DEPARTMENT OF JUSTICE

Over the last five years, the United States' Corporate Leniency Program ("Amnesty Program") has been responsible for detecting and cracking more international cartels than all of our search warrants, secret audio or videotapes, and FBI interrogations combined. It is, unquestionably, the single greatest investigative tool available to anti-cartel enforcers. An effective Leniency Program will lead cartel members, in some cases, to confess their conduct even before an investigation is opened. In other cases, it will induce organizations already under investigation to abandon the cartel stonewall, race to the government, and provide evidence against the other cartel members. While the availability of some investigative techniques, such as consensual monitoring or the compulsion of sworn testimony, may be limited or nonexistent in jurisdictions where hardcore cartel activity is not a criminal offense, Leniency Programs can potentially be utilized in any jurisdiction where such conduct is treated as a criminal, civil or administrative offense.

This paper discusses the reasons behind the success of the U.S. Amnesty Program and identifies three hallmarks for both a successful Amnesty Program and an enforcement program that deters cartel activity. The issues of detection and deterrence are closely aligned because the bedrock principles that apply to effectively preventing cartels are also at the core of implementing a successful leniency program for detecting cartel activity once it does occur. The three critical ingredients are as follows. First, one's antitrust laws must provide the threat of stiff sanctions for those who participate in hardcore cartel activity. Second, organizations must perceive a significant risk of detection by antitrust authorities if they engage in cartel activity. Third, antitrust authorities must provide transparency, to the greatest extent possible, throughout the anti-cartel enforcement program so that prospective cooperating parties can predict with a high degree of certainty their treatment following cooperation. These three major cornerstones -- stiff potential penalties, heightened fear of detection, and transparency in enforcement policies -- are at the heart of both our Amnesty Program and our deterrence efforts.

Before addressing these three principles, it's important to share some background information on how the Amnesty Program was developed and how it works in order to put its impact fully into perspective. The original version of our Amnesty Program actually dates back to 1978. Under that program, violators who came forward and reported their illegal activity before an investigation was underway were eligible to receive a complete pass from criminal prosecution. The grant of amnesty, however, was not automatic and the Division retained a great deal of prosecutorial discretion in the decision making process. Unfortunately, for reasons discussed below, it became clear over time that this program was flawed. It resulted in relatively few amnesty applications and did not lead to the detection of a single international cartel.

In 1993, the Antitrust Division dramatically expanded its Amnesty Program to increase the opportunities and raise the incentives for companies to report criminal activity and cooperate with the Division. The Amnesty Program was revised in three major respects. First, the policy was changed to ensure that amnesty is automatic if there is no pre-existing investigation. That is, if a corporation comes forward prior to an investigation and meets the program's requirements, the grant of amnesty is certain and is not subject to the exercise of prosecutorial discretion. Second, the Division created an alternative amnesty, whereby amnesty is available even if cooperation begins after an investigation is underway. Third, if a corporation qualifies for automatic amnesty, then all directors, officers, and employees who come forward with the corporation and agree to cooperate also receive automatic amnesty. (See attached Corporate Leniency Policy). One aspect of the program that did not change is that amnesty applies only to criminal sanctions, and firms accepted into the Amnesty program are required, where possible, to make full restitution to their victims in the United States.⁽¹⁾

Because of the novelty and uniqueness of this policy, it took some time before the private bar and business community gained confidence in the program. However, over the last five years, the results have been staggering. There has been more than a ten-fold increase in amnesty applications during this time frame. In the last two years alone, cooperation from amnesty applications have resulted in scores of convictions and well over $1 billion in fines. More than anything else, the expansion of the Amnesty Program has been responsible for the success that we have had in cracking international cartels.

