Obesity and hiring discrimination.

We implement a correspondence audit study that uses a non-paired design to test the effect of obesity on employment discrimination in Mexico. In Mexico it is common practice to include a photograph in a resume, and job advertisements often specify personal and physical characteristics. We use both types of information to evaluate discrimination against the obese and identify its potential channels. We send two fictitious resumes, one of a woman and another of a man, one obese and the other non-obese, in response to advertisements for job openings. The obese photograph is a digital manipulation of the picture of the same non-obese person. We send a total of 3202 resumes in response to 1696 job advertisements. There is clear evidence of discrimination against obese women, but not obese men. The callback rate for the non-obese women is 29.1 %; for obese women it is 21.3 %. The difference is statistically and economically significant. An obese woman would need to send 37 % more resumes to obtain the same number of callbacks as a non-obese woman. We explore different channels that could explain this result, including customer discrimination (employer discrimination based on perception of customer expectations), productivity and salary, the gender of the person making the hiring decision, and the proportion of women in the occupation. We find no strong evidence in favor of any one channel. Advertisements with men identified as the contact person discriminate more than those with women, but replacing those men with women would reduce the observed gap in callback rate no more than 25 %.