Carl Shapiro provides his round-two comments on the draft Merger Guidelines.

To read more from the ProMarket Merger Guidelines Symposium, please see here.

The symposium Round One authors most supportive of the draft Merger Guidelines appear to agree that the draft announces a dramatic shift in merger policy. Unlike me, however, they see that shift as a long-overdue return to what Congress intended in 1914, when the operative language in Section 7 of the Clayton Act first became law, or in 1950, when Congress expanded the statute to include asset purchases and non-horizontal mergers and acquisitions.

If the Agencies believe that Congress has instructed them to challenge some class of mergers that does not harm counterparties based on enhanced market power, they should state that explicitly. I am skeptical of that claim, based on the statutory language and decades of antitrust case law.

For example, suppose a firm acquires an asset (a physical facility, some intellectual property, or an ongoing business operation) that allows it to enter into a new market more rapidly and more efficiently than would otherwise be possible. Suppose the acquisition will benefit customers in that new market and there is no prospect that it will tend to create a monopoly. Both common sense and economics tell us that this acquisition is a normal, healthy part of the competitive process that injects more competition into that new market. Yet the draft regards many such acquisitions as illegal extensions of a “dominant position” that “lessen competition.” That conclusion makes no sense. For centuries, “competition” has been understood to embrace just this type of dynamism, which has been central to past economic growth and is the fuel for future prosperity.

My focus in this piece is on how changes proposed in the draft guidelines would actually work in practice under current law. My assessment here builds on my Round One submission, which expressed concern about the Agencies abandoning the central goal of protecting customers from mergers that harm them due to enhanced market power. 

Horizontal Mergers

I applaud the Agencies for moving in the direction of stricter enforcement for horizontal mergers. My comments here are directed at the question of whether the draft is likely to be effective and durable in stopping truly harmful mergers while allowing beneficial ones.

Being in favor of stricter enforcement, I am concerned by the absence of several important concepts that have strengthened the Agencies’ hand in court. Here are a few examples.

• Where is the language warning courts not to accept overly broad markets in which the merging firms’ market shares are small, if there is direct evidence that the merger is likely to harm customers? The 2010 Guidelines, which I helped draft as Deputy Assistant Attorney General for Economics at the Antitrust Division, states: “Where analysis suggests alternative and reasonably plausible candidate markets, and where the resulting market shares lead to very different inferences regarding competitive effects, it is particularly valuable to examine more direct forms of evidence concerning those effects.”

• Why does the draft downplay the role of the hypothetical monopoly test (HMT) as a method for defining relevant markets by identifying “reasonable” substitutes? The HMT is well-established in the case law and is often used in court by the Agencies to establish narrower markets than the merging firms propose based on the “practical indicia” from the Brown Shoe case. Recognizing this, the 2010 Guidelines added language to explain how and why the HMT often leads, correctly, to narrow markets, including markets for targeted customers. The past 13 years has proven the effectiveness of that strategy.

• Where is the language explaining that market shares calculated in broad markets can falsely suggest that a merger will not harm customers? The 2010 Guidelines states: “Defining a market broadly to include relatively distant product or geographic substitutes can lead to misleading market shares. This is because the competitive significance of distant substitutes is unlikely to be commensurate with their shares in a broad market.” This critical idea is then illustrated using an example involving a motorcycle merger. The 2010 Guidelines further explains: “The measurement of market shares and market concentration is not an end in itself, but is useful to the extent it illuminates the merger’s likely competitive effects.” These statements have strengthened the Agencies’ hand in court over the past 13 years.

• Why are there no examples illustrating how the Guidelines operate in various settings? The 2010 Guidelines provides 24 concrete examples designed to show the manner in which mergers may be harmful and how to identify those effects. These examples have proven valuable and instructive. Related, does the 2006 Commentary on the Horizontal Merger Guidelines remain a valuable supplement to these guidelines, or not?

The important ideas highlighted above helped the Agencies prevail in challenges to mergers relating to tax preparation software, food distribution, health insurance, office supplies, hospitals, and book publishing, among others. Their absence from the draft is worrisome and perplexing.

I see no tension between retaining these important pro-enforcement ideas and updating the Guidelines to reflect new learning and the accumulation of Agency experience. The most valuable new material in the draft—such as the parts that discuss acquisitions of potential competitors, multi-sided platforms, and serial acquisitions—could easily be included in updated guidelines without changing their unifying theme or removing the ideas just identified.

So far as I can tell, these important ideas were removed as part of the overall shift away from evaluating mergers based on whether they harm customers due to enhanced market power. Whatever the reason, their removal would weaken merger enforcement. Hence my concern.

