The Effects of For-Profit College Training on Earnings

In this paper, I obtain the estimates of the effects of for-profit training and credentials on students' annualized earnings. I differentiate for-profit students by the program level and account for students' self-selection into for-profit sector. I formulate the evaluation as the series progressing in the assumption on the source of selection bias: a basic specification of the Mincer earnings model is followed up by the rich-covariate model of selection on observables and finally by the multinomial model of selection into for-profit postsecondary training. To identify the selection into for-profit training, I use two exclusion restrictions: state-specific community college tuition, and county-specific share of Title-IV eligible for-profit schools. I find that selection into employment is not a pressing issue with for-profit college trained workers. Also, after controlling for self-selection into for-profit sector I find that on average a receipt of a for-profit certificate is associated with an increase in individual annualized earnings. However, this finding varies by gender, so I investigate various explanations for this gender divide. I observe that the differences in earnings effects between men and women are not systematically related to occupational differences by gender. Also, there may be some merit to the claim that the lack of the significant effects of for-profit Associate degrees for males are due to the unrelated vocational training.