Judge Sullivan and the UPP: Much Ado About Nothing or Articulating the Real Problem with the New HMGs?

Josh Wright —  14 May 2010

Much has been made of Judge Sullivan’s recent decision in City of New York v. Group Health Incorporated and its implications for the UPP test and market definition in merger cases under Section 7 of the Clayton Act.  Given the 2010 Proposed Horizontal Merger Guidelines’ (2010 HMGs) shift toward diversion ratios and margins and away from market shares, the blogosphere has sold Judge Sullivan’s decision as sign that the agencies might have a tough time selling the UPP to federal courts in the post-2010 HMG world.  For example, the WSJ Law Blog writes:

Now, we have the reaction from the first federal judge to actually be presented with UPP as the basis for attacking a merger: Phooey. Throwing out a court challenge by the City of New York to a merger of two health insurers, District Judge Richard Sullivan said he’d never heard of a single federal case in which UPP had been used.

Is this really a rejection of the UPP or the value of diversion as a test relevant to market definition?  I don’t think so.  In the case, the plaintiff moved to amend their complaint to, amongst other things, add the UPP test as an “avenue of proof” to establish the proffered market definition.  In a footnote, and in response to an expert declaration stating that the UPP could provide evidence “in addition to or instead of” traditional market definition analysis, Judge Sullivan noted that it could not find a federal court decision applying the UPP test and its invocation was insufficient to avoid summary judgment largely because additional discovery would cause undue delay and prejudice the defendants.  It should be noted that under the 2010 HMGs, the value of diversion is relevant to the question of market definition, but it is also independently relevant to the issue of unilateral effects. 

Now, to be sure, Judge Sullivan’s opinion states that a plaintiff in a Clayton Act case must define a market and that “the preferences of a single purchaser cannot define a product market.”  The statement about the insufficiency of a single purchaser market as a matter of law makes for interesting discussion in light of the language in the new HMGs that “the hypothetical monopolist test may suggest relevant markets that are as narrow as individual customers.”  With respect to the whether or not market definition is a necessary condition under Section 7, if one adopts the view of Professor Crane and myself that in their current form the 2010 HMG are at best equivocal as to whether the agencies must define a market, then the opinion may preview impending hostility to such an approach in federal courts.  But while I’ve blogged that I do not think the 2010 HMGs are clear enough of the necessity of market definition, I do not think that those with genuine concerns about the new HMG approach are really concerned that the Agencies will bring cases in which they do not define a relevant market.

Indeed, the real problem with the 2010 HMGs is not that the Agencies will avoid defining markets is at all.  The Agencies want to win cases.  And to the extent that federal courts expect markets to be defined, you can bet the Agencies will do so as part of their case in chief.  It is true that diagnostics for unilateral effects are based on the value of diverted sales and can be done without defining a market, but so long as this is part of an analysis that also defines a market at some stage, the Agencies can comply with the requirement that markets be defined under Section 7 of the Clayton Act.

Instead, the real problem arises because the Agencies now believe that narrower markets are more accurate because “the competitive significance of distant substitutes is unlikely to be commensurate with their shares in a broad market.”  The value of diversion test endorsed by the new HMG tends to lead to narrower markets.  Defining narrower markets will inevitably lead to circumstances in which the consumers in the narrowly defined markets are harmed, but others are benefitted.  I suspect the true concern of those skeptical of the “narrow” market approach adopted by the HMGs, which has not been articulated, is that the narrower markets obscure competitive benefits of the merger that are “outside” the market.  Thus, the new approach could lead to Section 7 liability for mergers that result in net increases in consumer welfare.

Consider the case when Firms A and B will merge and there is convincing evidence that harms will occur to a narrow group of customers but that prices will fall to other groups.  Further assume that the benefits are significantly greater than the harms.  As the new HMGs endorse narrower market definitions, this will occur more and more frequently and can be expected to arise over a broad swath of mergers involving alleged unilateral price effects.  Under current merger law, the merger of A and B will violate Section 7 despite the fact that it increases consumer welfare because Philadelphia National Bank precludes counting efficiencies outside the relevant market.  The merging parties cannot point to the consumer gains outside of the narrowly defined product market to defend the merger.

