Competition policy has been no obstacle to the rise of dominant firms in e-commerce, social media, online search and other important aspects of the modern digital economy. The well-documented results of these trends are increasing market concentration, entrenched dominance, diminished competition and entry, and harm to consumers and businesses alike. Competition agencies, policymakers, academics, interest groups, and others have proposed various ways of addressing the weaknesses of past policy. Most of these proposed policies involve more vigorous application of conventional tools, which, however, are unable to address current levels of market concentration.
However, the most obvious solution – breaking up such firms — is generally dismissed as impractical, the equivalent of trying to unscramble eggs. The authors disagree in this paper, available here. The rationale for breaking up companies is straightforward: where the essential competitive problem with a company is its structure, in the sense that its anticompetitive behaviour flows inexorably from that structure and is otherwise difficult to prevent, it follows that the necessary solution lies in altering that structure.
Further, the authors show that there have been substantial numbers of successful breakups of firms. These include several break ups of dominant firms in U.S. antitrust, but also, more recently, break ups of much of the regulated telecom and electricity sectors in the U.K. and U.S. Many of these experiences with actual break ups are in fact successful restructurings, with major widely recognised benefits in terms of price, innovation, and investment. Further, the private sector regularly observes divestitures, which provide lessons with respect to how best to break ups of firms.
Section II begins by reviewing the decline of competition in major economies.
Mergers and acquisitions have contributed directly to increased market concentration throughout the economy. Studies have documented the rise in concentration in hundreds of industries across all sectors of the economy, generally beginning in the 1990s. Increases have also been documented in firm mark-ups, which are indicative of market power.
Simultaneously, the dramatic growth of the five major tech companies–Amazon, Apple, Facebook, Google and Microsoft—has been fuelled by their enormous numbers of acquisitions. This history of mergers and concentration, of tech acquisitions and expansion, only happened because competition policy has been unable or unwilling to contend with these developments. Of the 800 acquisitions by GAFAM since 2000, the U.S. antitrust agencies have investigated only a limited number, challenged exactly one, and ultimately blocked none. The same pattern holds in other jurisdictions. Worldwide, approximately 97% of these tech company acquisitions have not even been vetted by any competition authority. To date, not a single one has been blocked anywhere.
Compounding the problem, once approved few mergers and none of these tech acquisitions seem to be subject to meaningful oversight or remedy. In a few instances, competition authorities have sought to rein in the anticompetitive behaviour of these companies, but these experiences underscore the overwhelming difficulty faced by competition agencies seeking to control the behaviour of large, profitable and tightly structured companies through rule-making and conduct remedies.
Section II also details the inadequacy of traditional tools of competition policy in contending with the forces provoking this decline in competition.
Where competition agencies have acted, their approach has typically been to identify and prohibit the specific offending conduct through rules and conduct remedies imposed on the firms. In the context of merger enforcement, such rules have often been used to permit a merger to proceed while subjecting it to prohibitions on specific anticompetitive conduct made possible or profitable as a result of a merger. Despite there being a widespread view among policymakers that behavioural remedies are inferior to structural divestiture, agencies regularly accept behavioural remedies in lieu or as a counterweight to limited structural remedies.
Attempts at resolving structural and informational problems through the use of conduct remedies have not met with much success, as the authors demonstrate by means of a review of US and EU examples. To the contrary, reliance on behavioural remedies has been one important reason for the persistence and even growth of the dominant tech companies and many anticompetitive mergers. Dominant tech companies have the very properties that make rules and conduct remedies least likely to work. Their product is not a simple, homogeneous, static commodity, but rather complex and flexible, and subject to rapid change due to the underlying technology.
Section III considers a policy of breaking up certain dominant firms and consummated mergers.
The costs of break up is routinely said to be prohibitive. In fact, there is very little evidence of that. The prototypical scare story is the divestment of AT&T, but in two years this company had been broken up into nine companies, including the reallocation of 70 million customer accounts, 200 million customer records, 24,000 buildings, 177,000 motor vehicles and one million employees. Breaking up consummated mergers will likely prove to be simpler, since the necessary action might consist merely of reversing the merger by divestiture.
