Wednesday, December 21, 2016
Communiqué that has been prepared for the last edition of the Rome Antitrust Forum (organized by Alberto Heimler, Giorgio Monti and Mel Marquis)
Communiqué of 25 November 2016
ROME ANTITRUST FORUM
Summary of the third annual meeting of the Forum on 25 November 2016
The Rome Antitrust Forum is an initiative co-organised by the Scuola Nazionale dell’Amministrazione (Rome) and the Law Department of the European University Institute (Florence). The purpose of the Forum is to bring together distinguished antitrust practitioners (both lawyers and economists), leading figures in antitrust enforcement from abroad and representatives of the Italian Competition Authority to discuss, on a non-attributable basis, the way in which the Italian system of antitrust enforcement operates. The Forum facilitates an exchange of views among participants in order to identify key challenges confronting antitrust enforcement in Italy. The Forum ends with some suggestions and recommendations which could lead to a more effective system of antitrust enforcement in Italy.
This year the Forum considered two issues: the role of priority setting in antitrust enforcement and the challenges associated with the ex post evaluation of antitrust enforcement decisions.
Below we set out some of the major questions that emerged from the discussions among the Forum participants. We conclude with some policy suggestions.
Priority setting: what are the options for the AGCM?
The Italian Authority does not set explicit priorities. Since the initial implicit priority to promote competition in public utility industries (1990-1997), the AGCM has prioritised rigorous enforcement of antitrust law based on European established practices (1998-2005), an adoption of a consumer oriented enforcement agenda (2005-2011) and a return to addressing more established practices, such as price-fixing cartels (2011 to present). Furthermore, over the years the Authority tried in its annual report to rationalise the enforcement record by identifying a common thread across decided cases, but this has been more an ex post exercise than a guide for action.
Other authorities do it differently. They explicitly identify a set of priorities which are then made public and are meant to solicit complaints and signal likely ex officio cases. A prominent example is set by the UK Competition and Markets Authority (CMA). The CMA formalises its priorities by assessing each possible action in light of four factors: the likely market impact of the action, its costs, the risks that would be incurred, and the strategic significance of the matter at hand. This approach ensures that all likely action is assessed internally using a common metric. In addition to this prioritisation of specific actions, the CMA also sets out its annual priorities. For example, in 2016 the CMA identified the following as strategic priorities: favouring consumer access to markets and eliminating unjustified barriers to their freedom of choice, and promoting competition in online and digital markets, in regulated and infrastructure markets, in markets for public services and in sectors that are important to economic growth. More specifically, the CMA “seek[s] a balanced portfolio of cases – including large cases that have wider impact, and smaller, more local cases, that send the message that no business is beyond the reach of competition enforcement”.
There are great advantages in making public the priorities that are being set by the Authority. First, the process of identifying priorities and making them public may lead to a debate on the competitive constraints of the Italian economy and on the benefits of antitrust interventions. Furthermore, making priorities public may induce the public to report to the Authority possible violations. At the same time, firms would be discouraged from violating the law in sectors where the Authority’s interventions may be more probable. Finally, setting priorities may make the Authority more accountable as regards both the quality of the priorities it sets and its capacity to sustain them.
The process by which priorities are set is particularly important. Priorities should not simply be declared based on a priori convictions. The Authority should make an effort to consult producers and consumers in critical sectors, together with the relevant regulators, in order to better understand the competition issues involved and in order to intervene in a more focussed manner. Publishing its draft priorities may help the Authority to acquire relevant information from the public and arrive at greater precision and coherence.
Of course, the AGCM has a general obligation to pursue any case brought to its attention. Priority setting should not be an instrument for denying a complainant’s request for a necessary intervention; rather, it should be an instrument for enlarging the information set at the disposal of the AGCM, and for inducing complainants who might not otherwise do so to report their case to the AGCM. Setting priorities thus expands the possibilities of the Authority, it does not reduce them.
Ex post evaluation: case selection and lessons learned
The OECD Competition Committee has produced guidance for ex post evaluation exercises and has conducted two seminars on practical experiences with regard to ex post evaluation. In the course of the drafting process, most OECD Member State delegations showed a great interest in the issue, but then, when the guidance was made public, very few decided to undertake ex post evaluations. Indeed, ex post evaluation of antitrust decisions is strongly advocated by the academic community but it is unfortunately still considered almost irrelevant by most authorities, case handlers and policy makers. At most, the perception seems to be that ex post evaluation mainly has a marketing function: it can show how good and relevant the Authority has been, and it ensures that the Government will support of the institution.
But there are many other benefits associated with ex post evaluation, and these are of a strategic nature.
First of all, ex post evaluation can answer questions that are outside the sphere of control of the judge. In a world of mechanical rules - it is prohibited to drive faster than 50 km per hour - the only role for ex post evaluation of decisions is to promote a change in the rules. But in a world of qualitative prohibitions - it is prohibited to drive dangerously - ex post evaluation can challenge existing preconceptions or hypotheses, which may relate, for example, to dominance and to restrictions of competition. Ex post evaluation can also measure the benefits resulting from enforcement. Such an exercise differs from that undertaken by a judge evaluating the legality of a decision on the basis of rules, facts and precedents.
