Who (if Anyone) Should be Liable for Injuries from Generic Drugs?

Two recent Supreme Court decisions (PLIVA, Inc. v. Mensing (2011), and Mutual Pharmaceutical v. Bartlett (2013)), have essentially removed the threat of liability from generic drug manufacturers. In this paper, we consider four possible liability regimes in a simple model of drug market competition and safety research. Specifically, we compare a baseline regime of No Liability (NL) with the Everyone Liable (EL) regime (generics face the same liability as branded manufacturers), the No Liability for Generics (NLG) regime, resulting from PLIVA and Mutual (generics face no liability), and a novel Brand Fully Liable (BFL) regime (branded developer faces liability from injuries caused by a generic version of a drug it has developed). We find that the BFL regime typically provides the most efficient incentives to identify side effects and develop an efficient warning. However, the BFL regime can lead to overconsumption of the generic drug by patients who should not take the drug at all. For this reason, the EL regime may be preferable for a drug where the danger of side effects may outweigh the clinical value, as was alleged in Mutual Pharmaceutical v. Bartlett . We find that the NLG regime that resulted from the recent Supreme Court decisions is unlikely to be optimal, because it is dominated by BFL when the consumption decision is not important, and inferior to EL when it is very important.