Institutions and indirectness in intellectual property.

• Author: Smith, Henry E.

INTRODUCTION I. PROPERTY AND INFORMATION COSTS A. Property and Property Rights in the New Institutional Economics B. Information Costs and the Structure of Entitlements II. UNCERTAINTY, OPTIONS, AND ENTITLEMENTS IN INFORMATION A. Options and Modular Entitlements B. Managing Complexity Involving Information III. APPLICATIONS A. Licensing B. Remedies CONCLUSION INTRODUCTION

Institutions are important to intellectual property. Information is a major subject of exchange, and the special challenges of contracting over information have long been at the heart of economic theories of contracting. Exchanges involving information are difficult because a buyer will be reluctant to make a purchase without knowing what he is buying, but once the seller reveals the information, the buyer will no longer need to pay for it. (1) Contractors can also face challenges from asymmetric information, and some of the limits on people's ability to contract stem from the problems of incomplete information.

Where does this leave property? Although intellectual property is technically a form of personal property and some of its mechanisms are shared with regular property, commentators have found the notion of intellectual property problematic in a way that regular property is not. Because information itself is nonrival, in that one person's use of it does not diminish the value of another's use of the same information, excluding others from information when they could use it at zero marginal cost seems wasteful. And while intellectual property is one device among many that could provide incentives to people to create information, the nonrival nature of information is a count against intellectual property in comparison with rewards, kudos, lead times, and other alternatives to appropriating the returns from inventive and other creative activity. Should intellectual property be property at all?

One might expect the New Institutional Economics (NIE) to be helpful in addressing this question. In this Article, I will argue that the NIE is very helpful, but in a less straightforward way than one might think. The NIE shows promise because it is concerned with information and institutions, including "property rights" and the problems of contracting. But the concern in using the NIE to explain intellectual property lies in its generally thin notion of property. In the NIE, property rights are a subset of institutions, which comprise all the "rules of the game of a society," including both formal law and informal norms, all of which are "humanly devised constraints that shape human interaction." (2) But currently a very broad range of institutions would count as "property rights" in the NIE. The standard definition of property in the NIE is a right to engage in a set of actions with respect to a valuable resource, or the expectation of deriving utility from the use, exchange, or transformation of an asset. (3) This is fine as far as it goes, but it leaves several issues unaddressed. In particular, the NIE commonly adopts an extreme version of the bundle-of-rights theory of property, under which property is a collection of "rights, privileges, duties, and so forth," (4) with no particular defining characteristics. (5). The list of uses over which the owner has authority corresponds to this collection of sticks from the bundle-of-rights picture inherited from the legal realists. In the case of land, such sticks would include the right to build, the right to plant, the right to prevent development, and so on. Using this assumption of a list of uses, the NIE does not have much of an explanation for why property rights are, at their core, rights to things good against the world (in rem). This problem is especially acute in intellectual property. A system of rewards would be a "property rights system" in the NIE, but it would be one very different from patent, copyright, trademark, and the other intellectual property regimes with which we are familiar.

Because the NIE incorporates information costs and bounded rationality, it has the tools to provide a rationale for intellectual property as property. In particular, because assets are collections of valued attributes that are hard to measure, (6) actors face the economic question of how to group them--how to aggregate them, or more accurately, how to carve them up--for purposes of delineating property rights. In theory, one could build property up out of the smallest sticks, in an additive fashion. Thus, fee simple ownership of Blackacre would be delineated as the sum of the right to exclude, the right or privilege to farm (in many different ways), the right to park cars, the right to prevent development (nature preservation of various kinds), and so on.

There are two problems with this bottom-up, stick-by-stick, legal-realist approach to delineation, and both problems relate to how directly rights should make reference to valued attributes of assets and the uses that people make of them. First, the bottom-up delineation of asset uses is a highly duplicative and needlessly cumbersome way to describe the set. If the law gives a right to exclude (as it does prominently through the doctrine of trespass), supplemented and modified with respect to a limited set of uses (i.e., in the law of nuisance, the doctrine of necessity, etc.), one can get an extensionally very similar set of legal relations at a much lower cost. Consider situations of necessity such as the hiker stranded on a snowy mountainside who needs the food in an unoccupied cabin or those on a ship who need another's dock in a storm. The bottom-up method would treat the "breaking in and eating the food" stick or the "docking during the storm" stick as no different from any other stick except in its assignment to another party--the one facing the necessity. In actuality, the law gives the owner a presumptive right to exclude and then modifies this by removing protection against one facing certain emergency situations. In other words, the law starts out with a cheap and crude exclusion regime and in high-stakes situations moves over to a governance regime that focuses on particular uses. (7) Where contracting is not cost-effective, the governance regime is supplied off the rack. This method of presumptive exclusion with refinement through governance economizes greatly on delineation and processing costs for dutyholders and courts. (8) Many problems are taken care of with one stroke through the exclusion strategy, and the delineation effort is concentrated where it is most needed. Overall, the set of sticks in the bundle is numerous and factoring out some of their common elements is economizing.

This Article emphasizes methods of economizing on delineating property rights that sound in indirectness between the interests served and the mechanism by which the interests are secured. Our interest in resources relates to potential and actual uses (i.e., farming, parking, etc.). The exclusion strategy protects these interests in use indirectly, without having to delineate them individually most of the time. The right to exclude can also be regarded as the right to determine use, but when it comes to delineation, the exclusion strategy often economizes on cost through its lack of reference to specific uses at all. By invoking the law of trespass, the farmer can prevent damage to crops and enjoy the use of the land for farming without the law having to make reference to use. Indeed, trespass law, unlike nuisance, does not require allegations of actual harm (9) and, in the absence of proof of harm, will afford nominal damages. (10) And although injunctions for trespass to land are nominally an exceptional remedy, in practice they routine. (11) Trespass law prohibits entry into the column of are quite space defined by the ad coelum rule, (12) which sends a simple message of "keep off" to outsiders and requires very little contextual information for courts. (13) Further, much of the detail that trespass avoids is not relevant to outsiders and is not useful much of the time. So there is no sense in delineating various uses separately, and doing so would just lead to information overload for outsiders like dutyholders and courts. Delineating rights to aggregates of sticks instead of stick by stick is more indirect, and it conserves delineation and processing costs.

There is another sense in which the indirectness of property law avoids problems of delineation in terms of the valued uses of resources. One issue that receives scant attention by bundle-of-rights theorists is that their bottom-up strategy presents no principled way by which to specify the sticks. How fine-grained should the sticks be? Is it the right to farm, or the right to plant crops plus reaping, or planting different crops, and on and on? Furthermore, once delineated, these sticks can be combined and recombined for many different purposes, not all of which are known to every actor--or even, at any given moment, known to any actor. As yet undiscovered attributes and uses lead to all sorts of timing decisions regarding when to discover attributes and when to put them to which uses. Thus, the owner as decision maker holds a complex set of options. Moreover, the uses of an asset are not just risky (e.g., with a variance in outcomes forming a probability distribution), but uncertain, in the Knightian sense. (14) That is, the set of uses of an asset may not correspond to a known probability distribution, and nonowners may not even know the members of the set. Property law helps manage this uncertainty by not making knowledge of the uses or even the probability distributions of their values relevant to dutyholders. In previous work, I have argued that Knightian uncertainty is more conducive to property rules than to liability rules, which do require more knowledge of probabilistic information by officials or courts. (15)

Crucially, much of the time property forces nonowners not to know...