The Structural Presumption: Guideline 1

Herb Hovenkamp and I have called for strengthening the structural presumption that has applied to horizontal mergers for 60 years under the Supreme Court’s Philadelphia National Bank decision. The structural presumption is critical in practice because the Agencies typically prevail in merger litigation by using it to establish their prima facie case. I welcome efforts by the Agencies to utilize the structural presumption more effectively. Nancy Rose and I have provided specific suggestions on how to do this. Downplaying the HMT is not one of them.

Guideline 1 adds a structural presumption if the market share of the merged firm exceeds 30% and the increase in the HHI is at least 100, citing Philadelphia National Bank. The 30% figure comes straight from that case, but I do not understand the basis for the 100 point increase in the HHI. The market shares of the merging firms in Philadelphia National Bank were roughly 20% and 15%, so the increase in the HHI was about 600, not 100. This is not a small discrepancy.

Guideline 1 also lowers the HHI thresholds that trigger the structural presumption back to the levels found in the 1982 and 1992 Merger Guidelines. I can support that change, but only if the Agencies are going to consistently enforce near these levels, which are far lower than the HHI levels found in litigated cases over the past 20 years. If not, then lowering the thresholds will merely provide misleading guidance and undermine the credibility of the Merger Guidelines. After all, the HHI thresholds were raised in 2010 based on the good-government principle that official guidelines should accurately reflect Agency practice, as Assistant Attorney General Christine Varney explained in 2010.

So far as I can tell, all of the horizontal merger cases litigated by the Biden Administration have involved HHI levels well above the 2010 thresholds. The HHI increases all exceed 800.

Merging PartiesRelevant MarketPost-Merger HHIIncrease in HHI
United States Sugar/Imperial SugarProduction and sale of refined sugar in the Southeastover 2800over 800
United States Sugar/Imperial SugarProduction and sale of refined sugar in Georgia and its neighboring statesover 3100over 1100
Pengiun Random House/Simon & SchusterU.S. publishing rights to anticipated top-selling books3111891
UnitedHealth Group/Change HealthcareSale of first-pass claims editing solutions in the United Statesat least 5625at least 2500
Booz Allen/EverWatchSale of signals intelligence modeling and simulation services to NSA through the OPTIMAL DECISION contract10,000merger to monopoly
ASSA ABLOY/Spectrum BrandsPremium mechanical door hardware in the United Statesat least 4000at least 1600
ASSA ABLOY/Spectrum BrandsSmart locks in the United Statesat least 3000at least 1200

Given this track record, and limited resources, I wonder whether the Biden Department of Justice  and Federal Trade Commission have been constrained by the HHI thresholds in the 2010 Guidelines, and if so in what way.

Of course, HHI thresholds only apply after a relevant market has been defined. In 2010, my colleagues at the Antitrust Division and I recognized that the most important step to establishing the structural presumption was to put forward a relevant market that would hold up in court. See above. We did not relish trying to convince a federal judge that a merger involving low market shares warranted a strong structural presumption, especially given the sliding scale established in the widely cited Baker Hughes case: “The more compelling the prima facie case, the more evidence the defendant must present to rebut it successfully.” The path to victory for the Agencies typically involves establishing narrow markets, not relying on low market concentration thresholds.

Loss of Head-to-Head Competition: Guideline 2

I like the idea behind Guideline 2. Economists have long recognized that the loss of competition between the merging firms alone can cause substantial harm to customers. The concept of unilateral effects was introduced into the Guidelines in 1992 and further developed in 2010.

Economists have also recognized for many years that unilateral effects are best diagnosed based on diversion ratios and margins and related methods, not based on HHI levels. The 2010 Guidelines introduced Upward Pricing Pressure into the Guidelines to reflect advances in economic learning. The strategy in 2010 was pragmatic: to put unilateral effects on a more solid footing in the case law. That goal has been achieved.

The Agencies should explain how Guideline 2 will operate independently of Guideline 1. Suppose the court settles on a relevant market in which the merging firms have market shares of 8% and 4% (say), and the merging firms argue that their merger cannot violate Section 7 given these low market shares, citing Philadelphia National Bank and the thresholds found in Guideline 1. What will the Agency say in response? My guess is that the Agencies will assert that the merger is nonetheless illegal based on direct evidence that it will lead to higher prices. Is that the idea?

Trend Toward Concentration: Guideline 8

Guideline 8 lacks an economic foundation and would make it harder for industries to make beneficial adjustments to changing conditions, especially given that the Agencies will flatly ignore proven merger synergies when applying Guideline 8. In my Round One submission I stated: “Guideline 8 considers only structural factors, not economic effects. The fundamental problem is that Guideline 8 does not inquire into why there is a trend toward concentration or whether that trend has helped or hurt customers.”

