Listen to this article
Richard A. Epstein: Must we live in an all-or-nothing world?
European geekdom is in a frenzy. The cause? A key vote on a proposed Software Patent Directive is due to be held in the European Parliament on July 6th.
Except, depending on whom you listen to, the directive is not really about software patents at all. Some argue that as its official name indicates, it is only about “computer implemented inventions.” They claim it will only cover systems that use software to control something in the external world, traffic lights say, but not software that just runs to be.. well.. software, such as an internet browser or a spread sheet program.
Opponents strongly disagree. They claim the directive not only allows expansive “pure” software patents, it also might allow for the toxic US innovation of “business method patents.” If a normal patent covers a new design for a corkscrew or a way of cooking burgers, a business method patent covers the idea of opening a bottle of wine, or of selling fast food.
The debate has played out noisily over the last few years, complete with accusations of procedural impropriety, intense arguments about legal definitions, national government statements pro and con, the appearance of “tech celebrities” (if that isn’t an oxymoron) and a level of intensity that one might expect in arguing the ethics of stem cell research but which seems strange in an area like this. What is going on?
On the substantive level, what is going on is a fight over the basic conditions of competition in the digital world. Both sides agree that intellectual property is important for innovation but they disagree about how much – at what point more rights start to hurt rather than help – and how enterprises of different sizes will be affected. A particular and very real concern is the effect of software patents on so-called “free or open source software” such as Linux which is created by a fascinating process of decentralised development.
It does not help the debate that well-intentioned people simply disagree about what the proposed directive actually means; an indication of the dangers involved in law-making about property whose boundaries are marked, not by fences or walls, but by vaporous ideas and concepts. A property right crafted with one goal in mind can spread, amoeba-like, to cover situations its drafters never intended. That was the certainly the experience in the US.
Which side in the European debate is right? One of the problems is that, as normal, the argument proceeds almost entirely without evidence. There have been some empirical analyses, but they have generally been conducted by advocates rather than policy makers, or have had methodological shortcomings.
A Business Software Alliance study apparently showing reliance on software patents in Europe by small companies was released and received stinging media criticism for statistical inaccuracies and for a press release which was interpreted – rightly or wrongly – as vastly overstating the results. An interesting German-commissioned study of a large number of software developers showed a consistent belief in companies of all sizes that software patents would harm rather than help the development of new software. This is a striking result but studies based on attitudes and beliefs are notoriously difficult and the author was careful to caution that the group might not be statistically representative. There are good studies which show that patents can act as a useful signal to investors, but also that they can allow dominant firms to block new entrants and developments.
In the absence of further evidence, sound bites prevail. Proponents of the directive are left claiming that “stronger rights will mean more innovation.” Opponents quote Bill Gates’ 1991 words about the expansion of software patentability in the US: “If people had understood how patents would be granted when most of today’s ideas were invented and had taken out patents, the industry would be at a complete standstill today.”
There is, however a very impressive empirical literature of the expansion of patentability on the US software industry that some European policy makers seem to have missed.
For example, Professor James Bessen and Robert Hunt of the Federal Reserve Bank found that the increase in the level of software patenting in the US was associated with a significant decline in investment in research and development by software companies. As more and more patents were granted, companies spent less on R&D. Correlation does not prove causation, as the authors appropriately caution. Nevertheless their conclusions are clear about the assumption that granting stronger property rights in software will stimulate innovation. “Our evidence suggests this assumption may be incorrect in the case of software patents. If, instead, the legal changes create patent thickets, the result might well be less innovation.”
Certainly that is what the numbers on R&D dollars suggest. Funnily enough, Microsoft gave financial support to two economists to write a critique of the article. Has Mr. Gates changed his mind about the anti-innovative effects of expanding software patents? The data indicate he may have been right the first time. The European Parliament should take note.
This writer is William Neal Reynolds Professor of Law at Duke Law School, founder of the Center for the Study of the Public Domain and a board member of Creative Commons and Science Commons.
Richard A. Epstein: Must we live in an all-or-nothing world?
James Boyle’s provocative column highlights once again, as if further illustration were needed, the divisive tendencies within the high-tech community over the protection of intellectual property (IP) rights. The use of the term” frenzy” suggests, misleadingly in my view, that the debate in question is one that has to be discussed in an all-or-nothing fashion. Either one believes that patents are necessary to spur innovation, or that the creation of additional patents necessarily creates a patent “thicket” which blocks the very innovation that the patent system is intended to spur.
One reason for the heated debate is that both sides have a point. The grand claims on both sides of the debate refer to two conflicting tendencies that both necessarily operate within the IP space. One man’s castle is another man’s roadblock, and the ultimate question often turns on the rate at which voluntary transactions can correct the imperfections in the intellectual property world. On this score, it is generally easier for one person to negotiate the license for the use of key IP than it is for an innovator to contract with the rest of the world to respect his IP rights to gain universal protection for his innovation. It is this fundamental asymmetry that in general drives us to property based solutions. But the conclusion is not so clear as it seems if the new innovator has to procure multiple licenses from, literally, hundreds of different sources in order to continue his own work. It is thus highly unclear as a general matter whether the level of software innovation will rise or fall with IP protection.
In the face of this massive uncertainty, it seems unlikely that either side has the full truth on the matter. If that is the case, then perhaps we ought to back down from the hard line rhetoric that has arisen over the proposed Patent Directive that will shortly come before the European Parliament. In dealing with this issue, Boyle notes that many opponents of the new directive cringe at the thought that it “might allow for the toxic US innovation of ‘business method patents.’” And herein lies a cautionary tale. There is no question that the expansive reading given to some business method patents at the time of the State Street case a half a dozen years ago, did strike terror in the hearts of many thoughtful observers. But since that time the fevered debate over business method patents has cooled down, as the courts have prevented the worst excesses from becoming law.
The point to remember here is that patent law is inherently a two tier system. The first tier asks whether software inventions or business methods are “patent eligible.” If not, then they are out from the patent system. But if found patent eligible, then any particular application has to meet the standards of utility, nonobviousness, novelty and the like. Perhaps the right solution in this debate is to recognize that both software and business methods are patent eligible, while casting a closer look at the particular applications made to see if they meet all second-tier requirements. It might not be a perfect solution, but it could lower the flame under a debate that now runs the risk of generating more heat than light.
Richard A. Epstein is the James Parker Hall Distinguished Service Professor of Law at the University of Chicago, and the Peter and Kirsten Senior Fellow at the Hoover Institution