Scribd Logo
Upload

Welcome to Scribd!

  • 182 views

    Are Greg and Emily More Employable Than Lakisha and Jamal

    Uploaded by

    Ruhi Sonal
    This field experiment examined racial discrimination in the US labor market. Researchers sent roughly 5,000 fake resumes in response to job postings, varying only the applicant's name to signal race (African American or white). They found African American applicants received fewer callbacks than equally qualified white applicants. Having a white name yielded as many callbacks as an additional 8 years of work experience. The racial gap in callbacks existed across occupations and industries, and improving resume quality did little to close this gap. The results suggest racial discrimination persists in US hiring.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PPT, PDF, TXT or read online from Scribd

    Are Greg and Emily More Employable Than Lakisha and Jamal

    Uploaded by

    Ruhi Sonal
    182 views20 pages
    This field experiment examined racial discrimination in the US labor market. Researchers sent roughly 5,000 fake resumes in response to job postings, varying only the applicant's name to signal race (African American or white). They found African American applicants received fewer callbacks than equally qualified white applicants. Having a white name yielded as many callbacks as an additional 8 years of work experience. The racial gap in callbacks existed across occupations and industries, and improving resume quality did little to close this gap. The results suggest racial discrimination persists in US hiring.

    Copyright

    © Attribution Non-Commercial (BY-NC)

    Available Formats

    PPT, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This field experiment examined racial discrimination in the US labor market. Researchers sent roughly 5,000 fake resumes in response to job postings, varying only the applicant's name to signal race (African American or white). They found African American applicants received fewer callbacks than equally qualified white applicants. Having a white name yielded as many callbacks as an additional 8 years of work experience. The racial gap in callbacks existed across occupations and industries, and improving resume quality did little to close this gap. The results suggest racial discrimination persists in US hiring.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PPT, PDF, TXT or read online from Scribd
    Download as ppt, pdf, or txt
    182 views20 pages

    Are Greg and Emily More Employable Than Lakisha and Jamal

    Uploaded by

    Ruhi Sonal
    This field experiment examined racial discrimination in the US labor market. Researchers sent roughly 5,000 fake resumes in response to job postings, varying only the applicant's name to signal race (African American or white). They found African American applicants received fewer callbacks than equally qualified white applicants. Having a white name yielded as many callbacks as an additional 8 years of work experience. The racial gap in callbacks existed across occupations and industries, and improving resume quality did little to close this gap. The results suggest racial discrimination persists in US hiring.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PPT, PDF, TXT or read online from Scribd
    Download as ppt, pdf, or txt
    of 20

    Are Greg and Emily more Employable than Lakisha and Jamal?

    A Field Experiment on Labor Market Discrimination

    Marianne Bertrand Sendhil Mullainathan American Economic Review, 2004

    Are African Americans discriminated against in the labor market?

    Differing views on whether discrimination exists or not

    Role of affirmative action programs. Profit maximization is the motive of employers. Reverse discrimination due to stringent enforcement of affirmative action programs.

    Previous Research on discrimination in labor markets

    Census: Shows persistence of inequalities, but suffers from the problem of unobservable characteristics in census data. Audit Studies: These cannot account for inherent differences in individuals. This method is not double blind.

    Features of this experiment

    Uses resumes instead of employing human auditors: comparability maintained across race. Random assignment of resumes to names. No demand effects: In case of audit studies, the auditors are not actually job seekers. Large sample size: nearly 5000 resumes sent in response to over 1300 ads.

    Experiment Design

    Creating a bank of resumes:


    Four occupational categories: Sales, administrative support, clerical services and customer services. Resumes prepared on the basis of resumes of actual job seekers in Chicago and Boston. Resumes modified so that they werent identical to actual job seekers. Resumes modified to accentuate quality difference, creating high quality and low quality resumes.

    Identities of Fictitious Applicants:

    Name frequency data on babies born in Massachusetts between 1974-79 used to ascertain which names are unique to African Americans and Whites. Survey conducted in Chicago to check for race associated with the names. Resumes of the same race, sex, city and resume quality were allotted the same phone number. Voice mailbox system was used. Fictitious addresses constructed based on real streets in Boston and Chicago. Addresses allocated randomly to resumes. E-mail accounts created and addresses mentioned in higher quality resumes.

