Tacit Collusion: The Neglected Experimental Evidence

• Published date: 01 September 2015
• Date: 01 September 2015
• Author: Christoph Engel
• DOI: http://doi.org/10.1111/jels.12080

Tacit Collusion: The Neglected

Experimental Evidence

Christoph Engel*

Both in the United States and in Europe, antitrust authorities prohibit merger not only if

the merged entity, in and of itself, is no longer sufficiently controlled by competition, but

also if, post merger, the market structure has changed such that “tacit collusion” or

“coordinated effects” become disturbingly more likely. It seems that antitrust neglects the

fact that for more than 50 years, economists have been doing experiments on this very

question. Almost any conceivable determinant of higher or lower collusion has been tested.

This article standardizes the evidence by way of a meta-study, and relates experimental

findings as closely as possible to antitrust doctrine.

I. The Issue

Normally, hard scientific evidence on legal issues is a seriously scarce resource. More

than one empirical publication on a precisely defined doctrinal question is a rare event.

Against this backdrop, a problem in antitrust is salient. Both in the United States and in

Europe, antitrust authorities prohibit merger not only if the merged entity, in and of

itself, is no longer sufficiently controlled by competition. The authorities also intervene

if, post merger, the market structure has changed such that “tacit collusion” becomes

disturbingly more likely. This can be seen as a precautionary measure. Collusion in gen-

eral, and tacit collusion in particular, is hard to detect, and even harder to control by

authorities bound by the rule of law. Therefore, the antitrust authorities do not wait

until market participants actually team up against the opposite market side, and against

society at large; rather, they preempt later anticompetitive behavior by prohibiting the

further concentration of a market.

Antitrust authorities mainly use the traditional doctrinal approach to assess whether

a proposed merger would significantly increase the risk of tacit collusion. Based on their

experiences in dealing with mergers in narrow markets, both in the United States and in

Europe, the authorities have published catalogues of potentially relevant factors (U.S.

*Max Planck Institute for Research on Collective Goods, Kurt-Schumacher-Straße 10, D 53113 Bonn; email:

engel@coll.mpg.de.

537

Journal of Empirical Legal Studies

Volume 12, Issue 3, 537–577, September 2015

Merger Guidelines;

1

E.U. Merger Guidelines

2

). On the initiative of the chief economist,

the E.U. Commission also relies on game theory for this purpose (Ivaldi et al. 2003).

There also is rich econometric evidence (Feuerstein 2005) and a host of case studies in

the economics literature (Levenstein & Suslow 2006). Yet this does not exhaust the avail-

able evidence. On this prognostic question, economists, for more than 50 years, have con-

ducted experiments. Almost any conceivable determinant of higher or lower collusion has

been tested. This solid body of evidence is almost untapped by the legal community. Anti-

trust authorities appear to take little notice of the fact that one of their thorniest empiri-

cal questions has already been thoroughly investigated by a neighboring discipline.

This article makes the evidence available. It standardizes the findings by way of a

meta-study, and it matches the experimental findings as closely as possible to doctrine. In

preparation, the key concepts of the doctrine of tacit collusion in the United States and in

Europe are reported (Section II), and the methodology of the experiments, and of the

meta-study summarizing this literature, is explained (Section III). Antitrust can capitalize

on the experimental evidence at two levels. It is most helpful in evaluating and further

developing the guidelines that authorities on both sides of the Atlantic have issued for

assessing the “coordinated effects” of mergers. The individual criteria in these checklists

can be related to the experimental evidence. More importantly, the relative importance of

these criteria can be determined (Section IV). Individual cases combine these criteria in idi-

osyncratic ways. Although many parameter constellations have been tested experimentally,

an experimental result on exactly the parameter combination of a given case is often not avail-

able.Basedonthemeta-study,itissometimespossibletopredicttheimportanceofspecific

factor combinations for the risk of tacit collusion, but the data on multiple interactions are

usually not significant. A statistically less demanding approach is multiple regression. It intro-

duces the features of the case as controls and checks whether a reduction in the number of

suppliers still substantially increases the risk of collusion. The power, and the limits, of the

approach are demonstrated with respect to the European landmark case on tacit collusion,

the Airtours ruling (Section V). Meta-analysis often is used to check whether the existing body

of evidence suffers from discernible bias since studies that have found the desired effect were

more likely to be published. Happily, with these data, there is not much reason for this con-

cern. However, since experimental data are, by design, only analogous to the issue in the field

one wants to understand, external validity warrants discussion (Section VI).

