I’ve been really impressed with the job that Regulation, Cato’s quarterly journal of regulatory policy, has been doing covering patent policy of late. A year ago, I highlighted a fantastic exchange between legendary libertarian legal scholar (and Cato adjunct scholar) Richard Epstein and Berkeley law professor Peter Menell over the legal and philosophical status of the patent system. Epstein, drawing a close parallel between traditional property rights and patent rights, argued that courts should give patent holders the same kind of strong enforcement powers—including the power to obtain injunctions against infringers—that are available to the holders of traditional property rights. Menell, for his part, emphasized the differences between patents and traditional property rights, and argued that in light of the patent system’s various deficiencies, it’s a good idea to give trial judges wide discretion about whether to award injunctions or monetary damages in infringement cases.

In the latest issue of Regulation, two of my favorite patent scholars, James Bessen and Michael J. Meurer, contribute something that’s all too rare in patent debates: empirical evidence. They argue that at root, Epstein and Menell’s dispute isn’t so much a philosophical disagreement as an empirical one: do patents, in fact, operate in the same beneficial fashion as traditional property rights? That is, do they enhance legal certainty and increase incentives for innovation, or do they confuse and discourage potential innovators? This is a question that can’t be settled in the abstract; it must be answered by looking at the performance of real patent systems and seeing what effects they have in real industries.

Bessen and Meurer’s answer to the question “do patents promote innovation?” is “it depends.” In particular, there appears to be wide variation in the efficacy of the patent system across nations, industries, and time periods. Historically, the patent system appears to have worked somewhat better in the United States than the UK, although its performance in the 19th century was mixed in both cases. Today, the patent system appears to work well for the pharmaceutical and chemical industries and poorly for most other industries. In most non-chemical industries, the costs of litigation are so astronomical as to completely swamp the patent system’s benefits. That is, the threat of litigation due to inadvertent infringement discourages research and development more than the patent system’s rewards to inventors encourages it. If Bessen and Meurer’s data are right, then the public would be better off if those industries did not have access to the patent system at all.

Bessen and Meurer stop there in their Regulation piece, but in their book, they argue that the fundamental problem is that outside of the chemical industries, the patent system does a poor job of defining the boundaries of patent rights. That is, in industries like software, it’s difficult to determine which patents cover any given product or technology. It’s analogous to a physical property regime in which there were numerous, overlapping claims for any given piece of land, and no clear procedure for sorting out the true owner. Such a “property” system would not have any of the beneficial features that libertarians correctly attribute to well-designed property regimes.

Friedman prize winner Hernando de Soto made a name for himself by advocating reforms to third-world property systems to make them work more like Western property systems. I think we should regard patent reformers like Bessen, Meurer, and Menell as doing something similar: seeking to reform the patent system to bring something like the predictability found in traditional property systems. De Soto understood that until that can be achieved, it’s crucial that the old “property” system not be strictly enforced, because the laws on the books are so far out of sync with the facts on the ground. It’s not fair to a Guatamalan squatter to raze the home he’s lived in for a decade because some bureaucrat decides the land rightfully belongs to someone else. By the same token, it would be unfair to an innovator like Vonage to force it to shut down the Internet telephony network it has constructed because it accidentally violated one of Verizon’s overly-broad patents. Before insisting that people respect patents, we need to make sure that the patent system is respectable.