Antitrust & Competition Policy Blog is hosting a symposium on Competition in Agriculture.  Mike’s post from yesterday is available here.   So far in the symposium there are also posts by Ron Cass (BU Law), Jeff Harrison (Florida Law), Peter Carstensen (Wisconsin Law), and Kyle Stiegert (Wisconsin Applied Econ).  Additional posts should be forthcoming from Christina Bohannan (Iowa Law), Andrew Novakovic (Cornell Applied Economics), and the great George Priest (Yale Law), who I hope gets the blogging bug.

Josh, Scott Kieff and I have posted a short comment based on our submission to the DOJ/USDA Workshops on Agricultural Competition, co-authored by us and Mike. The comment should be available for download from the DOJ webpage when the public comments are posted (someday . . . ).  A copy is also available here (www.laweconcenter.org), and comments are most welcome at gmanne@laweconcenter.org Please leave comments on this post over at the A&CP Blog.

Regarding firm size and integration, it must be kept in mind that the agriculture industry in the U.S. has, for good reasons, moved beyond the historic, pastoral image of small family farms operating in quiet isolation, devoid of big business and modern technologies. The genetic traits that give modern seeds their value—traits that confer resistance to herbicide and high yields, for example—are often developed through processes that are technologically-advanced, time- and money-intensive, risky investments, and subject to various layers of regulation. It doesn’t take expertise in industrial organization to imagine why at least for some participants in this market these processes are likely to be more efficiently and effectively conducted within large agribusiness companies having enormous research and development budgets and significant expertise in managing complex business and legal operations, than they are by the somber couple depicted in the famous 1930 Grant Wood painting, “American Gothic.” Nor is such expertise required to imagine why complex contracting across firms, of any size, is likely to be of significant help in supporting the specialization and division of labor that is useful in allowing some businesses (even a small family farm is a business) to be good at planting and harvesting while others are good at inventing, investing, managing, developing, testing, manufacturing, marketing, and distributing the next wave of innovative crop technologies. This requires on the one hand that the government give reliable enforcement to contracts and property rights whether tangible or intangible (extremely important in this industry are patents, trade secrets, and even trademarks), while on the other hand it allows firms wide flexibility to decide for themselves which of these contracts and property rights they would like to enter into or obtain pursuant to the applicable bodies of contract and property law.

When courts and regulatory agencies like the DOJ Antitrust Division adopt special approaches to the body of antitrust law to address concerns that may arise from these property rights and contracts, they run the risk of crafting doctrines that inappropriately override well-established bodies of law that are informed by longstanding judicial and scholarly thought and consideration of each area, and creating the potential to reduce innovation and economic growth. A central countervailing concern is that the putative antitrust injuries that might arise are rooted in stylized economic models that are heavily dependent on a narrow set of assumptions, leaving significant room for erroneous antitrust enforcement. A modest but fundamental safeguard to protect against this concern of “false positives,” is an approach to antitrust that requires a strong demonstration of actual anticompetitive effect as a precondition for a monopolization violation.

Not only are patents not presumptive proof of market power in any static sense, but patents can also meaningfully improve both competition and access to patented technologies over time, in the dynamic sense. From the public record it appears that the driver of much of today’s antitrust enforcement in the agricultural industry boils down to intervention into business disputes between large and sophisticated parties. The inherent uncertainty regarding the economic consequences of specific conduct, coupled with competitors’ poor incentives and the huge costs of error, counsel strongly against antitrust intervention without strong empirical evidence that the conduct has reduced competition and harmed consumers in the form of higher prices, lower quality, or reduced innovation.

Alan Meese (William and Mary) and Barak Richman (Duke), have an op-ed over at the Huffington Post on the Ticketmaster Live nation merger and settlement.  They frame the DOJ decision to approve the merger as a victory of principle over politics and economic populism.  Here is an excerpt:

Many hoped that the Live Nation-Ticketmaster merger would fall prey to a new economic populism. When the companies announced their plans to merge, some characterized the merger as a consolidation of “entertainment powerhouses” designed to inflate ticket prices and squeeze consumers. Public figures, including none other than Bruce Springsteen, condemned the combination. Members of Congress piled on, characterizing the transaction as a naked combination of industrial titans and demanding action from antitrust enforcers.

Meese and Richman go on to argue, based on their own analysis, that the merger would have been pro-competitive:

The Live Nation-Ticketmaster merger would have been another procompetitive victim to an angry public. Our careful analysis of the proposed merger reveals that it is much more a response to Schumpeterian technological change than an effort to concentrate market power. In other words, the companies are combining forces to pursue an innovative business model, one that pursues new consumer demands and responds to the rise of electronic music. It is not an attempt to acquire a stranglehold over an industry that technological change has made increasingly resistant to strangleholds.

The populist anger directed at the proposed merger — which was in no small part fueled by the companies’ smaller competitors who feared having difficulty competing effectively against the new company– characteristically did not discern the complexities of the industry and evaluate the merger’s likely competitive impact. Of course, few in Washington brake for complexity. Which is why it is a relief the Obama administration did.

