Book review “The more economic approach to antitrust law”
Overview of the book’s main argument
Anne Witt’s book “The More Economic Approach to Antitrust Law” deals with the “more economic approach” (MEA) in antitrust law, which the Commission proclaimed a decade ago. The book’s main argument is expressed eg the following passage:
“The more economic approach has transformed the Commission’s competitive assessments to the point of unrecognisability. […] The Commission operates on the basis of a different concept of harm, a different concept of countervailing effects and uses different tools to prove the former.” (250)
In regard to each of these three main “differences”, the book’s main part (Part II) describes the situation before and after the introduction of the MEA in regard to Article 101, 102 and merger control. According to the book, the “key to the Commission’s more economic interpretation of the law lies in its new understanding of the antitrust rule’s purpose” (246). This new understanding of the law’s purpose is described as follows:
“The more economic approach … is based on the principle that EU antitrust law should be guided by the objective of economic theory, namely the creation of economic wealth, to the exclusion of any other aim. However, the Commission chose to adopt a more restrictive concept of economic wealth than that favoured by economic theory, and decided that EU antitrust law should be guided by the aim of enhancing consumer welfare rather than total welfare.” (246)
Part I briefly describes “triggers and catalysts” leading to the adoption of the MEA, and Part III describes what the author views as the MEA’s “advantages”, as well as “concerns” that it raises.
Overall, I found the book to be an interesting read. This is particularly because of its systematic description of the “before” and the “after” situation in regard to all three areas of EU antitrust law. However, I also found that the book suffers from major methodical shortcomings, which sadly undermine this potential strength of the book completely. These shortcomings — especially when it comes to the relationship between economics and law — are, I believe, characteristical for much of European legal scholarship on competition law and economic law, and this is why I discuss them at length in this review. Because the review became very long, and addresses methodical problems that haunt competition law literature in general (and not just the book under review), I formulated a number of key methodical “takeaways”. These are indicated throughout the text, and which will be recapitulated at the end. My overall evaluation of the book is a mixed one: its methodical shortcomings significantly undermine the credibility and plausibility of the book’s analyses and evaluations. However, the underlying effort is manifest, and I believe — if these methodical shortcomings are remedied — the book’s extensive before-after analysis could become useful. In this sense, I wish for a second, completely reworked, edition (with the emphasis on “completely”).
Research question and method
The book’s research question can be identified, for example, in the following passage:
“The analysis focuses on the substantive changes that the Commission has made in the interpretation and application of the EU antitrust rules under the catchphrase of the more economic approach.” (3)
Thus, the book’s research question is whether and how the Commission interprets and applies antitrust rules under the MEA differently than before. I believe that the term “catchphrase”, while maybe not used with that intention, is important to understand one of the central methodical challenges the author faced. Surely, the MEA may describe a substantively different interpretation and application of the law. However, it also served or still serves an important communicative function; it is, in other words, also a public relations (PR) instrument of the Commission. Why can the MEA plausibly be conceptualized as a communicative strategy? The book describes, in its Part I, the significant political (from the US government) and academic pressure (from mostly US academics) that the Commission was subjected to in the 2000s. After the US and the EU investigations led to different outcomes in the Boeing/McDonnell Douglas and GE/Honeywell mergers as well as the Microsoft case, the Commission faced considerable criticism: it was alleged that the decisions were protectionist, and that they were developed on the basis of an deficient analysis based on an outdated understanding of competition law and economics. The MEA was — in my understanding — part of a strategy to counter these claims, and to re-establish the credibility of the Commission as an antitrust authority. Establishing a prominently visible “chief economist”, employing an “economistic” vocabulary (i.e., using terms and concepts that aim to give the impression of economic competence) and filling up its publications and decisions with complicated-sounding and sophisticated-looking concepts and analyses could deflect further criticism. Such understanding of the MEA should not at all be viewed as far-fetched: a review of the Commission’s work shows in most areas of activity a high degree of strategic consciousness about the communicative dimension. It routinely employs legal-technical terms that are manifestly chosen on the basis of communicative considerations (this is illustrated by terms such as “area of freedom, security and justice” or “stability and growth pact” which mobilize positively connoted terms). The Commission also employs complex-sounding, economistic vocabulary in a number of other areas of European policy, including international trade and the budgetary rules. This indicates that the MEA may have been adopted with a certain strategic-communicative objective in mind. Consequently, the description of the MEA as a “catchphrase” is fitting, though maybe unintended.
This double nature of the MEA — as an interpretative approach to antitrust law, and as a strategic-communicative tool — implies a potential tension: what is said that is done may not (fully) correspond to what is actually done. Methodically, these two dimensions of the MEA require two distinct approaches — a legal-doctrinal approach on the one hand, and discourse analysis on the other. Of course, a legal study does not have to fully engage in both forms of research. However, even if — as I assume the author would describe her intentions, given that the book aims to identify the “substantive” changes — the book is concerned with the legal-doctrinal question alone, it is still necessary to account for the other dimension as well, because what institutions do and what they say they do is simply not the same thing, and we simply cannot infer the former from the latter without taking the just-mentioned tension into account. To provide an example, think about a certain, observable human behavior, which may have numerous causes, including psychological and neurological ones. A psychologist who researches this behavior may not wish to engage in e.g. neurological research, and instead wants to focus on the psychological dimension alone. She still has to be clear about the limits of her psychological research premises and methods in explaining that behavior, and about how the insights about the psychological causes she gained through her research may be undermined by findings about neurological causes. Methodical accuracy in legal research requires, first of all, to be precise about the methodical limitations of the research project. And, in my view, the book fails to do that. Take, for example, the following passage, which appears in the chapter that attempts to establish the Commission’s “agenda”:
“A number of sources are capable of shedding light on the Commission’s intentions and aspirations in its review process. First and foremost, there are the Commission acts introducing the more economic approach and the public statements of the key Commission officials under whose watch these instruments were created. As the following shows, none of these sources is particularly clear as to the actual implications of the approach. However, if one reads them in their broader context and takes into consideration the academic debate accompanying the reform process, it is possible to gain an adequate idea of the issues that the Commission intended to address.” (54)
I agree that the listed sources can be, potentially, helpful to establish the Commission’s “agenda”, if this is understood as what the Commission actually plans to do. However, they may also just express the Commission’s PR communication, and the Commission’s “agenda” may in fact be a broader, more limited, or completely different one (for example, it stands to reason that the Commission always takes industrial policy of the EU into account, regardless of whether it admits to do so or not).
As a work of legal research, I accept that the book does not have to extensively employ non-legal methods (e.g. discourse analysis) in order to identify the Commission’s agenda. Rather, the Commission’s objectives may potentially also be inferred by insights made with the legal method proper. It is precisely the identification of a mismatch between the Commission’s proclaimed agenda and its practical implementation that would potentially make such inference possible. However, this possible tension between what is said and what is actually done is not accounted for in the book: there is no methodical strategy identifiable that would allow the author to exclude the possibility that the book is merely describing the Commission’s PR, rather than its legal approach proper. The book does not even acknowledge this possibility. Of course, the two dimensions may overlap or fully correspond; but if they do not, we still wouldn’t know after reading the book. Consequently, the book’s methodical imprecision undermines the credibility of its very findings.
