The Measured Degree of Hiring Discrimination and the Level of Standardization of the Job Applicants´ Qualifications in Field Experiments

The idea with using field experiments for measuring discrimination in hiring is basically making all variables of a job applicant that are observable to the employer also observable to the researcher. This in turn should provide scope for measuring the true level of discrimination in hiring, which is very challenging if traditional ex post regression analysis of public microdata is used. However, most of the conducted field experiments have so far ignored that at what level the observable characteristics of the job candidates are standardized by the experiment might influence the measured degree of discrimination. In the current paper, a simple framework is first presented to illustrate the issue and then data from a field experiment conducted in the Swedish labor market is utilized to empirically analyze the question. The analysis show that the predicted difference in callback rate to a job interview between applicants with a typical Swedish and a typical Arabic name varies significantly over applications with different attributes attached. The conclusion is that studies which standardize the characteristics of the job applicants at a particular level might obtain very non generalizable results. At the end of the paper, we give some suggestions for how the field experimental methodology might be improved.