Reconciling the Conflict: Antitrust Leniency Programs and Private Enforcement
Thomas Knight
University of Florida
[email protected]
Casey Ste. Claire
University of Florida
[email protected]
ABSTRACT
Following the recent passage of Directive 2014/104/EU in the European Union,
there has been a renewed concern that private antitrust enforcement reduces
the efficacy of cartel leniency programs. We discuss the benefits of private
enforcement and use a simple dynamic model of collusion to examine whether
the introduction of antitrust liability in private suits reduces the attractiveness
of leniency programs, or overall deterrence. We show that private
enforcement may reduce the number of firms that seek leniency, but that those
firms that choose not to seek leniency cease to collude nonetheless. They
defect on the cartel agreement, but choose not blow the whistle to avoid
follow-on claims. Private enforcement can always be used to increase overall
deterrence.
1
I. Introduction
Price-fixing cartels are regularly viewed as the most egregious of all antitrust
violations. These agreements involve a concerted effort to restrict competition to
the detriment of consumer welfare and market efficiency. It is difficult, if not
impossible, to identify an economic justification for allowing these practices. Stated
by Lande and Davis (2011), “if a country were to have only one type of antitrust
violation, surely it would be against horizontal cartels.” Accordingly, these
agreements are per se violations of Section 1 of the Sherman Act in the United States,
and of Article 101 of the Treaty on the Functioning of the European Union.1
Antitrust law, in the United States, is enforced by both public antitrust
authorities and private parties. Recent developments in the European Union mark a
rising trend in the prevalence of private antitrust enforcement. Following the
Courage Ltd. v. Crehan ruling, however, European citizens have a right to
compensation for harm suffered due to competition violations.2 This ruling
establishes a right to sue for violations of European competition policy, but it also
restricts the European Commission and member state governments from extending
leniency in private suits. This creates a serious concern regarding the efficacy of
cartel leniency programs — where some combination of fine reductions and
reduced criminal sanctions are offered to companies and individuals that confess
For a more thorough review of antitrust enforcement differences between the
United States and the European Union, see Ginsburg (2005).
2 Case C-453/99, Courage Ltd. v. Bernard Crehan (E.C.J. Sept. 20, 2001)
1
2
participation in cartel agreements — because private follow-on claims liability could
dissuade potential leniency applicants from blowing the whistle.3
While leniency programs and private antitrust lawsuits both provide a
deterrent effect individually, numerous antitrust scholars and practitioners have
voiced concern that private lawsuits undermine the efficacy of leniency programs.
Wils (2003), Wils (2009), Canenbley and Steinvorth (2011), Cauffman and Philipsen
(2014), Migani (2014), Kirst and Van den Bergh (2015), and Knight and de Weert
(2015) all argue that when private damages liability exists – for example, with
follow-on claims – seeking leniency may be less desirable. Blowing the whistle and
seeking leniency could increase the likelihood of facing a private antitrust suit.
Accordingly, the incentive to seek leniency may be weakened in the presence of
private damages liability. This argument, however, ignores the point that private
damages lawsuits reduce the probability of sustained collusion. When cartels are
being privately monitored, their continued operation becomes less likely. This effect
increases the incentive to defect on a cartel agreement. In this paper, we show that
private antitrust lawsuits can enhance deterrence and improve the profitability of
seeking leniency.
We proceed in Section 2 by identifying and describing the possible benefits of
increased private enforcement of antitrust policy. In Section 3, we review the
deterrent effect of antitrust leniency programs. Section 4 presents a simple dynamic
A follow-on claim is a private suit in which the plaintiff seeks damages from a
defendant that has already been found guilty of an antitrust violation.
3
3
model of collusion and employs that model to demonstrate that private cartel
enforcement can always enhance overall deterrence and may enhance the
attractiveness of leniency programs. We show that it is necessary to distinguish
between deterrence and the attractiveness of leniency, because some firms may
choose to defect on a collusive agreement, but not seek leniency, to avoid potential
follow-on claims.