This is not to say that the book does not identify multiple instances where the Commission’s practice stands at odds with what it proclaimed it would do. At numerous points the book’s in-depth analysis of Commission decisions or documents produces a much more ambiguous picture of how the MEA is interpreted and applied than the Commission’s bombastic statements would suggest; however, the book does not systematically pick up on these tensions, and leaves them un-theorized; instead, at multiple instances, the book actually glosses over the contradictions. But it would precisely be a systematic representation of these contradictions which would be particularly helpful to understand the MEA: the rule becomes often clear only through its exceptions. The book repeatedly leaves the impression that the contradictions between announced MEA and the Commission’s actual practice are considered a nuisance, which are subsequently explained away, ignored, or which just remain unexplained. Consider, for example, the following passages:
“The key to the Commission’s more economic interpretation of the law lies in its new understanding of the antitrust rules’ purpose. During the first 40 years of EU antitrust law, the Commission did not operate on the basis of a clearly and exhaustively defined legal objective. On the contrary, it took the view that the aims of EU antitrust law could not be encapsulated by one sole objective … The more economic approach changed this position radically. It is based on the principle that EU antitrust law should be guided by the objective of economic theory, namely the creation of economic wealth, to the exclusion of any other aim.” (Conclusion to Part II, 246). And: “(t)he Commission does not fully follow the premise that EU antitrust law is guided by the exclusive aim of enhancing economic consumer welfare. While the aim of consumer welfare clearly dominates the decision practice and interpretative guidelines, the Commission also recognises and relies on the aim of market integration …” (296) And: “Why make an exception for the objective of market integration, but not for any of the many other Treaty objectives?” (254)
We thus have a description of the Commission’s MEA apparently aiming at implementing an exclusive, narrow consumer-welfare standard, and not doing so at the same time. But both cannot be true: either the MEA is narrow and conceptually pure but thereby fails to take all other of the Union’s regulatory objectives into account, or, it does the latter but is therefore not conceptually clear, and thereby not qualitatively different to the Commission’s approach of the past 40 years (which, however, is — as we saw above — what the book claims), where it also took all kinds of regulatory objectives into account. The book’s methodical imprecision undermines its legal analysis and findings to the extent that it provides contradicting accounts of whether the MEA’s implementation does or does not conform to its characteristics as proclaimed by the Commission. As a methodical takeaway, I would therefore propose the following takeaway:
Takeaway 1: there is a difference between what an institution says it does and what it actually does. The legal method is limited in its ability to establish the former; an author must be transparent about the limits of the methods she employs.
As a side note, by blurring over the distinction between proclaimed agenda and actual practice, the MEA is set up for ideological abuse: the Commission can celebrate its approach for its alleged conceptual purity, which in turn allows people to claim (based on ideological motivations) that numerous important regulatory objectives (e.g. animal welfare, environmental or social concerns) cannot be considered in an antitrust investigation (because the MEA is allegedly concerned with a single objective), and at the same time allows for all kinds of deviations in the practical application. I understand that the Commission is happy to have it both ways, just because it gives the Commission more strategic options. However, proper legal research should not just give the Commission a pass by glossing over the difference between what the Commission says it does and what it actually does.
Dealing with economic knowledge in legal discourse
Law and (social or natural) science are different forms of discourse, and knowledge produced by the latter cannot easily be adopted in the former. Numerous problems of “translation” between the different discourses arise, and a failure to acknowledge them leads to methodically deficient legal scholarship, which is potentially liable to ideological abuse. I will now discuss a number of problems arising from a failure to recognize this difference.
What does it even mean to claim that an approach is “more” economic?
To start with a personal observation, I find it deeply suspicious that, in 2016, a publication can claim to scientifically address a concept termed the “more economic approach”, and not immediately challenge its implicit assumption that there is one correct economic approach that can be applied gradually (“more” or “less”). However, obviously and manifestly the Commission’s approach cannot just be “more” or “less” economic because there is more than one economic approach, given that there is more than one paradigm in the discipline of economics. This problem must be obvious to anybody who has opened a newspaper at any time in between 2008 and today; I simply refuse to accept that anybody — least of all scholars of economic law — would not have become aware of the existence of multiple, conflicting paradigms within economics. I also, prospectively, refuse to accept any imaginable justification for this regrettable omission, e.g. along the lines that the existence of conflicting paradigms would allegedly not affect antitrust. Of course it does, and massively so.
The author is of course aware of the existence of different economic perspectives on matters concerning antitrust. She refers, though only in a few paragraphs and footnotes, to both post-Chicago and contemporary German positions. But this knowledge should have been enough for challenging the implicit assumption of a concept termed “more economic approach”, namely that there is only one (correct) economic view, of which there can be a “more” and a “less”. And if the crisis has shown anything, than that the hegemonic economic paradigm of the 2000s is not the “correct” one, and therefore not superior to the others. Consequently, alternative economic paradigms must be acknowledged when economic concepts are imported into legal discourse (according to Joseph Stiglitz, neoclassical and new classical economics were the “two churches” of mainstream economics; I would add that neoliberalism is the policy approach connected to them; alternative economic views are e.g. Keynesianism or post-growth economics). When a concept from the 2000s suggests that there is one correct economic view, then it clearly refers to the neoclassical/new classical/neoliberal position. Back then, most of us were under the delusion that these views were in fact the “correct” ones (me included). But post-crisis, this cannot be accepted. And yet, a reader of the book could assume that there exists only one correct economic view. Consider, for example, the following passage:
“It is a ‘more economic’ objective than the Commission’s previous understanding of the law’s objective, because it, partially at least, embraces the aim of economic theory, which is to maximise economic welfare.” (109)
The implicit presumption of this passage is that there is something like one correct “economic theory”, which is a clearly and manifestly wrong claim. Otherwise, what sense would the statement make if the book acknowledged that there in fact are multiple, conflicting “economic theories”? There are numerous passages like this throughout the book, including (but not even remotely limited to) the following:
- “US antitrust experts found the European Commission’s analyses to be strangely out of sync with contemporary theory” (26)
- “In their [i.e., US antitrust experts] view, a modern antitrust regime had to be guided by the objective of economic theory, ie the maximisation of economic welfare, be it in the form of general welfare or consumer welfare.” (33)
- “Bringing the legal provisions into line with contemporary economic thinking did not happen overnight, nor did it happen in one step.” (40)
- “Bringing Article 102’s concepts of dominance and abuse into line with mainstream economics raised fundamental questions as to the objectives and key concepts of EU competition law.” (42)
- “The more economic approach … is based on the principle that EU antitrust law should be guided by the objective of economic theory, namely the creation of economic wealth, to the exclusion of any other aim.” (246)
This is a massive methodical shortcoming, comparable to a legal scholar who refers to a controversial view in e.g. medicine without indicating the controversy, and who instead continuously speaks of “contemporary medical science” when she refers to this controversial view. This would be considered a massive methodical flaw, and it must be considered a massive methodical flaw in the context of a discussion of economics and antitrust law, too. Legal scholarship that deals with non-legal knowledge has to deal with the conditionality of the insights adopted from non-legal disciplines, and has to account for inner-discipline disagreement; not only, but in particular when it is so prominent as in the case of economics. This leads us to a second methodical takeaway:
Takeaway 2: If non-legal knowledge is employed, legal scholars must acknowledge the existence of inner-discipline dissent.
