Divestiture: Doctrinal Development and Modern Application

A. Origins and Early Application: 1911–1950

From its first application in 1911 until the Clayton Act was amended in 1950, divestiture was primarily administered in cases brought under the Sherman Act. ¹⁰ As a data point, the Supreme Court heard twenty cases between 1911 and 1950 that contemplated divestiture, sixteen of which were brought under the Sherman Act alone compared to two grounded solely on Clayton Act violations. ¹¹ More specifically, most divesture orders were issued against defendants found liable under § 2 of the Sherman Act (hereinafter “§ 2”) for actual or attempted monopolization. ¹² However, the monopolization claims that resulted in divestiture during this time period did not center on unilateral conduct as § 2 violations typically do today. Instead, as the Court’s jurisprudence reflects, divestiture was applied exclusively to remedy unlawful acquisitions and combinations. Prior to the 1950 amendment of the Clayton Act, enforcers seeking such a remedy necessarily relied on § 2.

Divestiture was first applied as an antirust remedy in Standard Oil, which was decided in 1911. ¹³ There, the Supreme Court held that the oil conglomerate violated §§ 1 and 2 of the Sherman Act and ruled that the company’s ongoing monopolization of the domestic petroleum market “require[d] the application of broader and more controlling remedies” than mere injunctive relief to limit future restraints on competition. ¹⁴ Prior Sherman Act enforcement centered on the § 1 prohibition of multilateral anticompetitive conduct, under which injunctive relief could nullify existing contracts and combinations between firms and restrict the creation of such agreements in the future. ¹⁵ In Standard Oil, the Court recognized that such an injunction would not remedy Standard Oil’s already successful and illegal monopolization of the market. ¹⁶ While the Sherman Act did not expressly authorize divestiture as a remedy, the Court found statutory authority for it in Congress’s intent that the Act address “injury to the public by the prevention of an undue restraint.” ¹⁷ Its conclusion that only divestiture could repair the harm caused by Standard Oil’s monopolization led the Court to affirm a break-up of Standard Oil into thirty-four independent companies. ¹⁸

On the same day that the Court issued its decision in Standard Oil, it also ruled in American Tobacco that the American Tobacco Company had violated §§ 1 and 2 of the Sherman Act. ¹⁹ On the issue of remedy, three factors were identified as relevant to the Court’s decision to dissolve American Tobacco into four separate and independent entities, namely:

(1) The duty of giving complete and efficacious effect to the prohibitions of the statute; (2) the accomplishing of this result with as little injury as possible to the interest of the general public; and, (3) a proper regard for the vast interests of private property which may have become vested in many persons as a result of the acquisition, either by way of stock ownership or otherwise, of interests in the stock or securities of the combination without any guilty knowledge or intent in any way to become actors or participants in the wrongs which we find to have inspired and dominated the combination from the beginning. ²⁰

Thus, the Court’s remedial analysis in American Tobacco identified the same overarching objective as in Standard Oil, which was to prevent and correct harm to the public interest caused by unlawful restraints on competition. In pursuing that objective, the Court acknowledged the difficulty in ascertaining a specific remedy’s likely effectiveness in restoring competition and its concomitant concern that any such remedy could actually further undermine competition in the relevant market(s). ²¹ Even in the early years of antitrust enforcement, the risks of overdeterrence and unintended harm to the public interest led the Court to opt for measures less drastic than divestiture. With those concerns in mind, the Court overturned a lower court’s divestiture order in Terminal Railroad Association only one year after Standard Oil and American Tobacco. ²² Having determined that railroad association members had coordinated to deny competitors access to essential facilities in violation of the Sherman Act, the Court concluded that injunctive relief would better serve the public interest than divestiture. ²³ The Court reasoned that the railroad association was not an inherently illegal combination, and instead had acted illegally through exclusionary contract agreements. Opining that it was important to “preserve to the public a system of great public advantage,” the Court ordered the parties to modify their existing agreements in seven specific ways to remove the anticompetitive terms, on pain of dissolution. ²⁴ The Court thus offered the railroad association an opportunity to cure the unlawful aspects of its agreements while making clear that it would be dissolved if the parties did not adhere to the Court’s order.

The Court’s central focus on public interest concerns expressly excluded consideration of the remedy’s potential harm to the defendant’s private interests. The likelihood that the remedy would create or exacerbate a defendant’s financial difficulties was in of itself irrelevant, as reflected in the Court’s observation in the 1944 Crescent Amusement Co. decision that “Those who violate the Act may not reap the benefits of their violations and avoid an undoing of their unlawful project on the plea of hardship or inconvenience.” ²⁵ However, the defendant’s potential financial difficulties may be relevant where the public interest would be harmed as a result. For example, in affirming a lower court’s rejection of divestiture in United States v. U.S. Steel, the Court found an absence of public benefit in dissolution while noting that structural relief would harm the public interest by causing “a material disturbance of, and . . . serious detriment to, the foreign trade.” ²⁶ The salient point here is that the Court’s concern is with the public interest, not the impact on the private defendant, although public and private interests may coincide in disfavoring divestiture.

Section 2 was the primary vehicle for divestiture through 1950. However, the Court made clear that divestiture’s suitability largely depends upon the nature of the conduct underlying liability rather than the statutory source of liability, stating in Crescent Amusement Co., “Dissolution of the combination will be ordered where the creation of the combination is itself the violation.” ²⁷ During the first half of the twentieth century, such illicit combinations were often achieved through acquisitions identified as exclusionary acts that monopolized the market, such as in Standard Oil and American Tobacco. Violative combinations were also found where horizontal competitors coordinated through a central organization, such as a holding company or trade association, which then exercised monopoly power on their collective behalf to set prices, boycott outside competitors, and so forth. ²⁸ In the context of combinations, divestiture results in the dissolution of the central organization, not the break-up of any of individual member. Whether applied to unlawful acquisitions or combinations, divestiture was developed as a form of restitution or market correction to restore competition, rather than to punish offenders. Through the present day, as discussed below, divestiture continues to be applied primarily to illegal mergers and combinations, albeit under the banner of § 7 of the Clayton Act following the statute’s 1950 amendment.

B. The Shift to Clayton Act § 7 Divestiture: 1950–1976

C. The Hart-Scott-Rodino Act of 1976

The role of divestiture in antitrust enforcement experienced another major shift with the enactment of the Hart-Scott-Rodino Act (“HSR”) in 1976, which ushered in “the replacement of merger control through litigation with a comprehensive scheme of merger regulation.” ⁴⁵ The statute established requirements for premerger notification to the FTC and the Department of Justice (“DOJ”) for proposed transactions “that affect commerce in the United States and are over a certain size.” ⁴⁶ In turn, premerger filing requirements enabled federal enforcers to more easily identify anticompetitive acquisitions prior to consummation, thereby reducing the frequency of post-consummation litigation and subsequent court-ordered divestitures, and increasing the prevalence of negotiated settlement agreements imposing partial divestitures. ⁴⁷

Under the HSR, the FTC and DOJ are to review the premerger filings before allowing the parties to consummate the proposed acquisition. The HSR further requires that the valuation thresholds that trigger premerger filing requirements be revised and filed every year. The most recent thresholds were published by the FTC on February 1, 2021. For mergers affecting commerce in the United States, the 2021 premerger notification thresholds required filings either where (1) the value of the transaction exceeded $368 million or (2) the value of the transaction exceeded $92 million and the smaller and larger parties were valued at a minimum of $18.4 million and $184 million, respectively. ⁴⁸

