This paper attempts to reconcile the goals of competition and equality within antitrust, and suggests that this task corresponds to John Rawls’ efforts to reconcile liberalism and equality within his principles of justice. Applying a Rawlsian analysis, I propose the adoption of inclusivity as a secondary objective within competition law (not as an additional primary objective, as under a public interest test), and suggest that this approach might be labelled Rawlsian Antitrust. In locating this approach among the existing established schools of thought, I emphasize both its economic basis, and the clarity of its values, and how the reconciliation between those might be operationalized. I identify pro-enforcement and more cautious factions within this approach, and suggest that between them lie the tools that will hardwire inclusivity into competition policy. These will help make competition policy play its role within a coherent policy response to the need for greater inclusivity, arguably the great challenge of our time, which was first stirred, and then made urgent by successive crises that have shaken our world, and which now threaten to burn down the technocratic antitrust chateau.
By Chris Pike1
I. INTRODUCTION
As the full impact of COVID-19 continues to shake our world, we find ourselves retreating to our homes, and unsure of the long-term implications for our health, our societies and our economies. However, in competition policy, as in other policy areas, we are at the same time, still observing the effects of the aftershock of the last financial crisis, and are only just beginning to see the emerging outline of a realignment triggered by that shock. Whether this has time to solidify, or is reinforced, or indeed washed away by the impact of COVID-19 on the debate remains to be seen at this stage.
However, what is evident is that we are in a moment where globalisation and digitalisation, and the policy response to these phenomena have helped to deliver huge improvements to our lives, but also huge disruption and a degree of fragility that these twin crises have ruthlessly exposed. This had unfortunately already led to a loss of trust in the efficiency and fairness of markets, even before COVID-19.2 This loss seems to be visible in a number of areas.
Firstly, there is rising resentment at the inequality that our markets have generated, both in terms of income and wealth inequality, and specifically in the interplay between that and racial and gender-based inequalities that result from structural discrimination.
Secondly, there has been a dramatic productivity slowdown that some have attributed, fairly or not, to the growth of market power,3 and which suggests that the palpable unfairness of outcomes has not even managed to buy us greater efficiency.
Thirdly, there is a popular feeling that our markets, and our meritocracy, are in fact rigged.4 Certainly, the bail-out of banks during the financial crisis (however necessary for systemic stability), sent a message that some would be protected from the market forces that were deemed inescapable in other industries. However, the same conclusion has been drawn by those observing the support provided to state-owned enterprises in China, and the low tax rates paid particularly by large digital firms.5 Absent recognition of a truly level-playing field, the legitimacy of market mechanisms is therefore inevitably weakened.
Then, fourthly, as we struggled (painfully under the weight of misguided austerity measures) to slowly emerge from the wreckage of the financial crisis, we found that the quickly growing digital economy and its superstar firms, seemingly our best ticket for future prosperity, was also creating huge complications. We watched as it increased the precarious nature of work for many, manipulated our attention and distorted our public debate, and increasingly monopolized markets, creating concentrated centres of power (both market power and lobbying power). This is nowhere more evident than in the threatening behavior that certain firms have unleashed against nation states, most recently in Australia.6
In parallel to all this, over the decade, recognition of the sustainability problem has finally turned into a much broader-based alarm at the climate emergency, and the apparent inability of markets and market-based policymaking to address it.7
Taken together this loss of trust has resulted in a damaging shift towards over-simplistic and ineffective populist answers to often-difficult policy questions. Walls, tariffs, anti-immigration, isolationism, national champions, market share caps, price caps and nationalisations have bounced back onto the agenda. This in turn has created a threat both to our shared prosperity, and in some cases even to our democracies.
Against this background of a loss of trust in markets, the role of antitrust is perhaps a modest one. However, alongside most other market-based policy fields, it finds itself buffeted by the same winds, and so its future path will be shaped by the way that these forces, and the debates that they stir, ultimately play out.