I’ve not yet read the analysis, but plan on doing so soon.  In the meantime, here is how Meese & Richman close:

Even while the Obama Administration might engage in antitrust saber rattling, its approval the Live Nation-Ticketmaster and the associated consent decree shows the triumph of economic reasoning that is often counterintuitive to policy advocates. Its settlement further extracts concessions that further enhances competition, promotes innovation, and protects consumers. It is the commendable product of careful analysis reflects a deliberate navigation across the minefield of antitrust politicization.

I’m left with two questions.

The first, which I hinted at here, is that the combination of the structural fix coupled with the non-retaliation provisions strikes me as somewhat odd.  If the structural fixes restore competition, then why the need for restricting vertical contractual arrangements as between Ticketmaster and venue owners/ customers?  Here is DOJ Competitive Impact Statement.  Meese and Richman describe the settlement as extracting concessions that are pro-competitive and the product of careful analysis.  I agree with the authors that refusing to succumb to economic populism when the data and analysis go the other direction is a victory for rigorous antitrust analysis.  But I’m not yet convinced that the conditions are to be celebrated.  So, I wonder what the authors mean when they say the  concessions are pro-competitive, and in particular, whether the conduct fix can be said to be pro-competitive if the structural fix restores any perceived competition problem?  Or is the DOJ just hedging?

The second is a rhetorical question: I wonder whether Intel feels like despite the Obama administration’s “saber rattling,” they can rest assured that the invocation of Section 5 is a sign that economic reasoning will triumph over populism and politicization?

Yesterday the final Horizontal Merger Guidelines Review workshop was held and, among other antitrust luminaries, our own Josh Wright participated.  We look forward to a report from the front lines.

Meanwhile, Assistant Attorney General Varney’s comments are available on the interwebs.  Overall her remarks seem uncontroversial, especially following on the heels of the agency’s (surprising?) clearance of the Live Nation/Ticketmaster merger with conditions (but see the agency’s challenge of the consummated Dean Foods/Foremost Farms merger, about which I will have more to say in a subsequent post).  But I did find one section quite a bit troubling.  Acknowledging that agency practice did not hew slavishly to the Guidelines’ “five-step analytical process” for assessing markets and market share, Varney noted that:

Implicit in deemphasizing the sequential nature of the Guidelines inquiry is a recognition that defining markets and measuring market shares may not always be the most effective starting point for many types of merger reviews. Remember, the purpose of defining a market and assessing shares is to assess potential harm. When it is clear, for instance, that either certain vulnerable customers are likely to be harmed by a merger, or that certain customers have in fact been harmed by a consummated merger, the need to define a market to assess likely competitive effects is diminished. For instance, the consumer harm that followed from the consummated Evanston hospital transaction lessened the importance of the Commission’s market definition and market share analyses in that matter. Our panelists have largely confirmed the view that market definition should not be an end-all exercise. Rather, it is something to be incorporated in a more integrated, fact-driven analysis directed at competitive effects.

I am among the many commenters who have criticized the Guidelines’ approach to market definition and market share–my submission to the workshops is here.  There has also been a strong movement recently to do away with market definition in some unilateral effects analysis and to replace it with the UPP analysis promoted most recently in the Farrel & Shapiro article (pdf).  Interestingly, while Varney is previously  on record opposing this movement, elsewhere in this speech she seems to endorse it:

There is a growing body of evidence that measures of upward pricing pressure, which focus on diversion ratios, and price-cost margins, can be highly informative in assessing the likelihood of unilateral pricing effects.

But this is in a different section of the speech, UPP remains an analytical approach (as opposed to the class of cases Varney is concerned with here where harm to certain consumers is simply “clear”), and it does not seem to be what she’s talking about in the quote above.  Here she seems to mean something else–and I fear it is something troubling.

Taken literally, what Varney is saying is that an ad hoc (ok, fine–an “integrated, fact-driven”) determination that some customers (”vulnerable” ones, whatever that means) may be made worse off by a merger lessens the need for a more comprehensive assessment of overall competitive dynamics within a relevant market.  But I don’t know what this means, frankly.  In the first place, how is the agency supposed to know that some customers are likely to be harmed if it hasn’t assessed the availability of substitutes and the extent of diversion?  One can certainly criticize the method by which this assessment is made, but a conclusion of harm absent this assessment seems absurd.  Moreover, if Varney really means that all that is required to condemn a merger is that any customers may be harmed, no matter how many are also benefited, at a minimum it sounds like she’s writing the efficiencies defense out of the Guidelines, but she may even be justifying condemnation of any and all mergers–after all, how many actions in the marketplace impose a cost on literally no one?  If, as seems likely, it is inframarginal consumers who are “likely” “vulnerable” to price increases (where “vulnerable” may be a synonym for “having inelastic demand”), then this test is a repudiation of the entire economic edifice of modern merger analysis (parallel to my discussion of the DC Circuit’s Whole Foods decision here).

And Varney’s reference to the FTC’s Evanston Northwestern case is a bit of a sleight of hand.  That was indeed a consummated transaction, where the requisite harm was shown by direct pricing evidence following the merger.  That’s quite a bit different than tossing out the Merger Guidelines in a non-consummated merger case because it is “clear” that “vulnerable” consumers are “likely” to be harmed.  And even the Evanston Northwestern case is not without controversy, precisely because forsaking the Guidelines’ analytical framework also forsook clarity in the analysis (see, for example, the strong criticism of the case here).

