Why Do Corporations Merge and Why Should Law Care?


Mergers and acquisitions are extraordinarily prevalent in the United States, generating massive expenditures every year. However, a serious empirical puzzle lies at the heart of all that activity. That empirical phenomenon’s most remarkable feature by far is that even though it is well established in an extensive literature and implies far-reaching policy consequences, American law ignores it entirely.

Generations of researchers have failed to find evidence that merger and acquisition activity generates any lasting benefits for the combining firms’ owners or anyone else. No one seriously doubts that efficiencies of scale or technological integration are real or that acquisitions sometimes achieve them. Still, the evidence strongly implies that they are mostly available in small deals among small firms. While the results are no longer very seriously contested, no one has any conclusive explanation for the several puzzles they pose.

This paper comprehensively reviews the empirical literature and works through the policy implications for the major bodies of law it affects. The most important insights are two. First, the evidence should undermine the confidence among scholars of corporation law in the regulatory self-sufficiency of market institutions, on which corporation theory depends so heavily. In particular, a major hope for controlling managerial agency cost—one of the theory’s chief preoccupations—remains the hope that the so-called “market for corporate control” governing hostile takeovers will discipline underperforming managers. The evidence discussed here suggests that the market produces no meaningful benefits at all. The second lesson is simpler. The other major legal regime governing mergers and acquisitions—antitrust—should return to the simpler, pro-enforcement presumptions by which it once more effectively limited market concentration. The fear of jeopardizing pro-competitive gains, on which antitrust merger law has been rendered largely inert, is no longer very defensible. The balance between false positive and false negative should be reset.

Why Do Corporations Merge and Why Should Law Care?

Merger and acquisition activity is an extremely busy engine in the U.S. economy. At its heart, however, lies a serious empirical puzzle. It remains not only unsolved but largely ignored, and it reflects an incoherence throughout much of American business-regulatory policy and its theoretical foundations.

Namely, though corporate combinations are hugely common and have been for over a century, a large body of empirical evidence now suggests that they do not do any good for the combining firms’ owners or anyone else. While the problem is not unknown among academics and the result is no longer very seriously doubted, no one has any good explanation for it.

Not only do corporate managers continue in this apparent omnipresent irrationality, but they spend massive amounts of money on it.1Total annual value of deals in most recent years has met or exceeded trillion, and
the record for annual value was recently set at nearly trillion. Overall value since 2013 is well
over trillion. See Deloitte, The State of the Deal: M&A Trends 2020, at 2, 5 (2020), https://www2.deloitte.com/content/dam/Deloitte/us/Documents/mergers-acqisitions/us-mna-trends-2020-report.pdf; Stephen Grocer, A Record .5 Trillion in Mergers Were Announced in the First Half of 2018, N.Y. Times, (July 3, 2019), https://www.nytimes.com/2018/07/03/business/dealbook/mergers-record-levels.html [https://perma.cc/2EF9-TS2N]; Bourree Lam, 2015: A Merger Bonanza, The Atlantic (Jan. 9, 2016), https://www.theatlantic.com/business/archive/2016/01/2015-mergers-acquisitions/423096/ [https://perma.cc/V874-FF2P]. The annual record was set at nearly trillion in 2021. Michelle F Davis, Bankers Weigh What’s Next After Historic Trillion M&A Record, Bloomberg (Dec. 28, 2021) https://www.bloomberg.com/news/articles/2021-12-28/bankers-weigh-what-s-next-after-historic-5-trillion-m-a-record#:~:text=Companies%20announced%20a%20mammoth%20%245,right%20up%20until%20the%20holidays.
Most remarkable—and a fact that will turn out to be significant in this study—is that acquiring firms pay large “premiums” for their targets. A takeover proponent ordinarily must spend substantially more for a controlling percentage of shares in a target firm than the shares’ current trading price.2In some years the average premium in excess of the target firms’ pre-acquisition share prices has been as much as 50%, and premiums in individual deals sometimes more than double it. See, e.g., Bubble Stocks Top the Takeover Premium Charts, Seeking Alpha (Sept. 23, 2016), https://seekingalpha.com/article/4008095-bubble-stocks-top-takeover-premium-charts [https://perma.cc/5H6E-ZDAL]. Assuming that large premiums reflect rational estimates of cost-saving synergies, pricing power, or some other advantage that will come from a merger, they imply very large gains, either for shareholders or society, or both. Based on that reasoning, merger premia under current circumstances would indicate that acquisitions contribute more than $300 billion in shareholder or social gains per year, an amount that would represent a non-trivial fraction of the entire U.S. GDP.3Takeover premiums currently average well over 30%, and annual total deal value has been running at about trillion. Jens Kengelbach & Alexander Roos, Boston Consulting Group, Riding the Next Wave in M&A 10 (2011) https://www.bcg.com/publications/2011/riding-the-next-wave-in-m-and-a [https://perma.cc/EMR3-PLQV] (finding average premium of 36% from 1990 to 2010, with annual peaks as high as 45%). Therefore, aggregate merger premiums in recent years have been about 0 billion.

But therein lies the empirical puzzle. Over several decades of research, the dozens of researchers who have looked for those gains have been unable to find any meaningful evidence of them at all. Even the large premium payments enjoyed by target-firm shareholders represent merely a one-way wealth transfer from acquiring-firm shareholders. And aside from that one-time payday, the empirical evidence strongly suggests that growth by acquisition, on average, causes both acquiring and target firms to lose value or break even at best, generating no demonstrable benefit to anyone else. No one seriously doubts that efficiencies of scale or technological integration are real or that acquisitions can achieve them. Still, the evidence strongly implies that these benefits are mostly available only in small deals among small firms. The only deals relevant to important policy issues are very large in relative terms—that is, deals are typically policy-significant only when they are between firms that hold substantial market shares in concentrated markets. Surprising as these results may seem, they just confirm with empirical rigor what was already longstanding Wall Street wisdom: most deals fail, and they fail with capricious unpredictability.4See, e.g., Andrew Ross Sorkin, Is Merger Advice Worth the Price?, N.Y. Times (Aug. 6, 2007), https://www.nytimes.com/2007/08/06/business/worldbusiness/06iht-banker.4.7004868.html [https://perma.cc/EE2Y-JK9W] (referring to “[t]he old adage that half of all deals destroy value”); see also The Trouble With Mergers, Cont’d, The Economist, Nov. 28, 1998, at 15 (noting academic consensus that acquiring firms lose value, apparently because efficiencies prove elusive). As many as 70% of acquisitions are said to fail on some measure,5Clayton M. Christensen, Richard Alton, Curtis Rising & Andrew Waldeck, The New M&A Playbook, Harv. Bus. Rev., Mar. 2011, at 49. and easily a third or more fail in the ultimate sense that the acquired assets are later resold.6See, e.g., David J. Ravenscraft & F.M. Scherer, Mergers, Sell-Offs, and Economic Efficiency 1–2, 159–65 (1987) (finding divestiture to follow from as much as a third or more of mergers); Michael E. Porter, From Competitive Advantage to Corporate Strategy, Harv. Bus. Rev., May 1987, at 43, 46 (finding that among a sample of large firms as many as 50% of assets acquired in conglomerate merger wave of 1960s were later divested); see also Steven N. Kaplan & Michael S. Weisbach, The Success of Acquisitions: Evidence from Divestitures, 47 J. Fin. 107, 107 (1992) (finding that in a large sample of mergers, 44% of acquired assets were divested, but arguing that not all divestitures are evidence of failure).

These results are starkly at odds with presumptions underlying American business regulation and its theory. Leading theoretical models in law, finance, and economics presume the rationality of business firms and thus take for granted that mergers and acquisitions create value for someone.7Firm-level rationality— the idea that firms make decisions to maximize the stream of profits owned by the firm, regardless what might be the individual interests of its managers or other stakeholders—is presumed from the simplest microeconomics, see Dennis W. Carlton & Jeffrey M. Perloff, Modern Industrial Organization 35–78 (4th ed. 2015), to the most sophisticated formal models of merger behavior, see Joseph Farrell & Carl Shapiro, Horizontal Mergers: An Equilibrium Analysis, 80 Am. Econ. Rev. 107, 110 (1990) (influential merger model of substantial mathematical complexity, assuming that firms “maximize . . . profits”). Otherwise, the thinking goes, rational firms would not pursue them. It is lost on no one that executives might disserve their shareholders through selfishness or incompetence—this is the problem known as “agency cost” and it is taken as uncontroversial in corporate theory.8This problem of agency cost is often described as “[a] core problem—if not the problem” in corporation theory, Kobi Kastiel & Yaron Nili, Competing for Votes, 10 Harv. Bus. L. Rev. 287, 293 (2020), and it has been at least since the seminal Michael Jensen & William Meckling, Theory of the Firm: Managerial Behavior, Agency Costs and Capital Structure, 3 J. Fin. Econ. 305 (1976). The roots of the idea are older yet. Robin Marris captures what has been, since Adolf Berle & Gardiner Means, The Modern Corporation and Private Property (1932), a standard consensus: “Under any conditions of less than perfect competition, profit maximization is rarely compulsory, and when the decision-taker is not directly the profit-receiver, profits need not necessarily represent even one dimension of motivation.” Robin Marris, A Model of the “Managerial” Enterprise, 77 Q.J. Econ. 185, 186 (1963). However, it is widely presumed that market forces control that behavior pretty well, especially when companies adopt governance tweaks recommended in the academic literature of the day.9Within the massive literature on agency costs produced since Jensen & Meckling, supra note 8, there is a predominant view that favors reliance on market forces to align management and shareholder interests, as opposed to regulatory intervention or liability rules. Any number of specific market-oriented solutions have been suggested. The most prominent proposal happens to be one that is central to this paper: that hostile takeovers should be encouraged, so that society benefits from the so-called “market for corporate control.” If incumbent managers know their firms can be taken from them when they underperform (and their stock prices therefore fall, making hostile takeover a bargain), they will work to keep share prices high. See discussion infra Part II.C. Other proposals include many variations on shareholder empowerment or institutional investor oversight, meant to tie shareholder self-interest more meaningfully to long-term corporate performance, see, e.g., William W. Bratton & Michael L. Wachter, The Case Against Shareholder Empowerment, 158 U. Pa. L. Rev. 653, 655–659 (2010) (describing some such proposals), steps to make proxy contests more like competitive markets, Kastiel & Nili, supra note 9, at 293, and executive compensation designs that align managers’ personal interests with share performance. Kevin J. Murphy, Executive Compensation: Where We Are, and How We Got There, in Handbook Of The Economics Of Finance 211 (George Constantinides et al. eds., 2013). Interestingly, one very well-known suggestion is to make corporate governance better by using market forces to shrink the amount of government regulation itself. Roberta Romano, Empowering Investors: A Market Approach to Securities Regulation, 107 Yale L.J. 2359, 2361 (1998), recommends that firms be allowed to choose whether to be subject to federal or state securities regulation, and predicts that competition among regulatory regimes would result in an optimal amount of regulation. One imagines that it would result in much less regulation, and so a market solution would lead to more reliance on market institutions. Thus while corporate, securities, antitrust, and some other bodies of law oversee merger conduct, they each do so only nominally and, on the presumption of that conduct’s rationality, wind up leaving it effectively unregulated. Admittedly, these laws’ combined effect is a very expensive compliance obligation, and no doubt merging parties and their lawyers would be amused to hear that their deals are “unregulated.” But indeed, the whole regulatory edifice rarely stops or seriously limits their ability to merge. It merely imposes costs.

Certain puzzles remain unexplained. Above all, if mergers really do no good even for the merging firms’ shareholders, some explanation is needed for why their managers would still seek them out. That puzzle has been a preoccupation of management scholars and economists for a generation or more.10See, e.g., Yaakov Weber, Shlomo Yedidia Tarba & Ziva Rozen Bachar, Mergers and Acquisitions Performance Paradox: The Mediating Role of Integration Approach, 5 Eur. J. Int’l Mgt. 373 (2011); see generally discussion infra Part IV.A.1. A separate puzzle stems from the presumption that mergers are at least sometimes anti-competitive. There is an assumption that some mergers create profitable market power by reducing the number of competitors in a market and empowering those that remain to raise their prices. The whole antitrust law of merger and acquisition is predicated on that assumption. The mechanisms by which anti-competitive effects could occur are well understood,11See infra notes 107–09 and accompanying text. and evidence of it happening has recently grown.12See generally Jonathan B. Baker, The Antitrust Paradigm (2019) (reviewing recent literature). But it remains a mystery how mergers could generate market power, enabling supra-competitive prices, without systematically benefitting shareholders. If mergers give firms market power, why do firms struggle to make money from them?

While dozens of explanations have been offered for the empirical mysteries surrounding mergers and acquisitions, many can be rejected. The best-supported explanation is that large-firm managers are comparatively free from shareholder interests. In other words, the explanation for merger riddles that best fits the evidence turns out to be the familiar problem of agency cost. Managers can make decisions to some degree for their own reasons. In fact, empirical “puzzles” of this kind are common throughout corporate governance and economic regulation. They often seem puzzling only because of strong assumptions, like firm-level rationality or the strength of market institutions that will keep firms rational. That same slippage of incentives probably explains counter-intuitive trends like the seemingly counterproductive design of governance institutions,13Specifically, on the advice of academics and management advisers, corporations have widely adopted governance innovations that have been shown to be ineffective, and sometimes even harmful. Prominent examples are “board independence” reforms, see Urska Velikonja, The Political Economy of Board Independence, 92 N.C. L. Rev. 855, 858 & nn.7–8, 863–73 (2014) (documenting extensive adoption of board independence reforms and evidence that they have often been irrelevant or even harmful); Sanjai Bhagat & Bernard Black, The Uncertain Relationship Between Board Composition and Firm Performance, 54 Bus. Law. 921, 942 (1999) (evidence that board independence does little or no good); Usha Rodrigues, The Fetishization of Independence, 33 J. Corp. L. 447, 452 (2008) (same); Paul Rose, The Corporate Governance Industry, 32 J. Corp. L. 887, 909–15 (2007) (same), and efforts to involve institutional investors, see Ronald J. Gilson & Jeffrey N. Gordon, The Agency Costs of Agency Capitalism: Activist Investors and the Revaluation of Governance Rights, 113 Colum. L. Rev. 863, 889–95 (2013) (explaining how the rise of institutional investors might just replace traditional agency costs with new and different agency costs); K.J. Martijn Cremers & Simone M. Sepe, Institutional Investors, Corporate Governance, and Firm Value, 41 Seattle U. L. Rev. 387, 394 (2018) (noting the general underperformance of institutional investors as governance constraints); Mark R. DesJardine & Rodolphe Durand, Disentangling the Effects of Hedge Fund Activism on Firm Financial and Social Performance, 41 Strat. Mgt. J. 1054, 1078 (2020) (finding large short-term gains in targeted firms’ financial performance followed by substantial long-term declines). excessive executive compensation,14See generally Kevin J. Murphy, Executive Compensation: Where We Are, and How We Got There, in 2 Handbook Of The Economics Of Finance 211 (George Constantinides et al. eds., 2013) (canvassing background and evidence of incentive compensation). and the widespread pursuit of celebrity CEOs.15See generally Jonathan A. Knee, Bruce C. Greenwald & Ava Seave, The Curse of the Mogul: What’s Wrong with the World’s Leading Media Companies (2d ed. 2011) (presenting empirical evidence that pursuit of high profile, empire-building CEOs has been bad for media companies, which nonetheless continue systematically to pursue them). There is no need to believe in systematic irrationality or failed economic theory to explain these things if market institutions do not work very well to control managerial opportunism, slack, or incompetence. This Article’s concluding section will offer ideas for retooling policy without making such presumptions.

This Article comprehensively reviews the empirical literature and the several confusing puzzles it raises and works through the policy implications for the bodies of law it affects. It begins by briefly recounting the histories of American merger activity and merger regulation, both of which disclose a key theme. Popular concern over rapid consolidation drove periodic reform efforts that sought to control it, but in every case the reforms were ineffective because enforcement proved politically costly. The Article then briefly reviews the theoretical foundation on which merger activity has been justified. That is, it considers the body of economic theory that has mostly persuaded American policymakers to leave merger activity alone, the central thrust of which has always been theoretical arguments that business activity must be rational and, therefore, beneficial. That review is important in understanding the popular and government ambivalence that frustrates enforcement and in judging the policy recommendations that will be made here. The Article then turns to the heart of its study, the empirical mergers and acquisitions literature and the several puzzles it raises. It concludes by working out some implications for the legal rules that mostly deal with acquisitions: the law of corporations, securities, and antitrust.

These policy lessons are not extensive or revolutionary and they are secondary to the Article’s main focus, which is on the empirical puzzle. But they are nevertheless fundamental. The lesson for corporate and securities law is simply that market institutions cannot constrain managerial agency costs in the way the literature has overwhelmingly presumed. The literature’s key remedy for agency costs remains the market for corporate control, and many scholars cling to a commitment that it can outperform legislation or judicial intervention. But that market institution depends on an efficient world of hostile takeover that the empirical evidence discredits entirely.

As for antitrust merger law, there is no longer any compelling reason for the caution by which the law has been rendered so ineffective. The policy judgment driving it—that we should be cautious in stopping mergers because of the good they might do—turns out to have effectively no support. Meanwhile, the theory that mergers can do harm is well established, and empirical corroboration is growing and real. The balance between false positives and false negatives should be reset.

