What Do We Do with Presumptions in Antitrust?

Cite this Article
Daniel J. Gilman, What Do We Do with Presumptions in Antitrust?, Truth on the Market (January 25, 2024), https://truthonthemarket.com/2024/01/25/what-do-we-do-with-presumptions-in-antitrust/

Winter was coming, as it does. We knew the agencies were going to issue new merger guidelines, and then they did. On Dec. 18, 2023, the Federal Trade Commission (FTC) and U.S. Justice Department (DOJ) jointly issued merger guidelines, supplanting 2023’s draft guidelines, the 2010 Horizontal Merger Guidelines, and the 2020 (partially withdrawn) Vertical Merger Guidelines

That’s big news in antitrust, even though the guidelines do not have the force of law. There’s more on the merger guidelines below. But what else is new? 

For one thing, the agencies have won a few. The FTC secured a preliminary injunction against the proposed IQVIA Holdings/Propel Media merger in the U.S. District Court for the Southern District of New York. It also succeeded in blocking the Illumina/Grail merger by partly winning in the 5th U.S. Circuit Court of Appeals, which rejected some of Illumina’s constitutional claims, left others for the U.S. Supreme Court, and agreed that the FTC had made out a prima facie case under the burden-shifting framework commonly attributed to Baker Hughes.

The decision was not, however, an FTC sweep. The court held that the FTC had applied the wrong legal standard in evaluating Illumina’s “open offer” (see me here; Alden Abbott here and here; Jonathan Barnett here; and the International Center for Law & Economics’ (ICLE) amicus brief here). The open offer was a contractual tool (already in force with some parties) designed to eliminate the risk of harm (however great or slight) that the FTC alleged.

Contra the FTC, the court ruled that the open offer should have been considered at the liability stage, not the remedy stage, and that the FTC had therefore analyzed the open offer under a more stringent standard than it should have. Thus, the case was remanded to the FTC. Illumina abandoned—no doubt for various reasons, including an estimation that continuing the dispute would be costly. This was perhaps not least due to an inkling that the FTC would view its own case favorably on remand. 

The FTC also touted its settlement of the Amgen/Horizon matter (here and here) although, as I explained in December, the core of the consent order—reached on the eve of trial—had been proposed by Amgen all along. The settlement of a case that shouldn’t have been brought seems roughly the right outcome (see ICLE’s amicus brief here). Tanking the Illumina/Grail merger seemed the wrong one, although it’s worth noting that the FTC’s case was based on an established theory of harm and did not depend on anything terribly novel from, e.g., the new merger guidelines or the FTC’s expansive (if not downright fanciful) Section 5 statement

Just last week, on Jan. 16, Judge William G. Young  of the U.S. District Court for the District of Massachusetts ruled in favor of the DOJ in enjoining JetBlue’s acquisition of Spirit Airways. The decision is, in many respects, grounded in established law, applying the Baker Hughes burden-shifting framework, and considering price effects (and consumer welfare); likelihood of entry; and, indeed, likely merger efficiencies.

Yet, in one very fundamental way, it’s unfortunate. Judge Young recognized that, as a matter of fact, the merger would be procompetitive, on net, on a national level. But because he agreed with the government that the merger was likely to harm competition “in at least some relevant markets”–that is, on some of the specific city-to-city routes among the hundreds identified in the government’s complaint–he ruled it a violation of Section 7 of the Clayton Act. That decision was not baseless, in law or fact, but it was hardly necessary, and it’s unfortunate for both competition and consumers. On that sort of balancing, see me and my ICLE colleagues Brian Albrecht and Geoff Manne on out-of-market effects here

Much has been made of the antitrust agencies’ win/loss record under the Biden administration (even as we remind ourselves that the FTC is—or is supposed to be—an independent agency headed by a bipartisan commission). Initially, it seemed dismal. It’s improved. Jan Rybnicek—a former FTC attorney advisor—has a useful thread about it here. As he observes, the win/loss ratio has improved, but it remains unremarkable, both in terms of cases brought and in terms of cases won. He also notes that the wins have tended to rest on established theories of harm.

And three years are but three years; that is, not that long and not enough cases to make much of a trend, one way or the other. The future of the agencies’ more creative endeavors remains to be seen. Still, losses and wins both count. 

Back to the agencies’ December 2023 gift to the world of antitrust (if not to consumers and competition): many think it’s an improvement over the July draft. In some ways it is, although views on the magnitude of improvement vary, ranging from slightly less bad to notably better but still problematic. 

