This report focuses on several reasons for questioning the appropriateness of the standard measures of returns to education in deciding about education programs and evaluating their comparative advantages vis-a-vis other investments in developing countries. The goal is to provide some guidelines in selecting future research options in the field of education and development. This paper concentrates on two sets of issues arising on the "demand side" of education. The first is concerned with the demand for education by households, and with the demand for educated labor by the productive sector of the economy. The second issue focuses on the public, or "collective," aspect of the returns to education and is more speculative. The first of two parts of this document summarizes the mechanics of "educational transition" as this transition could be derived from a competitive model integrating the labor market and educational system. This general framework provides a reference to evaluate the biases that result from adopting a static or stationary approach to the measurement of educational returns. Explained are how the shortcomings of this approach can be overcome by several research orientations. The remaining part of the paper addresses issues linked only to purely productive externalities, the fundamental issue being that of identifying them empirically. Several research directions are investigated, one of which relies on the well-known Solow-Denison macroeconomic framework. An attempt is made to show the important analytical potential that lies in the availability in many countries of repeated cross-sectional surveys on earnings, occupation, and schooling. (40 references) (RR)