I. A Brief History and Legal Context: Mergers Go
Systematically Unregulated

There were not always corporate acquisitions in the United States, at least in routine numbers. Before the turn of the twentieth century, state corporation laws sharply limited acquisitions.16Merger, combination, and ownership of one corporation’s stock by another was mostly illegal in the United States until a set of New Jersey statutes of the 1880s and 1890s. The restrictions were driven in part by popular hostility to accumulations of capital or private power. See Fletcher Cyclopedia of Corporations § 2 (2013) [hereinafter Fletcher Cyclopedia]; Nelson Ferebee Taylor, Evolution of Corporate Combination Law: Policy Issues and Constitutional Questions, 76 N.C. L. Rev. 687, 695–753 (1998). Acquisitions were also probably hindered by a lack of limited-liability incorporation and modern capital markets.17See George J. Stigler, Monopoly and Oligopoly by Merger, 40 Am. Econ. Rev. Papers & Proc. 23, 28 (1950). Even during the late 1800s, when mergers were finally freed up by less restrictive state laws and easier access to capital, they were met with antitrust challenges and popular opposition. However, mergers finally arrived in numbers around the turn of the twentieth century, and when they did, they exploded. A decade-long wave of mergers began in 1895 that is still known as “the Great Merger Movement.” Historians disagree over its causes,18Several possible explanations of the Great Merger Movement are advanced in careful histories on an extensive record, including that it was an efficiency enhancing hunt for scale economies and vertical rationalization. See Alfred D. Chandler Jr., The Visible Hand: The Managerial Revolution in American Business (1993). For an argument that it was at least partly a scramble to reduce excess capacity following the depression of 1893, see Herbert Hovenkamp, The Opening of American Law: Neoclassical Legal Thought, 1870–1970 (2014); that it was just the understandable reaction of an entire society struggling to find order in what seemed like worsening chaos, see Louis Galambos, The Emerging Organizational Synthesis in Modern American History, 44 Bus. Hist. Rev. 279, 282 (1970); that indeed it may have been inadvertently caused by antitrust law itself, see George Bittlingmayer, Did Antitrust Policy Cause the Great Merger Wave? 28 J. L. & Econ. 77 (1985); or that, after all is said and done, it really was just the hunt for supracompetitive monopoly profits that most people always thought it was, see Naomi R. Lamoreaux, The Great Merger Movement in American Business, 1895–1904 (1988). Whatever its deeper motives, it appears to have been coordinated in part by a speculative, oversold, Wall Street investment scheme. See Gabriel Kolko, The Triumph of Conservatism: A Reinterpretation of American History, 1900–1916 (1963); Ron Chernow, House of Morgan (1990); Stigler, supra note 17, at 30 (merger “permitted a capitalization of prospective monopoly profits and [their] distribution . . . to the professional promoter,” enabling “a Morgan or a Moore to enter a new and lucrative industry: the production of monopolies”). but in its course it caused the elimination of thousands of firms and was the largest business consolidation in history.19See generally Lamoreaux, supra note 18.

The experience since has been one long stream of merger activity, punctuated every few decades by further massive waves. There have been perhaps five or more major waves since the Great Merger Movement, each on some measures larger than the last, reshaping whole sectors.20See F. M. Scherer, A New Retrospective on Mergers, 28 Rev. Indus. Org. 327, 328 fig.1 (2006). While determining whether there has been a “wave” at any given time is itself a complicated empirical problem,21See generally Klaus Gugler, Dennis C. Mueller & Michael Weichselbaumer, The Determinants of Merger Waves: An International Perspective, 30 Int’l J. Indus. Org. 1 (2012). Indeed, perceptions that there have been waves appear sometimes to be wrong. Scherer, supra note 20, at 328 fig. 1. no one doubts that there have been a half-dozen or so. Interestingly, each merger wave has had its own peculiar character. After the Great Merger Movement, which mainly consisted of horizontal mergers and often generated near-monopoly in particular sectors, a wave in the 1920s consisted of smaller deals that have been called “merger for oligopoly.”22See Stigler, supra note 17, at 31–33. Conglomerate deals dominated another wave during the 1960s, and another wave in the 1980s involved large deals and new financial or strategic novelties, including heavy leverage and hostile takeover. Another wave in the 1990s culminated in the dot-com bust of (circa) 2001, and perhaps one or two more have occurred since.23See Grocer, supra, note 1. In addition to the fact that mergers come in waves, there is the interesting and poorly-explained fact that waves are, to some degree, industry-specific. In each given wave, merger activity will be most intense in one or a few industries.24See infra notes 183–84 and accompanying text.

The law, for its part, has always imposed various limits on this enthusiasm, though it probably only meaningfully constrained mergers during short episodes. Mergers have predominately been addressed through three areas of law: corporate, securities, and antitrust law. The history of these laws’ merger rules is a fascinating topic on its own, fraught with confusion and ideology. Waves of mergers have generally led to some public controversy and have often prompted legal reforms. Reform efforts have occasionally been so major that they influenced presidential elections or whole political eras. In every case, however, reforms have faded, often quickly, when enforcement proved politically costly. These varying efforts and the political challenge they pose reflect America’s complex, changing, and often confused efforts to reconcile popular fear of corporate power with a generally liberal political philosophy that favors private enterprise.25These themes are discussed at much greater length in Chris Sagers, United States v. Apple: Competition in America 43–53 (2019), and Gary Gerstle, The Protean Character of American Liberalism, 99 Am. Hist. Rev. 1043 (1994). And so it is that after a century of popular concern and legal efforts to constrain them, mergers are more prevalent now than at any previous time and are largely free of meaningful legal constraints.

The rest of this Part explores these policy trends and sets out the major legal rules that currently govern mergers and acquisitions.

A. Corporate and Securities Law: Shareholder and Constituency Protections

For the most part, corporate and securities law rules governing mergers serve the routine goal of protecting shareholder value. Passive investors are thought to face various threats in change-in-control transactions. State corporate law attempts to protect passive investors with procedural requirements, the “appraisal” remedy, and the general demand that fiduciaries take care for shareholder value. Hostile takeovers in particular are subject to special fiduciary protections.26See generally Theresa A. Gabaldon & Christopher L. Sagers, Business Organizations 697–792 (2d ed. 2018) (canvassing the statutory procedures and fiduciary protections applicable to corporate control transactions). However, passive shareholders are not the only concern of corporate and securities law. State anti-takeover laws and Securities & Exchange Commission (SEC) oversight under the Williams Act protect other constituencies.27The Williams Act, Pub. L. No. 90–439, 82 Stat. 454 (1968), (codified, as amended, at 15 U.S.C. §§ 78m(d)-(e), 78n(d)-(f)); see infra notes 39–43 and accompanying text.

Corporate control transactions generally require approval by target-firm shareholders and directors.28See generally Del. Code Ann. tit. 8, subch. IX; Rev. Mod. Bus. Corp. Act chs. 11-12 (2003). More generally, the law protects shareholders by imposing fiduciary duties on officers and directors and insists that management make their decisions in the interests of the shareholders.29See, e.g., Shlensky v. Wrigley, 237 N.E.2d 776, 180–81 (Ill. Ct. App. 1968); Dodge v. Ford Motor Co., 17 N.W. 668 (Mich. 1919); Blasius Indus., Inc. v. Atlas Corp., 564 A.2d 651 (Del. Ch. 1988) (citing Schnell v. Chris Craft Indus., 285 A.2d 437 (Del. 1971)); Fletcher Cyclopedia, supra note 16, § 1035.50 (2010). Consequently, managers negotiating, approving, and recommending sale-of-control transactions must ensure that shareholders receive “fair value.”30Smith v. Van Gorkom, 488 A.2d 858, 864 (Del. 1985); Weinberger v. UOP, Inc., 457 A.2d 701, 710–15 (Del. 1983). Even where full business-judgment-rule deference is available, the managers are obliged to investigate fair value adequately, and, in principle, can be responsible for a breach of their fiduciary duty to shareholders if they fail.31Van Gorkom, 488 A.2d at 872–74. Shareholders are also protected by the statutory appraisal remedy when their firms are acquired, under which dissenting shareholders can sometimes ask the courts to revalue their shares.32See generally 15 Fletcher Cyclopedia, supra note 16, ch. 61. Though it is a limited remedy in several ways, appraisal litigation has blossomed in recent years.33See Audra Boone, Brian Broughman & Antonio J. Macias, Merger Negotiation in the Shadow of Judicial Appraisal, 62 J. L. & Econ. 281 (2019); Wei Jiang, Tao Li, Danqing Mei & Randall Thomas, Appraisal: Shareholder Remedy or Litigation Arbitrage?, 59 J. L. & Econ. 697 (2016); Charles R. Korsmo & Minor Meyers, Appraisal Arbitrage and the Future of Public Company M&A, 92 Wash. U. L. Rev. 1551 (2015).

In the last several decades, a set of narrower state fiduciary rules evolved to deal with the particular problem of hostile takeovers and the entrenched managements who often resist them. Hostile takeover is thought to be desirable to target-firm shareholders, both because they can enjoy large premiums and because hostile deals might unseat ineffective managers. Incumbent management, however, usually attempts to stop those deals to protect their interests; thus, there is an inherent tension. Courts recognized the tension by mid-century, and by 1964 they developed a modest rule to curtail it.34Cheff v. Mathes, 199 A.2d 548, 554–55 (Del. 1964). They took renewed interest during the period of hostile takeovers of the mid-1980s and substantially toughened their oversight of incumbent managers. This was in part because the 1980s takeover wave generated substantial public concern and, in part, because it spawned a complex new world of anti-takeover defenses, deal protections, and hard-fought, multi-party acquisition battles.35Takeover defenses and deal protections are steps that firms take to protect themselves from hostile takeover. These can be as simple as self-tenders to buy back some of the firm’s own stock, and as complex as elaborate “poison pill” arrangements. See generally Richard A. Booth, The Promise of State Takeover Statutes, 86 Mich. L. Rev. 1635, 1659–66 (1988). Starting in 1985,36Unocal Corp. v. Mesa Petroleum Co., 493 A.2d 946 (Del. 1985). the Delaware Supreme Court enhanced the mild restrictions it had traditionally put on management self-defense.37Id. at 955. Under Unocal, directors are in breach of their fiduciary duties if they adopt anti-takeover defenses that are not proportionate and based on reasonable investigation. The Unocal rule applies to all defensive measures, whenever adopted, even if adopted when no hostile takeover attempt is pending. Moran v. Household Int’l, Inc., 500 A.2d 1346, 1350, 1355–57 (Del. 1986). A special rule applies to defensive measures that interfere with shareholders’ voting rights, such as by packing board seats or moving election dates to thwart proxy efforts. Such actions require some “compelling justification.” See Blasius Indus., Inc. v. Atlas Corp., 564 A.2d 651, 661–63 (Del. Ch. 1988). In certain special cases, directors are entirely precluded from stopping a takeover and can only take action to ensure the highest premium price.38Revlon, Inc. v. MacAndrews & Forbes Holdings, Inc., 506 A.2d 173 (Del. 1986).

A very interesting part of this history, and further evidence of how American business regulation is frustrated by political ambivalence, is that merger and acquisition rules are often internally contradictory. The rules discussed so far effectively favor changes in control, especially the fiduciary rules for hostile takeovers. The rules encourage changes in control because shareholders stand to gain financially from control premiums.

But separate reforms arose around mid-century, when America saw its first real incidence of hostile takeovers—first during the 1960s and then with greater intensity during the 1980s39Strictly speaking, despite popular perception, most control transactions in the 1980s were not hostile; the percentage of transaction that were hostile was probably no more than about 14%. See, Gregor Andrade, Mark Mitchell & Erik Stafford, New Evidence and Perspectives on Mergers, 15 J. Econ. Persp. 103, 104 (2001). More of them occurred as compared to the past; however, several takeovers were very large and achieved extraordinary attention, which drove a variety of legal reform efforts. Along with them came certain other innovations that seemed to many jarring and dangerous, including heavily leveraged going-private transactions and seemingly risky financial innovations, such as takeovers funded with junk bonds.—which sought to protect incumbent managers and local communities. This body of rules cuts the other way. These separate rules seek to discourage takeovers, or at least slow them down and control them, because of the damage they might do. The rhetoric driving these rules could sometimes be vitriolic, and anti-takeover animus often ran high. While debating the bill that eventually became the Williams Act in 1968, one co-sponsor called tender offerors “pirates,”40111 Cong. Rec. 28, 257 (1965) (statement of Sen. Harrison A. Williams). and another compared hostile takeover to rape.41113 Cong. Rec. 857–58 (1967) (statement of Sen. Thomas H. Kuchel). For what it is worth, under pressure from academics and securities regulators, that particular bill took on a policy of neutrality, seeking only to protect target shareholders from non-disclosure and coercion. Under that law as adopted, plus several SEC rules interpreting it during the 1980s, meaningful acquisitions of public company stock require certain disclosures, and must comply with a timeline and various procedural protections.42Namely, tender offer proponents must make disclosures with the SEC and target shareholders, 15 U.S.C. § 78n(d)(1); Rule 14d-6, 17 C.F.R. § 24014d-6 (2010); tender offers must be held open for twenty business days after the offer is first made, SEC Rule 14e-1(a), 17 C.F.R. § 240.14e-1(a) (2010); tendering shareholders are given a period during which their tender can be withdrawn, § 14(d)(9), 15 U.S.C. § 78n(d)(9); shares must be purchased pro rata from each tendering shareholder if the offer is oversubscribed, § 14(d)(6), 15 U.S.C. § 78n(d)(6); and offerors must follow a “best price” rule, under which if the proposed purchase price is increased during the pendency of the offer, that higher price must also be paid, even to shareholders who may already have tendered, § 14(d)(7), 15 U.S.C. § 78n(d)(7). Tender offers are also subject to a special anti-fraud rule, § 14(e), 15 U.S.C. § 78n(e). See generally Arnold S. Jacobs, The Williams Act: Tender Offers and Stock Accumulations (2020); Edward F. Greene, Regulatory and Legislative Responses to Takeover Activity in the 1980s: The United States and Europe, 69 Tex. L. Rev. 1539, 1544–56 (1991).

Similarly, since the perceived takeover crisis of the 1980s, many states have adopted “anti-takeover statutes” to limit control changes. Generally, they sought to protect communities and local interests from the dislocations of hostile takeover or movement of firms out-of-state.43The number and variety of these statutes is large. The first wave, which proliferated during the 1980s, were mostly “control share” statutes, which imposed additional procedures or target-shareholder rights wherever a bidder acquired some specified minimum of the shares of a firm incorporated in the particular state. Later generations of anti-takeover statutes empowered incumbent management to defend against unfriendly acquisitions, as by specifically disavowing Delaware’s Unocal and Revlon duties, or by authorizing specific defensive measures. See generally Michal Barzuza, The State of State Antitakeover Law, 95 Va. L. Rev. 1973 (2009); Booth, supra note 36, at 1670–81. On the various concerns that motivated these many laws, and the particular concern of many of them to protect local communities and non-shareholder constituencies, compare Alan E. Garfield, Evaluating State Anti-Takeover Legislation: A Broadminded New Approach to Corporation Law or “A Race to the Bottom”?, 1990 Colum. Bus. L. Rev. 119 (1990), with Lyman Johnson & David Millon, Missing the Point About State Takeover Statutes, 87 Mich. L. Rev. 846 (1989), and Lyman Johnson & David Millon, Misreading the Williams Act, 87 Mich. L. Rev. 1862 (1989). Various statutory and regulatory controls have been proposed e that would limit or discourage takeovers, including miscellaneous tax rules,44See, e.g., Kevin J. Murphy, Executive Compensation: Where We Are, and How We Got There, in 2A Handbook Of The Economics Of Finance 212, 302 (George Constantinides et al. eds., 2013) (noting tax penalties on some sizeable golden parachute payments in takeovers). an interpretation of a Federal Reserve lending rule modestly limiting leveraged deals,45See Greene, supra note 42, at 1553–54 (discussing interpretation of Regulation G). and a spate of attempts by Democrats in Congress to augment the Williams Act.46For example, Jensen lists more than twenty separate bills introduced in 1985 alone that would have limited or added new oversight of hostile takeovers. Michael C. Jensen, The Takeover Controversy: Analysis and Evidence, in Knights, Raiders, and Targets: The Impact of Hostile Takeover 314 (John C. Coffee, et al., eds. 1988) [hereinafter Jensen, The Takeover Controversy]; see also S. Rep. No. 265–100 (1987).

This mish-mash of disparate rules has occasionally imposed some limits on mergers, and during a few recent periods, they have generated significant amounts of important litigation. For the most part, however, these rules have made little impact; at the moment, they are mostly effectively unenforceable. So long as no one negotiating a deal is conflicted, garden-variety fiduciary challenges to takeover are analyzed under the business-judgment-rule. Fiduciary Challenges under that standard are all but literally unwinnable. When plaintiffs once scored a victory in a closely watched case at the beginning of the 1980s takeover wave,47Smith v. Van Gorkom, 488 A.2d 858, 858 (Del. 1985). state legislatures quickly rendered it almost completely inert.48As leading examples, Del. Code Ann. tit. 8, § 102(b)(7) (West 2022) and Rev. Mod. Bus. Corp. Act § 2.02(b)(4) (2003) both authorize corporations to bar money damages recovery against directors for mere, unconflicted breaches of the duty of care. Most Delaware corporations have elected to protect their directors under section 102(b)(7). Accordingly, money damages are unavailable unless plaintiff can allege some conflict or illicit gain, and the incentive to sue is drastically reduced. See, e.g., James D. Cox & Randall S. Thomas, Delaware’s Retreat: Exploring Developing Fissures and Tectonic Shifts in Delaware Corporate Law, 42 Del. J. Corp. L. 323, 334 (2018). Even in the presence of an explicit conflict of interest, as when minority shareholders are misled or coerced into sale by a parent corporation,49See, e.g., Weinberger v. UOP, Inc., 457 A.2d 701, 710–12 (Del. 1983). recent doctrinal innovations in Delaware have made it possible for merging firms in essentially all circumstances to secure business-judgment treatment and make all claims against them easily dismissible.50Namely, Kahn v. M & F Worldwide Corp., 88 A.3d 635 (Del. 2014), held that in a parent-subsidiary merger—in which by definition the parent is a fiduciary acting under conflict of interest—defendants can avoid the plaintiff-friendly duty of loyalty standard by conditioning merger approval on certain specific procedures. The business judgment rule applies if the merger is conditioned on negotiation by a fully independent board on behalf of the subsidiary and approval by the subsidiary’s fully informed shareholders. Id. at 644. Accordingly, in the wake of all these changes, Delaware fiduciary litigation has ground down to nearly nothing. Meanwhile, the Williams Act was never very restrictive to begin with, and state anti-takeover statutes are mainly compliance obligations that rarely impose meaningful limits.

