Tuesday, March 6, 2012
Posted by Ariel Katz
The last chapter in Bohannan & Hovenkamp’s excellent book deals with post-sale restraints, those “imposed by a seller on how a purchased good can be used and resold after the initial sale.” Post-sale restraints have attracted treatment by both antitrust and IP law: antitrust, through its treatment of (mainly contractual) vertical restraints, and IP law, through the first-sale doctrine, which limits the exclusive rights that survive the initial authorized sale of an item protected by IP rights. While antitrust has relinquished its historic aversion to post-sale restraints and has grown to recognize that they are often benign, if not beneficial, IP law seems to adhere to a stricter first-sale rule, which prevents IP holders from enforcing post-sale restraints by availing themselves to remedies for the infringement of IP. In B&H’s view, this disparate treatment of post-sale restraints is hardly justifiable. Therefore they ask: “does the first-sale rule serve any useful purpose at all?” (p. 393) and their answer is a resounding “no”.
Their logic is straightforward: if post-sale restraints are generally efficient, then generally they ought to be enforceable. And because “[p]ost-sale conditions enforced by infringement actions seem superior to contract suits when … potential violators are numerous, difficult to identify, or not in privity of contract” (id) such infringement actions should be available, provided that proper notice of the restraint is given. If a need arises to control a few harmful restraints, this should be done by antitrust or misuse rules and on the basis of fully articulated and demonstrated harm to competition or innovation (id).
I agree with much of B&H’s analysis regarding the efficiency of many post-sale restraints, but my answer to the question “does the first-sale rule serve any useful purpose at all?” is “yes”. In my opinion, the first-sale doctrine is not a relic from an era in which the economics of vertical restraints were not well understood, and there is no real contradiction between the teachings of modern antitrust and the first-sale doctrine. There are important reasons for maintaining a distinction between the scope of arrangements contracting parties can agree on and the scope of conditions that property owners can impose on third parties. That is why the law may generally permit contractual restraints in appropriate cases, while remaining reluctant to incorporate a right to impose such restraints into the IP bundle. Contrary to B&H’s suggestion that IP remedies should generally be available, except when it could be demonstrated that they are harmful, I would posit that IP remedies should not be generally available, except when their benefit to innovation or welfare can be fully articulated and demonstrated.
Modern antitrust law and economics recognizes a variety of pro-competitive purposes that post-sale restraints might be imposed to achieve. The common thread it that post-sale restraints may facilitate efficient contracting between collaborating firms by reducing risk and controlling opportunism, thereby helping to organize sophisticated distribution systems when a manufacturer is not fully integrated into distribution and retail. Obviously, in those situations the restraints should be enforceable, and treating IP rights as default entitlements and permitting contracting around them to achieve efficiency in a profit-neutral way makes sense. However, it hardly follows that post-sale restraints should be generally enforceable in situations that do not involve of joint production, and remain enforceable—for the entire duration of the IP—against anyone possessing the item in which the IP is embodied. The economics of joint production cannot normally justify long-lasting post-sale restraints or restraints imposed on third parties who are not joint-producers. The trouble is that while restraints that can be enforced in rem are more effective, they can be enforced not only against the collaborating firms, but also against anyone who possesses the item. Moreover, they may be enforced even when and where the reasons for imposing them no longer exist, or in situations where enforcing them may not be socially beneficial.
See the following diagram.
The diagram demonstrates this point graphically. It describes circles of relationship between an IP owner and interested parties through the life-cycle of an innovative good. The inner circle represents the IP owner and its co-producers, or the several firms that participate in the production or distribution of an innovative good. The second circle consists of consumers who buy the goods produced in the inner circle. The third circle is populated by the users who may at one point be interested in obtaining and using the innovative good but do not obtain it directly from the inner circle. Two vectors—distance and time—are also presented. Distance reflects the transactional proximity between the IP owner and the user. The time vector reflects temporal proximity between production, distribution and use. As the distance between the user and the IP owner increases and as time goes by, the need to control use diminishes.
While the marginal benefit from having enforceable restraints diminishes as we move along the vectors of distance and time, the marginal social costs associated with goods encumbered by restraints could easily increase over distance and time. Any use that is inconsistent with the restraint would require the IP owners’ permission, while the cost of obtaining such permission could easily increase over time and distance. The result may be net increase in future transaction costs, leading to inefficient asset use. Unfortunately, neither the IP owner, nor its immediate transacting parties are likely to internalize these costs or even consider them when they design their initial transactions because these costs will be borne by future unknown users.
