This paper addresses two questions: The first is What are the benefits of social science research?; the second is "How should they be measured?" The response to the first is that, as with research in the physical sciences, the benefits should be identified in terms of changes in economic surplus for different groups. It may be useful to use a framework that considers the incidence of the effects of social science research on firms, households, and govenment agencies. The response to the second question is that estimating returns to social science research using conventional econometric techniques may be particularly difficult. Instead, it may be necessary to resort to a case study approach, but care must be taken to ensure that the cases selected for study are genuinely representative.