Estimation of a Life-Cycle Skill Accumulation Model with Comparative Advantages and its Economic Implications∗

We estimate a structural model of schooling, labor supply and life-cycle earnings in which some individuals have comparative advantage in producing skills in school while others are better at producing skills in the labor market and analyze some of its key implications. The model is estimated using a sample of white males taken from the 1979 cohort of the National Longitudinal Survey of Youth (NLSY). We find that schooling induces more skill formation (at an average rate of 5.9% per year) than full-time work (at a rate of 4.9% per year) in the early phase of the life-cycle (until age 35). Although the correlation between individual specific rates of skill production in school and in the market is positive (0.19), 24% of the US white male population is more effective (ex-ante) at producing skills in the market than in school and has already accumulated as much skills in the market as in school by age 35. We show how comparative advantages reduce the explanatory power of individual differences in the utility of attending school in favor of more educational selectivity based on the skill production technology and transforms the notion of Ability Bias into a positive correlation between realized schooling and subsequent wage growth. Comparative advantages also tend to damp the portion of wage inequality explained by initial conditions (about 50% of observed inequality at age 35 is explained by heterogeneity). Finally, we show how comparative advantages may help explain the incidence of negative IV estimates of the return to education.