Treating Cartel Activity As A Crime. The first element common to both deterring cartel activity and creating a successful Leniency Program is the threat of severe sanctions for violators. In the United States, hardcore cartel activity -- such as price-fixing, bid-rigging, and customer and market allocation agreements -- is a felony violation of our criminal laws, and both corporations and individuals may be held liable. Corporations risk heavy criminal fines with the maximum potential fine being the greatest of $10 million, twice the gross gain to the cartel, or twice the gross loss suffered by the victims of the conspiracy. Relying on the twice the gross gain/loss formula, the Division has obtained criminal fines of up to $500 million. The maximum sentence for individuals is three years imprisonment and a fine which is the greatest of $350,000, twice the gross gain to the cartel, or twice the gross loss suffered by the victims of the conspiracy.

Of course, the United States is in the minority, albeit a substantial and growing one, in treating hardcore cartel activity as a crime and prosecuting offending corporations and individuals as criminals. I will not take this opportunity to explore all of the arguments in favor of prosecuting cartel activity criminally except to make the following point relevant to the discussion on deterrence. Based on our experience, there is no greater deterrent to the commission of cartel activity than the risk of imprisonment for corporate officials. Corporate fines alone are simply not sufficient to deter some would-be offenders. For example, in some cartels, such as the graphite electrode cartel, individuals personally pocketed millions of dollars as a direct result of their criminal activity. In our view, a corporate fine alone, no matter how punitive, is unlikely to deter such individuals.

The Perceived Risks Must Outweigh The Potential Rewards. While there is no current consensus as to whether cartels should be prosecuted criminally, there should be no dispute that cartel activity will not be deterred if the potential penalties are perceived by firms and their executives as outweighed by the potential rewards. If the potential sentences are not sufficiently punitive, then the potential sanctions will merely be seen as a cost of doing business. To ensure that this is not the case, we have recently "upped the ante" by obtaining record-breaking fines against firms who engage in cartel activity. At the time the Amnesty Program was revised in 1993, the highest antitrust fine ever obtained in the United States was less than $3 million. Today, fines of $10 million or more have become almost commonplace with more than 30 imposed in the last five years. In fact, the Division has obtained fines of $100 million or more against five corporate defendants, including a $500 million fine against F. Hoffmann-La Roche (HLR) for its participation in the worldwide vitamin cartel. This dramatic leap in the level of criminal fines, however, is more than just a reflection of our aggressive approach for deterring cartel activity. Rather, because fines in the United States are based in large part on a company's sales in the United States affected by the conspiracy, the record fine levels demonstrate the mammoth size of the international cartels that we have been uncovering, largely through the amnesty program; cartels that simply dwarf the domestic conspiracies that we have previously encountered.

The HLR case also offers interesting, if not sobering, insight as to the risks companies are willing to take in order to profit from cartel activity and to how inadequate fines may be to ensure specific, let alone, general deterrence. In 1997, roughly two years before HLR's participation in the vitamin conspiracy was exposed, HLR was convicted of participating in a separate international cartel in the citric acid industry. The firm entered into a plea agreement, agreed to cooperate, and was ordered to pay a $14 million fine -- which at the time was the third largest antitrust fine ever obtained in the United States. As part of its pledge to cooperate, HLR was told of the Division's then-covert investigation into the vitamin industry and offered the opportunity to cooperate. The Division interviewed two top executives from HLR who participated in the citric acid conspiracy and who also had dual responsibilities in HLR's vitamin business. HLR and its top executives denied to our investigators any knowledge of, or participation in, a vitamin cartel. Of course, we would later learn that HLR executives had, in fact, engaged in a decade-long, worldwide vitamin cartel, and continued to lead that cartel even after it pled guilty in the citric acid investigation and learned that its vitamin business was under investigation. Instead of being deterred, top-level HLR executives orchestrated false statements to enforcement authorities, took steps to further conceal the firm's illegal activities, and continued to lead the world's other producers in a global cartel. This decision will end up costing HLR well over a billion dollars in criminal fines and civil settlements. In addition, three European executives from HLR, including the two executives who blew their opportunity to come clean in the citric acid investigation, served time in U.S. prisons for their participation in the vitamin conspiracy. Clearly, the $14 million fine in the citric acid prosecution was not nearly enough to deter HLR or its top executives from continuing to participate in the vitamin cartel. Time will tell if the $500 million fine and jail sentences for the HLR executives will be a better deterrent.