Finally, while there has been something of a renaissance in appraisal litigation and while early evidence suggests it may have generated shareholder benefits,51Boone et al., supra note 33; see also Albert H. Choi & Eric Talley, Appraising the “Merger Price” Appraisal Rule, 34 J. L. Econ. & Org. 543 (2018) (setting out a formal model predicting that the availability of a robust appraisal remedy leads to higher merger prices, notwithstanding market efficiency). serious caveats apply. First, appraisal is subject to several limits that will make it unavailable or unrealistic in many cases,52Appraisal, generally, is not very useful for non-institutional shareholders, because of its cost. While it can be brought on contingency terms, and while courts can award attorney fees at their discretion, it is not generally brought on a formal class basis, and fees are not guaranteed. See Fletcher Cyclopedia, supra note 17, at § 7164.10. It is also subject to certain exclusions. Most importantly, both Delaware and Model Act jurisdictions generally deny appraisal rights where the target is publicly traded, unless the deal requires its shareholders to accept something other than liquid securities or cash. See Del. Code. Ann. tit. 8, § 262(b)(1) (2022); Rev. Mod. Bus. Corp. Act § 13.02(b)(1) (2003). Thus, appraisal is unavailable for many large transactions. and there is indication that courts may soon restrict it further.53See DFC Glob. Corp. v. Muirfield Value Partners, L.P., 172 A.3d 346 (Del. 2017) (accepting conditionally actual deal price as strong evidence of fair price where circumstances suggest market efficiency). More importantly, hedge funds bring a large proportion of current appraisal litigation, and there is real uncertainty about whether it benefits other shareholders or anyone else. But fundamentally, even effective appraisal—and, for that matter, most of the corporate law of mergers and acquisitions—ignores the social problem at the heart of this Article. An appraisal claim aims to prove that a takeover premium is too low. Its effect would exaggerate its estimate of the social value of acquisitions even further when that value appears not to exist.

B. Competition Protections

A separate and very different regime regulates mergers and acquisitions under antitrust law, which serves different goals and constituencies. Since 1914, section 7 of the Clayton Act has prohibited business combinations whose “effect . . . may be substantially to lessen competition, or to tend to create a monopoly.”5415 U.S.C. § 18.

Over its long and complicated history, antitrust merger law has probably not been very effective. A recurrent theme over the course of this history has been a struggle between Congress and the courts over whether enforcement would be vigorous and meaningful—as Congress has repeatedly indicated was its intention—or restrained and cautious. Every few decades Congress intervenes to shore it up, but each time the courts have re-construed the law in ways favorable to defendants. That was true of the initial adoption of the basic merger law, Clayton Act section 7, as well as each significant reform thereafter.55Congress has revisited merger legislation many times, essentially always to strengthen enforcement, and often enough to reverse perceived judicial resistance to it. In an often-told story, § 7 itself was meant to correct the courts’ failure to control concentration with the only federal law then available, the Sherman Act. See William Letwin, Law and Economic Policy in America: The Evolution of the Sherman Antitrust Act 267–78 (1965); Martin J. Sklar, The Corporate Reconstruction of American Capitalism, 1890–1916: The Market, the Law, and Politics 289–332 (1988); Lawrence A. Sullivan, Warren S. Grimes & Christopher L. Sagers, The Law of Antitrust: An Integrated Handbook § 8.1a (3d ed. 2015). Over its first decades, however, neither section 7 nor the Sherman Act was used effectively as a merger law. Popular frustration with weak enforcement and the perceived growth of concentration culminated in the 1941 final report of the Temporary National Economic Committee and an influential study of the Federal Trade Commission. Temp. Nat. Econ. Comm., Final Report and Recommendations 38–40 (1941); Fed. Trade Comm’n, The Merger Movement – A Summary Report in Annual Report for the Fiscal Year Ended June 30 1948, at 16–22 (1948) [hereinafter Fed. Trade Comm’n, The Merger Movement]. Matters perhaps reached their breaking point after a much-criticized decision, United States v. Columbia Steel Co., 334 U.S. 495 (1948), approved what was perceived to be another large and dangerous steel merger. Note, Section 7 of the Clayton Act: A Legislative History, 52 Colum. L. Rev. 766, 768 & n.10 (1952). There followed the storied adoption of the Celler-Kefauver Amendments of 1950, ch. 1184, 64 Stat. 1125 (1950), widely understood to intend stricter enforcement, and a string of decisions during the years of the Warren Court giving those amendments life that were stricter than federal merger control had ever been. See infra notes 57–60 and accompanying text. But barely more than ten years after this jurisprudence began to take root, a Supreme Court repopulated by the large number of Richard Nixon’s appointees undermined it in a set of decisions that would largely end merger enforcement as a meaningful institution. See infra notes 61–67. Indeed, even as Congress was in the midst of adopting another historic, pro-enforcement reform—the system of pre-merger review created by the Hart-Scott-Rodino Antitrust Improvements Act of 1976, Pub. L. No. 94–435, 90 Stat. 1383 (1976), now codified as Clayton Act § 7a, 15 U.S.C. § 18a— that same Supreme Court announced one of the more surprising retrenchments in the law’s history. Though section 7 had been in effect for sixty years and substantially amended to expand its coverage, the Burger Court purported to discover in 1975 that the law had in fact been misread as applying too broadly. The Court held that it applied only to mergers that were genuinely interstate in character, in that the firms involved were in different states. United States v. Am. Bldg. Maint. Indus., 422 U.S. 271, 276 (1975). Learned commentators met that announcement with surprise and consternation, see, e.g., Antitrust Procedural Improvements and Jurisdictional Amendments: Hearings Before the H. Subcomm. on Monopolies and Com. L. of the H. Comm. on the Judiciary, 108, 111-12, 96th Cong., 1st Sess. (1979) (statement of Eleanor Fox), and Congress quickly reversed it, Antitrust Procedural Improvements Act of 1980, Pub. L. No. 96–349, § 6(a), 94 Stat. 1154, 1157 (1980) (codified as amended at 15 U.S.C. § 18). Most significant for the law’s modern life was a 1950 amendment to section 7,56Celler-Kefauver Act, Pub. L. 81–899, 64 Stat. 1125 (1950). on the heels of what was perceived as a wave of consolidation in the 1940s and growing popular dissatisfaction with underenforcement.57See, e.g., Fed. Trade Comm’n, The Merger Movement, supra note 56 (government study influential during the period of Clayton Act reform, documenting rise in concentration). The Supreme Court, finding in the legislative history a desire to toughen merger law,58Brown Shoe Co. v. United States, 370 U.S. 294, 315–23 (1962). For similar views of the legislative history, see Derek C. Bok, Section 7 of the Clayton Act and the Merging of Law and Economics, 74 Harv. L. Rev. 226, 231–38 (1960). deployed the amended law in a series of decisions during the 1960s that adopted simplified, streamlined, and extremely enforcement-friendly prophylactic rules. Most importantly, in the 1963 United States v. Philadelphia National Bank (PNB) decision,59United States v. Phila. Nat’l Bank, 374 U.S. 321 (1963). the Court adopted a still-used burden-shifting rule under which a merger can be found illegal on nothing more than the increased concentration in an already concentrated market.60Id. at 362–63. Nominally, the plaintiff’s demonstration of concentration statistics was only a prima facie showing, and defendant could rebut it with evidence that concentration alone did not prove market power. See id. The courts applied the rule very strictly, however, and for the first decade of its life the government always won on the prima facie showing. For a time, that innovation was very effective. Over the course of the next decade, the federal government won twelve merger cases before the Supreme Court and lost none.61Eleanor M. Fox, Antitrust, Mergers, and the Supreme Court: The Politics of Section 7 of the Clayton Act, 26 Mercer L. Rev. 389, 396–97, 408, 410 (1975).

However, vigorous antitrust merger enforcement met political hostility and quickly proved unsustainable. By 1974, just over a decade after PNB, a Supreme Court newly populated with Richard Nixon’s unusual number of appointments62For elaboration, see generally Chris Sagers, #LOLNothingMatters, 63 Antitrust Bull. 7, 20–21 n.69 (2018). announced a set of closely divided decisions reaching much more defense-friendly results.63The cases were United States v. Citizens & S. Nat. Bank, 422 U.S. 86 (1975) (6-3), United States v. Marine Bancorporation, Inc., 418 U.S. 602 (1974) (5-3), and United States v. Gen. Dynamics Corp., 415 U.S. 486 (1974) (5-4). Lower courts took them to significantly soften the PNB presumption and demand much more proof from merger plaintiffs.64As something of a historical accident, modern merger law has been made exclusively by the lower courts. The Supreme Court has not taken a merger case in many, many years. See Sullivan et al., supra note 56, at 476–79. Among the most influential lower court decisions were United States v. Baker Hughes Inc., 908 F.2d 981 (D.C. Cir. 1990), decided by a D.C. Circuit panel including then-Judges Clarence Thomas and Ruth Bader Ginsburg, and United States v. Waste Mgt., Inc., 743 F.2d 976 (2d Cir. 1984). Because merger policy is almost inherently speculative—enforcement actions are usually brought prospectively before deals are consummated—the setting of evidentiary burdens largely determines which parties win or lose. At length, while PNB nominally remains the law, and is still cited as seminal,65Recent appellate opinions from a variety of circuits cite PNB as leading authority. See, e.g., FTC v. Advocate Health Care Network, 841 F.3d 460, 464, 467, 468, 469, 470, 474, 476 (7th Cir. 2016); FTC v. Penn State Hershey Med. Ctr., 838 F.3d 327, 344, 348 (3d Cir. 2016); Saint Alphonsus Med. Ctr.-Nampa Inc. v. St. Luke’s Health Sys., 778 F.3d 775, 783, 786 (9th Cir. 2015); Polypore Int’l, Inc. v. FTC, 686 F.3d 1208, 1213-18 (11th Cir. 2012). its rule is honored only in the breach. With some exceptions, only the very largest horizontal mergers are now challenged,66 John Kwoka, Mergers, Merger Control, and Remedies: A Retrospective Analysis of U.S. Policy 24–33 (2014) (showing that the effective threshold of concentration beyond which agencies will challenge mergers has substantially risen). and not infrequently even those claims lose67The government admittedly still wins merger cases, but the cases it wins are exclusively horizontal and involve massive concentration numbers. See, e.g., United States v. Anthem, Inc., 855 F.3d 345, 351 (D.C. Cir. 2017) (ruling for government, where merger would result in HHI as high at 3675); United States v. Aetna Inc., 240 F. Supp. 3d 1, 42–43 (D.D.C. 2017) (ruling for government, where merger would result in HHIs in most local markets of more than 5000). And plaintiffs still lose even the biggest cases. Consider, for example, T-Mobile’s acquisition of Sprint in 2020, a four-to-three merger in a market with high entry barriers and inelastic demand, see New York v. Deutsche Telekom, 439 F.Supp.3d, 179 228 (so finding), generating very high concentration numbers, id. at 206 (finding that merger would increase HHI by 679 points, to 3186). However, after the Justice Department approved the deal, on a transparently feeble package of remedial asset divestitures and problematic circumstances, a group of state Attorneys General brought their own suit and lost. Id. or just barely win.68Consider FTC v. Staples, Inc., 190 F. Supp. 3d 100 (D.D.C. 2016). Not so many years earlier, the FTC had blocked a merger between the same parties in a celebrated courtroom victory, FTC v. Staples, Inc., 970 F. Supp. 1066 (D.D.C. 1997), but the parties attempted to renew the deal in 2015, arguing that circumstances had changed. The new deal would generate breathtaking concentration numbers. On the market definition accepted by the court as fact, the merger would create a firm holding nearly 80%, and it would be fifteen times larger than its nearest rival. (In technical terms, the initial HHI of 3270 would increase by nearly 3000 points, to 6265). Staples, 190 F. Supp. 3d at 128. But what is truly amazing about the case is that the FTC barely won. As proof how weak the law has become, the defense was so confident they would win that they rested without putting on any evidence. Id. at 110. But even with no defense case at all, the court wrote that the case was close, that the decision posed serious challenges, and hinted that had there been just somewhat more evidence of likely entry, the government would have lost. See id. at 136 (“If Amazon Business was more developed [and more clearly ready to enter], the outcome of this case very well may have been different.”).

* * *

In the end, America’s long history of fretting over its hundreds of thousands of mergers and the tension between the fear of mergers’ bad consequences and the risk of mistakenly upsetting desirable business behavior, has left us in a remarkable state. The United States administers a regime of merger law that keeps thousands of attorneys, economists, and government officials employed. It entails substantial social costs,69Cf. Joe Sims & Deborah P. Herman, The Effect of Twenty Years of Hart-Scott-Rodino on Merger Practice: A Case Study in the Law of Unintended Consequences Applied to Antitrust Legislation, 65 Antitrust L. J. 865, 890–92 (1997) (early estimate of the administrative costs of the antitrust pre-merger clearance system). but virtually never constrains any transaction in any meaningful way. Given this extraordinary state of affairs, one would imagine not only that mergers do some good, but that they must generate really exceptional, dramatic, and measurable social gains. It would seem unreasonable to bend so far backwards to facilitate them unless they do. And yet, at least so far as any evidence has been able to establish, they do not.

II. The Theoretical Foundation on Which Mergers Become
Unregulated: That There Must Be Something Good About Them

One major contributor to the ambivalent politics of merger history is the diverse body of theories that predict social benefits from mergers. They are of basically three varieties: (a) the argument that, first and foremost, mergers generate “efficiencies” or “synergies” or in some other way deliver net social gain; (b) closely related theoretical controversies over whether concentration in itself can indicate social injury or risk—whether “big” is “bad”—as was implied in antitrust policy under the Warren Court; and finally, (c) the rather different argument that policy should foster a “market for corporate control,” because the possibility of hostile acquisition disciplines incumbent management and directs assets to the most capable owners.

These theories are related. Each arose or was significantly advanced during the revolution among American economists commonly thought of as a “conservative” revolution, and each came to exert extensive influence on the law. But while each came to face substantial academic critique and none now holds more than a controversial position within academia, their influence on the law has remained unchanged.

A. Efficiencies

The chief theoretical justification for growth by acquisition has always been efficiency. In various ways, proponents of merger activity claim that firms can make their goods better or more cheaply if they can do it at a larger volume, and that for various reasons growth by acquisition might be preferable to internal growth. Theorists have suggested a huge range of possible mechanisms by which growth could generate efficiency.70The specific mechanisms by which growth or consolidation could drive efficiency are, at least in principle, many. Producing at larger volume can generate scale economies where fixed costs are comparatively large. See F.M. Scherer & David Ross, Industrial Market Structure and Economic Performance 97 (3d ed. 1990). Integrating different technical processes in one firm can sometimes make the processes more technologically efficient, regardless of whether scale changes. See, e.g., M.A. Adelman, Integration and Antitrust Policy, 63 Harv. L. Rev. 27, 31–32 (1949) (describing the potential savings in steel making if the furnace for making pig iron and the furnace for finishing it into steel are located in the same factory, avoiding the cost of heating the same materials twice). Economists and business theorists have suggested any number of other mechanisms that might lower cost or raise quality of production, many of which in principle could be improved through growth or consolidation. No one seriously doubts that efficiencies can occur with growth or consolidation, and real-world evidence exists.71See, e.g., Aloke Ghosh, Increasing Market Share as a Rationale for Corporate Acquisitions, 31 J. Bus. Fin. & Acct. 209 (2004); Joachim Schwalbach, Directorate General for Economic and Financial Affairs, European Comm’n, Economies of Scale in Intra-Community Trade 1998 Econ. Papers No. 68, http://aei.pitt.edu/36976/1/A3011.pdf [https://perma.cc/48V9-7ZSC]; Aubrey Silbertson, Economies of Scale in Theory and Practice, 82 Econ. J. 369 (Special Issue) (1972). However, there is serious reason to doubt that efficiencies are significant or common in mergers of the size and competitive risk relevant to public policy, much less that they commonly outweigh the risks posed.

Before the mid-20th century, traditional arguments for merger efficiencies tended to be simple-minded generalities. Even now, it is often said without elaboration that mergers will make bigger firms “stronger competitors” or better able to confront buyers or suppliers with market power.72See, e.g., Jamie Henikoff Moffitt, Merging in the Shadow of the Law: The Case for Consistent Judicial Efficiency Analysis, 63 Vand. L. Rev. 1697, 1698, 1702 (2010). In general, such claims have not been taken too seriously because simple economics implies that competition improves efficiency more than a lack of it does.73Economists have cited the efficiency benefits of competition at least since Adam Smith: Adam Smith, The Wealth of Nations 132 (Vintage Books 2020) (1776) (“Monopoly…is a great enemy to good management, which can never be universally established but in consequence of that free and universal competition which forces everybody to have recourse to it for the sake of self-defence.”). See also F.M. Scherer, Industrial Market Structure and Economic Performance 38 (2d ed. 1980); see also infra notes 195-201 and accompanying text (discussing the X-efficiency literature). However, economists began working out merger efficiencies with more theoretical rigor in the 1960s, first in a key 1968 paper by Oliver Williamson.74Williamson’s first paper was Oliver E. Williamson, Efficiencies as an Antitrust Defense: The Welfare Tradeoffs, 58 Am. Econ. Rev. 18 (1968) [hereinafter Williamson, Efficiencies]. He corrected and expanded on the original paper in Oliver E. Williamson, Economies as an Antitrust Defense: The Welfare Tradeoffs, in Readings in Industrial Economics (C. Rowley ed. 1972), and Oliver E. Williamson, Economies as an Antitrust Defense Revisited, 125 U. Pa. L. Rev. 699, 699–700 (1977) [hereinafter Williamson, Revisited]. See generally Alan A. Fisher & Robert H. Lande, Efficiency Considerations in Merger Enforcement, 71 Calif. L. Rev. 1580, 1583 (1983) (noting the seminal role of Williamson’s papers in the handling of merger efficiencies). Williamson’s insight was that under certain assumptions and for a simple mathematical reason, efficiencies could sometimes outweigh the deadweight loss of increased market power. He set out a formal demonstration, along with a table of striking numerical results, showing that mergers should usually be net socially desirable. The central theoretical insight was simple: deadweight loss affects only those sales lost by the monopolist’s price increase, whereas efficiency gains are spread across the firm’s entire remaining production.