According to the Guidelines themselves,

The unifying theme of the Guidelines is that mergers should not be permitted to create or enhance market power or to facilitate its exercise. Market power to a seller is the ability profitably to maintain prices above competitive levels for a significant period of time.

The presence of some harm (how much, by the way?) to some consumers does not necessarily equate to market power, unless the definition is simply tautological.  Under the Guidelines approach, this would require a market definition so narrow (defined to include only the harmed customers) that it would be economically meaningless (the classic “red-haired, bearded, one-eyed, man-with-a-limp classification” condemned by Justice Fortas in his Grinnell dissent).  Sidestepping the Guideline’s analytical framework by equating the exercise of market power with a theoretical price increase that wouldn’t be cognizable under the Guidelines (and wouldn’t exist in the real world) is not merely an analytical shortcut, it is a subversion of the whole analysis.  Again, see the fundamental errors of the Whole Foods case.

Now, in the end, all she may be saying is that sometimes there is direct evidence of harm, properly-statistically attributable to the merger, many years after a transaction has been consummated.  Or that the risk of harm is so self-evident that a formal analysis isn’t required–say, when there are simply no other competitors in a relevant geographic area, no timely entry is possible (for some reason . . . ) and a significant number of customers is affected.  I suppose this could happen.  I would expect in such circumstances that the parties wouldn’t even bother attempting the merger, but maybe once in a while the situation could arise.  But I just can’t fathom that this could be a significant enough possibility that it would rise to the level of an important policy speech on the Merger Guidelines by the AAG.

So what is Varney saying?  Anyone?

Thanks to all of our participants in the Merger Guidelines Symposium.  We hope many of you, as well as our readers, will look back over the collected posts and engage in an ongoing dialogue in the comments over the many interesting ideas raised here.  You will find all of the posts from the symposium by clicking on the “merger guidelines symposium” link to the right, under the “blog symposia” heading.  For you reference, we have also collected all of the posts below.

For our friends in the agencies, we hope this symposium gives you some food for thought and a sense of where an informed subset of the antitrust community see issues meriting attention.  It is interesting to note (we received all of the submissions before any were posted) the extent to which so many of these comments focused on the problems of the structural presumption (including market definition and market share calculations), HHI thresholds and unilateral effects–all obviously related matters, most informed by a concern for aligning the Guidelines with agency practice.

Although there was not a lot of discussion of the related matter of the recent Farrell/Shapiro proposal on unilateral effects analysis without market definition, many of the posts provide a backdrop against which discussion of the proposal is sure to arise.  For my own parting thought on the issue, I would note this:  Whatever the problems with market definition in unilateral effects analysis, I hope we will not be too quick to jettison the exercise in exchange for the Farrell/Shapiro proposal (or similar).  Although imperfect, the market definition exercise exerts some check on merger enforcement that is absent from the Farrell/Shapiro proposal.  And in the absence of extremely good empirical data (as in Staples),  I fear that the UPP analysis may permit enforcers to find competitive effects in a much wider range of cases than would be desirable.  As essential background to assessing this trade-off in the modern economy, I refer our readers to the Sidak/Teece submission to the symposium and to their forthcoming paper.

Thanks again!

• Joe Farrell, Welcome from the Agencies
• Dennis Carlton, Revising the Guidelines
• Luke Froeb, Merger Enforcement without Market Definition
• Robert Gertner, Revising the Guidelines
• Steve Salop, Comments on Updating the Merger Guidelines
• David Teece and J. Gregory Sidak, Dynamic Competition and Antitrust Law
• Einer Elhauge, Do the Merger Guidelines Need Revision?
• Joshua Wright, The Guidelines Should be Revised to Reject the PNB Presumption
• Andrew Gavil, Andrew Gavil on Revising the Merger Guidelines
• Daniel Crane, Let Sleeping Dogs . . .
• Herbert Hovenkamp, Herbert Hovenkamp on Revising the Merger Guidelines
• Thom Lambert, “Standardizing” The Horizontal Merger Guidelines
• Geoffrey Manne, Fix the Supply Side
• Danny Sokol, Small Changes
• Joseph Simons, Revisions to the Merger Guidelines: Above All, Do No Harm

My sense is that there is no need to revise the DOJ/FTC Horizontal Merger Guidelines, with one exception.  As Greg Werden points out, “a thorough revision would take up to three years and occupy some of the agencies’ best people for a total of more than two thousand hours.” The current guidelines lay out the general framework quite well and any change in language relative to that framework are likely to create more confusion rather than less.  Based on my own experience, the business community has had a good sense of how the agencies conduct merger analysis.  The only exception is the Guideline’s concentration thresholds, which have been obsolete in practice since at least the mid 1990s.  Those thresholds should probably be changed to reflect actual practice.  If, however, the current administration intends to materially change the way merger analysis is conducted at the agencies, then perhaps greater revision makes more sense.  But even then, perhaps the best approach is to try out some of the contemplated changes (i.e. in actual investigations) and publicize them in speeches and the like before memorializing them in a document that is likely to have some substantial permanence to it.

