The role of cost-sharing as a signal of project quality in the federal funding of academic research: An application to the National Science Foundation.

摘要

One of the traditional justifications for the use of cost-sharing in the federal funding of academic research is an equity argument. Since universities accrue real, monetary gains from research, either directly from the generation of patent revenues or indirectly from the gaining of increased prestige that translates into increased tuition revenues or more external research funding, they should be required to pay their fair share of the costs. The focus of this thesis is to explore another potential justification for the use of cost-sharing, that cost-sharing provides information to the funding agency about the quality of the proposal. A theoretical model is developed to establish a framework for understanding the cost-sharing element in the federal funding of academic grants. The model predicts that in some funding environments, cost-sharing would allow the agency to weed out low type projects, leaving only high type projects in the proposal competition. Under looser agency budget environments, lower type projects could apply and be funded, but higher cost-sharing commitments would be required from all proposals which had lower preliminary assessments.;In testing these predictions, the empirical model employed data on 1168 continuing grants that were funded by the National Science Foundation between 1993 and 2000. The results provided very minimal evidence that cost-sharing varies negatively with the preliminary assessment. This evidence includes results from the 1993-1994 and 1997-98 periods, results from the individual NSF directorates, and the relationship between total cost-sharing and grant award. Thus, we cannot conclude with any certainty that there is any informational content in the cost-sharing level. However, the testing did show that greater cost-sharing is a function of the threshold of the project, whether this threshold derives from the university, department, or researcher level characteristics. This is supportive of the traditional leveraging explanation of cost-sharing. Agencies seem to act in a manner similar to that of a price-discriminating monopolist, requiring varying levels of cost-sharing (or “prices”) based on the ability of the proposer to provide cost-sharing. It was also found that cost-sharing was affected by agency’s threshold, which varied over time and across field or directorate.