It is worth dwelling on the serious problems that complicate Williamson’s model, because it remains a cornerstone of modern merger policy even though its weaknesses are well known.75For criticism, see Herbert Hovenkamp, Appraising Merger Efficiencies, 24 Geo. Mason L. Rev. 703, 715–24 (2017); John Kwoka, The Changing Nature of Efficiencies in Merger Analysis, 60 Antitrust Bull. 231, 233 (2015); Richard A. Posner, The Social Costs of Monopoly and Regulation, 83 J. Pol. Econ. 807, 821 (1975). Williamson acknowledged some of the weaknesses himself.76See Williamson, Revisited, supra note 74, at 710–13; Williamson, Efficiencies, supra note 75. First, he assumed a perfectly competitive pre-merger market that develops market power after just one merger. That is extremely unlikely empirically, and it is also legally irrelevant because under antitrust law—as it has existed for more than forty years—mergers in competitive markets are effectively per se legal. But it also seriously distorts the striking numbers he was able to generate.77Hovenkamp, supra note 76, at 715–24; Lars-Hendrik Röller, Johan Stennek & Frank Verboven, Efficiency Gains From Mergers, in European Merger Control: Do We Need An Efficiency Defence 108 (Fabienne Iizkovitz & Roderick Meiklejohn eds., 2006). As has now been shown several times with a useful graphical representation,78The graphic is reproduced in an Appendix. For earlier presentations, see Hovenkamp, supra note 76, at 717; Frank Mathewson & Ralph Winter, The Analysis of Efficiencies in Superior Propane: Correct Criterion Incorrectly Applied, 20 Can. Comp. Rec. 88, 96 (2000). it matters quite a lot in measuring the effects of a merger whether the pre-merger market was competitive or not. Price increases imposed on a perfectly competitive market cause relatively little harm because marginal consumers were already nearly indifferent between having the good and not having it. But if the pre-merger market already reflects substantial market power—as will be the case in all mergers likely to be challenged under antitrust law—even consumers at the margin would have paid more than marginal cost for the good. When output goes down after the merger, that benefit they enjoyed by having the product will be a deadweight loss. Real-world cases have shown that this distortion can be very large.79In a closely watched Canadian antitrust case, Comm’r v. Superior Propane Inc., 2000 Comp. Trib. 15, the court permitted a merger to an effective monopoly on a finding that efficiencies hugely outweighed deadweight loss. But it could only find that disparity because it assumed that pre-merger prices were competitive. A subsequent study more plausibly estimated a pre-merger overcharge and found post-merger deadweight loss to be 8.5 times larger than that measured by the Court, largely wiping out the estimated efficiency. See Mathewson & Winter, supra note 78, at 89, 91. Matthewson and Winter could calculate the pre-merger overcharge from an econometric estimate of demand elasticity at the pre-merger price (accepted by the court in its decision). Demand elasticity greater than one implies that price is above marginal cost. See id. at 90–91.

A second major problem is that Williamson hypothesized mergers in which there would be both a price increase (which will reduce quantity) and some cost reduction.80See Williamson, Efficiencies, supra note 74, at 18 n.2. But it is empirically unlikely that gains from scale would accompany a market power increase. For one thing, scale benefits likely won’t improve when output goes down. Though a merged firm can simultaneously increase its own scale while reducing total output, any scale benefits gotten that way are emphatically the kind that competition itself is supposed to produce.81Imagine that two firms each produce fifty units, but minimum efficient scale (MES) is seventy-five. If they merge, total output will go down, by at least twenty-five, but so will the merged firm’s average cost. But competition is precisely the mechanism that is supposed to generate this result. The two firms should fight for sales until one of them reaches MES, and if the other must exit the market, then that is an ordinary consequence of competition. Circumstances are likely unusual in which MES is so large that a market will support only a small number of firms. See infra notes 165–67 and accompanying text.It is also possible that an output-reducing merger could generate efficiencies other than scale economies, like improvements in management. Efficiencies of that kind tend to be very difficult to verify and unlikely to be merger specific. See U.S. Dep’t of Justice & Fed. Trade Comm’n, Horizontal Merger Guidelines § 10 (2010); Hovenkamp, supra note 75, at 711. Williamson himself claimed that the most likely efficiencies from merger would be transactional savings (he apparently thought scale efficiencies were unlikely). Those savings might be had where combination into a hierarchical firm was cheaper than complicated, contingent, long-term contracting. Williamson, Revisited, supra note 74, at 723–26. But as he acknowledged, complex contracting to achieve pro-competitive ends will occur mainly in vertical relationships, and perhaps conglomerate ones. Id. It is hard to imagine contracting problems solved by most horizontal mergers. Moreover, efficiencies from growth tend to be exhausted at a comparatively small scale.82See infra notes 165–67 and accompanying text. There was already some evidence to that effect at the time of Williamson’s article. See, e.g., Joe S. Bain, Advantages of the Large Firm: Production, Distribution, and Sales Promotion, 20 J. of Mktg. 336 (1956) (reporting results of study of 17 mostly concentrated industries, finding little evidence that MES could explain existing levels of concentration in most of them). Admittedly, the evidence that existed at the time faced the controversy typical of such developing ideas. Bain’s 1956 study, for example, has been criticized for its small and probably biased sample selection. See Stephen Davies, Minimum Efficient Size and Seller Concentration: An Empirical Problem, 28 J. of Indus. Econs. 287, 291 n.9 (1980). The model conceals other problems as well.83See Kwoka, supra note 75, at 233 (listing other problems).

In essence, Williamson made a perceptive mathematical observation about a rare, idiosyncratic set of hypothetical transactions that might never occur and would not be legally challenged if they did. Indeed, no such case has ever been identified in American merger law.84See Herbert Hovenkamp, Progressive Antitrust, 2018 U. Ill. L. Rev. 71, 90–92 (2018). Williamson predicted that such cases would be only “occasional”85See Williamson, Efficiencies, supra note 74, at 18. and repeatedly warned that his model was “naive.”86Williamson states as follows:“To be sure, the partial equilibrium welfare economics apparatus upon which I relied to display the welfare tradeoffs is a blunt instrument that can be used in an intimidating way. To forestall the risk that subtle and complex policy issues might be resolved in an undiscerning manner, I specifically labeled the simple welfare economics model as ‘naive’ and went on to introduce a number of economic and extra-economic qualifications that must be considered.”
Williamson, Revisited, supra note 74, at 701.

Nevertheless, the mere possibility of efficiencies still largely governs American merger policy. Courts, commentators, and policymakers still take for granted that very significant improvements in production cost or quality are commonly available through corporate combinations and consider it a serious risk that government interference in those transactions will inadvertently jeopardize gains that could have been realized. This idea infuses corporate and securities merger law, and, while its role in antitrust has been somewhat complex, it overwhelmingly drives antitrust as well.87Antitrust has long been ambivalent about efficiencies as a defense in litigated antitrust cases, and for much of its history refused to permit any demonstration of efficiency gains at all. See Timothy J. Muris, The Government and Merger Efficiencies: Still Hostile After All These Years, 7 Geo. Mason L. Rev. 729, 730–32 (1999). However, courts and agencies have incorporated the possibility of efficiency in a different way. They set the initial concentration threshold for an antitrust merger plaintiff’s prima facie showing under the PNB standard (see supra notes 62–71 and accompanying text) by asking at what level of concentration the likelihood of harm outweighs likely gains. This was the approach of the 1982 Merger Guidelines, see Fisher & Lande, supra note 74, at 1584–85, and observers since have continually urged that that the threshold be raised. See, e.g., Daniel A. Crane, Rethinking Merger Efficiencies, 110 Mich. L. Rev. 347, 348–50 (2011). That has in fact happened—the threshold has increased substantially. See supra note 69 and accompanying text.

B. The Disputed Relevance of Concentration: SCP, Entry Optimism,
and the Contestability “Uprising”

An argument for permissive merger policy closely related to the efficiency claims just discussed is that the size of corporations and the concentration of markets are themselves not relevant to public policy. The argument relates mainly to competitiveness and antitrust, and the claim is that size and concentration in and of themselves do not generate market power. “Big,” in other words, is not “bad.” In extreme forms, important theoretical movements have gone so far as to claim that even outright monopoly might generate no market power at all and claims even that radical have at times held sway in policy.

Perhaps more than any other issue, it was crucial to the mid-century economic revolution to argue that market concentration and firm size could not prove social harm. Up until that time, economists of the influential “Structure-Conduct-Performance” tradition (SCP) argued that concentration alone could predict bad outcomes from merger.88SCP was perhaps the first major contribution of industrial organization economics as an independent discipline, and was associated with founding figures in the field, principally economists Edward Mason of Harvard and his student Joe Bain of Berkeley. See generally Ewald T. Grether, Industrial Organization: Past History and Future Problems, 60 Am. Econ. Rev. Papers & Proc. 83 (1970); Timothy F. Bresnahan & Richard Schmalensee, The Empirical Renaissance in Industrial Economics, 35 J. Indus. Econ. 371 (1987). The SCP movement was said to be important to Warren-era merger policy because its specific premise was that concentration leads reliably and systematically to market power. See Herbert Hovenkamp, United States Competition Policy in Crisis: 1890–1955, 94 Minn. L. Rev. 311, 355–56 (2009). The movement was mainly empirical and rested on a well-established correlation between concentration and profit. The idea implied the significant policy argument that the law might regulate concentration for its own sake, without further investigating actual consequences.89For what it is worth, Mason, Bain, and their ilk were not critics of mainstream economic theory. They just found plausible a causal link between concentration and market power, which they thought was explained by collusion facilitated by concentration. They also believed they’d substantiated it empirically. See Hovenkamp, supra note 88, at 349–59; Richard Schmalensee, Collusion Versus Differential Efficiency: Testing Alternative Hypotheses, 35 J. Indus. Econ. 399, 399 (1987). The idea was basic to the PNB presumption and the strict antitrust law of the Warren Court era. During a harsh reactionary period, however, conservatives argued that government should disregard concentration entirely and regulate only conduct.90See, e.g., Yale Brozen, Concentration, Mergers and Public Policy (1982); Harold Demsetz, Two Systems of Belief About Monopoly, in Industrial Concentration: The New Learning 164 (Harvey J. Goldschmid, H. Michael Mann, & J. Fred Weston, eds. 1974). Those critics were so successful that not only did SCP fall out of favor, but empirical study of industry structure ended more or less altogether.91See Bresnahan & Schmalensee, supra note 88.

The critics did not doubt the basic empirical result underlying SCP—that concentration correlates with accounting profit. They generally acknowledged it as fact.92See, e.g., Sam Peltzman, The Gains and Losses from Industrial Concentration, 20 J. L. & Econ. 229, 229 (1977); see Schmalensee, supra note 90 (“Since the pioneering work of Joe Bain . . . a positive correlation between industry concentration and accounting measures of industry-average profitability has generally been accepted as a stylized fact.”). Instead, they attacked SCP’s causal reasoning.93Their critique very often began by observing that mid-century economists uncritically assumed that behavior inconsistent with perfect competition must be explained by monopoly. See, e.g., R.H. Coase, Industrial Organization: A Proposal for Research, in 3 Economic Research: Retrospect and Prospect: Policy Issues and Research Opportunities in Industrial Organization 59, 67 (Victor R. Fuchs ed., 1972), reprinted in R.H. Coase, The Firm, the Market, and the Law 57–74 (1988) (mocking mid-century economists when such “an economist finds something . . . that he does not understand, he looks for a monopoly explanation[,]” and because “in this field we are very ignorant, the number of ununderstandable practices tends to be rather large, and the reliance on a monopoly explanation, frequent.”); Harold Demsetz, Barriers to Entry, 72 Am. Econ. Rev. 47, (1982) [hereinafter “Demsetz, Barriers to Entry”] (“[T]he research custom of industrial organization economists during the post-World War II period . . . was to seek monopoly explanations for data not obviously or directly implied by the perfect competition model.”). Most importantly, they argued that no inference of market power could be drawn from accounting profits, which might entail no economic profit at all.94Accounting profits are simply revenue less the costs of inputs. Economic profits are revenue less the costs of those inputs (known by economists as “accounting costs”) plus a competitive rate of return to capital. The distinction is critical in economic theory because it is the reason firms might sometimes seem to earn excess “profits” without market power. Even firms in perfectly competitive markets might earn revenues in excess of accounting costs, but by definition their economic profits are still equal to zero. That is so because the investor returns needed to keep capital invested in the enterprise might vary with risk or other factors. The critics explained the concentration-profits correlation by efficiency or other advantages that might correlate with firm size,95See supra note 92; see also Michael Salinger, The Concentration-Margins Relationship Reconsidered, in Brookings Papers on Economic Activity 289 & n.7 (Arthur M. Okun & George L. Perry, eds. 1990) (discussing various explanations for apparent profits among critics of SCP). and they generated preliminary empirical support for it.96See generally Harold Demsetz, Industry Structure, Market Rivalry, and Public Policy, 16 J. L. & Econ. 1 (1973) (presenting descriptive statistical evidence that rates of return to small firms do not increase with concentration, as Demsetz expected if the prevailing explanation were correct that concentration facilitates collusion); see also Peltzman, supra note 92. For a time, it also appeared that the traditional concentration-profits correlation might have shrunk or disappeared during the 1970s. More recent evidence finds it to have returned in the early 1980s. If true, that result nicely coincides with the structural transformation and the return of substantial corporate profits so heavily implicated in more recent evidence. See Baker, supra note 12, at 13–18 & nn.30–35. Salinger suggests the temporary loss of the correlation may have been caused by import competition, which increased during the 1970s. Salinger, supra note 95, at 306–07. They further argued that if the profits typical of concentration were caused by anti-competitive behavior, they should invite entry.97See Peltzman, supra note 92. Rents could persist only if entry barriers were high, and conservative critics were skeptical that they could be.98See George Stigler, The Organization of Industry 67, 113–22 (1968); see also Demsetz, Barriers to Entry, supra note 93, at 47. They also contributed to the evolving industrial-organization theory of imperfect competition, purporting to show that the cartels or oligopoly interdependence that could generate economic profits should be difficult to sustain.99See generally George J. Stigler, A Theory of Oligopoly, 72 J. Pol. Econ. 44 (1964) (classic article formulating this influential view).

Confidence in the irrelevance of concentration led to one final theoretical elaboration that, in retrospect, seems like an almost bizarre extreme of the new market optimism. A theory known as “contestability”100For a key work introducing the theory, see William J. Baumol, John C., Panzar, & Robert D. Willig, Contestable Markets and the Theory of Industry Structure (1982).It seems not coincidental that the theory was developed by economists with ties both to the most powerful monopolist of the day and the political movement for deregulation. Its progenitors included Baumol and Willig of Princeton, who devised the theory in part to defend AT&T’s monopoly, and to urge that it be free from competition even after deregulation. Two other progenitors, Elizabeth Bailey and John Panzar, spent time at the economics unit of AT&T’s Bell Laboratories, and some of Panzar’s relevant research appeared in the Bell Journal of Economics. See William G. Shepherd, Robert D. Willig, in Pioneers of Industrial Organization: How the Economics of Competition and Monopoly Took Shape 294–95 (Henry W. de Jong & William G. Shepherd, eds. 2007). Meanwhile, Bailey would go on to serve as a member of the Civil Aeronautics Board at a time when that agency actively advocated its own deregulation, and her contributions to the theory mainly concerned contestability in deregulated airline markets. purported to explain how concentrated markets could mimic competition and, under the right circumstances, generate perfectly competitive outcomes even in total monopoly. In those markets, the mere threat of entry would constrain incumbent prices.101The most important condition for contestability was that the fixed costs of entry not be sunk, such that not only is entry easy, but exit is as well. In that case, any entrant whose variable costs were comparable to the incumbents’ costs could quickly enter in times of positive profits, and then exit again as soon as incumbent reaction made the strategy no longer profitable. Other factors thought to enhance contestability are that incumbent prices are relatively sticky, and consumers can switch suppliers relatively quickly. Entry is more profitable the more that an entrant can steal fickle customers while enjoying some period of protection from the incumbent’s price response. This theory had an implicit but deep influence on the law.102While the literature itself was not cited directly in the merger caselaw, it was part and parcel of the freewheeling judicial speculation of the time that even in already massively concentrated markets, further consolidation was not a cause for serious concern. On the theory’s influence on the courts, see Gregory J. Werden, Inconvenient Truths on Merger Retrospective Studies, 3 J. Antitrust Enf’t 287, 287–88 (2015); Richard Schmalensee, Ease of Entry: Has the Concept Been Applied Too Readily?, 56 Antitrust L.J. 41 (1987). Much of its influence probably flowed from its wide and early application in transportation sectors, where it was used to justify substantial consolidation. See generally Joseph F. Brodley, Potential Competition Under the Merger Guidelines, 71 Cal. L. Rev. 376, 394 n.76, 401 (1983) (airlines). Notably, contestability was typically ignored when its application suggested a need for stricter antitrust or regulatory intervention.103For example, contestability counters Stigler’s rejection of fixed costs as entry barriers. If fixed costs are sunk, they can deter entry significantly, even when they are fairly small. Indeed, even the incumbent’s own sunk expenses might deter entry by establishing credible commitment to repel entry with vigorous price reaction. See Baker, supra note 12, at 194–95; William J. Baumol & Robert D. Willig, Fixed Costs, Sunk Costs, Entry Barriers, and Sustainability of Monopoly, 96 Q.J. Econ. 405 (1981); Joseph E. Stiglitz, Technological Change, Sunk Costs, and Competition, 3 Brookings Papers on Econ. Activity 883, 891 (1987). That, however, tended not to convince antitrust critics that non-contestable markets would experience significant entry protection, and so should be the focus of stricter antitrust enforcement. More generally, given that conservatives thought entry would ordinarily be likely and effective, one might have expected them to support antirust challenge to acquisitions of potential competitors, but indeed they did not. Conservatives tend to be harshly critical, for example, of the Supreme Court’s decisions in United States v. Falstaff Brewing Corp., 410 U.S. 526 (1973) and F.T.C. v. Procter & Gamble Co., 386 U.S. 568 (1967), while strongly supportive of United States v. Marine Bancorporation, Inc., 418 U.S. 602 (1974). See also Brodley, supra note 102, at 852–58 (noting that the former airline regulator, the Civil Aeronautics Board, approved several horizontal mergers on its view of easy entry, but it also approved acquisitions of likely entrants, sometimes over arguments from its own staff and the Justice Department, because contestability theory convinced it that potential entrants would not have entered anyway).

These theoretical propositions remain very influential. They have been so fully absorbed that even now some advocates find “simply no general theoretical relationship between market concentration and price,” at least where products are differentiated, and differentiated products are found in most markets in the economy.104Douglas H. Ginsburg & Joshua D. Wright, Philadelphia National Bank: Bad Economics, Bad Law, Good Riddance, 80 Antitrust L.J. 377, 385 (2015).

