictitious but realistic resumes to more than 1,700 entry-level, white collar job openings — positions such as managers, business and financial analysts, sales representatives, customer service representatives, and administrative assistants. The two resumes were very similar in terms of the applicant’s qualifications, but one resume for each opening mentioned that the applicant had been part of a gay organization in college.
“I chose an experience in a gay community organization that could not be easily dismissed as irrelevant to a job application,” Tilcsik writes. “Thus, instead of being just a member of a gay or lesbian campus organization, the applicant served as the elected treasurer for several semesters, managing the organization’s financial operations.”
The second resume Tilcsik sent listed experience in the “Progressive and Socialist Alliance” in place of the gay organization. Since employers are likely to associate both groups with left-leaning political views, Tilcsik could separate any “gay penalty” from the effects of political discrimination.
The results showed that applicants without the gay signal had an 11.5 percent chance of being called for an interview. However, gay applicants had only a 7.2 percent chance. That difference amounts to a 40 percent higher chance of the heterosexual applicant getting a call.
Has this methodology not been used before when looking at discrimination against gays? If not, I’m a little surprised.
The methodology of the audit study is interesting: it essentially is an experiment where one detail between the resumes or applications is changed to isolate its effect. In this study, the detail is what college organization the applicant worked for. It was clever to differentiate between these two particular organizations in order to rule out a political effect.
I assume the next step would be to expand this study to more states/locations and to more job categories beyond “entry-level, white collar job openings”? How much would the results differ if the study involved more blue collar jobs or managerial, white collar jobs?
When I’ve told students about audit studies, several have raised issues of deception: aren’t these fake applicants? Yes, but the harm to the companies is minimal outside of some time spent looking at them. (However, from what I have read about how much time hiring people look at the hundreds of job applications they see for these types of positions, looking at one or more applications wouldn’t seem to matter.) I do wonder if hiring people have ever spotted these applications from researchers – or the pool of applications is simply so broad that they see all sorts of interesting things.