Chairman Leibowitz in his Fordham speech last month stated “From my perspective, the current Guidelines do not explain clearly enough to businesses how the agencies review transactions.” Won’t that always be the case? In a specialized area of complex regulatory law, there is whatever guidelines an agency (or in our case agencies) will promulgate and then small group of insider lawyers who will have enough repeat business to really understand the meaning of the guidelines via their agency contacts. For others, the guidelines will remain unclear unless you create a 600 page set of merger guidelines – not a good idea.

Maybe the lack of clarity is a problem, particularly with judges. After all, judges play a role in enforcing the law. The Guidelines cover this aspect fairly early on in Section .01. However, who is the end user of the Guidelines? Is it the firms practicing before the agencies or for judges to use? If the Guidelines have a purpose for judges, there is an important institutional issue at play as we have not had a merger case before the Supreme Court for many years. This leads to a meta-question — Is the law Section 7 or is it the Guidelines as defined by the agencies at any given time?

I do have three points as to the substance of the Guidelines:

1. The Guidelines have only a brief mention of R&D (Section 4), in which they state that even if synergies are substantial, they “are generally less susceptible to verification and may be the result of anticompetitive output reductions.” This treatment seems incomplete. After all, the rise of the new economy and of a greater role for IP suggests that there should be more guidance with regard to R&D efficiencies.

2. Let us start using HHIs that have some remote relationship to actual enforcement decisions. Although HHIs are only rough preliminary screens, and other factors make the real difference, the agencies ignore the HHI thresholds in the Guidelines in making enforcement decisions. The rub is that the agencies then cite the HHIs in district court to show how bad the violation is.

3. The treatment of unilateral effects needs to be revisited. The concept as used in practice needs to be better reflected in the Guidelines.

Do the Merger Guidelines need revision? No.  Thanks for having me.

OK. Yes–and market definition/market shares, and in particular the effective incorporation of supply-side effects, seem to me like the most pressing issues in need of attention.

Judge Posner, in the first edition of his justly celebrated Antitrust Law, noted that market definition was an unfortunate means to an end, a necessity given the inability of our analytical tools effectively to assess the effects of mergers beyond a circumscribed boundary.  As Posner noted,

The importance attached to defining a market in which to appraise the competitive effects of a challenged merger is on more example of the law’s failure to have developed a genuinely economic approach to the problem of monopoly.  If we knew the elasticity of demand facing a group of sellers, it would be redundant to ask whether the group constituted an economically meaningful market. . . . It is only because we lack confidence in our ability to measure elasticities, or perhaps because we do not think of adopting so explicitly economic an approach, that we have to define markets instead.

I doubt we would shun the approach for being too explicitly economic today, but I think a very small number of cases permit us to identify competitive effects with sufficient confidence—Staples being the paradigmatic case.  We may be stuck with market definition, but the outdated conception enshrined in the Guidelines–of measuring concentration in a static market defined essentially by demand-side elasticities–can surely be improved upon.

In the first instance, we should remove any hint of a concentration/price presumption so intimately tied to measuring market definition and HHIs.  Although the Guidelines attempt to cabin the market definition and market share analyses as “starting points,” in practice they are the beginning and end of most merger cases.  The Merger Guidelines (and, importantly, agency practice) should be revised to limit reliance on market definition and market share, at a minimum by stating explicitly that the definition of the market and the calculation of market shares are not sufficient to indicate adverse competitive effects, and perhaps also by removing HHI threshold discussions which seem to imply the same.  Even if the agencies and the Guidelines don’t mean these tools to be used in this way, courts haven’t really gotten the message.

At the same time, the Guidelines should explicitly incorporate supply side elasticity into the market definition inquiry.  There is little defense of the Guideline’s statement that “Market definition focuses solely on demand substitution factors—i.e., possible consumer responses.”  While the Guidelines and actual practice do attempt to make some allowance for supply-side effects, these allowances seem like afterthoughts, and I think it is rare that HHIs are calculated to incorporate production capacity not currently devoted to the narrowly-demand-defined product market, especially outside of the commodity realm.  Meanwhile, even a small bias against supply substitution, entry and unforeseen competitors (and/or new products) is a particular problem in fast-shifting, innovative industries where this is precisely whence the most significant competitive threat will come.

To the extent (and it is a large extent) that market definition and market share are far-removed from competitive effects, they should be more carefully circumscribed by the Guidelines.  The economic irrelevance of much of the evidence used to define markets and the general disregard for supply-side response help to ensure that market definition, while incredibly important in litigation, is not actually all that helpful.  At the same time, the general lack of correlation between concentration and unilateral effects makes reliance on the calculation of market shares (of crabbed markets, often disregarding supply-side effects) similarly misleading and prejudicial.

I’m confident that my esteemed colleagues, who have far more expertise about the merger guidelines than I, will offer all sorts of terrific ideas for revising the substance of the guidelines. While I would certainly advocate a few specific changes (i.e., revise the HHI thresholds to reflect actual agency practice), I’ll leave the devilish details to the experts and concentrate on one “modest” (quite literally!) revision:

I would encourage the antitrust agencies to clarify, within the actual text of the guidelines (i.e., not in mere commentary like that issued in 2006), that the guidelines are not the law, should not be treated as such by the courts, do not exhaustively specify when a merger will or will not be anticompetitive, and should be flexibly implemented to account for case-specific factors that cannot be specified ex ante. In short, the guidelines should explicitly acknowledge that the ultimate question in any horizontal merger case requires the application of a standard, not a rule.

