On the Failure of R&D Projects

There is an extensive literature on the success/failure of firm-funded R&D projects, but growing policy interest focuses on publicly funded R&D projects. We discuss this literature, which builds on a long stream of research of which Albert Rubenstein was a prime investigator, and use the literature to motivate the theoretical and empirical parts of the paper. Using data from 1878 Phase II R&D projects funded through the U.S. Small Business Innovation Research (SBIR) program, of which 624 had been discontinued prior to technical completion, we provide for the first time quantitative findings on the success/failure of publicly funded firm-performed R&D projects. Using a single-equation qualitative choice model, we find that prior R&D experience with the technology funded by SBIR projects, the amount of the SBIR award, and having a female as a principle investigator, other factors held constant, are all negatively related to the probability of project failure. In contrast, firm size is positively associated with project failure. We did not find evidence that human capital, experiential knowledge embodied in the firm conducting the R&D, or university involvement in the R&D project affected the probability of project failure. We discuss the implications of these findings for practice, policy, and further research.