Clarification from Prof Sendhil Mullainathan

In one of my earlier post, I had quoted a very interesting study on discrimination in the US labor market.

I mailed one of the authors of the paper, Prof Mullainathan, for a clarification and the question I had was..

"By applying to 1300 REAL jobs posted in newspapers and adding 5000 NON-EXISTENT resumes to the job seekers pool during the period of study, do you think the study would have changed the hiring process of atleast a few jobs and thus affected the actual outcome/dynamics of those jobs for a few REAL applicants?'"

Prof Mullainathan replied,

"One of the things we did was to check with several HR managers before the study. We were worried that by giving one of our people an interview, we'd be crowding out a real interview. But they told us that there is actually a non-trivial percent of people who they try and call for an interview but cannot reach to schedule one. In these cases, they simply interview a few more people. As such, when they call one of our fake applicants and can't reach them, the next best candidate would simply get their chance at an interview. I think the distortions moreover are far more severe at the interview stage. At that stage, there is the potential for actually crowding out one real applicant, but we never send anyone in for a real interview."

I would like to THANK Prof Mullainathan for the clarification. More of his working papers can be obtained here.

Some good reads include..
"The Market for News" (with Andrew Shleifer), 2004
"Behavioral Economics" (with Richard Thaler), 2000