Typically, the merging parties must wait thirty days after filing before consummating the merger to permit the FTC and DOJ to each complete a review of the transaction. ⁴⁹ This initial agency review will result in either an approval of the transaction or a second request for further information and documentation from the merging parties. The latter provides for an additional review period of thirty days once the information is fully submitted, or ten days for cash tender offers and bankruptcy sales. The agencies also have procedural mechanisms in place that allow the regulated parties to contest those aspects of the second request for information as “unreasonably cumulative, unduly burdensome, or duplicative.” ⁵⁰

Prior to the adoption of the HSR Act, federal enforcers not only struggled to identify mergers that might violate § 7, but also lacked the information-gathering capabilities often needed to secure preliminary injunctive relief delaying consummation. ⁵¹ As a result, premerger relief initially remained difficult to obtain after the Celler-Kefauver Act’s enactment in 1950. ⁵² Difficulties in identifying potentially unlawful transactions before their occurrence, and then assembling the information necessary to act preemptively, necessarily limited the government’s ability to file enforcement actions prior to the completion of the transaction. In instances where enforcers lacked the notice and information to initiate premerger § 7 enforcement, the government had to also account for the attendant risks and costs associated with pursuing post-consummation relief, rather than merely considering the likelihood and degree of anticompetitive effects posed by the acquisition. ⁵³

With its enactment in 1976, the HSR Act addressed the informational deficiencies that had previously limited the frequency and success of premerger challenges. As former Assistant Attorney General William Baer has explained, “HSR gave the antitrust agencies the tools they need to enforce Section 7 as Congress intended—to stop potential problems in their incipiency.” ⁵⁴ The 1976 statute finally equipped enforcers with “adequate notice of proposed mergers and a powerful investigative tool,” as well as sufficient time “for the most part” to evaluate whether and how to best proceed. ⁵⁵ In the years that followed the HSR Act’s enactment, § 7 enforcement shifted further away from post-consummation divestiture actions, and increasingly occurred via premerger regulation. By 1997, “the vast majority of the mergers the agencies investigate[d] [were] . . . reported and examined at the premerger stage, and the vast majority of merger challenges [were] initiated at the premerger stage.” ⁵⁶ More than two decades later, § 7 actions continue to be utilized primarily in the form of premerger enforcement. However, as Section III explains, divestiture remains a critical mechanism of merger control, notwithstanding the decline in post-consummation § 7 challenges.

A. Premerger Challenges and Settlement Agreements

Although federal authorities began relying on the revamped § 7 as their preferred mechanism for merger control following the 1950 enactment of the Celler-Kefauver Act, informational obstacles continued to limit premerger challenges. ⁵⁷ Upon its enactment in 1976, the HSR Act permitted federal enforcers to pursue preemptive injunctive relief more effectively against proposed anticompetitive mergers. However, the expanded use of premerger enforcement has not reduced the frequency of divestiture. Instead, it has resulted in the increased use of divestiture as a component of negotiated settlement agreements and a decrease in the imposition of divestiture in post-trial final orders. The obvious explanation for this trend is that companies often opt to negotiate for more limited divestiture terms in exchange for merger approval, rather than risk costly litigation and more expansive divestiture terms.

The FTC and DOJ have promulgated guidelines explaining their criteria and processes for assessing and challenging transactions under § 7 since 1968, most recently publishing horizontal merger guidelines in 2010 and vertical merger guidelines in 2020. ⁵⁸ As explained in the guidelines, either agency may seek an injunction in federal court to prevent a transaction’s consummation when the HSR premerger filings suggest a sufficient likelihood of a § 7 violation. Unlike in § 2 cases, federal enforcers do not have to establish that a defendant possesses monopoly power or prove a but-for causal connection between the merger and anticompetitive effects. Rather, they have the more limited burden of showing that substantial anticompetitive effects are likely and will outweigh any cognizable efficiencies asserted by the defendant. ⁵⁹

Premerger filings made it easier for enforcers to preemptively identify and permanently enjoin acquisitions prior to consummation under the relatively lenient standards of § 7. One such case was the DOJ’s 2011 challenge of a proposed merger between H&R Block and Tax Act on § 7 grounds. Following trial, the court ruled that the DOJ had “established a prima facie case indicating that anticompetitive effects are likely to result from the merger,” while the defendants failed to provide “evidence that rebuts the presumption of anticompetitive effects,” leading the court to issue a permanent injunction blocking the merger. ⁶⁰ The threat of such injunctions, as well as the monetary and collateral costs of lengthy litigation, can lead parties to abandon contemplated acquisitions before an enforcement action is adjudicated. For example, after the DOJ filed a complaint seeking to enjoin AT&T’s acquisition of T-Mobile under § 7, the two companies opted to scuttle the merger rather than plunge into what promised to be a protracted and uncertain legal battle. ⁶¹

Perhaps unsurprisingly, since the enactment of the HSR Act, § 7 challenges have most often been resolved through settlement agreements between the government and the merging parties, which are then entered as consent decrees once court approval is obtained. Premerger settlement agreements typically permit an acquisition’s consummation while addressing the government’s anticompetitive concerns. In addition to behavioral relief, such agreements commonly feature some form of divestiture, with the concordant result that settlements have become the primary vehicle for divestiture. ⁶² The DOJ has identified six underlying principles that ultimately guide its approach toward determining the appropriate remedy, including the scope of any contemplated divestiture, whether sought through a negotiated settlement or adjudication:

(1) Remedies must preserve competition. (2) Remedies should not create ongoing government regulation of the market. (3) Temporary relief should not be used to remedy persistent competitive harm. (4) The remedy should preserve competition, not protect competitors. (5) The risk of a failed remedy should fall on the parties, not on consumers. (6) The remedy must be enforceable. ⁶³

Of course, that an agreement was reached and includes divestiture does not necessarily mean that the agency was likely to succeed at trial. Settlement negotiations may be initiated by defendants hoping to avoid further litigation and uncertainty. ⁶⁴ They may be particularly incentivized by the government’s advantages in bargaining power, as well as the importance of securing the merger, avoiding delay, and limiting the costs of litigation. ⁶⁵ Put simply, the dual interest in certainty and cost containment are often compelling reasons for a party to compromise early if it can achieve an essential goal. Moreover, the scope of divestiture imposed via settlement is far more narrow than post-consummation divestitures that unwind completed mergers and are certainly less draconian than those that broke up acquired monopolies.

In actions brought by federal enforcers, the process is typically completed through a sale of assets to a government-approved buyer, which must “include[ all assets necessary for the purchaser to be an effective, long-term competitor.“ ⁶⁶ Regulators most favor divestitures that incorporate a standalone business belonging to one of the merging entities “because it has demonstrated success competing in the relevant market.” ⁶⁷ A standalone business may refer to a particular product line or business unit, or perhaps production facilities within specific geographic markets where the merger would place competition at risk. ⁶⁸ In some instances, the settlement may require divestiture of more than a single standalone business and could include additional assets falling under other business verticals, such as intellectual property or machinery, while other mergers may be approved on the basis of a consent decree limited to asset carve-outs from a standalone business. ⁶⁹ Settlement divestitures will typically be limited to either assets already held by the acquirer or a partial divestiture of the assets being acquired. ⁷⁰ Should the parties fail to reach a settlement and § 7 liability is later established at trial, a permanent injunction may be issued to block the merger’s consummation. However, courts also retain the discretion to instead order a narrower divestiture or behavioral relief while allowing the merger to proceed.

