| The first two chapters of this dissertation examine different aspects of the identification and the estimation of the causal effect of education on labor market earnings. The last chapter appraises the extent of abusive behavior in the Unemployment Insurance program, one of the major social programs in the United States.; The first chapter quantifies the extent to which the rise in the measured return to education between 1979 and 2000 reflects a change in the causal effect of education on labor market earnings. As a basis for the empirical analysis, I develop a two-factor model of ability, schooling and earnings that emphasizes the role of unobservable ability and heterogeneous returns to education, and try to uncover the contribution of both factors to the changing earnings-schooling relationship. Using data from the Current Population Survey for cohorts of men born between 1930 and 1970, I find that the causal return to education increased by about 30% between 1979 and 2000.; The second chapter examines the causal link between family background characteristics, like parental education and family size, and the returns to education. In order to formalize the conceptual issues, I set out a model of schooling and earnings that explicitly recognizes the possibility that the returns to schooling may vary across individuals, and that part of this variation is generated by differences in familial environment. Using data from the 1973 Occupational Change in a Generation Survey, I find that men raised in larger families have substantially lower returns to education, while the combined effects of parental education are more modest. The analysis concludes that different aspects of familial environment affect schooling and earnings through different mechanisms.; The third chapter, written in collaboration with Orley Ashenfelter and David Ashmore, reports the results of a series of unique randomized trials designed to measure whether stricter enforcement and verification of work search behavior would lead to lower unemployment claims and benefit payments in the U.S. Unemployment Insurance program. Our results provide no support for the view that the failure to actively search for work has been a cause of overpayment in the UI system. |
