COMPETITION ECONOMICS AND ANTITRUST IN EUROPE _____
1
Competition economics and antitrust
in Europe
Damien J. Neven
1
(Graduate Institute of International Studies, Geneva and CEPR)
1. INTRODUCTION
This paper aims to evaluate the influence that economic analysis has had on competition policy in
Europe
2
over the last twenty years. It uses evidence from the involvement of economists in competition
investigations, as well as from the evolving content of competition decisions, to argue that there has
been a significant increase in the economic sophistication of competition enforcement. However, at a
number of points enforcement has appealed to economic reasoning in flawed or speculative ways. The
paper discusses procedural reasons why this may have occurred, and evaluates current and potential
reforms with an eye to ensuring this occurs less often in the future.
Why does any of this matter? At the outset, it is worth emphasizing that the state of competition in a
modern economy has an appreciable effect on economic efficiency – though, as we discuss below, the
extent to which the state of competition can be determined by conscious policy is a matter of some
debate. There is also an important constitutional issue surrounding competition policy. It is one of the
few areas in which competence was ceded very early to the European institutions from the member
1 I would like to thank a number of lawyers and in particular D. Gerardin, P. Mavroidis, Nicolas Petit, A. Sykes and W. Wils for guidance on the
legal framework, B. Bishop, C. Mayer and M. Williams for their insights on the market for economic consultants, K. Metha, L.-H. Röller, S. Evenett
for useful comments and discussions, John Vickers and two anonymous referees for comments on a previous version of this paper and S. Baler and
S. Boffa for excellent research assistance. The managing editor in charge of this paper was Paul Seabright.
2 We will focus on antitrust policy at the level of the European Union. Considering the antitrust policies of the member states is a book-length
project which is beyond the scope of this Paper.
COMPETITION ECONOMICS AND ANTITRUST IN EUROPE _____
2
states, probably because it was considered (somewhat simplistically) to be a largely technocratic
domain in which important political trade-offs were unlikely to be considered necessary. Yet in recent
years there has been a tendency to delegate enforcement to the member states, largely because more and
more member states have developed active and sophisticated enforcement regimes. There is no doubt
that the evolution of economic reasoning in policy-making has played an important part in this
interesting and unusual constitutional development.
Judge Learned Hand once observed that "Possession of unchallenged economic power deadens
initiative, discourages thrift and depresses energy…Immunity from competition is a narcotic and rivalry
a stimulant to industrial progress."
3
Over the last twenty years, a significant body of evidence has
accumulated which confirms his intuition, indicating that competition matters for economic efficiency
and in particular for productive efficiency and incentives to innovate
4
. For instance, in one of the early
papers in this literature, Nickell (1996) considered a sample of UK firms and evaluated whether their
productivity growth was affected by competition. He measured the lack of competition by the
importance of the profits accruing to firms. His estimates allow for a comparison of the productivity
growth for firms at the 80th percentile and firms at the 20th percentile of the distribution of profits in
the sample. The difference is a remarkable 4 percentage points, confirming that competition matters in
providing adequate incentives to control cost and improve productivity over time. Very large effects
have also been observed in transition economies that provide a natural laboratory to consider the effect
of competition (see Djankov,and Murrell, 2002, for a survey). Ahn (2002) considered a large sample of
studies on the link between competition and innovation and concluded that competition encourages
innovative activities and has a significant sorting effect between efficient and less efficient firms over
time.
