This interview is part of a series following the Supreme Court's Bilski decision, which left the laws on what you can get a patent on largely as they are, after a four-justice minority failed to ban "business method" patents.
Columbia Law School professor Eben Moglen heads the Software Freedom Law Center (SFLC) and wrote the group's amicus brief in Kappos v. Bilski. [PDF]
Moglen's position on the subject of software patents—that they should be banned—is, to say the least, outside the mainstream in legal circles. It has, however, garnered support among software developers and other techies, especially those who work in the world of open-source and free software.Moglen's critique of the patent system extends well beyond the software issues he writes about, however. He suggests, for instance, that the 20-year monopoly granted by a patent is the product of a bygone era. And though he rejects the notion that he is "anti-patent," he says that the patent monopoly grant should be subject to a rigorous cost-benefit analysis, not simply handed out at the "monopoly window" that he believes the current Patent and Trademark Office represents.
The Prior Art interviewed Moglen about his views on the Bilski decision, the future of software patents, and how patents affect innovation.
Q: What's your opinion of the Bilski decision?
The decision is one of those that clearly shows why it has come down on the last day of the term after being argued early. When you see a case like that come down, one of the things that occurs to you is that a majority has fallen apart. The justices agreed about less than they thought.
Everybody has to adjust to the fact that a question we believed was going to be decided didn't get decided after all.
In the case of Bilski, the question was: "How does patent law work in the twenty-first century?"
The job of patent law was an eighteenth-century job—encouraging people to immigrate to the U.S., bringing skills with them. This was a big continent which the white people here thought of as "empty." Patent law—the way the Continental Congress wanted to make it, the way the Constitutional Convention saw it—was a way of rewarding inventors.
In the nineteenth century, that eighteenth-century system worked well. America filled up with people who used technology to tame the natural environment.
Right now, we have a patent system that's extensively used only by two industries: information technology and pharmaceuticals. Anybody else, the patent system doesn't help them that much. Most of the world's manufacturers now use trade secret law. [And] the two industries that use patent law heavily use it for reasons irrelevant to its formation.
In the information technology [IT] industry we live in now, where software [is often] made by people who want to share rather than own, patent law imposes very significant problems for innovators. That patent law doesn't maximize innovation in software has been the observation of engineers and lawyers who work in the part of the world I work in, sometimes called the free and open source software community. But what it really is is part of science. Knowledge about the world is collected and organized for other people to learn from. Unlike almost all other forms of science, because writing it down in computer terms can also be writing down a program a computer could execute, basic scientific communications are made the subject of ownership by current patent law, unintentionally.
The Congress that in 1954 admitted for thte first time the idea of patenting processes didn't mean by that act you make it possible to own mental steps or algorithms or facts of nature. They didn't mean to make mathematics patentable. They were talking about the kind of process the Federal Circuit was talking about in making the machine-or-transformation test.
Q: And what about the pharmaceutical industry?
Pharma uses the patent system because the ownership of molecules, which is practically what they've reduced patent law to, is unbelievably profitable. It is largely exploited by pharmaceutical companies and causes enormous harm in the form of unnecessary human death in order to reward the greed of investors in pharmaceutical companies.
But it's not good for the human race to allow the ownership of molecules. It's not the best system for creating excellent drugs, including excellent new drugs.
Q. That sounds like a much broader critique of patents than what's in your brief. But Congress knew that patent rights would keep drug prices high, for a time—that's the balance it sought to achieve with the Hatch-Waxman act in 1984.
Pharma uses the results of socialized research [from taxpayer-funded sources such as grants from the National Institute of Health]. Without government subsidization of basic research, developed pharma couldn't exist. It takes the output of government spending, and monopolizes it by finding molecules it can own. It's a very peculiar use of patents but it's the most important use of patents in the world.
Pharma has thousands of bad patents that are costing america hundreds of millions of dollars a year. Pharma has gone beyond the limits of the patent law. They use it in ways that are profitable to them, but not authorized.
But you won't publish that. If you publish that, you'll get a hostile call.
Q: Back to Bilski, If the Federal Circuit's machine-or-transformation test had been kept as the exclusive test, would that be a preferable situation? Would that have achieved SFLC's goal of a world without software patents?
Properly applied, [the machine-or-transformation test] would have the effect of putting several hundred thousand existing patents on software out of patent scope, and would have substantially reduced the ability to issue software patents in the future. It would particularly have had the effect of eliminating large numbers of bad software patents, which, when you get down to it, are nothing more than an algorithm.
Everybody on the Supreme Court agrees that Rand Warsaw's patent application was bad. There are tens of thousands of patents presently being used to clog innovation and prohibit useful work that depend upon the precise kind of claim that is the claim in the Warsaw patent.
