texte original http://eupat.ffii.org/10/03/bessen/
Jim Bessen: Do Patents Work as Property?
Review of a Lecture at Duke Law School
Jim Bessen and colleagues found by statistical analysis that innovators are nowadays, unlike 20 years ago, losing more money by patent litigation than they are gaining from patent royalties. Bessen correlates these findings to changes in patent law which made the boundaries of patents more fuzzy.
1 Replace “Fuzziness of Boundaries” with Breadth and “Boundary Ascertaining Costs”!
In his lecture at Duke Law School, Jim Bessen presents the findings of a recent work which we discussed on our mailing list a while ago. We found this that this work unprecedentedly high quality evidence to the debate about patent reform, but had doubts about the message regarding some of the conclusions drawn therefrom. In particular in the lecture Jim Bessen seems to overestimate the importance of the question of clarity vs fuzziness of the boundaries of invention-based property claims.
As an example of fuzzy boundaries, Bessen cites the Freeny patent which its owner E-Data Inc used after it had been dormant for 15 years to squeeze money out of thousands of e-commerce developpers. The patent claims contained an abstract notion “point of sale” which some courts construed as referring to dedicated music vending machines whereas others said that it covered any computer to which digital content might be downloaded. Other spectacular cases of this kind include the patent that British Telecom asserted against hyperlinks and the “Rembrandts in the Attic” that were asserted against the JPEG standard.
In the discussion after the lecture, some members of the audience express doubts about the correlation of the findings from the statistical analysis with the appearance of fuzziness due to abstract claims. It is suggested that
- * The scope of exclusion granted to patent holders has always been unstable and that this is thought of positively (as not a bug but a feature of the system) by people who argue for the introduction by additional fuziness in terms of variable durations of patents.
- * The patent system is a system that works well as long as it is not being put to serious use; when many people obtain patent rights and get serious about litigating, the system breaks down.
- * The development of the patent system is subject to cycles of technological development; patent are used intensively at the end of a cycle, when companies have accumulated large patent portfolios and have lost much of their innovative dynamism.
It is tempting to argue that only property rights with clear boundaries are well-functioning property rights. We at FFII have argued, somewhat differently, in favor of “fast, cheap and narrow” intellectual property rights. We never asked for “clear” IP rights, nor for “high quality patents”. According to our line of thinking, copyright is better than patents, and patents for inventions in applied natural science tend to be better than software patents, and one of the arguments against software patents has been that, due to the abstract nature of software they lead to function claims which tend to foreclose an unoverseeably large field of later innovation, as happened e.g. in the cases of Freeny, BT Hyperlink, JPEG and many others. It has been pointed out that it was the admission of function claims that paved the way both to software patenting and to patenting of large unexplored fields of future technology. Yet, even so, function claims are not necessarily any less clear than traditional claims. The frequent lack of clarity in patent claims, functional or otherwise, seems to be a result of the imperfectness of human language rather than of the abstractness. What really matters here is not so much the lack of clarity about the breadth/extension of the claimed territory as the breadth itself and the high costs associated with ascertaining the exact extension, both of which are incommensurate to the ease and lightness of software innovation. Clarity comes in only at a secondary level.
Focussing the discussion on the question of clarity is tempting, because it allows you to capture the benevolence of that large part of the audience that is not familiar with the patent system but has very strong feelings about “property” being the basis of a free market that produces the wealth of nations. It also allows you to capture the benevolence of those who believe in optimised patent systems that cheaply produce “high quality patents” with clear boundaries and are looking for a reform that can achieve this goal. Having captured the benevolence of this audience, which comprises most current-day legislators, you can then proceed step by step to lead the underlying assumptions ad absurdum, but at least the present lecture doesn’t advance very far in this direction.
If you want to convince yourself that patents usually do not have clear and stable boundaries, you just have to read up on the discussions regarding the “doctrine of equivalents” (Äquivalenzlehre). A patent’s real scope of exclusion emerges only after several rounds of litigation. In the case of designs and copyright, the boundaries are even less clear, as anyone who has to do with litigation about derivative works will notice. What really matters is thus not so much the borderline-clarity as the breadth/extension of the claimed property realm. The broader an immaterial property realm becomes, the more worthwhile it becomes to spend time and money on ascertaining its boundaries. Only such broad property titles as patents come with explicit (and thus relatively clear) claims.
Thus we share the intuition of those in the audience who question the importance of clear property boundaries and point to other factors as being more important. In the case of patents there has always been the problem, which Bessen also mentions, that only 5% of them bring the revenues. These 5% are successful because their claims have, after a few rounds of litigation, turned out to be clear and broad. In other words, the patent system is by design an expensive game which converts inventions into property titles according to more or less arbitrary rules, which can lead to more or less clear and stable boundaries only after investment of tens or hundreds of expert manhours that in turn have to be paid for by the revenues from broad claims. The problem is exacerbated when this game is extended to fields where the broad claims, while still being commensurate to the high costs of the expert manhours needed to ascertain them (boundary ascertaining costs), are no longer commensurate to the low costs of producing and practising the claimed invention (production costs).
