Courts and commentators are sharply divided about how to assess "reverse payment" patent settlements under antitrust law. The essential problem is that a PTO-issued patent provides only a probabilistic indication that courts would hold that the patent is actually valid and infringed, and parties have incentives to structure reverse payment settlements to exclude entry for longer than this patent probability would merit. Some favor comparing the settlement exclusion period to the expected litigation exclusion period, but this requires difficult case-bycase assessments of the probabilities of patent victory. Others instead favor a formal "scope of the patent" test that allows such settlements for nonsham patents if the settlement does not delay entry beyond the patent term, preclude noninfringing products, or delay nonsettling entrants. However, the formal scope of the patent test excludes entry for longer than merited by the patent strength, and it provides no solution when there is either a significant dispute about infringement or a bottleneck issue delaying other entrants.
This Article provides a way out of this dilemma. It proves that when the reverse payment amount exceeds the patent holder's anticipated litigation costs, then under standard conditions the settlement will, according to the patent holder's own probability estimate, exclude entry for longer than both the expected litigation exclusion period and the optimal patent exclusion period, and thus will both harm consumer welfare and undermine optimal innovation incentives. Further, whenever a reverse payment is necessary for settlement, it will also have those same anticompetitive effects according to the entrant's probability estimate. This proof thus provides an easily administrable way to determine when a reverse payment settlement is necessarily anticompetitive, without requiring any probabilistic inquiry into the patent merits. We also show that, contrary to conventional wisdom, patent settlements without any reverse payment usually (but not always) exceed both the expected litigation exclusion period and the optimal patent exclusion period, and we suggest a procedural solution to resolve such cases.
Reverse payment patent settlements have led to widespread legal controversy. In such settlements, the patent holder agrees to make a payment to an allegedly infringing potential entrant (called a "reverse" payment because traditionally settlement-payment flow was from alleged infringer to patent holder) and the potential entrant agrees to stay out of the market until a later date.1 Such settlements have anticompetitive potential because they can exclude entry for longer than the expected litigation exclusion period, which would have reflected the often significant likelihood that the patent holder would have lost.2 Indeed, unless constrained by the risk of antitrust liability, settling parties would (no matter how weak the patent) always have incentives to set the settlement entry date at the end of the patent term because that maximizes joint profits (by precluding competition for as long as possible), and they can use the reverse payment to split those joint profits in a way that leaves both better off. However, if antitrust liability could be designed to prevent settlements that exclude entry for more than the expected litigation exclusion period, then reverse payment settlements could theoretically avoid litigation costs without causing any anticompetitive effect.
Such reverse payment settlements have been a huge issue in the multitrillion dollar pharmaceutical industry. But the issue is even bigger than that because reverse payment settlements can occur in any market where the patent holder would have greater market power if the entrant were excluded.3
The federal courts of appeals are in utter conflict on when reverse payment settlements violate antitrust law. The Sixth Circuit has held that reverse payment settlements are per se illegal. …