Northwestern University Law Review : Colloquy : 2010 : The Riddle Underlying Refusal-to-Deal Theory
The Riddle Underlying Refusal-to-Deal Theory
May a dominant firm refuse to share its intellectual property (IP) with its rivals? This question lies at the heart of a highly divisive, international debate concerning the proper application of the antitrust laws. In this short Essay, we consider a profound, yet previously unaddressed, incongruity underlying the controversy. Specifically, why is it that monopolists refuse to share their IP, even at monopoly prices? To resolve this issue, some have recommended compulsory licensing, which would require monopolists to license their IP in certain circumstances. This proposal, however, entails an inescapable contradiction, one rooted in the issue of monopolists’ seemingly inexplicable refusal to share their IP.
The policy tensions implicated by the compulsory licensing debate are straightforward, involving a tradeoff between short- and long-term considerations. Assuming that the monopolist being targeted has not committed a separate violation of the antitrust laws, First, forced sharing will render the market more competitive in the short run. Second, by increasing access to the relevant IP, the law may accelerate cumulative innovation and invent-around. Both of these benefits seem reasonable, yet they portend a serious long-term cost: antitrust incursions into IP threaten to diminish incentives to invent.
The tension between the long and short run moves the debate in a predictable but inconclusive direction. Consider, for example, the differences between the U.S. and European approaches to compulsory licensing. Courts in the United States have ruled that monopolists may refuse for almost any reason to license their IP to a rival, reasoning that the long-term harm to incentives counsels against compulsory licensing. Conversely, the European Court of First Instance has ruled that compulsory licensing is appropriate when necessary to bring "viable" competition to uncompetitive markets, choosing to discount (or ignore) the long-term incentive effects.
A secondary argument, but an important one nonetheless, also follows predictable lines. U.S. courts are skeptical about the ability of regulatory bodies or judges to determine, over time and changing circumstances, the proper fee and other terms of a compulsory license. They believe that embroiling agencies or judges in the details of writing and interpreting ongoing licensing agreements will generate large costs with little or no commensurate benefit. Instead, they regard the market as the best mechanism for organizing transactions and setting prices. Their European counterparts, more confident about the abilities of regulatory regimes and less trustful of market mechanisms, believe otherwise. They would thus entrust those complicated tasks to judges and administrative agencies.
This debate over compulsory licensing is remarkable, and for good reason. Its very existence illustrates deep and seemingly unbridgeable rifts between the U.S. and European competition regimes regarding the proper application of basic elements of antitrust law and economics. Its indeterminacy—there is no answer to the critical question whether the short-term benefits of compulsory licensing exceed the long-term costs of reduced incentives to innovate—provides a vivid demonstration of the limits of economic analysis. And its resolution in the respective jurisdictions bears witness to the important role of culture and history in providing a basis for decision-making when the explanatory power of economics is of no avail.
We have written about this elsewhere and will not rehearse it here. Instead, we consider perhaps the single most remarkable feature underlying the debate: the decision of the dominant firm to refuse to license its IP. This refusal would appear to be a business mystery of sorts. If the goal of every business is to reach the point where it can maximize profit by charging monopoly prices, as would certainly seem to be the case, then why would a dominant firm ever refuse—completely refuse—to license its valuable IP to rivals? Why would it not offer to license at the monopoly price? Why would cases arise in which a dominant firm refuses even to name a price at which it would license its IP to rivals?
There are risks, after all, in refusing to license. Granting a license at the monopoly price would ensure for a time the rival’s use, and the market’s continued use, of the dominant firm’s technology. Such an agreement would certainly pose risks for the dominant firm—for example, the rival might improve on the technology—but contract (license) terms could minimize or eliminate most of those risks. On the other hand, a refusal to license at any price, while harming the rival in the short-term, or maybe even eliminating it, would compel the next generation of inventors to invent around the dominant product. If they succeeded, their new technology would leap-frog the dominant firm’s IP, and might thus (depending on switching costs) make that IP obsolete in the next term. Licensing now at the monopoly price could forestall that later leap-frog invention and extend the life of the dominant firm’s market power. Why not license it, then?
There are at least two potential answers to this question. First, the dominant firm might lack the information needed to determine the monopoly price. Perhaps the IP has great value now in certain markets or applications, but may have even greater value in other markets or applications not yet identified. A license at today’s monopoly price might not fully capture tomorrow’s value, allowing one’s rival to profit in those yet-to-be-identified markets and encouraging it to identify those markets for itself, thus gaining a first-mover advantage in them. Nor would it necessarily be a simple matter to contract around this problem. Contract is an imperfect vehicle, incapable of adequately specifying solutions to all potential problems that might arise during its term. In addition, contracting costs might be high. And, of course, the terms needed to prevent the rival from using the IP to create or discover a profitable new market—or requiring it to share that market with the licensor—might create legal risks of their own.