A rule is a legal mandate that entails an advance determination of what conduct is permissible and leaves only factual issues for the adjudicator (e.g., “Do not drive faster than 65 m.p.h.”). A standard, by contrast, is a mandate that leaves to the adjudicator both factual issues and some judgment about what conduct is permissible (e.g., “Do not drive at an excessive speed.”). Rules provide superior guidance to the governed and the adjudicator, but they can misfire if over- or underinclusive, and they therefore require ex ante specification of all factors that might be relevant to a sound decision. Standards provide less guidance, but they are more likely to generate a correct adjudication in any particular case, for the adjudicator is free to account for unforeseen, case-specific quirks.

The legal question at issue in a horizontal merger case — “might the business combination substantially lessen competition or tend to create a monopoly?” — requires the ultimate adjudicator to apply a standard, not a rule. It is simply impossible to specify ex ante all the considerations relevant to answering this question. Accordingly, to the extent the merger guidelines are viewed by courts as rules for separating pro- from anticompetitive mergers, they are bound to generate incorrect adjudications when they inevitably misfire.

Now I realize the merger guidelines, as currently drafted, do not purport to bind courts and do state that their “mechanical application . . . may provide misleading answers” and that they should be applied “reasonably and flexibly to the particular facts and circumstances of each proposed merger.” I don’t believe this is enough. The fact is, generalist judges asked to resolve the complicated economic disputes in a horizontal merger case are reluctant to veer beyond the guidelines’ prescribed analysis and, as a practical matter, treat the guidelines as though they are, in fact, the law. In the D.C. Circuit’s Whole Foods decision, for example, the majority chided the dissent for having incorrectly applied the merger guidelines. Given that the merger guidelines simply cannot exhaustively specify all the considerations relevant to evaluating a proposed merger, courts’ treatment of them as the final word implies that relevant considerations will get left out.

Take Whole Foods for example. A key fact in that case was that the vast majority of shoppers who buy from so-called “premium natural and organic supermarkets” (PNOS) also shop regularly at conventional grocers. Thus, if a combined Whole Foods/Wild Oats were to raise prices on items available at conventional supermarkets, buyers would likely just start buying those items on their conventional grocer outings rather than on their PNOS outings. Unfortunately, nothing in the merger guidelines calls for a consideration of “cross-shopping,” and this important argument therefore got short shrift. Had the guidelines explicitly stated: “This is not the law. We can’t state up front all relevant considerations. Courts should credit plausible arguments based on factors not stated herein,” this important argument might have gotten the attention it deserved.

Or take considerations relevant to dynamic (Schumpeterian) competition. While I am sympathetic to the Sidak/Teece arguments that the merger guidelines should account for dynamic competition concerns, I simply can’t figure out how one would write a rule that would do that. How can we specify ex ante all the considerations that are relevant to whether a business merger will enhance dynamic, though not necessarily static, competition? There may be an answer to that question — and I much look forward to hearing from Sidak and Teece on this issue — but I don’t know what it is. An alternative approach would be to free up the ultimate adjudicators — the courts — to account for dynamic competition considerations by disabusing them of the notion that they must treat the merger guidelines as law. Parties could then articulate their dynamic competition arguments on a case-by-case basis, and the courts could credit those that appear to have merit.

The main objectives of the merger guidelines, I assume, are (1) to deter combination attempts that would harm competition (i.e., those that would clearly be subject to challenge); (2) to avoid deterring combinations that would not harm competition (i.e., those within a safe harbor); and (3) to assure some consistency across the regulatory agencies and across administrations. These objectives could still be attained — and greater accuracy in outcome could be achieved — if the merger guidelines specified that the ultimate inquiry involves application of a standard rather than a rule.

1. Yes, the Merger Guidelines should be revised; in particular:

a.  The discussion of concentration thresholds for collusion facilitating mergers must be aligned more closely with both recent case law and actual enforcement practices; otherwise they fail to provide guidance.  The current Guidelines indicate that concentrations greater than 1800 HHI and a post-merger increase exceeding 100HHI presumptively indicates a challenge. In fact, mergers with post-merger HHIs in excess of both these numbers are routinely permitted. While the standards in the current Guidelines are too aggressive, the George W. Bush administration policy was too lenient.  More fundamentally, the HHI creates an illusion of precision in coordinated effects analysis that is simply not warranted, particularly not when market definitions are ambiguous or when the merger market is subject to product differentiation.  Further, the “other factors” portion of the Guidelines tends to dominate the analysis.  A better approach is reduced reliance on the HHI and more on simpler observations about who the 3 or 4 largest firms in the market are, the effects of eliminating the acquired firm as an independent market entity, and the likely responses of rivals to an output reduction by the post-merger firm.

