Kevin Collins on Creation Without Restraint

Posted by Kevin Collins

In Creation Without Restraint: Promoting Liberty and Rivalry in Innovation, Professors Christina Bohannan and Herbert Hovenkamp offer a wide-ranging and insightful commentary on how to most effectively promote innovation using both intellectual property (“IP”) law (actually, patent and copyright law in particular) and antitrust law. In this short commentary, I ignore much and focus myopically on only one of their principal theses: their “IP Injury” thesis. This thesis suggests that IP reform has much to learn from the recent history of antitrust law. In the 1970s, the addition of a new doctrinal hurdle that requires complaining parties demonstrate a particular type of antitrust injury led to an efficiency-enhancing decrease in the conduct that is actionable under antitrust law. The “IP injury” thesis posits that changing IP law to require IP owners to demonstrate something the authors refer to as an “IP injury” would lead to a parallel, efficiency-enhancing decrease in the conduct that is actionable under IP law. I hope to make a two-fold argument about the authors’ “IP Injury” thesis. First, I believe that the notion of an “IP injury” is a very useful trope or rhetorical device. It potently expresses why the rights in the contemporary IP regime should be pared down, and it therefore shores up many IP-minimalist proposals for IP reform. Second, however, I am concerned that an “IP injury” requirement in IP law may not prove to be particularly helpful on an instrumental, doctrinal level. Positioning an “IP injury” requirement as the means that courts could use to administer a slimmed-down IP regime cannot yield the benefits that the authors promise.

The recent evolution in antitrust law that the authors suggest should serve as a model for IP reform is well known. In the before picture, antitrust law was bloated and overweight from an efficiency perspective. It did not focus solely on anticompetitive practices that harmed consumers; it sometimes protected smaller firms from larger firms when the larger firms simply enjoyed benefits, such as economies of scale, which allowed them to produce goods more cheaply. In the after picture, antitrust law is leaner and meaner. Its doctrine is more closely tailored to the normative goal of promoting competition, i.e., restraining only anticompetitive practices. One way in which antitrust law was slimmed down was the implementation of the antitrust injury requirement. For a firm to allege an actionable antitrust suit today, the firm must demonstrate that it in particular has suffered a particular type of injury—roughly an injury that results from decreased rather than increased competition. The antitrust injury requirement functions can function as a gatekeeper. Injuries from increased and decreased competition are animals of different stripes that courts can readily distinguish at low cost and with a low error rate based on the pleadings, and the enforcement of the antitrust injury requirement obviates the need to address complex, uncertain issues about markets and market power. As the authors note, “[i]f a plaintiff is complaining that a merger caused more rather than less competition in a market, why bother with the difficult substantive analysis?” [34]

The authors posit that courts could model an “IP injury” requirement on the antitrust injury requirement, and that it could slim down the contemporary, over-bloated IP regime just like antitrust injury requirement helped to slim down the antitrust regime. In order to prevail in an IP infringement action, they argue that an innovator/rights-holder must demonstrate that the allegedly infringing user is engaged in a use of an innovation that, if it were to go uncompensated, would cause the kind of social harm that the IP laws are designed to remedy, namely a decrease in the rights-holder’s ex ante incentive to produce the innovation in the first place. Inversely formulated, an “IP injury” requirement would mean that “[i]f a particular use is unlikely to affect the IP holder’s decision to produce a work, then the IP holder is not entitled to a remedy.” [58] The authors posit that an “IP injury” requirement can be a gatekeeper in IP law that allows courts to cut off complex and time-consuming questions at the pass, just as the antitrust injury requirement does in antitrust. That is, “we can correct problems of IP overreaching by” implementing an “IP injury” requirement and “avoiding the truly difficult problems of formulating substantive patent and copyright rules.” [34]

