--summarized by Robert M. Hunt* (Federal Reserve Bank of Philadelphia)

Summary
Antitrust authorities have been re-evaluating the desirability of “patent pools.” Patent pools usually arise when different firms join together to license their patents as a joint package. For many years, patent pools have been discouraged because they inhibit competition, but recent scholarship has highlighted circumstances where such pools may be socially valuable. Using a straightforward theoretical model, Jay Pil Choi points out that patent pools may have an effect not recognized in this recent literature: they may change the incentives that firms have to invalidate low quality patents.

The problem arises because the patent office sometimes grants patents that are not truly valid and will not be upheld in court. Patent pools can improve social welfare if most of these patents are properly granted. When many patents are erroneously granted, however, cross-licenses and patent pools reduce the incentive for firms to challenge them through litigation and this reduces welfare.

Choi concludes: “ . . . the most serious case arises when both firms have weak patents and do not have any incentive to challenge each other’s. In such a case, public policy should be geared towards providing incentives to challenge weak patents from other concerned parties.”

Details
Many theoretical models assume that patents are only granted when they should be and that their claims are well delineated. In practice this is not always the case.

Choi’s main analysis is based on an example in which two firms own patents on technologies that are necessary to produce a single good. These inventions are essential and complementary. In such a scenario, each firm can block production of the good simply by refusing to license their technology to the other firm. But firms can resort to litigation to both prevent infringement and to challenge the validity of a rival’s patent. When a patent holder engages in litigation, he takes the risk of having his patent invalidated, or its claims narrowed.

Choi first considers a case where both patents are unambiguously clear and valid. He shows that, in the absence of transactions costs, both firms would prefer to form a pool rather than setting individual licensing fees in an uncoordinated fashion. A patent pool reduces the total royalty fees charged to downstream producers and this reduces consumer prices.

When patents are no longer unambiguously clear or valid, the firms have the option to challenge the validity of their rival’s patent. Each firm is more likely to sue the other if it believes its own patent is of higher quality—i.e. more likely to be upheld—and its rival’s patent is of lower quality, and more likely to be invalidated.

Consider a case where, in the absence of a pool, either one or both firms are willing to litigate. So long as litigation is costly, the firms will prefer to form a patent pool over litigation, with the resulting profits divided according to the relative quality of the two firms’ patents.

Would it be better to prohibit a patent pool in this situation? Choi shows that a pool should be allowed if the two patents are sufficiently likely to be held valid. If patent quality is too low, however, prohibiting a patent pool will induce litigation that is likely to result in one or both patents’ being invalidated. While profits would be lower in this case, consumers are better off than with a pool.

Choi also shows that when the quality of both firm’s patents is sufficiently low, neither will sue. Instead, the two firms will form a pool and divide the resulting profits equally between them. In this case, prohibiting a patent pool cannot affect the likelihood that one or more patents will be invalidated. From the standpoint of social welfare, then, it is better to permit the patent pool.

In total, the outcome is hardly desirable. As Choi puts it, “we thus have a paradoxical result that a patent pool should be allowed and the monopoly preserved exactly when patents are most suspect and have little value.” Choi argues that this is rationale for creating incentives for third parties to attempt to invalidate patents.

Other Results
Choi presents several more interesting results. He examines the effect of a patent pool in a case where the underlying inventions are substitutes rather than complements. And he also considers the effect of a patent pool on the incentive of a second stage inventor whose improvement will infringe a patent contained in the pool. It turns out that a patent pool improves incentives for second stage innovation, possibly at the expense of incentives for inventors in the first stage.

Conclusion
Choi reveals a new interaction between patents and antitrust law. The quality and clarity of patents that are granted have implications for the circumstances in which cross-licensing agreements and patent pools should be permitted by antitrust authorities. If patent quality is low, these agreements may warrant additional scrutiny. If patent quality is too low, there is an argument for providing an incentive for third parties to dispute patents.

*The views expressed here are those of the author and do not necessarily represent the views of the Federal Reserve Bank of Philadelphia or the Federal Reserve System.