B. Consummated Mergers

While the Celler-Kefauver and HSR Acts reduced their necessity and frequency, post-consummation divestitures remain an unexceptional form of § 7 merger control, largely subject to the same standards of liability and remedial analysis as premerger enforcement. Thus, as the FTC has explained, § 7 liability in a post-consummation challenge is assessed pursuant to a legal analysis that “does not differ significantly” from that applied in a premerger action. ⁷¹ That said, the acquirer’s integration of the target firm’s assets can pose practical obstacles to divestiture, which can be further aggravated by delays in enforcement. As time passes after a merger’s consummation without enforcement, the merged entities may become increasingly integrated and thereby complicate efforts to later unwind the transaction. There is a dearth of recent caselaw addressing the intersection of delayed enforcement and divestiture, but it appears that the greater the degree of integration, the more reluctant courts are to unscramble the eggs.

To be clear, the completion of a merger is not an accrual date and there are no formal limitations on the government’s ability to challenge or seek to unwind consummated mergers. The question of timeliness was squarely presented in the Facebook action, which was not filed until 2020 even though it concerned acquisitions made in 2012 and 2014. The district court rejected Facebook’s argument that the government’s complaint was untimely, noting that the Supreme Court had defined “the term ‘acquisition’ as used in Section 7 of the Clayton Act . . . [not as] ‘a discrete transaction but [rather] a status which continues until the transaction is undone’.” ⁷² More to the point, “Section 7’s ‘ban against [certain] acquisitions . . . include[s] a ban against holding certain assets’, not just ‘obtaining’ them in the first place.” ⁷³ The court succinctly summarized this point in Facebook, stating, “The rule under Section 7 is thus that so long as an acquiring company continues to hold acquired assets, the Government may at any time argue that such company is violating Section 7.” ⁷⁴

Liability may exist even where decades have passed between the purchase of another company’s assets and the filing of an action. In 1957, for instance, the Court ruled in United States v. E.I. du Pont that the defendant corporation’s 1919 stock acquisition remained subject to liability. ⁷⁵ Although the E.I. du Pont and Facebook opinion specifically concerned whether the passage of time can circumscribe liability, there is no formal bar to divestiture once liability has been established. The court has discretion to order the remedy that is most likely to effectively restore competition, including divestiture. In the context of antitrust actions, federal enforcement actions for injunctive relief are not subject to a statute of limitations defense, and so any “formal limitations period often has little relevance if the government seeks to enjoin either an ongoing violation or the possibility of recurrence of particular past conduct.” ⁷⁶ This is not to say that the passage of time is irrelevant to the question of whether divestiture is an appropriate remedy, but rather that it is simply a contextual factor for a court’s consideration.

Recent enforcement statistics indicate that divestiture continues to be a readily applied post-consummation remedy. Between 2001 and 2020, the DOJ and FTC collectively filed fifty-one enforcement actions against consummated mergers, thirty-seven of which led to some form of divestiture achieved at either trial or adjudication. ⁷⁷ While most of the actions were resolved through settlement, at least seven cases produced written opinions that addressed divestiture. An examination of those seven cases indicates that courts ⁷⁸ have continued to exercise their remedial discretion in accordance with the same principles applied in § 7 post-consummation challenges prior to the HSR Act. ⁷⁹ This continuity of courts’ remedial analyses has bridged the gap in divestiture doctrine from its initial application under § 2 to its now predominant use under § 7, regardless of whether the action is premerger or post-consummation. In both premerger and post-consummation actions, the overarching objective of § 7 relief “is to restore competition lost through the unlawful acquisition.” ⁸⁰ Thus, the central consideration is “whether [divestiture] will effectively [restore competition] under the facts of each case,” which encompasses the likelihood that divesture will produce a “viable, independent entity” capable of competing with the defendant and other market participants. ⁸¹ In addition, courts continue to apply the principle stated in du Pont that while “remedies should not be punitive . . . ‘courts are authorized, indeed required, to decree relief effective to redress the violations, whatever the adverse effect of such a decree on private interests’.” ⁸²

As the decisional history makes clear, once § 7 liability has been established, divestiture is likely to be ordered, given that it is “the customary form of relief in § 7 cases.” ⁸³ That is because liability gives rise to a presumption that the proper remedy requires a total divestment of the assets acquired in the consummated merger, rather than a partial divestiture or only behavioral relief. ⁸⁴ Full divestiture may only be avoided if the defendant rebuts the presumption by proposing an alternative remedy that would “adequately redress any violation which is found.” ⁸⁵ In so doing, “defendants bear the burden of showing that any proposed remedy would negate any anticompetitive effects of the merger.” ⁸⁶ The underlying reason for this presumption is that full divestiture is typically the remedy most likely to restore competition hindered by an unlawful combination or merger, and that more limited remedies carry greater risk of failing to restore competition at the expense of consumers. ⁸⁷

An acquisition’s consummation does not itself affect this evaluation, just as dissolution’s financial consequences for a defendant are not material absent evidence of potential injury to the public interest. ⁸⁸ If held liable at trial, a defendant thus faces a tall order in overcoming a presumption of full divestiture. The defendant must propose an alternative remedy and provide “clear and convincing” proof of its “probable efficacy.” ⁸⁹ Furthermore, behavioral or conduct remedies are regarded as more costly and intrusive than divestiture, and are therefore disfavored. ⁹⁰ Nor is the government held to the same standard in seeking complete divestiture, given that once “the Government has successfully borne the considerable burden of establishing a violation of law, all doubts as to the remedy are to be resolved in its favor,” ⁹¹ and that “if an order of divestiture appears to the Commission to be in all likelihood the most effective available remedy, the Commission need not justify its order beforehand by showing that it will unquestionably restore competition.” ⁹² There are also limitations to the materiality of post-merger evidence favorable to the defendant, which must be a form of evidence that the defendant is incapable of controlling. ⁹³ Under the subject to manipulation standard, courts assign limited probative value to post-merger pricing and market share indicative of continued competition, since defendants are arguably capable of manipulating those forms of evidence in their favor ahead of anticipated government investigation. ⁹⁴ On the other hand, post-merger price increases and similar evidence of anticompetitive effects is probative, so long as the government establishes causation. ⁹⁵ Even if the defendant presents sufficient evidence of a proposed remedy’s likely efficacy in restoring competition, the trial court is under no obligation to favor the more narrow alternative and retains the discretion to order divestiture. ⁹⁶