II. Private Enforcement
Private antitrust lawsuits, in which individuals and companies claim
compensation for harm suffered due to antitrust violations, are an important
component of antitrust enforcement. Private damages lawsuits account for 90-95%
of antitrust enforcement in the United States, but have historically played a minimal
role in European jurisdictions.4 Recent efforts by the European Commission,
however, have led to a framework for developing the private enforcement
mechanism in EU Member States. Signed into law in November 2014, Directive
2014/104/EU provides a set of guidelines for the EU Member States. These
guidelines are to be fully implemented in 2016. These guidelines include
prescriptions for discovery of evidence, the establishment of joint and several
4
See Canenbley and Steinvorth (2011)
4
liability in cartel cases, and the role of leniency statements can play in follow-on
claims.5
Private damages suits offer a number of benefits to the overall efficacy of an
antitrust regime by raising the probability that an antitrust violation is discovered
and by increasing the sanction applied to ones that are.6 Lande and Davis (2011)
argue that private enforcement provides a much stronger deterrent effect than
public enforcement in American cartel cases. These private suits address three
issues that arise with public enforcement. First, as pointed out by Motta and Polo
(2003); Aubert, Rey, and Kovacic (2006); and Chen and Rey (2013), public
enforcement may be limited by an administrative budget constraint. Private
enforcement can expand overall enforcement activity when the antitrust authority
is resource constrained. Second, Shavell (1984); Kovacic (2000); Segal and
Whinston (2006); McAfee, Mialon, and Mialon (2008); and Motchenkova and
Leliefeld (2010) argue that plaintiffs likely possess information of violations that the
government does not. Third, private antitrust suits compensate the victims of
antitrust violations. The desirability of establishing compensatory justice has been a
key motivator in the recent passage of Directive 2014/104/EU and the acceptance
within that directive of the so-called passing-on defense.
A leniency statements is an admission of guilt in which a cartel member provides
evidence of the collusive agreement. Directive 2014/104/EU states that these
statements shall be inadmissible in follow-on lawsuits. However, in Pfleiderer AG v.
Bundeskartellamt, the ECJ ruled that evidence provided in a leniency application is
not shielded from discovery in national courts.
6 See Becker (1968).
5
5
III. Leniency Programs
Antitrust leniency programs are viewed as the most important cartel
enforcement tool.7 These programs have attracted considerable scholarly attention.
Scholars have convincingly demonstrated that these programs enhance overall
deterrence and reduce instances of cartelization.
The theoretical leniency literature identifies a tradeoff between ex ante
deterrence and ex post desistence. Reduced fines or criminal liability may
destabilize cartels (desistence effect), but they also reduce the expected cost of
cartel participation (deterrence effect). Motta and Polo (2003) identify this tradeoff
and introduce the question of whether leniency programs foster collusion by
reducing the expected cost of an infringement or if they destabilize existing cartels
by increasing the benefit of blowing the whistle. Harrington (2008) further
examines the former effect, referring to it as the Cartel Amnesty Effect, the channel
through which leniency fosters cartel formation by reducing the expected cost of
collusion. Harrington (2008), however, argues that this effect will only dominate
when leniency is weak – as opposed to absolute leniency, in which the leniency
recipient is granted full amnesty. When full amnesty is available to leniency
applicants, leniency reduces overall collusive behavior. Spagnolo (2003) and Bigoni
et. al. (2012) find, similarly, that leniency programs have a positive deterrent effect.
A small empirical literature examines the characteristics of detected cartels
after substantial changes in US and EU leniency programs and seeks to identify
7
See Sokol (2012).