Whether competition policy, as currently practiced, stimulates competition is another, possibly more
controversial, matter. Much of the evidence on this issue relates to the US and relies on accounts of
particular cases in which decisions have had effects on competition and others where it is has not (see
for instance, Baker (2003) for a vigorous case in favor of antitrust enforcement and Crandall and
Winston, (2003) for a more skeptical view). Some insights can be gained from international cartels: the
effects of the Vitamin cartel for instance appear to be stronger (in terms of price increases) in those
countries without antitrust enforcement (relative to those with enforcement)
5
. Exploiting cross-country
differences, Connor (2003) also finds that fines have a deterrent effect on cartels (but not one that will
ever be sufficient to deter all of them) and that leniency programs increase the probability that cartels
will be uncovered. The record of the EU in terms of the prosecution of cartels certainly confirms that
effective cartels can be harmful with long lasting and substantial increases in prices
6
. The record also
suggests that leniency programs may lead to prosecutions of cartels that may otherwise have remained
secret and possibly in operation but of course, the very frequency of cartel prosecution also indicates
that deterrence is currently far from sufficient. With respect to mergers, Duso, Neven and Röller (2007)
use stock market reactions for the merging firms and their competitors to construct a benchmark against
which EU decisions can be assessed
7
. They find that the EU prohibits very few mergers that the stock
3
United States v. Aluminum Co. of America,148 F.2d 416, 427 (2d Cir. 1945).
4
See Evenett, 2005, for a survey (from which Judge Learned Hand’s statement is borrowed).
5
Clarke and Evenett (2003)
6
See Connor (2003) and the Annual reports of the European Commission.
7
The basic intuition behind this approach being that at least in some circumstances, mergers which harm consumers should benefit competitors (and
vice-versa).
COMPETITION ECONOMICS AND ANTITRUST IN EUROPE _____
3
market perceives as pro-competitive (it makes few type I errors) but may still fail to prohibit quite a few
mergers that the stock market perceives as anti-competitive (the frequency of type II errors may be
greater)
8
.
The implementation of competition rules is a core European policy. Competences with respect to anti-
competitive agreements and the abuse of dominance were explicitly allocated by the founding treaty
(respectively Article 85 and Article 86 of the Treaty of Rome, later renumbered as Article 81 and
Article 82). It was conceived as an essential component of the internal market and unusual powers of
enforcement were granted to the Commission (by delegation from the Council). In one of its early
decisions
9
, the Court of Justice (ECJ) made this clear: “The treaty, whose preamble and content aim at
abolishing the barriers between states,…, could not allow undertakings to reconstruct such barriers.
Article 81(1) is designed to pursue this aim”. In addition, the Council adopted procedures in which
implementation was centralized; regulation 17
10
established that in order to obtain the benefit from an
exemption under Article 81(3), firms had to notify their agreements to the Commission, which
accordingly became a “Passage oblige”. Further competences for merger control were granted in 1989,
through the merger regulation
11
(ECMR), again with a centralized mechanism of implementation.
Competition is also an area of in which competences are shared with the member states, which have
developed their own antitrust rules. Jurisdiction is allocated by formal rules and has not been an
important source of conflict. Finally, a few years ago, the Council replaced regulation 17 by a new set
of rules which partly delegate the implementation of EU law to the competition authorities of the
member states
12
. This delegation is not immune from incentives problems as member states have no
clear interest in considering effects which take place outside their jurisdiction (see for instance, Neven
and Mavroidis (2001)). Still, this architecture of enforcement is unusual among EU policies and as
experience accumulates, its functioning may be a useful source of inspiration in other areas.
The fact that economics has become more important in EU antitrust policy and practice since this
Journal was first published is hardly controversial. One of the objectives of this essay will be to attempt
some quantification of the relative importance of economic inputs in antitrust practice. Focusing on the
fees earned by economic consultants, we will observe that the EU may be converging towards the US in
terms of the relative importance of economics and law as inputs in cases. By comparison, economic
resources at the level of the EU commission remain meager, and the asymmetry in resources between
the authorities and the businesses they regulate is a cause for significant concern..
Evidence that economists have been hired increasingly to provide advice is merely an indication that
parties and their legal advisors have found economists useful in order to prevail. It provides only limited
evidence with respect to the role that economics, as a discipline, has played. The role that economic
insights, in terms of theory and empirical evidence, have played can only be inferred from decisions and
judgments and the reasoning that supports them, as well as the evolution of the legal framework
8
For the first few years of merger control, Neven, Nuttal and Seabright (1994) still find some case of possibly anti-competitive mergers that have
been allowed because of political pressure.
