Page content

The Pros and Cons of Counterfactuals

The Swedish Competition Authority organized an international seminar in Stockholm on the 6th December 2013. Invited speakers and opponents gave their presentations on the topic: "The Pros & Cons of Counterfactuals".

2013 year’s “Pros & Cons” conference addressed the issue of counterfactual analysis in anti-trust cases with a panel of distinguished practitioners and academics keeping up the tradition of high-quality presentations followed by stimulating debate.

Contributors (from the left): Eduardo Pérez Motta, Sten Nyberg, Wojciech Dorabialski, Arvid Fredenberg, Damien Geradin, Cento Veljanovski, Miguel de la Mano, Bruno Lassere, Vivien Rose, Šarūnas Keserauskas, Linda Orvedal, Richard Whish.

Summery of the seminar

The former chief economist of the Swedish Competition Authority, Professor Sten Nyberg (Stockholm University), was the moderator of the event. Professor Nyberg kicked off the proceedings by stressing the importance of counterfactual analysis in instances such as the recent Telia-Sonera case as well as cautioning the audience of the caveat that a casual application of counterfactual analysis may easily lead to too many degrees of freedom (economics jargon for “too many scenarios to choose from”).

Professor Richard Whish (Kings College) was the first of the speakers to step up and his presentation titles “What if we did not use counterfactuals?” centered around three key questions: i) When is it necessary to use a counterfactual?, ii) When is it not necessary to use a counterfactual?, and iii) When is it useful to use a counterfactual? According to Professor Wish the answer to the first question is when dealing with either article 101(1) TFEU (restriction by effect) or article 101(3) TFEU cases, in merger control as well as in the assessment of damages. It is not necessary to use counterfactuals in article 101(1) TFEU restriction by object cases as well as in certain article 102 TFEU abuses, such as predatory pricing and abuse of regulatory procedures, as well as dealing with fining policies. In general, Professor Wish argued that it is always useful to use counterfactuals when trying to move to more effects-based analysis but one should be careful with abandoning simple “bright-lines” rules when those are appropriate for the sake of more complex analyses while a competition policy world without counterfactual analysis would be one still stack in form-based investigations.

Professor Damien Geradin (Covington & Burling) was second in line and gave a presentation about “Counterfactual Analysis in Merger Control”. Professor Geradin stressed the fact counterfactual analysis has always been an indispensable part of merger control (even if the term itself is rather new) since any potential effects following a merger need to be compared to the most likely alternative scenario. Traditionally, the standard case has been to assume that the alternative would be the status quo in the market before the parties entered into negotiations for merging. A more dynamic (and modern) approach to counterfactuals would need to also take into consideration and evaluate the chance of occurrence of alternative market exit and market entry scenarios (that have a reasonably high chance of taking place). A failing firm scenario and the existence of an alternative buyer would be typical examples of counterfactual market exits, which Professor Geradin finds the European Commission’s analysis satisfactory whereas he considers their analysis of counterfactual market entry too static.

Dr. Cento Veljanovski (Managing Partner, CASE Associates) took the floor next to share his thoughts on the “Uses and Abuses of Counterfactuals in Competition Law”. Dr. Veljanovski took a rather critical stance against the use of counterfactuals. According to him, counterfactual analysis is a potentially useful tool, consistent with effects-based approach that suffers, however, from many drawbacks. Most notably introducing a great degree of vagueness in most case evaluations since there can always exist several alternative counterfactual scenarios with no way of proving which one is more likely or more suitable for the case at hand. This, explains Dr. Veljanovski, leads to great legal uncertainty and by asking a question but giving no definite answer. His main conclusion was that counterfactual approach is in urgent need of clarification and specification in order to alleviate these shortcomings.

The fourth speaker was Miguel de la Mano (Head of Unit, DG Markt) who presented his thoughts on “The Counterfactual Revolution in Antitrust Law”, what he referred to as “the long journey from deductive to abductive reasoning”. Deductive reasoning refers to traditional form-based analysis that by “giving primacy to form over likely or actual effects risks that enforcement decisions will lack economic logic, evolve as a random walk, and remain unpredictable and easy to circumvent”. The increasing role of economic reasoning in anti-trust led to a shift from form to effect-based approaches which begged the question of which effect was of importance giving birth to the concept of “theory of harm”. The establishment of credible theories of harm and the execution of an effect-based analysis depend on the introduction of counterfactual analyses. Most crucially, according to Miguel de la Mano, the analytical methods must be relatively straightforward, robust and predictable so that dominant firms have a reasonable chance to draw the line between legal and illegal behavior with some certainty.

Last, but not least, Vivien Rose (Chairman, Competition Appeal Tribunal), in a change of tune, took the floor to give a presentation titled “Predicting the Past: Constructing Counterfactuals in Antitrust Damage Claims”. Unlike all the previous speakers that dealt with the use of counterfactuals in assessing anti-trust cases, Vivien Rose, discussed the aftermath once a condemning infringement decision has taken place, by looking at four recent cases, three abuse of dominance cases and one non anti-trust case. Vivien Rose raised several interesting questions such as: i) what is the counterfactual to a cartel’s collusive pricing? What prevents cartelists from independently setting monopoly prices in the case the cartel was never formed? What would be a “lawful” pricing range? ii) is a proper counterfactual what “would” have happened “but for”… or what “should” have happened “but for”…? iii) how do you treat causation in the loss of a “chance”? what is the proper counterfactual in the case of loss of expected profit? iv) when applying strict microeconomic models, is the decision of a perfectly rational utility maximizing economic agent the most appropriate counterfactual if the actual companies in question clearly behave differently? Vivien Rose concluded by stressing how a court may not simply claim the problem at hand “too difficult to judge” and parties that advance unrealistic, implausible counterfactual scenarios may find themselves doing more self-harm than good.

Presentations from the webinar