For instance, we have already seen the re-emergence in the U.S. of the Brandeisian structuralist school of antitrust, who reject even the Harvard school approach as an inadequate countervailing force to the influential Chicago perspective. In contrast, in Europe, we have the longstanding ordo-liberal school, and the more economic approach (which includes, but should not be confused with, a Chicago faction), and a greater openness to regulatory solutions. That openness is for instance particularly evident in the current debate over competition in digital markets, where the neo-Brandeisians and the Chicago school find themselves sometimes aligned in opposing the type of pro-competitive regulations to which Europe is turning (albeit for different reasons).8
As these different groups collide across zoom panels and social media, the goals of antitrust once again find themselves centre-stage. The uneasy agreement on consumer welfare was always loosely drawn in a fashion that enabled many different groups to each see their own position reflected within it, but this has been upended by the direct challenge posed by the neo-Brandeisian structuralists. In addition, the economic nature of antitrust law, the role of worker welfare, the impact of non-controlling ownership of a firm, the evidentiary standards used, and the proportionality of different remedies have all been opened up for debate.
II. WHERE THEN DOES INCLUSIVENESS FIGURE IN ALL THIS?
Firstly, it can be found anywhere. Certainly, those in the structuralist school are concerned about inequality. However, while Chicago-school antitrust may sometimes coincide with a distaste for any policy with a re-distributionary flavor, there are Chicago-school thinkers that are happy to see re-distribution through fiscal policy (which they see as the appropriate policy mechanism for addressing the concern), while arguing against the use of antitrust to pursue such goals. Many of these might even welcome the re-distributionary consequences of the reduction of rents that the OECD has identified as an inevitable consequence of an effective antitrust policy,9 though they would presumably see this as a perhaps fortunate side-product that should be studiously ignored, and certainly not prioritized.
However, for some, inclusiveness can, or should, be located at least in some form within the goals of antitrust. Inclusive antitrust (the title of this collection), must therefore include each of these different groups, which I attempt to describe below.
For some, inclusive antitrust might mean the use of public interest tests to ensure that competition policy delivers inclusiveness, regardless of the impact on competition, efficiency, or consumer or citizen welfare. This is evident both in South African antitrust, and indeed in many OECD countries’ media antitrust policies (which prize plurality above all else).
For the Neo-Brandeisian school, inclusive antitrust will mean a structuralist approach, which looks to protect consumers, workers and small businesses, and to guard against inequality by preventing the concentration of power, regardless of whether that is market power, or a more nebulous concept of corporate power (which we discuss further in section 4).
For some in Europe it will mean an ordo-liberal approach that looks to protect the equal opportunity for smaller firms to compete, regardless of effect.
Meanwhile in the U.S. it might still mean a Harvard school approach that worries about market power rather than corporate power, but which looks to prioritize the protection of consumers (or sellers, in labor and other purchasing markets) by assuming that such market power will be misused.
Finally, Inclusive Antitrust can also mean an approach that is economic, but which explicitly sets out inclusiveness as a secondary objective in antitrust policy. Here an inclusive secondary objective simply means achieving the primary objective in the most inclusive way possible. Indeed competition can itself be a secondary objective for some regulators, such as the Bank of England, who seek to achieve their primary objective of financial stability in the most pro-competitive way possible.10]
This last approach is a strand within the post-Chicago and ‘more economic’ schools of thought, and therefore focuses on the effects on welfare.11 This welfare, as Hovenkamp (2019a) sets out however, is not simply consumer welfare (which he identifies as badly mislabelled and hence falsely equated with consumer surplus), but also includes the welfare of workers to the extent that this is inefficiently restricted (for instance through monopsony market power).12 This might therefore be more helpfully referred to as an ‘efficient citizen welfare standard.’13
Therefore, this approach is effects-based, but conscious of, and for some, predominantly concerned about, the risk that under-enforcement will lead to anticompetitive effects, and the use of market power against consumers, workers and other sellers in purchasing markets. For example, for some, this concern reflects the view that under-enforcement errors tend to have a higher inequality cost than over-enforcement errors. This is because under-enforcement errors result in higher prices (or poorer quality) for low-income purchasing groups amongst others, rather than resulting in missed opportunities for the owners of firms to reduce their costs (which may or may not be passed on to low-income purchasers).
For others however, there is no focus on concerns over under-enforcement, and for them the secondary objective can manifest in terms of an appetite to prioritize resources and nuance the detail of its analysis in order to maximize its impact on inclusivity. In practice this approach can already be discerned in the OECD’s pioneering work on a gender inclusive competition policy, in the CMA’s explicit strategic focus on vulnerable consumers, and, as Pike & Santacreu-Vasut (2019) and the Acting Chair of the U.S. FTC, Commissioner Slaughter, have flagged, it also offers a promising means to reflect anti-racism efforts within antirust policy.