6
support for a positive deterrent effect. Harrington and Chang (2009) argue that the
introduction of a leniency program should be followed by: 1) an increased number
of detected cartels, 2) an increase in the average duration of detected cartels in the
period directly following the introduction of leniency, and 3) a decrease in the
average duration of detected cartels in the longer run. Miller (2009) identifies these
characteristics and reports that the US leniency program is responsible for a 59%
lower cartel formation rate and a 62% higher detection rate. Evidence from the EU,
on the other hand, is mixed. Brenner (2009) does not identify an increase in cartel
duration after the introduction of the EU leniency program; Zhou (2013), however,
identifies an increased duration of detected cartels in the short run and a decreased
duration of detected cartels in the longer run. Zhou (2013) explains the
inconsistency between his results and those of Brenner (2009) by stating that the
latter did not correctly operationalize the short and longer runs. While it is
impossible to observe cartel formation and collusive behavior explicitly, recent
theoretical and empirical analyses provide convincing evidence that leniency
programs reduce the number of cartel agreements, which warrants an inquiry into
whether their efficacy is diminished by private enforcement.
IV. Private Damages and the Efficacy of Leniency
Scholars and policy-makers alike express concerns that antitrust leniency
programs and private enforcement may be in conflict. More precisely, there is a fear
that private damages suits undermine the efficacy of leniency programs. When
7
considering whether to apply for leniency, colluding firms may fear that doing so
would expose them to an increased risk of private damages liability. We employ a
simple dynamic model of collusion to examine whether private damages suits
actually weaken the efficacy of leniency programs. We find that private damages
lawsuits likely reduce the number of leniency applications, but that very low and
very high private damages awards reduce cartel stability.
This section proceeds by presenting a simple dynamic model of collusion,
introducing public enforcement and leniency into that model, exploring when a
“defecting” firm would choose to seek leniency in the presence of private
enforcement as opposed to simply defecting on the cartel agreement without seeking
leniency, and characterizing the conditions under which public enforcement with
leniency and private enforcement together increase the likelihood of cartel
agreements breaking down.
Baseline Model: No Antitrust Enforcement
We employ a simple dynamic model of collusion, in which infinitely
repeated play may allow firms to overcome the Prisoner’s Dilemma. Markets
characterized by imperfect price competition usually place firms in a Prisoner’s
Dilemma. Firms enjoy the greatest possible amount of joint profit when they agree
to charge a high price, but each individual firm has an incentive to undercut its
rivals. This individual incentive leads to a price war, in which prices fall in the
8
direction of firms’ marginal production costs. Firms can avoid this tendency – and
enjoy more profit – if they form a cartel.8
Table 1 illustrates provides a normal form representation of a two-player
2x2 Prisoner’s Dilemma game. We define the payoffs in this game as follows: 1) 𝜋 𝐶
denotes the profits each firm earns when both firms uphold the cartel agreement, 2)
𝜋 𝐷 denotes the profit earned by a firm that defects on the cartel agreement when
the other firm upholds the agreement, and 3) 𝜋 𝑁 denotes the profit earned by both
firms when neither upholds the collusive agreement. By assumption, the market in
question is characterized by a Prisoner’s Dilemma, and 𝜋 𝐷 > 𝜋 𝐶 > 𝜋 𝑁 . This condition
implies that (Defect, Defect) is the Pure Strategy Nash Equilibrium of the singleinteraction game, and 𝜋 𝑁 can be considered the level of competitive profits.
Firm 1 \\ Firm 2
Collude
Defect
Collude
𝜋𝐶 , 𝜋𝐶
___ , 𝜋 𝐷
Defect
𝜋 𝐷 , ___
𝜋𝑁 , 𝜋𝑁
Table 1: One-Off Prisoner’s Dilemma Game
When this Prisoner’s Dilemma game is indefinitely repeated, firms may be
able to sustain a cartel agreement. That is, it may be incentive compatible to collude.
This possibility arises, because firms can punish defections in subsequent time
periods. We consider the Grim Trigger punishment strategy. The Grim Trigger
Strategy is a strategy in which a firm agrees to collude in each time period if their
8
See Green and Porter (1984).