Note that this is not directly a problem that originates in the new HMG.  Both the 1997 (Section 4, n. 36) and 2010 HMGs (Section 10, n.11) contain notes indicating that the Agencies sometimes might consider efficiencies outside the relevant market if they are “inextricably linked with it that a partial divestiture or other remedy could not feasibly eliminate the anticompetitive effect in the relevant market without sacrificing the efficiencies in the other market(s).”  However, while the Philadelphia National Bank problem is nothing new, and certainly not created by either the 1992, 97 or 2010 HMGs, the value of diversion approach of drawing smaller circles around particular groups of harmed consumers is likely to dramatically increase the number of cases in which other groups of consumers benefit but those benefits are systematically excluded from merger analysis though they would otherwise meet the requirements of Section 10 (efficiencies) of the new HMGs.

To be sure, eliminating Philadelphia National Bank limitation on cross-market balancing would create some problems.  As a simplifying procedural assumption, it has some benefits.  For example, relaxing the rule would give rise to efficiencies claims outside the narrowly defined relevant market and therefore create the need for a balancing analysis.  Of course, these “out of market” efficiencies would still have to satisfy the requirements of Section 10 of the new HMGs.  While such balancing makes for a more complex analysis, that should not get in the way of incorporating cognizable efficiencies associated with the merger.  Also, the PNB rule flies in the face of the modern trend in favor of looking at actual competitive effects instead of simplifying and potentially misleading proxies like market shares.  The intellectual case in favor of excluding cross-market efficiencies is not a strong one, and it becomes even weaker when the Agencies adopt an approach of ever-narrowing market definitions.

As such, my view is that n.11 in the new HMGs should be expanded to discuss how they will apply their prosecutorial discretion in such cases.  This leaves an interesting question about what exactly the Agencies should do about this.  If the HMGs are meant merely to provide guidance, n. 11 does say that the Agencies will largely ignore “out of market” efficiencies but count them if the anticompetitive effects in the market are small and the efficiencies are large.  That’s some guidance.  And it is echoed in the 2006 Merger Commentaries.  But the value of diversion approach adopted by the new HMGs is likely to increase the demand for guidance on this score.  What does it mean for efficiencies to be “inextricably linked”?  How disproportionate do the efficiencies gains have to be relative to the anticompetitive effects?

But in addition to the guidance issue, I believe the HMGs should go further.  Just as the HMGs are searching for a more intellectual rigorous approach on the anticompetitive effects side, shedding imperfect market share proxies for a view that reflects the competitive realities of competition between close substitutes, the HMGs should adopt the same analytical approach with efficiencies.  Specifically, I believe the new HMGs should more strongly commit to taking into account the pro-competitive effects outside the narrowly defined markets.  What I have in mind is language that would indicate that the Agencies would not bring cases when they believe that the merger, taking into account competitive effects in the narrowly defined relevant market and elsewhere, is pro-competitive.

Such a commitment might go a long way towards relieving some of the skepticism about the narrower market approach by committing to not bring cases where it is clear that consumer welfare increases as a result of the merger even if some narrow group of consumers is harmed.  It might also explain the reluctance of federal judges, like Judge Sullivan, to sanction single customer market definitions despite Agency guidance and commentary to the contrary.  I do not believe such a commitment is likely under the current Administration.  But Republican administrations have also squandered the opportunity to address this issue in previous iterations of the Merger Guidelines and (as Steve Salop has pointed out) the Commentaries and so should not be let off the hook.  But perhaps now is the time.  If we are going to take the opportunity to make changes to HMGs to eschew misleading proxies for analyses that more accurately reflect competitive realities, it makes analytical and practical sense to do so more efficiencies and not just anticompetitive effects.