Why trash don't pass? pharmaceutical licensing and safety performance of drugs

This paper examines how asymmetric information in pharmaceutical licensing affects the safety standards of licensed drugs. Pharmaceutical companies often license potential drug molecules at different stages of drug development from other pharmaceutical or biotechnology companies and complete the remaining of research stages before submitting the new drug application(NDA) to the food and drug administration. The asymmetric information associated with the quality of licensed molecules might result in the molecules which are less likely to succeed to be licensed out, while those with greater potential of success being held internally for development. We identify the NDAs submitted between 1993 and 2004 where new molecular entities were acquired through licensing. Controlling for other drug area specific and applicant firm specific factors, we investigate whether drugs developed with licensed molecules face higher probability of safety based recall and ultimate withdrawal from the market than drugs developed internally. Results suggest the opposite of Akerlof's (Q J Econ 84:488-500, 1970) lemons problem. Licensed molecules rather have less probability of facing safety based recalls and ultimate withdrawal from the market comparing to internally developed drug molecules. This suggests that biotechnology and small pharmaceutical firms specializing in pharmaceutical research are more efficient in developing good potential molecules because of their concentrated research. Biotechnology firms license out good potential molecules because it increases their market value and reputation. In addition, results suggest that both the number of previous approved drugs in the disease area, and also the applicant firms' total number of previous approvals in all disease areas reduce the probability that an additional approved drug in the same drug area will potentially be harmful.