John Kwoka writes that the antitrust agencies under President Joe Biden made thoughtful revisions to the Merger Guidelines that will strengthen enforcement and competition. However, they still fall short in their treatment of the structural presumption and efficiencies defense, where both economics and the law provide grounds for strengthening. Current practices strain agency resources and permit anticompetitive mergers and acquisitions. The next administration must revisit these two issues.
Editor’s note: This article is part of a symposium that asks how the next presidential administration and its antitrust agencies should reorient competition policy for the next four years. Contributions from Herbert Hovenkamp, John Kwoka, Steven Salop, and Ginger Zhe Jin and Liad Wagman can be read here.
One of the biggest accomplishments of the Biden administration’s antitrust policy was the promulgation of an updated Merger Guidelines to more rigorously scrutinize mergers and acquisitions after decades of underenforcement leading to industry concentration and decreases in competition. (Full disclosure: I worked on the new Merger Guidelines while serving as chief economist to Chair Lina Khan of the Federal Trade Commission.) The new Merger Guidelines, finalized last December by the Federal Trade Commission and the Justice Department (the agencies), have been rightly celebrated for explicitly addressing several issues that previously got short or no shrift in earlier versions. These include a focus on labor markets and buyer power generally, platform mergers, vertical mergers, mergers that eliminate potential competitors, and mergers that entrench dominant firms. But there is more work to be done, notably in two areas of longstanding controversy and of crucial importance to merger enforcement: the structural presumption and the treatment of efficiencies.
Why do I call these controversial? It is in no small part because of what the Supreme Court said, or seemed to have said, about these issues. In its 1963 Philadelphia National Bank decision, the Supreme Court declared that for mergers of significant size in concentrated markets, the resulting higher concentration is sufficient to presume that the merger is anticompetitive. In its 1967 Procter & Gamble case, the Supreme Court stated that under the law supposed efficiencies from a merger cannot rescue an otherwise anticompetitive consolidation. Together, these two Supreme Court cases dictate the structural presumption and treatment of efficiencies that the agencies and lower courts must follow.
However, mainstream economics has often ridiculed the benighted judiciary that would make such statements and has convinced the lower courts to largely dismiss them. Instead, economists have developed elaborate methodologies for analyzing the very questions that the Supreme Court has said do not really matter: the actual effects on prices, output and quality of a merger in a high-concentration market, and the magnitude of efficiencies in all but extreme cases. This process has resulted in high administrative costs for the agencies to review these mergers, which has strained resources and forced the agencies to wave through anticompetitive mergers. At the same time, the process has produced little benefit in terms of clarity or accuracy of enforcement.
The new 2023 Merger Guidelines attempted to contain the damage from existing practice with respect to these issues and made important progress but ultimately fell short. Finally and fully fixing this problem should be high on the new administration’s antitrust agenda. So, how did negligence of the structural presumption and treatment of efficiencies come about? Where do we stand? And exactly what is to be done?
The Structural Presumption
Consider the role of concentration in merger analysis. The proper interpretation and role of concentration had long been matters of both research and disagreement. Free market advocates viewed high shares and concentration as reflecting meritorious superiority whereas critics pointed to their association with excess profit margins. But before that controversy fully erupted, the Supreme Court made its own view known. In Philadelphia National Bank, the Court observed that, due to its predictive nature, “whether the effect of a merger ‘ may be substantially to lessen competition’…is not the kind of question which is susceptible of a ready and precise answer in most cases.”
Despite numerous advances in economic analysis, this remains a fair statement regarding many mergers, and especially those where the concern is with post-merger coordination among the remaining firms. For such mergers, theory does not predict the precise number of firms where this threat suddenly becomes substantial. Enforcers are left hoping for incriminating documents but often have to rely on a checklist of factors affecting the likelihood of coordination. As the Court rightly observed, in such cases the effect of the merger cannot be precisely predicted, but the Court provided the practical solution. It stated:
…[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.
Put differently, a showing of high and rising concentration creates a strong presumption of illegality. Remarkably, this ruling has been essentially disregarded by lower courts. Those courts have been seduced by the false economic promise that precise effects can be demonstrated for most mergers. The courts have happily accepted that promise, and defendants, even more happily, have argued that when plaintiffs, generally the antitrust agencies, cannot meet this standard, the plaintiffs should lose. The result is that agencies must expend resources in efforts to prove what sometimes is simply unprovable, as the Philadelphia National Bank Supreme Court acknowledged, and at other times is self-evident. All this despite the Supreme Court’s explicit ruling to the contrary.
Worse yet, detailed investigations are not only costly, but they routinely fail to get it right. My meta-analysis of dozens of published merger retrospectives found that in more than 80 percent of carefully studied mergers, all subject to detailed investigation by the agencies, the mergers have resulted in price increases net of other factors, indicating a substantial underprediction of competitive harms from case-by-case analysis.