Should the defendant’s consummated merger be found unlawful, the courts’ remedial analysis is inclined toward divestiture. However, a defendant may be able avoid or limit the scope of divestiture where the facts present “unusual circumstances” under which such a remedy would “be impracticable or inadequate.” ⁹⁷ Such circumstances are more likely to be found where the defendant has achieved significant post-merger integration of the acquired assets. Thus, while the passage of time between an acquisition’s consummation and the government’s enforcement action does not in of itself undermine the likelihood of divestiture, a delay can allow for further integration that affects the court’s remedial analysis. The Evanston Northwestern Healthcare Care Corp. case aptly illustrates this point. ⁹⁸ There, the FTC overturned an ALJ’s divestiture order in favor of behavioral relief. In that case, seven years had passed since the merger’s consummation, by virtue of which the defendant hospital had acquired one of several regional competitors. The FTC affirmed the ALJ’s determination that the transaction violated § 7, finding that the resulting anticompetitive effects were reflected in post-merger price increases and that the effects were not “offset by merger-specific efficiencies.” ⁹⁹ However, the Commission overturned the ALJ’s divestiture order, stating that, due to the significant integration that had occurred in the years following consummation, divestiture could potentially produce inefficiencies and diminish certain post-merger improvements in the hospitals’ healthcare services that were not readily replaceable. ¹⁰⁰ The FTC’s decision underscores that the greater the delay between consummation and enforcement, the more potential there is for increased integration that can make it “very difficult, or impossible, to unscramble the eggs and reconstruct a viable, divestable group of assets.” ¹⁰¹ The relationship between the length of elapsed time and the degree of integration is consistent with the available data. Most of the thirty-seven divestitures secured between 2001 and 2020 concerned acquisitions that had been consummated within two years of the government’s complaint, and the longest period of time from closing to complaint was five years and one month. ¹⁰² Concurrently, the value of the transaction is relevant as it may reflect the complexity of the merger and the ensuing integration, and the attendant difficulties in unwinding the merger once completed. The transaction value is publicly available for twenty-seven of the 2001–2020 divestitures: twenty concerned deals valued at less than $50 million, while seven transactions were valued at more than $100 million, only three of which were valued at more than $1 billion. ¹⁰³ Actions regarding each of the three consummated mergers valued at over $1 billion were filed ten months, ¹⁰⁴ eleven months, ¹⁰⁵ and eighteen months after the closing date. ¹⁰⁶ Recent trendlines regarding transaction value are partly a result of mergers falling outside HSR filing requirements, only for regulators to later take notice of ensuing anticompetitive effects. Still, it is not unreasonable to infer that relatively larger transactions and longer delays in enforcement are also each indicative of a greater degree of integration and thus less likely to result in full divestiture. The absence of enforcement actions involving large-scale acquisitions and lengthy delays may be a matter of happenstance, but likely also reflects concerns on the part of regulators that there may be a point where unringing the bell is no longer feasible.

While the practicalities of integration may impede divestiture, it must be emphasized that it is rare for the degree of integration or other potential complexities to preclude divestiture upon a judgment of § 7 liability. In Evanston, the FTC stressed the exceptional nature of the remedy in noting that “this is the highly unusual case in which a conduct remedy, rather than divestiture, is more appropriate.” ¹⁰⁷ It further clarified the narrow applicability of its reasoning to future cases, stating:

Nor will our reasoning here necessarily apply to consideration of the appropriate remedy in a future challenge to a consummated merger, including a consummated hospital merger. Divestiture is the preferred remedy for challenges to unlawful mergers, regardless of whether the challenge occurs before or after consummation. ¹⁰⁸

Indeed, the Commission rejected the invocation of Evanston in a later healthcare merger case and upheld the lower court’s divestiture order upon finding that neither the degree of integration nor the merger’s benefits to consumers were comparable. ¹⁰⁹

Although the enactment of the HSR Act has brought about a reduction of post-consummation divestiture actions, the doctrinal principles upon which courts rely in addressing post-consummation divestiture have remained consistent. Thus, precedent regarding divestiture’s use under § 2 has continued to be applicable in § 7 challenges to consummated mergers, and determinations of liability will normally result in a full divestiture of the acquired assets when sought by federal enforcers. That said, most such post-consummation divestitures in the past two decades have involved lower-valued transactions and actions commenced within two years of consummation.

A. United States v. Alcoa

The 1950 district court decision in Alcoa and the 1953 district court ruling in United Shoe are two of the few cases that analyze divestiture as a remedy to unilateral conduct. Notably, both courts rejected dissolution on the basis that the defendant company’s structure made divestiture too difficult and risky a process. The reasoning underlying these rulings illustrates regulators’ limitations in utilizing § 2 to compel divestiture against a single-firm defendant.

The salient history of Alcoa begins with the Second Circuit’s 1945 determination that Alcoa had violated § 2 by monopolizing the market for virgin aluminum ingot, the raw material used to manufacture aluminum products. ¹¹⁵ Due to its ownership of key patents, the corporation had gained control of over 90 percent of the virgin aluminum ingot market. In the years leading up to the 1945 decision, wartime manufacturing needs had increased demand for aluminum ingot, resulting in an expansion of its production and market share. Competitors struggled to challenge Alcoa’s dominance due to its patent control and the burdensome expense of electrical energy sufficient to power production at a competitive scale. Alcoa also owned a separate company called Limited, to which it had transferred ownership of its international properties and through which Alcoa entered into cartel agreements with international competitors that facilitated output and price restrictions.

In what was a landmark decision on § 2 liability, the Second Circuit began by identifying a narrower product market than had the district court. ¹¹⁶ The Circuit provided a definition of monopoly power that remains in use today, which has been summarized as “the power to control prices and exclude competition,” ¹¹⁷ and determined that Alcoa possessed monopoly power. It then held that monopolization occurs where a firm with monopoly power has willfully carried out conduct that inhibits competition, even in the absence of a specific intent to maintain or expand its control of the market. ¹¹⁸ It follows, the Circuit explained, that a monopoly can only be attained or maintained legally through natural means, such as shifts in consumer preferences, a superior product or an ability to profitably scale production at lower costs than competitors. ¹¹⁹ The Circuit’s reasoning was subsequently adopted by the Supreme Court one year later in American Tobacco (1946), thereby expanding the scope of § 2 liability and laying the foundation for the modern formulation of monopolization articulated by Grinnell in 1966. ¹²⁰

On the preceding basis, the Circuit found Alcoa liable for violating § 2. However, the evidence introduced at trial had not accounted for the government’s creation and continued ownership of additional aluminum plants to support wartime production, which Alcoa and other producers operated pursuant to government contracts. ¹²¹ Thus, with the exception of narrowly crafted injunctive relief, the Second Circuit tabled any further remedy pending the expiration of the leases over the upcoming years, which were expected to clarify the disposition of the government-owned plants. In 1947, Alcoa petitioned the district court for a decree that it no longer possessed a monopoly on the market for virgin aluminum ingot. The government countered that that the district court ought to find that Alcoa had continued its monopolization and should therefore be divested to the extent necessary to create a new entity capable of restoring competition. ¹²²

The district court issued its ruling in 1950. The court concluded that while further relief was necessary, divestiture of Alcoa’s physical assets was not. In reaching this determination, the court did not cite any prior cases or economic analysis, seemingly relying instead on its “considered and firm judgment” as to divestiture’s obstacles and risks. It began by stating that a “strong and resourceful domestic aluminum industry is a vital necessity” in the interests of national security and the “peacetime welfare of the general public.” ¹²³ The court then explained that divestiture of an integrated company structured as a single firm was inherently “a highly speculative—and even hazardous—venture” because it could not be “dismembered . . . without a marked a loss of efficiency.” ¹²⁴ Alcoa’s uneven allocation of resources and production to different plants prevented the court from simply divesting it of certain facilities. Moreover, the court reasoned, the viability of a new competitor emanating from the divested assets would be undermined by a limited supply of “efficient and experienced management” due to Alcoa’s longstanding status as the sole domestic producer of virgin aluminum ingot. Finally, the court expressed concern that divestiture would dissolve Alcoa’s research department, thereby impairing continued innovation within the industry.

Having determined that Alcoa was too integrated to break up, the court declined to order a broad dissolution, directing instead that Alcoa’s shareholders divest their stock in either Alcoa or Limited while also issuing behavioral relief. In limiting the scope of divestiture, the court emphasized the company’s unitary structure and capacity for innovation, as well as the public’s interest in a healthy aluminum industry. It appears facially evident that the court was far more hesitant to apply divestiture than it would have been had Alcoa achieved or maintained its monopoly by acquiring a competitor.