Corporate Amnesty Can Mean Zero Fines and No Jail. Of course, the Amnesty Program offers companies and their executives an alternative to these potentially harsh sentences. If a company detects an antitrust violation, it has to chose whether to report the violation and seek leniency or to remain silent and hope for the best.

The question is often raised as to whether an Amnesty Program will work in a jurisdiction where there is no individual liability and, therefore, no possibility of incarceration for culpable executives. Clearly, the opportunity to avoid imprisonment for corporate officials is a major inducement for firms to seek amnesty. However, an Amnesty Program can still succeed if the threat of heavy fines is significant enough. This belief is supported by our experience with foreign-based firms who have sought and obtained amnesty in international cases at a rate almost equal to their domestic counterparts. For example, the worldwide vitamin cartel was cracked by the cooperation provided by French-based Rhône-Poulenc SA.⁽²⁾ The company made the decision to come forward even though the culpable French executives resided outside of the United States and our extradition treaty with France does not cover antitrust offenses. So, the opportunity to avoid incarceration for its culpable executives was probably not the major inducement to Rhône-Poulenc's decision to come forward, but rather the desire to avoid a criminal conviction and heavy fine for the corporation. And, indeed, while Rhône-Poulenc paid zero dollars in fines, its principal co-conspirators, HLR and BASF, paid fines of $500 million and $225 million, respectively.

Building A Strong Enforcement Record. Of course, if firms perceive the risk of being caught by antitrust authorities as very small, then stiff maximum penalties will not be sufficient to deter cartel activity. Likewise, if cartel members do not fear detection, they will not be inclined to report their wrongdoing to authorities in exchange for leniency. Therefore, antitrust authorities must cultivate a law enforcement environment in which business executives perceive a significant risk of detection by antitrust authorities if they either enter into, or continue to engage in, cartel activity.

Over the last five years, we have built a strong enforcement record that demonstrates our commitment to fighting cartel activity. Moreover, the fines obtained in these cases have increased dramatically. Still, the question remains whether the $100 million-plus fines are having a deterrent effect by either stifling the formation of cartels or disbanding them if they already exist? There is no way to quantify the number of cartels that are deterred before they are ever formed, but we can observe how our enforcement efforts are influencing cartels that already exist. To put it plainly, cartel members are starting to sweat, and the Amnesty Program feeds off that panic.

The Race To The Courthouse. The more anxious a company is that its cartel participation may be discovered by the government, the more likely it is to report its wrongdoing in exchange for amnesty. Of course, amnesty is only available to the first one in the door. If you are second, even if only by a matter of a few hours, which has happened on a number of occasions, the second firm and all of its culpable executives will be subject to full prosecution.

This "winner-take-all" approach sets up a race, and this dynamic leads to tension and mistrust among the cartel members. For example, consider a scenario where five members of a cartel are scheduled to hold an emergency meeting. When the meeting starts there is an empty seat at the table -- one of the conspirators has not returned calls and has unexpectedly not arrived at the meeting. One of the cartel members at the meeting starts to get nervous. Has the missing cartel member had a change of heart and abandoned the cartel? Has he gone to the Feds? Or, did he just miss his plane? In this environment, with the risk of detection and the stakes so high, who can you trust? Or consider the very common situation when a cartel first learns that it is under investigation. Each member of that cartel knows that any of its co-conspirators can be the first to come forward in exchange for total amnesty. A decision that will seal the fate of the rest. Imagine the vulnerability of being in that position and asking yourself, "Can I really trust my competitors?"

