06-414 Ch 08.qxd

larger scale than that for which their estimated effects were originally obtained, occurs across many different applications, programs, and disciplines. Economic models of scale-up, which are the concern of this chapter, have focused on particular types of scale-up effects that occur frequently in interventions where economic outcomes are the major interest and where program beneficiaries—in this case, students, rather than teachers or schools—are usually the actors making the decision of whether to take up the intervention. However, because they have the ambition to provide a general model of individual choice behavior, economic models have a much wider applicability than to economic outcomes alone and to the decisions of students alone. Economists have indeed begun in recent years to apply their models to broader sets of outcomes and issues, such as the effect of educational interventions on noncognitive outcomes, the importance of peer effects within schools, and the effects of special education programs. However, this work is still relatively immature as a subfield and in addition, economists have done almost no work, and have developed almost no models, for certain types of scale-up effects, particularly those concerning the change in the nature of the intervention itself, the effect on which program operators most often focus. It is argued here that economics has nevertheless much to contribute to the problem of scale-up. First, the economic model of production processes provides a natural framework within which to discuss the problem of scaleup in general, and to develop a taxonomy of different types of scale-up effects. As an example of its usefulness, I argue below that it allows one to provide alternative explanations for one of the most common findings in