And yet, as with Williamson’s model, critiques of SCP themselves came under doubt. The original argument that concentration is profitable because it is efficient implied certain logical tensions.105See Salinger, supra, note 95 at 288, 293–94. The contestability thesis too has been rendered all but irrelevant to real-world cases. It remains theoretically significant on some level of abstraction, but observers realized that the conditions for strong contestability were oddly specific and implausible, and that outcomes were sensitive to them.106Above all, some costs of entry usually are sunk, and even small sunk costs could defeat contestability. It should also be unusual that consumer switching is easy but incumbent prices are sticky. There were other problems as well. See generally William G. Shepherd, Potential Competition Versus Actual Competition, 42 Admin. L. Rev. 5 (1990) (discussing these problems); Stiglitz, supra note 103, at 888 (same). The empirical evidence was against it,107See generally Steven A. Morrison & Clifford Winston, Empirical Implications and Tests of the Contestability Hypothesis, 30 J. L. & Econ. 53 (1987) (summarizing existing evidence, and presenting new econometric results inconsistent with perfect contestability); Stephen Martin, The Theory of Contestable Markets (2000) (working paper), https://www.researchgate.net/publication/228432634_The_theory_of_contestable_markets [https://perma.cc/W339-BDNM]. and even its progenitors came to acknowledge that it had been overstated.108William J. Baumol & Daniel G. Swanson, The New Economy and Ubiquitous Competitive Price Discrimination: Identifying Defensible Criteria of Market Power, 70 Antitrust L. J. 661, 678–79 (2003).

Meanwhile, the theoretical case that concentration does correlate with competitive injury has grown much stronger, and most conservative critics acknowledge it.109For accessible explanations of a commonly accepted model applicable to common circumstances, see Jonathan B. Baker, Market Concentration in the Antitrust Analysis of Horizontal Mergers, in 4 Antitrust L. and Econs. 234 (Keith Hylton, ed. 2010); John Kwoka, The Structural Presumption and the Safe Harbor in Merger Review: False Positives or Unwarranted Concerns?, 81 Antitrust L. J. 837 (2017). That theoretical literature poses its own problems—it is large and complex, and it often generates indeterminate outcomes and copes with some theoretical challenges.110Within any market, putting together the collection of firms through merger that might be needed for market power is thought to pose a hold-out problem, just like building a lasting and disciplined cartel. If some competitors in one’s market merge to raise price, it may be profitable to remain independent and expand output at some lower price. See Stigler, supra note 99, at 46; Volker Nocke, Mergers, Endogenous, The New Palgrave Dictionary of Economics (2016). Separately, merger under some peculiar circumstances could reduce profits. See Stephen W. Salant, Sheldon Switzer & Robert J. Reynolds, Losses from Horizontal Merger: The Effects of an Exogenous Change in Industry Structure on Cournot-Nash Equilibrium, 98 Q. J. Econ. 185 (1983). And yet it is now widely accepted that concentration will diminish competition under common circumstances. Finally, the most important rebuttal to the conservative attack on SCP has been a return to empirical study of industry structure. The evidence that concentration correlates with market power has grown significant enough that the original SCP movement’s core results—so soundly rejected by the academic consensus—have come to seem plausible again.111See Baker, supra note 12, at 13–18 & nn.30–35; Jonathan B. Baker, Market Power in the U.S. Economy Today 2–3 & nn.8-12 (March 2017), https://equitablegrowth.org/market-power-in-the-u-s-economy-today/ [perma.cc/6KC5-MG68].

And yet, as with Williamson’s now much-contested efficiency argument, the influence on the law of the SCP critique has not caught up with the academic state of play.

C. The Market for Corporate Control

Finally, a theoretically distinct justification for mergers has been that acquisitions themselves occur within a market—the “market for corporate control”—that is desirable and should be encouraged.112The model is most closely associated with a seminal 1965 journal article by Henry Manne. Henry G. Manne, Mergers and the Market for Corporate Control, 73 J. Pol. Econ. 110 (1965). The idea first appeared in Marris, supra note 8. Within that market, acquirors take advantage of undervalued companies because their management has run them poorly. That process allocates ownership of business assets efficiently, the argument goes, and more effectively regulates management than fiduciary litigation or SEC enforcement.113Manne, supra note 112, at 113. Advocates of the theory argue that takeovers best serve these goals only if they are left free from regulatory constraints.

This justification has been a prominent and explicit influence throughout merger policy. It surfaced first in congressional consideration of the Williams Act, where its advocates persuaded Congress to drop the strong anti-takeover bias of the law’s initial proponents.114Manne himself wrote scathingly of the bill, Henry G. Manne, Cash Tender Offers
for Shares—A Reply to Chairman Cohen
, 1967 Duke L. J. 231 (1967), and decried Congress’s failure to
call him for testimony, Interview with Henry Manne Conducted on August 6, 2012 by James Stocker ,
Secs. and Exch. Comm’n Hist. Soc’y, (Aug. 2012), http://3197d6d14b5f19f2f440-5e13d29c4c016cf96cbbfd197c579b45.r81.cf1.rackcdn.com/collection/oral-histories/20120806_Manne_Henry_T.pdf [https://perma.cc/2LBX-LZJ7]. Additional pressure from the SEC and others was apparently persuasive, see Manuel F. Cohen, A Note on Takeover Bids and Corporate Purchases of Stock, 22 Bus. Law. 149 (1966) (article by SEC Chair supporting the bill but noting a variety of changes the Commission successfully urged to put target and tender offeror management on “equal footing”); Full Disclosure of Corporate Equity Ownership of Securities under the Securities Exchange Act of 1934: Hearings on S. 510 Before the Subcomm. on Banking and Com. of the S. Comm. on Banking and Currency, 90th Cong. 16–17, 32, 182 (1967) [hereinafter “Senate Hearings”] (statement of Manuel F. Cohen, Chair, Securities and Exchange Commission) (emphasizing that federal law should be neutral as between tender offerors and targets, and should neither encourage nor discourage tender offers); Memorandum of the Securities and Exchange Commission to the S. Comm. on Banking and Currency, on S. 2731, 89th Cong., 112 Cong. Rec. at 19,003 (1966); Senate Hearings, supra note 114, at 114–28 (testimony of various academics, who were predominately opposed to restrictions on tender offers, based in large part on corporate-control theory), and Congress ultimately adopted a policy of formal neutrality, see, e.g., S. Rep. No. 550, at 3–4 (1967) (advising that “[hostile] takeover bids should not be discouraged because they serve a useful purpose in providing a check on entrenched but inefficient management.”); H.R. Rep. No. 1711, at 3–4 (1968).
The idea then found its way into antitrust law during the 1980s, 115The theory featured in the legislative history of the Hart-Scott-Rodino pre-merger review reform, see H.R. Rep. No. 94-1373, at 12 (1976), and it has played occasional roles in antitrust ever since. When the theory pops up now in antitrust enforcement, it is usually in connection with § 1 challenge to collusion among takeover bidders. Early cases said that, in deference to efficient takeover concerns, antitrust did not apply, see, e.g., Finnegan v. Campeau Corp., 915 F.2d 824, 825 (2d Cir. 1990); Kalmanovitz v. G. Heileman Brewing Co., 769 F.2d 152, 155–58 (3d Cir. 1985), but later cases suggest the early ones have been overruled by Credit Suisse (USA) Sec., LLC v. Billing, 551 U.S. 264 (2007), see, e.g., Penn. Ave. Funds v. Borey, 569 F. Supp. 2d 1126, 1130, 1134–35 (W.D. Wash. 2008). See generally Edward B. Rock, Antitrust and the Market for Corporate Control, 77 Calif. L. Rev. 1365 (1989) (discussing the early caselaw and calling for antitrust oversight of conduct among takeover bidders); Thomas A. Piraino, Jr., The Antitrust Implications of “Going Private” and Other Changes of Corporate Control, 49 B.C. L. Rev. 971 (2008) (same). and the Reagan administration prominently deployed it in its antitrust,116The theory was given a prominent place in the Reagan administration’s first major antitrust gesture, the 1982 Merger Guidelines revision, see U.S. Dep’t of Just., Merger Guidelines § 1 (1982), and remained part of its long-running program to limit antitrust generally, see Merger Law Reform: Hearings on S. 2022 and S. 2160 Before the S. Comm. on the Judiciary, 99th Cong., 26–29 (1986) (statement of Edwin Meese, III, Att’y Gen. of the United States); Oversight Hearings on the Antitrust Division of the Department of Justice Before the Subcomm. on Monopolies and Com. L. of the H. Comm. on the Judiciary, 97th Cong. 6 (1983) (statement of William Baxter, Assistant Att’y Gen. for Antitrust) (“There are a variety of good reasons why horizontal mergers occur. Horizontal mergers can represent an important discipline of the capital market, upon poor management of a particular company through the takeover phenomenon.”).

corporate and securities, 117Namely, it was the foundation of the administration’s challenge to state antitakeover statutes as unconstitutional. See Johnson & Millon, Misreading the Williams Act, supra note 43, at 1882 (detailing the amicus briefing work of the Reagan SEC). and general economic policy.118See, e.g., H.R. Doc. No. 99-19, 187–216 (1985). At the same time, it would become a driving spirit of the Delaware judiciary’s important fiduciary innovations of the 1980s, which were designed to protect desirable takeovers from the machinations of entrenched management.119While the Delaware courts never went so far as strong corporate control proponents wanted, see, e.g., Frank H. Easterbrook & Daniel R. Fischel, The Proper Role of a Target’s Management in Responding to a Tender Offer, 94 Harv. L. Rev. 1161 (1981); Ronald J. Gilson, A Structural Approach to Corporations: The Case Against Defensive Tactics in Tender Offers, 33 Stan L. Rev. 819 (1981), the courts relied explicitly on those proponents’ academic work. See, e.g., Unocal Corp. v. Mesa Petroleum Co., 493 A.2d 946, 954–56 & nn.9–11 (Del. 1985) (citing Easterbrook & Fischel, supra note 119).The theory is also used in corporate law more generally to justify greater deference to corporate fiduciaries under the business judgment rule. See, e.g., Cuker v. Mikalauskas, 692 A.2d 1042, 1046 (Pa. 1997) (quoting Rosenfeld v. Metals Selling Corp., 643 A.2d 1253, 1262 (Conn. 1994)) (“Not only do businessmen know more about business than judges do, but competition in the product and labor markets and in the market for corporate control provides sufficient punishment for businessmen who commit more than their share of business mistakes.”). And finally, the theory played a role in a brief, strange episode within the Supreme Court, where it narrowly avoided adoption as a doctrine of federal constitutional law.120In a pair of closely watched cases in the 1980s, considering the wave of “anti-takeover” statutes adopted by several states, the Court seemed to imply that sometimes the Commerce Clause is offended by state action restraining the market for corporate control. First, in 1982, Edgar v. MITE Corp., 457 U.S. 624 (1982), held an Illinois anti-takeover statute in violation of the dormant Commerce Clause, explicitly noting that its consequences were to disrupt the market for corporate control, and relying for that view on leading theoretical scholarship on point. Id. at 643–44. But just a few years later the Court seemed to limit or perhaps undo MITE Corp. entirely, in CTS Corp. v. Dynamics Corp. of Am., 481 U.S. 69 (1987). CTS Corp upheld a somewhat less restrictive Indiana anti-takeover statute. The CTS Corp. majority, recalling Justice Holmes’ famous dissent in Lochner v. New York, 198 U.S. 45, 65 (1905), noted that “[t]he Constitution does not require the States to subscribe to any particular economic theory,” and held that merely “decreas[ing] the number of successful tender offers . . . would not offend the Commerce Clause.” CTS Corp., 481 U.S. at 90–93. CTS Corp. moreover reversed an opinion by Judge Posner elaborately founded on market-for-corporate-control theory. That said, CTS Corp. ultimately ruled only on the less restrictive nature of the Indiana statute, and lower courts have continued to find anti-takeover statutes unconstitutional if they “significantly deter nationwide tender offers . . . and impede the market for corporate control.”
Tyson Foods, Inc. v McReynolds, 865 F.2d 99, 103 (6th Cir. 1989).

Corporate control theory, like the other theories already discussed, also invited severe attack. As will be seen, the evidence now seems to corroborate a simple logic problem with the theory, with which it was confronted early on. Holdout shareholders should frustrate efficient takeovers, at least where the shares are dispersed among small shareholders,121Sanford J. Grossman & Oliver D. Hart, Takeover Bids, the Free-Rider Problem, and the Theory of the Corporation, 11 Bell J. Econ. 42, 44–45 (1980). To be clear, while Grossman and Hart built this influential paper around the holdout or free-riding problem, they meant ultimately to defend market-for-control theory. Others have taken the paper to support permissive takeover policy—because, for example, relaxing Williams Act disclosure requirements would allow raiders to act in secret and avoid free-riding minority-shareholder holdouts. See, e.g., Jensen, The Takeover Controversy, supra note 46, at 347. Grossman and Hart’s major theoretical conclusion is that founding shareholders will balance the benefit of encouraging takeover by sacrificing takeover premiums (which disciplines current managers) against the benefit of retaining high takeover premiums when the threat of takeover is not a very effective constraint on management. Knowing which benefit is larger in a given case requires an elaborate analysis, even under simplifying assumptions. See Grossman & Hart, supra note 121, at 47–53.At length, however, the evidence has better corroborated the simple statement of the underlying collective action puzzle with which they began their essay: in the absence of ex ante commitments to encourage them, efficiency-enhancing takeovers should routinely be frustrated by free-riding minority shareholders, especially where the shareholders and the raider can both accurately estimate whatever value the raider can bring to the firm. The evidence corroborates it because, in fact, target-firm shareholders do on average confiscate all the gains available from change in control. and the evidence suggests that does, in fact, occur.122Because the empirical evidence has failed to corroborate the expectation that mergers enhance efficiency, as explained in Part III below, a likely explanation for the strikingly high premiums paid to target-firm shareholders is just that a takeover corners the market in a scarce commodity. Knowing that, shareholders can hold out and inflate the takeover price. So the motive for takeover seems likely to be something other than rationally anticipated gains from superior management or reorganization. Separately, unfriendly takeovers are very costly and uncertain, and should be justified only by severe mismanagement at target firms. But as we shall see, the evidence is that takeover targets tend to be in better health than their peers, especially during merger waves.123See infra notes 135–36 and accompanying text. Perhaps most significantly, merger waves turn out to correlate with stock market over-valuations, contrary to what might be expected were there an efficient market for corporate control.124See infra notes 148-51 and accompanying text.

While some still cling to the idea as a justification of merger activity, its foundations are theoretically and empirically very shaky. And yet, as with each of the other traditional theoretical justifications, this theory still holds strong sway. It is still cited in support of merger activity and restrained antitrust enforcement, and indeed it remains a primary argument for corporate law scholars who hope to discipline managers without judicial or legal interference.

III. In Fact, They Appear To Do Little Good at All

All that background finally brings us to the heart of the Article, which is a survey of the empirical evidence that has looked for the gains predicted by the theoretical tradition, and for which the law observes so much caution.

The empirical study of mergers and acquisitions is of longstanding interest. Whether or not consolidation was desirable was a raging preoccupation of lawyers and economists for some decades around the turn of the 20th century.125See generally Charles J. Bullock, Trust Literature: A Survey and Criticism, 15 Q. J. Econ. 167 (1901) (surveying evidence and academic literature). Like most of the more sophisticated research that would follow years later, those early studies found that consolidation generated little or no benefit at all.126See Thomas F. Hogarty, Profits From Merger: The Evidence of Fifty Years, 44 St. John’s L. Rev. 378, 378–83 (1970) (reviewing four studies that measured merger profitability in the early 20th century, and two more from mid-century; finding that they were not generally profitable or successful). As Hogarty observes, while all the early studies suffered serious flaws, it is striking that in fifty years of study, no one was able to find systematic gains from merger. Id. at 389. Interest quieted for some time, however, and though there were a few subsequent studies,127See, e.g., Shaw Livermore, The Success of Industrial Mergers, 50 Q. J. Econ. 68 (1935) (one of the few subsequent studies). as late as 1955 a famous survey complained that most of this work remained divorced from both theory and data.128Jesse W. Markham, Survey of the Evidence and Findings on Mergers, Bus. Concentration and Price Pol’y 141, 143–146 (NBER 1955) (surveying literature and making this complaint).

Since then, the empirical state of play has advanced dramatically. As it stands, a large body of research, much of it dating from mid-century but still relied upon by some researchers and policymakers, purported to show substantial gains from merger based on stock market reaction to merger announcement. Ranged against it, however, is another now-dominating body of more modern evidence that seriously undermines the earlier stock market studies, measuring merger consequences directly and finding them to generate no systematic benefits at all.