B. United States v. United Shoe

Only three years later, the application of divestiture to single-firm conduct was again front and center in district court, and was again rejected as far too challenging and risky an endeavor. In an opinion summarily affirmed by the Supreme Court, the court determined that United Shoe had violated § 2 by monopolizing the market for shoe machinery. ¹²⁵ The court concluded that United Shoe possessed control of the market for shoe machinery and that its dominance was at least partly attributable to exclusionary practices, primarily in the form of leasing agreements that functioned as barriers to entry for potential competitors. ¹²⁶ In this sense, the government was able to establish both the existence of a monopoly and a corresponding exclusion of competition. However, with respect to the appropriate remedy, the court rejected the government’s request that United Shoe be divided into three separate and independent entities, and instead issued behavioral relief and a far narrower divestiture order. ¹²⁷

The court opened its discussion of remedy by stating that its primary objective was “to extirpate practices that have caused or may hereafter cause monopolization, and to restore workable competition in the market.” ¹²⁸ The court then articulated a need for judicial restraint in fashioning § 2 remedies, citing an absence of “economic or political training” among the judiciary, a lack of representative authority relative to other branches of government, and the need to cautiously exercise the unique economic authority conferred by antitrust law. ¹²⁹ Without a single cite to prior divestiture decisions, the court concluded that “the Government’s proposal [to] dissolve United into three separate manufacturing companies is unrealistic.” ¹³⁰ It explained that United Shoe’s use of a single facility to manufacture its machinery, “with one set of jigs and tools, one foundry, one laboratory for machinery problems, one managerial staff, and one labor force,” meant that, as a practical matter, the company “cannot be cut into three equal and viable parts.” ¹³¹ Beyond the challenges posed by United Shoe’s single-firm structure, the court pointed to the government’s failure to offer “a more formal commitment by the Attorney General” and sufficient evidentiary support that divestiture would effectively restore competition as a factor relevant to its determination.

The court did, however, impose a divestiture more limited than that sought by the government. In addition to behavioral relief tied to leasing terms, the court ordered the divestiture of United Shoe’s “subsidiaries and braches [sic] which produce supplies in fields which United has monopolized. . . . [of which] [t]he clearest examples are nails and tacks, and eyelets for the shoe machinery market.” ¹³² The court explained that this specific relief—contrary to the broader corporate divestiture requested by regulators—“is the kind of dissolution which can be carried out practically.” ¹³³ Here, the court explicated, the company’s supply chains could not be further separated because:

in each of the other cases where this might at first blush seem reasonable, the supply is technically so intimately related to a machine as to be naturally manufactured by the maker of the machine, or the supply is sold in such small annual volume, or the difficulties of enforcing divestiture of part of a plant are so obvious, as to make the extension of the decree to those instances undesirable. ¹³⁴

But beyond the court’s conclusion that the narrowed divestiture presented “no similar practical difficulties” as those identified with respect to dissolution, the most important distinction may be that the supply chains subjected to divestiture were those belonging to United Shoe subsidiaries “which [were] not part of United’s organization.” ¹³⁵ It is unclear whether the supply chains ordered to be divested would have escaped divestiture had they simply been owned directly by United rather than through its subsidiaries. ¹³⁶

The United Shoe decision provides one of the few opinions addressing divestiture as a remedy to unilateral violations, and like Alcoa, reflects a reluctance to order dissolution in single-firm cases. Together, the two decisions suggest that courts are disinclined to comprehensive divestiture orders against single-firm monopolists because of the practical obstacles to facilitating viable competitors, as well as the associated risks and potential costs accompanying such orders. In each case, the perceived risks of single-firm divestiture appear further aggravated by the broader economic stakes involved and the judiciary’s limited experience in economic analysis and decision-making. Thus, United Shoe and Alcoa indicate that even where monopolization has been established, courts are unlikely to divest single-firm monopolists of their directly owned assets.

C. United States v. AT&T

In 1982, telecommunications giant AT&T entered into a historic agreement settling a § 2 action brought by the DOJ in 1974. After nearly a century of monopolization over all aspects of the domestic telephone industry, AT&T agreed to divest itself of 22 regional Bell operating companies, thereby exiting the market for local telephone service. As explained below, AT&T monopolized long-distance and local telephone service and equipment markets through an aggressive acquisition strategy aimed at local independent exchanges, the government’s endorsement of the company as a natural monopoly, and the company’s ability to leverage network effects across geographic and product markets. ¹³⁷ Although the divestiture originated from a negotiated settlement, the district court’s opinion approving the proposed agreement offers guidance as to how single-firm divestitures are analyzed and considered.

The origins of AT&T’s dominance can be traced back to the invention of the telephone. In 1876, Alexander Graham Bell secured a patent on the telephone and then co-founded the Bell Telephone Company in 1877, which was later reorganized as AT&T. ¹³⁸ AT&T’s patents expired in 1894, at which point independent companies entered the market. But by then, AT&T had already established a nationwide network of exchanges connected by its long-distance service and secured other advantages that further limited new entrants’ abilities to compete. With the independent companies “unable to create a comprehensive parallel long-distance network,” AT&T “expanded its system through acquisitions” ¹³⁹ while denying competitors access to its network and engaging in “intense price competition.” ¹⁴⁰

After the company’s run of acquisitions led the DOJ to file an antitrust complaint in 1913, AT&T and the government entered into an agreement known as the Kingsbury Commitment. Under its terms, AT&T effectively became a government-sanctioned, regulated monopoly. ¹⁴¹ The government permitted AT&T to maintain its control of telephone services while the company agreed to “operate as a public utility, eventually providing high-quality phone service to the vast majority of Americans regardless of income or geography.” ¹⁴² In short, AT&T could continue to monopolize telephone service, but had to provide independent local companies with access to its long-distance network and accede to government rate regulation.

In the decades following the Kingsbury Commitment, AT&T continued to acquire independent exchanges and leverage its long-distance network. The Willis Graham Act of 1921 formally extended the 1913 agreement while exempting acquisitions of local exchanges from antitrust scrutiny. ¹⁴³ Under the statute, AT&T’s acquisitions of local exchanges simply required approval from the Interstate Commerce Commission, which by 1924 had approved 223 out of 234 acquisitions pursued by the company. ¹⁴⁴ After its creation in 1934, the Federal Communications Commission (FCC) regulated AT&T’s interconnectivity and telephone service rates, required AT&T to interconnect with independent exchanges, and limited the entrance of new competitors into the long-distance service market. ¹⁴⁵ The interconnectivity requirement further cemented independent exchanges’ reliance on the AT&T network. At the same time, AT&T made it impossible for exchanges to connect to its network without using equipment produced by Western Electric, a vertical subsidiary that manufactured and supplied telephone equipment. ¹⁴⁶ Antitrust concerns regarding AT&T’s vertical integration led to a 1956 consent decree that permitted the company’s continued ownership of Western Electric, in exchange for which AT&T was prohibited from entering the computer and computer services markets. ¹⁴⁷