III. A ROLE FOR RAWLS?
Indeed, what I want to suggest here is that this last, as yet nameless approach, in seeking to reconcile the goals of competition and equality within antitrust, can be seen to correspond to the efforts that John Rawls made to reconcile liberalism and equality within his principles of justice, and that this approach might therefore be labelled Rawlsian Antitrust. This would distinguish it from the broader Post-Chicago school, and the “more economic” approach, each of which are also economic effects-based, and which also recognize the risks of under-enforcement, but which, in their orthodox forms remain strictly technical, in that they do not necessarily have a view on inclusiveness, nor on how to balance the risk of under or over-enforcement. Indeed, some of the eminent post-Chicago scholars have expressed reservations about reflecting concerns over inequality within the goals of antitrust.14
In contrast, as the Acting Chair of the U.S. FTC has eloquently set out: “I do not believe antitrust can be value-neutral… When we make decisions about whether and where to enforce the law or how to deploy our enforcement resources, we are making decisions that will have an effect on structural equity or inequity. Our decisions can either reinforce existing structural inequities or work to break them down.”15 In this context, a post-Chicago economic approach that has nothing to say about inclusiveness is not value-free or value-neutral, but rather its values are focused solely on efficiency (maximising the size of the pie) or perhaps on maximising total consumer surplus (the size of the pie that consumers as a whole obtain). This is a perfectly coherent set of values, but it is a set of values nonetheless. The question may therefore be posed: which set of antitrust values best reconcile liberalism and equality?
On this question, those of an orthodox post-Chicago position would profess indifference, at least within the context of antitrust, or point instead towards fairness of processes (see for instance Akman, 2018).16 However, other schools of thought within what might generally be labelled as Inclusive Antitrust, would have their own views on how to do so, and we will come to that in section 4, but first, let us elaborate on what I have labelled Rawlsian Antitrust.
To begin with let’s take Rawls’ thought experiment, which was to ask us to consider which principles we would use for the basic structure of society, if we had no knowledge ahead of time of what position we would end up having in that society. This means we have no idea of our ethnicity, social status, gender or of our individual perspective on how to lead a good life. We do not know what we consume, where we work, if we work and we do not know if we benefit from anti-competitive rents of any form.
For Rawls, the answer was that we would set out three principles: firstly to achieve the greatest equal liberty for all; secondly, to achieve equal opportunity, and thirdly, to deliver the greatest benefit to the least advantaged members of society, consistent with what he called a “just savings” principle. How might these principles apply to Antitrust and Competition Policy?
A. The Greatest Equal Liberty Principle
First, we might interpret the greatest equal liberty principle as requiring that when operating within markets, people should have access to the best value possible in terms of price, quality and innovation, certainly in respect of primary goods, since the cost of these eat into an individual’s ability to invest in themselves, their community and their experiences.17,18 This might mean prices should be low, quality should be high, innovation should be plentiful, and wages should be at the marginal revenue product of labor. The greater value offered by such markets would give individuals the greatest freedom to achieve their goals and live a fulfilling life. This might correspond to the ‘efficient citizen welfare’ standard that is described above.19
In traditional selling markets, this citizen welfare standard equates to a consumer welfare standard, but in purchasing markets, it would mean protecting the welfare of workers and small businesses to the extent that this does not reduce efficiency. As many have now recognized, the common assumption that any and all cost-cutting by firms is efficient, is demonstrably false, and the exercise of monopsony market power against workers of suppliers is an efficiency risk that must be guarded against. Equally, however, a total surplus standard is unacceptable since it would credit the anti-competitive rents that incumbents accumulate when they improve their ability to exploit buy or sell-side market power.
B. The Equal Opportunity Principle and the Difference Principle
We might then interpret the equal opportunity principle (2b) as meaning that, in addition to the equality of opportunity for individuals that Rawls focused upon; there is also an equal opportunity to compete. This might be taken to mean that competitors should be free from being excluded from the opportunity to compete on the merits (as is standard under antitrust law), and that there must be a level playing field (as set out in the OECD’s recently agreed recommendation on competitive neutrality).20
However, in antitrust these are each instrumental offences which are problematic because they threaten to cause an increase in prices (or lower quality, innovation or wages) and hence these are already likely to be inconsistent with the greatest equal liberty principle. For example, conduct by a firm (or a regulator) that excludes rival black or female entrepreneurs from apparently highly competitive markets might be of concern because it risks having an effect on market outcomes, for example by reducing innovation, alternative perspectives, and the introduction of products that cater to under-serviced communities.