The Safe Harbors Are Shrinking. Of course, antitrust authorities in Canada, Europe and around the world share the credit for this heightened fear of detection. The Brighton conference will highlight many of the changes in antitrust laws, the enforcement victories, and the commitment to anti-cartel enforcement that is taking hold throughout the world, so I will not try to summarize them here. Suffice to say, the world is changing, antitrust authorities are working more closely together than ever, and the safe harbors for cartel activity are shrinking.

However, if you want any more proof of how significantly attitudes have changed toward anti-cartel enforcement or how seriously the risks are now perceived by the international business community, consider the case of the six high-level European executives from HLR and BASF who led their companies' decade-long participation in the worldwide vitamin cartel. Each of these foreign executives voluntarily traveled to the United States to plead guilty and to serve time in a U.S. prison. They were all citizens of either Switzerland or Germany, residing abroad, with no family or other ties to the United States. Therefore, the United States courts had no personal jurisdiction over any of these individuals so long as they did not set foot on U.S. soil. Moreover, the United States does not have an extradition treaty with either Switzerland or Germany covering antitrust offenses. And so, you may wonder why did these six international executives agree to leave their families behind, submit to the jurisdiction of a U.S. court, agree to plead guilty, cooperate with our investigation, and serve time in a U.S. jail? I am sure there are many diverse and intensely personal reasons why these individuals arrived at the decisions they made. However, I'm convinced that all of these executives were driven, in large part, by the recognition that Europe and the rest of the world are changing their attitudes toward cartel activity. They did not want to live their lives as international fugitives in this changing world. The risk is too great. To me, this is the best indication that cartel members are getting the message.

Transparency In Enforcement Policies Maximizes Cooperation. The third and final hallmark of both an effective Amnesty Program and an anti-cartel enforcement program is the need for transparency in enforcement policies. Self reporting and cooperation from offenders have been essential to our ability to detect and prosecute cartel activity. Cooperation from violators, in turn, has been dependent upon our readiness to provide transparency, to the greatest extent possible, throughout our anti-cartel enforcement program. If prospective cooperating parties cannot predict, with a high degree of certainty, their treatment following cooperation, then they are less likely to come forward.

Our experience has been that transparency must include not only explicitly stated standards and policies, but also clear explanations of prosecutorial discretion in applying those standards and policies. The Division has sought to provide transparency in the following enforcement areas: (1) transparent standards for opening investigations; (2) transparent standards for deciding whether to file criminal charges; (3) transparent prosecutorial priorities; (4) transparent policies on the negotiation of plea agreements; (5) transparent policies on sentencing and calculating fines; and (6) transparent application of our Amnesty Program.⁽³⁾ However, this paper will focus on how transparency is critical to an effective Amnesty Program.

Transparency In The U.S. Amnesty Program. The Division has a written Amnesty Policy and has published a number of papers in order to clarify the Division's application of its Corporate Amnesty Program.⁽⁴⁾ In addition, representatives from the Division regularly speak about the Amnesty Program before national and international bar associations, trade groups, other law enforcement agencies, and the media. However, in order for an Amnesty Program to work, an antitrust authority must do more than just publicize its policies and educate the public. It has to be willing to make the ultimate sacrifice for transparency - the abdication of prosecutorial discretion.

Our Amnesty Program by its nature is transparent because we have eliminated, to a great extent, the exercise of prosecutorial discretion in its application. Obviously, this is a very difficult thing to do, and we have had to swallow hard on a number of amnesty applicants that we would have preferred to prosecute. However, remember we had roughly 15 years of experience with an Amnesty Program that was designed to maintain a greater degree of prosecutorial discretion, and it simply did not work. Prospective amnesty applicants come forward in direct proportion to the predictability and certainty of whether they will be accepted into the program. Uncertainty in the qualification process will kill an amnesty program.

Below are three examples of instances where our program, as well as other programs, have sacrificed prosecutorial discretion in order to maximize predictability and transparency in the qualification process. In each instance, this approach is contrasted with the alternative approaches exemplified by our prior 1978 program and, as I understand it, the present EU program.

