Earnings, Occupation, and Human Capital Investment

In recent years the human capital earnings model has been widely used as a framework for examining the determination of earnings. A specification of the human capital earnings function developed by Chiswick and Mincer has been employed to examine earnings determination across individuals [Mincer, 1974b], states [Chiswick, 1974], metropolitan areas [Hirsch, 1978], occupations [Rahm, 1971], and over time [Chiswick-Mincer, 1972]. However, empirical studies utilizing the human capital framework have assumed that the effects of workexperience (postschool investment) on earnings are identical for individuals, regardless of occupation. This assumption runs counter to the current emphasis of many labor economists on the importance to lifetime earnings of gaining early access to certain occupational job ladders and of receiving firm specific on-the-job training [Doeringer and Piore, 1971 ; Thurow, 1975]. Likewise, human capital economists stress the importance of postschool training investments and at least recognize the fact that these investments vary across occupations. Indeed, Mincer has stated that [Mincer, 1974a, p. 33] : "an analysis in which these effects [of schooling and of experience] are allowed to differ [across occupations] would be desirable for a number of purposes, not the least of which is an insight into differential post-school job skill investments." This paper examines whether schooling, experience, and weeks worked have different effects on individual earnings in nine separate occupational categories. The Chiswick-Mincer spec*University of North Carolina at Greensboro. The research is drawn in part from my doctoral dissertation on which William R. Johnson and Roger Sherman provided helpful comments. Support by a doctoral dissertation grant from the U.S. Department of Labor, Employment and Training Administration, is gratefully acknowledged. ification of the human capital earnings function is modified in order to focus on differences across occupations in their earnings-experience profiles. This makes possible several inferences regarding the intensity, length, and rate of return to postschool human investments across occupational job ladders. In addition, we examine the manner in which occupation interacts with schooling in the earnings generation process and discuss analytical problems which arise in modeling this process. Finally, the relationship between wage rates, weeks worked, and occupation is examined. The Theoretical Framework