b.  The unilateral effects analysis requires elaboration that brings it into line with actual agency analysis.  Some of the economists would like to jettison a market definition approach and attempt to measure price increase potential more directly.  While I am sympathetic with that, there is too much water over the bridge; it could lead to courts’ simply rejecting the Guidelines.  A better approach is to write market definition criteria that permit narrower market definitions in situations where a unilateral price increase is likely.  This requires that some thought be given to situations where the merger is between one firm and its second closest rather than its closest rival.  Right now it is simply too easy for defendants to argue that the closer rival will steal sales and competition will not be affected.  So the government is currently losing cases that it should not be losing.  Finally,the Guidelines should quell discontent in lawyers’ perception about unilateral effects analysis that it is more speculative or less rigorous than traditional coordinated effects analysis.  The fact is that both types of analysis rely on many assumptions, some of which are no more than conjectures about future firm behavior.  If anything, the robustness edge belongs to the unilateral effects approach.

I feel no great urgency to revise the Guidelines.  True enough, they’re more of an analytical thought experiment than an accurate description of how merger review takes place in the agencies, but who’s really fooled?  Perhaps some business people think that the Guidelines are a computer program waiting for the introduction of the relevant data to spit out the answer, but most sophisticated executives contemplating a merger will understand that the Guidelines are just a beginning point for conversation.

Could the beginning point be clearer or conform more closely to agency practice?  Sure, but that doesn’t mean that revision of the Guidelines is justified.  With hindsight, the First Amendment could be a little better worded, but no one wants to tinker with it now–who knows what would result?  I’m sufficiently satisfied with current merger practice in the agencies that I don’t care that much what’s in the Guidelines and I am worried about the unpredictable results that could obtain if we started tinkering.  Let sleeping dogs . . .

But if we are going to revise, then my pet issue is the asymmetrical treatment of the probabilities on anticompetitive effects and offsetting efficiencies–a point on which Joe Simons and I are planning a fuller analysis.  The Guidelines seem to suggest that if the probability of anticompetitive effects of magnitude 100 is 50% and the probability of offsetting efficiencies of magnitude 100 is 50%, then the merger should be challenged, since a greater quantum of proof is required for efficiencies than for anticompetitive effects.  This makes no sense to me–everything else being equal, efficiencies that would be passed on to consumers and market power increases should be given equal weight and not assigned separate probability standards.

1.  Do the Merger Guidelines Need Revision?

Yes.  Conceptually, the current Guidelines incorporate multiple strands of intellectual and legal history with respect to merger analysis that have been layered one upon the other over time, but never effectively integrated.  This now encumbers the application of the Guidelines and may be inhibiting the government’s capacity to effectively and efficiently initiate merger challenges.

The current Guidelines remain strongly tethered to the “structural” school, with roots in the 1950s.  The influence of the structural school was evident in both the Philadelphia National Bank presumption and the 1968 Guidelines, and the 1982-97 Guidelines have retained the essential features of that tradition — reliance on market definition, market share calculation, and inferences about market power drawn from those shares.

The 1982 Guidelines also began the process of layering on top of that tradition, injecting elements of the “new learning,” which focused largely on expanding upon General Dynamics and the various bases for rebutting a structural case, and oligopoly theory.  As the Guidelines further evolved through 1997, various dimensions of oligopoly theory were added, especially in the sections on anticompetitive effects, reflecting the work of Stigler, Nash, Cournot, and Bertrand.  Yet additional economic thinking was inserted in the revised and expanded discussions of entry and efficiencies.  Today’s Guidelines, therefore, are an intellectual collage of various traditions in law and economics.

2. If yes, what is the most important revision that you would recommend and why?

Although it is a tall order, the agencies could undertake to clarify the overall conceptual framework of the Guidelines.  Currently, even though it does not describe actual agency practice, the Guidelines are perceived by courts as outlining a linear process, in which its five steps are mechanically undertaken in sequence, each proceeding only after the previous step is completed.  This in effect makes structural analysis a prerequisite to the evaluation of every merger and it impairs reliance on any other approach to evaluating likely anticompetitive effects or efficiencies more directly.  This approach wrongly elevates the status of market definition above competitive effects, which is the core concern of merger analysis – indeed of all antitrust analysis.

A related goal would be to harmonize the intellectual underpinnings of merger analysis with other areas of antitrust law.  This is especially important with respect to courts’ willingness to rely on direct effects evidence since the Supreme Court’s decisions in NCAA and Indiana Fed’n of Dentists.  Depending on the kind of evidence available, unilateral effects theory can be understood as an application of the direct effects approach to mergers.  In cases like Staples, for example, where natural experiments provided a basis for predicting the likely future effects of the merger, it is a sort of “predicted actual effects” doctrine.  As with NCAA and Indiana Fed’n of Dentists, in such cases defining markets and inferring market power indirectly should be understood as a surrogate that is unnecessary when more direct and more reliable indications of future market power are available.

