Litigation and Settlement in Patent Infringement Cases
by Claude Crampes (GREMAQ, IDEI and Toulouse) &
and Corinne Langinier (Iowa State)
FULL TEXT

Significance
A patent cannot protect an invention unless the patent holder can detect whether another firm has infringed the patent. The patent holder must monitor for possible infringement and this may be costly. This paper explores the effect of such monitoring on the actions of patent holders and potential challengers.

This is an important and often ignored subject. Researchers frequently assume that patent infringement can be detected easily and without cost. But in a real world of complex technologies, a single product may involve thousands of different patents and determining whether any one of these patents is infringed may require reverse engineering. This can be expensive. For example, Texas Instruments’ legal and engineering staff often spend a year to determine the extent of a suspected infringement [Grindley and Teece, 1997, p. 19]. Such costs raise a host of questions about how patent protection actually works. Do monitoring costs alter the effects of patent policy? Do patents give large companies an advantage because inhouse legal staff reduces monitoring costs? Is open source software at a particular disadvantage because reverse engineering is easy?

This paper makes a first foray into the subject.

Setting and Assumptions
The authors look at just a simple setting of a patent holder and a prospective imitator. The patent holder considers what resources to spend on monitoring infringement and the prospective imitator decides whether to enter the market or not. Given a level of monitoring, the patent holder only has a probability of determining whether the entrant has infringed. If infringement is determined, the patent holder has three possible actions: to ignore the infringement, to sue the infringer or to negotiate a license with the infringer. The paper considers cases where monitoring and entry decisions are made simultaneously and sequentially.

The authors make some limiting assumptions. There is no uncertainty about litigation outcomes (as in other litigation research); once infringement is detected, the outcome of a suit is certain. Also, antitrust regulations are assumed to prevent collusive licensing. And dynamic considerations, such as the effect of a reputation for aggressive patent enforcement are not included.

Results
Many of the results are straightforward and intuitive: imitators are more likely to enter with lower legal costs and with greater product differentiation (allowing more profits for the entrant); entry is less likely when monitoring costs less.

But one counterintuitive result also emerges: within a range, larger penalties for infringement may actually increase the entry of imitators. This is because higher penalties encourage licensing rather than litigation (they increase the bargaining surplus). If potential entrants know that the patent holder will offer them a license, then the threat of a suit cannot credibly prevent the imitator from entering. Monitoring costs greatly extend the range of circumstances for which this is true. So with monitoring costs, stronger penalties for infringement may not have the presumed effect. Even in this simple setting, strong infringement penalties may work against the goal of patents to increase innovation incentives.

References
Grindley, Peter C. and David J. Teece. 1997. "Managing Intellectual Capital: Licensing and Cross-Licensing in Semiconductors and Electronics," California Management Review, v. 29, pp. 8-41.