Justice News

International Anti-Cartel Enforcement
United States
Sunday, November 21, 2004
This document is available in three formats: this web page (for browsing content), PDF (comparable to original document formatting), and WordPerfect. To view the PDF you will need Acrobat Reader, which may be downloaded from the Adobe site. For an official signed copy, please contact the Antitrust Documents Group.


Presented by

Assistant Attorney General
Antitrust Division
U.S. Department of Justice

Before the
2004 ICN Cartels Workshop
Sydney, Australia

November 21, 2004


It is a pleasure to be here in Sydney for the ICN workshop on cartel enforcement. This cartel enforcement workshop has long been a key part of global anti-cartel efforts. The fact that the ICN is now involved is particularly gratifying. Because cartel enforcement is the core priority of antitrust law, it is vital that the ICN's work involve cartel enforcement. I can think of no better or more valuable substantive contribution to cartel enforcement than supporting this workshop. As a lawyer whose background is primarily in civil cases, I would not presume to lecture this gathering of the world's top anti-cartel experts on how to do your job. Rather, I am here primarily to thank you for your efforts on behalf of competition and consumers. No aspect of antitrust enforcement is more important than the fight against cartels.

In fact, we at the U.S. Justice Department see antitrust enforcement as a three part hierarchy. At the top of this hierarchy is enforcement against cartels, conduct that is devoid of any efficiency justification and inflicts tremendous harm on our economy. Our Supreme Court, in its recent Trinko decision, described collusive behavior as "the supreme evil of antitrust." Obviously, this is our core priority at the Antitrust Division. Second, we review mergers using the best analytical tools available, and make judgments on whether the effects of the merger may be "substantially to lessen competition or to tend to create a monopoly." If so, we must back up that judgment with a suit in court to block the merger. Third, we analyze unilateral conduct, as well as agreements subject to rule of reason analysis, in a cautious and objective manner. We do this mindful that it is often difficult to tell the difference between good, hard competition and anticompetitive conduct, but ready to challenge conduct that is harmful to competition.

There is now significant global consensus that cartel enforcement deserves our most focused attention. At the recent Fordham international antitrust conference, my friend Mario Monti described cartel enforcement as the highest priority work against private anti-competitive conduct, equal in importance to combating government restraints on competition. I agree entirely. Especially for countries that have taken up antitrust enforcement relatively recently, there can be no sounder way to develop a strong competition culture than to place primary emphasis on cartel enforcement. Consumers will benefit. Businesses that rely upon commodity inputs will benefit. Taxpayers and governments will benefit when bid-rigging is curtailed. And unlike other more subtle or controversial areas of competition law, there is no danger here that government intervention might have anti-competitive effects.

This morning I would like to address three topics. First, I will trace the development of the US anti-cartel enforcement program. Second, I will discuss the importance of amnesty programs in the fight against cartels. Finally, I will talk about the reasons why the United States continues to be a strong proponent of criminal sanctions as the most effective deterrent of cartels.

Development of the US Cartel Enforcement Program

The past thirty years have seen many milestones in the development and enhancement of U.S. anti-cartel laws and policies. The Antitrust Procedures and Penalties Act was enacted in 1974, making violations of the Sherman Act a felony and increasing the maximum corporate fine from $50,000 to $1 million. In 1978, our original Corporate Leniency Policy was announced. In 1990, the Sherman Act maximum corporate fine was increased from $1 million to $10 million. In the early 1990s, the Division recognized a need to revamp its leniency program, and in August 1993 a new Corporate Leniency Policy was issued. The amnesty program was enhanced further in 1994 with the issuance of an Individual Leniency Policy. As business became more globalized in the 1990s, the focus of the Division's enforcement efforts shifted to the detection and prosecution of international cartels, which inflict the greatest harm on U.S. businesses and consumers. The $100 million fine obtained in 1996 from Archer Daniels Midland Company for its participation in the international lysine and citric acid cartels marked the beginning of this new era in antitrust prosecutions. This record fine was surpassed beginning in 1998 by even larger fines against members of the international graphite electrode cartel, in 1999 by the $500 million and $225 million fines against F. Hoffmann-La Roche and BASF for their participation in the international vitamin cartel, and most recently by the $160 million fine agreed to by Infineon Technologies AG in our first prosecution against the international DRAM cartel. To date, the Division has obtained corporate fines at or above the former $10 million statutory maximum in 46 cases and fines of $100 million or more against seven corporations.