A third and also related goal might be to more openly acknowledge that the Guidelines influence how courts assign burdens of production and proof.  It has become tradition for all government Guidelines to pronounce that they are not intended to specify burdens of proof or production; that instead, they simply outline the government’s internal processes for exercising prosecutorial discretion. But these assertions are at odds with actual practice once the government enters the courthouse.  The Merger Guidelines have been read to specify burdens: Steps 1 and 2 (market definition and anticompetitive effect) shift the burden to the defendant, who can then seek to rebut the prima facie case based on entry, efficiencies, or a failing firm defense.  And courts such as the D.C. Circuit in Baker Hughes and Heinz, have promoted the use a “sliding scale” to establish variable-strength presumptions and to evaluate burdens, often relying heavily on structural evidence as a burden shifter, a tradition rooted in Philadelphia Nat’l Bank.  Given that the Guidelines are applied this way, the agencies should consider saying a bit more about what should be sufficient to shift a burden of production in either direction and whether certain kinds and quantities of evidence shift a burden more emphatically than others.

The specification of burdens also has implications for the welfare standard being used to assess mergers.  This is most obvious in the efficiencies discussion, which currently tries to nuance the issue.  The text strongly suggests a commitment to a consumer welfare standard defined as giving greater weight to consumer surplus.  But there is a hint of equivocation in footnote 37, which contains some support for the argument that aggregate welfare may be a relevant consideration.  More clearly specifying the relative burdens of production could eliminate any remaining ambiguity regarding the welfare standard.  Although it is unlikely that very many cases would be influenced by the choice of a consumer welfare or aggregate welfare standard, the courts have generally applied the consumer welfare approach now dominant in the Guidelines and this should be more clearly acknowledged.

Finally, post-consummation merger challenges have become more common and yet the Guidelines are focused entirely on making predictions about likely future effects.  The agencies might consider either adding a section on post-consummation mergers or generating a separate guidance document that would do so.  The goal would be to explain how the post-consummation analysis of a merger might differ from those done pre-merger.  This again highlights the need to integrate thinking about the role of actual effects evidence, which becomes more central in post-consummation challenges.  Such a discussion also could include consideration of limited reliance on abbreviated or “quick look” types of analysis to shift burdens in appropriate post-consummation cases and reduce the costs of merger review.

The Day 1 posts are up and available.  But we’re not done yet.  We have seven more coming tomorrow from Dan Crane, Andy Gavil, Herbert Hovenkamp, Joseph Simons, Thom Lambert, Geoff Manne, Danny Sokol and Paul Yde.

In the meantime, please peruse the first installment of posts and feel free to comment!

• Joe Farrell, Welcome from the Agencies
• Dennis Carlton, Revising the Guidelines
• Luke Froeb, Merger Enforcement without Market Definition
• Robert Gertner, Revising the Guidelines
• Steve Salop, Comments on Updating the Merger Guidelines
• David Teece and J. Gregory Sidak, Dynamic Competition and Antitrust Law
• Einer Elhauge, Do the Merger Guidelines Need Revision?
• Joshua Wright, The Guidelines Should be Revised to Reject the PNB Presumption

Yes, the Merger Guidelines should be revised.

The Guidelines primary purpose is to “articulate the analytical framework the Agency applies in determining whether a merger is likely substantially to lessen competition.”   While the Guidelines have been very successful in articulating a useful economic framework for analyzing mergers, their performance in terms of satisfying that goal has been largely mediocre.   If the goal articulated by the Guidelines is an important one, and I do believe that providing some modicum of legal certainty to businesses contemplating merger decisions is valuable goal, then the Merger Guidelines do need revising in order to fulfill their purpose.

The mere fact that agency practice is much more predictable now than it used to be does not warrant inaction when one considers the Guidelines in light of their primary purpose.  Indeed, that the agencies consistently apply different standards to the Clayton Act 7 standard than what appears in the Guidelines is no more a defense of the Guidelines than the fact that nobody cites Vons Grocery or Pabst is an argument for keeping those decisions on the books.

So in my view, the Guidelines ought to be revised with an eye  toward changes that will bring the Guidelines in line with actual agency practice and update the Guidelines to reflect modern economic thinking and empirical evidence.

What is the first revision I’d make and why?

Following these principles, I think revision discussions should focus on low hanging fruit that satisfies these criteria.  There are a number of areas that arguably qualify, but let me start with one simple one:  The Agencies should revise the Guidelines to affirmatively abandon use of the structural presumption to demonstrate a substantial lessening of competition.  Of course, the Guidelines tell us that they are not designed to inform parties “how the Agency will conduct the litigation of cases that it decides to bring,” use of the structural presumption is fundamentally about economics and not litigation strategy.

Indeed, in its formulation in Philadelphia National Bank,the Supreme Court held that:

A merger which produces a firm controlling an undue percentage share of the relevant market, and results in a significant increase in the concentration of firms in that market is so inherently likely to lessen competition substantially that it must be enjoined in the absence of evidence clearly showing that the merger is not likely to have such anticompetitive effects.

While the structural presumption has been eroded from the days of Philadelphia National Bank when it was invoked in the context of a merger creating a firm with 30% share of the market, it also has not disappeared.  And it is important to note that the justification for the presumption was its economic content.  Indeed, the Court noted that the size of mergers above the presumptive level of concentration made them “inherently suspect” and that the test was “fully consistent with economic theory,” and forged “common ground among most economists.”

There are two important reasons why the Agencies should formally reject the structural presumption.

