This cross-disciplinary, cross-institution collaborative research project combines economic analysis with state-of-the-art methods from statistical machine learning, to assess the relative efficiency and efficacy of research and development expenditures across the U.S. National Institutes of Health (NIH) portfolio of extramural projects. The novel combination of econometrics, topic modeling, and document classification permits analysis of massive collections of grant abstracts and scientific publications, identification of latent research topics present in NIH-funded research, assessment of possible spillover effects across research topics, and evaluation of causal linkages between changes in NIH funding by research topic and scientific advances. Two specific research outcomes are considered: scientific publications, classified by topic; and pharmaceutical innovation, measured by drugs entering into clinical development to treat specific diseases. Planned research also includes refinement of existing economic theory to produce normative evaluations of the allocation of public research spending.
Broader Impacts: This research will inform key policy questions related to federal funding of biomedical research. First, fiscal austerity requires careful attention to the nation's research portfolio and investments. This project will describe and evaluate the productivity of those investments, as a first step towards policy recommendations for rebalancing the portfolio, to maximize society's expected return on investment. Second, if NIH funding for basic research spurs increased pharmaceutical innovation, NIH possesses an important policy tool to promote pharmaceutical R&D in areas of high therapeutic importance or significant health disparity.
