Abstract
In this paper we reconsider the well-known paper of Angrist and Krueger (1991), who introduced a method to estimate the effect of education on income. The presence of latent capa¬bilities of individuals and measure¬ment errors complicate this estima¬tion, so that an ordinary regression is not reliable. First, we look at the approach and results of Angrist and Krueger (1991) for American data. Then, we take a closer look at these results: it appears that the reported estimate of a 10% increase in income per year of education is completely determined by data of a few Southern states. Given the large economic dif¬ferences between the regions of the US, the estimate may be far from representative for the average return on education across the US.
Original language | Undefined/Unknown |
---|---|
Pages (from-to) | 6-13 |
Number of pages | 8 |
Journal | Medium Econometrische Toepassingen |
Volume | 15 |
Issue number | 1 |
Publication status | Published - 2007 |