A. The Event Studies

Modern merger study mainly began in the mid-1970s with work by academic financial economists, often explicitly designed to substantiate the market for corporate control hypothesis.129See, e.g., Jensen, The Takeover Controversy, supra note 46, at 314 (vigorous, extended argument of event-study progenitor that, as of 1988, existing evidence still supported corporate-control hypothesis). Their work employed the “event study” approach, in which stock returns130A stock “return” is the amount that a shareholder earns from holding a share of stock over some period of time, measured by both dividends and increase in value of the stock. Alan Hughes, Dennis Mueller & Ajit Singh, Hypotheses About Mergers, in The Determinants and Effects of Mergers: An Int’l Comparison 27, 51–52 (Dennis C. Mueller, ed. 1980). for both the acquired and acquiring firms were measured over a brief period surrounding the acquisition. If returns improved, it was thought, it would indicate the market’s judgment that the takeover was a good idea. Implicit is a strong commitment to capital market efficiency. The idea is that capital markets effectively integrate relevant information into security prices and can very quickly measure the consequences of ongoing events for firm value.131See generally A. Craig MacKinlay, Event Studies in Economics and Finance, 35 J. Econ. Lit. 13 (1997) (summarizing history and reasoning of event-study approach).A perspective widely held among corporate law academics and policymakers is that public securities markets are highly “informationally efficient,” meaning that securities prices very quickly adjust to available information. If they do, then current trading price should already fully incorporate the value of assets and future income stream. And yet, publicly traded firms sometimes trade at share prices below the market value of their assets. Likewise, the very fact of trading in securities itself implies that stock market participants value shares of a given company differently, but the value of a company’s share price assigned by potential buyers of stock should not rationally differ if markets are highly efficient. See generally Reinier Kraakman, Taking Discounts Seriously: The Implications of Discounted Share Prices as an Acquisition Motive, 88 Colum. L. Rev. 891, 891 (1988); Lynn A. Stout, Are Takeover Premiums Really Premiums? Market Price, Fair Value, and Corporate Law, 99 Yale L. J. 1235 (1990). For definitive explanations of the efficient capital markets claim, and leading syntheses of the literature, see Ronald J. Gilson & Reinier H. Kraakman, The Mechanisms of Market Efficiency, 70 Va. L. Rev. 549 (1984); Eugene Fama, Efficient Capital Markets: A Review of Theory and Empirical Work, 25 J. Fin. 383 (1970). If that is true, then stock returns should accurately and quickly reflect gains from merger. Dozens of event studies emerged, eventually synthesized in several extensive surveys, and the studies found on the whole that the shareholders of merging firms earned large, statistically significant gains.132Most famously, Michael C. Jensen & Richard S. Ruback, The Market for Corporate Control: The Scientific Evidence, 11 J. Fin. Econ. 5 (1983). There were some tantalizing subsidiary findings as well. For example, some scholars identified evidence that takeover announcements produce target-firm gains even when a buyer drops a takeover attempt. This was taken to show that if someone wanted to buy a firm, they must have discovered non-public evidence that the firm was undervalued or that buying it promised untapped synergies.133See generally Michael Bradley, Interfirm Tender Offers and the Market for Corporate Control, 53 J. Bus. 345 (1980); Peter Dodd & Richard Ruback, Tender Offers and Stockholder Returns: An Empirical Analysis, 5 J. Fin. Econ. 351 (1977); Michael Firth, Takeovers, Shareholder Returns and the Theory of the Firm, 94 Q. J. Econ. 235 (1980). The researchers also found evidence they thought was inconsistent with market power as the motive for merger. They took it as further proof that the likely motive was efficiency.134Jensen & Rubeck, supra note 132, at 24 (citing prior event studies that measured the effect of horizontal merger on stock returns of the merging firms’ rivals and finding no increase in those returns; arguing that the merged entity therefore did not gain market power and rather gained efficiency). Some other empirical analyses have corroborated the event study literature in one way or another, such as evidence that governance structures that deter hostile takeover correlate negatively with firm performance.135See Lucian Bebchuk, Alma Cohen & Allen Ferrell, What Matters in Corporate Governance?, 22 Rev. Fin. Studs. 783, 788 (2009) (finding that a short list of governance provisions for incumbent “entrenchment,” including some that likely deter unfriendly takeover, correlate with lower Tobin’s Q).

But the event study literature always suffered certain problems, and a few serious ones cropped up early on. Its results could be sensitive to minor differences in experimental design and data selection.136See Fisher & Lande, supra note 74, at 1618–19 (discussing a series of studies finding methodological problems). It also posed a problem of case-by-case unpredictability that would frustrate use of the evidence in policymaking.137Fisher and Lande showed through study of specific cases that while many mergers succeed and many fail, they could not predict on their review which would be which. Their point was that even if some mergers succeed and should be encouraged, to make that fact part of antitrust law on a case-by-case basis, enforcers would need some way to predict which cases would succeed. Since they found no greater likelihood that mergers would succeed either when theory would predict success or when theory would predict greater competitive risk, they believed that inclusion of efficiency on a case-by-case basis would do little good. See id. at 1619–24. Their study was only anecdotal, but it has not been refuted by any more rigorous empirical demonstration of when efficiencies can be predicted. But more to the point of this Article, even at its peak, the literature’s proponents had trouble explaining the gains they discovered.138See Jensen & Ruback, supra note 132, at 25 (acknowledging, in a meta study otherwise claiming significant successes, that “identification of the actual source of the gains in takeovers has not yet occurred.”). Among other things, it was known early on that a major prediction of corporate-control theory—that target firms should be undervalued—was often untrue.139For a contemporary formulation of the argument, see Boyan Jovanovic & Peter L. Rousseau, The Q-Theory of Mergers, Am. Econ. Rev. Papers & Proc., May 2002, at 198. On the one hand, acquiring firms tend to have higher valuations than the firms that they buy, and some evidence shows that merged firms enjoy better stock returns when the acquirer is better-valued than the target. But on the other hand, the firms targeted for acquisition also tend to outperform their peers and often have high valuations.140Sources finding both results are reported in id. at 198. Likewise, researchers had difficulty documenting other effects that should be present if mergers really are driven by efficiency. For example, there was no evidence that merger announcements affected the merging firms’ bond prices.141Jensen & Ruback, supra note 132, at 14 n.7 (italics added). While study of the issue had been more limited, there was little evidence that mergers drove any meaningful measure of productivity.142See Scherer, supra note 20, at 337–40 (presenting regression results, finding very weak and mostly insignificant correlations between merger activity and productivity, and noting lack of evidence to the contrary). Any number of other miscellaneous results seemed hard to explain, including this strange and striking result: in a sample of “clean up” deals conducted by tender offer, in which the bidding firm already had majority control of the target, the bidders wound up paying premiums comparable to those in any takeover. That could not be explained by any synergy, special information, or replacement of inefficient management.143Dodd & Ruback, supra note 133. It seems better explained by the problem of opportunistic holdout behavior that has always been a major critique of efficient merger theories.144See supra notes 121-22 and accompanying text.

But the most serious problem of all was that gains by target firm shareholders massively outweighed gains to acquiring firm shareholders. That target firm shareholders enjoyed large, immediate gains from takeover announcements was beyond empirical doubt as early as the mid-1980s.145See, e.g., Richard Caves, Mergers, Takeovers, and Economic Efficiency: Foresight vs. Hindsight, 7 Int’l J. Indus. Org. 151, 152–53 (1989) (summarizing literature and consensus view); Ellen B. Magenheim & Dennis C. Mueller, Are Acquiring-Firm Shareholders Better Off an Acquisition?, in Knights, Raiders, and Targets: The Impact of the Hostile Takeover 171, 171–72 (Louis Lowenstein et al., eds. 1988) (views of skeptics of event-study literature, acknowledging that this result had been demonstrated). But gains for acquiring firm shareholders were hard to find. So far as the event studies could show, the effect on their shares was small in magnitude and statistical significance, and ambiguous in sign.146Magenheim & Mueller, supra note 145, at 171–72; Bernard S. Black, Bidder Overpayment in Takeovers, 41 Stan. L. Rev. 597, 602–05 (1989); Jensen, The Takeover Controversy, supra note 47, at 316. Theorists proposed several explanations for the anomaly and tested them empirically. Several suggested that it could result from bargaining between acquiring and target firms, or from competition among rival bidders for the same target.147Caves, supra note 145, at 154 (noting these arguments in the literature). Still, they had no good explanation why target firms should enjoy such a large advantage.148It might make perfect sense that the gains in a particular acquisition are captured by one side because that side has some negotiating advantage. It could be consistent with the deal being efficiency-enhancing and both privately and socially desirable. Imagine that a selling firm has a particular asset that will generate a particular kind of synergy or efficiency if it combines with another firm, but several competing bidders could each benefit from it in the same way. The bidding firms might vie with each other to acquire the target, and they might negotiate away nearly all the gains of the deal to the target-firm shareholders. But there could also be cases in which one bidding firm wants to make a particular acquisition, and several selling firms could offer the same asset to it that would generate the efficiency. In that case, we should see the selling firms compete away all the gains, and the bidding firm take them. It seems very unlikely that selling firms would systematically enjoy this kind of negotiating advantage so much more often that they take all the gains in deals overall. Unless there is some such reason, this theory does not explain the imbalance in gains from takeover. See id. Others tried to explain the difference by the structure of transactions or the means of payment. For example, many studies have found significant differences in bidding-firm gains depending on whether acquisitions were paid for with cash or equity, and in general that acquirers suffer larger losses when they pay with their own equity.149See Gregor Andrade, Mark Mitchell & Erik Stafford, New Evidence and Perspectives on Mergers, J. Econ. Persp., Spring 2001, at 103 (reviewing literature). Some have explained this fact with a theoretical model under which markets disfavor purchases with equity.150See generally Sara B. Moeller, Frederik P. Schlingemann & René M. Stulz, Do
Shareholders of Acquiring Firms Gain from Acquisitions?
18 (NBER, Working Paper 9523, 2003), http://www.nber.org/papers/w9523 [https://perma.cc/A7C4-E5RW]; cf. Jensen, The Takeover Controversy, supra note 46, at 316 & n.7 (citing Yen-Sheng Huang & Ralph A. Walkling, Target Abnormal Returns Associated with Acquisition Announcements: Payment, Acquisition Form, and Managerial Resistance, 19 J. Fin. Econ. 329 (1987)). The reason equity payments are devalued is thought to be that management must know some non-public reason that their own stock is overvalued, or else they would pay in cash. See Stewart C. Myers & Nicholas S. Majluf, Corporate Financing and Investment Decisions When Firms Have Information That Investors Do Not Have, 13 J. Fin. Econ. 187 (1984).

But in any case, a bigger problem was emerging evidence that shareholder gains might not be permanent for either firm. The original event studies used very short windows, often measuring stock price changes over a few days. As researchers began to expand the event window, they found that a year or so after acquisition, even target-firm gains often dissipated.151See Andrade et al., supra note 149, at 103 (reviewing literature). Later research undercut the tantalizing early finding that unsuccessful takeover attempts caused permanent target-firm gains. In fact, those gains dissipate fairly quickly unless another takeover bid comes through and succeeds.152Michael Bradley, Anand & E. Han Kim, The Rationale Behind Interfirm Tender Offers: Information or Synergy?, 11 J. Fin. Econ. 183, 185–86 (1983). In retrospect, that intriguing result just seems to confirm that the gains to target firm shares merely reflect the anticipation that another takeover is forthcoming.

Another promising finding would be the relation of merger “waves” to systematic stock market overvaluations. Since the beginning of heavy merger activity, observers have remarked that mergers are cyclical, occurring in large, periodic waves, and, to early observers, they seemed to coincide with business cycles.153See, e.g., Markham, supra note 128, at 146 (noting the frequent observation of merger “cycles” and citing examples from as early as 1901); Ralph L. Nelson, Merger Movements in American Industry, 1895–1956 (Princeton Univ. Press. 1959). The “business cycle” is the rise and fall in output of goods and services, usually measured by gross domestic product, adjusted for inflation. While it was discovered early on that mergers don’t actually correlate well with business cycles,154See, e.g., Markham, supra note 128, at 146–52. mergers are correlated with swings in stock prices (which do not necessarily coincide with cycles in the real economy).155See id. at 151–53. More recent confirmations of the relationship include P.A. Geroski, On the Relationship Between Aggregate Merger Activity and the Stock Market, 25 Eur. Econ. Rev. 223 (1984), and Ronald W. Melicher, Johannes Ledolter & Louis J. D’Antonio, A Time Series Analysis of Aggregate Merger Activity, 65 Rev. Econ. & Stat. 423, 423, 429 (1983) (finding merger activity specifically correlated with rising stock prices and bond yields, and not with output in the real economy). See also Gugler et al., supra note 21, at 2 (results of large study of data from United States, United Kingdom, and Continental Europe, finding that merger waves occur almost exclusively among publicly traded firms, implying that they reflect stock market phenomena and not changes in the real economy). Moreover, they tend to correlate specifically with over-valuations.156See generally Devra L. Golbe & Lawrence J. White, Mergers and Acquisitions in the U.S. Economy: An Aggregate and Historical Overview, in Mergers and Acquisitions 25, 42–44 (Alan J. Auerbach, ed. 1987). That in itself seems somewhat counter-intuitive. All things being equal, one might expect more interest in acquisitions when capital markets undervalue firms, a specific prediction of market-for-control theory being that firms would be bought when they were bargains.157See Manne, supra note 112. The evidence is to the contrary. This can be shown with a common valuation measure known as aggregate “Tobin’s q.”158Single-firm Tobin’s q is the ratio of a company’s current market valuation to the replacement value of its assets. In other words, Tobin’s q is the stock market’s estimate of the firm’s going-concern value, the value of its business over and above its value in liquidation. Tobin’s q can also be measured on an aggregate basis, for a whole sector or a whole economy. See Golbe & White, supra note 156, at 40. Where Tobin’s q is high, it implies that the firms in a given sector are systematically overvalued. And it so happens that public-market acquisition activity, and especially merger waves, appear to be correlated with high aggregate Tobin’s q.159Golbe and White, supra note 156, at 42–44. Golbe and White control for the possibility that high stock valuations might be caused by increased merger activity, instead of the other way around. In other words, they were concerned that merger waves correlate with high stock valuations, but not for any reason that would be at odds with market for control theory, and rather because increases in M&A caused by other factors might themselves drive up stock prices. Controlling for that factor, they found that M&A activity itself could not account for the correlation between merger waves and stock market overvaluations. Id. at 44. Not only that, but merger waves have almost always ended with precipitous declines in equity prices.160Matthew Rhodes-Kropf, David T. Robinson & S. Viswanathan, Valuation Waves and Merger Activity: The Empirical Evidence, 77 J. Fin. Econ. 561, 562 (2005). This has come to be a fairly robust finding that merger waves coincide with stock market bubbles. That result has not yet been very well explained, but it is in substantial tension with any theory of merger efficiency.

During the 1980s, various theorists, including the event-study proponents themselves, began to question whether target-firm shareholder gains really reflect any social value.161Cf. Jensen & Ruback, supra note 132, at 42 (“We are . . . reaching the point of rapidly diminishing returns from efforts that focus solely on stock price effects.”). See generally Caves, supra note 145 (summarizing literature as of 1989). The event studies might merely have measured the likelihood of control-premium buyouts, and that whatever explains those premiums, it was not the rational expectation of durable efficiencies. As early as 1988, one of the progenitors of the event-study effort, Richard Ruback of the Harvard Business School, wrote “[r]eluctantly” that “we have to accept this result—significant negative returns [over the longer term] following a merger—as a fact.”162Richard S. Ruback, Comment, in Corporate Takeovers: Causes and Consequences 260, 262 (Alan J. Auerbach, ed. 1988). For some, the confusion cast doubt on capital market efficiency more generally, and especially the extravagant assumption that stock prices could really impound information as well and quickly as event-study proponents had thought.163For criticism of the time that stock-return event studies were unreliable, see, e.g., Edward S. Herman & Louis Lowenstein, The Efficiency Effects of Hostile Takeovers, in Knights, Raiders, and Targets: The Impact of the Hostile Takeover 211 (Louis Lowenstein, Susan Rose-Ackerman & John C. Coffee, eds. 1988). As for skepticism of efficient capital markets theory more generally, there was already evidence at the time that short-term stock price changes might be highly unreliable estimates of actual future changes in corporate performance. See, e.g., Robert J. Shiller, Do Stock Prices Move Too Much to be Justified by Subsequent Changes in Dividends?, 71 Am. Econ. Rev. 421 (1981) (finding swings in stock price as much as five times higher than should have occurred, given actual changes in dividend streams). That emerging evidence was still quite controversial, see, e.g., Jensen, The Takeover Controversy supra note 46, at 320 & n.17 (acknowledging Shiller’s evidence and other disputes, but claiming that, as of 1988, “there is no better documented proposition in the social sciences” than the efficient markets hypothesis), but even then, it could not be ignored.Since then, other evidence has emerged to cast short-term stock-return studies in serious doubt. For example, evidence was found that stockholder returns on takeover announcements during the 1960s conglomerate wave were substantially better for diversifying acquisitions than for same-industry deals, see Andrei Schleifer & Robert W. Vishny, Takeovers in the ‘60s and the ‘80s: Evidence and Implications, 12 Strategic Mgt. J. 51, 52 (1991) (discussing John G. Matsusaka, Takeover Motives During the Conglomerate Merger Wave, 24 Rand J. Econ. 357 (1993)), but on further analysis the conglomerate merger wave has proven one of the worst wrong turns in American business history. See generally Ravenscraft & Scherer, supra note 6; Porter, supra note 6.

B. Direct Study of Merger Effects

Meanwhile, other approaches to studying mergers were gaining ground. The most significant development in the empirical effort was the growth of new methods of research design that gained prominence while the event-study literature was beginning to falter. The newer studies—which had previously been rare because they are demanding—used accounting and engineering data for merging firms, rather than stock market performance, measured much longer time periods and controlled for effects across industries (and sometimes over whole economies).164Accounting-data studies are more difficult because the available data are usually too highly aggregated. These studies attempt to relate performance measures like profit or growth to a firm’s having merged or not. But many factors would surely contribute to those measures, including tax benefits, accounting choices, and the effect of control premia, that are unrelated to the substantive costs and benefits of mergers as such. To measure those extraneous effects directly, researchers need firm-level data (which is usually confidential), or some other way to measure internal performance. As late as the 1980s, observers could report that because of these data and methodology problems the accounting-data literature remained “quite mixed” and inconclusive. Fisher & Lande, supra note 74, at 1610–11. It has grown substantially since then, however. That literature began in part with a large, seven-country study in 1980 that found very little evidence of benefits from consolidation.165That study was Determinants and Effects of Mergers, supra note 130.

Research in this newer vein discovered certain significant peculiarities early on. For example, even though U.S. firm sizes had increased over the twentieth century, to some large degree as the result of merger and acquisition, plant sizes stayed the same.166Alan Hughes & Ajit Singh, Mergers, Concentration, and Competition in Advanced Capitalist Economies: An International Perspective, in Determinants and Effects of Mergers, supra note 130, at 1, 15–17. That is, firms seemed to be growing not by consolidating and synthesizing their production processes, as one might expect if their goal were technological or operating efficiency. By appearances, they were just gathering free-standing businesses and putting them under common ownership. Likewise, it seemed significant that when U.S. merger enforcement became vigorous in the 1960s and aimed its attack mainly at horizontal and vertical deals, managers turned to conglomerate mergers. Thus ensued the first and only conglomerate merger wave in history, and it was truly a massive wave.167Many have argued that the conglomerate merger wave was driven by antitrust in this way, because vigorous enforcement against horizontal and vertical deals left merger-hungry managers only conglomerate deals as their option. See, e.g., Frederick M. Rowe, The Decline of Antitrust and the Delusions of Models: The Faustian Pact of Law and Economics, 72 Geo. L. J. 1511, 1527 (1984); George J. Stigler, The Economic Effects of the Antitrust Laws, 9 J. L. & Econ. 225, 232–33 (1966) (presenting evidence that horizontal mergers declined substantially during period of strict enforcement, while conglomerate mergers grew); Melicher, et al., supra note 155, at 423 (acknowledging this relationship); Shleifer & Vishny, supra note 163, at 52. If conglomerate deals are just as good as horizontal or vertical ones, even though they differ substantially in their economic substance, then perhaps managers are not doing deals because consolidation itself is synergistic or desirable. They might just do deals because, for whatever reasons, they want to do deals.