What the Supreme Court did in Bilski—if you take seriously the language of Kennedy's opinion—is what they did with pornography in 1975. That was the case when they decided, they knew it when they saw it. Once the court decided that they know it when they see it, they had to sit every year and look at the all the pornography in the U.S. and decide whether they knew it or not.
Q: What if the minority opinion authored by Justice Stevens, who sought to ban patents on "business methods" but garnered only four votes, had won a majority? Would we be better off, in your view?
I do not know what it would have said and what it would have done. It's not a finished opinion yet. If it had been law it would have been "something else." The "something else" that would have been added would also have been material.
Bilski represented the Supreme Court's opportunity for leadership in this area. They declined.
You're asking me, what if they had taken leadership? Then, history would have unfolded, with the Supreme Court in the lead.
[The Supreme Court said] we all agree this is a crappy patent that can't issue but we don't agree about why.
That's not good judging really. It's not what you want. They should have done what judges do, which is make decisions. Instead they did what football teams in trouble do. They punted.
Q. But it seems there was a vigorous debate about the value of a ban on "business methods," vague as that may be. Would a ban on business method patents have been a step in the right direction, in your view?
These categories, software patent, business method patent, they're not categories of patent law. They're categories of discussion. The real question isn't "Are business methods in or out? Is software in or out?"
What Americans need to understand is that the government has opened a window where rich people can get monopolies that they can use like real estate. "You daren't have my idea! I own that idea."
When monopolies are given away that cheaply, people will try to use it.
Patent law is ancient. But in the twentieth century, we should do cost-benefit analysis. We require that when the government comes into the economy to build or road, or a fishery, or to drill for oil in the Gulf of Mexico. We expect regulation to be based on an educated decision that it does more good than harm. We expect to be able to test that as a cost-benefit matter beforehand.
Patent law doesn't do cost-benefit analysis. It assumes that the agency can't figure that out. It's a nineteenth-century administrative system. Patent law treats the benefit of the twentieth-century monopolies as infinite. Patent law is unique in its enormous influence over the public by regulation.
Patent law is about manufacturing property for owners. It's another example of government privilege turned into property. It doesn't need to be abolished, or recreated—it needs to be subject to the test of the public interest.
Q: As someone who advocates limits on software patents, and believes they harm innovation, do your goals look farther away after the Bilski decision?
I don't know. We have not told people that we thought litigation was going to bring an answer to their problems in the U.S. Supreme Court. We were hopeful, and optimistic, that some justices wanted to reform the patent law. And there were some.
Q. Trade groups representing software businesses, like the Software and Information Industry Association, say they need software patents. If software patents are bad for innovation why do so many small companies pursue them?
If you are a publicly traded company, responsible to your shareholders for managing their wealth, how could you say anything other than that?
The people who own those corporations are freed from any obligation other than the interest of their shareholders. That's capitalism. It's a system for freeing people from the public interest. They defended their property. I don't know what else they could have done or should have done.
That makes sense to me. IBM doesn't want a lot of people getting software patents. They're happy, because their property is safe, and other people's property is to be negotiated.
In the end [software patents] won't serve the public interest. But in the near term, the managers of the businesses have to behave in the interest of the businesses, not of the public.
The problem we have in patent law isn't that IBM isnt doing its job. It isnt that Pfizer isn't doing its job. The problem rests with legislatures and judges who have abandoned the public interest.
Q. Again, this sounds like a much broader critique of the patent system.
Every day I see either clients of my own, or others [working in free software] who are intimidated into not posting code on a website....People are intimidated out of giving learned papers or speaking at conferences. Money is demanded from people for distributing software from which they make no money. People are prevented from describing their own ideas, that they had themselves, on their own website. All because some officious intermeddler claims he has "ownership" of the idea.
Those things happen every day, because the public interest is not protected in the patent law. They happen so that people who claim to have property can hold it to the public's disadvantage. They posture themselves as having the moral entitlement of "owners." But the words "my patent" say: the government orders you to pay me. And the question arises about the substantive justice of that order.
Q: Yet when I speak to companies—even smaller companies who feel victimized by patent trolls—even they aren't asking for broad reform. They're not mad about the structure of the patent system, they're mad at the troll who is attacking them.
Look, right now, there are people all over the U.S. who lost their houses, their jobs, and their pensions. Their hope for the future and their stake in the future has been destroyed—and they're mad at a Mexican.
Why are you surprised that's what politics does? It helps people to get mad at the wrong people. Power exists because it can get people mad at other people.