Technological evolution led to the emergence of new competitors during the 1960s and 1970s, and the FCC “slowly and unevenly” began to “allow competitive entrance into various portions of the market.” ¹⁴⁸ By that point, AT&T had consolidated its local exchanges into 22 regional Bell operating companies (often referred to as “Baby Bells”) while offering long-distance service under its Long Lines division, selling telephone equipment through Western Electric, and operating Bell Labs as its research and development division. However, AT&T proved resistant to the prospect of competition in the local and long-distance service markets, and “often denied” network access to its competitors, or “at least delayed [access] through aggressive exploitation of the regulatory processes available to it” which “invited . . . renewed scrutiny from the [DOJ].” ¹⁴⁹ Thus, in 1974, the DOJ filed a new antitrust action against AT&T. ¹⁵⁰ The complaint alleged that AT&T and its regional operating companies had violated § 2 by engaging in exclusionary and anticompetitive tactics to preclude competition in both the long-distance and local telephone service markets, as well as by tying network access to the purchase of “inferior and overpriced telecommunications products from its subsidiary manufacturer, Western Electric.” ¹⁵¹ At the onset, the government sought relief in the form of “divestiture from AT&T of the Bell Operating Companies as well as the divestiture and dissolution of Western Electric.” ¹⁵²

AT&T initially resisted the DOJ’s efforts to break up its monopoly, resulting in lengthy and complex pretrial proceedings and discovery disputes that delayed the start of trial until 1981. ¹⁵³ Following opening statements, the court granted the parties’ request and stayed the proceedings for six weeks to allow for settlement negotiations. When those discussions failed, the trial resumed, with the government calling hundreds of witnesses and presenting “many thousands of documents and additional thousands of stipulations.” ¹⁵⁴ When its mid-trial motion to dismiss was denied, AT&T presented its defense, calling some 250 witnesses and introducing tens of thousands of documents. ¹⁵⁵ Judicial comments made in the course of the denial of AT&T’s motion to dismiss led the company to “realize it probably could not prevail in court.” ¹⁵⁶ Spurred on by the threat of defeat, AT&T revisited the issue of settlement. The company was concerned that losing would expose it “to ruinous private litigation” and behavioral relief that would prevent its continued operation. ¹⁵⁷ Further incentivizing AT&T to settle was its desire to maintain ownership of Western Electric and Bell Labs, and the opportunity to eliminate the 1956 prohibition against entry into the computer industry. ¹⁵⁸ Moreover, while divestment of the regional operating companies meant that AT&T would no longer be providing local telephone service, the company could continue operating as a long-distance service provider under scaled-back regulatory oversight. ¹⁵⁹

With these considerations in mind, a settlement agreement was reached with the government prior to the DOJ’s rebuttal presentation by which AT&T agreed to divest itself of the regional operating companies. ¹⁶⁰ The basic contours of the settlement, as reflected in the modified final judgment issued in August 1982, can be summarized as follows:

Under its terms, AT&T will divest itself within eighteen months of its local telephone exchange operations. The new AT&T will provide long-distance telephone services in markets that eventually will be deregulated, and it will be allowed for the first time to diversify into data processing activities . . . [T]he former Bell local operating companies [are] to be separated into seven regional telephone companies, each with projected assets of $17 billion to $21 billion. The regional companies are precluded from entering long-distance communications, or any other competitive market . . . AT&T will retain ownership of its Long Lines department and will obtain the long-distance telephone facilities of the divested local exchange companies to form an integrated national long-distance network. AT&T also will retain Western Electric and Bell Laboratories and will be allowed to compete for the equipment business of the new regional operating companies. The currently installed base of Bell-owned customer premises equipment will be transferred from the operating companies to AT&T. ¹⁶¹

Under the Tunney Act, ¹⁶² the district court was obligated to review the proposed consent decree to determine whether its entry was in the public interest. ¹⁶³ Thus, the fact that divestiture was imposed under a settlement agreement rather than adjudication does not diminish the significance of the court’s opinion to future cases considering single-firm divestiture. Moreover, the factors delineated in the Tunney Act largely mirror those typically considered in divestiture analyses in merger control and combination cases, with the exception that courts have discretion to consider broader public interest concerns under the Tunney Act. ¹⁶⁴ To that end, the court stated that “the decisions granting relief after a finding of liability form the most relevant yardstick for determining whether the proposed consent decree will further antitrust policies, and [it] will therefore use these decisions as its basic standard.” ¹⁶⁵ Accordingly, the court’s analysis centered on the nature of AT&T’s anticompetitive conduct and its relation to market concentration, the likelihood that divestiture would restore competition, and the suitability of alternative remedies. ¹⁶⁶ In exercising its broad discretion to weigh additional competitive and public interest factors, the court accounted for the strength of the trial evidence throughout its analysis and considered the sociopolitical effects of economic concentration. ¹⁶⁷

Based on the record, the court found that AT&T’s control of the regional operating companies had facilitated its myriad forms of exclusionary conduct. Each iteration of anticompetitive behavior that AT&T employed to monopolize long-distance and local service markets had been made possible “in large part because of its control over the local exchange facilities.” ¹⁶⁸ In addition, by restricting network access to purchasers using Western Electric equipment, AT&T leveraged its control of the regional operating companies to force purchasers “to buy products from Western Electric even though other equipment manufacturers produced better products or products of identical quality at lower prices.” ¹⁶⁹ Because its control of the regional operating companies was central to AT&T’s exclusionary conduct, the court found that divestiture would “make it impossible, or at least unprofitable, for AT&T to engage in anticompetitive practices.” ¹⁷⁰ The court did not use the term causation anywhere in the decision. However, its explanation of the necessity and likely effectiveness of divestiture pointed to AT&T’s ownership of the regional operating companies as the but-for cause of its monopoly power.

The court next examined and rejected alternative remedies less drastic than the proposed divestiture. It found that a proposed divestiture of Bell Labs and Western Electric, even when accompanied by the divestiture of some (but not all) of the regional operating companies, “would not eliminate AT&T’s ability and incentive to take anticompetitive actions against its competitors in the intercity market . . . [and] would thus be ineffective in the very area in which the government’s proof . . . was strongest.” ¹⁷¹ In addition to finding the alternative divestiture proposal less responsive and effective, the court also stated that it would harm Bell Labs’ research and development capabilities, and undermine its continued ability to support “innovation in the telecommunications industry and, more broadly, in industrial research.” ¹⁷² Behavioral relief was also rejected as a standalone remedy, given the scope and variation of AT&T’s “anticompetitive behavior that are claimed to have occurred over a considerable period of time, in various geographical areas, and with respect to many different subjects.” ¹⁷³ Upon noting that the company had employed its vast financial resources, skilled personnel, and technological acumen to evade effective oversight, the court concluded that “it is unlikely that, realistically, an injunction could be drafted that would be both sufficiently detailed to bar specific anticompetitive conduct yet sufficiently broad to prevent the various conceivable kinds of behavior that AT&T might employ in the future.” ¹⁷⁴

From there, the court considered the possible effects of divestiture on consumers’ access to telephone service and whether it would result in higher prices or lower quality service. The consent decree had included funding mechanisms for local telephone service that the court believed could maintain pricing at existing levels. ¹⁷⁵ Greater confidence was expressed in the likelihood that the quality of service would be maintained under divestiture, given the regional operating companies’ assets and commercial incentives for continued research and development. ¹⁷⁶ Ultimately, the court viewed divestiture as the remedy best suited to promote consumer pricing and quality control interests.

