Explaining Schooling Returns and Output Levels Across Countries

This paper develops and calibrates a model with endogenous schooling and returns to schooling to explain Mincerian returns variation across countries, and assess the importance of schooling on output variation. The calibrated model is able to explain forty percent of the variation in the data, substantially more than linear regressions explain. Variation in the direct costs of schooling driven by government funding levels relative to enrollment rates and fertility rates contribute the most to the explanatory power of the model. Nevertheless, high effective discount rates are needed to reconcile the high level of Mincerian returns in the data. In the calibrated model, schooling contributes on average one-third to output per worker. ∗I am grateful to Virgiliu Midrigan for his exceptional work as a research assistant. I have benefited from comments from George Alessandria, Paul Evans, Jim Heckman, and Tom Mroz, as well as those received in presentations at Northwestern, Notre Dame, Ohio State, and the NEUDC 2003 and SED 2003 meetings. I thank the the Ohio State University for financial support. All mistakes are, of course, my own. †Contact: [email protected], 415 Arps Hall, 1945 N. High St., Columbus, OH, 614292-5588 (ph), 614-292-3906 (fax)