Human capital externalities and proximity evidence from repeated cross-sectional data

Does schooling generate social returns in excess of the private returns captured by the individual who makes the human capital investment? As a strategy to detect human capital externalities I use Dutch survey data to estimate the impact of the average human capital stock in a region on individual wages, considering regional human capital as a local public good. Indeed, the regional fraction of high-skilled workers or the region’s average educational attainment appears with a positive and statistically significant coefficient in an augmented Mincer specification. However, the impact on individual wages completely vanishes when the firm’s human capital stock is included as an additional control. This may suggest that human capital externalities predominate within firms, though alternative explanations, in particular selection of high-skilled workers in high-productive companies and imperfect substitutability across skill groups, cannot be ruled out.