NY Times (and maybe Professor Hovenkamp?!) Confused About the Merger Guidelines

Josh Wright —  12 September 2011

The NY Times starts its op-ed against the AT&T / T-Mobile transaction with a false proposition about antitrust analysis of mergers: “The analysis begins with a mathematical formula for calculating the deal’s effect on competition.”  Any antitrust lawyer or economist will recognize the error.  A major change from the 1997 Horizontal Merger Guidelines to the 2010 version is that the former observes that agency analysis must begin with market definition and evaluation of concentration:

First, the Agency assesses whether the merger would significantly increase concentration and result in a concentrated market, properly defined and measured

However, the it is widely recognized that the 2010 Guidelines shed the cookie cutter, algorithmic approach to merger analysis in favor of a fact-intensive analysis involving multiple tools of which market definition and calculating market shares and evaluating concentration levels and changes is just one.  Indeed, the 2010 Guidelines expressly state:

The Agencies’ analysis need not start with market definition.

This is not trivial detail; these changes were at the very core of the changes in the new Guidelines promulgated by the Obama administration’s antitrust enforcement agencies.  The NYT analysis simultaneously relies exclusively upon market concentration statistics while appealing to the 2010 Guidelines which rejected that approach as authority.  Odd.  But not unsurprising.

What is more surprising is Professor Hovenkamp’s quote, whom we certainly can expect more from than the NYT.  Hovenkamp observes:

“It’s only a slight overstatement to say that if they weren’t going to block this one, the Justice Department might as well just throw the antitrust guidelines out the window,” said Herbert Hovenkamp, professor of law at the University of Iowa, who is considered by many to be the dean of American antitrust law. “This merger clearly seems to violate them.” …

“It was becoming legendary that the Bush administration wasn’t enforcing the old guidelines,” Mr. Hovenkamp said. “What good is a guideline that doesn’t provide any guidance? The Obama administration conceded that perhaps the old guidelines were too strict. So it made it easier, but at the same time said, ‘We’re going to enforce this.’ ”

I’ve got to believe Hovenkamp was quoted out of context here because, frankly, this doesn’t make much sense.  I doubt Hovenkamp would argue that the Guidelines’ thresholds were treated as gospel by any administration regardless of political ideology.   But what is absent from Hovenkamp’s discussion is the primary reason why the Guidelines expressly shifted away from concentration and toward direct analysis of competitive effects.  The answer doesn’t lie in politics.  Put simply, antitrust economists and lawyers at the agencies and elsewhere simply do not believe the HHI thresholds in the Guidelines provide a useful predictor for competitive effects.  The persistence of the HHI thresholds are at least somewhat a result of path dependency; despite some prodding, it proved too tempting for the agencies to keep the thresholds in given their appeal and general acceptance in merger precedent emerging in the 1960s and 70s.  But that was the age when those types of market structure arguments were in fact the economic state of art.  That is no longer true — and rejection of that general approach is a key (if not they key) component of the Guidelines’ evolution toward the current approach.

The theme of the NY Times article and the omission of any sense at all that the shift at the agency level has been the polar opposite of what is claimed — that is, away from treating HHI thresholds as gospel or even related to analysis of competitive effects and toward an analysis more directly focused upon competitive effects — I’m left puzzled by a few things in Hovekamp’s quote.  When the agencies have screamed from the rooftops that competitive effects and not market structure and market definition is what matters in merger analysis, the idea that not blocking a merger that nominally crosses otherwise meaningless thresholds in agency Guidelines threatens the rule of law or means that we ought not have Guidelines is at the very least overstated.  Of course, one could interpret the statement as a critique of leaving the thresholds in the Guidelines at all if one is not going to enforce them.  I agree with that.  But they’ve always been there and often been ignored when the agencies’ analysis concluded the merger would not harm consumers.

And of course, that interpretation is difficult to square with the statement that this “merger clearly seems to violate them.”    Violate them?  The Guidelines do not have the force of law.  If this merely means something like “the merger appears to be one that the agencies’ analytical framework articulated in the Guidelines indicates that they will challenge” — that’s fine.  But, that statement suffers the same analytical flaws described above. Violating the Guidelines would require a showing that the merger was likely to create market power and produce anticompetitive effects — to do so under the new Guidelines requires more than a simply counting the number of firms.  That type of analysis no longer passes muster in antitrust analysis at the agencies.  To claim a merger “clearly seems to violate” the Guidelines  by sole reference to the HHI thresholds at the same time the agencies have distanced themselves from them(in favor of more fact-intensive and direct analysis of competitive effects) is not consistent with the economic letter or spirit of the new Guidelines.

One response to NY Times (and maybe Professor Hovenkamp?!) Confused About the Merger Guidelines


    The Times author is obviously out of his depth — even if we still had the old guidelines, no one who knew what they were doing would start with HHI. Now it’s hardly worth mention. Nonetheless, it’s also pretty obvious that the author misunderstood Hovenkamp (he should have sent him a draft of the article to give the Professor a chance to spare him the mess) and just quoted him the way he thought supported the article. If you take Hovenkamp’s quote on it’s own, it makes sense — it’s a pretty solid argument to make that the current guidelines suggest this merger is anticompetitive. Moreover, the way the agencies have been doing things for some time suggests that a 4-3 (really a 3-2 among the majors) merger is going to get push back.