| Abstract: | This article estimates the returns to education for households who derive part of their income from household based non-farm self-employment ventures in rural Peru. While education is an individual level variable, earnings are observed at the household level. This asymmetry complicates both the estimation and the interpretation of the returns to education. This article is the first jointly to incorporate three channels through which education affects household earnings. Education affects earnings through the marginal productivity of labour (worker effect), labour allocation across activities (between-activity allocative effect) and its production externality effect (spillover effect). The results suggest that the between-activity allocative effects of education dominate the returns. This article also makes novel use of economic geography to proxy for the role that access to markets plays in determining these returns. In particular, altitude is a strong predictor of activity choice and the returns to education in this mountainous country. |
