2. To test the effectiveness of a job training program on the subsequent wage of workers, we specify the model: wagei=0+1traini+2educi+3experiencei+ui,i=1,,n. where traini is a binary regressor equal to 1 if the i-th person participated in the training program. Think of the error term as containing unobserved worker ability. Discuss in what kind of situation the OLS estimator of 1 is likely to have bias and show the sign of the bias. In what kind of situation there will be no bias for the OLS estimator of 1 ? 2. To test the effectiveness of a job training program on the subsequent wage of workers, we specify the model: wagei=0+1traini+2educi+3experiencei+ui,i=1,,n. where traini is a binary regressor equal to 1 if the i-th person participated in the training program. Think of the error term as containing unobserved worker ability. Discuss in what kind of situation the OLS estimator of 1 is likely to have bias and show the sign of the bias. In what kind of situation there will be no bias for the OLS estimator of 1