We might interpret this principle as suggesting that exclusionary conduct and non-level playing fields are not only instrumentally important, but that they can also be seen as intrinsic offences against an equal opportunity principle for entrepreneurs.21 This would mean that exclusion from a highly competitive market that had no effect on market outcomes, would still offend this principle due to its impact on the opportunity of the entrepreneur (so not because of the effect on consumers, but because everyone deserves the opportunity to make the most of their talents). Notably this might be particularly important for entrepreneurs from disadvantaged groups that might face barriers to entry that prevent them enjoying the same opportunities that others have (even if these may not always translate into harm to consumers’ and workers’ liberty).
However, if there is no effect on market outcomes, then this takes us outside the primary goal of antitrust, and into the possible secondary objectives. We are therefore in the realm of policy, advocacy and prioritisation. More importantly perhaps, we will often arrive at this question of secondary objectives in the all-too-common cases of uncertainty over effects. Specifically, the heightened uncertainty that exists before an agency has decided to devote resources to the case or market in question.22 At this stage, the effects and hence the strength of the case under the primary objective is often highly uncertain. Competition agencies therefore have to make decisions over which cases to take, and which markets are worth studying, all with little or no information on the effects at stake.
A Rawlsian approach to these problems would be to propose that where the potential effects are not significantly different, these decisions be made on the basis of a secondary inclusiveness objective. For example, which cases are likely to improve equal opportunity, and if that does not help, then on the basis of which cases are likely to deliver the greatest benefit to the least advantaged members of society (the difference principle). This “difference principle” (or maxi-min strategy in game theoretic terms) would therefore require that when there is great uncertainty over effect, a focus should be placed upon those cases in which any reduction of anti-competitive rent would be likely to have the most re-distributionary consequence (if it were to be substantiated).
Note that none of this is to disagree that equal opportunities and re-distribution are each policy goals that require their own specific policies and tools, and that those dedicated policy tools offer the most effective way of achieving those goals. However, this is not to say that competition policy cannot and should not contribute to forming a coherent policy response to those objectives by defining equal opportunities and re-distribution as secondary policy objectives. For example, as noted, achieving financial stability requires its own policy tools, but as part of a coherently pro-competitive policy response across the breadth of government, it does help that secondary competition objectives are defined for regulators that operate the policy tools that focus primarily on delivering financial stability (and indeed other public policy objectives).23
C. Summary
In summary, this would suggest that the primary objective of competition law should be to deliver greater liberty for all through lower prices, better quality, efficient wages and more innovation. But that secondly, given the often high degree of uncertainty prior to opening an investigation or market study, that it should then seek to deliver an equal opportunity to compete on the merits on a level playing field, and a redistribution towards the least advantaged members of society.
Taking this further, this might also lead us to ask whether decisions over the outcome of a case should ever be made on the basis of a secondary objective? Arguably, this would not be necessary as after an investigation the agency will always be in a position to make a balanced (though not infallible) assessment of the expected effect.
However, if the secondary inclusivity objective is easier to assess than the primary objective, then there might be a good case that the secondary objective should be used to establish a presumption, or to inform the evidentiary standard that applies. For instance, in a market like pay-day loans, if there is any anticompetitive effect from an abuse of dominance or a merger, then the inequality cost of that behaviour or agreement will clearly be extremely high (since the anti-competitive rent would inevitably be redistributed from some of the poorest in society towards the shareholders of pay-day loan firms). In such cases where the inclusivity impact is high, a rebuttable presumption of an anti-competitive effect might help to ensure that the benefit of any doubt sits with the marginalised users. Similarly, a rebuttable presumption of an anti-competitive effect might be adopted in cases where that would improve equality of opportunity, if a competitive effect were indeed to be shown. For example, behaviour that allegedly excludes or discriminates against minority-owned SMEs might be identified as having a high potential inclusivity impact, and hence begin with a rebuttable presumption of an anti-competitive effect. Crucially however, the test should remain one that is based on effects, since this determines whether the primary objective (and the first principle of greatest equal liberty) is met or not. However, as has become clear in the debate on the risks of over and under-enforcement, an effects-based test need not start from a presumption of no effect.24
Alternatively, if the secondary inclusivity objective is more likely to be harmed by under-enforcement than by over-enforcement (as suggested in section 2), then the secondary objective might justify agencies erring on the side of over-enforcement. For example, in practice this might either mean targeting a non-zero loss rate in court, or the adoption of a rebuttable presumption of an anti-competitive effect as the starting point for the analysis.