9
rivals have colluded in the previous time period. Otherwise, the firm defects in the
current time period and in all subsequent time periods. The complete breakdown of
collusion appears most appropriate when defections are coupled with a leniency
application. If 𝛿 denotes a firm’s individual discount rate, it is in the firm’s best
interest to uphold a cartel agreement if:
𝜋𝐶
1
𝛿
≥ 𝜋𝐷 + 𝜋𝑁
(1 − 𝛿)
(1 − 𝛿)
That is, if the discounted payoffs of receiving 𝜋 𝐶 in the current time period and all
subsequent time periods exceeds the discounted payoffs of receiving 𝜋 𝐷 in the
current time period and 𝜋 𝑁 in all subsequent time periods, indefinite collusion is
incentive compatible.
We can solve for a threshold discount rate, 𝛿̃, above which a firm would
continue to uphold the collusive agreement. A collusive agreement is sustainable
whenever all firm’s individual discount rates exceed 𝛿̃. In the remainder of this
section, we examine how different antitrust policies affect 𝛿̃. When this threshold
increases, collusion becomes less likely. In the absence of antitrust enforcement, the
threshold is given as:
𝛿̃ =
𝜋𝐷 − 𝜋𝐶
𝜋 𝐷 − 𝜋𝑁
We obtain this critical threshold by treating the preceding inequality as an equality
and solving for δ. For the remainder of this section, we normalize 𝜋 𝑁 = 0 for
tractability. This assumption does not affect the results that we obtain.
10
Public Enforcement with Leniency
We introduce public enforcement with leniency to the simple dynamic model
presented above. As before, we assume that a firm can either continue to collude or
defect on the cartel agreement. The firm, however, now has the added consideration
that if it chooses to collude, it faces the possibility of being discovered by the
antitrust authority and being sanctioned. This makes collusion less desirable for two
reasons: 1) if a firm chooses to collude, it faces the possibility of being sanctioned,
and 2) there is a possibility that collusion breaks down to due detection by the
antitrust authority. This second concern is only apparent when examining collusion
in a dynamic framework.
Introducing public enforcement requires the addition of two new variables to
the model. We define p as the probability that a firm is detected and sanctioned by
the antitrust authority, and we define F as the applied sanction. After normalizing
𝜋 𝑁 = 0, the discounted expected value of colluding in a particular time period is
given as:
𝐸𝑉 𝐶 = 𝑝[𝜋 𝑁 − 𝐹] + (1 − 𝑝){𝜋 𝐶 + 𝛿[𝑝(𝜋 𝑁 − 𝐹) + (1 − 𝑝)𝜋 𝐶 ] + 𝛿 … ,
which simplifies to:
𝐸𝑉 𝐶 =
(1 − 𝑝)𝜋 𝐶 − 𝑝𝐹
1 − 𝛿(1 − 𝑝)
In the absence of private enforcement, we assume that if a firm chooses not
to collude, the firm seeks leniency. If a firm is going to break from the cartel
agreement, it should also seek leniency. Comparing 𝐸𝑉 𝐶 to payoffs associated with
11
defecting on the cartel agreement and receiving leniency, we identify a threshold
discount rate. If all firms in the agreement have a discount rate that exceeds this
threshold, collusion persists. Otherwise, the cartel breaks down. This threshold
discount rate is given as:
𝛿̃𝑃𝑈𝐵 =
𝜋 𝐷 − 𝜋 𝐶 + 𝑝(𝜋 𝐶 + 𝐹)
(1 − 𝑝)𝜋 𝐷
(1)
We make two observations regarding this expression for 𝛿̃𝑃𝑈𝐵 :
1) If there is no chance of being detected by public enforcement (i.e., p = 0), the
discount threshold with public enforcement and leniency simplifies to the
discount threshold without enforcement.
2) 𝛿̃𝑃𝑈𝐵 is strictly greater than the baseline value for all positive values of p.