The enormity (in every sense of the word) of these cartels and the harm they inflict has recently required that our cartel laws be revised again to increase the maximum sanction that could be imposed for antitrust violations. In June, President Bush signed the Antitrust Criminal Penalty Enhancement and Reform Act of 2004, increasing the maximum corporate Sherman Act fine to $100 million. The Act also enhanced individual deterrence by increasing the maximum individual Sherman Act fine to $1 million and the maximum Sherman Act prison sentence to 10 years to make it more consistent with maximum prison sentences for other white collar crimes. And the legislation increased the incentives for firms to blow the whistle on cartels they have participated in by enhancing the Corporate Leniency Policy. Under the new legislation, the damages paid by a corporate amnesty applicant to private plaintiffs are detrebled and reduced to actual damages attributable to the applicant's sales of the affected product or service if the applicant cooperates with the plaintiffs in their efforts to recover joint and several treble damages from the other cartel members.

As we shifted our focus to the prosecution of international cartels in the 1990s, we also realized that our law enforcement relationships with other governments had to be strengthened to meet the investigative and prosecutorial challenges of international cartels. We stepped up our efforts to enter antitrust cooperation agreements with foreign enforcers, a process that had begun in 1976 when we entered a cooperation agreement with Germany. From 1991 to 2000, we entered cooperation agreements with the EC, Canada, Israel, Japan, Brazil, and Mexico and our first antitrust enforcement assistance agreement with Australia. During this same period, the Department of Justice negotiated a number of Mutual Legal Assistance Treaties with foreign nations, which are frequently used for evidence gathering in criminal antitrust investigations. Currently, the United States has more than fifty MLATs in force. In 1999, the Division organized the first International Anti-Cartel Enforcement Workshop, which brought together enforcers from twenty-seven jurisdictions to identify and share best practices in the investigation and prosecution of cartels. The workshop was such a success that it became an annual event, with later conferences in the United Kingdom, Canada, Brazil, and Belgium. This year Australia hosts the sixth annual workshop. The dialogue and cooperation we have developed with foreign authorities are essential to the obtaining of evidence from foreign locations, avoidance of conflicts, and the coordination of enforcement activities in multiple jurisdictions to avoid premature disclosure of an investigation and the destruction of evidence.

We have engaged in other multilateral efforts to promote the development of sound antitrust policy and practice, through the Organization for Economic Co-operation and Development, World Trade Organization, United Nations Conference on Trade and Development, and most recently the International Competition Network. The need to combat globalized cartels resulted in the landmark 1998 OECD Recommendation Concerning Effective Action Against Hard Core Cartels which recommended that member countries ensure that their competition laws provide for effective sanctions to deter participation in cartels and adequate enforcement procedures and institutions to detect and remedy cartels.

The Importance of Amnesty Programs

One key area in which governments have developed enforcement procedures for the detection of cartels has been the development and use of corporate amnesty policies. As amnesty policies were developed in multiple jurisdictions, it became clear that convergence in effective policies was needed. Over time, we learned that occasionally members of international cartels did not apply for amnesty in one jurisdiction because they had greater exposure in another jurisdiction that did not have a transparent and predictable amnesty policy. Recent convergence in amnesty policies in multiple jurisdictions, however, has led to many simultaneous amnesty applications, which has enhanced enforcement by providing opportunities for coordinated raids, interviews, and service of subpoenas.

