The Uneasy Cohabitation of Law and Economics in Competition Regimes
Competition policy has always had to steer a wary course between economic rationality and justiciability. There is thus a tension in all competition regimes between prescription and analysis, the wish to provide certainty by condemning (or increasingly often, blessing in advance) nominated forms of business behaviour versus the desire to fit regulatory responses as closely as possible to the circumstances of particular cases. The first approach, which we may call formalist, assumes without further proof that particular economic outcomes (be these desirable or undesirable) follow so frequently from particular acts or omissions that the transaction costs associated with investigating the link between the two in each and every case are not worth incurring. The second approach, which could usefully be dubbed empiricist, requires a case-by-case showing of anticompetitive purpose or effect.1
The tension between empiricism and formalism is not a question of law versus economics. Bright-line rules, where possible, should be justified on economic grounds before they become part of regulatory practice. As part of this process, the transaction costs attached to over- and under-regulation need to be assessed properly. Once a bright-line rule is in place, however, economics is effectively banished from the courtroom. Conversely, it has to be admitted that much of the economic evidence given in competition cases will seem peculiarly indeterminate to the trier of fact, partly because of the ex ante nature of the questions economic witnesses are required to answer, partly because of a lack of consensus among economists as to the conclusions to be drawn from such facts as can be proven, and partly due to the important but easily blurred distinction between deductive and inductive reasoning that lies at the heart of everything economists say and think. The point that needs to be emphasised here is that while this often means that fact finding in competition cases has to be done in the face of an empirical deficit, this also applies (if less visibly) to the adoption of bright-line rules. Bright-line rules do not fill factual lacunae, they ignore them.
While the language of economics provides a convenient and seemingly ubiquitous lingua franca in which competition lawyers can converse easily across jurisdictional boundaries, we need to remember that the openness of individual competition regimes to economic reasoning has evolved over time at different speeds and within different legislative frameworks. There is also the lingering persistence of non-economic values in competition policy everywhere,2 a persistence that provides periodic opportunities for a legislatively driven populist resurgence. The different rates of economic take-up are the product of different views as to how (and how far) legal and economic thought processes can properly be made to mesh in the application of what are, in the end, legal rules. These differences owe as much to history and politics as to logic. All feed into the way in which individual jurisdictions treat refusals to license.
In the earliest period of antitrust history, a period in which economic analysis played little or no part, the rule of reason became established as the appropriate mechanism for assessing anti-competitiveness. Despite its present-day centrality in the internal taxonomies of competition laws around the world, the concept of a rule of reason has always been rather fluid. In some contexts it means no more than that when such a rule applies, antitrust liability is fault-based rather than absolute. More usually it purports to describe a process or a methodology in which judges remain free to mould liability to the circumstances of individual cases in the pursuit of relatively diffuse goals such as ‘competition’ or ‘efficiency.’ When invoked in this way it is invariably counterpoised against its supposed opposite, the per se rule, a not entirely apposite dichotomy, at least in relation to United States antitrust law. There is nothing, however, about the concept of a rule of reason that mandates an economic input into either its formulation or application.