That is, public enforcement with leniency makes collusion less likely than
when there is no sanction for collusion. This finding arises, because public
enforcement with leniency reduces the expected value of choosing to collude
in a particular time period, but it does not affect the expected value of
choosing to defect.
We proceed by introducing private enforcement into this setting. More specifically,
we examine how the threshold discount rate changes when public enforcement with
leniency and private enforcement operate side-by-side.
12
Leniency and Private Damages
We now direct our attention to a setting with both public and private
enforcement. When firms choose to uphold a collusive agreement, they face the
possibility of detection by the antitrust authority and detection by a private firm or
citizen. If either or both of these forms of detection occur, the cartel breaks down
without any firm explicitly choosing to defect. We present the possible outcomes
and probabilities that each arises, when a firm chooses to uphold a collusive
agreement, in Table 2.
Probability
Payoff
p
(1-p)q
(1-p)(1-q)
𝜋𝑁 - F - D
𝜋𝑁 – D
(collusion ends)
(collusion ends)
𝜋𝐶
(collusion continues)
Table 2: Possible Outcomes When Colluding
When evaluating the incentive to uphold a cartel agreement in the presence
of public enforcement with leniency and private enforcement, one must specify
whether the firm seeks leniency if it does not choose to uphold the agreement. Since
seeking leniency increases the probability that a firm faces a private antitrust suit, it
may be reasonable to defect on the cartel agreement, but not blow the whistle. We
suppose, here, that if a firm seeks leniency, they are liable for damages D with
probability q=1. If a firm does not seek leniency, it faces a fine F with probability p
and damages liability D with probability q. We find that a firm that chooses to defect
on a cartel agreement seeks leniency if:
13
𝐷<
𝑝
𝐹
(1 − 𝑝)(1 − 𝑞)
(2)
Otherwise, a firm that defects on a cartel agreement will do so without seeking
immunity under a leniency program. If the size of follow-on claims is sufficiently
large, the net gains of seeking leniency become negative. By blowing the whistle, the
firm exposes itself to private damages liability.
When private damages liability are sufficiently low, firms that choose to
defect on a collusive agreement will seek leniency. The resulting threshold discount
rate above which firms choose to uphold the cartel agreement is:
𝛿̃𝑃𝑈𝐵+𝑃𝐷 =
𝜋 𝐷 − (1 − 𝑝)(1 − 𝑞)(𝜋 𝐶 + 𝐷) + 𝑝𝐹
(1 − 𝑝)(1 − 𝑞)(𝜋 𝐷 − 𝐷)
We observe that when this threshold discount rate is less than one, it is decreasing
in the amount of private damages liability, D, at an increasing rate. If a firm chooses
to seek leniency, it will face a follow-on claim of D. Therefore, any increase in D
enhances cartel stability, assuming that a defection is followed by an application for
leniency. The threshold above is increasing in q. That is, as the probability of being
successfully sued for collusion – which results in the breakdown of collusion – rises,
the expected value of collusion declines.
𝑝
When 𝐷 < (1−𝑝(1−𝑞) 𝐹, a firm that defects on the cartel agreement chooses to
seek leniency. It is easy to demonstrate that when D=0 and q=0, introducing private
damages neither increases nor decreases the threshold discount rate, relative to the
14
threshold provided by Equation 1. As q increases above zero, this threshold discount
rate increases above that provided by Equation 1. That is, for all positive values of q,
when D=0, introducing private damages decreases cartel stability. This arises,
because even for small values (or zero) of D, the possibility of being detected by the
private enforcement mechanism reduces the appeal of deciding to sustain collusion.
We have assumed that if a colluding firm is discovered, collusion breaks down.
