Monday, March 5, 2012
Posted by Tom Cotter
In the introduction to their insightful and thought-provoking new book, Creation without Restraint: Promoting Liberty and Rivalry in Innovation, Christina Bohannan and Herb Hovenkamp [hereinafter BH] cite Jared Diamond’s book Guns, Germs, and Steel for the proposition that “Those societies that have enjoyed the greatest amount of innovation progress have provided substantial freedom for innovators to build on the work of their predecessors”—a sentiment with which I fundamentally agree. In reading through their book, however, I found my mind turning at times to Diamond’s other major work for a popular audience, Collapse: How Societies Choose to Fail or Succeed. To be sure, Diamond’s work focuses on factors (such as environmental degradation, climate change, problems with or lack of support from neighboring groups, and inadequate social responses to these problems) that have influenced the collapse of entire societies, such as the Rapa Nui culture that created the Easter Island megaliths, while BH’s topic is the (weighty, but comparatively less weighty) dysfunctional nature of much of U.S. intellectual property law. And, to be fair, neither BH nor I think the U.S. IP system (much less U.S. society generally) is on the verge of imminent collapse. But I was struck at the ways in which many of the problems inherent to our current system appear to be largely resistant to rational reform. For example, BH discuss in some detail the public choice factors that have led, over the past few decades, to the statutory overprotection of IP rights, and they make some intelligent proposals for responding to these trends (e.g., construing ambiguous statutes against their content-industry drafters). But I wonder if some of our problems do not go even deeper. Various structural features of our litigation system, for example, encourage patent trolling: the constitutional right to a jury trial in civil cases, which sometimes leads to exorbitant damages awards; an “American rule” that permits nonprevailing plaintiffs to impose heavy litigation costs on their opponents with impunity; an aversion to the use of court-appointed experts; and so on. Moreover, as a nation we have committed ourselves to various international obligations relating to IP—many of them of our own making—that also can render it very difficult to change things once they are in place, even when subsequent analysis suggests that they do more harm than good: consider, for example, the excessive copyright term, the elimination of most copyright formalities, and the (supposed) obligation to resurrect certain defunct copyrights that was recently reaffirmed in the Supreme Court’s Golan opinion. Indeed, Golan itself, by severely limiting the courts’ ability to engage in any serious constitutional scrutiny of copyright and patent legislation that negatively impacts the public domain, could serve as Exhibit Number One in a catalogue of what has gone wrong with the U.S. system (though it would have many rivals for that dishonor).
But perhaps I grouch too much. The recently enacted America Invents Act, while surely imperfect, marginally improves upon existing patent law by moving the U.S. to a first-to-file system more or less consistent with the rest of the world, and by introducing opposition proceedings intended to screen out invalid patents at an early stage. And many of BH’s proposed reforms, directed as they are to the courts instead of to Congress, are in theory plausible and would, in my view, improve the workings of the system in various ways. In the short space accorded me for these remarks, I will focus mostly on two such reforms that I also have touched upon in some of my previous work.
First, BH argue that courts should not presume that every unauthorized use of protected subject matter harms the IP owner, but rather (borrowing a page from competition law’s concept of “antitrust injury”) should require IP plaintiffs to plead and prove injury of the type the IP laws were intended to prevent. At the margin, this would mean, for example, that some unauthorized uses of copyrighted material that today would be viewed as unauthorized derivative works—for example, the unauthorized use of plots and characters from the Seinfeld t.v. show to create a Seinfeld quiz book, which was at issue in Castle Rock Entertainment, Inc. v. Carol Publishing Group)—would be viewed as nonactionable, on the theory that the use at issue was unforeseeable at the time of the creation of the underlying work and does not threaten to undermine the creation of such underlying works in the first place. On BH’s view, the unauthorized use in such a case does not harm the owner of the Seinfeld copyright. I largely agree with their analysis, though as I have argued elsewhere I think that the concept of harm is sometimes less clear than their analysis suggests; harm to the copyright owner is not a natural category, like physical harm, but rather (from a legal positivist perspective, at any rate) depends upon the baseline entitlement that the law itself provides. I would argue therefore that an IP owner suffers (cognizable) harm if (1) she prefers that the use not occur absent her permission; (2) transaction costs are not so high as to render negotiations with the user impractical; and (3) the law backs up the owner’s preference with the threat of legal sanctions. On this view, the fact that the author of the quiz book did not pay the owner of copyright in the underlying work a royalty probably satisfies conditions (1) and (2), but whether it also satisfies condition (3) depends upon the scope of the owner’s legal entitlement. Determining the scope of that entitlement is the crucial issue. In my view, and consistent with BH, where the statute is ambiguous or open-ended, as is the provision of the Copyright Act relating to derivative works, courts can and should define that entitlement narrowly if reasoned analysis suggests that an expansive right would generate more social cost than benefit: that is, if so construing the right would pose no threat to the creation of future works and would reduce corresponding costs of access. Of course, one has to be careful here; not every case will be susceptible to such simple cost/benefit analysis, and courts probably should not engage in such analysis except in cases of true ambiguity—that is to say, when they have no choice but to act as policymakers. At the end of the day, however, despite some differences in approach, I think that BH and I largely reach the same conclusions; and I find BH’s analysis of what rights should, and should not, fall within the scope of the derivative right (pp. 222-24) compelling.