The authors do not wish to understate the difficulties that may be posed by some break ups. Nonetheless, a successful break up operation can replace an otherwise lost effective competitor and avoid thorny and often ineffective aspects of conduct remedies.
At this stage, the authors provide a large number of examples of the feasibility of breaking up firms, including:
- In the US, an average of three consummated mergers are challenged each year. On most of these cases little operational integration had occurred at the time of challenge, so break ups posed few if any practical problems. Even when operational integration occurred, divestiture has been pursued and the divested entities survived as going concerns.
- In the UK, the competition authority can conduct market investigations, which may lead to the adoption of remedies to tackle market characteristics which adversely affect competition. Among these remedies, one can find the breaking up of firms, which have been used, if only sparingly (e.g. following market investigations into open banking, airports, private healthcare and cement).
Section III looks more narrowly at how to break up dominant companies.
Possible break ups of dominant firms raise different issues depending on whether the competitive problem involves the firm’s related businesses or the core area of its dominance. Related businesses typically involve a vertical relationship, a complementary product or service, or even a potentially competitive product. Often, this related business is the result of a merger or perhaps internal expansion into a distinct business operation. As a result, the “fault lines” between that business operation and the core platform suggest how separation can be achieved. In practice, the process would be much the same as with breaking up a consummated merger.
The alternative possibility is that the competitive problem concerns the core business. In certain circumstances, particularly if markets are likely to tip, breaking up the core business of an incumbent would probably not result in the long-term viability of multiple competing platforms. Instead, while there might be some period of vigorous competition among the rivals, the end-result would most likely be a shakeout of all but one dominant firm. Instead, alternative policies towards firms with core dominance may prove helpful in sustaining rivalry and competition. These include data portability, open access, interoperability, and the like.
The authors then provide a number of examples of dominant companies being broken up, including:
- US antitrust cases that broke up dominant companies, such as Standard Oil and American Tobacco (1911), and, more recently, AT&T and Microsoft (overturned on appeal). In addition to the remedy adopted, it is noteworthy that some type of anticompetitive behaviour was necessary to trigger the imposition of the break-up remedy.
- Mandated separation of dominant companies in regulated industries worldwide. In regulated sectors – particularly in industries such as telecommunications, electricity, and railways, characterised by a vertically-integrated incumbent operator that provides an essential input (access) to its retail rivals –, breaking up incumbents has been consensually accepted as part of the policy toolkit. In addition, these breakups have occurred without prohibitive costs or permanent damage to the companies. If anything, the problem has been the lack of perceived effectiveness of break-ups, which has meant that ex post regulation was still necessary.
Section V looks at self-initiated private break ups (personal note: I think a most common term for this is ‘business spin off’).
Break ups initiated by companies are quite common—nearly as frequent as mergers and acquisitions. One compilation of 86 of the Fortune 100 companies reported a total of 2,307 mergers and acquisitions in the 1990s, and fully 1,611 divestitures during the same period. The average company engaged in about three acquisitions and two divestitures per year. A considerable literature is devoted to evaluating the relative merits of restructuring alternatives. What is relevant about these experiences for our purposes is that they provide guidance on how best to engage in divestitures – i.e. on how divestitures are designed and implemented, and what have been the correlates of a successful and efficient process.
A number of commonalities regarding divestments can be identified in this literature. Break ups often occur along the fault lines marking different parts of the company. Furthermore, successful spin offs require a full package of assets in order to succeed post-divestiture. Studies have identified five different areas requiring attention and action in support of successful breakup. These include business unit strategy, target P&L, operating model, processes and talent, and culture.
Section VI provides recommendations on the practical aspects of a policy of breaking up consummated mergers and dominant tech companies.