Employment of the concept of the rule of reason in judicial decision making long pre-dated the routine acceptance of economic evidence as a guide to factual outcomes or the content of legal rules. Rather than being the handmaiden of economics, the rule of reason was in its earliest manifestations simply a mechanism for defending freedom of contract. It was first invoked by the United States Supreme Court in Standard Oil of New Jersey v United States 3 to read down the otherwise bald and seemingly uncompromising prohibition set out in section 1 of the Sherman Act. By this means, the expression ‘every restraint in trade is illegal’ was softened into ‘every unreasonable restraint in trade is illegal’: a methodology taken directly from common law notions of restraint of trade which, at this early stage of antitrust law, the judges imagined themselves to still be applying. In time, the rule of reason made the passage to section 2 of the Sherman Act, again without much in the way of a direct contribution from economics.4
In contrast to the United States, where different economic models and methodologies have, one after another, cut a swathe through its antitrust law, European competition law regimes have by and large managed to avoid economics-driven oscillations in enforcement policy. Courts and regulators in the European Union continued for several decades to apply an approach5 based on formalistic categories of permitted (or not permitted) behaviour,6 an approach that to its critics seem to be as much concerned with the impact of that conduct on individual competitors as with its effect on the competitive process as a whole. They remained, until relatively recently, unreceptive to the infiltration of economic theories of any school.7
It is only in the last decade or so that it can be said that economists and economic theory have gained a firm foothold in the decision making of the European Commission itself. The formal appointment in 2004 of a Chief Economist to the Commission and an increasing reliance on purportedly ‘scientific’ market-measuring tools such as the Herfindahl–Hirshman Index in the Commission’s analysis of market power8 have been followed by the announced integration by the Commission of more econometric analysis into abuse cases (including refusals) via various soft law initiatives.9 These signal the adoption by the Commission of a more ‘effects-based’ approach to enforcement, one under which concepts such as ‘economic efficiency’ and ‘consumer welfare’ are intended to play an increasing part in assessing anti-competitive harm. Given that both concepts are fraught with ambiguities,10 this effects-based approach is yet to be properly tested in the courts. Neither is it clear that the Court of Justice or the General Court is entirely persuaded by these new perspectives.
Courts and regulators in our other three jurisdictions have been blessed (or cursed) with a degree of legislated detail that simultaneously limits and reinforces the economic element in their decision making. Thus while on one hand legislators have been happy to pick up11 economic terms of art such as ‘market’, ‘consumer’ and ‘efficiency’, and thereby leave would-be interpreters to choose among competing theories on these subjects, they have also not hesitated on occasion to seek to insert a particularly economic viewpoint into the legislation, generating some fairly creative statutory interpretation in response. The line between per se rules and rules of reason in these jurisdictions is laid down in the statute itself. Limitations and exemptions for particular types of activity are also legislatively mandated, particularly in relation to intellectual property. In Canada and New Zealand there are also provisions laying down the objectives of the competition statute which further seek to tie the courts’ hands in various ways. The end result of this close legislative attention, however, turns out not to be greater clarity but rather a raft of illogical distinctions that few economists of whatever stripe would be able to endorse or justify.
While the modern rule of reason relies heavily on economic analysis to move its operation beyond mere judicial hunch, there are still occasions on which lawyers and economists appear to be talking past each other when the results of that analysis are fed into the mincer of legal and regulatory processes. The reasons for these mutual misperceptions are not hard to find. They arise from the fact that, while the logic of economics appears to be unified, its conclusions are actually based on two quite different forms of reasoning. The first is inductive; proceeding by way of the ‘last hypothesis standing’ methodology of all sciences (hard and soft) in which empirical evidence is gathered and sifted not to demonstrate that ‘X is true’ but that ‘X has yet to be proven untrue’. Economics’ second mode is deductive; reasoning down logically from stated assumptions, a process usually presented mathematically in the form of a model. In its inductive aspect, economics is not markedly more developed or rigorous than other behavioural sciences. It is the deductive side that both gives economics its rigour and limits its ability to provide definitive once-and-for-all answers to the factual and hard legal questions thrown up by competition proceedings, questions which require immediate answers with little or no time for empirical or statistical analysis before or during a formal hearing or investigation. This empirical shortfall would be there even if all economists were to agree on predicted outcomes (which they largely do not) and methodology (where they tend to). It is scarcely surprising, then, that disputes as to theory in competition cases tend to take the form of a battle between the current starting assumptions or past predictive reliability of competing models.12 In theory, it should be possible to bring induction and deduction together by using the former to test the latter as predictions against actual outcomes. For reasons that are as much cultural as methodological, this seldom happens outside the courtroom, much less within it.13 Erecting and demolishing models is what most economists do most of the time (such activity providing a surer path to career advancement, as a public choice theorist would no doubt observe). Empirical studies tend to be on a smaller (and thus more easily assailable) scale than the models which they seek to test. Conducting them is therefore not without professional risk.
One of the consequences of elevating the forensic use of models in this way is the marginalisation of that which is not easily modelled. As one commentator has pointed out, the propensity of modern industrial economics to emphasise internal logical consistency and internal rule following14 is both a symptom of its disconnect from the other behavioural sciences and a ‘response to the lack of a secure empirical base’.15