To test the effectiveness of a job training program on the subsequent wages of workers, we specify the model log(wage;)
Posted: Mon May 02, 2022 8:26 am
To test the effectiveness of a job training program on the subsequent wages of workers, we specify the model log(wage;) = Bo + Bitraini + B2educ; + Bzexperi + ui = Where train is a binary variable equal to 1 if a worker participated in the program, and 0 otherwise. Think of the error term u as containing unobserved 1 worker ability. If less able workers have a greater chance of being selected for the program, does the OLS estimator of B1 (Ⓡ1) over-estimate or under-estimate B1? Show steps.