Monday 29 April 2013

Patents and Cumulative Innovation: what is the impact of Federal Circuit patent invalidations?

"Patents and Cumulative Innovation: Causal Evidence from the Courts" is  paper by Alberto Galasso (University of Toronto) and Mark Schankerman (London School of Economics and the Center for Economic Policy Research, CEPR).  This paper will shortly be distributed as CEPR Discussion Paper 9458, looks really interesting  The authors' concluding comments are as follows, with citations omitted:
"In this paper we estimate the causal effect of patent rights on cumulative innovation, using patent invalidation decisions of the [United States] Federal Circuit. The identification strategy exploits variation in the propensity of judges to invalidate and the fact that the three-judge panels are generated by a random computer algorithm. There are three key empirical findings. First, invalidation leads to a 50 percent increase in subsequent citations to the focal patent, on average. Second, the impact of patent invalidation is highly heterogeneous, with large variation across patents and technology fields. Third, we find that this effect only occurs between patents owned by large firms that appear to block small innovators. Thus, invalidation of large firm patents ‘democratizes’ innovation by small firms.

These findings suggest that some licensing deals are not taking place in the presence of patent protection. There are two main reasons why this might occur. First, it might be optimal for a patent owner to restrict access if licensing reduces joint profits (e.g., because it intensifies downstream competition). Second, information asymmetry and uncoordinated, multilateral bargaining can lead to licensing failures even when it would increase joint profits. It is important to distinguish between these explanations because they differ in terms of their implications for welfare and policy (even putting aside the effect on consumer surplus). Our empirical findings help to do this.

The impact of patent invalidation is concentrated on a small subset of patents which have unobservable characteristics that suggest the presence of asymmetric information that induces bargaining failure in licensing. Our results also help to pin down where the bargaining failure occurs. The effect is concentrated in fields characterized by two features: complex technology and high fragmentation of patent ownership. This reinforces the market failure interpretation, since earlier studies identify these features as key determinants of licensing breakdown ... . We find no evidence of blocking in non-complex fields such as chemicals, mechanical, and pharmaceuticals. [...]

Overall, our findings show that patent rights block cumulative innovation only in very specific environments, and this suggests that government policies to address this problem should be targeted. However, scaling back patent rights may not be the most appropriate policy. Theoretical models of cumulative innovation show that such policies have ambiguous effects on overall innovation incentives.[...] It may be preferable to design policies and institutions that facilitate more efficient licensing and thereby promote cumulative innovation ...".
This blogger simply does not have the economics background to enable him to comment on the paper. He will only confine himself to making the usual caveat that data is always a problem.  Litigated patent disputes are always going to be easier to identify and analyse because of their public nature, and patents that are the subject of disputes which are appealed to the Federal Circuit cannot even be assumed to be typical of litigated patents as a whole. It would be interesting to match research on US data with any available data from the more litigation-averse Europe, to see whether the findings here are reinforced or weakened and, if so, how and to what extent. It would also be good to know whether the authors' conclusions chime in with the expectations of the economics-friendly European Commission as to what the effect of patent invalidation decisions of the new Unified Patent Court will be.

1 comment:

Anonymous said...

The paper clearly asks very interesting questions about patent invalidation. However I think the authors need to be wary of overanalysis of data. For example a recent IPWatchdog article (http://www.ipwatchdog.com/2013/05/02/google-we-dont-sell-to-patent-trolls/id=39927/) alleges that Apple and Microsoft will not even consider licencing until a party begins litigation. That's an irrational and unhelpful strategy which is really caused by humans being human. Analysis of human behaviour often assumes a rational basis which is unfortunately is not present, and so is ultimately flawed.