  • Alden Abbott—former FTC general counsel—had a recent ToTM piece called “The Porcine 2023 Merger Guidelines (The Pig Still Oinks).” I recommend it (and put it in the well-populated camp that rates the guidelines still-terrible-but-slightly-less-bad).
  • Herb Hovenkamp has a more mixed view, suggesting both “a vast improvement” and significant remaining problems. And as Hovenkamp notes, “even in final form, the new Merger Guidelines are a revisionist document.”
  • Again, there’s ICLE’s recent post on out-of-market effects in merger enforcement, which also touches on the new guidelines. 

For a refresher on issues to do with the July draft guidelines, see . . . lots. ICLE’s comments on the draft are here, and my posts here and here identify (with links) useful commentary from many others. 

The agencies have given me much to kvetch about (thanks?). For today, I’ll leave myself stuck on first base, with Guideline 1. 

Guideline 1: The Largely Debunked SCP Paradigm Strikes Back

Guideline 1 alters the treatment of structural presumptions in merger analysis in several ways. Most conspicuously, it changes the thresholds for structural presumptions. The 2010 Horizontal Merger Guidelines identified three categories of market concentration: “unconcentrated markets” (those with a Herfindahl–Hirschman index (HHI) below 1,500), “moderately concentrated markets” (those with an HHI between 1,500 and 2,500 (inclusive)), and “highly concentrated markets” (those with an HHI above 2,500). 

Under the new guidelines, markets with an HHI above 1,800 are deemed “highly concentrated.” The new guidelines do not define “unconcentrated” or “moderately concentrated” markets. 

The new guidelines also announce that “mergers raise a presumption of illegality when they significantly increase concentration in a highly concentrated market.” Under the guidelines, there are two ways to raise that presumption. First, there is a structural presumption of illegality if the post-merger HHI is (a) greater than 1,800 and (b) the change in HHI is greater than 100. Second, the presumption is triggered when the merged firm’s market share is (c) greater than 30% and (d) the change in HHI is greater than 100. 

The 2023 guidelines, like the 2010 Horizontal Merger Guidelines (and the 1992 guidelines) use the HHI as a measure of concentration. For any given product (or service) and geographic market, the HHI is simply the sum of the squares of each market participant’s market share. So, for a true one-participant (100% market share) monopoly, the HHI is 10,000. If there are two participants, each with a 50% market share, the HHI is 5,000.

So, the 2023 Merger Guidelines, like the 2010, 1992, and even 1982 guidelines, employ structural presumptions and use HHI as a measure of concentration. But much has changed. 

One obvious change is in the simple number of structural thresholds: one, under the 2023 guidelines, versus three, under the 2010 guidelines.  Another is a change in the variety of presumptions. Under the 2010 guidelines, either of two presumptions might be applied to mergers in or to highly concentrated markets, depending on the change in HHI:

Mergers resulting in highly concentrated markets that involve an increase in the HHI of between 100 and 200 points potentially raise significant competitive concerns and often warrant scrutiny. Mergers resulting in highly concentrated markets that involve an increase in the HHI of more than 200 points will be presumed to be likely to enhance market power. The presumption may be rebutted by persuasive evidence showing that the merger is unlikely to enhance market power. (emphasis added)

Moreover, the 2010 guidelines told us that small changes in concentration (HHI deltas smaller than 100 points) were “unlikely to have adverse competitive effects and ordinarily require no further analysis.” Similarly, “[m]ergers resulting in unconcentrated markets are unlikely to have adverse competitive effects and ordinarily require no further analysis.” Mergers that resulted in “moderately concentrated markets” (HHI greater than 1,500 but not greater than 2,500) and an increase in HHI of more than 100 points “potentially raise significant competitive concerns and often warrant scrutiny.” 

As described above, there’s only one defined structural presumption: a presumption of illegality. That is, while the 2023 guidelines expressly identify a single structural presumption, the 2010 guidelines identified four different presumptions, depending on the post-merger HHI and the change in HHI. 

Plainly, there’s been a substantial drop in the threshold for a “highly concentrated” market (from 2,500 to 1,800), while the change in concentration triggering the strongest presumption in the 2010 guidelines (and the only presumption in the 2023 guidelines) has been cut in half (from 200 to 100). 

This is not about mergers to monopoly or two-to-one mergers. As we illustrate in our out-of-market piece, under the new guidelines, a market is deemed “highly concentrated” if, e.g., seven competitors have market shares of 30%, 20%, 15%, 15%, 9%, 8%, and 3% (HHI = 1,904). If the largest firm were to acquire the firm with 9% market share, the HHI would jump 540 points, to 2,444. What was deemed a merger from and to a moderately concentrated market under the 2010 guidelines would trigger both structural presumptions of illegality under the new guidelines.

