“The system is broken”: Larry Lessig on patents. And a riposte
From the Financial Times.
Lawrence Lessig: Exclusive rights to stagnate
By Lawrence Lessig
Published: February 20 2003 20:47 | Last Updated: February 20 2003 20:40
As pressure mounts on the European Parliament to extend patent protection to software, a crisis is developing in US patent law that Europe would do well to consider. The system in America is broken – to the great detriment of software developers generally – and there is no reason to believe the Europeans could do any better.
The claim that the US patent system is in crisis is nothing new. What is new is the identity of those making it.
From the start, Americans have been sceptical of the government-backed monopoly that is the patent. The first head of the US Patent Office, Thomas Jefferson, was a vocal critic of the very idea. Nonetheless, since the founding of the American republic, Congress has had the power to grant “exclusive rights” to “inventors” for their “discoveries”. And for most of that period the scope of those “exclusive rights” was limited.
But since the middle of the last century, both the Patent Office and the courts, encouraged by Congress, have increasingly expanded the range of “discoveries” subject to patent. Most recently and controversially, the courts surprised most practitioners by concluding that both software and “business methods” can be subject to patent protection.
This last shift took place formally in 1998, although clever patent attorneys had been effectively patenting software for many years before that. Critics had been arguing for some time that this extension of patent protection would do more harm than good. As Bill Gates, Microsoft chairman, wrote in a 1991 memo: “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.”
Yet throughout the administration of President Clinton the Patent Office insisted that the system worked just fine. Patents were being granted for truly novel inventions only, the office said; and innovators had no trouble in identifying who owned what invention. Claims that the system was in crisis were little more than the ravings of Chicken Little. The system would work itself out. It always had.
Yet now the Patent Office is singing a different tune. As its new head, former Republican Congressman James E. Rogan, said in an interview with the L.A. Times on February 7, 2003: “This is an agency in crisis and it’s going to get worse. It doesn’t do me any good to pretend there’s not a problem when there is.”
The reason is the mess created by the last administration’s patent office, especially in the context of business method patents (the type of patent, for example, that gives Amazon an exclusive right to its “one-click” method for selling merchandise online). “Some of [these] were fairly broad,” Mr Rogan told L.A. Times reporter David Streitfeld. “We’ve gone from a 75 per cent acceptance rate to a 75 per cent rejection rate.”
This early and easy acceptance rate led to an explosion in patent applications and patents granted – and, in turn, in the costs that software developers face. “Developing software is [now] like crossing a minefield,” says Richard Stallman, the originator of the free software movement that has developed the GNU/Linux operating system. “With each design decision, you might step on a patent that will blow up your project.”
This is the most surprising fact about software patents: they are generally opposed most strongly by the people they are intended to benefit. But such opposition is not difficult for a conservative like Mr Rogan to understand. Patents are a form of regulation. They represent a government decision on who gets a monopoly over what invention. Republicans like to claim that Democrats regulate first and ask questions later. They are therefore more eager to ask the right questions up-front.
Yet the questions here have no good answers. Like any form of government regulation, patents make sense only if their benefits outweigh their costs. The public benefit from patents is presumably the inventions that otherwise would not have been made. The costs include the extraordinary burden of knowing just what innovation is and is not subject to a government monopoly. These costs are borne both by innovators seeking a patent and by those just writing code. Both must wade through incomprehensible claims about who owns what inventions to avoid the inevitable hold-up if their code proves successful.
Software developers are quite aware of these costs. Yet economists have found it very hard to reckon any net benefits. And thus conservatives are increasingly sceptical of this form of regulation. No doubt it has produced “a whole cottage industry of shysters,” as Mr Rogan admits. It is harder to show that patents have produced any gain that would justify their costs.
The issue is not just a problem of implementation. The weakness runs much deeper. It may well be that software development requires some form of government protection. It does not follow, however, that patents are the protection that software needs. Software already receives the protection of copyright and trade secret. (The “code” of software is a kind of writing that copyright protects; and the properly hidden secrets that stand behind software can be protected like any other business secret.)
These too have their critics: the term of software copyright is effectively perpetual; and trade secrets tend to hide, not spread, knowledge. But if these forms of protection are inadequate or misinformed, then the solution is to find a form that better fits software. No one really believes that patents are well designed for this type of invention. Yet no government has adequately explored the alternatives.