Second, if the dominant firm is able to determine the monopoly price and assess the relevant costs, it might conclude that the costs outweigh the benefits and that agreeing to license would not therefore be the profit-maximizing strategy. Alternatively, the firm might not know the "correct" price, but might know that the pertinent costs are high. Or, knowing neither the price nor the costs, it might fear that high costs could swamp the profitability of a license agreement.These conclusions yield an important insight. On any of these premises, the dominant firm should be free to refrain from licensing, not because its freedom is an important antitrust value in itself, but because no regulatory authority or judicial decree could create a licensing agreement that would plausibly offer to improve social welfare. This conclusion follows from the assumptions that the dominant firm is a self-interested profit-maximizer: that it has every incentive to identify the monopoly price and the costs—transactional, legal, and otherwise—of a licensing arrangement; and that when it either cannot adequately identify those factors, or identifies them and concludes that licensing is ill-advised, its conclusion is bound to be superior to any different conclusion that an agency or court might reach.
*. Distinguished Research Professor, DePaul University College of Law; Dartmouth College, B.A., 1968; Yale Law School, J.D., 1971.
**. Law Clerk to the Honorable Richard D. Cudahy, United States Court of Appeals for the Seventh Circuit; University College Dublin, B.B.L. (Int’l), 2004; University of Chicago, LL.M., 2005; University of Chicago, J.S.D., 2006; Stanford Law School, J.D., 2007.
1. If the monopolist has infringed the competition laws through unrelated conduct, see, e.g., United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001) (link), compulsory licensing might be an appropriate remedy.
2. See, e.g., Verizon Commc’ns. Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 407–16 (2004) ( ); In re Indep. Serv. Orgs. Antitrust Litig., 203 F.3d 1322, 1325–28 (Fed. Cir. 2000) (link); Image Technical Servs., Inc. v. Eastman Kodak Co., 125 F.3d 1195, 1215 (9th Cir. 1997) ( ).
3. See Case T-201/04 R, Microsoft v. Comm’n, 2004 E.C.R. II-4463.
4. See William J. Kolasky, Conglomerate Mergers and Range Effects: It’s a Long Way from Chicago to Brussels, 10 Geo. Mason L. Rev. 533, 533–34, 537 (2002).
5. Cf. id. at 534–35.
6. See Alan Devlin & Michael Jacobs, Antitrust Divergence and the Limits of Economics, 104 Nw. U. L. Rev. __ (forthcoming 2010) ( ).
7. See Verizon, 540 U.S. at 407 (explaining that "[t]he opportunity to charge monopoly prices . . . is what attracts ‘business acumen’ in the first place").
8. For instance, exclusive grantback provisions in patent licenses, by which licensees are compelled to assign back to the licensor any improvement patents, may run afoul of the antitrust laws. See, e.g., Transparent-Wrap Mach. Corp. v. Stokes & Smith Co., 329 U.S. 637, 646–48 (1947) (link). More generally, market-sharing agreements, even within the scope of a patent, are subject to antitrust scrutiny. See, e.g., In re Ciprofloxacin Hydrochloride Antitrust Litig., 604 F.3d 98, 104 (2d Cir. 2010) (per curiam) (link).
9. One possibility is that the relevant IP is of greater value to the inventing monopolist than it would be to its rivals due to the endowment effect. See Cass R. Sunstein, Switching the Default Rule, 77 N.Y.U. L. Rev. 106, 112 (2002). This cognitive bias, which might lead monopolists to place an artificially elevated value on their IP, causes dominant firms to refuse to license at what would be deemed a "monopoly" price by the larger market. On that basis, one might contend, a regulator or court may improve social welfare by compelling access at a "market" rate. Notwithstanding monopolists’ potential failure to act as rational actors, a "market" price—so defined—will under-compensate the inventor. Subjective value, as Judge Posner has pointed out, is an important economic criterion. See Richard A. Posner, Economic Analysis of Law 55–58 (2003).
Copyright 2010 Northwestern University
Cite as: 105 Nw. U. L. Rev. Colloquy 1 (2010), http://www.law.northwestern.edu/lawreview/colloquy/2010/17/LRColl2010n17Devlin&Jacobs.pdf.
Persistent URL: http://www.law.northwestern.edu/lawreview/colloquy/2010/17