9
Consten and Grundig vs Commission, case 56-58/64
10
EC Reg. 17/62 of 6 February 1962, OJ 21/02/62, pp 204-211
11
Council Regulation N° 4064/89 of 21 December 1989
12
EC Reg. 1/2003 of 16 December 2002, OJ L/1/1 of 4 January 2003.
COMPETITION ECONOMICS AND ANTITRUST IN EUROPE _____
4
(including soft laws like guidelines) and policy statements. This essay will thus evaluate whether
economic insights have had an effect on policy and case law and whether some insights have been
neglected.
We observe that economic analysis has had a strong impact in a number of areas:
the analysis of agreements between firms, in particular vertical agreements under Article 81,
has increasingly focused on effects;
the assessment of competition has moved away from the formal notion of dominance
towards effective competition;
the analysis of the factors that determine effective competition has become more
sophisticated, in particular regarding the definition of the relevant markets, bidding markets,
the proximity of competitors’ position and buyer power;
the concept of collective dominance has been progressively developed in terms of the theory
of collusion in repeated interactions;
quantitative methods have become more important.
enforcement procedures, like the leniency programs, which find some foundation in
economic analysis, have been implemented.
Both the Commission and the Courts seem to have played a role in enhancing the role of economic
analysis. There are, however, two areas of concern. The first is the implementation of Article 82, on the
abuse of dominant positions, which has remained rather formalistic. The Commission has however
launched a debate in this area and has published a discussion paper which moves some way towards an
effects based approach
13
. Hence, it may only be a matter of time for economic analysis to have an
stronger impact on the implementation of Article 82. The second area is a matter of process and
procedure. The process through which the concept of collective dominance has emerged has involved
the annulment of a Commission decision, in which the Commission’s treatment of economic theories
and evidence has been criticized. Another two important merger prohibitions have been annulled
14
by
the CFI (Tetra Laval/Sidel
15
and Schneider/Legrand
16
), and another one largely annulled on similar
grounds (GE/Honweywell
17
). We also observe more generally that the record of the Commission in
Court may not be all that impressive.
The way in which the Commission develops and uses economic analysis therefore deserves attention.
We develop a framework to think about antitrust procedures and identify the factors that will influence
how theories and evidence are handled. We identify the system of proof taking implemented by the
Commission, which is mostly inquisitorial with a prosecutorial bias, and discuss this system of proof-
taking at greater length in light of the literature. We observe that the reforms implemented by the
Commission go in the right direction, suggest some additional reforms and discuss the implementation
of an alternative, adversarial, regime of proof taking.
13
DG Comp discussion Paper on the application of Art. 82 or the Treaty to exclusionary abuses, December 2005, available from
http://ec.euroPa.eu/comm/competition/antitrust/others/Art.icle_82_review.html#14062006
14
In what follows, we will give references of the decisions and judgments only when they appear in the discussion for the first time.
15
Case COMP/M2416 and judgments Case T-5/02 at the CFI and Case C-12/03 P at the Court of Justice
16
Case M 2283 and judgment C 380/01
17
Case COMP/M2220 and judgment Case T-210/01
COMPETITION ECONOMICS AND ANTITRUST IN EUROPE _____
5
The paper is organized as follows. Section 2 attempts to provide some quantitative measure of the role
that economists have played in European Antitrust. Section 3 discusses the economic insights that have
had an effect on case law and policy. Section 4 provides a framework to analyze anti-trust proceedings
along five dimensions: namely, the scope of the decision (what has to be proven?), the system of proof
taking (how is the proof gathered?), the standard of proof (what should be the degree of confidence in
the proof?), the type of evidence which is deemed sufficient to meet the standard of proof (what
elements of proofs should be considered as sufficiently telling to conclude that the required degree of
confidence is reached?) and the standard of review (how is the proof assessed in case of appeal
18
?).
