VoxEU Column EU policies Productivity and Innovation

Improving efficiency in the 'market for innovation'

The EU is considering two major policy initiatives to improve patent enforcement – patent litigation insurance and establishment of a centralised patent court. This column studies the US experience in centralising patent decisions and concludes that a European Patent Court would likely bring clarity to patent rights, making the market for innovation more efficient.

Patent rights can be a valuable incentive to innovation if effectively enforced. In practice, enforcement can be costly, uncertain, and create substantial delays in the use and transfer of technology. Research on patent litigation shows clearly that exposure to litigation is not random – litigants target more valuable patents, creating an unlevelled playing field for small firms (Lanjouw and Schankerman, 2001, 2004). To make enforcement less costly, more certain, and more even, the EU is considering two major policy initiatives to address these concerns – patent litigation insurance and establishment of a centralised patent court.

European patents are currently litigated in the national courts of the jurisdiction in which alleged infringement occurs. Because countries differ in procedural laws and other aspects of patent enforcement, outcomes of infringement cases for the same patent can differ across jurisdictions and be very uncertain. In the current environment, as Lemley and Shapiro (2005) emphasise, property rights in patents are highly probabilistic.

During the past ten years, European states have worked intensively on a European Patent Litigation Agreement aimed at providing more uniform treatment and reducing uncertainty in patent enforcement. The current draft of the agreement includes the establishment of a European Patent Court that would have exclusive jurisdiction in cases dealing with infringement and revocation of European patents. If implemented, this institutional change will closely resemble the establishment of the Court of Appeal of the Federal Circuit (CAFC) that took place in the US in 1982. As with the proposed European Court, the creation of CAFC was intended to bring uniformity and predictability to patent decisions (Gallini, 2002). Did that occur with CAFC? Would a European Patent Court improve patent enforcement?

The US experience with a centralised patent court

In recent research, we analyse the impact of CAFC on the settlement of patent infringement disputes in the US (Galasso and Schankerman, 2008). We are interested in settlements because an effective market for innovation requires that patents disputes are resolved as quickly as possible. In fact, because it is common for patents to be licensed as part of settlement agreements, delay and uncertainty in the settlement process mean slower diffusion of patented technology. Moreover, longer delays would typically be associated with higher transaction costs for the negotiating parties. This is especially important for small firms for whom patents are often their most important asset, and the ability to license or sell them effectively is critical to preserving their innovation incentives and access to venture capital finance.

To study this issue, we use comprehensive data on the timing of settlements in patent disputes filed in US courts during the period 1975-2000. We find that there was a sharp decline, of about 40%, in the length of patent disputes for cases filed in the lower (federal district) courts after the formation of CAFC. The average delay between filing a suit and settling it declined from 29 months before CAFC was established to only 17 months.

Moreover, this reduction in dispute duration is associated with a decline in the fraction of cases reaching final adjudication at trial. Prior to the introduction of CAFC, 17.2% of all patent suits reached final court adjudication, as compared to only 5.9% afterwards. This reduction occurred in all technology fields. At the same time, we observe a substantial increase the number of patent suits filed each year. These facts suggest that the observed reduction in dispute duration is due to earlier settlements and not to an increase in the rapidity of court decisions.

We interpret this decline in settlement time as a consequence of the greater clarity of property rights which the centralised appellate court created. Many scholars have argued that CAFC generated a distinct `pro-patent' shift that took the form of tougher evidentiary standards to invalidate patents and increased likelihood of large damage awards. Our view is that this pro-patent shift at the appellate level facilitated settlement of disputes in the lower courts by reducing the uncertainty about damage awards.

Before the establishment of CAFC, there were sharp differences across the 10 different appellate jurisdictions (circuit courts, in American parlance) in their enforcement of patent rights. Henry and Turner (2006) show that there was a lot of variation in the frequency of validity and infringement findings, both across circuit courts of appeal and across district courts within any given circuit. If our view of CAFC is correct, the biggest drop in dispute duration, after CAFC was established, should be registered in circuit courts where the uncertainty over damages was the largest. This is exactly what we find: the estimated impact of CAFC is almost twice as large for the four circuits where the pre-CAFC variance in court outcomes was larger.