The main objective of ex post evaluation in antitrust is to improve decision-making: learning from experience, building on techniques that work, and correcting mistakes. Two large series of ex post evaluations by the US Federal Trade Commission (hospital mergers and the petroleum industry) led to policy improvements in both areas. Results from retrospective studies helped the UK CMA to shape institutional details (e.g., merger assessment guidelines). At the level of the EU, the Commission in 2005 generated a valuable ex post study on merger remedies which informed significant revisions of the Commission’s Remedies Notice. More recently, the Commission completed a detailed analysis of all its interventions in the energy market which will probably affect the future approach of the EU in that area. In the telecoms sector, the Netherlands Authority for Consumers and Markets partnered with the Austrian sectoral regulator and with the Commission to study price effects following two mergers, approved in 2006 and 2007, which affected each of the two countries concerned.
Econometric techniques are particularly suitable for comparing situations where antitrust interventions have resulted in a change: a merger has been authorized, a cartel has been prohibited, or an abuse has been stopped.
The choice of the data to be used in the analysis is crucial, as the data heavily influence the exact design of the evaluation framework. Very often, data are available but they may be particularly costly to gather. This is why rigorous quantitative analysis could be complemented by qualitative analysis that could shed some light on the validity of past decisions. Sometimes qualitative analysis can be a substitute for rigorous econometric analysis.
For example, the Competition Authority of New Zealand recently decided to examine whether or not anticipated market developments that were key to a number of its merger decisions materialised as predicted. The purpose was to refine and improve the techniques and types of evidence used in forming those expectations.
The Competition Authority selected all merger decisions in which mergers were cleared primarily because strong expectations were formed around one or more of four clear-cut issues:
- barriers to entry in the market were low and entry was likely;
- there would have been enough effective competition in the market after the merger;
- divestiture would satisfy competition concerns; or
- buyers had countervailing power because they could sponsor entry.
The study thus does not seek to determine whether the decisions were appropriate; instead, the aim is to test the validity of the main expectations about the evolution of the market that had led to the merger clearances. The Competition Authority considers this to be easier to achieve in terms of time, data availability and resources, and it expects that the exercise will lead to an evaluation, although incomplete, of the appropriateness of the decision making process.
These qualitative analyses are much less costly and burdensome than rigorous econometric ones. However, if a competition agency should decide to undertake them, especially because of the lack of sufficient resources (both financial and technical), the results achieved should still be made public in appropriate forms and subject to academic and scientific control.
In brief, the recommendations by the Forum on priority setting and ex post evaluation are as follows:
- Undertaking a participatory process of priority setting would help the Italian Authority to identify (what are perceived to be) the most severe restraints of competition in the Italian economy;
- Market players, academics, governments and EU Commission officials could all be asked to provide input for the process of priority setting, and to comment on initial drafts;
- Setting strategic priorities could be undertaken for example every seven years when a new Chair is nominated and the Authority’s priorities could be adapted on a yearly basis according to what has been learned from experience;
- Publishing priorities would signal to the public the importance of complaints and the fact that such complaints will be part of a strategic process undertaken by the Authority;
- Publishing priorities would make the Authority more aware of the importance of ex officio cases;
- Publishing priorities would make the Authority more accountable and enhance its ability to deliver value for money;
(Ex post evaluation)
- Outside the context of cartels, most decisions and most remedies are based on some speculation as to how the market will respond. Therefore, ex post evaluation of antitrust decisions should be undertaken in order to learn from experience and to be better able in the future to anticipate market responses;
- Ex post evaluation is complementary to judicial review, and it is particularly relevant in understanding the effectiveness of remedies and commitments and in evaluating the effects of an authorised merger;
- Quantitative ex post evaluation is possible when data are available, both before and after the triggering event (i.e., the antitrust decision), and the use of econometrics techniques can help to draw rigorous conclusions about what happened;
- While the decision regarding which cases to subject to an ex post evaluation should be made internally, econometric analysis can also be done by external consultants - with the direct involvement of the Authority in all stages of the process;
- In any case, in order to guarantee its quality, the whole process and results should be subject to the control by the scientific community;
- Qualitative ex post evaluation is also very important, and it can verify or falsify the hypotheses on which the Authority’s decisions were based: barriers to entry, the extent of the geographic and product markets, countervailing market power, etc.;
- Qualitative analysis is best done internally. Its results should be made public so as to provide a guide for future decisions and a strategic tool for evolution in antitrust enforcement.
 See Prioritisation principles for the CMA, 2014, https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/299784/CMA16.pdf.
 See Competition and Markets Authority Annual Plan 2016/17, https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/508136/AP2016-17-final_PRINT.pdf, paragraph 2.7.
 See DG COMP, Merger Remedies Study (public version), 2005, http://ec.europa.eu/competition/mergers/legislation/remedies_study.pdf; Commission Notice on remedies acceptable under the Council Regulation (EC) No 139/2004 and under Commission Regulation (EC) No 802/2004,  OJ C 267/1.
 See Commission, The economic impact of enforcement of competition policies on the functioning of EU energy markets, 2016, http://ec.europa.eu/competition/publications/reports/kd0216007enn.pdf.
 See https://www.acm.nl/en/publications/publication/15090/Effect-study-of-two-telecom-mergers-in-Austria-and-the-Netherlands/.
 See OECD, Reference guide on ex-post evaluation of competition agencies’ enforcement decisions, 2016, http://www.oecd.org/daf/competition/reference-guide-on-ex-post-evaluation-of-enforcement-decisions.htm.