Finally, we might also note the important caveat over the just savings principle. This is used by Rawls to require that respect is left for future generations, and that redistribution should not ignore the intergenerational dimension. This might therefore be interpreted as creating precisely the type of sustainability as a relevant efficiency defence that Holmes (2020) and others have identified as an important component of an inclusive competition policy.25
IV. WHAT WOULD OTHER SCHOOLS OF THOUGHT MAKE OF THIS?
Naturally, rivals schools of thought will disagree with the reasoning of those that might subscribe to this Rawlsian Antitrust school. However, might they also dispute the claim to the label?26
For instance, structuralists and supporters of a public interest might each well contend that they also share Rawls’ concern for inequality. However, fundamental to Rawls’ principles is the idea that equal opportunity takes priority to reducing inequality. This suggests that the opportunity for a firm to operate more efficiently without damaging the liberty of others by raising prices (or reducing quality) should be permitted, even if it increases inequality (though redistribution might nevertheless occur via fiscal policy). This principle would therefore clash with the application of public interest tests that seek to prevent mergers that result in efficient rationalization of the workforce (that is, job losses that do not flow from monopsonistic behavior).
Now those favoring a public interest test might retort that equal opportunities should be protected even if they result in higher prices and lower quality for all (that they should not be a secondary objective). For example, providing subsidies or asymmetric regulation that helps to provide equal opportunities for historically disadvantaged persons as entrepreneurs even if this raises prices (or reduces quality) for historically disadvantaged and other persons as consumers or workers. This is certainly a coherent position, and one that Michael Sandel might well agree with given his critique of the importance that Rawls places upon meritocracy. However, it would not fit within this analysis because equality of opportunity is second order to the greatest equal liberty principle, and so would not be traded-off against it.27
However, other public interest tests might be adopted in order to protect more fundamental liberties that also sit within, and perhaps higher within, the first principle of greatest equal liberty. For example, as noted, public interest tests for plurality in media markets are commonly part of competition policy in many OECD countries. This therefore perhaps reflects an acknowledgement that there are more fundamental liberties (in this case the fact that preserving a plural media is a necessary condition for freedom of speech) that are more important than obtaining better value. As a result the primary objective of obtaining better value is displaced, not by one of the suggested secondary objectives (equal opportunity or the interests of the least advantaged members of society), but by another aspect of Rawls’ first principle, the greatest equal liberty.
The same is true in relation to national security, which again constitutes a defence of the greatest equal liberty principle. How policymakers should decide which parts of the economy should be governed by the need for better value (the primary objective for competition policy) and which should be governed by these different aspects of the greatest equal liberty principle is beyond the scope of this paper. However, it is not difficult to imagine that the categorisation of absolute, limited, and qualified rights that are used within human rights law might be helpful in sub-dividing Rawls’ greatest equal liberty principle.
Finally, those with a more structuralist approach might also feel their perspective to be Rawlsian in some respects, since they too are concerned for justice and fairness.
Indeed the harm created by corporate power can be important for marginalised consumers and workers even if it does not stem from an ability to increase rent by restricting substitutability. For example, mergers that increase the merged entity’s ability or incentive to lobby might have an indirect effect on substitutability and market power by foreclosing rivals in the same way that mergers that make a firm too-big-to-fail can (see Cowgill et al, 2021 on the first, and Hellwig, 2017 on the second of these). However, in each case the lobbying power might also harm consumers (or workers) via non-market power channels. For instance, prices might rise (or quality or wages fall) because a merger increases the systemic importance of a firm (perhaps by increasing its importance as an employer within an area, its influence on political debate, or the contagion risk that its failure poses to others), each of which might increase its bargaining power with government.28 In such cases, the creation or abuse of this lobbying power might allow a firm to extract rent, for example by reducing the protection from exploitation that is offered to consumers or workers by regulations.
However, a non-rebuttable structuralist approach would for example focus not on the effects of a merger, but more narrowly on the change in structure, e.g. the HHI (albeit in the expectation that this would often have an adverse effect on different stakeholder groups). This would therefore entail restricting a firm’s scope to act to attempt to improve its efficiency (e.g. by merging to obtain economies of scale), even when the specific evidence of the case suggests this would not lead to higher prices, lesser quality and innovation or inefficient wages. It would therefore involve action against mergers (or conduct), the effects of which are not expected to create a realistic prospect of harm, and hence are unlikely to reduce liberty.
