The Disintegration of Intellectual Property?: A Classical Liberal Defense

Richard A. Epstein - New York University Law School

Posted in , ,

The Conceptual Counteroffensive Against Private Property

This basic theory of property rights identifies three central components of private ownership:  the rights to possess, use, and dispose of property. Taken together, these elements facilitate the creation of complex voluntary arrangements to coordinate the activities of multiple actors. The operational success of this system, however, has not spared it from major attacks by those writers who think that it leaves the state with too little control over resources. I address two of these critiques—by Thomas Grey and Margaret Radin—in this Editorial.

Grey. Thomas Grey’s famous essay, The Disintegration of Property, attacks the traditional notion of property: “It seems fair to conclude from a glance at the range of current usages that the specialists who design and manipulate the legal structures of the advanced capitalist economies could easily do without using the term ‘property’ at all.”1 The “modern specialist” instead views property as a “bundle of rights,” which Grey dismisses as more “shadowy” conception.2 To illustrate the confusion, he notes the ambiguity that resides in the simple statement that we “deposit our money in the bank.”3 The asset holder no longer has, of course, ownership of the discrete dollars in question, but a “set of abstract claims against an abstract legal institution.”4

These points offer in some sense an accurate description of how the classical theory of property rights works. But they do not in any sense count as a serious criticism of its intellectual coherence or economic worth. Quite the contrary, bank deposits, like trusts and corporate shares, are all created by a set of contractual movements that satisfy one basic condition. These institutions arise because the gains from trade exceed the transaction costs of putting these deals together. So long as we have accurate rules of conveyance and recording, so that all individuals know who holds what fraction of the original bundle at any particular time, there is no reason to be anxious over the philosophical question of “who owns what.” On those occasions where it may be necessary to identify an owner, as for the purposes of real estate taxation, a simple social convention can handle the problem. Let the party in possession of the land bear the responsibility for the taxes during the period of occupation, which will of course switch upon foreclosure—itself a procedure designed to allow for the reunification of property interests in the lender in the event of default. Far from showing why property systems labor under some deep conceptual confusion, rightly understood, Grey’s argument paves the way for understanding the adaptability of this system of property rights to technological advances.

Radin. A similar analysis explains the shortfall in Margaret Radin’s criticisms of my views of eminent domain, as developed in Takings: Private Property and the Power of Eminent Domain. Radin decries my analysis for its excess reliance on the notion of “conceptual severance.”  That term refers to the strategic effort of a property owner to break a single piece of land into its constituent parts in order to increase the likelihood for obtaining compensation from the government for a “regulatory taking” under the rules of Penn Central Transportation Co. v. New York City. The current constitutional tests require that compensation only be offered if there is a substantial loss in value to “investment-backed” expectations that the owner has in the regulated parcel. That case thus gives rise to the well-known “denominator” problem, whereby the original landowner divides the original parcel into ever smaller slices of property, in order to make the government taking seem like a larger fraction of a smaller thing, thereby increasing the odds of receiving compensation under the Penn Central rule.

I have no disagreement with Radin’s description of the desultory tactics that private owners will adopt to secure compensation for regulatory losses. But it is a mistake to treat this outcome as the result of some dubious “conceptual” move by the defenders of private property. The real culprit is the takings doctrine itself, which refuses to recognize what the private law has always understood—that the loss of any incident or fraction of a large interest in property is a partial taking that should be compensated. The rule in these cases is that the government should have to pay for what it takes, and should not be allowed to escape its duty to compensate by pointing out that it has left something behind for the owner. Once constitutional law adopts the right public rules for deciding when property is taken, the incentive to divide property into smaller interests, conceptually or practically, disappears. The air rights, for example, that were taken in the Penn Central case remain fully compensable regardless of whether they were retained by the surface owner or sold off to a third party.

The great defect of modern takings law is that it does not track the sophistication of the private system of multiple interests in single things, but instead imposes artificial limitations on property rights in “things” that only increase the power of the state to take. That change in approach is not without dangerous social consequences. The ideal of the just compensation requirement was to insure that government coercion was used only in the service of positive sum games. The current “ad hoc” rules of compensation for regulatory takings routinely invite the government to initiate negative sum games from which social welfare suffers in both the short and long run.

Intellectual Property

The central thesis of the second portion of this Editorial is that the schemes of property rights developed for tangible forms of property carry over to intellectual property, most notably patents and copyrights.

