There is surprisingly little macroeconomic empirical research which would support a presumed link between education and development. I identify three major reasons why it remains difficult to estimate the economic relevance of education as a determinant of growth and development. First, most empirical research has ignored some of the crucial productivity aspects of education as proposed by new growth models. Second, measuring the contribution of education to economic development has largely ignored international differences in rates of return and the quality of education. Third, the allocation of resources within the education sector usually does not follow considerations of efficiency, which implies that additional spending on education cannot be expected to produce substantial output effects.