Europe's 'me-too' patent law: Copying US legislation on intellectual property rights threatens to inhibit software innovation, warns Lawrence Lessig
The Americanisation of European patent law continues apace, with Germany's Supreme Court moving closer to the US practice of granting patents for software. This autumn, the European parliament will consider a proposal to extend the rule to the European Union as a whole.
If the measure is approved, software in Europe, as in the US, will enjoy both the protection of copyright and patent law. But though the apparent beneficiaries of this change are software developers, those same developers are beginning to resist this expansion in their rights. Companies such as Adobe have opposed the change, as do many smaller developers. They fear patents will make coding far more costly, and that this cost may drive them out of the market.
When the beneficiaries of a government monopoly object to it, that should suggest that the politicians have got it wrong. But the lawyers pressing for patent expansion in Europe seem reluctant to listen, and Europe's parliament is racing to mimic the US without pausing to think.
Patents are a form of regulation. They are a government-backed monopoly that can extend for 20 years. This regulation is sometimes thought necessary to help spur innovation. Without it, ideas would be too easily stolen; investors, in turn, would not invest in new technologies. Patents, the argument goes, help create incentives to innovate.
But is such an incentive needed for software? The answer is not obvious. As economists have long argued, while it is sometimes true that a patent will spur innovation, it is not always the case. Patents are both an input and an output in the production process. Increasing patents thus may increase incentives, but it will also increase costs.
Whether on balance the incentives outweigh the costs is a hard question. In many contexts, they clearly will. But for at least some kinds of innovations economists agree that they will not. In those cases, expanding intellectual property rights will harm the incentive to produce.
So you might think that before Europe paints its patent law with the Stars and Stripes, someone would have studied the question whether there is any good economic reason to believe that software patents will induce more innovation. No such luck. The language of the European Commission is just lawyers' talk - about "certainty" and "consistency" and how software is "just like" any other technology.
There are good grounds to be sceptical about the proposed change. One important study released late last year concludes that software patents in the US actually harmed investment in software research and development. Technologist James Besson and Harvard economist Eric Maskin show that R&D in software actually fell after patents became common. Their study offers a powerful model to show why in this type of industry - where innovation is sequential and complementary - patent protection will slow innovation, not speed it. Other historical studies about innovation in similar fields have also demonstrated that in industries like software, innovation was greatest when monopoly protections were least.
There is one area of software development, however, where the consequence will be very clear. This is the effect of patents on the open source, or free, software movements. These movements produce software whose code is not controlled by any single entity. The source code remains free for others to take and develop; the only strong requirement (in some of the projects at least) is that the developers keep the source code free for others to build on later. The Linux operating system, launched by Linus Torvalds, building on Richard Stallman's GNU Project, and seen by Microsoft as its strongest competitor, is an example.
Many software producers in Europe have adopted the open source model for their projects, and the growth in open source in Germany and France is huge.
But software patents are particularly damaging for open code projects. With the source code available for anyone, the co-ordination costs in negotiating a licence become extremely high. Thus, open source coders in particular fear that patents will force them to close their code. Why pay for a licence on a bit of code that others can get for free?
If there were good reason to believe that patents would improve innovation in software, the arguments of the open code movement might be weak. But no good reason has yet been offered. The push for patents comes from lawyers, not technologists or economists. And the only sure effect of the change will be to increase the work for lawyers, as they increase the burden on technologists.
If patenting software will induce more innovation in software development, then let proponents demonstrate it, through careful and convincing economic evidence. But until they do, Europe should wait. Legislation needs a better reason than that lawyers like it, and that America does it.
The writer is a law professor at Stanford Law School and a fellow of the Institute for Advanced Study, Berlin