Red Hat is sponsoring a symposium today at the University of North Carolina School of Law on “Frontiers in Empirical Patent Law Scholarship.” The symposium features some twenty distinguished scholars with backgrounds in law, economics, technology, and other fields with research interests in the patent system.

The word “empirical” is a very useful bit of vocabulary, although it is not used much in everyday conversation. It means, roughly, based on experience or observation. An intelligent being newly arrived on the planet Earth might be surprised to learn that not every human institution is guided and improved by reference to observation and experience, and that an empirical approach to problems is by no means the norm. Sad to say, we native Earthlings are well accustomed to institutions that run without any reference to empirical reality, whether due to tradition, vested interests, or simple ignorance.

There is probably no governmental institution that could not benefit from a more empirical focus, but for the free and open source software (FOSS) community, the patent system is an excellent place to begin. The patent system is founded on a theory with little empirical support. One of the basic premises of the patent system is that innovation is promoted by offering a monopoly on inventions. The inventor, according to this theory, will be more inventive because he or she can obtain a patent that excludes others from making, using, or selling the invention. In return for a temporary monopoly, the inventor discloses the invention to the public, which can use such knowledge as a basis for further advances. In short, patents are thought to promote innovation. The premise is so familiar that it seems almost beyond question.

But the success of FOSS shows that, at least in the software area, the patent system is seriously flawed. FOSS innovations have changed the world of technology dramatically, but is implausible to suppose that they have been inspired by hope of receiving a patent. No FOSS developer wishes to exclude others from their invention by obtaining a patent. Instead, FOSS developers share their code freely. Moreover, the public gains nothing from a FOSS patent, because, by definition, the source code is freely available without reference to a patent disclosure. In terms of motivating FOSS inventors or spreading FOSS knowledge, the patent system contributes precisely nothing.

For the FOSS community, the problem is not so much that patents are failing to serve their intended purpose of promoting innovation. The history of FOSS shows that innovation happens in spite of poorly conceived legal rules. The real concern is that patents may actually hinder software innovation. There are now more than 200,000 software patents, and there is no practical way to be sure that a new program does not infringe one or more of them. A patent lawsuit can cost several million dollars in attorneys’ fees. The risk of patent litigation is one that the FOSS community has learned to live with. But it hardly seems likely that the risk is doing anything other than inhibiting software innovation.

The issue here is more than academic. The patent system we have today was based on a theory of innovation that, at least in the FOSS area, is simply wrong. If we want to encourage innovation, we need to test the theory with empirical research. We should bring the tools of economics and other social sciences to bear to clarify what forces encourage innovation and what forces hinder it. Where the patent system is not performing its intended function, what changes are needed? Are there effective incremental reforms for the software patent problem, or do we just need a new system?

Hail to the scholars who have undertaken empirical research on these matters! We want to encourage them and those inclined to join them in investigating the important issues relating to the system regarding software patents. The symposium is an excellent opportunity for us to learn from them. We hope and expect that the dialogue will be productive and will continue.