There clearly is ample precedent and experience with breaking up dominant firms. Most breakups seem to have resulted in structurally more competitive markets and stronger competition. Strikingly, there seem to be no examples where breaking up a dominant firm has been attempted but failed, in the sense that this was attempted but literally could not be done, or permanently damaged the firm as a going concern in the process. Nor are there obvious examples where break ups were accomplished but led to harm in market competition. That remarkable fact by itself suggests that a break up policy is viable, that procedures to enact it are adequately understood, and that some measure of success is a plausible outcome.
Reflecting this, the authors identify a number of principles for the practical success of break up policies. In short, they amount to selling a viable business – either by reversing a merger or by selling a business unit with a bundle of assets to ensure its viability.
If you would like an overview of divestitures in the EU and the US, this is a nice starting point. Of more relevance to the authors’ objectives, the paper provides a neat argument for considering divestment remedies alongside other options – both in merger and in antitrust cases.
However, I think the paper needed to go further to fulfil its ambitions.
The authors first argue for the need to adopt divestment remedies more often, because failing to do so has led to increased market concentration. However, it is not clear to me – in general, and from this paper – that mergers are the sole or even main cause of this increase in concentration. Should mergers be the cause of market concentration, then a divestment remedy would seem to be a poor alternative to simply prohibiting them.
The authors also explain that the current state of affairs of market concentration is a consequence of no merger involving a dominant platform ever having been prohibited. While this implicitly suggests that market power in tech is at least partially the result of anticompetitive mergers, the reality is that either the cleared transactions did not involve any competition issues, in which case no remedy is called for; or the competition community at large has been played, in which case one can doubt the wisdom of now empowering it to start breaking up firms left and right.
In any event, breaking up companies ex post also strikes me as a rather poor remedy for the competition evil that the authors are tackling. In itself, market concentration is not a competitive harm unless it is obtained illegitimately (e.g. through a merger or anticompetitive conduct). Competition authorities cannot break up a company unless an illegitimate act occurs. For example, I can think of a systematic pattern of purchases to preclude competition amounting to an infringement of competition law – however hard this may be to prove -, in which case divestments may be suitable as a remedy. Nevertheless, the authors seem to go beyond merely arguing that structural remedies should be used in such circumstances, to argue that break ups could be considered whenever market concentration becomes excessive. Absent specific powers to that effect, as under the UK’s market investigation regime, this is simply not an option open to agencies. The examples the authors identify in regulated sectors are, well, about regulated sectors – which used to be state owned at that.
Further, breaking up individual mergers following ex post analysis is unlikely to dent market concentration in the digital sector. In effect, and as I have argued elsewhere, if there is an issue here it probably lies with patterns of systemic (but lawful) acquisitions. Such a problem requires a systemic response. Now, breaking up companies may well be a suitable part of this response, but it would not be applied in the context of competition law as we usually understand it. Instead, it would be about remedying unacceptable levels of market concentration, wherever and regardless of why it arises – and would probably involve state-mandated restructuring of whole (digital) sectors.
In addition, market concentration seems to be a natural consequence of the structure of digital markets. In such a scenario, breaking up companies is unlikely to prove a particularly effective tool. The authors acknowledge that ‘it is widely understood that […] digital markets present a combination of forces–high economies of scale and scope, tipping, and network effects–that drive the core platforms toward dominance’. However, they dismiss this by focusing on ‘how their mergers, acquisitions, and expansions have exploited, extended and cemented their dominance and thereby harmed competition’; and by correctly limiting the scope of this remedy to peripheral/related businesses.
In short, and given this, it strikes me that a thorough discussion of break up as an effective remedy to the problem of market power digital markets needs to be coupled with a discussion of how such a remedy would relate with the regulatory frameworks that should govern ‘natural’ market power. Hence, I am surprised that the authors so easily dismiss rules and conduct remedies – which they seem to have unnecessarily conflated with behavioural remedies imposed on infringing firms by competition agencies, as opposed to behavioural regulation more widely considered.