The Antitrust Division's amnesty policy has become the cornerstone of our international anti-cartel enforcement program. It has led to the detection and prosecution of more international cartels than all of our search warrants, consensual monitoring, and FBI interrogations combined. Because cartel activities are hatched and carried out in secret, obtaining the cooperation of insiders is the best, and often the only way to crack a cartel. Obtaining the cooperation of knowledgeable insiders at an early stage of an investigation may shorten an investigation by many months, if not years. This saves scarce government resources, leads to the earlier termination of cartels, allows conviction of defendants that might otherwise never be prosecuted, and assists in securing recovery for the victims of the crime.

Through the years, we have faced the deficiencies in our amnesty program and revised it to make it more effective. Our original 1978 amnesty program lacked transparency and predictability and retained considerable prosecutorial discretion. It was difficult for corporations to predict with much certainty whether they would receive amnesty if they chose to apply. Our policy was designed to give "serious consideration" to refraining from prosecuting a company that confessed its wrongdoing before we began investigating the cartel. But we expressly refused to limit our prosecutorial discretion or to make amnesty automatic. There was no written policy, adding to the lack of transparency. A further deficiency, in addition to the unpredictable nature of the policy, was that amnesty was not available once an investigation had begun. Frequently, however, corporate counsel or a board of directors would not discover a cartel until counsel undertook a focused investigation as a result of the corporation's receipt of a Division document subpoena at the beginning of a grand jury investigation. In many cases, a company's authoritative representatives for legal matters might not even be aware of the need for amnesty until an investigation was under way. As a consequence of the unpredictable nature of the policy and lack of availability of amnesty once an investigation had begun, we had roughly one amnesty application per year from 1978 until the policy was revised in 1993.

Under the revised policy, we increased the transparency of the program and created an alternative amnesty for companies that come forward after an investigation has begun. Under Part A of the policy, which applies before an investigation has begun, amnesty is automatic if the applicant can meet six objective criteria. The requirements under Type A are (1) that the Division has not yet received information from any other source about the conspiracy, (2) that the applicant takes prompt and effective action to terminate its involvement in the conspiracy upon its discovery of the conspiracy, (3) that the applicant reports the wrongdoing with candor and completeness and provides full, continuing and complete cooperation throughout the investigation, (4) that the applicant's confession is truly a corporate act as opposed to isolated confessions of individual executives or officials, (5) that the applicant make restitution to injured parties, and (6) that the applicant did not coerce another party to participate in the cartel and was not the leader or originator of the cartel.

If an applicant cannot meet the above Type A factors, it might be able to qualify for amnesty under the alternative Type B amnesty in the revised 1993 policy. In addition to the Type A termination, corporate confession and restitution obligations, Type B amnesty also requires that the applicant be the first one to qualify for amnesty with respect to the cartel, that the Division not yet have evidence against the applicant that is likely to result in a sustainable conviction, that the applicant reports the cartel with candor and completeness and provides full, continuing and complete cooperation that advances the Division in its investigation, and that granting amnesty would not be unfair to others, considering the nature of the cartel, the applicant's role in it and when the applicant comes forward. In return for creating this additional opportunity for amnesty, the Division did retain some discretion under Type B amnesty, but this was deemed necessary because Type B generally comes into play after an investigation has begun. In addition to issuing a written policy, we have further clarified the application of the policy by publishing several speeches to address questions that have arisen regarding the policy.(1) Since the policy was revised in 1993, the application rate has soared to an average of approximately two per month.

In addition to revising the Corporate Leniency Policy in 1993, we also adopted an Individual Leniency Policy in 1994. While we have not received individual amnesty applications at the same rate as corporate applications, the availability of individual amnesty is an important option for individuals and enhances our corporate policy. If an individual executive urges his company to apply for corporate amnesty, under which he could receive nonprosecution coverage, the company knows that if it refuses to apply for amnesty, the individual may decide to approach the Division on his own to apply for amnesty and thereby escape prosecution. In the case of individuals, this incentive is very strong, for it means not just avoiding a substantial fine, but most importantly for the individual, a jail sentence. With the individual amnesty policy in place, the company likely would decide not to take that risk and apply for amnesty itself.