    Responding to Ads:

    Employment ads that sought resumes in The Boston Globe and The Chicago Tribune were selected. Four resumes- high and low quality of both races were sent in response to each advertisement. Resumes were altered according to job specifications wherever required. Nearly 5000 resumes sent in response to over 1300 ads.

    How to ascertain discrimination?

    Equal treatment: When a callback is received on


    all four resumes none of the resumes both high quality resumes both low quality resumes

    Results of the Experiment

    Impact of Resume Quality

    Lower incentive to invest in skills: The difference in callback percentage (probit regression) for high quality resumes is statistically significant for African American names, but remains lesser than the whites. Impact of applicants address: Living in neighborhoods that are whiter, have higher income earners and more educated people increases callback rate, but there is no evidence that the African Americans benefit more than the whites. A white name is equivalent to the return on additional 8 years of experience.

    Effect of Job Requirements:

    Several ads mentioned specific requirements: communication skills (12%), organization skills (7%), education level(11%). No significant effect of job requirements on racial difference in callback was observed. Information about employer characteristics was collected through the ads, and web reverse lookup services: size, ownership, equal opportunity employer, federal contractors, etc.

    Effect of Employer Characteristics:

    Callback rates from Equal opportunity employers does not differ significantly from the rest of the firms.

    Differential treatment does not vary with employer size. Point estimates show less differential treatment in the not for profit sector.

    Employers were classified into six industry categories; jobs into six occupation categories:

    Positive gap in callback observed in all occupation and industry categories (except transport and communication). No systematic relationship between occupation earnings and the racial gap in callback. Differences in gap in callback rates across industries and occupations was not statistically significant.

    Significant positive effect of employer location (zip code), but this effect was very small.

    Does a high callback rate for whites imply discrimination?

    Since the names were assigned randomly, the callbacks for both races should not differ on the average. We can also get these results if a disproportionately high number of African Americans apply to the jobs in the sample, while the firms target a particular number of African American employees. E.g. Equal Opportunity Employers. But this is unlikely, as callback rates across industries and occupations do not differ significantly.

    Weaknesses of the Experiment:

    It measures gap in callback for interviews rather than gaps in hiring or earnings. Resumes suggest rather than directly report race. The race specific names may not be representative of the average African American who may not have a racially distinct name. Newspapers represent only one channel of job search: it is possible that African Americans use mostly social networks for job search or employers who hire using such networks discriminate less.

    Potential Confounds

    Apart from race, the names chosen may signal some personal trait. E.g. Latonya and Tyrone are names that might be interpreted as coming from disadvantaged backgrounds.

    But, African Americans benefit lesser than whites when they are from well off neighborhoods. This gap should be more for jobs requiring soft skills. The chosen names are more prevalent among the disadvantaged. But the correlation between the education level of mothers and the callback rates is not significant.

    Reverse Discrimination: Employers of small firms may assume that highly qualified African Americans have better opportunities. This is unlikely, as the gap in callback is not significantly smaller among better jobs.

    Relation to Existing Theories

    Taste based discrimination: Customers, coworkers or employers may have prejudiced tastes.

    Discrimination by customers and co-workers is not an explanation as the amount of customer contact and fraction of white employees differ significantly in different industries and occupations. Employer discrimination: Can explain the general difference in callback, but cannot explain why return to credentials is lower for African Americans.

    Statistical Discrimination

    Employers may use observable characteristic (race) as a proxy for an unobservable characteristic. This cannot explain the credentials effect.

    Affirmative action makes it easier for African Americans to acquire skills, thereby these are discounted at a higher rate by employers. Lexicographic searches by employers: Employers may follow rules such as read no further after coming across a resume with an African American name.

    Conclusion

    African Americans face differential treatment while searching for jobs, across occupations and industries. Improving observable skills and credentials does not seem to overcome this problem.

    You might also like