Richard A. Epstein: Two ways of viewing IP

Listen to this article

00:00
00:00

James Boyle: Disclaiming Liability?

Here are two ways to think about intellectual property. The first views it as a subset of property relations generally, to which the general principles of property law apply. Alternatively, it counts as a special kingdom of its own, governed by rules that operate at right angles to ordinary property rules. In truth, any exhaustive treatment of the subject will reveal both points of similarity and tension between the two fields. But the safer approach starts with the happy assumption that sound rules of property law will on average carry over to IP. That proposition is emphatically true in the debate that pits property rules against liability rules in structuring property arrangements.

This basic distinction is as old as the common law. When faced with clear violations of property rights, judges always have to choose the appropriate remedy for the innocent party. Typically, in the name of complete and perfect relief, they issue injunctions that stopped the violation and award cash payments for interim damage. An alternative strategy, which could be adopted even for deliberate wrongs, awards damages equal to the plaintiff’s loss, but allows the plaintiff to keep the property taken. In general, most legal systems opt for the first of these legal regimes. The key consequence is to deny the defendant a private right of eminent domain - taking property upon payment of just compensation - in order to force a voluntary transaction, which generates mutual gains, between the parties.

A great 1972 article of Guido Calabresi and Douglas Melamed dubbed the first of these two regimes as one with a “property rule,” and the second as one with a “liability rule.” It hinted broadly that a liability rule - where property is taken for a fair valuation against the consent of the owner - might be preferable to a property rule because it prevents a current owner from “holding out” for an exorbitant payment for a resource to which it attaches little or no value. The owner of scrub land, for example may deny the right of passage to the owner of a nearby mine, if some enormous payment is not forthcoming. The liability rule eases that problem, even if it creates another - the valuation of the interest that is lost.

The common law balked even in this case, in the absence of legislative authorisation. This extreme example shows how far it tilted in favour of strong property rights. Only cases of abject necessity upset the usual rule that transfer of property interest requires mutual consent. That strong presumption in favour of property rules has, however, come under increasing academic attack by IP scholars who favour a liability rule as a way to overcome the holdout problem, especially with respect to pharmaceutical patents. In these cases, abject necessity is not the motivating factor. But the use of a liability rule for drug patents has a different motivation. The patent system encourages the patent owner to price its product above marginal cost in order to recover the fixed costs of its initial investments, which by some estimates run to over $800m per new chemical entity. That high price excludes from the market users who can pay for the marginal cost of their own pill, but not the monopoly price. The liability rule is said to create the best of both worlds. The patent owner gets the fair value of the patent in a lump sum, so its incentives to produce are left undisturbed. The individual consumer pays only marginal cost, so no one is excluded.

Unfortunately, however, there are no free lunches in this game. Think of the alternatives: condemning the patent outright today imposes intractable questions of valuation not found with scrub lands. No private voluntary agreement for patent sharing relies on an outright sale; all are sharing arrangements by way of licenses filled with terms. The once and for all valuation is too uncertain, given that entry by new drugs, changes in US Food and Drug Administration warnings, discovery of new adverse side effects or unanticipated uses can completely wreck the initial valuations. Better it is to keep patents strong, and to encourage innovation of new drugs that are substitutes for the incumbent, which won’t happen if low-ball estimates for compensation are used.

Bad as the outright purchase system is, worse has been proposed. Recently, a Committee in the District of Columbia has reported out a measure (on which I have consulted with the Pharmaceutical Research and Manufacturers of America) that require drug patentees to issue non-exclusive licenses to firms designated by the District for production drugs for District representatives. The firm, or firms, chosen would then go into competition with the incumbent in a program, with a political promise to save millions to District consumers. The valuation issues in these selective licenses are mind-boggling, even if one gets over the initial hump, that only firms that have received FDA licenses can make drugs, and then only after the initial license has expired. For our purposes, the key insight is that the family of liability rules in all its forms cannot achieve the economic nirvana of competitive pricing without imposing enormous administrative and incentive costs.

David Hume had it right long ago when he said that every sound system of property rights, whether for tangible or intellectual property, should preserve the settled expectations of all members of the community.

Richard A. Epstein is the James Parker Hall Distinguished Service Professor of Law at the University of Chicago, and the Peter and Kirsten Senior Fellow, the Hoover Institution


James Boyle: Disclaiming Liability?

How do we judge whether it a good or bad idea for poor countries to be able to issue compulsory licenses for AIDS drugs? Should we allow anyone to use the basic tools of genetic research on payment of a flat fee? Is the response to “biopiracy” of genetic resources in developing countries to grant access, but require payment? Each of these proposals has something in common: they give an innovator the right to payment for innovation but not a right to exclude consumers and competitors from the innovation. Some of these rules are called “compulsory licenses.” Others are called “liability rules.” The terminology varies but the underlying idea is neither unusual nor novel. In the US, both the arms industry and the recording industry work comfortably around significant compulsory licenses and liability rules.

This idea is attracting interest now because it promises to give us the advantages of intellectual property rights without some of their disadvantages. Say one is worried about those with monopolies over technologies with strong “network effects” exploiting their control into dominance in other markets. Right now we deal with this through cumbersome antitrust actions. For the future, one answer might be to create a rule that required the monopolist to grant access to competitors but to be paid for it.

So Epstein is to be commended for focusing attention on the question. His answer to it is that liability rules are bad ideas. In any particular case, they might well be. But what arguments support the general conclusion? Epstein offers three. Intellectual property is just like tangible property. I am not sure I agree. In any event it is irrelevant because we have these rules in both areas. That leaves two. Liability rules will upset settled expectations, and are anyway bad because we should leave valuation to the market.

But when are these rules being imposed? Three situations need to be distinguished. First, the state gives you a patent, and then surprises you by imposing a compulsory license with no basis in current law. Here Epstein’s expectations critique is well taken; an unpredictable pattern of retrospective interventions may deter future investors. Second, there is a compulsory license option in the law before you invest. The state says that under certain circumstances – a public health emergency, for example – it will impose a compulsory license. You invest under those conditions. That is a known “risk” of uncertain magnitude; there might be an epidemic or an anthrax attack. There might not. Markets are pretty good at dealing with risks like those. Richard’s critique applies weakly, if at all. Third, there is the situation where the state sets up a system that prospectively applies a liability rule; if you want to produce a greatest hits of the 70s album you select your songs, and pay your standard fees under section 115 of the Copyright Act. It is in this third group that most of the scholarly proposals he refers to actually lie – for example those of my colleague at Duke University, Jerry Reichman. Here the expectations critique is completely irrelevant.

That leaves the question of valuation. This is a real concern. In general, we prefer markets to set prices. But while important, this concern is not disabling to every proposal. The other dangers – the costs of monopoly control to future innovation, the resolution of patent thickets, the human cost of monopoly pricing – may or may not outweigh it. The question is whether the costs of this kind of liability rule pricing in this circumstance will outweigh its benefits – for current consumers and future innovators. That is a hard question to answer, but the answer is not always ‘yes.’ Sometimes, exclusive rights are needed to provide the incentive. Sometimes, we need liability rules to grant access and prevent blockages. And sometimes the whole body of work should be in the public domain in the first place.

back to the top

Copyright The Financial Times Limited 2017. All rights reserved. You may share using our article tools. Please don't copy articles from FT.com and redistribute by email or post to the web.