Automatic Amnesty If No Pre-Existing Investigation. If a company comes forward and reports its wrongdoing before an investigation has begun and meets the program's other conditions, the grant of amnesty is automatic and the organization pays zero dollars in fines. There is no prosecutorial discretion involved. As we understand it, the Canadian, United Kingdom (UK), and German Leniency Policies have also adopted this approach. The policy creates the maximum incentives for companies to defect from the cartel and report the conduct. It's clear that without the certainty of obtaining the maximum reward, a number of massive global cartels that have recently been identified and prosecuted would still be acting with impunity today.

The European Union (EU) has taken a different approach with its Leniency Policy. Its policy provides a sliding scale whereby a company that reports wrongdoing before an investigation has begun, and meets all of the other requirements of the program, is eligible for a reduction in fines ranging from 75 percent to 100 percent. (The potential reduction drops down to between 50 percent and 75 percent if it meets all of the program's conditions, but informs the Commission after an investigation has started.) However, since an organization's potential fine exposure in the EU is 10 percent of the worldwide turnover for all products sold by the company, the difference between total and partial amnesty may be very significant. For example, a multi-national corporation with worldwide annual revenue of $1 billion still faces a potential maximum EU fine of $250 million, even if it brings the illegal conduct to the attention of the EU. Of course, this may well overstate the amount of the fine that would be imposed in such a case. Still, if a company cannot accurately predict how it will be treated as a result of its corporate confession, our experience suggests that it is far less likely to report its wrongdoing, especially where there is no ongoing government investigation.

Measuring The Value Of Cooperation. Under the U.S., Canadian, and U.K. programs, applicants must report their involvement with candor and completeness and must provide full, continuing, and complete cooperation that advances the investigation. In some cases, an amnesty applicant's cooperation will provide sufficient evidence to convict all of the remaining cartel members. However, in other cases, the applicant's full cooperation will not, in and of itself, amount to "decisive" evidence of the existence of the cartel, but it may lead to such evidence. Therefore, the U.S., Canadian, and U.K. programs do not have a "decisive evidence" requirement in their leniency programs. In fact, some of the best results under our amnesty program were achieved with the help of amnesty applicants who would not have qualified under such a subjective evidentiary standard. For example, in our graphite electrodes investigation, the amnesty applicant played only a peripheral role in the conspiracy and thus was not able to provide us with "decisive evidence." In this respect, the company's mitigating role made it more attractive as an amnesty applicant than the other members of the cartel that were far more culpable. While the amnesty applicant did not attend the key conspiratorial meetings, it provided the Division with information that allowed us to obtain warrants to search the offices of several of the cartel members. The execution of the warrants together with the other cartel members' knowledge of an insider's cooperation quickly led to the cooperation and guilty pleas of the remaining six conspirators and nearly $300 million in fines. So, while an amnesty applicant's cooperation alone may not always add up to decisive evidence, it can provide us with leads and opportunities that will take us to additional evidence and, ultimately, result in successful prosecutions.

In contrast, the EU Program requires that the leniency applicant come forward "with decisive evidence proving the existence of the cartel." Our experience suggests that, this high standard will automatically disqualify potentially valuable leniency applicants. But there is an additional problem with the use of a "decisive evidence" standard. Because the standard is necessarily subjective, potential amnesty applicants may not be able to predict with significant certainty whether an antitrust authority will consider its proffer of evidence to meet the standard. To the extent potential applicants view this standard as subjective and cannot predict the outcome of their application, the requirement runs the substantial risk of dissuading potential applicants from coming forward.