If instead, a legal framework for addressing such non-market power concerns were available, then economists would be able to test different potential theories of harm that stem from the creation or abuse of lobbying power (as opposed to market power or dominance). This would allow them to work towards distinguishing the harmful from the harmless rather than simply prohibiting size. After all, it is not clear that a structuralist approach that sets bright line tests to limit a firm’s size and which cannot be rebutted would achieve either of the secondary objectives discussed above, since like price, quality, innovation and wages, neither are straightforwardly correlated with market structure.
V. WHAT WOULD RAWLSIAN ANTITRUST MEAN IN PRACTICE?
It might be argued that in practice we already have a cautious and limited version of Rawlsian Antitrust. For instance, as Bill Kovacic points out, we do not typically see competition agencies focusing their limited resources on anti-competitive conduct in luxury yacht or sports car markets.29 Indeed as noted, in the UK, the CMA, since the financial crisis, has explicitly targeted markets that matter to those vulnerable consumers and those on low incomes.
However, as recent examination of the treatment of gender and racial inequality within antitrust makes clear, this has for most agencies not extended to other disadvantaged groups. There would therefore seem to be considerable scope for prioritising resources and cases for those markets that matter to the least advantaged members of society. As I have argued elsewhere (Pike & Santacreu-Vasut, 2018 and 2019), for gender inequality there is a strong case for focusing upon child, elderly and social care markets, and predominantly female labor purchasing markets, as well as the variety of markets that impact upon unpaid work within the home (the burden of which falls disproportionately upon women across the world). There will no doubt also be a list of important markets that might be tackled by those seeking to address racial inequality. Again this is might include some labor markets, but might also include certain housing markets, in addition to other markets that might be disproportionately important for particular minorities, either as consumers, as entrepreneurs, or as workers. Engaging with inequalities researchers here might offer promising insight into which services or capabilities these communities are lacking, and which they rely upon but might overpay for, or suffer the provision of poor quality products (for instance in publicly-funded markets these communities are likely to suffer poor quality rather than high prices).30
Other proposals have included the idea that within an investigation a gender or racial lens should be used to ensure that due consideration is given to differences in preferences and behavioral biases between disadvantaged groups and the average consumer. This is therefore not about case selection or prioritisation, but instead an analytical measure that can result in a more careful analysis of markets that matter to these groups. For instance, it might result in different market definitions being adopted, the identification of different competitive constraints between products for different groups, and the need for different remedies.31
Finally, in recent years there has been increasing recognition of the case for a recalibration of the relevant tests and evidentiary standards for antitrust intervention. This has been proposed in digital mergers (CMA, 2020, Klobuchar, 2021), start-up acquisitions (Valletti, 2018, Motta & Peitz, 2019), potential competition cases (Caro de Sousa & Pike, 2020, Simons, 2020), all merger cases (Baker et al, 2020, Delrahim, 2021) and in anticompetitive conduct cases (as established under the new German competition law amendments, and under the EU’s proposed Digital Markets Act).32 These are motivated by past under-enforcement, and a view that the established standards of evidence and burden of proof have been set up in a way that gives the benefit of the doubt to a narrowly drawn shareholder base in order to obtain the benefit of possible (often speculative and illusory) efficiencies, rather than to consumers and workers.
Rawlsian antitrust might therefore be seen to include both a pro-enforcement wing, and a more cautious wing. The pro-enforcement wing see scope for the adjustment of the relevant tests in order to facilitate more active enforcement. They might also consider that under-enforcement errors tend to have a higher inequality cost than over-enforcement errors, and therefore see the need for a thumb on the scale in favor of consumers. In contrast, the more cautious wing might be keen for competition policy to obtain the double-dividend of both greater efficiency and an impact on inclusiveness, but are broadly content with the system as is, though many would like to see faster decision-making and more effective remedies when problems are identified.
VI. CONCLUSION
This paper argues that a Rawlsian approach to reconciling inclusivity and competition would suggest that the primary objective of competition law should be to deliver greater liberty for all through lower prices, better quality, efficient wages and more innovation. But that secondly, greater inclusivity should feature (ideally in legislation) as an explicit but secondary objective of antitrust and competition law (not as an additional primary objective, as for example under a public interest test). Greater inclusivity here would mean an equal opportunity for marginalised groups to compete on the merits on a level playing field, and a redistribution towards these less advantaged members of society.