But at length, one thing these dozens of papers have not found is systematic evidence either of lasting shareholder gain or increases in the performance in merged firms. As is now shown in a large number of papers and several surveys, merged firms actually appear to lose value in the longer-term following merger, or at best, see no systematic gain.168See, e.g., David R. King, Dan R. Dalton, Catherine M. Daily & Jeffrey G. Covin, Meta-Analyses of Post-Acquisition Performance: Indications of Unidentified Moderators, 25 Strat. Mgmt. J. 187 (2004) (summarizing consensus and presenting meta-study supporting it); Deepak K. Datta, George E. Pinches & V. K. Narayanan, Factors Influencing Wealth Creation from Mergers and Acquisitions: A Meta-Analysis, 13 Strat. Mgmt. J. 67 (1992) (summarizing consensus and presenting meta-study supporting it); Louis Kaplow & Carl Shapiro, Antitrust, in 2 Handbook Of Law And Economics 1073, 1154 (A. Mitchell Polinsky & Steven Shavell eds., 2007) (summarizing evidence that acquiring firms do not benefit from mergers on average); Röller et al., supra note 77, ch. 3 §2.1.3 (concluding from a review of economic studies that “mergers have but modest average effects on the profitability of the merging firms” and “a large proportion of mergers reduces profitability”). The large, one-time gains to target-firm shareholders cannot measure the social value of acquisitions. At most, they can only reflect inframarginal shareholders’ different estimates of future earnings under uncertainty.169See Stout, supra note 131, at 1269.

Certain studies in this tradition have come to be particularly large and striking. One analyzed all friendly acquisitions between exchange-listed firms over a thirty-year period. It found that shareholders in acquiring firms suffered a statistically significant loss of about 10% over the five-year post-merger period.170Anup Agrawal, Jeffrey F. Jaffe & Gershon N. Mandelker, The Post-Merger Performance of Acquiring Firms: A Re-Examination of an Anomaly, 47 J. Fin. 1605, 1606 (1992). Another study, on a very large set of finely disaggregated data and deploying certain important methodological improvements, found that U.S. manufacturing mergers led to large and statistically significant price increases in the merging firms’ product markets, without evidence of efficiency gains.171Bruce A. Blonigen & Justin R. Pierce, Evidence for the Effects of Mergers on Market Power
and Efficiency
Fed. Res. Bd. Fin. & Econ. Discussion Series No. 2016-082 (2016), available at https://doi.org/10.17016/FEDS.2016.082 [https://perma.cc/HHX5-E5SH]. Blonigen and Pierce use plant level data for all firms, both public and private, in the U.S. manufacturing sector. They use difference-in-difference tools to measure change in productivity and mark-ups for mergers among them between 1997 and 2007. Because the study is done at the plant level, there is no need for market definition and the problem of using SIC codes does not arise. They find large, statistically significant increases in average mark-ups—between 15 and 50% larger than control firms—but no statistically significant change in productivity or any other efficiency measure.
Some important findings of other kinds have also emerged. For example, while the effort to relate merger activity to broader economic trends is challenging and controversial, one study measuring a large dataset over a long period found that even the biggest merger waves had no effect on overall economic productivity, even when measured for periods as long as three years.172Scherer, supra note 20, at 338–40.

It has not been fully explained why mergers produce such disappointing results, but there are likely answers. First of all, the technological or other efficiencies that could plausibly generate social gains are not that likely, except in small deals between small firms. The empirical evidence that exists suggests that minimum efficient scale173“Minimum efficient scale” (MES) is the point of output at which average costs stop falling. Technically, MES is just the point at which the marginal cost curve intersects the average total cost curve, because by definition that is the lowest possible point of ATC. That in itself, however, tells us nothing about the actual shape of real-world cost curves. Most evidence suggests that average cost curves tend to be L-shaped—initially falling and then flattening out to become constant—or U- or “bathtub” shaped, with a flat range of constant returns at the bottom, between the initial fall in costs and a later point at which they rise again. There is therefore a point of output early in the curve—known as MES—beyond which further cost savings cannot be obtained solely by expanding output.Bain pioneered the concept. See Joe S. Bain, Economies of Scale, Concentration, and the Condition of Entry in Twenty Manufacturing Industries, 44 Am. Econ. Rev. 15 (1954). Over time, a belief in constant returns after some initial range of falling average costs became a rough empirical consensus. See, e.g., Silbertson, supra. note 71, at 369, 376. in most U.S. markets is probably fairly small.174See generally Frederic M. Scherer & David Ross, Industrial Market Structure and Economic Performance, 97–151 (3d ed. 1990); cf. Stigler, supra note 17, 26–27 (“casual observation suggests that the economies of scale are unimportant over a wide range of sizes in most American industries, for we commonly find both small and large firms persisting.”). The question is when and how they can occur, and whether they are likely to follow from the large and heavily concentrating deals that are usually of policy concern.Direct study of efficiencies is difficult. Studies usually review laboriously collected engineering data on a product-by-product basis or use statistical proxies that have been criticized. See Stephen Davies, Minimum Efficient Size and Seller Concentration: An Empirical Problem, 28 J. of Indus. Econ. 287, 290–91 (1980) (explaining that leading proxies for MES, being based on observed average plant and firm sizes, likely were really just estimates of concentration, and so when used as econometric explanatory variables they simply regressed one concentration measure on another, and so found strong correlations); George J. Stigler, The Economies of Scale, 1 J. L. & Econ. 54, 54 (1958) (early study noting the lack of evidence and the methodological difficulty; introducing the “survivor” method but acknowledging its weaknesses). But see Leonard W. Weiss, The Survival Technique and the Extent of Suboptimal Capacity, 72 J. Pol. Econ. 246, 248–49 (1964) (presenting survival-technique estimates of MES for several industries previously tested with engineering estimates in Bain, supra note 174, and finding close agreement). Interestingly, Davies’ critical review of 1980 suggests that the leading MES proxies probably overstate MES. Davies, supra note 82, at 292–93.
There are some exceptions. Bain, for example, found that there seemed to be substantial scale returns in advertising and other promotional efforts in some differentiated consumer products, even at large scale, though he suggested several reasons that the results might be misleading or might actually reflect first-mover or anticompetitive advantages. Joe S. Bain, Advantages of the Large Firm: Production, Distribution, and Sales Promotion, 20 J. of Mktg. 336 (1956).
The most obvious and easily measurable gains—scale, scope, and technological integration175The important point is that most gains are likely associated with scale, but there is probably some comparatively low level of production beyond which all the available scale gains have been realized. The most obvious gains are productivity from increased specialization and reductions in average cost from increased scale that don’t increase fixed cost. See F.M. Scherer, Economies of Scale and Industrial Concentration, in Industrial Concentration: The New Learning 16, 22 (Harvey J. Goldschmid et al., eds. 1974); Silbertson, supra note 71. Firms also save when vertical consolidation integrates functions that are cheaper to combine in some way. See M.A. Adelman, Integration and Antitrust Policy, 63 Harv. L. Rev. 27 (1949) (giving examples). But as with more familiar scale economies, benefits of scope and vertical integration are directly associated with scale of output. In order to efficiently produce related products (that is, achieve benefits of scope) and link successive stages of production (that is, achieve vertical integration), a firm must be able to do those things without sacrificing simple scale economies at each stage. See Silbertson, supra note 71, at 374.—should mostly be exhausted at comparatively small scale.176In principle, any number of other efficiencies are conceivable, like managerial improvements and economies in marketing and finance. Those benefits tend to be more speculative, and even where they exist, they are probably less important than technical economies, see Silbertson, supra note 72, at 386, and in any case there is no clear reason that merger would be needed to achieve them. Occasionally merger studies have found efficiency gains, even on large samples, but if anything, those exceptions prove the rule. Whether they mean to show it or not, they imply that the gains are restricted to small firms doing small deals. For example, one large study, taken by some to prove substantial efficiencies, really seems to show the opposite. It lends support to the view that efficiencies are real but are usually available only among very small deals.177Ghosh, supra note 71. While Ghosh’s study cannot directly prove this point, it tends to show it indirectly (and, one imagines, unintentionally). Many, many other studies have failed to find merger efficiencies, and they appear to be found only when one does as Ghosh did: he takes average measurements across a whole sample that turns out to be overwhelmingly skewed to tiny deals.The study’s major finding was that acquisitions of all kinds generate increases in market share, and that the market share gains correlate with operating performance. However, the mean market shares of the acquiring and target firms in the sample was about 9.5% and 4%, respectively, and the medians were about 2.5% and 0.6%, implying that the sample skewed heavily toward very small deals among very small firms. One cannot tell if the findings apply across the whole range of the sample, or if they are concentrated at one end or the other of the size distribution of observations. That might be known, for example, if the study were re-run with a sample only of large firms, or if it included a regression variable for firm size. Instead, it regresses performance measures only on a variable for increased market share. See id. Moreover, if, to any degree, efficiencies do explain merger activity, then one would expect mergers to be more common among small firms. They appear not to be. They are more common among large firms, which are less likely to gain efficiency by growth, at least by way of traditional technological integrations.178See Dennis C. Mueller, Mergers: Theory and Evidence, in Mergers, Markets and Public Policy 13 (Giuliano Mussati ed., 1995) (citing papers finding that merging firms tend to be larger than average firms in their industries, implying that they had already reached MES before mergers); see also Michael Gort, An Economic Disturbance Theory of Mergers, 83 Q.J. Econ. 624, 632 (1969).

Moreover, even where efficiencies would be available in principle, combinations typically suffer serious implementation problems. A long and well-documented history of failed implementation appears to reflect both clashes of corporate cultures and diseconomies of large size or scope.179See Fisher & Lande, supra note 74, at 1619–24; see also Christensen et al., supra note 5, at 56–57 (discussing evidence of failed implementation); see also Scott A. Christofferson, Robert S. McNish & Diane L. Sias, Where Mergers Go Wrong, 2 McKinsey Q. 93, 93 (2004) (same); see also Deloitte, supra note 1, at 13–14 (reporting corporate dealmaker survey results that substantially blame implementation factors for underperforming mergers). Scholars in management and economics have struggled to find the determinants of post-acquisition success. They have found that the implementation strategies most likely to succeed are rarely followed.180See, e.g.,, King et al., supra note 168, at 188; Weber et al., supra note 10, at 375. So it may be that however real size-related benefits might be, mergers are a bad way to achieve them; it is just too hard to pull off. Accordingly, it should often be that firms’ incentives to achieve them by internal growth would be as great as to achieve them by merger. The odds are poor that firms will do any better or have any more incentive to obtain efficiencies by the costly and fraught work of consolidation than through internal growth.181This point captures an important value that has long been part of merger policy—that the law should favor internal growth over growth by acquisition. The former is competition but he latter is just tkaing control over something consumers already had. See United States v. Anthem, Inc., 855 F.3d 345, 356 (D.C. Cir. 2017); F.T.C. v. H.J. Heinz Co., 246 F.3d 708, 722 (D.C. Cir. 2001).Admittedly, there could be cases in which some efficiency or synergy requires access to a particular asset that is scarce and difficult to come by. Perhaps in some cases the cheapest or only what to get the asset is by acquisition of a target that already has it. But such cases should be rare. That seems particularly true of gains other than scale improvements, like the need to replace inefficient management or redesign products or marketing.182Obviously enough, as many have pointed out, inefficient management could just be replaced by the target firm’s board or shareholders, without the need for takeover, and the same is true of improvements in product design, marketing, and other operating matters that merger proponents often offer as benefits of their deals. There should be internal means to replace management cheaper and less risky than hostile takeover, though corporate-control advocates most often explain hostile deals as a means to replace bad managers.

Nowadays, even the most sympathetic literature reviews, which attempt to cast mergers in the friendliest light, tend only to claim that shareholders should be indifferent to them. The argument goes that acquiring firm shareholders are usually left with net present returns of roughly zero, so at least they are no worse off than if a merger didn’t happen.183See, e.g., Robert F. Bruner, Does M&A Pay? A Survey of Evidence for the Decision-Maker, J. Applied Fin., Spring/Summer 2002, at 48 (making this argument as evidence that mergers generate “success”). That is not a ringing endorsement, but it may be the best that can be said.

IV. The Remaining Puzzles

So all of this leaves a few serious and not well-answered questions. The biggest is why firms and shareholders would systematically engage in conduct that should seem irrational from both their points of view. Many theories attempt to explain the persistence of acquisitions despite their disappointments, but none yet has decisive support.

A. Why Do They Happen at All?

1. If There’s Nothing to Gain, Why Do Buyers Continue to Buy? A variety of explanations have been offered for why firms would engage in activity that seems not to benefit shareholders. Some can be rejected fairly easily. It was once thought that tax savings might be the real goal, in which case mergers might seem rather benign, at least in terms of effects on shareholders and the competitiveness of markets. That now seems very unlikely, as this argument fails to explain much of the empirical evidence.184Namely, tax savings cannot explain variation in merger rates across industries, or variation over time that doesn’t correlate with tax law changes. See Gort, supra note 178, at 625. Studies have found tax motivations to be important only in a very small range of cases. See Mueller, supra note 178, at 9, 14. Separately, a “bargains theory” of takeover presumed that acquirers would snatch up firms when they were undervalued, and it was a core prediction of corporate-control theory. It seems quite clearly rejected by evidence that target firms tend to be profitable and valued by investors.185See supra note 140 and accompanying text. Likewise, the common suggestion that acquisition might sometimes just be cheaper than internal growth186See, e.g., Gort, supra note 178, at 628; Merger Law Reform: Hearings Before the Comm. on the Judiciary, United States Senate, 99th Cong., 2d Sess. 58 (1986) (statement of Malcolm Baldridge, Secretary, U.S. Department of Commerce). seems at odds with the substantial history of failed implementation and shareholder losses. If growth by acquisition is so difficult to pull off and likely to generate losses, would it not be cheaper to generate the same growth internally? There was some empirical support for the related idea that some efficiencies require acquisitions because some inputs are “lumpy.”187See Caves, supra note 145, at 155. An activity might require some asset that can be gotten only in discrete lumps and cannot be shared between independent firms. An example of an asset like this might be know-how.188See, e.g., Caves, supra note 145, at 155–56 (discussing empirical evidence for this hypothesis). In any case, though—except perhaps where overall demand is significantly contracting—acquisition is probably an uncommonly risky, disruptive, and expensive means of growth or synergy. Whatever might be gained through scale or sharing lumpy assets across broader activities, it is hard to imagine they are so much more expensive to achieve through internal growth that the evidently costly alternative of acquisition is really better.

The strong correlation between merger waves and stock market overvaluations suggests other possible motives, and these perhaps have stronger support. It may be that during bubbles, managers exploit the opportunity to use their own overvalued stock to acquire real assets, essentially arbitraging for their shareholders’ benefit.189See Andrei Shleifer & Robert W. Vishny, Stock Market Driven Acquisitions, 70 J. Fin. Econ. 295, 297–299 (2003) (modeling such an argument). That theory, however, is at odds both with evidence that firms with large shareholders are less likely to make acquisitions190See Gugler et al., supra note 21, at 8–9. and that the percentage of acquisitions financed with cash and debt (as opposed to the acquiror’s own equity) is about the same during waves and during normal times.191See Gugler et al., supra note 21, at 13. A different but closely related explanation, which is analytically elegant and perceptive, is that “economic shocks”—technological innovation or other disruptions that alter expectations of future earnings—increase the range of disagreement over the value of particular firms. As disagreement grows between owners and non-owners of any given firm over its future profitability, the possibility increases that some non-owner will begin to value it sufficiently more highly than its owners so that gain from trade becomes possible.192See Gort, supra note 178, at 626–29. This “economic disturbance” theory enjoys the corroboration that, in fact, merger waves tend to have industry-specific effects and correlate with higher trading volume, which might indicate increasing divergence of opinion among traders over true valuations.193Golbe & White, supra note 156, at 42-43. However, “disturbance” explanations are hardly arguments that disturbances generate correct decisions or accurately anticipate efficiency gains. One party that disagrees with the value of a stock is wrong, and there is no reason to believe it is always the seller. In any event, it appears that disturbances do not drive merger activity in non-wave periods.

The explanations that are left, and that fit the evidence best after these decades of empirical study, depend on simple agency cost. They have the advantage of explaining the evidence without relaxing the rationality assumption, and elsewhere in corporate theory, agency cost arguments are uncontroversial. A corporation is not a unitary entity with a single goal that it maximizes, and those who manage it often hold incentives at odds with their shareholders. So long as they are unconstrained by shareholders or stock price discipline, they can cause mergers that do not benefit their firms, and they might very rationally do so to serve their own interests. Agency explanations now enjoy the substantial validation that merger waves affect publicly traded companies but largely do not affect privately held ones.194See Gugler et al., supra note 21, at 5–7. Likewise, mergers are generally less likely where some stockholder or group of them holds a large share of voting stock.195Gugler et al., supra note 21, at 9–10. This result is inconsistent with efficiency-enhancing views of mergers, because large shareholders should favor efficient mergers. It also casts doubt on the theory that merger waves correspond with stock market bubbles because managers trade their own overvalued stocks for real assets. That would make sense, but if it were the explanation for merger, then managers of firms with large shareholders should do them just as often. These results should not follow if efficiencies generally drive merger and acquisition.