By contrast, a well-designed structural presumption has a much higher rate of success. Matching outcomes to the number of remaining competitors in the same database that I just described reveals that, while few in number, all mergers to monopoly (i.e. 2-to-1 mergers) were anticompetitive, which can hardly be a surprise. So too were all 3-to-2 mergers and a large but diminishing fraction of those with a greater number of remaining firms. The Court in Philadelphia National Bank had that right: when precise prediction is difficult, market structure does a very good job of anticipating the outcome.
Various editions of past Merger Guidelines have referenced the structural presumption, but the language in most versions has not indicated much reliance on it. The new Merger Guidelines read somewhat differently from past versions but the practical problem remains. The very first specific guideline states that “[t]he Agencies therefore presume, unless sufficiently disproved or rebutted, that a merger between competitors… that significantly increases concentration and creates or further consolidates an already concentrated market may substantially lessens competition.” It goes on as follows: “The higher the concentration metrics…the greater the risk to competition suggested by this market structure analysis and the stronger the evidence needed to rebut or disprove it.”
This sliding scale approach is entirely logical, and the Merger Guidelines were wise to lower the thresholds that trigger the structural presumption, but it should be further strengthened. To begin, for mergers to monopoly there should be no rebuttal. The merger retrospective database that I mentioned does not contain a single case of a merger to monopoly that is anything other than anticompetitive. Clever economics can demonstrate that cost savings or technological gains can theoretically outweigh the market power that is inherent in monopoly, but these are simply not realistic. In addition the original merger statute, the Clayton Act, specifically prohibits mergers that “may…substantially lessen competition or tend to create a monopoly” [emphasis added]. a prohibition that necessarily includes mergers to monopoly. For mergers to monopoly, there is no policy ambiguity.
But non-rebuttable presumptions should not necessarily only apply to mergers to monopoly. Extension at least to 3-to-2 mergers (mergers to duopoly) might be warranted by the evidence. Beyond that it might be appropriate to create a category of mergers resulting in perhaps 3-5 firms that is subject to a rebuttable presumption, such as that in the new Merger Guidelines. For mergers that result in yet more firms, no presumption would apply, but rather a full inquiry would be required. These categories would be determined by a commissioned study that compiles data on the outcomes and characteristics of a larger number of consummated mergers, and then determines the likelihood of anticompetitive outcomes by the number of remaining significant firms. Of course, the proper criterion for drawing the line is not that every single merger above the line is necessarily anticompetitive, but rather that the bright line makes no more errors than a case-by-case determination.
At present, the agencies face enormous difficulties in relying on any version of the Supreme Court’s structural presumption. These are vividly demonstrated in the Justice Department’s recent challenge to the JetBlue-Spirit merger. In its complaint, the DOJ cited the Supreme Court’s Philadelphia National Bank precedent to conclude that on numerous routes where the merger would substantially raise already high concentration, the merger was presumptively illegal without further showing. The JetBlue-Spirit court acknowledged that precedent, itself citing the Philadelphia National Bank opinion, and then proceeded simply to ignore it, although it still ruled in the DOJ’s favor. Instead the court cited and relied on lower court cases that insisted on evidence of anticompetitive effects in each case. Clearly, re-establishing the presumption will require persistence in bringing cases that cite the controlling law and the convincing evidence.
Efficiencies Defense
A second area where the new Merger Guidelines would benefit is further strengthening the so-called efficiencies defense. In this case the issue is not that the lower courts refuse to do what the Supreme Court instructed; rather, the lower courts are examining a merger issue that the Supreme Court said to limit or put aside. In its Procter & Gamble case, after reviewing the merger statute and legislative history, the Supreme Court concluded that “[p]ossible economies cannot be used as a defense to illegality. Congress was aware [when it enacted the Clayton Act] that some mergers which lessen competition may also result in economics, but it struck the balance in favor of protecting competition.”
This quote might appear to declare economics, or efficiencies, as essentially irrelevant to merger analysis, and critics of Procter & Gamble have interpreted this statement as prohibiting all efficiency claims. But the Court’s immediate endorsement of a “balance in favor of protecting competition” [emphasis added] has generally, and quite reasonably, been viewed as allowing for cost savings that are large enough to more than fully offset competitive harms from a merger. The latter outcome, where price actually decreases, would not, in the Court’s view, lessen competition.
This guidance directs attention to mergers resulting in substantial cost savings, and these are very much the exception. There is considerable economic evidence that cost savings from mergers average out to essentially zero or at most insignificantly positive. By itself, that might argue for dismissing all claims of efficiencies, but the empirical evidence also suggests that in a very small fraction of cases, mergers do in fact appear to achieve significant, sometimes substantial, efficiencies. Rejection of all efficiency claims would therefore sacrifice the small number of important gains, and so, as the Supreme Court perhaps intuited, in those cases efficiencies should be recognized.