If inner-discipline dissent is not acknowledged, legal scholarship is liable to forward an ideologically biased view within antitrust law. It is from this perspective that the communicative dimension of the MEA (MEA as a “catchphrase”) should be considered once more: why did the Commission, and in particularly its economically trained members, such as Giorgio Monti or the various chief economists, would possibly forward something termed the “more economic approach”, despite the existence of multiple, conflicting economic paradigms? Of course, they may have been convinced by their own approach (this is what the author repeatedly claims, without much substantive evidence); however, this does not justify that EU policymakers present their academic and political preferences or convictions as the sole correct, “sound” economic approach. This is an outright, ideological abuse of their powers. In this sense, the use of the term “more economic approach” may also be understood as a strategic communication to forward a specific economic and ideological position, but disguise it as the objective or “correct” view. We have to conclude that, in any reasonable understanding of the concept, there is no such thing as a “more economic approach”. The term is ideologically loaded, and legal scholars should not accept this.
Problematic use of economic concepts in legal discourse: “efficiency” and “welfare”
This brings us to a specific problem in the use of economic knowledge in legal discourse, namely that economic concepts may carry problematic (possibly ideological) connotations, and even more so once they are removed from their academic context, and transplanted into a different debate. If this is not accounted for, these concepts may unduly influence legal discourse and decision-making. From a methodical perspective, as already mentioned, this problem relates to the broader issue of the use of non-legal knowledge in legal discourse. I will discuss the problem on the basis of the concepts of “efficiency” and “welfare”, which the book employs — in my view in a methodically completely problematic form— throughout the book.
Here, we have to start at the very beginning, with definitions, in order to then point out the contradictory character of the concepts of efficiency and welfare. What is the point of any economic activity? Maybe individual survival, reproduction of the human race, joy, comfort, material wealth, status, building massive churches or temples? Possibly all of them, plus many more. There are obviously a million possible reasons why humans engage in economic activity (which, moreover can essentially be any activity). Because of this, some economists of the 19th century developed the abstract notion of “utility” to acknowledge the multiplicity of reasons why humans engage in economic activities. Utility is just a fancy way of saying “making somebody feel better”. They defined the maximization of this “utility” as the driving force of human activity (which already is an ideologically problematic assumption), and consequently also of policy. As there are limited resources (time, labor, raw materials, land, etc), choices must be made as to which utilities are maximized. “Utility”, in an abstract form, is obviously unhelpful as a guide for policymaking, including antitrust policy, because it can, in principle, denote anything that somebody might want (including things like love or a meaningful life). The concept of “utility” can be made operational in two ways:
- either, somebody decides what the objective(s) of policymaking should be (e.g. GDP maximization). This must logically also be based on a decision of how different objectives relate to each other (if the utilities of different individuals are assumed to be comparable and that they can be aggregated, people speak of “cardinal utility”). However, obviously such choice can never be uncontroversial (even if made democratically), in particular if we assume that people do not desire the same things, and cannot fully be convinced otherwise (e.g., I might agree that childcare services are generally important and therefore I pay taxes, and still would prefer to use the resources on the acquisition of a ginormous, wildly polluting car). If we assume that utilities of individuals are ultimately not commensurable, people speak of “ordinal utility”.
- This problem leads us to the other possible, very different approach to guide policymaking: if there is no way that policy choices could maximize all utilities of all persons, maybe there is some institutional setting in which everybody can maximize their utility themselves: this institutional setting, of course, is the market (clearly this is an ideologically guided assumption, as it claims that markets are per definition superior to other forms of regulation in maximizing everybody’s utility). The idea is the following: if people can exchange limited goods and services freely, they will do so with the intent to maximize their utilities; at the end, they will receive those goods and services that maximize their individual utility under conditions of limited resources (ie, everybody will make the most of what they have).
These two ideas of utility (ie, commensurable+uncommensurable) lead to two ideas of efficiency (which is another way of saying “making the best of a given amount of resources”):
- If utilities are commensurable, ie, if we can agree on a common understanding of what is important (e.g. maximization of material wealth), “efficiency” means reaching this objective in the best way (e.g., “productive efficiency” means making the most of a given resource input). This form of efficiency is, at least in principle, measurable.
- By contrast, if utility is incommensurable, efficiency cannot be measured! Instead, it is a quality of an institutional setting that follows exclusively from the theoretical premises of incommensurable and maximizable utilities. In the literature, this type of efficiency is usually further specified in the form of Pareto efficiency, which describes a fictitious situation where nobody’s position can be bettered without somebody else being worse off (this is obviously a massively ideological concept, because its premises prohibit any kind of redistribution). Obviously, Pareto efficiency is a quality that only markets can have, and state intervention can never lead to Pareto efficiency: because, on markets, if somebody’s situation could be bettered, she would make the necessary exchanges herself. Actions on markets are called (e.g. by Paul Samuelson) “revealed preferences” (simply because we cannot measure people’s utilities, which are “revealed” only by their individual actions on the market). By contrast, public policy choices will always leave somebody worse off — all of this is purely theoretical, and as such not open to empirical proof.
If somebody speaks of “efficiency” (or welfare, which is the same), she may in fact speak of two different things without clearly distinguishing between the two, or without indicating which one she means. In practice, we see competition law scholars and authorities constantly meandering between the different versions of efficiency without being transparent about it. However, in terms of policy guidance, the two types of efficiency are completely different:
- The first, commensurable type of efficiency can, at least in principle, be measured: e.g., by a welfare indicator such as GDP growth. If policy-making (such as antitrust policy) is guided by this form of efficiency, its maximization is, in principle, a question that is accessible to empirical study: the overall welfare effect of an antitrust intervention may be superior to non-intervention, or the other way around.
- By contrast, the incommensurable type of efficiency is a theoretical assumption and cannot be measured even in principle, because of its very own premises. It is simply a description of what markets are, deduced from the chosen premises. Public interventions are, again exclusively by definition, always less efficient.
It can be seen that the two versions of efficiency guide public policy very differently:
- Commensurable efficiency may be achieved, in principle, by any type of institutional setting, whether market- or non-market based. This means that, in regard to antitrust, we cannot say in the abstract what kind of market settings and what kind of public interventions are more efficient — it must be established for the individual situation. The antitrust authority can employ, again only in principle, empirical methods to find out.
- By contrast, incommensurable efficiency always suggests that markets are more efficient. The only exception is when the market systematically fails, ie, when market failures are present.
We will have a closer look at the specific problem of measuring efficiencies further below. For now, it is important to understand that the two types of efficiency are conceptually completely different, and guide policy in a completely different form. However, in antitrust discourse, these two types of efficiency are rarely, if ever, distinguished, and their different implications for policy-making are hardly ever made explicit. This is the case in the present book as well, which also meanders between different definitions of efficiency. However, the confusion between the two types of efficiencies make antitrust discourse liable to ideological abuse.