When private damages liability is sufficiently high that firms that choose to
defect on a collusive agreement opt not to seek leniency as to potentially avoid
follow-on claims, the threshold discount rate above which firms choose to uphold
the cartel agreement is given as:
𝛿̃𝑃𝑈𝐵+𝑃𝐷 =
𝜋 𝐷 − (1 − 𝑝)(1 − 𝑞)𝜋 𝐶
(1 − 𝑝)(1 − 𝑞)[𝜋 𝐷 − 𝑝(𝐹 + 𝐷) − (1 − 𝑝)𝑞𝐷]
If this discount rate is less than one, it is increasing in, D, at an increasing rate. It is
𝑝
also increasing in q. Interestingly, when 𝐷 = (1−𝑝(1−𝑞) 𝐹, the two thresholds are
equal, implying that there is no discontinuity at the threshold discount rate.
Figure 1 presents the threshold discount rate as a function of private
damages liability, D. We also demonstrate how this discount rate changes as the
probability of being liable for such damages, q, rises. Whenever this threshold
discount rate is above the threshold provided by Equation 1, private enforcement
reduces cartel stability. Figure 1 provides both the threshold discount rate before
the introduction of private enforcement, 𝛿̃𝑃𝑈𝐵 , and the threshold discount rate after
the introduction of private enforcement, 𝛿̃𝑃𝑈𝐵+𝑃𝐷 . It demonstrates how 𝛿̃𝑃𝑈𝐵+𝑃𝐷 is
15
affected by changes in q. We let 𝛿̃′𝑃𝑈𝐵+𝑃𝐷 denote the threshold discount rate when q
𝐷
𝐶
−𝑝𝐹[𝑝𝜋 −(1−𝑝)𝜋 +𝑝𝐹]
= 0 and 𝛿̃′′′𝑃𝑈𝐵+𝑃𝐷 denote the threshold discount rate when 𝑞 = 𝜋𝐷[𝜋𝐷+𝑝(1−𝑝)𝐹] .
Importantly, this value of q, above which private enforcement reduces cartel
stability for all values of D, is less than one. 𝛿̃′′𝑃𝑈𝐵+𝑃𝐷 denotes the threshold discount
rate for some intermediate value of q.
̃
𝜹
𝛿̃′′′𝑃𝑈𝐵+𝑃𝐷
𝛿̃𝑃𝑈𝐵
𝛿̃′′𝑃𝑈𝐵+𝑃𝐷
𝛿̃𝑃𝑈𝐵
𝛿̃′𝑃𝑈𝐵+𝑃𝐷
𝛿̃𝑃𝑈𝐵
𝛿̃𝑃𝑈𝐵
̃’
𝐷
̃ ’’
𝐷
̃ ’’’
𝐷
𝑫
Figure 1: The Effect of D on Deterrence
As Figure 1 depicts, there are combinations of private enforcement
parameters q and D for which private enforcement enhances deterrence and
increases the attractiveness of the leniency program. When D is very low, private
16
enforcement reduces the attractiveness of upholding a collusive agreement, because
it introduces an increased likelihood that collusion breaks down. As D increases,
however, this effect may become dominated by the potential for large follow-claims.
For very large values of D, defecting to avoid liability in future time periods causes
private enforcement to enhance deterrence once again.
We find that private enforcement can enhance deterrence by making
collusion less profitable. Interestingly, we find that very low and very high levels of
damages liability maximize deterrence, whereas intermediate levels are less
effective. We summarize our results as:
1) There exists a threshold value of q, above which collusion is made less
likely for all values of D.
2) For all positive values of q, there is a range of D over which private
enforcement reduces the likelihood of collusion.
3) For all positive values of q, very low values of D make collusion less likely
and leniency more attractive, because an increased risk of collusion
breaking down makes defecting more attractive.
4) For all positive values of q, very high values of D make collusion less
likely, but makes leniency less attractive. Firms avoid high damages
liability in subsequent time periods by defecting on the collusive
agreement, but they do not blow the whistle.
These results have important policy implications. They dispel the notion that private
enforcement will reduce the efficacy of leniency programs and weaken the deterrent
17
effect of cartel policy. Moreover, they offer insight into policy parameterizations that
lead to the greatest level of deterrence. The most important implications of these
results are that private damages suits do not necessarily reduce deterrence and the
attractiveness of leniency programs, and that heightened deterrence can be
achieved with very low amounts of damages liability.