Future reforms of the IP system will probably have to acknowledge and address the fact that the boundaries of invention-based property titles need to be ascertained gradually by means of expensive litigation. Rather than cultivate an illusion of “cheap and high quality patents for everyone”, the reformers must see to it that it is rarely necessary to ascertain the boundaries and that this will normally happen only in areas where the boundaries ascertaining costs are commensurate to the production costs. Economic studies should refine their observations of the patent battleground so as to take account of elements such as claim breadth and boundary ascertaining costs, even if those are difficult to quantify. By doing so it will hopefully become possible to bridge the gap between the economist’s and the “patent practitioner’s” perspective on patent policy.
2 Influence on Europe
As the moderator says at the beginning, with reference to the European software patent debate, that James Bessen may have had greater influence on policy debates than any other scholar. The work of Bessen & Maskin of 2000 came in time to serve the detractors of the pro software patent initiatives of the European Patent Office and European Commission as an argumentation basis, and it was, just as the later work of Jim Bessen and his colleagues, the best-underpinned research work of its kind. As we know, the European Patent Office continues to grant software patents against the letter and spirit of the written law and the European Commission continues to work to make this practise enforceable throughout Europe. However, thanks to the research work of Jim Bessen and colleagues, they are no longer able to present their policy as a way of promoting innovation in Europe. That may not yet be a huge influence, but it is more than researchers can usually hope to achieve, and there is a potential for further growth.
3 Patent legislators in China & EU increasingly wary of broad claim interpretations
The Chinese patent law reform that entered into force last year were preceded by a phase of critical discussion, involving thousands of scholars, of the effects of the patent system as they had been emerging. The new Chinese patent law encourages judges to construe claims narrowly according to the literal wording together with the examples provided in the specification, thereby even reducing their scope in cases of opportunistic application of old claims that were based on a different context. They have also adopted rules such as allowing the imposition of an obligatory license when the patent proprietor fails to implement his invention in a way that satisfies market demands. It is to be seen to what extent these potentially far-reaching rules can be translated into judicial practise.
In Europe the view of UK courts which tend to adhere literally to claim wordings and reject the doctrine of equivalents has gaining the upper hand. After a few rounds of discussions in legal journals between British and German top level judges, the latter seem to have adopted the former’s reasoning, and use of the doctrine of equivalents has become a rare exception. Yet, due to the constraints of human language, literal interpretation can also not be a solution to the problem of lack of clarity. The problem that our top judges have been solving rather seems to be one of excessive breadth and an excessive patent hype fuelled thereby
4 Remedies against the current situation
Jim Bessen mentions a few possible measures that could help alleviate the current problems. Certainly the most helpful of them would be the reversal of some of the landmark decisions by which courts extended patentability since 1987 (and earlier). Another potentially incisive proposal is that the cost burden on the patentee, e.g. in terms of annual renewal fees, should grow steeply as time passes, so as to discourage the maintenance of “Rembrandts in the Attic” (i.e. patents maintained only for speculative purposes). Some others proposals are too modest to be of real help:
- * Post-grant opposition is not significantly different from invalidity litigation; just the fact that the patent office offers this service does not make it less costly in terms of manhours; more importantly, the resulting patents may be somewhat narrower and clearer, but they are also more dangerous weapons than the untested broad patents.
- * The issuance of non-infringement certificates issued by the patent office, such as currently done experimentally by UKPO, is a way of making the patent office take more responsibility for the patents which it grants, but the fact remains that the clarification of boundaries is costly and these costs will not change very much just because the service is offered by the patent office rather than by a civil court, as is done nowadays e.g. in Germany with an action for declaratory judgment of non-infringment (negative Feststellungsklage).
We have been advocating some more radical reforms that could move us toward a system “fast, cheap and narrow IP rights”:
- * Private parties who have found a patent claim set to be invalid can, by a pre-trial cease-and-desist admonition procedure (‘Abmahnung’ in Germany) request the patent owner to remove or correct his claims and to reimburse the costs of examination and admonition incurred by the sender.
- * Patent offices become mere registries, similar to internet domain name registries; examination becomes a non-obligatory service on a free market of insurance service providers (which may be required by courts as a prerequisite for certain enforcement measures)
- * Patent descriptions are immediately published upon registry; an initial set of claims is added 1 year later; these can subsequently be narrowed, but not broadened.
These proposals take account of the fact that patent offices are unable to provide the service of determining clear and stable boundaries for invention-based property titles at a reasonable price. Rather than burden every patent applicant or even the public with these unreasonable costs, it is better to place the burden only on those applicants who want to assert dangerously broad claims, to provide cheap and efficient enforcement to those right holders who are satisfied with narrow claims, and to reward those members of the general public who are able to sweep away patent mines or to provide validity insurance services at a competitive price.
- * Duke Law @ Youtube: James Besson - Empirical Evidence on Patents: Do They Work Like Property?
- * EUPAT mailing list for followup discussions
- * Influential papers of Jim Bessen in this area that we previously discussed
- o Bessen & Meurer 2008: Patent Failure – How Judges, Bureaucrats and Lawyers put Innovators at Risk
- o Bessen & Hunt 2003
- o Bessen & Maskin 2000