Education policy is a fascinating and controversial subject. Part of the reason is that it is an area where jurisdiction is divided. The analysis and recommendations of professional educators are often quite different from those of economists, and purely political factors also playa major role, both because large amounts of public funds are used, and because access to education appears to be an important determinant of the distribution of income. Furthermore, as I will discuss below, even if we confine attention to the economic analysis of educational issues, there are different approaches with substantially different policy conclusions. Given what is at stake, in terms of resource use and the fundamental importance of education for the future of a society, it is very important that we “get it right” in studying educational issues. And for the policy-maker, “getting it right” doesn’t just mean making sure that our conclusions are consistent with the assumptions of our models; making sure that we are using the right model in the first place is even more important. There is, therefore, a strong need for empirical work that would enable us to discriminate between competing analytical approaches. Yet, as I will discuss later, education is an area where it has proved particularly difficult to use empirical observation for this purpose.