The first is that it is an an explicitly economic test that is no longer justified by modern economics.  Modern economic learning and empirical evidence does not support the notion that mergers that generate post-merger firm with greater than 30 percent share substantially lessen competition.  It is a convenient litigation tool to shift the burden to defendants when courts are not persuaded by a competitive effects story.  But that is no excuse.  If the Agencies believe the best available evidence does not support the economic foundation of the structural presumption, they should abandon it.

Note that I’m not saying that the Agencies could not discuss alternative presumptions based on sound economics.  Perhaps economic theory and empirical evidence provide the basis of alternative bright line standards which can appropriately be made the basis of a presumption that a particular merger will substantially lessen competition and harm consumers.  But the presumption as it currently exists no longer comports with modern economics or Agency thinking about merger analysis and should be abandoned.

The second reason to explicitly abandon the presumption is that it is far too sensitive to the market definition exercise.   I think its important to note the tension between the movement toward and sometimes celebration of unilateral and coordinated effects analysis at the Agencies and away from market definition (even if not completely) and use of the structural presumption which depends so heavily on the identification of the relevant market.  The market definition exercise may well have its virtues in constraining agencies from bringing cases with overly narrow markets.  However, when the critical lesson of the modern economic approaches to mergers is that post-merger change in pricing incentives and competitive effects are what matters, it is difficult to justify the structural approach either with reference to the Guidelines articulate purpose or modern economics.

When it comes to the structural presumption, the Agencies should not get to have it both ways.  If we’re going to take the modern economics of mergers seriously, and reduce the importance of market definition in favor of competitive effects, we should also do away with the structural presumption.  The Supreme Court is not likely to take a merger case any time soon and do away with the presumption themselves.  It is up to the Agencies to eliminate the presumption as it currently exists in order to restore consistency with economic learning and empirical evidence.

The merger guidelines should be revised, not only to provide clearer guidance, but because the current version makes it harder for the agencies to win cases when they do challenge a merger.  The reason is that the guidelines often don’t fit actual agency practice or modern economic theory.  For example, part of the reason the DOJ lost the Oracle/Peoplesoft merger case was that its guidelines on unilateral effects do not match modern practice or theory.  Such a mismatch makes it harder to convince judges that merger enforcement satisfies traditional rule of law concerns about providing advance notice of legal standards that are clear enough to guide private conduct and meaningfully constrain legal discretion.  Here are the areas on which I would focus.

Unilateral Effects. The unilateral effects portion of the merger guidelines should be rewritten because the economics on unilateral effects have no connection to the HHI or 35% market share thresholds used in the guidelines.  For example, given that unilateral price effects were established in Staples, it really shouldn’t have mattered whether the court concluded Staples and Office Depot operated in an office superstore market or were simply close to each other in a differentiated office supply market.  Their market share was only 5.5% under the latter market definition, but market labels don’t alter the price effects, and why should a merger that significantly increases consumer prices be tolerated just because we can define a large market?

Market Definition. The guidelines thus should not suggest that market definition is required in unilateral effects cases, but should instead provide an alternative methodology for proving such effects.  The guidelines could rely directly on diversion ratios, as suggested by Professors Farrell and Shapiro in their prior academic roles.  Or, to use an alternative that would look more familiar to courts used to market definitions, the agencies could posit a product space around the merging parties and directly measure the relevant demand elasticities from that space to other spaces, and the supply elasticities within and into that space, and predict price effects directly.

Indeed, this approach seems better than market definition even in oligopoly effect cases because market definitions rest on all or nothing assessments that distort the economic analysis.  For example, market definition equates substitutes that could constrain prices increases of 6% with having no substitute at all, and equates substitutes that could constrain price increases of 4% with having a complete substitute, when neither is really accurate.  In practice, courts and agencies adjust the weight given to HHI or market share calculations based on judgments about the extent to which buyers could switch to outside the market and the extent to which rivals could enter or expand.  But those judgments are simply subjective assessments about demand and supply elasticities that would be better to make openly, and once we make them we don’t really need market definition to predict price effects.

Oligopoly Effects. The guidelines now provide a lot of guidance on when a market is likely to engage in oligopolistic coordination, but little guidance on when a merger is likely to worsen coordination, other than for mergers that involve the acquisition of a maverick.  Thus, the guidelines lend themselves to Catch-22 effect.  If the guidelines indicate coordination is unlikely, then courts conclude the merger cannot produce oligopoly effects.  But if the guidelines indicate coordination is likely, then courts sensibly ask why coordination isn’t equally likely before the merger, in which case the merger won’t worsen anything.  The guidelines need to be clearer about defining the incremental increase in coordination caused by mergers.

Partial ownership. Although the agencies sometimes challenge acquisitions of partial stock ownership, we don’t have any clear guidelines about when partial ownership will count as sufficiently active to be assessed as a merger.  Further, some enforcement actions have made noises about challenging partial stock acquisitions on the theory that they might lessen firm incentives to compete with each other.  This later theory could apply even when the investment is entirely passive, but so far has not been, and we don’t have any guidelines clearly indicating whether this theory would ever be applied in a case where the investment is entirely passive or what the criteria would be for distinguishing how much passive ownership is too much.  Without such advance guidelines, it is hard to imagine winning an enforcement case based on such a theory.