Nowhere does this potential of ideological abuse of an intransparent use of the different concepts of efficiency become more manifest than in the writings of the Chicago school. Consider two of their central demands: first, market behavior should be evaluated as to its individual effects only; and second, markets are assumed to be generally able to dismantle cartels and monopolies, so that antitrust enforcement should be limited to “plain vanilla cartels” and “merger to monopoly” (as Easterbrook stated, cited at 64, Fn 52). These two demands are completely at odds with each other, because the first refers to a commensurable concept of efficiency (ie, the antitrust authority can measure pro- and anti-competitive effects and balance them), whereas the second refers to the incommensurable concept of Pareto efficiency, ie, markets are superior in maximizing utilities, with the exception of market failures (here, “plain vanilla cartels” and “mergers to monopoly” have to be conceptualized as market failures — because, if they weren’t, then the market would dismantle them on its own, and no public intervention would be necessary). But if the first assumption is taken seriously, the second would have to be dismissed, and the other way around! If efficiencies are incommensurable, then an individual assessment of the competitive effects of market behavior is impossible (because there is no measurable standard to establish whether this market behavior is pro- or anticompetitive). By contrast, if efficiencies are commensurable, then we cannot operate with any kind of generalized assumption or legal presumption, because effects must be established individually! This makes the Chicago approach completely incoherent, and liable to ideological abuse. And this is what many of the Chicago school-affiliated scholars — who, by and large, were staunch neoliberals — in fact did: if they don’t like a certain legal presumption, they demand individual assessments; if their preferred outcomes cannot actually be proven by individual assessments, they apply a legal presumptions to their liking, based on their understanding of when markets are Pareto efficient and when they aren’t.
The problems relating to the concepts of efficiency have been extensively discussed in economics, and brought into a comprehensible form eg by authors like Amartya Sen, Joseph Stiglitz and Paul Krugman. Given that the three named authors have all received the Nobel prize, their views can hardly be called marginal or irrelevant in “mainstream” economic discourse. And yet, the present book refers to the efficiency and welfare concepts as being held by “modern”, “contemporary” or “mainstream” economics, without any indication of their problematic and disputed character. This is methodically deeply troublesome, because it misrepresents contested concepts as the scientifically correct ones. This leads us to the following takeaway:
Takeaway 3: non-legal concepts cannot simply be adopted from other disciplines into legal discourse without fundamentally scrutinizing their inherent presumptions. To the very least, the inner-discipline criticism must be acknowledged, and its implications for the legal discipline spelled out.
The author, according to herself, is not (fully?) prescribing to the Chicago view. And yet, the book completely fails to point out the contradictions of their approach, even though this would be, in turn, decisive for the evaluation of the MEA and its application. Consider the following example: both Chicago and the MEA demand individual assessments of pro- and anticompetitive effects; at the same time, both apply a presumption of illegality for hardcore cartels; the contradiction between these two approaches are not at all addressed by the book, which argued, for example, the following: “Cartel decisions are therefore as a rule far less informative on the Commission’s understanding of competitive harm than decisions on less obviously anticompetitive conduct.” (122) But if cartels constitute, as the statement implies, a very obvious form of anticompetitive conduct, this should be particularly easy to measure and to establish in an individual assessment, instead of defining a legal presumption. And yet, this perfectly obvious contradiction is not picked up in the text, creating the impression that it would be perfectly coherent to demand individual assessments and apply legal presumptions at the same time under the premises of the Chicago school.
The problem of measurement
This brings us to a follow-up problem, namely that of measurement and empirical proof. The MEA is fundamentally based on the assumption that the anti-competitive effects of a certain behavior would be measurable in the individual case. This is also, as the book indicates, the view of the author, who states at one point: “Ideally, both the anticompetitive and pro-competitive effects should be quantified.” (298). However, I believe that this view — ie, that quantification is “ideal”, and implicitly, also achievable — is based on a fundamental misunderstanding of what science can and cannot do in general and of economics in particular, and — by extension — on a misunderstanding of the relationship between science (including economics) and law. First, we need to go back to the distinction between two types of efficiencies that was made previously, namely between commensurable and incommensurable concepts of efficiency.
- The first (commensurable) type of efficiencies are, at least in principle, observable, and therefore, again in principle, open to empirical study and measurement.
- By contrast, the second (incommensurable) type of efficiency is, prima facie, an assumption only, and can therefore not be measured. (More indirectly, the presence of market failures can be open to empirical research, though under severe limitations.)
However, even if these efficiencies are, directly or indirectly, open to method-guided (and potentially empirical) research within the framework of the premises within which economists choose to operate, this does not mean that their insights are in any form “objective” data, or in any form employable within legal discourse, for the following reasons:
- first, regarding the incommensurable efficiency concept, the existence of certain types of market failures — on which legal presumptions such as the anti-competitive effects of “hardcore cartels” are necessarily based — may be made relatively convincingly in a general form (see e.g. Akerloff’s article on “lemons”, or Stiglitz’ research an information asymmetries). However, it strikes me as largely impossible to do so unambiguously for individual assessments, which is precisely where the MEA claims to employ economic insights. More generally speaking, insights about aggregate economic phenomena or tendencies can simply often not be translated into individual situations. In other words, a macro-level analysis does usually not hold conclusive information about individual situations, and can usually not directly be reproduced for the micro-level. Moreover, the discussion on functioning and failing markets depends so much on the contested premises employed by the various economists that their insights cannot be unambiguously picked up in legal discourse (e.g., some renowned economists believe that market failures almost never happen, whereas others believe that they are ubiquitous — which view should the Commission follow?).
- Second, even the “commensurable” efficiencies can often not unambiguously be established or measured. How should, for example, innovation be measured? It is certainly possible to develop a measurement for them, but they will rely on numerous premises which, in turn, cannot unambiguously be translated into legal discourse, and certainly not for individual assessments. For example, it might make sense, on an aggregate basis, to understand research spending as a proxy to innovation. However, this is certainly not directly translatable into a legal assessment, not least because research spending as an indicator would certainly be easy to manipulate. Finally, once multiple efficiencies (innovation, price, etc) are recognized, we face the additional problem of weighing these factors against each other (how does innovation relate to product quality etc). Again, we encounter the difficulty of macro-to-micro translation (what can be researched in the aggregate may often not be researched for individual companies, even much less for individual transactions).
- Finally, if a welfare standard is used that exclusively focuses on one individual factor (it will likely be price effects), we still cannot evade the previously discussed problems. Any kind of measurement must still be based on numerous premises: the price effects accruing within which timeframe shall be considered? How “directly” related to the transaction should these effects be? Do we “translate” eg social and environmental costs into monetary terms? Morever, the macro-to-micro translation problem also remains: even if general economic effects could be established in principle (not very realistic, given how badly economists predict eg future GDP development), believing that this could be done for one specific transaction is just a ridiculous assumption.
This means that legal discourse must be really careful in referring to findings of economic research — not because economic research is per se unreliable (though some is), but because of the significant difficulties of employing non-legal knowledge in law. In particular, the following issues are relevant:
- First, we generally cannot incorporate economic insights into legal discourse, without taking into account the premises they are based on. However, economic premises — transposed into law — constitute policy choices. For example, if economists chose not to consider environmental costs for a certain calculation, then their findings will provide a certain outcome (though potentially a useless one, if we want to be informed about the overall costs and benefits of some phenomenon, and not just the ones that economists want or can measure). By contrast, an antitrust authority deciding not to take environmental costs into account makes a policy choice in that regard. The implication is that, if the “welfare” standard is employed, the antitrust authority is in fact making numerous policy choices in regard to numerous regulatory objectives.
- Second, most efficiencies are simply not calculable for the individual case — there is an unsurmountable macro-to-micro translation problem.