Finally, the court concluded divestiture was necessary to address AT&T’s consolidation of power within the telecommunications industry. This was a concern that extended beyond the economic ramifications of the company’s anticompetitive conduct to include potential harm to the nation’s political systems. The court turned to the legislative history of the Sherman Act and the 1950 Celler-Kefauver Act, ¹⁷⁷ as well as U.S. Supreme Court jurisprudence in its invocation of the principle that the Sherman Act “is founded on a theory of hostility to the concentration in private hands of power so great that only a government of the people should have it.” ¹⁷⁸ The district court explained its view of the relationship between economic and political power as follows:

Our political system is designed so that the power of one group may be checked by the power of another. The antitrust laws require this same approach in the economic sphere. Obviously, if one company controlled an essential part of the economy, it would be in a position to gain an undue influence over economic decisions and, as a result, most likely over political decisions. Thus, the antitrust laws seek to diffuse economic power in order to promote the proper functioning of both our economic and our political systems. The significance of these concepts is accentuated by the context in which the Court must consider the public interest in these cases. The telecommunications industry plays a key role in modern economic, social, and political life. Indeed, many commentators have asserted that we are entering an age in which information will be the keystone of the economy. ¹⁷⁹

The court’s discussion of economic concentration was tied to the Tunney Act’s objective of protecting the public interest. However, given district courts’ broad remedial discretion and the dearth of precedent regarding single-firm divestitures, it is plausible that economic concentration could be factored into a court’s consideration of divestiture as a remedy in a contested action. What is far more certain is the relevance of causation, alternative remedies, and consumer welfare to judicial determinations in future single-firm cases, as demonstrated by the absence of divestiture in the government’s 2002 settlement with Microsoft.

D. United States v. Microsoft

Like AT&T, technology giant Microsoft settled with the DOJ to resolve the government’s § 2 action against it. But unlike AT&T, Microsoft was unwilling to entertain divestiture and successfully negotiated an agreement limited to behavioral relief that was then judicially approved in 2002. ¹⁸⁰ Prior to the settlement agreement, the district court had found Microsoft liable under §§ 1–2 of the Sherman Act and subsequently ordered the company to divest either its operating systems business or software applications business, with the objective of severing them into separate and independent companies. ¹⁸¹ On appeal, the D.C. Circuit affirmed Microsoft’s § 2 liability in part but reversed the finding of § 1 liability and vacated the divestiture order, sending the case back to the district court for further proceedings. Microsoft then settled with the DOJ before the outstanding liability issues and remedy could be adjudicated on remand. While the vacatur was predicated on procedural grounds, the Circuit’s opinion provided substantive guidance, albeit in dictum, concerning divestiture. In articulating its posture toward single-firm divestiture, the D.C. Circuit placed particular emphasis on the unitary company considerations that anchored Alcoa and United Shoe, and expressly referenced the requirement of causation implicit in AT&T.

By way of background, Microsoft had been the focus of several antitrust investigations prior to the DOJ’s 1998 enforcement action and the resulting 2002 settlement. Founded in 1975, the company grew rapidly and quickly established industry dominance as the leading developer and manufacturer of computer software. ¹⁸² By the early 1990s, the company had captured more than 70 percent of the market for personal computer operating systems (hereafter referred to as “PC” and “OS”), aided in part by the market’s high barriers to entry and network effects. ¹⁸³ Microsoft’s growing control of the OS market led to antitrust scrutiny, which resulted in a 1995 consent decree prohibiting the company’s continued inclusion of exclusionary terms in its licensing agreements with computer manufacturers. ¹⁸⁴ Two years later, the DOJ filed an action seeking to hold Microsoft in contempt for alleged violations of the 1995 consent decree. ¹⁸⁵ The district court denied the petition but granted preliminary injunctive relief and referred the government’s motion for a permanent injunction to a special master. ¹⁸⁶ On appeal, the D.C. Circuit found in June 1998 that “the district court erred procedurally in entering a preliminary injunction without notice to Microsoft and substantively in its implicit construction of the consent decree on which the preliminary injunction rested” and remanded the case to the district court for further action, thus effectively ending the action. ¹⁸⁷

In 1998, the DOJ brought its most significant antitrust action against Microsoft, and it is this case that squarely addressed the issue of divestiture. The DOJ’s action was consolidated with another case filed by the Office of the Attorney General for Washington, D.C. and Attorney General Offices for nineteen states. ¹⁸⁸ In 2000, after a seventy-six-day bench trial, the district court issued its findings of fact. Following several months of unsuccessful settlement negotiations, the court then issued its conclusions of law. ¹⁸⁹ The court ruled that Microsoft had violated § 2 by monopolizing the OS market and attempting to monopolize the Internet browser market, and violated § 1 by unlawfully tying its Internet browser software to its OS. ¹⁹⁰ In a separate opinion, the court ordered Microsoft to divest either its OS business or software applications business, the latter of which was responsible for its Internet browser. Dedicating only three pages to its remedial analysis, the court stated that divestiture was the proper remedy on the basis that Microsoft (1) had engaged in wide-ranging anticompetitive conduct that was undeterred by prior injunctive relief, (2) appeared otherwise likely to continue its unlawful conduct, and (3) had acted dishonestly and in bad faith with respect to prior enforcement measures. ¹⁹¹

On appeal, the D.C. Circuit affirmed the district court’s findings that Microsoft had violated § 2 by maintaining its monopoly on the OS market. ¹⁹² However, the Circuit reversed the lower court’s determination that Microsoft had illegally attempted to monopolize the Internet browser market. ¹⁹³ The Circuit remanded the § 1 tying claim, having found that the district court erroneously applied a per se standard and instructing the court to instead apply a rule of reason analysis. ¹⁹⁴ Having reversed the lower court’s determinations of liability on two of the three federal claims, the Circuit vacated the divestiture order, ruling:

In short, we must vacate the remedies decree in its entirety and remand the case for a new determination. This court has drastically altered the District Court’s conclusions on liability. On remand, the District Court, after affording the parties a proper opportunity to be heard, can fashion an appropriate remedy for Microsoft’s antitrust violations. ¹⁹⁵

Although the Circuit’s vacatur was based on procedural grounds, the panel also commented at some length on the applicability of divestiture. Before a dissection of the court’s remedial guidance, it is necessary to first provide a more detailed explanation of the court’s appellate review regarding Microsoft’s § 2 liability.

The D.C. Circuit affirmed the district court’s ruling that Microsoft’s exclusive dealing arrangements with Original Equipment Manufacturers (“OEMs”), Independent Software Vendors (“ISVs”), and Apple had helped Microsoft maintain its monopoly on the OS market in violation of § 2. ¹⁹⁶ Under these agreements, OEMs were forced to limit user accessibility to other Internet browsers competing with Microsoft’s Internet Explorer browser. Similarly, Microsoft required Apple, which utilized its own OS, to promote Internet Explorer over competing browsers as part of a deal between the two companies to improve the compatibility between Microsoft’s Office products and Apple’s OS. ¹⁹⁷ Finally, Microsoft’s arrangements with Internet Access Providers (“IAPs”) placed limitations on the share of users able to use browsers other than Internet Explorer. At the same time, ISVs were contractually obligated to make Internet Explorer the default browser. ¹⁹⁸

Although the contractual arrangements most directly concerned the market for Internet browsers, the Circuit determined that the relevant terms also helped maintain Microsoft’s OS monopoly due to network effects. As a result of Microsoft’s early dominance in the OS market, software developers prioritized compatibility between their products and Microsoft’s OS. This generated broader access to software applications for Microsoft OS users, which strengthened consumer preferences for Microsoft’s OS and thereby further reinforced software developers’ incentives to continue prioritizing such compatibility. The result was a positive software-OS feedback loop that helped Microsoft maintain its dominance in the OS market. ¹⁹⁹ However, the growing emergence of rival Internet browsers posed a threat to this self-perpetuating circle, due to their ability to run compatible software applications (i.e., applications compatible with the rival browser) regardless of the applications’ compatibility with the underlying OS. But first, these alternative browsers needed to establish a user base sufficiently large to incentive software developers to write compatible products. ²⁰⁰ Microsoft thus violated § 2 by leveraging anticompetitive licensing terms to expand Internet Explorer’s usage share at the expense of Netscape and other competitors, as part of a broader scheme to maintain its OS monopoly. ²⁰¹