This summer, as I mentioned earlier, another milestone in the amnesty program was achieved when President Bush signed legislation detrebling the damages paid by a corporate amnesty applicant to private plaintiffs if the applicant cooperates with the plaintiffs in their efforts to recover damages from the other cartel members. We frequently heard that applicants did not apply for amnesty because of the prospect of U.S. treble damage litigation. The removal of this disincentive ­ in fact, the offering of a significant incentive ­ will lead to more amnesty applications in the United States, and more simultaneous applications in multiple jurisdictions.

We at the Antitrust Division are sensitive to the fact that the unique US system of private treble damages litigation should not be allowed to impede global cartel enforcement efforts. Earlier this year, I had the pleasure of appearing before the Supreme Court to argue the United States' position in Empagran. Empagran involves the global price-fixing and market-allocation conspiracies among American and foreign manufacturers and distributors of bulk vitamins. This cartel sold billions of dollars of vitamins in the United States and other countries around the world at fixed prices. The United States' investigation and prosecution of the cartel resulted in the criminal conviction of twelve corporate defendants and thirteen individual defendants and the imposition of criminal fines exceeding $900 million. American private parties sued the vitamin companies seeking treble damages under American antitrust law based on the conspiracies' effects on U.S. commerce; they have obtained a settlement in excess of $2 billion.

The plaintiffs in Empagran are foreign corporations who purchased vitamins abroad for delivery abroad. They brought suit in a U.S. district court under the U.S. antitrust laws. The district court held that it lacked subject matter jurisdiction over their claims under the Foreign Trade Antitrust Improvements Act of 1982. The 1982 Act provides that the Sherman Act shall not apply to non-import foreign conduct unless it has a direct, substantial, and reasonably foreseeable effect on U.S. commerce and that such effect gives rise to a claim under the Sherman Act.(2) The court of appeals, however, reversed and held that the 1982 Act allows foreign plaintiffs injured by anticompetitive conduct to sue whenever the conduct's harmful effect on U.S. commerce gives rise to a claim by anyone, even if not the foreign plaintiff who is actually before the court.(3)

The United States, joined by Germany, Belgium, Canada, Japan, the United Kingdom, Ireland, and the Netherlands, participated as amici and urged the Supreme Court to reverse.(4) Our position was that the 1982 Act's requirement that the effect on U.S. commerce gives rise to a claim means that the effect must give rise to a claim by the particular plaintiff before the court. We also argued that the expansive reading by the court of appeals is "highly likely to have the perverse effect of undermining the [United States'] efforts to detect and deter international cartel activity" and to impair similar efforts by foreign nations.(5) We were very pleased that the Court adopted this rationale for its 8-0 decision in the case. Further litigation is ongoing on remand, and the Division continues to be involved to help assure that the private interests of plaintiffs and their attorneys not undermine international enforcement cooperation.

Many other governments have benefitted from the lessons we learned in developing a successful amnesty program. Principally, uncertainty in the application process will kill an amnesty program. Many other governments have followed the model of the revised U.S. policy and developed effective amnesty programs of their own. Most notably, the EC issued a revised program in February 2002, which improved the transparency and predictability of the policy, created the opportunity for full immunity after an investigation has begun, and has already proven to be a great success. However, many governments still lack amnesty policies or have policies that are stymied by ambiguous, discretionary elements that create uncertainty regarding the application of the program.

Governments that do not have transparent, predictable amnesty policies will not receive many applications, and they will not be able to participate in the information sharing that occurs among jurisdictions with effective amnesty policies. Of course, the Division and other jurisdictions have well known confidentiality policies regarding amnesty applications. The Division does not disclose an amnesty applicant's identity, absent prior disclosure by or agreement with the applicant, unless authorized by a court order. The Division has adopted a policy of not disclosing information obtained from a leniency applicant to foreign authorities pursuant to cooperation agreements, unless the applicant first agrees to the disclosure. This confidentiality policy is a necessary inducement for amnesty applications. Leniency applicants routinely consent, however, to the sharing of information among jurisdictions in which they have obtained conditional amnesty, so that those jurisdictions may conduct coordinated investigations.