Imagine a tort case of a smoker against a cigarette producer, and the judge asks the medical expert: did smoking cause cancer in this patient? The expert will only be able to say: we see a probability of this kind of cancer in this kind of patient with this kind of lifestyle of this age of this racial and socio-economic background who smokes this much at very roughly xx%, but for this specific patient it might be completely different. It’s a probability, at most, that can be provided. By contrast, some economists act as if they could precisely predict such causation in the economic realm. The ex-chief economist Röller, for example, holds that the Commission under the MEA is “assessing the likely competitive impact of a particular transaction” (cited at 56). Let us relish, for an instance, in such hybris. Whereas a serious medical expert witness would presumably refrain from drawing conclusions in the individual case based the discipline’s general knowledge of causation, the Commission’s chief economist believes that he can evaluate an at least comparably complex phenomenon for the individual transaction? This is brilliantly ridiculous. Antitrust economists like Röller are simply promising too much. In US courts, academic-looking evidence that cannot seriously be backed up by sound science is called “junk science”. Trying to precisely establish pro- and anti-competitive effects of an individual transaction, maybe even in monetary terms, would be just that.
To sum up, the fact that economists do empirical studies or quantify things should not lead legal scholars into believing that they now have “objective” data at their disposal that can simply be picked up in legal discourse. As shown, doing so implies numerous policy choices. An antitrust authority which claims to pursue an exclusive welfare standard, and employs economic data to evaluate it, is merely being intransparent about the multiple policy choices that this implies. Moreover, it is blind to the macro-to-micro translation problem, which essentially means that individual assessments as to efficiency effects (regardless of how they are defined) will usually not be able to rely on credible economic data (at least if “junk science” is excluded). Ultimately, an antitrust authority will always have to rely on legal premises, which implies that major policy choices that touch on numerous regulatory objectives are inevitable.
How plausible is the claim to superiority of “welfare” as an antitrust standard? (No)
This, now, leads to the question in how far the promises of the MEA hold up, once the problems of “translating” economic concepts into law are accounted for. The answer is no, because of the following reasons:
- First, the individual assessment of antitrust cases requires the presumption of a commensurable efficiency standard (otherwise an individual assessment of pro- and anti-competitive effects would not make sense). Some authors — e.g. Posner, if I remember correctly — are intellectually consequent insofar as they propose one singular standard (Posner picks the maximization of wealth, even though this choice is in turn more problematic than he is ready to acknowledge). But such singular standard is hardly justifiable given the breadth of policies that all polities — including the EU — pursue through economic regulation (this critique is also briefly addressed by the book in its concluding chapters). And this is also why no antitrust authority — as the book shows (e.g. 65 for the US) — in fact applies just one standard, and instead all scrutinize “welfare” effects on the basis of a plurality of factors. But as soon as multiple factors are considered, we face the problem of how these should be weighed: how do lower prices relate to lower innovation, etc. Consequently, even if an antitrust authority would in fact claim to be singularly concerned with a welfare standard (which, as the book rightly states, the Commission does not, and instead also scrutinizes the effects on the internal market), this in fact implies multiple objectives of unclear relative weight. This is no different than the Commission’s “old” approach.
- Second, the presence of multiple objectives imply that continuous policy choices must be made by the antitrust authority. Moreover, we face the problem of measurability discussed above, which in turn implies policy choices. But this necessity of multiple policy choices completely undermines the claim that, as it allegedly pursues only one, measurable standard, the MEA would be superior to the Commission’s previous approach.
- Third, as many objectives related to commensurable efficiency are not as such measurable (e.g. innovation), and those which could be measured in principle can still usually not unambiguously measured (much less predicted) for the individual case, presumptions are inevitable. This is the case, for example, when the Commission pursues anticompetitive behavior “by object”, which conflicts with the requirement that the pro- and anticompetitive effects of each act should be evaluated individually. The same is true for the Chicago claim — voiced e.g. by Easterbrook — that antitrust enforcement should be limited to “vanilla cartels” and “merger to monopoly”. These also constitute legal presumptions (here, relating to market failures), which equally conflict with the individual assessment demand. Finally, the same is true for the US’ courts application of “quick look rules” or “bright line tests”, because they also apply general presumptions instead of engaging in individual assessments. Thus, we encounter numerous “form-based” concepts, which the Chicago people and the MEA allegedly attempted to get rid of! But if the MEA operates under legal presumptions which are not scrutinized individually, the MEA is also no longer different from the Commission’s previous approach in that regard.
- Fourth, the general presumption that markets will usually undermine cartels+monopolies, which is the basis for the proposition that antitrust authorities should act only in limited, and clearly established cases of anti-competitive behavior, is based on the Pareto efficiency presumption — and the Pareto efficiency presumption also conflicts with the individual assessment requirement. The same is true for the demand that anti-competitive behavior should not be prohibited unless a relatively high standard of proof is reached, because it also implies the presumption that markets generally are Pareto efficient. The theoretical weakness of the presumption that markets are Pareto efficient is also indicated, for example, by the fact that the Chicago people found it necessary to invoke an additional argument in favor of their preferred high standard of proof: Bork (I believe) argued that it was “unfair” to prosecute market actors unless there is sufficient proof for the anti-competitive effects of their behavior. However, this invokes an additional objective of antitrust investigations — namely one of procedural fairness — which is not connected to efficiency, and instead is based on a certain idea of justice (which, however, the MEA and the Chicago people presumably want to push out of antitrust!). This, in turn, undermines the claim that the MEA would be based on a single objective, and makes it no different — in qualitative terms — from the “old” approach.
- Finally, if a “consumer” instead of a “total” welfare standard is applied, an additional problem is faced, namely how to distinguish between whose costs and benefits are taken into consideration, and whose are not. This is a distributive question that the antitrust authority has to decide on. Moreover, implementing this for individual investigations appears beyond what is feasible even in theory. Consequently, the “consumer welfare” standard implies distributive choices, which also conflicts with a welfare standard as conceptualized by the Chicago folks.
To sum up, we still encounter numerous legal presumptions, which conflict with the demand for individual assessment and the critique of “form-based” antitrust enforcement. Moreover, we encounter numerous regulatory objectives (though some of them are re-packaged in the welfare vocabulary) of unclear relative weight, which conflicts with the Chicago/MEA claim of one singular standard of evaluation that would free the antitrust authority from making policy choices. But if this is the case, the MEA is in fact no longer qualitatively different from the approach the Commission pursued in the previous decades. This leads to the following takeaway:
Takeaway 4: the MEA requires the Commission to evaluate multiple regulatory objectives, and make numerous policy choices. Consequently, there is no basis to claim that the MEA is in any way superior (or qualitatively different) from the Commission’s previous approach.