Acquisition. First, it is important to acknowledge the one key difference between traditional forms of property like land and intellectual property: duration. For land there are no social gains that come from the artificial truncation of the duration of the ownership interests obtained by first possession. For patents and copyrights, however, there are substantial gains. In these cases the removal of the protection allows for open and nonrivalrous use by all individuals simultaneously, without having to incur the transactions cost of licensing the use of a patented technology or a copyrighted literary work. We cannot of course go the whole way toward a public domain regime that removes all protection for writings and inventions, because this would eliminate the main incentive for private investment in creating these technologies or works. But the peculiar nature of information goods strongly militates to an intermediate term of protection. That term should be shorter for patents than copyrights, given that scientific advances are generally more easily replicable with time than literary ones. Elisha Grey invented the telephone about the same time as Alexander Bell, perhaps sooner. No one was about to write Hamlet if Shakespeare had fallen ill.

Exclusion. Once the duration of intellectual property rights is properly limited over time, the general strong rules on the rights to exclude and dispose carry over to the law of intellectual property. On exclusion the usual rule allows the owner of the property to repel any entrant who wishes to make temporary or permanent use of a protected thing. This rule avoids the need for owners to have to deal with multiple trespassers, who in the absence of injunctive relief would have every incentive to bypass the voluntary arrangements that allow for efficient cooperation between various rights holders. The same arguments apply to the patent world, so that the 2005 decision in eBay v. MercExchange represents a large step backward in the development of a coherent set of patent rules. The ad hoc nature of the eBay tests increases uncertainty as to the level of protection, which in turn diminishes the incentive to innovate in the first place.

The danger is especially acute with nonexclusive licenses: under the Supreme Court’s standard, the licensor may not be able to obtain an injunction because he does not practice the patent, while any particular licensee may not be able to obtain the injunction because others practice the patent. True disintegration of the property system is a real possibility if potential licensees think it more prudent to circumvent the licensing system and take their chances on damage litigation in a post-eBay world. To be sure, the pre-eBay law did put in place all sorts of limitations on the injunctive right. One simple device is the ability to postpone injunctions to allow infringers to make the needed adjustments for small components of some large assembly. Another is to deny injunctions to patent holders guilty of laches and estoppel. Once these are in place, there is no reason to gratuitously weaken the structure of the overall system.

Alienation. Recent case law developments have also undermine the powers of disposition, thereby compromising the effectiveness of voluntary licenses. Licenses are, if anything, more important for intellectual property than they are for tangible property. Intellectual property usually forms one component of some larger venture, whose success depends on the incorporation of multiple devices created by all sorts of people. The difficulties of up-front valuation usually drive most patentees to commercialize through nonexclusive licenses subject to complex royalty arrangements. Even the firms that practice their own technology will tend to license it as well.

It is therefore incumbent to place as few obstacles to the licensing process in order to maximize the gains from trade. Placing licenses in pooling agreements should of course be subject to the standard antitrust prohibitions against horizontal cartelization, but otherwise assembling patent portfolios should be understood as an important technique for countering the gridlock problem.

It follows therefore that carrying over the usual rules that allow for the disposition of tangible property makes sense. This decision has three concrete applications. First, the general rules of estoppel by license that apply for tangible property should carry over to intangible property. Contrary to the current Supreme Court law on this point, a patent licensee should not be allowed to use information that it acquired from the licensor to attack the validity of the patent. It is too costly to adopt a rule that slows down licensing across the board in order to increase the odds of discovering a few invalid patents whose validity can always be attacked by a stranger.

Second, it is unwise to insist that the copyright or patent owner only have direct claims against the party with whom it does business. In particular, the “first sale rule” as it applies to copyrights needlessly forces the original owner to recast sale transactions as licenses in order to impose restrictions on reuse that should be allowable in either form of transaction. One risk, for example, is that the inability to impose contractual restraints on alienation will lead to a reduction in, for example, promotional efforts. It is therefore imperative to allow these restrictions to be put in place, so long as clear notice against further resale or licensing (which is also required in land transactions) is also given in copyright sales and licenses. The same rules should apply to patents, where no “inherent” reason explains why third persons with notice cannot be required to pay royalties. Most patentees have little reason to incur the costs of seeking royalties from multiple parties. But in those cases where it happens—as when it is necessary to coordinate individual with blanket licenses—no reason of public policy should thwart those voluntary arrangements.

The same logic makes it unwise to impose any legal restrictions on efforts to assign or sell rights of action for the violation of intellectual property rights. The legal restrictions imposed in these cases only hamper the efficiency of the overall system by making it harder for some firms to specialize in development and others in collection, which is a common practice for other types of assets. The legal rules need only guard against the unilateral splitting of a cause of action, which could result in the harassment of a debtor by multiple creditors. That practice is not permitted in ordinary contract cases and should not be permitted here.