Then there’s the question of the nature of the presumption. While the 2010, 1992, and 1982 guidelines did not quite specify safe harbors (they disavowed “rigid screens” at either end of the spectrum), they came close to it in identifying classes of mergers that are “unlikely to have adverse competitive effects and ordinarily require no further analysis.” 

What about the strongest structural presumption? In 2010, the most worrisome mergers were “presumed likely to enhance market power.” In 2023, we are told that mergers to highly concentrated markets (a lower threshold) with HHI changes of at least 100 points (a smaller change) are “presumed to substantially lessen competition or tend to create a monopoly.” The phrase “substantially lessen competition or tend to create a monopoly” is lifted straight from Section 7 of the Clayton Act. And the heading to Guideline 1 tells us that the presumption is a “[p]resumption of illegality.” 

On the one hand, antitrust has long been concerned with the acquisition and exploitation of market power; that’s recognized in the case law, and in the 2010 guidelines. Mergers in or to highly concentrated markets that involved HHI changes greater than 200 raised red flags and were bound to be subject to careful scrutiny at either agency. So, apart from the substantially revised thresholds, the difference in language may seem a small difference—perhaps much smaller than the difference in the numbers. 

Or not. There’s that phrase “presumption of illegality,” which does not appear in the 2010 Horizontal Merger Guidelines (or the 1992 guidelines, or the 2020 Vertical Merger Guidelines, for that matter). If we dial the clock back 40-plus years, we come close to it, as the DOJ’s 1982 guidelines identified mergers the agency was “likely to challenge.” But that was the DOJ in 1982, with 42 years and lots of water (agency experience, economic research, and case law) under the bridge. 

And what about the more basic question: what to make of structural presumptions as a general matter? As many have noted, economic learning and agency experience have tended to diminish the role of structural presumptions over the course of several decades (at least). My ICLE colleagues and I spent a good many pages (and citations) on this in our response to the draft merger guidelines. The structure/conduct/performance paradigm has been largely abandoned, because it’s widely recognized that market structure is not outcomedeterminative. The view is shared, as we note, by scholars across the political spectrum.

To take one prominent recent example, professors Fiona Scott Morton (deputy assistant attorney general for economics in the DOJ Antitrust Division under President Barack Obama, and now teaching at Yale University), Martin Gaynor (former director of the FTC Bureau of Economics under Obama, now serving as special advisor to Assistant U.S. Attorney General Jonathan Kanter, on leave from Carnegie Mellon University), and Steven Berry (an industrial organization economist at Yale) surveyed the industrialorganization literature and found that presumptions (and analyses) based on measures of concentration are unlikely to provide sound policy guidance:

In short, there is no well-defined “causal effect of concentration on price,” but rather a set of hypotheses that can explain observed correlations of the joint outcomes of price, measured markups, market share, and concentration.…Our own view, based on the well-established mainstream wisdom in the field of industrial organization for several decades, is that regressions of market outcomes on measures of industry structure like the Herfindahl-Hirschman Index should be given little weight in policy debates.

We (and they) were hardly alone. The Global Antitrust Institute (GAI) filed three distinct comments on the draft merger guidelines, with one focused specifically on “the 2023 Draft Merger Guidelines emphasis on structural antitrust.” As they observe, the structuralist approach of the 2023 draft (which was maintained in the final 2023 guidelines) harkened back to that of the 1968 Merger Guidelines, which “was also supported by ‘the Structure-Conduct-Performance (SCP) paradigm dominant in industrial organization economics at the time.’” Since 1968, however:

[A]dvances in economic knowledge exposed the flaws associated with the SCP paradigm, undermining the empirical basis for structural antitrust contained in the 1968 Guidelines. This research demonstrated that the SCP paradigm’s empirical methodology could not identify the causal effect of industrial concentration on market performance. The correlations between market concentration and economic performance produced by SCP studies could not be used to predict the causal effect of a merger nor relied upon by Agencies or courts for the purposes of rational antitrust merger review. Moreover, economic analysis also exposed the lack of robustness of the paradigm’s main empirical underpinnings. (internal citations omitted) 

And there are many more such examples, including Gregory Werden (former chief counsel for economics at DOJ) here, and Nathan Miller et al. (many co-authors, including Aviv Nevo, director of the FTC’s Bureau of Economics, eight former directors of the FTC’s Bureau of Economics, several former chief economists at the DOJ Antitrust Division, and leading academics) in 2022 here

Notably, comments submitted to the FTC just last year by John Asker, Kostis Hatzitaskos, Bob Majure, Ana McDowall, Nathan Miller, and, again, the FTC’s Aviv Nevo had the following to say in 2022 comments to the agencies:

As noted by DOJ and FTC staff and front office economists in late 2020, the 2010 HMG continue to accurately reflect the practices of the agencies and highlight practices and techniques of continued relevance to modern practice. We agree. Despite concerns voiced by commentators and raised in some academic studies, our read of the evidence is that it does not support large deviations from the approach in the 2010 HMG or the 2020 VMG. Some proponents favor strengthening structural presumptions and lowering the presumption thresholds. This, they argue, would be an important step toward strengthening enforcement and reducing the number of harmful mergers passing unchallenged. Our view is that this would not be the most productive route for the agencies to pursue to successfully prevent harmful mergers, and could backfire by putting even further emphasis on market definition and structural presumptions. If the agencies were to substantially change the presumption thresholds, they would also need to persuade courts that the new thresholds were at the right level. Is the evidence there to do so? The existing body of research on this question is, today, thin and mostly based on individual case studies in a handful of industries. Our reading of the literature is that it is not clear and persuasive enough, at this point in time, to support a substantially different threshold that will be applied across the board to all industries and market conditions. 

I don’t present that as “gotcha” on Nevo: agency guidelines are just that, and even a director of the Bureau of Economics cannot exert complete control over the particulars. Rather, like some of the comments and articles cited above, it’s a serious appraisal from established economists with a demonstrated willingness to support vigorous enforcement of the antitrust laws, and to consider pro-enforcement reform. And it’s very much at odds with the new guidelines. 

That’s not to say that nobody thinks there’s any signaling value to HHI measures. Some recent research suggests there might be, although it points more toward the significance of changes in HHI than the absolute post-merger measure, and it suggests different results depending on the nature of the transaction and its associated (likely) efficiencies.

But what sort of signaling value? This is not simply a question about whether or not to return to the 1992 HHI thresholds. It’s a question of the presumption that’s triggered at any given threshold. The new guidelines, unlike the 2010 guidelines, or even the 1992 guidelines, describe a “presumption of illegality,” and they do so:

  1. without any limiting principles;
  2. without anything akin to a safe harbor;
  3. with a very cramped notion of cognizable merger efficiencies; and
  4. most generally, with no clear recognition that the very large majority of mergers are either procompetitive or benign and, not incidentally, lawful. 

Collectively, that seems a very big change, indeed.

How Do the Agencies Justify Their Presumptions?

So what do the agencies have to say to justify the new presumptions of illegality being attached to very old thresholds? In a footnote, they acknowledge the 2010 thresholds and say, “based on experience and evidence developed since, the Agencies consider the original HHI thresholds to better reflect the law and the risks of competitive harm suggested by market structure and have therefore returned to those thresholds.” 

What experience and evidence? There are no supporting citations to the literature, or descriptions of agency analysis of particular merger matters from the intervening years. None. They do cite three cases decided prior to 2010: Chicago Bridge & Iron Co. N.V. v. FTC; FTC v. H.J. Heinz Co.; and FTC v. Univ. Health, Inc. While they are correct in noting that those decisions cite prior editions of the guidelines as persuasive authority, not one of the cases depends on anything remotely like the new thresholds. 

  • In Chicago Bridge, the court noted that post-merger HHIs for the relevant markets were 5,845, 8,380, and 10,000 (while also noting that “[t]he HHIs are just one element in the Government’s strong prima facie case.”).
  • In FTC v. HJ Heinz, there was a 510-point change in HHI, and a post-merger HHI of 5,285.
  • In FTC v. Univ. Health, the court noted that “the proposed merger would increase the HHI by over 630 to approximately 3200”; that the district court had “improperly assumed” that nonprofit hospitals would not act anti-competitively; that state certificate-of-need law would impede competitive entry; and that the defendants had offered a poorly substantiated failing-firm defense, with no evidence of merger efficiencies. 

Three cases, no post-merger HHIs below 3,200, considerably larger changes in HHI than any specified in the 2010 guidelines or the 2023 guidelines, and more going on than just the HHI measures.

That all seems like scant evidence to double down on structural presumptions, borrowing “highly concentrated” numbers from 1992 and a presumption of illegality at least as strong as that signaled by the DOJ in 1982. I use “scant” here as a term of art: really, no good evidence at all. And frankly, given the considerable controversy raised by the draft guidelines, that’s just plain lazy. 

Once upon a time, there was a TV sitcom called The Odd Couple, based on Neil Simon’s play of the same name. In one episode, there’s a courtroom scene that’s become a television classic (yes, there are such things—sorry, Professor Adkins). One of the characters, Felix, representing himself pro hac vice, gets the witness to admit that she “just assumed” something material. With great (self-sense of) drama, and the aid of a blackboard, Felix tells her that “you should never assume, because [as he writes the word “ASSUME,” and underlines its constituent parts in sequence] when you assume, you make an ASS of U and Me.” 

He said “assume,” not “presume,” and yet . . .