Highlighting the notion of “IP injury” in discussions of intellectual property rights is a useful way of bringing discussions about intellectual property back to their economic underpinnings. We do not grant IP rights to ensure that innovators get what is naturally theirs or reap what they have sewn. We grant IP rights to serve economic ends, and we should not grant rights that become so expansive that they fail to serve those ends. However, on an instrumental level, I am skeptical of the value of an “IP injury” requirement as a doctrinal matter. I doubt that it can serve the kind of gate-keeping role in IP law that the antitrust injury requirement plays in antitrust law. There is no way to avoid “the truly difficult problems of formulating substantive patent and copyright rules” with an “IP injury” requirement. Rather, to the extent that an “IP injury” requirement can be instrumentalized in IP law as doctrine at all, formulating an “IP injury” requirement is inevitably nothing other than crafting those very difficult, substantive patent and copyright rules themselves. Consider just two reasons why the role of the antitrust injury requirement as a gatekeeper in antitrust law cannot be paralleled by an “IP injury” requirement in IP law. First, it is extremely difficult to identify in any given case whether there has been an “IP injury,” making such an injury a poor candidate for a gatekeeper. Second, and more profoundly, whether there has been an “IP injury” in any particular case is often irrelevant in IP law, even if IP law is to be firmly grounded in economic concerns. An “IP injury” will often be difficult for a court to identify. To demonstrate an “IP injury,” an innovator would have to prove something like both (a) that the innovator’s particular innovation would not have been created in an optimally timely manner but for the expectation of an IP right of some kind, and (b) that this particular defendant’s conduct is the type of conduct that must be encompassed within the innovator’s IP right in order to create an IP right of sufficient scope to optimally incentivize the innovator. Demonstrating an “IP injury” is difficult because, as a theoretical matter, the socially optimal level of IP-induced ex ante incentives to invest in innovation is a contested subject. Assuming the theoretical problem is resolved, demonstrating an “IP injury” is difficult because, as an empirical matter, getting one’s hands on the data to demonstrate what would have happened but-for the lure of an IP right is next to impossible. This makes an “IP injury” very different from an antitrust injury in that the distinction between more and less competition is presumptively not difficult to identify.

Even assuming one could demonstrate that an individual innovator has not suffered an “IP injury,” this finding is not dispositive of whether an IP right should exist and what its scope should be. Antitrust analysis under the rule of reason requires a highly individualized, fact-intensive analysis of the economic ramifications of the particular conduct of a particular plaintiff and a particular defendant in a particular market. In contrast, the economic policy of IP plays out entirely on the basis of categories and rule-like proxies. IP rights are expected to be over-inclusive and under-inclusive, so there will inevitably be innovators who suffer no “IP injury” as a particularized factual matter and yet who should prevail in an IP infringement suit, even if IP rights are optimally calibrated. A particular copyright owner may have had unusually low sunk costs, yet that copyright owner is treated no differently from another copyright owner who had unusually high sunk costs. The nonobviousness determination is a technological proxy for a determination of whether an invention would suffer from a market failure absent patent rights, but it is clearly only a rough one.

Of course, the obvious response to my concerns about the difficulty of administering an “IP injury” requirement is to take a less literal and less slavish approach to the translation of the injury requirement from antitrust to IP. Given that IP is about categories, one could take a categorical approach to the “IP injury” requirement. For example, one could attempt identify in advance categories of innovators that, in most situations, do not need any IP-induced incentives to have socially optimal incentives to innovate. Similarly, one could attempt to identify (categories of) uses of (categories of) innovations that, under normal circumstances, need not be within the control of an IP owner under any reasonable interpretation of the socially optimal level of ex ante incentives. I believe that drawing lines like these is a critical part of achieving IP reform. However, grappling with categories like these is nothing more or less than engaging with “the truly difficult problems of formulating substantive patent and copyright rules.” [34] Applied in this categorical manner, an “IP injury” requirement does not serve as a gatekeeper as it does in antitrust law; it does not allow courts to avoid grappling with the difficult substantive rules of IP. Rather, it is only another way—albeit a potentially productive way—of talking about how we should tweak those substantive rules.