This approach would be based on an economic understanding of the competitive effects of a particular behaviour, and a clear set of values, and would be transparent on how the reconciliation between those might be operationalised.
The paper also argues that this approach would mean clarifying that the consumer welfare standard does not, or should not, simply identify harm to consumer surplus, but should also guard against inefficiently low wages for workers or sellers in input purchasing markets. While the paper labels this an “efficient citizen welfare standard,” in substance this aligns with Carl Shapiro’s recent description of a “protecting competition standard.”
The paper identifies the existence of both pro-enforcement and more cautious factions that may co-exist within a Rawlsian approach. Depending on the faction, you might therefore expect to see those from this Rawlsian school advancing different proposals for reform or for enforcement in order to deliver a more inclusive competition policy. However, what they would share would be a belief that inclusiveness is a welcome product of competition policy, and that, subject to the need for an effects-based analysis (perhaps with differing rebuttable presumptions) that prioritizes value, innovation and the other primary goals of antitrust, this side-product of greater inclusiveness should be maximized to benefit the most disadvantaged within society.
In contrast, you would not find them supporting public interest tests for non-absolute liberties (e.g. employment), or non-rebuttable structuralist presumptions that disregard the expected effects of a specific conduct or merger. In this respect, their values would lead them to seek to reconcile inclusiveness with the need to use the evidence available to help identify whether anticompetitive practices can be expected to result in higher prices, lower quality, lower wages or less innovation. The hope would be that this reconciliation might then help to make a small but perhaps significant contribution to achieving greater inclusivity, which is arguably the great challenge of our time, and which was first stirred, and then made urgent by successive crisis that have shaken our world, and which now threaten to burn down the antitrust chateau.33
1 The author is a competition expert, formerly of the OECD Competition Division, and an associate at the Centre for Competition Policy at the University of East Anglia. Corresponding address: chris.pike.wh@gmail.com.
2 Edelman, Trust Barometer, 2020.
3 Barkai 2017, De Loecker et al. 2020, Gutiérrez and Philippon 2017, Philippon 2020.
4 See Wolf, why rigged capitalism is damaging liberal democracy, 2019, and Reich, The System: Who Rigged It, and How We Fix It (2020). As noted, this can also be seen in the evident meritocratic failings within the job market that Sandel (2020) and others have identified as resulting from entrenched privilege, whether it be the result of private or selective schooling (see OECD, 2019) or unpaid and hence highly exclusive internships. All of which has become visible within the data on declining social mobility.
5 See for instance the lower tax rates obtained in exchange for decisions to locate headquarters and facilities, the advantages of not being liable to pay the business rates paid by rivals on high streets, and the selling of products that are sourced from off-shore sellers. See also the tax advantages of using misclassified “self-employed” contractors.
6 https://www.theguardian.com/media/2021/jan/22/google-threatens-to-shut-down-search-in-australia-if-digital-news-code-goes-ahead.
7 This is not to suggest that such a critique is necessarily fair, but to identify that it exists and is driving a loss of trust. In fact, while markets left to their own devices have shown themselves unable to address the issue (unsurprisingly given the inevitable market failure), many would argue that market-based policy solutions have not been adopted on the scale or magnitude that would have been required in order to succeed.
8 See Chilson, 2020.
10 Baker and Salop (2016) consider the possibility of introducing an inequality objective alongside consumer welfare and efficiency, however they identify a number of problems regarding trade-offs that would arise in doing so. While they are therefore sceptical of including inequality as an objective, introducing inequality or inclusivity more generally as a secondary objective might offer a way to address the difficulties they idenitfy.
11 Elsewhere within the post-chicago school of thought you can also find what Hovenkamp (2007) refers to as a chastised (by Chicago) Harvard school approach (located around the work of Areeda and Turner) which differs from early-Harvard school, but which has arguably been successful in capturing antitrust decision making in the courts and the enforcement agencies (See Kovacic, 2007, Hovenkamp, 2007, and Crane, 2009). This chastised and yet dominant Harvard school approach has therefore worked in what Kovacic (2007) describes as a double-helix with the Chicago school to deliver the low-enforcement antitrust environment that we’ve seen in the US. Shapiro (2021) describes this approach as Chicago-school (thereby emphasising its influence within the double-helix) and argues that it has failed to rein in market power, and therefore requires strengthening, he therefore proposes a modernist approach, rather than a populist (by which he refers to the neo-Brandesian or early Harvard structural approach). Shapiro’s modernist approach is broadly recognisable as a tougher post-Chicago approach, and like Hovenkamp’s (2019b) ‘Technical Antitrust’, it is sympathetic to what Hovankamp calls ‘collateral interests’ such as full employment and more egalitarian wealth distribution, but is unwilling to put these within the primary objectives of antitrust. The secondary objective status that this paper proposes for these inclusive objectives might therefore find favour amongst those attracted to Shapiro’s modernist approach and Hovenkamp’s Technical Antitrust.