It is unknown why corporate managers should like takeovers so much, and there could be different explanations. Managers, and their lawyers and financiers, might gain individually from increasing firm size, even when the companies do not, either by direct pecuniary benefit or some other, perhaps psychological gain from managing bigger firms.196British economist Robin Marris first argued in 1963 that corporate managers might seek to maximize firm size for its own sake. Robin Marris, The Economic Theory of “Managerial” Capitalism 10 (1964) (introducing the idea in 1963); see also William J. Baumol, Business Behavior, Value, and Growth 44–48 (1967). Obviously, firm size might generate direct pecuniary gains to individuals, like increased salary for acquiring-firm officers or fees earned by investment bankers, but there are other theoretical possibilities. Any number of strong believers in integration and the rationality of mergers have nonetheless acknowledged that managers may serve a desire for growth for its own sake. See, e.g., Jensen, The Takeover Controversy, supra note 46, at 321–22; M.A. Adelman, Integration and Antitrust Policy, 63 Harv. L. Rev. 27, 34 (1949). Even Manne foresaw the possibility, and argued that it would be consistent with a market for corporate control insofar as the larger a firm is, the more transaction cost there will be in acquiring control of it. The greater that additional cost, the higher that inefficient manager salaries could be before a proponent would seek control. See Manne, supra note 112, at 117 n.26. There are any number of related agency-cost suggestions. Common among them is the free cashflow hypothesis, under which managers are reluctant to pay out cashflows as dividends, even when they should, and hoard it or reinvest it inefficiently in internal projects.197See, e.g., Jensen, The Takeover Controversy, supra note 46, at 321–37; Shleifer & Vishny, supra note 163, 52, 55. Somewhat cruder but perfectly plausible ideas include the “hubris hypothesis,” under which managers systematically overestimate their own ability to identify undervalued assets.198See Richard Roll, The Hubris Hypothesis of Corporate Takeovers, 59 J. Bus. 197 (1986). That theory is corroborated by the otherwise unexplained fact that acquiring firms often experience substantial supernormal returns in the few years before they undertake acquisitions.199See Magenheim & Mueller, supra note 145, at 176–88 (setting out new evidence of that finding and summarizing prior studies). Perhaps managers who’ve done well for a while overspend on acquisitions, thinking their judgment is better than it is.

In any case, whatever it is that motivates acquisition activity, it is not a rational or reliable estimate of sustainable efficiency.

2. If Markets Are Efficient, Why Would Shareholders Hold Out? Shareholder behavior in takeovers poses a double-sided logic problem. On the one hand, some theorists have found it puzzling that target-firm shareholders would hold out for premium prices because that behavior is at odds with any strong efficient-markets commitment. On a strong efficiency view, shareholders should not need substantial price increases to be persuaded to sell.200See generally Stout, supra note 131, at 1259–75 (canvassing relevant scholarship, and offering her own interpretation of theory and evidence). Takeover premiums may seem familiar and unremarkable now, but strictly speaking, they are unexpected if public securities markets are as informationally efficient as they are often assumed to be. If they are, prices should already equal the present value of their expected returns, and a rational shareholder should sell to any buyer willing to pay more than current trading price. That problem has puzzled legal scholars and finance economists since the beginning of contemporary corporation theory, See, e.g., Black, supra, note 146; John C. Coffee, Jr., Regulating the Market for Corporate Control: A Critical Assessment of the Tender Offer’s Role in Corporate Governance, 84 Colum. L. Rev. 1145 (1984); Kraakman, supra note 131. It was for many scholars the very rationale for market-for-corporate-control rhetoric: takeover premiums must prove that takeovers generate value, because it would be irrational for any shareholder not to sell at any price marginally above current trading price, see, e.g., Easterbrook & Fischel, supra note 119, at 705. But on the other hand, if takeover proponents are willing to pay high premiums, it implies they believe they can make the stock more valuable than what they’ll pay. So why wouldn’t a target-firm shareholder prefer to keep their stock and share in those gains?

Shareholder behavior, however, seems easy to explain. A takeover, by definition, corners the market in a scarce commodity and inflates its price. The only thing necessary to believe shareholders would rationally hold out for higher prices is the relaxation of the most unrealistic efficient capital markets presumption.201The phenomenon could be explained in different ways. As Stout observed, shareholders vary in their estimates of intrinsic value. As in other markets for goods that consumers value differently, the pre-takeover price will be set by marginal shareholders, who interpret future earnings the most pessimistically. To get a control share, however, takeover buyers must secure some number of inframarginal shareholders, who interpret earnings more positively, and will hold out for a higher price. See Stout, supra note 131, at 1259–75. The phenomenon also might be explained by strategic conduct, as Grossman and Hart famously did in their analysis of market-for-control theory. In public firms, shareholders will hold out and will confiscate all the gains that acquirors could have achieved by their conquest. Grossman & Hart, supra note 121. To whatever extent capital market returns measure anything, it is something other than efficiency gains in the underlying transactions. Capital markets continue to support large takeover premiums despite decades of evidence that takeovers generate no real gains. By now, the markets should have priced out the possibility, and takeover premiums therefore must be measuring something else. It appears the markets are merely pricing takeover proponents’ willingness to spend their own shareholders’ money to pay large premiums. To that extent they merely measure the agency cost imposed by acquiring firm managers.

3. And Why Can’t They Make Money? One final puzzle remains. Whether mergers generate efficiency or benefit shareholders, it is generally thought that at least some mergers create market power. Antitrust law takes it for granted, and both standard economic theory202See supra note 109 and accompanying text. and a recently growing body of empirical evidence suggests it.203See supra note 112 and sources cited. If this is true, it would help explain the prevalence of mergers. But one would assume that increases in market power should also increase profitability. If acquisitions generate monopoly, then someone should at least be able to profit from it.

This final tension has received little attention, but a few likely explanations make it quite plausible that mergers could increase pricing power while failing to increase the firm value or accounting profits. Most importantly, profits from pricing power might just be wasted through “X inefficiency.” The theory predicts that lack of competitive discipline leads to slack and inattention to productive efficiency.204See Harvey Leibenstein, Allocative Efficiency vs. “X-Efficiency, 56 Am. Econ. Rev. 392 (1966) (coining “X-efficiency” and providing its first elaboration); see generally Scherer, supra note 73, at 38; Jean Tirole, The theory of Industrial Organization 46–48 (1988); Bengt Hölmstrom & Jean Tirole, The Theory of the Firm, in 1 Handbook of Industrial Organization 61, 95–97 (Richard Schmalensee & Robert Willig eds., 1989). When firms compete, all managers are pressured to find productivity savings to translate into lower prices. But since non-owner managers might prefer less effort, they do not try as hard when competition lessens. There follows the famous folk wisdom that “[t]he best of all monopoly profits is a quiet life.”205J. R. Hicks, Annual Survey of Economic Theory: The Theory of Monopoly, 3 Econometrica 1, 8 (1935); see also, supra note 73 (quoting Adam Smith’s famous observation to this effect).

In fact, if product-market competitiveness is a determinant of organizational slack, then mergers creating market power should tend to reduce efficiency. They should dissipate monopoly profits which might otherwise appear as shareholder gains. The idea has formal theoretical support,206See Jean-Etienne de Bettignies & Thomas W. Ross, Mergers, Agency Costs, and Social Welfare, 30 J. L. Econ. & Org. 401 (2014); Oliver D. Hart, The Market Mechanism as an Incentive Scheme, 14 Bell J. Econ. 366 (1983). The purely theoretical literature occasionally suggests that competition could sometimes reduce internal efficiencies. See Röller et al., supra note 77, at 96–98 (summarizing studies). So far, however, those claims seem at odds with the empirical evidence. and substantial empirical literature establishes corroborating facts.207For reviews of the empirical literature, see Chad Syverson, What Determines Productivity?, 49 J. Econ. Literature 326, 347–57 (2011), and Matthew Chiasson & Paul A. Johnson, Canada’s (In)Efficiency Defence: Why Section 96 May Do More Harm Than Good for Economic Efficiency and Innovation, 32 Can. Competition L. Rev. 1, 9–16 (2019). Productivity varies substantially among real-world firms, even within narrowly defined industries,208Syverson, supra note 207, at 326–27. and competition correlates with higher productivity.209Id. at 347–57. There could be more than one mechanism by which competition produces these effects,210It might pressure current managers to invest effort and resources into efficiency improvements, or it might just drive a Darwinian selection by which less efficient firms and their managers are forced out entirely. There is significant support for both explanations. On the one hand, managers do in fact have tools at their disposal to improve productivity and some evidence confirms that individual firms do so over time. Syverson, supra note 207, at 335–47, 352–53. On the other hand, the best-established result in the literature is that more productive firms are more likely to survive. Id. at 327. Which of the two mechanisms is more important remains unclear. but whichever best explains it, support has increased that competition correlates with productive efficiency.

A second, separate reason that mergers could generate market power without shareholder gain is wasteful expenditures taken to get or keep market power. The hunt for market power is generally thought to be socially wasteful and, under at least some circumstances, should lead firms to dissipate most or all the gains of monopoly in seeking it or protecting it once achieved.211The classic statement comes from Richard A. Posner, The Social Costs of Monopoly and Regulation, 83 J. Pol. Econ. 807, 809 n.4 (1975).

V. Policy Implications

What should the law do to accommodate the empirical results discussed here? To summarize, mergers and acquisition are unlikely to lead to social gains and might cause serious social loss. Efficiencies of scale and technological integration are probably real but available only in deals too small to be of policy relevance. Large takeover premia are apparently driven by target shareholder holdout and represent no more than one-time transfers of acquiring-firm wealth. While exactly what motivates this seemingly irrational conduct is unknown, large-scale acquisition activity seems best explained by management agency cost. Connections might exist to stock market overvaluations or economic shocks. Still, they are not well understood, and in any event, they would not support explanations based on efficient merger or rational expectations of efficiency. Meanwhile, the evidence increasingly supports the traditional expectation that concentration generates market-power harms and is worsened by mergers and acquisitions.

So what policy corrections could be useful?

First, there is no obvious call for new law, or at least no opportunity for new lawmaking that seems both politically plausible and up to any necessary task. Existing law seems adequate to pressing needs, so long as certain recalibrations are made.

The less important concern is in corporate and securities law. To be clear, there is room for a substantial theoretical re-examination in corporation scholarship. The large extent to which scholarship presumes market institutions are better able to control corporation problems than legal rules is badly at odds with the evidence here. At least one very significant class of market activity operates very ineffectively. It seems unwise to assume on a priori grounds that institutions are available that operate with automatic effectiveness. In particular, the confidence of that literature that agency cost or any other problem could be constrained by an efficient market for corporate control seems wholly misplaced.

But reform of existing legal rules in corporate or securities law would not accomplish much concerning the problems that acquisitions cause. Those laws are already basically inert, at least as they relate to acquisitions. Moreover, it is hard to imagine corrections in them that would solve any relevant problem. Admittedly, the empirical evidence discussed in this paper suggests that the caselaw on point indulges empirical presumptions that are incorrect. However, the rules in question don’t usually generate especially bad results because they generate hardly any effects at all. Substantive corporation law does not specifically encourage acquisitions, except that it makes all challenges to management decisions extremely difficult. In some sense it encourages deals because, absent conflict of interest, challenges face business-judgment-rule treatment and are effectively impossible to win. But that is just a consequence of the law’s extraordinary deference to all management decisions, not to acquisitions in particular. To some degree, the law might encourage hostile takeovers because the courts have created special, somewhat tougher rules in that context to limit how incumbent managers react to them. If hostile takeovers are as unlikely to do good as other deals, these rules may be undesirable. Perhaps incumbent managers should be largely free to thwart hostile deals if they like. But they are already largely free to do that on current law. So-called Unocal and Revlon challenges may be easier than business-judgment-rule cases, but they are still extremely challenging and plaintiffs rarely succeed.212See supra notes 47–50 and accompanying text.

Conceivably, there could be some value in making shareholder merger challenges easier to win, and there is no particular reason that corporate law reforms should not be made. The point is merely that they cannot remedy the real and potentially very serious harms likely done by mergers and acquisitions. On the one hand, maybe target-firm shareholders should more often win the fiduciary or appraisal lawsuits they usually bring, in which they claim that the prices they were paid were too low. Even though the evidence suggests that they are already substantially overpaid, a greater legal risk of liability to those shareholders might discourage merger activity overall, and maybe that would be a good outcome. Or, on the other hand, acquiring-firm shareholders should have an easier time challenging their managers’ decisions to make acquisitions since they are so often ill-advised. But these ideas just lead to the second problem.

But more importantly, corporate and securities laws probably cannot do much to address the real harms caused by mergers because those harms are felt by third parties. Corporate and securities laws are devoted on a fundamental level to protecting companies and their investors. Their basic mechanism is to impose liability on managers when they take culpable action that reduces shareholder value. Accordingly, the only parties who actively enforce corporate law are shareholders. They have very little incentive to stop inefficient mergers. Generally speaking, target-firm shareholders favor takeover regardless of efficiency, because of the large premium prices they generate. Likewise, except in closely held firms with large individual shareholders, no acquiring-firm shareholder likely has a sufficient financial interest to challenge ill-starred acquisition deals. These incentives are telling, in fact, because the real harm they cause is not to companies or their owners. Deals among large firms can definitely waste the acquiring firm’s money, but do not always, and on average they may leave shareholders more or less just as well off. The real harms of mergers are to third parties—the merging firms’ suppliers, labor, and customers. When firms do gain market power, they use it either to raise prices to those who buy their products or lower the prices they pay to those who supply their inputs. Those who are the real parties in interest to some policy problem tend to be the best enforcers of policy to correct that problem. The fact that corporate third parties can effectively never enforce corporation law is the key reason that that law is not well suited to the problems that mergers cause.

As a separate problem, the only routine corporation law challenge that could actually address the harm that mergers pose to companies or shareholders—claims by acquiring-firm shareholders to stop their managers wasting company money—would logically entail speculative, prospective judicial guesses on whether particular deals will be profitable. That is a judgment even managers themselves are manifestly bad at making, and the large empirical record sheds no light on which deals will fail and which will succeed. Admittedly, the major alternative this Article will suggest—revitalized antitrust enforcement—might also have to make efficiency determinations, but as will be discussed, antitrust is better suited to do it.

So, if there is any lesson for corporate or securities law, it is more abstract. It goes to the general approach or mood that has dominated that law and its theory for some time. It has become so deferential to businesses and their managers that scholars now often describe the “death of corporate law,”213For a few of the several recent essays on this theme, see James D. Cox & Randall S. Thomas, Delaware’s Retreat: Exploring Developing Fissures and Tectonic Shifts in Delaware Corporate Law, 42 Del. J. Corp. L. 323, 328–373 (2018) (documenting the deliberate retreat of Delaware courts from judicial control of management, based on confidence in market and shareholder constraints), and Zohar Goshen & Sharon Hannes, The Death of Corporate Law, 94 N.Y.U. L. Rev. 263 (2019). and many of them welcome it. That has taken form most recently in the dramatic rise in institutional ownership,214Gilson & Gordon, supra note 13, at 874–76 (describing recent rise in institutional ownership). Institutions now own more than half the equity of all public firms, and among the largest firms they own nearly three-quarters. See id. The institutions themselves are also concentrated, with a handful of the largest of them controlling most of those assets. which many scholars expect will bring the sophisticated oversight of institutional investors as alternatives to courts and law.215See, especially, Goshen & Hannes, supra note 213; see also Gilson & Gordon, supra note 13 (arguing, on a theoretical basis that, while most institutions have remained essentially passive, they might partner with activist hedge funds to pressure better governance). But that position is just characteristic of the dominant attitude in corporate scholarship for decades. The view is that rationality and strong market institutions—like the enlightened self-interest of institutional shareholders—will do the best job of regulating companies, including by constraining their managers’ self-interest. The fact that rationality and market institutions preserve an emphatically failed and wasteful universe of takeover transactions suggests otherwise.

But that is a side issue, because there is a real policy solution at hand and it is simple re-vitalization of existing antitrust law. In this area, the empirical evidence favors radical readjustment of priorities and enforcement, but it can be accomplished without any new legislation or changes to the basic law. All that is needed is for the federal judiciary to take notice of the evidence presented here. The existing law is entirely judge-made, and it was made on a priori judgments of a small number of lower-court judges many years ago that now seem demonstrably, emphatically false. They believed that mergers probably can’t cause much harm, and that constraining them threatens loss of substantial social gains.216See supra notes 64-66 and accompanying text. There is in fact so little evidence of social gain from any merger even distantly likely to face antitrust challenge that we could comfortably return to the stark prophylaxis of PNB with no fear of loss.

The fact that a revitalized merger law could entail some speculative efficiency determinations, of the kind that seem very hard for judges to make, is only a regrettable inevitability, and is best handled institutionally in antitrust if it must be handled at all. During the PNB regime of the Warren Court years, efficiency was disregarded entirely.217See supra note 89 and sources cited. That seems a reasonable enough trade-off if it really is true that mergers systematically generate harm but do not systematically generate any social gain at all. But even if efficiencies must be measured, there are several advantages in measuring them through antitrust. Antitrust merger law is primarily enforced by two federal agencies that both maintain large, sophisticated bureaus of economists. Moreover, antitrust theory and scholarship contain more than fifty years of elaborate, extensive analysis regarding whether and when merger efficiencies are possible, and how they can best be measured institutionally. And a key benefit would be the law’s assignment of party incentives. In antitrust, there will always be a party—the plaintiff—with a substantial interest in challenging a defendant’s efficiency claims. In corporation law there frequently will not be.

So, in mergers and acquisitions we have come fully around the circle. We have lived through a conservative revolution driven almost entirely by a priori theoretical critiques. Those critiques apparently reflected pre-theoretical, ideological pre-commitments to prove that “big” is not necessarily “bad,” and might often be desirable, and to oppose government interventions that don’t give defendants the benefit of extensive doubt. Those critiques and ideological priors never earned meaningful empirical corroboration, despite decades of robust inquiry. It seems that the apparently crude and empirically thoughtless presumptions of vigorous antitrust and non-deferential corporation law might have been pretty reasonable after all. “Big” is perhaps a pretty reasonable proxy for “bad,” such that concentration control using crude, prophylactic measures is more than adequate and desirable, and is not unduly costly.

The balance between false positive and false negative, at long last, should be reset.

Appendix: Difference in Total Welfare Loss

Depending on Pre-Merger Market Power

Figure 1 assumes that the pre-merger price, P1, was competitive—and therefore equal to the merging firms’ marginal cost, MC. When the post-merger price rises to P2, the deadweight loss is only the shaded triangle, representing the lost surplus that consumers would have enjoyed from the sales between quantities Q1 and Q2. Figure 2 more plausibly assumes that firms merging to significant market power were already charging a supra-competitive price. Deadweight loss is now the shaded trapezoid.

* James A. Thomas Professor of Law, Cleveland State University, c.sagers@csuohio.edu. Material in this paper was presented at the law schools of the University of North Carolina and the University of Pennsylvania. Thanks for feedback and discussion (over years and different versions of this project) to Neil Averitt, Jonathan Baker, Michael Borden, John Coyle, David Forte, Doron Kalir, Ken Kowalski, John Kwoka, Christa Laser, Joe Mead, Diana Moss, John Newman, Russell Pittman, Rick Su, Eric Tucker, and Jonathan Witmer-Rich. Particular thanks to Peter Carstensen for feedback and for allowing me to build on our separate work together.