Consequently, anybody who claims that welfare is or should be one or the sole standard of antitrust policy, without addressing which the continued presence of multiple objectives and the connected necessity of policy choices is liable to obscure the issue at stake, and therefore misrepresents the factual activity of the antitrust authority. This is, as already discussed, illustrated by the Chicagoans: demanding individual assessments while advocating certain legal presumptions and defining a certain standard of proof implies all kinds of policy choices as regards a variety of regulatory objectives, and their relative weight; and yet, their main line of attack against the previous antitrust practice was precisely this plurality of objectives, and their unclear relative weight. In effect, the so-called “antitrust revolution” (a term the book employs completely uncritically) leaves antitrust in a worse state than before: before, at least, the policy choices made by the antitrust authority were relatively visible; today, these choices still happen — yet, they are obscured by the claim of a quantifiable, objective, single standard. If a main task of legal scholarship (as defined e.g. by Kelsen) is to be clear as to what the law requires, and where policy choices are made, then the Chicago approach — and, to a more limited extent, the MEA — put the system at a much worse state than before. However, the book completely ignores this problem, and repeatedly suggests that the MEA is, at least in principle, much clearer than the previous approach of the Commission, about which the book states eg the following (these are a few examples of many):
- “An analysis of these acts reveals an incoherent and not particularly well-thought-out understanding of the aims of EU antitrust law.” (91)
- regarding the “old” merger control system: “A key problem with this definition of dominance is that it is broad and vague.” (129)
- “but what type of harm is actually caused by an undertaking’s ability to behave independently? This remained unclear.” (129)
- “What these decisions left open, however, was why exactly the Commission considered the inability of competitors to ‘stand up’ to the merged entity or to enter the market in the first place problematic and why it sought to prevent such a situation from arising.” (131)
By contrast, the “advantages” of the MEA are described as follows:
- the “logic and internal consistency of the theoretical principles underlying the more economic approach are compelling”, because it “proposes one concept of harm for all three pillars of EU antitrust law: a lessening of competition that results in a reduction of consumer welfare” (253).
- And: “the more economic approach has the potential to achieve very precise results. It operates on the basis of clear and largely quantifiable concepts of harm and countervailing effects. Changes in consumer welfare, at least insofar as they occur in the form of higher prices and lower output, are quantifiable in theory” (254).
This, however, is an important qualification, because, as the book itself states “decreases in variety, quality, choice and levels of innovation” are “less easily quantified, and less easily balanced accurately against cost savings” (255). But if there are actually multiple standards to evaluate, and if their relative weight is not clear, and if many of them are not measurable, how is the new approach actually in any way different from the “incoherent” and “unclear” old approach? Moreover, the author also correctly recognizes (255–256) that, even if such quantification was possible in principle (i.e., if a really narrow price-based approach is chosen), it fails in practice. There are usually no generally recognized methods to do so, and economists disagree on the applicable models and the data is usually insufficient and contested: “even the use of econometric tools does not necessarily guarantee the correct, let alone an undisputed, result” (256). At another point, the book holds:
“One of the attractions of a concept of harm that equals a reduction in economic welfare is that it can be measured, quantified and accurately balanced against countervailing efficiency effects. It is one of the great advantages over a freedom-based concept of harm. Including welfare factors in this concept that cannot be quantified greatly diminishes that advantage.” (297)
But if the welfare standard, in its broader form, refers to multiple (mostly unquantifiable) factors of unclear relative weight, if the available economic methods are not actually leading to uncontroversial evaluations of pro- and anticompetitive effects, and if the Commission explicitly pursues other objectives additionally to the welfare objective, how can the book come to the conclusion 1) that the new approach is qualitatively different from the old one, and 2) that the new approach is in any way more precise or accurate? I really, fundamentally, fail to see how the book’s findings would support such conclusions. The book names “logical and internal consistency”, “accuracy” (253–256) and “consistency” (296–297) as the MEA’s advantages; I completely disagree, and I do not believe at all that the book’s findings actually support such evaluation. Finally, consider the following statement by Bork (which the book cites):
“Antitrust policy cannot be made rational until we are able to give a firm answer to one question: What is the point of the law — what are its goals? Everything else follows from the answer we give.” (79)
Given this definition of a successful antitrust policy, the MEA (and similarly the Chicago school approach) is a failure, because it is still intransparent about the objectives that are pursued. The sad irony of the welfare standard in antitrust law is that the antitrust authority still decides on the basis of a multiplicity of regulatory objectives and makes numerous policy choices, but is no longer transparent about it. This, in turn, makes it liable to ideological abuse: if regulatory choices are made in an untransparent form, they are liable to be guided to a much greater extent by power and influence; this, in turn, may account for e.g. the fact that the Commission essentially stopped to prohibit mergers altogether, with a few (in the broader economic context) irrelevant exceptions. This leads to the following methodical takeaway:
Takeaway 5: antitrust authorities cannot evade to make policy choices (e.g. by reference to an alleged “sound economics”). Legal scholars are liable to import ideology into legal discourse unless they identify and highlight these policy choices.
The relationship between law and economics, reconsidered
This, finally, leads us to a point where we can reconsider the relationship between law and economics, in antitrust law and otherwise.
What are social sciences? They attempt to explain (and potentially predict) societal phenomena, usually by establishing causal relationships in some form. Economics — like sociology, psychology or linguistics — focuses on specific phenomena, but is also characterized by specific premises and methodical approaches. The correctness of the scientific insights produced depends on the correctness of its premises, and the functionality of the methods employed. Social sciences — like all sciences — have inherent limits, relating to the object of research and the available methods. The precision of prediction achieved by social sciences, including (as the crisis showed) economics, is a limited one, which also indicates that their explanations are subject to severe restrictions. This claim is further substantiated by the ever-present inner-discipline disagreements regarding the premises, methods and findings in all social sciences, again including economics.
By contrast, what is a legal norm? It is, roughly in Kelsen’s terms, a prohibition backed up by force. In this, it must be distinguished from the social sciences, because it prescribes behavior, and does not explain it. Of course, law and social sciences stand in an multi-dimensional relationship: ideally, law and its application is guided by “sound” social science. Moreover, whether we want it or not, social science influences the law and its application because it influences how humans see the world. It is in this sense that Keynes once commented that defunct economists still shape our understanding of the world. However, the guidance of law and its application by social sciences must be based on a recognition of the latter’s inherent limitations. Moreover, it must also be based on a recognition of how non-legal assumptions and views may, often unrecognized by us lawyers, shape the understanding of law (e.g. by influencing a judge’s or lawmaker’s ostensibly common-sensical evaluation of human behavior). This requires a methodical awareness of the distinction between law and facts (as we understand them).
In the discussion above, we have encountered some of the limitations of economics as a social science. To recap: economics is based on numerous, potentially problematic, premises. Virtually all of these premises are contested within the discipline itself. Moreover, its methods are severely limited in their ability to predict future developments, and consequently also of their ability to explain present economic phenomena. These limitations are not surprising, given the complexity of the phenomena observed. However, this insight should lead to a healthy degree of skepticism when lawyers are dealing with economic knowledge. To the very least, the conditionality of economic knowledge, and the presence of inner-discipline disagreements should be accounted for.
This is frequently not done by scholars of economic law. The book, for example, frequently states that the “objective” of “economic theory” is the maximization of welfare (246). This is simply not true. First of all, a theory does not, as such, have an objective — rather, it is a set of assumptions that guide the understanding of certain phenomena. If a neoclassical definition is employed, economics is about how welfare can be maximized (see eg Lionel Robbins). Even such view — look it up in Paul Samuelson’s textbook — does not say that it should be. Even in neoclassical scholarship, the potential tension between “distributive” justice vs “allocative” efficiency is emphasized. To what extent the two are in fact pursued is a political choice. Of course, certain neoclassical economists claim that distributive objectives can be better pursued outside of market processes (whatever that means); however, the question which regulatory setting a polity chooses is a political choice that economists usually emphasize they cannot answer on the basis of their scientific instruments alone.