A secondary, but very policy relevant result is that damage trebling may
weaken the efficacy of private enforcement. When damages are sufficiently low that
a firm that chooses to break with collusive agreement seeks leniency, damages
trebling may reduce deterrence, or enhance cartel stability. Increasing damages
liability makes firms less likely to blow the whistle and apply for leniency, because it
increases liability in follow-on claims.
IV. Concluding Remarks
We have shown that private suits do not necessarily reduce the efficacy of
leniency programs. Moreover, private enforcement can always offer enhanced
deterrence. Therefore, the public policy concern that private damages undermine
cartel enforcement is unwarranted.
In response to the adoption of Directive 2014/104/EU, which facilitates the
development of a stronger role for private antitrust enforcement, numerous
antitrust practitioners and scholars have questioned whether these suits reduce the
attractiveness of cartel leniency programs, and consequently reduce overall
deterrence. We respond to these concerns by demonstrating that private
18
enforcement can enhance overall deterrence, and in some of these instances, make
leniency more attractive.
Leniency programs share the common feature of providing reduced
sanctions – or even, total immunity – from an antitrust authority in exchange for
evidence of an illegal cartel agreement. In both the United States and the European
Union, successful leniency applicants cannot receive immunity from private
antitrust suits. This creates the possibility that cartel participants are less likely to
blow the whistle and seek amnesty, because they are fearful that they will face
follow-on claims. Accordingly, private antitrust enforcement may diminish the
effectiveness of leniency programs and reduce overall deterrence.
We present and discuss three reasons that private antitrust suits may
enhance enforcement, and thus be desirable. Issues related to resource constrained
antitrust authorities, information and cost asymmetries between private plaintiffs
and public antitrust authorities, and compensatory justice are discussed.
Additionally, we employ a simple dynamic model of collusion to demonstrate the
private enforcement can enhance deterrence, and strengthen leniency programs.
We show that private enforcement enhances deterrence when damages liability is
sufficiently small, because firms are not sufficiently dissuaded by the (small) followon liability. In this case, a firm that breaks with the cartel agreement chooses to seek
leniency. Additionally, deterrence is enhanced when damages liability is sufficiently
large, because firms wish to avoid damages liability in future time periods. In this
case, however, firms do not seek leniency; they simply break with the collusive
agreement. Lastly, when the probability of being detected by the private
19
enforcement mechanism is sufficiently high, any amount of damages liability
increases overall deterrence.
20
References
Aubert, C., Rey, P., & Kovacic, W. E. (2006). The impact of leniency and whistleblowing programs on cartels. International Journal of Industrial Organization, 24(6),
1241-1266.
Becker, G. (1968). Crime and punishment: an economic approach. Journal of Political
Economy, 76, 169.
Bigoni, M., Fridolfsson, S. O., Le Coq, C., & Spagnolo, G. (2012). Fines, leniency, and
rewards in antitrust. The RAND Journal of Economics, 43(2), 368-390.
Blair, R. D., & Knight, T. (2013). Compliance with corporate policy: an economic
approach. Managerial and Decision Economics, 34(7-8), 529-537.
Brenner, S. (2009). An empirical study of the European corporate leniency program.
International Journal of Industrial Organization, 27(6), 639-645.
Canenbley, C., & Steinvorth, T. (2011). Effective enforcement of competition law: is
there a solution to the conflict between leniency programmes and private damages
actions? Journal of European Competition Law & Practice, lpr026.
Cauffman, C., & Philipsen, N. J. (2014). Who Does What in Competition Law:
Harmonizing the Rules on Damages for Infringements of the EU Competition Rules?.
Maastricht European Private Law Institute Working Paper, (2014/19).
Chen, Z., & Rey, P. (2013). On the design of leniency programs. Journal of Law and
Economics, 56(4), 917-957.