On the other hand, the Court offered no guidance for how the lower courts or agencies should identify those few cases and thereby avoid the burden of evaluating the vast numbers of other mergers with little or no cost savings. The agencies and Guidelines have struggled with this operational challenge. Early Guidelines were said to set the concentration thresholds triggering agency challenge at a slightly higher level than otherwise would be the case, thereby allowing for modest routine efficiencies without full inquiry in all but extraordinary circumstances. This approach, it was argued, was analogous to income tax deductions where taxpayers could either take the standard deduction (akin to routine efficiencies built into the system) or itemize if they claimed large deductions.
This analogy was defective, however, since taxpayers choose which approach works to their advantage, whereas in the case of efficiencies, all merging parties get the credit and still can itemize if they wish. As a result, there is no longer any advantage to settling for the standard credit, and so of course, all merging parties claim their efficiencies to be extraordinary. Later Merger Guidelines compounded the problem by saying that only under a majority of circumstances would special consideration not apply, but that battle had already been lost.
The new Merger Guidelines do no better. They dutifully quote the above Supreme Court language, and then immediately pivot to the factual statement that “merging parties sometimes raise a rebuttal argument that, notwithstanding other evidence that competition may be lessened, evidence of procompetitive efficiencies shows that no substantial lessening of competition is in fact threatened by the merger.” There is no explanation for the apparent inconsistency of these two statements, but instead an explanation of how the agencies will evaluate such arguments. As a result, there are effectively no limits on when or which merging parties can claim efficiencies demanding agency attention.
The courts face the same conundrum. The presiding lower court in the 2020Deutsche Telecom case began by confessing its confusion, stating that “It remains unclear whether and how a court may consider evidence of a merger’s efficiencies.” After again dutifully quoting the Supreme Court’s Procter & Gamble decision, it went on to note that:
…lower courts have since considered whether possible economies might serve not as a justification for an illegal merger but as evidence that a merger would not actually be illegal. The trend among lower courts has thus been to recognize or at least assume that evidence of efficiencies may rebut the presumption that a merger’s effects will be anticompetitive…
While as already suggested, this is a reasonable interpretation of the Supreme Court’s opinion and a fair statement of current policy, it does nothing to change merging parties’ incentives to submit claims and studies of efficiencies in all cases, which is very much the current pattern. This is where current practice and guidelines need to be tightened.
The following three-stage process would go far towards that goal. The first or qualifying stage would set three substantive and procedural criteria to screen out efficiency claims insufficient to overcome likely competitive harms. The key to the first criterion is to mimic the revelation incentives of the income tax analogy to efficiency claims, that is, to provide merging firms with some kind of incentive not to claim efficiencies except in “extraordinary” circumstances, but rather to accept the standard amount. Most analogous—but hardly realistic —would be to have less stringent concentration thresholds triggering challenges for routine cases, but lower thresholds for cases where parties intend to present detailed claims. More realistically, the incentive could be a commitment by the agency somehow to expedite the merging parties’ review since no time would be wasted on efficiency claims.
The intent would be to limit necessary agency reviews of efficiencies to what are, in the eyes of the merging firms, truly extraordinary claims. The agency would review these but evaluate the overall effect of the merger against somewhat stricter guidelines thresholds purged of the standard deduction.
The second criterion would be some guidance for what constitutes “extraordinary” efficiencies, based on other aspects of the Guidelines. Perhaps the most straightforward technique would be to rely on the formula for upward pricing pressure. This could be used to deduce the magnitude of efficiencies likely necessary to prevent a price increase of some magnitude, e.g., five percent. Alternatively, if the no-efficiencies HHI concentration threshold might have been 2000 and that reflecting the standard efficiency allowance is 2500, then some simple analytics could be used to infer the magnitude of efficiencies, all else equal, implied by that difference.
This first qualifying stage also includes a procedural criterion. Detailed claims of substantial efficiencies would need to be supported by evidence from either recent past practice (e.g., past acquisitions resulting in cost savings) or from documentation prepared in the ordinary course of business at least one year prior to the merger announcement or filing. This latter provision would give no weight to the common practice in which parties commission economic studies of prospective efficiencies in preparation for their merger submissions to the agency.
Efficiency claims that satisfy these two criteria would qualify for the second stage: full agency review along the lines of current practice. For mergers that are approved because of decisive efficiencies, there would be a third stage. This would require the merging parties, for a period of perhaps two to four years, to provide the agency with data and other information sufficient to assess the degree of realization of claimed efficiencies. Where realization cannot be verified by clear and convincing evidence, the agency would be empowered to take remedial action up to and including divestiture of the acquired assets.
These two changes in policy and practice fully establishing the structural presumption, as well as limiting the efficiencies defense, would go far toward remedying two important weaknesses of the Merger Guidelines and merger enforcement. These would at the same time better reflect the law and economics, while easing the burden on the agencies. The opportunity to accomplish all these objectives simultaneously should not be missed.
Author Disclosure: The author worked on the new Merger Guidelines while serving as chief economist to Chair Lina Khan of the Federal Trade Commission. The views here are solely his own. You can read our disclosure policy here.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