Thus, and to repeat, the behavior that is prescribed by law is a policy choice. Surely, social sciences can play a role in that, e.g. by informing the lawmaking, and by informing its application (e.g. by providing the evidence). However, any kind of concept of law remain precisely that, namely a legal concept. Consequently, if the law defines eg a certain standard of proof, it is a legal standard, not a social science one. If Article 101 TFEU prohibits agreements “which may affect trade between Member States and which have as their object or effect the prevention, restriction or distortion of competition within the internal market”, then “have as their … effect” refers to a legal, not a social science concept of causation. Confusing these two is methodically wrong, but also has absurd consequences in practice. As an example, consider the delict of murder, which (in must jurisdictions) requires intent. However, judges do not establish intent of the suspect in psychological terms. We don’t even have psychological methods to empirically establish such intent (except junk science methods such as the lie detector); the reason is that it is a legal concept; attempting to test reality as to the presence of the legal concept of intent is like testing reality as to the presence of love as defined by the Bronte sisters. Instead, judges infer intent by comparison to the counter-factual “reasonable person” — that’s not a method that would be condoned by any social scientist. Intent in law and social science is not the same; the same is true for anticompetitive “effects”. What the MEA does is ignoring this difference between law and facts (as established by the sciences). It demands a social-science form of proof for a legal condition, which mixes up two distinct things. It would be comparable to demanding that the prosecutor in a murder case should prove intent on the basis of a social-science understanding of intent, and to do so exclusively on the basis of empirical methods. Under such requirements, no murderer could ever successfully be prosecuted for murder, precisely because there is no psychological method available to test a legal concept. Of course economics should inform competition law. However, this requires an awareness that legal concepts and social science concepts are not the same, which can be formulated as the following takeaway:
Takeaway 6: Concepts of competition law are legal concepts, and cannot directly be answered by reference to social science.
However, the delineation between the two is continuously obscured in the book, when it speaks of the MEA eg as “(b)ringing Article 102’s concepts of dominance and abuse into line with mainstream economics” (42).
Further considerations regarding the evaluation of the MEA
The above-made points are of a purely theoretical kind, based on the recognition of the methodical difficulties that arise when economic knowledge is picked up in legal discourse. Beyond that, the Commission’s actual decision practice gives rise to further doubts as to whether the MEA holds in practice what it promised, some of which I will discuss now.
Is it possible to make the MEA operational in practice? (No)
For me, one of the most interesting observations made in the book is that the Commission is in fact rarely engaging in full individual assessments of pro- and anti-competitive effects of a certain behavior. Instead, we see a surge in object-based prohibitions, commitment decisions and cartel settlement procedures (e.g. 2). The book observes:
“The proliferation of commitment decisions has resulted in a significant reduction in the number of decisions that carry out formal legal assessments of the investigated conduct.” (3)
On commitment decisions, the book explains:
“Commitment decisions do not carry out detailed competitive analyses that result in an unequivocal finding of competitive harm (or lack thereof), but merely express ‘concerns’ that are based on a preliminary assessment of the facts. these assessments are normally relatively superficial and informal.” (122)
Moreover, the number of prohibited mergers has dropped markedly (139). The book makes these observations — to my understanding — mainly because this development of the Commission’s decision practice means that there are only very limited sources that would allow the book to scrutinize the Commission’s understanding and application of the law in practice (which raises the question as to the robustness of the book’s findings, but this is beside the point here). For me, however, it raises the interesting question about the practical consequences of the MEA’s methodical innovations, and ultimately about the nature of antitrust law and enforcement.
Why would the Commission generally evade engaging in individual assessments? First of all, it is — as the book argues — an issue of limited resources: if individual assessments become much more complicated and require many more resources, the Commission can — with the same resources — engage in much fewer investigations than before. Second, if the required standard of proof is one where extensive empirical support for alleged anticompetitive effects is required, but if such empirical support is either theoretically or factually impossible to acquire (as I discussed at length above), then an antitrust breach cannot successfully be established in practice. This — third — may make the Commission focus on cases that can be successfully made, which is the case in particular where they can rely on a legal presumption (most notably, hardcore cartels), or where the Commission does not have to deliver conclusive proof according to a high standard of proof, as in the case of commitment decisions. This implies, however, that the supposed “more” of empirical grounding and individual assessment is not actually happening. Essentially, we merely observe a shift from a system based on certain legal presumptions towards a system that still operates on the basis of legal presumptions, but on partly different ones. Another really interesting observation of the book is that the empirical evidence provided by the parties is hardly ever accepted as useful by the Commission. All of these points suggest that the hope of the MEA — empirically grounded antitrust enforcement based on individual assessments — is misplaced. Of course, this could lead to a scholarly demand of more-of-the-same (i.e., the Commission should make more individual assessments, apply an even more limited standard of evaluation, etc). However, it could also — not implausible, in my view — lead to the conclusion that the MEA is based on methodically flawed presumptions (which we discussed extensively above), and therefore fundamentally dysfunctional. But this possibility is not considered by the book. Consider the following statement:
“Having a clear idea of the law’s objective is, of course, not the be all and end all when trying to interpret a prohibition provision in a coherent manner. There remains the question, for example, of how harm to X, once defined, is to be established when applying the rule. May it be inferred on the basis of a legal rule of thumb? Or should it be established in a detailed assessment of the particular circumstances of every single case? And if so, what is the appropriate standard of proof? Notwithstanding these issues, knowing that the law’s aim is to protect X is the indispensable starting point for a coherent interpretation.” (79)
If the law’s aim is defined by the MEA as maximizing welfare based on empirical evidence, but if this is neither achievable in practice nor even — as discussed at length above — in theory, wouldn’t the most plausible conclusion be that the MEA is simply a flawed concept that cannot be made operational in practice?
The book’s observation that the MEA requires the parties to engage in much more sophisticated efforts, deliver data, etc, essentially implies that the costs of antitrust challenges have massively increased. Essentially, we observe an arms-race between the Commission, the parties, and other affected actors. Such arms race will presumably be first lost by those parties that have the least resources available to engage in a costly legal battle. This implies that economic power — in the form of available resources for legal defense — will become more decisive for the outcome of an investigation. The Commission may think twice before targeting economically powerful companies (as there is a risk of long, costly and ultimately unsuccessful investigations); less potent companies may be ready to give in earlier in the process; potentially damaged competitors may reconsider engagement. All of this implies that the MEA may play out in favor of already powerful companies, and even may incentivize economic concentration. This is the opposite of what antitrust law should do (though Chicago people might disagree).
Moreover, as the book correctly argues, the MEA has not successfully managed to deal with the problem of legal certainty for companies. Unless the Commission investigates on the basis of a legal presumption (which conflicts with the alleged premises of the MEA), the answer to the question whether a specific behavior will be targeted by the Commission is even less clear than before, and the outcomes of an individual assessment are impossible to predict. Moreover, the expansive use of commitment decisions limits the available case law that could ensure a minimum of predictability. The alleged legal uncertainty of the “old” approach has been one of the central critiques of Chicago — and yet, the new approach makes things not better, but worse.