Amnesty applicants in one jurisdiction will not, however, apply for amnesty in other jurisdictions that have ineffective, unpredictable leniency policies, and they will not grant waivers allowing the sharing of information with governments that have deficient amnesty policies. Thus, the more jurisdictions that have effective amnesty policies, the greater the incentives that will exist for simultaneous amnesty applications in multiple jurisdictions and waivers from amnesty applicants allowing the sharing of amnesty information. Simultaneous amnesty applications and information sharing will lead, and have led, to greater opportunities for multi-jurisdictional cooperation, including the coordination of initial investigative steps, such as raids, interviews, and service of subpoenas, which can minimize the premature disclosure of an investigation and the potential for destruction of evidence. Better investigations lead to more efficient and effective prosecutions, and hence the termination and punishment of more cartels.

Amnesty programs are most effective not only when the amnesty policies are transparent and predictable in application, but also when the jurisdiction's laws provide the threat of stiff sanctions for cartel members and the jurisdiction's antitrust institutions offer a significant risk of detection of cartel activity.(6) In the United States we believe that cartel activity is deserving of criminal sanctions and that criminal sanctions provide the best deterrent to cartel activity. But, the lack of criminal penalties is not a reason for a jurisdiction to forgo adopting an amnesty policy. Amnesty policies have succeeded in jurisdictions without criminal cartel penalties if those jurisdictions have other stiff cartel sanctions. The EC, for example, does not have criminal sanctions for cartel behavior, yet it has had great success with its amnesty policy, particularly since its revised policy was adopted in 2002.

Criminalization - A Growing Trend

Simply put, we in the United States believe that criminal sanctions provide the greatest deterrent for cartel activity. Corporations engage in cartels only through the conduct of their employees. Therefore, the employee, as well as the corporation, needs to be deterred. Individuals are best deterred by the threat of being held accountable in ways that are not reimbursable by their employer. If an executive is only fined, his employer can find a way to reimburse the employee even if the company does not directly pay the fine. It is virtually impossible to prevent a company from finding a way to reimburse an employee for a cartel fine. Imprisonment, however, imposes a social stigma for white-collar offenders and a loss that cannot be "reimbursed" by corporations. Our defendants routinely offer to pay large fines in lieu of going to jail ­ a plea that we reject ­ but they don't offer to go to jail in lieu of paying a large fine.

The criminalization of cartel behavior is a growing area of study around the world. The OECD, in its Second Report On Effective Action Against Hard Core Cartels issued in January 2003, recommended consideration of the introduction of criminal sanctions for cartel activity in nations where doing so would be consistent with social and legal norms in order to enhance deterrence and incentives to cooperate through leniency programs.(7) A number of nations on at least five continents, including the major economies of Canada, Japan, the United Kingdom, France, and Korea, already have laws providing for criminal sanctions. Some of these jurisdictions provided for even greater maximum jail terms than the United States did until the recent passage of the Antitrust Criminal Penalty Enhancement and Reform Act of 2004. As I noted earlier, this legislation increased the maximum jail term in the United States for cartel behavior to ten years from the prior three year maximum. This increase brings the Sherman Act prison maximum more in line with other white collar offenses, such as bribery, embezzlement, securities fraud, mail and wire fraud, and money laundering.

Having criminal anti-cartel laws on the books is not enough, however. Prosecutors must be willing to seek stiff sentences, and courts must be willing to impose stiff sentences for antitrust offenses in order for such laws to have any deterrent force. In the United States, we routinely seek jail sentences for individuals, and courts routinely impose jail terms for cartel offenses.