II. Doctrine

Under Section 7 of the Clayton Act, mergers are prohibited if their effect “may be sub-

stantially to lessen competition, or to tend to create a monopoly” (15 U.S.C. 18). Under

1

Department of Justice, Federal Trade Commission, Antitrust Division, 1992 Horizontal Merger Guidelines of

September 10, 1992, 57 FR 41552, Preface; Horizontal Merger Guidelines of August 19, 2010, available at

<http://www.justice.gov/atr/public/guidelines/hmg-2010.html>.

2

Guidelines on the Assessment of Horizontal Mergers Under the Council Regulation on the Control of Concen-

trations Between Undertakings, OJ 2004 C 31/5.

538 Engel

Section 1 of the Sherman Act, mergers are prohibited if they constitute a “contract,

combination ..., or conspiracy in restraint of trade” (15 U.S.C. 1). Both the Antitrust

Division of the Department of Justice and the Federal Trade Commission have jurisdic-

tion to intervene. To increase predictability, in 1992 both authorities issued joint guide-

lines on the treatment of horizontal mergers. Although most of these guidelines were a

restatement of court decisions and administrative practice, the section on potential coor-

dinated and unilateral effects of mergers was meant to innovate (U.S. Merger Guidelines

1992, Preface). In light of both the experiences of the authorities and the evolution of

the accompanying scholarly debates, in 2010 the guidelines were revised and extended

(U.S. Merger Guidelines 2010).

The guidelines stress that the anticompetitive effects of merger are not confined

to increased market power for the new commercial unit (U.S. Merger Guidelines 2010,

#1). Besides “unilateral effects,” a merger may also be challenged if it creates a risk of

“coordinated effects” (U.S. Merger Guidelines 2010, #1). For coordinated effects, there

is no need to show that firms will explicitly team up; it suffices to make a case of

“parallel accommodating conduct.” This “includes situations in which each rival’s

response to competitive moves made by others is individually rational” (U.S. Merger

Guidelines 2010, #7). To that end, the agencies expressly engage in prediction: “The

Agencies examine whether a merger is likely to change the manner in which market

participants interact, inducing substantially more coordinated interaction” (U.S. Merger

Guidelines 2010, #7.1).

In Europe, under Article 3 II of the Merger Regulation, “a concentration which

would significantly impede effective competition, in the common market or in a sub-

stantial part of it, in particular as a result of the creation or strengthening of a domi-

nant position, shall be declared incompatible with the common market” (OJ 2004 L 24/

1). According to the case law of the European Court of Justice, the test includes situa-

tions of “collective dominance.” In the Airtours case of 2002,

3

coordinated effects were

center stage, with a rich echo in academic writing (Motta 2000; Christensen & Rabassa

2001; Haupt 2002; Overd 2002; Stroux 2002; Guerrero 2003; Nikpay & Houwen 2003;

Scott 2003; Spink & Ong 2003; Veljanovski 2004; Kokkoris 2005). However, the doctrinal

concept is as old as the Nestle/Perrier case of 1992.

4

It was approved by the European

Court of Justice in the Kali und Salz case of 1998

5

(Perez 1998; Ysewyn & Caffarra 1998).

In Airtours, the European Court of First Instance held that a merger may be pro-

hibited if, in light of “the relationship of interdependence ... the parties ... are in a

3

Court of First Instance, 6 June 2002, Airtours plc v. Commission of the European Communities, European Court

Reports 2002 II 2585.

4

Commission Decision, 22 July 1992, OJ 1992 L 356/1; see also Commission Decision, 14 December 1993, OJ

1994 L 186/38, Kali1Salz/MdK/Treuhand.

5

European Court of Justice, 31 March 1998, European Court Reports 1998 I 1519, para. 221; see also European

Court of First Instance, 25 March 1999, European Court Reports 1999 II 753, Gencor and on this case Albors-

Llorens (2000); Commission Decision, 26 October 2004, OJ 2005 L 218/6, Oracle/People Soft and on this case

Pflanz (2005).

539Tacit Collusion