Abstract: | Estimation of the demands for many services supplied by government and charitable organizations are hampered by two practices common to the supply of these services. The suppliers often employ non‐price rationing of these services, and they price discriminate. The supply of college training, for example, is rationed on the basis of academic ability as well as willingness to pay, and more able students are typically quoted lower net prices (associated with scholarships) than less able students. This paper suggests a method for dealing with both practices in the analysis of cross‐sectional data. This methodology is used to investigate the question of why university faculty members are expected to do research as well as teach. One answer supported by our empirical work is that the customers (e.g., the students) demand it. Thus, controlling for price and non‐academic features, better students will choose to attend a university where more scholarly research is performed. Moreover, our empirical findings also support a strong negative link between faculty time devoted to teaching and the supply of research. Copyright © 2003 John Wiley & Sons, Ltd. |