One of the beauties of the patent system is that it creates anger at competitors rather than at the structure of ownership itself. It causes people to fall to fighting over little gold coins. it causes them to forget that there is a public interest, or that they are part of it.
....................
More Post-Bilski coverage:
- Supreme Court Decides Bilski: Stevens and Allies Try to Ban "Business Method" Patents, but Fail
- Peter Brann and Internet Retailers: Why Some Tried, but Failed, to Ban "Business Method" Patents
If you talk to people about startups, they'll tell you that software patents are worthless to them. Companies will just violate your patent or work around it and they can copy your idea pretty easily. i4i was pretty much an outlier.
What VCs look for are things like a good team with strong domain expertise, which competitors can't acquire easily.
They don't care that much about how many patents you have. Yeah, they do affect your valuation, but that only matters if you fail and they have to sell off the company piecemeal. They'd much rather invest in a company they believe will be successful than one they can dismantle for a tiny profit, after all.
Posted by: Q | August 05, 2010 at 12:04 PM
How the hell does this buffoon get off reading the minds of the Supreme Court Justices. Reds get a free pass I guess.
Posted by: Abe Alony | August 05, 2010 at 07:18 PM
Abe, you ought to consider who you are talking about before you spout such rubbish. Moglen is an acknowledged expert in the area with years of experience. Slavish devotion to the views and judgments of the Supreme Court is precisely the kind of attitude which maintains the sort of unsatisfactory status quo that has endured for the last few decades..
Posted by: Laurence | August 06, 2010 at 03:02 AM
I disagree with Q.
A good patent is the only way a startup can prevent a Google or Microsoft from just copying their idea and putting them out of business.
The number of new ventures in which domain expertise is a sustainable advantage is, in my experience, vanishingly small. It is much easier (and cheaper) to hire a domain expert than to overcome a well-drafted patent.
It is silly to conclude that all software patents are a plague on society, simply because one can site examples of abuse of the system. Perhaps the recent economic crisis should lead us all to conclude that owning a home is evil as well??
Posted by: anonymous | August 10, 2010 at 01:19 PM
@anonymous: Patent cases require two things: years and hundreds of thousands of dollars. Now, guess what startups don't have.
Even if the startup won in court, just watch them try to commercialise a product when a bruised Goliath with thousands of patents is just waiting to get back.
The need to acquire and license patents is a tax on development. In some fields, there might be benefits that make this worthwhile, but the participation costs in software are so low that such taxes end up being a relatively huge cost. Further, the per-unit royalties clash with the zero marginal cost of software and break many software distribution models.
If you make an innovative word processor and it doesn't visually resemble existing word processors or it can't read and write the file formats commonly used, it's simply not a functional word processor. You can patent your innovative features, and the established companies will just skip those features, and they can patent their features and you're blocked from implementing something that users consider essential. There's no public interest case for patents in software.
Posted by: Ciaran O'Riordan | August 10, 2010 at 05:41 PM
@anonymous: Not only do startups lack the resources to enforce patents, as Ciaran points out, but their hand is usually tied by counter-claims.
Single patents are not effective. Patents which are narrowly scoped enough to stand up in court tend to be easy to avoid. Expansive patents which are hard to avoid are more costly to obtain, since you may need to make many attempts, and they tend to be rejected in review or get narrowed during litigation.
The only effective way to use patents is to create a minefield of mixed narrow and broad patents covering a whole subject area. This is a game that only the biggest players can play effectively.
If you're working in an area that a company like Microsoft is interested in they almost certainly hold _thousands_ of patents potentially covering the fundamentals of your field. Simply reviewing the patents to determine your exposure can be costly enough to put you out of business. If you enforce your patent claims against them the counter-suits will be devastating. Even if you are clear of actual infringement there will be enough 'close calls' that they can sink you with litigation long before you can demonstrate your innocence in court.
The question anyone should ask here is "Did I go into business to innovate or did I go into business to litigate?" If the answer is the former— the patent system is not helpful for you. If the answer is the latter then there exist better business models which involve no real invention (and thus no risk of infringement counter-claims).
Posted by: Gregory Maxwell | August 12, 2010 at 11:22 AM
If you have a novel idea, then obtaining a patent does not take hundreds of thousands of dollars. It may take time, but not so much that I would forgo the filing.
If your startup is successful to the point that a Google or Microsoft wants to copy you, then you can afford to assert your patent in order to protect your rights.
I agree that you may lose if Google or Microsoft decide to assert every patent they may own, but why would they bother with the expense? Out of spite? They are businesses looking at the bottom line.
Most likely a cross-licensing deal can be reached. If you have a patent, then you can bargain. If you do not have a patent, you get NOTHING.
Posted by: anonymous | August 16, 2010 at 09:40 AM