Guideline 8 should be dropped entirely. Industry trends would remain relevant, not for their own sake but rather to the extent that they illuminate a merger’s likely effects.

If the Agencies retain Guideline 8, they should explain how it reflects “the collected experience of the Agencies over many years of merger review in a changing economy.” I think not.

Entrenching or Extending a “Dominant Position”

Guideline 7 is fundamentally flawed and dangerous, especially because it is so broad, applying  even to “mergers that are neither strictly horizontal nor vertical, so the Agencies may seek to identify any connection suggesting the merger may entrench or extend the dominant position.”

Entrenching a “Dominant Position”

Under Guideline 7, the Agencies could challenge many non-horizontal acquisitions that make the acquiring firm a stronger, more effective competitor in a market where it already has a market share of at least 30%. Under Guideline 7, if a firm with a market share of at least 30% purchases an asset that allows it to grow its share by offering customers lower prices or better products, that would “reduce the competitive structure of the industry” and “entrench” its “dominant position.”

Critically, the Agencies indicate that they will not evaluate such acquisitions based on their effects on customers. Rather, the stated goal is to “preserve the possibility of eventual deconcentration.” Acquisitions that increase entry barriers or deprive rivals of scale economies or network effects are explicitly identified as involving entrenchment, without any inquiry into whether these acquisitions benefit or harm customers. This approach would protect rivals from stronger competition enabled by acquisitions. Count me out. 

I have to wonder what other countries will think if this becomes the official policy of the United States, after we spent decades teaching the rest of the world that efficiencies are not a dirty word.

Extending a “Dominant Position”

Overwhelming evidence teaches us that competition is normally enhanced when a capable firm expands into a new market by acquiring a participant in that market. The canonical business model is for the acquiring firm to leverage its core strengths to gain a competitive advantage in the new market and grow its market share there. Acquisitions often accelerate this process.

Yet Guideline 7 announces that the Agencies will challenge such acquisitions if the acquiring firm has at least a 30% market share in an existing market and if the acquisition “may tend to extend” that position into the new market, even if there is no prospect that the acquisition will tend to create a monopoly. Critically, the Agencies will not base their evaluation of the acquisition on whether it benefits or harms customers in the new market. They simply ask whether the acquiring firm may gain a 30% share in the new market. Guideline 7 does not explain what evidence (if any) would convince the Agencies not to challenge such an acquisition. Evidence that the acquiring firm has capabilities that will enable it to grow its share in the new market would seem to count against the acquisition. That is stunning to me.

If Guideline 7 were put into practice, a great many beneficial acquisitions that enable successful firms to enter and grow in adjacent markets, injecting more competition into those markets, would be at risk of challenge. If a large number of such acquisitions were deterred, the damage to the American economy would be severe.

Amazon provides a salient example of what is at stake. The report released by the House Judiciary Committee in October 2020 contains a section on Amazon’s acquisitions (pp. 219-224). A major concern expressed there is that these acquisitions have “effectively protected and expanded Amazon’s market power in e-commerce and helped Amazon extend that power to other markets.” The largest recent acquisitions viewed with suspicion are Whole Foods (2017), Ring (2018), and PillPack (2018). I have not studied these acquisitions in detail, but it strikes me as pro-competitive for Amazon to enter new markets (here, brick-and-mortar supermarkets, home security, and online pharmacies) by leveraging its strength in distribution, logistics and procurement, its trove of data about consumer purchasing patterns, and its extensive customer relationships. Yet these appear to be the types of acquisitions that Guideline 7 intends to challenge, without regard to whether they benefit or harm customers and without regard to whether they promote or suppress “competition” as that term is conventionally understood.

Challenging acquisitions that enable successful firms to inject more competition into related markets to the benefit of customers would be a grave error: doing that would protect incumbent firms in those markets from competition, harming consumers in the process. The best I can say about such challenges is that they would almost certainly fail in court.

If the Biden Administration believes these types of acquisitions need to be prevented, it should go to Congress to seek new legislation. In a democracy, that would be the proper way to make such a dramatic policy change, not by inserting that policy change into agency guidance that purports to enforce existing law.

More generally, if the Agencies persist in announcing radical changes in their new Guidelines, the broader danger is that the courts will not accept those Guidelines as persuasive authority. Moreover, by shifting the emphasis from economic principles consistent with the case law to a recitation of mostly outdated case citations, the draft cedes the high ground, the “principal analytical techniques” explained in prior merger guidelines, where the Agencies’ experience and deep expertise is likely to command the most respect from the courts.

Author Disclosure: I am currently consulting for parties that have a financial interest in the Merger Guidelines. Please see my full disclosure statement here

Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.