American software developers will continue to choke on software patents, especially as more and more get enforced in massively expensive litigation. These patents will in turn inhibit the work of independent developers and protect large developers over small. As Mr Gates rightly concluded in his 1991 memo, “established companies have an interest in excluding competitors.”Patents give them just one more tool.
Until software patents prove themselves safe and effective, Europe could gain a great deal by sparing its developers the same drug. Rather than copying a failed American policy, the Europeans could be exploring alternatives to patents that might provide protection without sinking the intended beneficiaries. No doctor would approve an untested drug for his or her patient. Nor should Europe inflict such a remedy on its already weakened software industry.
The writer is a professor of law at Stanford Law School and author of The Future of Ideas: The Fate of the Commons in a Connected World
……………………………………………………………………………………………………………………
Richard Epstein – Patents underpin innovation
I think that it is important to part company with Lawrence Lessig?s latest FT.com column, which pronounces the US patent system for software broken. This is a harsh indictment of a system that to my knowledge shows no fundamental signs of breakdown. As I have often stressed in this column, I claim no technical expertise in the software that operates under the hood of my computer. But for these purposes ignorance is a kind of bliss, for it is quite evident as I sit working away on my OS X by Apple that it represents a major advance in convenience, versatility and power that could not have taken place if strong intellectual property rights had Balkanised the intellectual universe. One looks for fundamental flaws in underlying institutions only when the progress starts to grind to a halt. When the market seems to be vibrant, then what is needed is an explanation as to why it works.
Here is one candidate. New patents offer both obstacles and opportunities. As a matter of law, the patent gives one the exclusive right to make and market the patented invention for a limited period after it falls into the public domain. But the exclusive economic right that it provides does not necessarily provide an economic monopoly for its holder. Nor does it necessarily present an obstacle to individuals who wish to take their own new product to market. Granted, the first new entrant in some niche may well be able to exact some monopoly rent from the innovation. Yet by the same token, the person who purchases that innovation at the monopoly price is better off than he would have been if the technology had not been invented in the first place. Better to face a monopoly today than to have to wait six months or a year in order to get free access to a new product.
This monopoly scenario, moreover, overstates the dangers that patents hold to innovation. Quite simply, someone else could invent a second device that offers the functional equivalent to the first, so that the former customer of the monopolist now faces a duopoly, in which both firms charge lower prices than before. ?Inventing around? has both costs and benefits, but the difficulty of getting the right balance on the question is, I think, reason enough to preclude any harsh judgment that the current system is broken.
Here is one recent piece of evidence that just came to my attention. Recently, John Walsh, Ashish Arora and Wesley Cohen wrote a short article in Science Magazine (vol. 299, 1021, Feb. 14, 2003), which asked the analogous question of whether the much heralded patent thicket had placed obstacles in the path of biomedical science. The answer: not really. Extensive interviews with professionals in the field revealed that the landscape had indeed become more complex but that people could work through it. The authors reported that “almost none of our respondents reported worthwhile projects being stopped because of issues of access to IP rights to research tools,” such as various kinds of probes or reagents which may have no close substitutes. A similar form of ingenuity exists in the software business. “Inventing around”, patent pools and individual licences may not be perfect, but they in combination they undercut the force of the ostensible blockade.
Prof Lessig writes as though the rise of business method patents under the famous 1998 State Street decision marks a death knell for innovation. But in reality it does nothing of the sort. The case itself only decided that business methods were patent eligible, ie that they constituted a manufacture, process, machine or composition of matter. It did not decide whether that particular algorithm met the individualised tests of non-obviousness, novelty and utility. In this regard, it is little different from the decisions on medical devices, all of which recognise that something more than an unprotected algorithm is at stake when the machine in question uses external data as the basis for measurement and diagnosis of bodily conditions. The real action here should not be with respect to patent eligibility but with these second-tier questions that go to the validity or scope of individual patents. The same logic applies to software.
Prof Lessig?s column illustrates the big philosophical gap between us. I supported and applauded his unsuccessful efforts to upend the 1998 Copyright Term Extension Act because that statute represented an unconscionable giveaway of public domain assets to private firms, without any compensation in return. But there is no giveaway at stake in the ordinary cases of patent and copyrights under the current law. The commons works after the expiration of the patent or copyright because something of value has already been created under the incentives of the patent system. But if we treat all inventions as though they belong in some commons at creation, then in the not-to-distant future there were will be few patents or copyrights left to fall into the public domain. The current system has it about right. Tinker maybe, overhaul no.