The conclusion is that allowing IP owners to impose in rem restraints might interfere without good reason with downstream users’ freedom to utilize goods in their possession as they see fit, to freely dispose of them or sell them to others. These freedoms, which drive competitive markets in the first place, and are also crucial for the ability of downstream users to utilize goods that they purchase as inputs in their own innovative endeavors. The first-sale doctrine preserves them. The first-sale doctrine is important not only because it reduces transaction costs that would create static efficiency losses at any given moment in time during the lifetime of the innovative good, but also because eliminating these transaction costs promotes dynamic efficiency. This becomes clear when we consider the growing acknowledgement of user-innovation as a major source of innovation. User innovation is driven by lead users having an intention to use rather than sell an innovative technology. These users have heterogeneous needs that are not easily satisfied by mass-market production. Their unique intellectual capital and idiosyncratic needs and experiences often allow them to encounter problems and devise solutions that no one else can. They frequently anticipate features that improve products, but for which no general demand yet exists. Lead users are thus distinguishable from the general consumer in that they possess both a need for optimization that impels modification of an existing product for their own use, and an expertise to accomplish this modification. User innovations commonly emerge from many small ideas, obtained through use and manipulation of an asset, over long periods of time. If non-producer-centric models of innovation are becoming important sources of innovation, as the literature on user-innovation suggests, then the design of optimal rules with respect to post-sale restraints should take into account these other sources of innovation. The first-sale doctrine fosters user-innovation because it allows a user who possesses a copy or an artifact to experiment with it without the need to obtain permission from the IP owners, and it permits users who do not innovate to resell or otherwise transfer the goods to the ones who do innovate. User innovation is tied to asset use and emerges from it. Few can predict what innovation will follow when a good is sold. Fewer still can predict what value such innovation may have for social, cultural, and political life and price it accordingly. Moreover, the innovator may not even be present when the terms of the transaction between the IP owner and its transacting parties are negotiated, or she may not know that she would become one when she purchases the good. Therefore, in a truly Coasian way, if we cannot tell ex ante who is in the best position to further innovate, rules allowing possessors to innovate without seeking permission reduce transaction costs and promote innovation. This is exactly what the first-sale does: it facilitates creation without restraint.
To support their claim that the ability to impose post-sale restraints should be part of the IP bundle, B&H draw an analogy to the law of servitudes in real property, which generally allows property owners to create restraints that would bind any subsequent purchaser who has notice of the restraint (pp. 393-94). From an organizational perspective, both land servitudes and post-sale restraints may promote productive asset use in situations when integration is incomplete or absent. Obviously, if one firm owned all tracts of land in a relevant area it would internalize all costs associated with land use and decide for itself how optimally to use the land. Similarly, a firm that is fully integrated into distribution and retail would not need to impose post-sale restraints because it could control all externalities between its units through managerial orders. Beyond these similarities, however, land servitudes and IP servitudes (that is, post-sale restraints enforceable as a matter of IP law), are mirror images of each other, and the analogy fails.
Land servitudes allow private parties to coordinate productive land use and reduce potential conflicts that are an inevitable consequence of the finite nature of land, increased urban density, and of the inevitable externalities that the use of one piece of land may impose on others. Further, the attributes of land suggest that the need for coordination is often as durable as the land itself, unlikely to decrease over time, and will often outlive the tenure of those who own or possess the land at the time the need to coordinate arises. Under such circumstances, the cost-inefficacy of reliance on contracts to enforce land-use restrictions is evident and the superiority of servitudes is clear.
Unlike land, whose efficient use may require durable restraints that would be enforceable against future buyers, the organizational problems in the context of collaboration in the production and distribution of innovative goods occur mainly at the early stages of the product life-cycle: production problems disappear immediately after the good is produced, and distribution problems largely disappear upon the distribution of the good or shortly thereafter. Notably, these problems cease to exist long before the IP rights expire (particularly in the case of copyright). Moreover, as co-producers, the firms are often in sufficient privity, which enables them to rely on contracts for addressing many of the organizational problems associated with efficient distribution. Thus, while IP remedies may increase the enforceability of post-sale restraints, the marginal benefit is small and must be weighed against the fact that the threat of IP remedies might prevent the efficient use of the good long into the future, even when they no longer serve their original purpose.
Lastly, the dampening effect that in rem restraints may have on future innovation also explains why the problems with IP-backed post-sale restraints to not disappear with proper notice, and why antitrust and misuse rules may not be sufficient to control undesirable restraints. The answer is simple: the harms caused by long-lasting in rem restraints may be substantial but totally unobservable. We cannot identify and quantify the harm resulting from the fact that a particular book could not be resold or a particular product freely tinkered with. The innovations that did not occur, the knowledge that was lost, or the ideas that were not developed have no one advocating for them.
In sum, preventing the need to prove such harm is the useful purpose that the first-sale doctrine serves. It promotes innovation when it permits IP owners to impose restraints when they might be necessary to coordinate efficient production and distribution, and it promotes future innovation when it does not allow them to burden innovative goods with long-lasting in rem restraints. The elusive useful purpose that the first-sale doctrine serve is the book’s title: Creation without Restraint: Promoting Liberty and Rivalry in Innovation.