The Impact of the Corporate Leniency Program on Cartel Formation and the Cartel Price Path

• DOI: https://doi.org/10.1016/S0573-8555(06)82003-1
• Published date: 01 April 2007
• Date: 01 April 2007
• Pages: 59-80
• Author: Joe Chen,Joseph E. Harrington , Jr.

CHAPTER 3

The Impact of the Corporate Leniency

Program on Cartel Formation and the

Cartel Price Path

Joe Chenaand Joseph E. Harrington Jr.b

aFaculty of Economics, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo,113-0033 Japan

E-mail address: joechen@e.u-tokyo.ac.jp

bJohns Hopkins University, Baltimore,MD 21218, USA

E-mail address: joe.harrington@jhu.edu

Abstract

Previous research exploring the effect of corporate leniency programs has mod-

eled the oligopoly stage game as a Prisoners’ Dilemma. Using numerical analy-

sis, we consider the Bertrand price game and allow the probability of detection

and penalties to be sensitive to firms’ prices. Consistent with earlier results, a

maximal leniency program necessarily makes collusion more difficult. However,

we also find that partial leniency programs—such as in the U.S.—can make col-

lusion easier compared to offering no leniency. We also show that even if cartel

formation is not deterred, a leniency program can reduce the prices charged by

firms.

3.1. Introduction

One of the most important policy developments in antitrust policy in recent

decades is the revision of the Corporate Leniency Program by the U.S. De-

partment of Justice (DOJ) in 1993. Originally instituted in 1978, this program

allows corporations and individuals, who were engaging in illegal antitrust ac-

tivity (such as price-fixing), to receive amnesty from government penalties. This

means that a corporation can avoid government fines, while individuals escape

fines and prison sentences. The 1993 revision made it possible for amnesty to

be awarded even when an investigation had been started and made it a condi-

tion that the DOJ “has not received information about the illegal activity being

reported from any other source.” This means that amnesty is limited to one firm

per cartel. Leniency programs have proliferated as the European Commission

instituted one in 1996 and an increasing number of industrialized countries have

some form of leniency program. While it is difficult to assess the role of these

programs on cartel formation and collapse, we do know that it has been widely

CONTRIBUTIONS TO ECONOMIC ANALYSIS © 2007 ELSEVIER B.V.

VOLUME 282 ISSN: 0573-8555 ALL RIGHTS RESERVED

DOI: 10.1016/S0573-8555(06)82003-1

60 J. Chen and J.E. Harrington Jr.

used. Notable examples include Rhône-Poulenc in the vitamins case, Christie’s

in the fine arts auctions case, and Carbide/Graphite in the graphite electrodes

case.1

In light of the influence of leniency programs, it is not surprising that there

has been a growing amount of research exploring how such programs destabi-

lize collusion. Recent work includes Aubert et al. (2003),Motta and Polo (2003),

Spagnolo (2003),Feess and Walzl (2004),Motchenkova (2004), and Harrington

(2005a). This research has generally shown that leniency does reduce cartel

stability. In the context of a stationary environment (that is, the probability of

conviction without use of the leniency program is fixed over time), Spagnolo

(2003) shows that, if there is a budget-balancing constraint, a first-best solution

can be achieved by giving the first firm to come forward a reward equal to the

fines levied on the remaining firms. Motta and Polo (2003) allow the probability

of conviction to stochastically change over time though it is restricted to take

only two values, one of which is zero (which corresponds to the event that there

is no investigation). Their analysis provides qualified support for leniency. In

some cases, waiving a sufficiently high fraction of penalties can prevent cartel

formation, in which case such a policy is optimal. However,if it cannot prevent

cartel formation then no leniency should be provided. Harrington (2005a) also

allows the probability of detection and conviction to vary over time but it can

take any value from [0,1]. This introduces a new effect absent from Motta and

Polo (2003) which has the implication that, under certain conditions, more le-

niency can enhance cartel stability. It is then possible for partial leniency to be

optimal though plausible sufficient conditions are provided for it to be optimal

to waive all penalties for the first firm to come forward. It is also shown that

restrictions should be placed on when amnesty is awarded, though it can be op-

timal to award amnesty even when the antitrust authority is very likely to win

the case without insider testimony.

A common limitation to all of this research is that the impact of leniency pro-

grams is explored in a restrictive setting: The stage game modeling oligopolistic

interaction is the Prisoners’ Dilemma. This means that the collusive price and

profit are fixed, as are the price and profit associated with a firm cheating on the

cartel. It has the implication that the probability of the cartel being discovered

is exogenous to how the cartel behaves as is the penalty in the event of discov-

ery and successful prosecution. All of these restrictions affect the influence of

a leniency program. While a leniency program may be able to prevent cartel

formation, by fiat it cannot impact the cartel price in the event a cartel forms.

Furthermore, while a leniency program may prevent cartel formation when the

collusive price path is fixed (as it is with the Prisoners’ Dilemma), it may not be

able to do so if firms can strategically adjust the cartel price path so as to coun-

teract the destabilizing effect of a leniency program. Previous work has shown

1A good review of the status of leniency programs is provided in Hard Core Cartels (2003).A

critical description of the U.S. program can be found in Kobayashi (2001).