Not only is criminalization important for deterrence and punishment, but criminalization also enhances investigative powers. Strong criminal investigative powers, such as search and seizure and subpoena power, help prosecutors obtain evidence of cartel behavior and create a significant risk of detection that can incentivize amnesty applications. Criminal sanctions for noncompliance with compulsory process, such as obstruction of justice or perjury, are also necessary for the proper functioning of criminal investigative powers. Investigatory assistance through Mutual Legal Assistance Treaties, such as the compulsion of testimony and document searches by foreign authorities, is also available in the investigation of criminal offenses. Criminalization of cartel violations in multiple jurisdictions also increases the possibility of extradition of international cartelists.

Criminalization, because it involves the deprivation of liberty, justifiably requires necessary legal protections for the accused. Criminalization warrants a higher standard of proof, such as guilt beyond a reasonable doubt in the United States. It also triggers a higher level of rights for defendants, such as a right to counsel, right to trial, right of confrontation of witnesses, and discovery rights. For jurisdictions that can provide the necessary legal protections for the accused, criminal sanctions provide the greatest potential for deterrence and the most appropriate punishment for individual cartelists.


History shows that jurisdictions must continue to engage in a review of the efficacy of their anti-cartel laws and policies and must constantly adapt to changing events to maintain the effectiveness of their anti-cartel enforcement. As the fight against cartels continues, experience will prove that amnesty policies can play a critical role in the detection and punishment of more cartels and that criminal sanctions provide the greatest and most appropriate deterrent and punishment for cartel behavior.


1. See, e.g., Gary R. Spratling, Making Companies An Offer They Shouldn't Refuse, Speech Before 35th Annual Symposium on Associations and Antitrust of Bar Association of the District of Columbia (February 16, 1999); Gary R. Spratling, The Corporate Leniency Policy: Answers To Recurring Questions, Speech Before ABA Antitrust Section 1998 Spring Meeting (April 1, 1998); see also Scott D. Hammond, When Calculating The Cost And Benefits Of Applying For Corporate Amnesty, How Do You Put A Price Tag On An Individual's Freedom?, Speech Before ABA Criminal Justice Section 15th Annual National Institute on White Collar Crime (March 8, 2001).

2. 15 U.S.C. § 6a.

3. Empagran S.A. v. F. Hoffman-LaRoche, Ltd., 315 F.3d 338 (D.C. Cir. 2003).

4. Brief of the United States as Amicus Curiae Supporting Petitioners in F. Hoffman-La Roche Ltd. v. Empagran (Sup. Ct. No. 03-724) (filed Feb. 3, 2004), available at http://www.

usdoj.gov/osg/briefs/2003/3mer/1ami/2003-0724.mer.ami.pdf; Brief of the Governments of the Federal Republic of Germany and Belgium as Amici Curiae in Support of Petitioners in F. Hoffman-La Roche Ltd. v. Empagran (Sup. Ct. No. 03-724) (filed Feb. 3, 2004); Brief for the Government of Canada as Amicus Curiae Supporting Reversal in F. Hoffman-La Roche Ltd. v. Empagran (Sup. Ct. No. 03-724) (filed Feb. 3, 2004); Brief of Japan as Amicus Curiae in Support of Petitioners in F. Hoffman-La Roche Ltd. v. Empagran (Sup. Ct. No. 03-724) (filed Feb. 3, 2004); Brief of the United Kingdom of Great Britain and Northern Ireland, and the Kingdom of the Netherlands as Amici Curiae in Support of Petitioners in F. Hoffman-La Roche Ltd. v. Empagran (Sup. Ct. No. 03-724) (filed Feb. 3, 2004).

5. Brief of the United States, supra n.36, at 8.

6. See Scott D. Hammond, Detecting And Deterring Cartel Activity Through An Effective Leniency Program, Remarks Before the International Workshop on Cartels (Nov. 21-22, 2000).

7. Oecd Competition Committee, Second Report By The Competition Committee On Effective Action Against Hard Core Cartels (2003), available at www.oecd.org/competition.

Updated June 25, 2015