Takings. The final piece of the intellectual property puzzle involves the exercise of the eminent domain power over patents. In this context, it is also appropriate to treat all “physical” and “regulatory” takings under a single framework. The artificial Penn Central distinction between government use and government restrictions that fails for tangible property is equally disastrous here. Any restriction on the use or alienation of intellectual property has to be either justified or compensated. These justifications are hard to find for patents whose use generates few or no nuisance-like externalities.

It is unwise therefore to carry over the standard current distinction between physical (or per se) and regulatory takings into intellectual property. The danger is especially dangerous because the narrow definition of physical taking could negate the use of per se compensation rules for all forms of intellectual property. To avoid that risk it is best to treat any effort by the government to use, or allow others to use, some intellectual property as a per se taking. This alteration of constitutional doctrine narrows the class of regulatory takings, by analogy to land, to situations where the government does not use the property itself, or authorize the use of the property by others, but only restricts the owner’s use. One virtue of this approach is that it condemns the ill-fated Check 21 bill, whereby banks sought to use valid patents held by nonbank firms without royalty on the ground that their uses were permissible so long as the patent holders could license out their technology to other parties. Any repeated application of this principle dooms the entire patent enterprise. Strong property rights against government expropriation are thus also needed in the intellectual property realm.


The origins of the modern property system date back to the Romans, if not earlier. These early solutions have been subject to ceaseless modifications on small points over time. But the basic distinction between common and private property has held firm over centuries, as have the early rules for the acquisition, protection, and disposition of property interests in tangible property. The success of this system is confirmed not only by its ubiquity and durability, but also by its utilitarian justifications which have taken far longer to uncover.

Once the operation of this system is understood in the physical realm, it is capable, with some modification, of serving as the basis for the law of patents and copyrights that only become feasible with the rise of modern technology. The best approach first adjusts the duration of the property interests, which allows the law to carry over to patents and copyrights the traditional rules on acquisition, exclusion, and disposition of property interests. Yet, ironically, just when the tools of economic and social theory begin to supply the needed explanations, critics, who are anxious to expand the scope of government power, have launched wholesale but misguided attacks on the  system of private property. The practical success of their efforts has already had negative impacts on the overall efficiency of our economic system, even though their theoretical objections to the classical liberal system of property rights are without merit. The obituaries that are written about its coherence and utility are premature and should be disregarded. Further progress does not depend on jettisoning traditional property principles. It depends on the conscious decision to strengthen modern technology within the traditional forms of property rights.


Copyright © 2010 Stanford Law Review.

Richard A. Epstein is the James Parker Hall Distinguished Service Professor of Law at the University of Chicago, the Peter and Kirsten Senior Fellow at The Hoover Institution, and a visiting professor of law at New York University Law School.

This Legal Workshop Editorial is based on the following Law Review Article: Richard A. Epstein, The Disintegration of Intellectual Property?: A Classical Liberal Defense, 62 STAN. L. REV. 455 (2010).

  1. Thomas C. Grey, The Disintegration of Property, in Property: NOMOS XXII 73 (J. Roland Pennock & John W. Chapman eds., 1980).
  2. Id. at 69.
  3. Id. at 70.
  4. Id.


  • From the article:

    “We cannot of course go the whole way toward a public domain regime that removes all protection for writings and inventions, because this would eliminate the main incentive for private investment in creating these technologies or works.”

    This is a statement of faith. They are imposing this restriction on how people use inventions and imitate, and then they pull this out as a justification … literally, out of their hat. There is no empirical evidence for it. In truth, copyrights and patents started out way way way less intrusive then they are now. They grew out of control precisely because they are unjust at the core.

    For example, consider a government that had no rules other than forbidding the word “fuck”, well in order to secure that “right” they would also need to monitor our speech and internet, but then people would use encryption, so then the government would need to have the right to our keys, or to forbid encryption, but then we could use stenography, so then the government would need to outlaw it. But then people could use symbols and “code words”, so then the government would need to regulate what gestures people could make, and what words we are allowed to use. … and on and on and on.

    Copyright and patent grow out of control for similar reasons, and a constant economic pressure for those that have “ip”, to expand it’s scope.

    But the core fallacy is this: property rights do not exist for the sake of profit or incentive. Profit and incentive are a common side effect of property, not an ends in itself Property exists so that humans may resolve differing opinions and conflicts over limited resources in a rational and respectful and peaceful way. Without property rights, all that is left is “might makes right” philosophies. With information and invention, the limiting factor is not the information itself, but the time one puts into discovering it. So a person is totally within their God given right to bill by the hour for their invention or creative services, but not entitled to control how information and knowledge is used all over the planet for the sake of their “incentive”.

    Posted by argoff, 07.20.10 

Post a Comment (all fields are required)

You must be logged in to post a comment.