Feess, E., & Walzl, M. (2004). Self‐reporting in Optimal Law Enforcement when there
are Criminal Teams. Economica, 71(283), 333-348.
Ginsburg, D. H. (2005). Comparing antitrust enforcement in the United States and
Europe. Journal of Competition Law and Economics, 1(3), 427-439.
Green, E. J., & Porter, R. H. (1984). Noncooperative collusion under imperfect price
information. Econometrica: Journal of the Econometric Society, 87-100.
Harrington, J. E. (2008). Optimal Corporate Leniency Programs*. The Journal of
Industrial Economics, 56(2), 215-246.
Harrington, J. E., & Chang, M. H. (2009). Modeling the birth and death of cartels with
an application to evaluating competition policy. Journal of the European Economic
Association, 7(6), 1400-1435.
21
Kirst, P., & Van den Bergh, R. (2015). The European directive on damages actions: a
missed opportunity to reconcile compensation of victims and leniency incentives.
Journal of Competition Law and Economics, nhv033.
Knight, T., & de Weert, W. (2015). On Implementing Private Damages in European
Competition Cases. University of Florida Working Paper.
Kovacic, W. E. (2000). Private monitoring and antitrust enforcement: paying
informants to reveal cartels. Geo. Wash. L. Rev., 69, 766.
Lande, R. H., & Davis, J. P. (2011). Comparative Deterrence from Private
Enforcement and Criminal Enforcement of the US Antitrust Laws. BYU L. Rev., 315.
McAfee, R. P., Mialon, H. M., & Mialon, S. H. (2008). Private v. public antitrust
enforcement: A strategic analysis. Journal of Public Economics, 92(10), 1863-1875.
Medinger, J. D. (2003). Antitrust Leniency Programs: A Call for Increased
Harmonization as Proliferating Programs Undermine Deterrence. Emory LJ, 52,
1439.
Migani, C. (2014). Directive 2014/104/EU: In Search of a Balance between the
Protection of Leniency Corporate Statements and an Effective Private Competition
Law Enforcement. Global Antitrust Review, 7, 81-111.
Miller, N. H. (2009). Strategic leniency and cartel enforcement. The American
Economic Review, 750-768.
Motchenkova, E., & Leliefeld, D. (2010). Adverse effects of corporate leniency
programs in view of industry asymmetry. Journal of Applied Economic Sciences, 5(2),
12.
Motta, M., & Polo, M. (2003). Leniency programs and cartel prosecution.
International journal of industrial organization, 21(3), 347-379.
Salop, S. C., & White, L. J. (1986). Treble Damages Reform: Implications of the
Georgetown Project. Antitrust Law Journal, 55(1), 73-94.
Segal, Ilya R., and Michael D. Whinston. "Public vs. private enforcement of antitrust
law: A survey." (2006).
Shavell, S. (1984). Liability Harm versus Regulation of Safety. Journal of Legal
Studies, 13, 357-374.
Sokol, D. D. (2012). Cartels, corporate compliance, and what practitioners really
think about enforcement. Antitrust Law Journal, 78(1), 201-240.
22
Spagnolo, Giancarlo. (2003). Optimal deterrence mechanisms against cartels and
organized crime, mimeo, Mannheim.
Vande Walle, S., (2013). Private Antitrust Litigation in the European Union and Japan:
A Comparative Perspective. Maklu.
Wils, W. P. (2003). Should private antitrust enforcement be encouraged in Europe?
World Competition: Law and Economics Review, 26(3).
Wils, W. P. (2007). Leniency in antitrust enforcement: Theory and practice. World
Competition: Law and Economics Review, 30(1).
Wils, W. P. (2009). The relationship between public antitrust enforcement and
private actions for damages. World Competition, 32(1).
Zhou, J. (2013). Evaluating leniency with missing information on undetected cartels:
Exploring time-varying policy impacts on cartel duration.
23