Second, BH correctly point out that some conduct that inhibits innovation or impedes access to the public domain may not violate antitrust law, in part because antitrust’s tools for evaluating alleged harms to innovation are not robust; but that IP law can and should prevent some such uses itself, using among other tools the fair use, idea/expression, and misuse doctrines. Again, I largely agree. An IP license, for example, may not clearly enable the licensor to maintain control over a market that is well-defined for antitrust purposes, due to the presence of substitutes. The license nevertheless may enable the licensor to extract rents that are disproportionate to the investment in the underlying work; or may (singly or in the aggregate) increase transaction costs to the point that subsequent innovation is impeded; or may in other ways unnecessarily exclude third party access to the work with no corresponding public benefit. Although one would not expect or desire courts to make open-ended decisions about social costs and benefits on an unguided, case-by-case basis, various IP doctrines (in theory) serve the purpose of promoting a surplus of benefits over costs—for example, by relegating ideas and facts to the public domain, and by (in the copyright context) permitting fair uses that either cause no harm at all due to the presence of prohibitive transaction costs or that promise disproportionate social benefits. In addition, however, BH also advocate a greater role for the misuse doctrine, and on this point I am somewhat less convinced. The best argument for the misuse doctrine is that it enables courts to deal with restraints that either impede access to the public domain, or that pose a threat to future innovation that is too remote to merit antitrust condemnation but at the same time is not worth risking for the sake of some minimal boost to the patent or copyright incentive scheme; a license provision that prohibits reverse engineering might be a good example of such a restraint. The best argument against the doctrine is that its penalty (unenforceability of the IP right in its entirety) may be excessive, particularly given the “standing” issue: under standard misuse doctrine, any alleged infringer (included one who is not a victim of the alleged misuse) can raise the defense. BH sensibly argue that the penalty for misuse should be limited (at least in some cases) to the unenforceability of the offending license provision, and on this point I wholeheartedly agree. Moreover, while they do not mention the standing issue as such, it seems to me that if courts were to alter the penalty for misuse from unenforceability of the patent to unenforceability of the offending license provision then only victims of the misuse would have any reason to raise the defense. If this is correct, however, I find it slightly curious that BH appear to disapprove of the Federal Circuit’s recent Princo decision, holding that an alleged agreement between two contributors to a patent pool to suppress one technology in favor of another provided no basis for an infringement defendant’s assertion of misuse (pp. 281-82). To be sure, under BH’s reformulation of the doctrine, the alleged suppression agreement might be unenforceable (p.282), but I don’t see how that result would provide any basis for immunizing the Princo defendants from liability for infringement; yet if not, why would they bother to litigate the matter at all? The example suggests to me the limits (or awkwardness) of the misuse doctrine as a tool for promoting innovation in circumstances like those at issue in Princo. At the same time, the example arguably strengthens BH’s argument that the FTC should play a more active role in policing some potentially innovation-inhibiting conduct that may be difficult to scrutinize under standard antitrust law (see pp. 401-02, specifically proposing that the FTC impose interconnectivity obligations under some circumstances).
There is much else to admire in the book, including the authors’ discussion of problems with patent notice and continuation abuse; issues relating to patent pools and standard setting; and a devastating critique of the Supreme Court’s retro Quanta Computer decision on patent exhaustion. One issue that I found somewhat less satisfying was their advocacy of an independent inventor defense in patent law (pp. 128-32). On one view, this reform would make a lot of sense. The patent system is often justified on the ground that according inventors with a right to exclude free riders is necessary to encourage the initial investment in the inventive process; but where there is no free riding this rationale may seem lacking. The counterarguments, however, are also serious, among them that an independent invention defense would reduce inventors’ expected returns on investment and thus might impact the patent incentive in industries in which that incentive really is important; that it might reduce the incentives for would-be users to read patents; that it would be difficult to administer; and that it might impede licensing. (I wonder too whether it might violate article 30 of the TRIPs Agreement, one of the treaties I alluded to in the introductory paragraph.) Without going into the merits of the arguments here, I note only that these issues have been discussed by commentators including, among others, Samson Vermont, Mark Lemley, Stephen Maurer and Suzanne Scotchmer, and Norman Siebrasse, and I would have welcomed some further engagement with this scholarship in the BH analysis.
Overall, however, the book is a tour de force. My quibbling on some points notwithstanding, I hope that courts and other policymakers will take seriously its recommendations. As BH stress throughout, because the gains from innovation are so vast, practices that raise the cost of future innovation pose an immense threat to social welfare. Devising and implementing effective responses to such practices can be enormously challenging, but the stakes can be soberingly high. Not to press the analogy too far, but it would be a pity if our descendants were to find themselves, like the inhabitants of Easter Island a few centuries ago, awaking one day to a barren landscape distinguished principally by its monuments to a glorious, innovative, but now long-gone past.