But if transparency and legal certainty are not actually increased, who benefits from the MEA? One possible, outrightly perverse answer could be: antitrust experts do. If the MEA incentivizes an arms race that requires extensive, economistic-sounding data, then the beneficiaries are those who can provide such data, namely specialized law firms and consultants (including law professors). Moreover, such expertization of antitrust implies that it becomes increasingly difficult for non-experts, such as the European Parliament or the general public (but also courts, as the book rightly states), to engage with antitrust issues. This might at first appear to be beneficial for antitrust experts, because they no longer have to engage with non-expert views on their subject; they can — as one can frequently see in other expert-driven areas, such as international trade and investment law — dismiss non-expert critiques as uninformed and therefore irrelevant. This means that an important countervailing force will be excluded from the debate. The Commission may also partly view the expertization of antitrust as a benefit in the short run, because the use of specialized and economistic-sounding arguments shields it from challenges from the courts, from generalist legal scholars, and, most of all, from the public. However, it may also be possible that, over the long run, the legitimacy of antitrust as such will be undermined: if eg the Commission essentially stops to effectively police mergers because it is methodically close to impossible to make a successful case for prohibition, parliamentarians and the general public may lose their confidence that the European system of antitrust enforcement does more than protecting the powerful.
The new chief economist of the World Bank, Paul Romer, has recently launched a staunching critique of the tendency of the discipline of economics to develop ever-more complicated forms of reasoning. In particular, Romer dismisses the use of logics-inspired forms of reasoning in economics as its “mathiness”, which — according to him — has achieved little scientific progress, and has mainly served the function of increasingly shielding the economic “insiders” from critique by the “outsiders.”
However, such dimension of the MEA are not at all considered in the book; I don’t insist that this should have to be done, but it could have been an interesting vantage point from which to evaluate the MEA. Instead, the book continuously insinuates (and often enough states explicitly) that the Commission, Giorgio Monti and the Commission employees were driven by their genuine convictions. While this may or may not be true, I believe that a book that aims to take a comprehensive look at the MEA and provide an evaluation should at least hint at the possibility that other factors (e.g. of an ideological or institutional kind) may have played a role.
“Approximation with US antitrust law” as an advantage of the MEA
The relation of EU antitrust enforcement with enforcement on the US side is a considerable concern to the book, which is very helpful to contextualize the MEA. The book engages significantly with the conflicts that arose in regard to e.g. the Microsoft and the Honeywell cases, which is laudable. The book also discusses at length the critiques launched by US academics against the Commission, which I also found interesting. However, the US critique that the book refers too is reproduced in a generally uncritical manner; it addresses the fact that the Chicago approach is itself massively under pressure in US debate only in a few paragraphs, and is otherwise comfortable in merely narrating the critiques (many of which I find polemical, e.g. the critique that the Commission was employing 19th century economic thinking. I constantly find economic arguments referring to even older authors such as Adam Smith and David Ricardo, without similar reactions). In particular, the author picks up and employs some of the polemical attacks as valid analytical categories. This is the case, in particular, with the claim that “EU competition law protected competitors while US antitrust law protected competition” (eg 17 and 26); the book employs this distinction repeatedly to evaluate the Commission’s “old” approach. Such dichotomy is overly simplistic (reminding of the relationship book Men Are from Mars, Women Are from Venus), and cannot be the basis for a differentiated analysis.
In the concluding chapters, the book argues that the approximation of US and EU antitrust regulation constitutes an “advantage” of the MEA. This is a deeply problematic claim. First of all, maybe US antitrust law is not working that well, either. While this possibility is (very) briefly mentioned in the book, it then moves on to establish the approximation with US antitrust policy as a benefit in itself:
“Irrespective of the substantive merits of the more economic approach, the harmonisation of EU and US antitrust law may have advantages of its own. The US companies export to and import from Europe. Likewise, European businesses export to and import from the United States. Given the extra-territorial reach of both EU and US antitrust law, this type of business conduct is likely to fall within the scope of both jurisdictions.” (259)
Such approximation “create(s) greater legal certainty”, and prevent that “US and EU antitrust authorities will take a different view of the conduct’s competitive nature. […] Given the political tension that ensued between the United States and the European Union during that and similar conflicts, this can only be considered a good thing.” (259)
It is interesting that the book thereby employs arguments relating to industrial policy (import and export objectives), as well as to diplomacy (“political tension”). But this would stand fundamentally in conflict with the MEA’s alleged objective of exclusively pursuing welfare-related objectives, to the exclusion of other concerns (in particular industrial policy!). In fact, it was precisely the reproach that political considerations were shaping antitrust policy that was a major critique launched against the “old” approach to antitrust! Again, we end up precisely where we were before, namely that multiple regulatory objectives shape competition policy — but now, these multiple objectives are covered up in “welfare” rhetorics.
But beyond that, why should it be beneficial to align policies with the US? Where is the proof that the relevant US institutions are not engaging in industrial policy that must be considered disadvantageous from a perspective of the EU, or welfare at large? Given that the MEA’s key demand is empirically grounded enforcement, the approximation objective should not be accepted unless there is significant evidence for its beneficial effects. Moreover, the book itself shows — in its first chapter — that US competition policy even after the so-called “antitrust revolution” is heavily influenced by political choices. The GE/Honeywell merger — touching on defense — is, on the US side, obviously shaped by military policy considerations (e.g. 15). The book also mentions that the Bush administration chose to stop the pursuit of the Microsoft case (20), and also discusses the involvement of the US senate in the issue (24). Under such circumstances it cannot generally be assumed that the alleged US system’s welfare orientation is not just PR, effectively covering up the multiple other regulatory objectives that the US antitrust authorities pursue. The book itself describes the political and ideological dimension of the US approach in the following passage:
“The Chicago Revolution was by no means limited to the US judiciary. The Chicago school’s position that very little state intervention is needed to achieve the aim of enhancing welfare because most anticompetitive practices are untenable in competitive markets, sat well with the Reagan and Bush administrations’ political agenda of cutting back business regulation. It strongly influenced the enforcement practice of the DoJ and FTC, which also adopted the consumer welfare aim as the exclusive objective of the US antitrust statutes.” (64)
But if this ideological bent of the Chicago (and subsequently the US) approach is recognized, I fail to understand why an alignment of EU practice could “only be considered a good thing.” A more nuanced evaluation would have been appropriate.
Here, I recap some of the main methodical takeaways:
- Takeaway 1: there is a difference between what an institution says it does and what it actually does. The legal method is limited in its ability to establish the former; an author must be transparent about the limits of the methods she employs.
- Takeaway 2: If non-legal knowledge is employed, legal scholars must acknowledge the existence of inner-discipline dissent.
- Takeaway 3: non-legal concepts cannot simply be adopted from other disciplines into legal discourse without fundamentally scrutinizing their inherent presumptions. To the very least, the inner-discipline criticism must be acknowledged, and its implications for the legal discipline spelled out.
- Takeaway 4: the MEA requires the Commission to evaluate multiple regulatory objectives, and make numerous policy choices. Consequently, there is no basis to claim that the MEA is in any way superior (or qualitatively different) from the Commission’s previous approach.
- Takeaway 5: antitrust authorities cannot evade to make policy choices (e.g. by reference to an alleged “sound economics”). Legal scholars are liable to import ideology into legal discourse unless they identify and highlight these policy choices.
- Takeaway 6: Concepts of competition law are legal concepts, and cannot directly be answered by reference to social science.