Patent thickets

We also study the interplay between the centralised appellate court and fragmentation in the ownership of patent rights (often referred to as `patent thickets'). Fragmentation of patent rights is widely viewed as an impediment to innovation because it can increase transaction costs, settlement delays, and the likelihood of bargaining failures in negotiations over technology licensing. The classic statement of the problem goes back to Heller and Eisenberg (1999), with particular focus on the biomedical field. Fragmentation of patent rights actually has two distinct effects on patent dispute settlement delays. The first, and most obvious, is that dispersion in the ownership of patents increases the number of licensors with whom bargains have to be struck, and this increases overall delay. But there is a second, countervailing effect: fragmentation reduces the value at stake in each individual settlement negotiation, and this makes settlement and licensing easier to reach. Thus on a priori grounds, we cannot tell whether fragmentation increases or reduces total settlement delay.

Our study shows that, after controlling for other relevant factors (including measures of patent value and age, the involvement of serial infringers and patentees, and other factors), patent suits do indeed take less time to settle when the alleged infringer is using patents in technology areas where ownership is more fragmented. The following table summarises the effect of fragmentation in a simple way, both for the pre- and post-CAFC regimes. The table shows, first, that greater fragmentation reduces settlement time per dispute, on average. Second, this effect was much larger before CAFC was established (that is, when uncertainty over enforcement of patent rights was greater).

Fragmentation and Dispute Duration
  Fragmentation Below Median Fragmentation Above Median
Average Dispute Duration    
Entire Period (1975-2000) 19.6 17.6
Before CAFC (1975-1981) 33.0 27.7
After CAFC (1982-2000) 18.3 16.4

What do these findings imply about how patent thickets affect the total settlement delay, taking into account both the number of disputes that end up in court and the expected duration of each dispute? The answer turns critically on how licensing settlement negotiations are sequenced. If licensees negotiate with the different patent holders at the same time, then even if they have to negotiate with more parties, the expected total delay will be smaller because fragmentation reduces the delay per dispute. However, if they negotiate sequentially, then there are two countervailing effects: more patent holders to negotiate with increases total delay, but fragmentation reduces delay per dispute. Our computations suggest that, even if negotiations are conducted sequentially (which is the worst case scenario in terms of settlement delay), it appears that total delay may have been reduced by fragmentation in some technology fields, at least before CAFC was instituted. But after the centralised appellate court was established, patent thickets do appear to have increased the overall settlement delay.

Making patent rights clear

This work emphasises the importance of maximum clarity in enforcing property rights. This by itself says nothing about whether we should favour patent holders more or less strongly (that is a more complicated, though critical issue), only that we need to have a steady and well understood judicial stance toward such rights. Our results also suggest that a European Patent Court, bringing uniformity and certainty to patent decisions, could have a substantial impact on the speed of technology transactions.

References

Heller, Mark and Rebecca Eisenberg (2001), "Can Patents Deter Innovation? The Anti-commons in Biomedical Research," Science, 280: 698-701

Henry, Matthew and John Turner (2006), "The Court of Appeals for the Federal Circuit's Impact on Patent Litigation," Journal of Legal Studies, 35: 85--117

Galasso, Alberto and Mark Schankerman (2008), “Patent Thickets and the Market for Innovation: Evidence from Settlement of Patent Disputes,” Centre for Economic Policy Research, Discussion Paper 6946

Gallini, Nancy (2002), "The Economics of Patents: Lessons from Recent U.S. Patent Reform," Journal of Economic Perspectives, 16: 131-154

Lanjouw Jean and Mark Schankerman (2001), "Characteristics of Patent Litigation: A Window on Competition," RAND Journal of Economics, 32: 129-151.

Lanjouw Jean and Mark Schankerman (2004), "Protecting Intellectual Property Rights: Are Small Firms Handicapped?," Journal of Law and Economics, 47: 45-74.

Lemley, Mark and Carl Shapiro (2005), “Probabilistic Patents,” Journal of Economic Perspectives, 19: 75-98.

210 Reads