12 This is therefore distinct from the public interest considerations that capture worker welfare more generally, such as those used with South African competition law.
13 This is intended to convey that the ordinary citizen’s welfare is the focus, but that there is also an efficiency requirement upon that the defence of that welfare, and therefore that citizens should not expect that their welfare will be generally protected. For instance, a worker’s welfare might suffer from the efficient removal of duplicated via a merger, in such a case the inefficient welfare would not be protected. Equally, a consumer’s welfare might be reduced by efficient price discrimination that increases the welfare of other purchasers (who are able to purchase), the welfare lost to such price discrimination would not be protected. Unfortunately, this label may conjure an image of an ultra-efficient robotic citizen, and this will no doubt further antagonise those that enjoy criticising the assumption of a hyper-rational agent. However, on balance, in the present context the risk of continued confusion over the consumer welfare standard is the more important to address.
14 Shapiro (2018).
15 https://www.ftc.gov/system/files/documents/public_statements/1583714/slaughter_remarks_at_gcr_interactive_women_in_antitrust.pdf.
16 Presentation at OECD Global Forum on Competition: https://www.slideshare.net/OECD-DAF/how-can-competition-contribute-to-fairer-societies-akman-november-2018-oecd-gfc-discussion.
17 Primary goods were those that the citizens need as free people and as members of the society.
18 This is not to suggest however that liberty is only about maximising residual income in order to use that to lead a fulfilling life, rather that this is the aspect of liberty that is most relevant when applying this principle within an antitrust setting. Quite clearly there are numerous other fundamental aspects to liberty that are not examined further here.
19 Again, the efficient citizen welfare standard is an attempt to better describe the fullness of the consumer welfare standard.
20 http://www.oecd.org/daf/competition/competitive-neutrality.htm.
21 Alternatively, given Rawls’ focus of the greatest equal liberty principle on primary goods, we might consider that this principle broadens the scope to anticompetitive conduct with significant effects in markets for non-primary goods.
22 This is different therefore from the uncertainty that agencies encounter during a case, particularly in potential competition cases (see Caro de Sousa & Pike, 2020).
23 See OECD (2019).
24 See Caro de Sousa & Pike (2020), Motta & Peitz (2019), and Valletti (2018).
25 https://academic.oup.com/antitrust/article/8/2/354/5819564?login=true.
26 Ayal (2012), contains a non-Rawlsian analysis on the thought experiment set up by Rawls in order to consider how to treat producer surplus and monopoly rents given the possibility that those choosing behind the veil of ignorance might ultimately benefit from them. Notably the author disagreed with a Rawlsian analysis of that choice and argued that instead individuals might in fact choose to create anticompetitive rents in the hope that they might benefit from it. As Ennis et al (2019) demonstrate that hope would be far-fetched for most.
27 As a result rigorous examination of the creation and abuse of monopsony market power against workers or small businesses would be consistent with this approach, but permitting practices that deliver anticompetitive rents to small businesses would not (though enforcement in those cases, particular to those where there is uncertainty on the effect, would be a lower priority since the cost in terms of inequality or inclusivity would be relatively small).
28 Furthermore, this power might not occur at the level of relevant antitrust markets (based on degrees of substitutability). Instead, it might exist at or across industries or administrative regions.
29 See https://www.youtube.com/watch?v=s7gg7w7M_9E.
30 See for example, recent market studies into elderly care and children’s social care. Education markets may also be relevant here (OECD, 2019).
31 See the presentations by Oxera and Analysis Group at the recent OECD workshop on Gender Inclusive Competition Policy (http://www.oecd.org/competition/gender-inclusive-competition-policy.htm).
32 The Digital Markets Act does so through ex-ante regulation of certain conduct, not through Antitrust.
33 Weyl (2018). https://ecp.crai.com/events/cra-annual-brussels-conference-economic-developments-in-competition-policy-2018/.