The contractual arrangements were not the only basis for § 2 liability upheld by the Circuit, which found that Microsoft had further monopolized the OS market by limiting the cross-platform viability of competitors’ Java programming. ²⁰² Having developed Java virtual machine (JVM) technology (which allows the Java software programming to run) that lacked compatibility with competitors’ OS, Microsoft then entered into “first wave agreements” with ISVs that required the exclusive promotion of Microsoft’s Java technology. ²⁰³ While the court found that Microsoft’s production of JVM technology incompatible with competitors’ OS was not unlawful, it held that the first-wave agreements were exclusionary and violated § 2. Microsoft had also distributed tools to Java developers, putatively to assist their production of applications with cross-platform compatibility while in fact intentionally deceiving them into producing applications compatible solely with Microsoft’s JVM. ²⁰⁴ Furthermore, Microsoft successfully coerced Intel into abandoning its support of a Microsoft competitor’s efforts to develop a JVM with cross-platform computability. ²⁰⁵ Against this record, the D.C. Circuit affirmed the lower court’s finding that Microsoft’s deceptive and coercive conduct was violative of § 2.

Although it upheld § 2 liability for Microsoft’s monopolization of the OS market, the Circuit had no option but to vacate the divestiture order, partially because it was separately reversing other liability findings upon which the remedial order had also been based. As summarized by the Circuit, the procedural reasons for the vacatur were that:

(1) [T]he court failed to hold a remedies-specific evidentiary hearing when there were disputed facts; (2) the court failed to provide adequate reasons for its decreed remedies; and (3) this Court has revised the scope of Microsoft’s liability and it is impossible to determine to what extent that should affect the remedies provisions. ²⁰⁶

Determination of the proper remedy was therefore also remanded to the district court, with directions to reassign the case to a different district court judge in light of the Circuit’s determination that the trial judge had acted improperly in his interactions with the media during the case. ²⁰⁷ The remand did not result in a new judicially directed remedy. First, following the remand order, the DOJ—under new leadership following the inauguration of a newly elected President—announced it was foregoing its request for divestiture and would instead limit its request for relief to behavioral remedies. ²⁰⁸ Shortly thereafter, the DOJ and Microsoft reached a settlement agreement that did not include divestiture. The agreement was largely approved by the district court in November 2002, whose ruling was then affirmed by the D.C. Circuit on an appeal brought by various state Attorney Generals. ²⁰⁹

Although the D.C. Circuit’s 2001 vacatur was necessary for procedural reasons, the Circuit also elaborated on divestiture’s applicability with the express purpose of providing guidance to the lower court. Echoing Alcoa and United Shoe, the court first highlighted Microsoft’s status as a unitary company, noting that the government’s arguments for divestiture relied on cases which largely “involved the dissolution of entities formed by mergers and acquisitions.” ²¹⁰ To that end, the court referenced du Pont for its statement that divestiture “has traditionally been the remedy for Sherman Act violations whose heart is intercorporate combination and control” and emphasized “combination and control” in a nod to divestiture’s nearly exclusive use as a remedy to anticompetitive acquisitions and certain forms of concerted action. ²¹¹

The court cited United Shoe for its proposition that a unitary company “cannot be cut into three equal and viable parts.” ²¹² Relying on Alcoa, and without reference to any specific data or evidentiary support for its reasoning, it explained that the absence of “preexisting internal lines of division” rendered divestiture impossible to implement “without a marked loss of efficiency.” ²¹³ While stating it was not opining on whether Microsoft itself was a unitary company, the court acknowledged Microsoft’s claim that it was “a unified company without free-standing business units” and “not organized along product lines” while referencing Microsoft’s demands for an evidentiary hearing on remedies. ²¹⁴ The court further advised that “if indeed Microsoft is a unitary company, division might very well require Microsoft to reproduce each of these departments in each new entity rather than simply allocate the differing departments among them.” ²¹⁵

Turning to the issue of causation, the court opined that an assessment of the proper remedy “should consider whether plaintiffs have established a sufficient causal connection between Microsoft’s anticompetitive conduct and its dominant position in the OS market.” ²¹⁶ In fashioning the remedy, liability for unlawful anticompetitive conduct does not translate into a basis for whichever relief will generate the greatest degree of competition. Instead, the remedy must be tailored to restore competition to where it would be if not for the defendant’s anticompetitive conduct. Thus, divestiture cannot be imposed merely because the record establishes that a defendant is liable for antitrust violations and that divestiture would produce the most significant increase in competition. For divestiture to be appropriately ordered, the Circuit delineated, there must also be sufficient evidence of a but-for causal connection between the defendant’s anticompetitive conduct and its monopolization of the market, with the corresponding relief limited to that necessary to undo the injuries worked by the unlawful conduct. ²¹⁷

In this regard, as explained in the Areeda & Hovenkamp treatise cited approvingly by the D.C. Circuit, “more extensive equitable relief, particularly remedies such as divestiture designed to eliminate the monopoly altogether . . . require a clearer indication of significant causal connection between the conduct and creation or maintenance of the market power.” ²¹⁸ After acknowledging that the trial court expressly declined to offer a position on whether Microsoft would have still maintained its monopoly on the OS market in the absence of its anticompetitive conduct, the Circuit suggested, “If the court on remand is unconvinced of the causal connection between Microsoft’s exclusionary conduct and the company’s position in the OS market, it may well conclude that divestiture is not an appropriate remedy.” ²¹⁹ Although the court did not reference the AT&T divestiture, its emphasis on causation and degree of evidentiary support for causation align with the mode of analysis set out in AT&T.

The decisions issued in Alcoa, United Shoe, AT&T, and Microsoft are four of the only judicial opinions to offer guidance on divestiture’s application in unilateral conduct cases. Beyond the obstacles to liability for unilateral conduct, the incidence of single-firm divestitures has been limited by judicial aversion to far-reaching economic intervention, particularly where the firm’s organization renders divestiture costly and difficult to implement with little certainty that it will restore competition. Unilateral conduct cases do not typically present markets with a recent record of meaningful competition or ready-made dividing lines to identify divestable assets in the way that merger control and combination cases often do. As a result, divestiture requires stronger evidence of liability in unilateral conduct cases, as well as proof that the defendant’s unlawful conduct is the but-for source of its monopoly power. In merger control and combination cases, on the other hand, the acquisition or association itself is the basis for liability, thereby rendering causation self-evident once liability has been established.

As demonstrated in Alcoa, United Shoe, and Microsoft, courts are particularly reluctant to order divestiture where the firm’s assets are owned or otherwise controlled by a single corporate entity. Divestiture appears more likely to occur where the targeted assets are organized as subsidiaries or auxiliary businesses of the defendant, such as the Limited shares in Alcoa or the regional operating companies in AT&T. In such circumstances, the preexisting divisions of ownership and institutional organization suggest opportunities for division that are not present in a singular, unified entity.

The considerations favoring judicial restraint, coupled with distinct hurdles to liability, have made single-firm divestitures an anomaly for the entirety of Sherman Act enforcement. For the same reasons, absent legislative reform or seismic shifts in the courts’ doctrinal approach, single-firm divestitures will likely remain a generational occurrence, at most.