The paper suggests that typical estimates of returns to primary education are over-estimated, because import costs to individuals are excluded. In calculations with Honduran data, private returns are found to drop significantly when private costs are included. It is suggested that lower private...