Overeducation and wages revisited: A two-cohort comparison and random coefficients approach

摘要

We examine the effect of overeducation on wages by comparing two cohorts from the 1979 and 1997 U.S. National Longitudinal Survey of Youth panels. Using an econometric technique uncommon to the literature, we allow for overeducation to have a disparate impact on wages across individuals by employing random slopes models. Overeducation has a positive marginal effect on wages. Yet, cohort comparisons reveal that the returns to overeducation declined dramatically over time. In the past, surplus schooling in full-time jobs returned nearly as much as the correct level of education. Presently it returns approximately 50% less than in the past. The effect of undereducation and required education on wages changed as well, but by far less.