Intellectual Property, Standard Setting, and the Limits of Antitrust

Cite this Article
Joshua D. Wright, Intellectual Property, Standard Setting, and the Limits of Antitrust, Truth on the Market (October 22, 2009),

[Cross-posted at, a blog devoted on standards and related issues]

One of the most significant challenges facing competition policy today is defining the appropriate role of antitrust law within the context of intellectual property right licensing by standard-setting organizations (“SSOs”).  Many commentators believe it is necessary to apply the full force of the antitrust laws, and sometimes special rules that would increase the scope of antitrust, to the standard-setting process in order to adequately oversee what they perceive as a unique opportunity for anticompetitive behavior.  Indeed, antitrust agencies both in the United States and around the world have expressed agreement with the notion that the standard setting process requires strong enforcement of antitrust liability rules in order to ensure efficient outcomes that benefit consumers.  However, this view largely fails to consider the costs of antitrust.  In particular, it fails to recognize the limits of antitrust when the marginal benefit of antitrust enforcement is slight and the potential for erroneous enforcement (“false positives”) and thus, the likelihood that procompetitive behavior will be deterred, is high.  The Supreme Court itself has emphasized repeatedly that the scope of the antitrust laws should be responsive to such a cost-benefit analysis.

The limits of antitrust are particularly discernable in the context of patent holdup.  The basic patent holdup problem is well known.  In one scenario, a SSO adopts a specific patented technology as part of a standard in exchange for a promise by the patent holder to license the technology at a reasonable and non-discriminatory price (“RAND”).  In other scenarios the patent holder may fail to notify the SSO that it owns the patent under consideration, or intentionally engage in deceptive conduct to avoid disclosure, until after the patent has been adopted into a standard.  In either case, a patent “holdup” occurs if the patent holder later demands supra-competitive prices from parties using the patented technology that has been adopted in the standard after specific investments have been made.  The first line of case includes only breach of the RAND commitment (or other terms) but does not involve deceptive conduct.  The second includes allegations of actual deception on the part of the patent holder.  Patent holders are able to extract supra-competitive prices under such circumstances because SSOs are often unable to easily switch to another standard after spending significant time and capital to develop the current standard.

Traditional antitrust principles are capable of dealing with the “deception” line of patent holdup cases.  As the D.C. Circuit correctly observed in overturning the Commission’s finding of liability against Rambus in such a case, plaintiffs must bear the burden of establishing the traditional elements of a monopolization case: that the defendant had ex ante monopoly power, that its deceptive conduct was “exclusionary,” and that it caused the maintenance or acquisition of greater monopoly power resulting in consumer harm.  The “breach” line of patent holdup cases is more problematic, in large part because antitrust agencies and some commentators wish to deviate from traditional antitrust principles in order to reach conduct that they view as problematic.  Again, the problem with this view is that it ignores or minimizes the costs of expanding the scope of antitrust rules to the standard setting process in this way without considering how such a scheme may obstruct innovation an economic growth.

Consider the Federal Trade Commission’s (“FTC”) recent decision to prosecute patent holdup without finding actual deception is a particularly egregious example.  In N-Data, the IEEE, an SSO, adopted a patent owned by National as part of a standard in exchange for National’s agreement to a licensing fee of $1,000.  National later assigned its right to Vertical.  Vertical attempted to amend the $1,000 licensing term to permit fair, reasonable, and nondiscriminatory (“FRAND”) pricing.  The SSO approved the change and posted the new agreement, along with the original, for all users to view.  Vertical then assigned its rights to N-Data.  The FTC found that N-Data had engaged in “oppressive” and “coercive” actions that “harmed consumers” and “undermine competition.”   In doing so, the FTC paved the way for finding antitrust liability whenever a patent holder breaches any licensing commitment that has the effect of increasing royalties—even when done in good faith.

I’ve criticized this decision at greater length in articles here and here, but there are three primary problems with the N-Data approach which makes such a breach a sufficient condition for access to antitrust remedies.  The first is that the approach deviates from the conventional role of antitrust has a set of rules governing the competitive process rather than particular outcomes.  Here, antitrust liability attaches not because of any deceptive conduct by the firm, but because the antitrust agencies’ ex post evaluation of the terms renegotiated in good faith seven years after the original contract deems them “oppressive.”  This approach invites expansion of antitrust doctrine to mundane contract disputes and disagreement that have price effects but no impact on the competitive process.  The second problem is that the FTC’s willingness to employ Section 5 to this end indicates a desire to circumvent these traditional principles as expressed under conventional monopolization law, to avoid problems of proof and take advantage of a vague standard with less rigorous requirements in regard to demonstrating actual consumer harm.  The third is, as I’ve argued here with Bruce Kobayashi, that existing contract and patent doctrine contain superior substantive provisions to deal with the patent holdup problem without imposing the heavy hammer of antitrust law which does not give the flexibility to distinguish between the paradigmatic hypothetical case of pure deception in the standard setting process with cases like N-Data which involve mundane contractual disputes and renegotiation.

The immediate effect of employing the antitrust laws to patent holdup problems that present as merely ex post contractual opportunism is that businesses will forgo procompetitive negotiations for free of antitrust enforcement that carries severe penalties.  A more reasonable solution would have been to allow the contracting parties to settle any dispute under an already existing alternative regulatory scheme, such as contract or patent law, with a comparative advantage in enforcement.  As the prevalence of intellectual property in standards grows, it is important to remember that SSOs are sophisticated organizations with a host of state and federal remedies at their disposal that do not involve the heavy hammer of antitrust law, which ultimately may deter procompetitive innovation.