Section 5 characterizes EU procedures in terms of these dimensions. Section 6 discusses in more detail
the system of proof taking in light of the law and economic literature. Section 7 summarizes our
findings and discusses the scope for further reforms. Section 8 concludes
19
.
2. ECONOMIC INPUTS
Economic advice was marginal in antitrust proceedings up until the late eighties. It was undertaken
mostly by individual academics (there are references to some of them in early decisions like Soda/Ash
20
or Wood Pulp
21
). With the implementation of the merger regulation in 1990, demand for economic
advice seems to have risen. NERA opened an office in London in 1984 and London Economics was set
up in 1986. Lexecon (Ltd) was set up in January 1991 and up until the mid nineties, Lexecon, London
Economics and NERA were the main suppliers with a total amount of fees around £ 2.5 million in 1995.
This turnover corresponds to EU related competition work but also to competition work in national
jurisdictions. UK related work accounts for the vast majority of the latter. The market for EU related
advice grew rapidly in the late nineties, as the number of merger notifications (as well as other types of
cases) grew but also following the preparation and implementation of the notice on market definition.
This notice
22
, inspired by the US practice, used economic concepts explicitly
23
. As indicated by figure
1
24
, for the following ten years, total turnover grew
25
at some 25-30% per year, reaching about £ 24
million in 2004
26
.
It is also interesting to consider the turnover of economic consultancy relative to the turnover for legal
advice. Lexecon Ltd estimated that economic consultancy amounted to about 5 % of the total amount of
fees (legal and economic) in 1995
27
.
18
Appeal is somewhat of an abuse of language as EU courts formally only exercise a judicial review. In what follows we will still use “appeal” for
ease of reference.
19
I have been involved in a number of cases discussed in the Paper and in Part.icular Volvo/Scania, Airtours/First Choice, EMI/Time Warner,
TotalFina/Elf and Tetra Laval/Sidel. My discussion of these cases relies on public information only.
20
Case T36/91
21
Joined cases C-89/85, C-104/85, C-114/85, C-116/85, C-117/85 and C-125/85 to C-129/85.
22
Notice on the definition of the relevant market for the purposes of Community competition law. OJ C 372 on 9/12/1997
23
The fact that a quantitative analysis was used for market definition in the high profile acquisition of Perrier by Nestlé in 1992 may also have been
significant in alerting legal advisors to the potential of economic analysis in this regard.
24
The aggregate turnover has been obtained by adding the antitrust turnover of Lexecon, NERA, London Economics, Frontier, OXERA, RBB
Economics, LBE, CRA and LECG. Figures for some individual firms are confidential and cannot be reported individually. Others have been
estimated on the basis of the number of staff. Some of the figures have been interpolated on the basis of a constant growth. Independent
consultancy firms on the continent, which have remained small over the period, have not been considered. The turnover of independent academics,
which was probably significant in the earlier years relative to the turnover of commercial firms has not been considered. Traces of the role played
by these academics (in particular B. Yamey, G. Yarrow and D. Morris) can be found in some UK cases.
25
This rapid growth is to some extent a consequence of the fact that different parties in a competition case often have different interests – or, in other
words, “where a single economist starves, two will make a living”.
26
This growth gives a biased estimate of the growth of competition work in Europe as some firms (like Lexecon) started to generate very
substantial fees from work outside Europe (in particular South Africa).
27
The turnover of legal advice was estimated as follows: at the time, law firms in the UK had to obtain insurance from a common industry scheme.
They had to publish their turnover for this purpose. In order to obtain the fees related to antitrust, it was assumed that each partner would generate
the same amount of fees (an assumption which was validated with law firms) and partners undertaking mostly antitrust work were identified). These