Consider the hypothetical case of a company that is deciding whether to report conduct to the EU before there is an open investigation. Assume the company has no "smoking gun" documents, but can provide the cooperation of several employees who directly participated in the cartel, and can detail the operation of the conspiracy, including the time and place of key meetings, the participants, and any agreements that were reached. In assessing whether to report the conduct, this hypothetical company has first to decide whether it can meet the decisive evidence standard. Aware of the EU's historical emphasis on documentary evidence, it may be unsure as to whether it will fall short of this standard. However, whether this standard is met can have huge consequences for the company. Assuming it meets the decisive evidence standard (as well as the program's other requirements), the company can qualify for up to a 100% reduction in its fine. However, if it fails to satisfy this standard, the company, which has brought the cartel to the EU's attention in the first place, is only guaranteed a 10 percent reduction in its fine. If a company is uncertain as to how it will be treated as a result of its cooperation, we believe that it will often make the decision not to cooperate at all.

Role In The Offense Requirement. A final area where leniency programs may differ in terms of the potential exercise of prosecutorial discretion relates to the amnesty applicant's role in the offense. In order to qualify under the U.S., Canadian, and U.K. programs, the applicant must not have coerced another party to take part in the offense and must not be "the" instigator or "the" leader of the illegal activity. In each of these programs, applicants will only be disqualified from obtaining total amnesty if they are clearly the single organizer or single ringleader of a conspiracy. If, for example, there are two ringleaders in a five-firm conspiracy, then all of the firms, including the two leaders, are potentially eligible for amnesty. Or, if in a two-firm conspiracy, each firm played a decisive role in the operation of the cartel, both firms may qualify for amnesty. Wherever possible, we have tried to tilt our program in favor of accepting the amnesty applicant into our program in order to provide the maximum incentives and opportunities for companies to come forward and report illegal activity. As a result, the antitrust bar has confidence that they can accurately advise clients as to whether they are likely to qualify for leniency under our program.

Our approach to the "role in the offense" requirement is distinguished from the EU program which requires that the applicant not have acted as "an" instigator or "a" ringleader, and the German program which requires that the applicant not have played a "decisive role" in the cartel. The language of these programs suggests that more than one company may be ineligible for full amnesty in a given cartel. Understandably, there may be serious cultural and legal issues which would lead a competition authority to adopt a leniency program that excludes any party that played "a" leading role in the offense, regardless of whether they were the sole leader or merely a co-leader. However, like the decisive evidence standard requirement discussed above, there are costs to retaining one's discretion in this area. If companies cannot confidently predict how an enforcement authority will apply this standard, then they may ultimately decide against self reporting and cooperation, and existing cartels will go unreported and unpunished.

There is global interest in the use and development of Corporate Leniency Programs as a means of detecting and cracking international cartel activity. The message has been clear that Leniency Programs can provide antitrust enforcers with an unprecedented tool for detecting and investigating cartel activity. The formula for a successful anti-cartel enforcement program should include equal parts of stiff potential penalties, high detection rates, and transparent enforcement policies. This recipe will serve to prevent most companies from ever engaging in cartel activity. However, when cartels are formed, we have found that the Amnesty Program, with its lure of leniency in exchange for self reporting and full cooperation, is the most effective investigative tool for cracking cartel activity.

FOOTNOTES

1. Under the U.S. program, corporate amnesty and corporate leniency are used interchangeably to mean a complete pass from criminal prosecution and zero dollars in fines for the anticompetitive conduct. This definition differs from other jurisdictions, such as the EU, that use "leniency" to refer to fine reductions from anywhere between 100 percent down to 10 percent.

2. The Division has a policy of treating the identity of amnesty applicants as a confidential matter, much like the treatment afforded to confidential informants. However, confidentiality is not required in this case, with respect to Rhône-Poulenc's amnesty status, because the company has already issued a press release announcing its acceptance into the Amnesty Program.

3. For a more detailed discussion on transparency in enforcement with references to Antitrust Division public statements in each of the areas outlined above, see, "Transparency In Enforcement Maximizes Cooperation From Antitrust Offenders," by Gary R. Spratling, Deputy Assistant Attorney General, Antitrust Division, before Fordham Corporate Law Institute (October 15, 1999).

4. See e.g., "The Corporate Leniency Policy: Answers To Recurring Questions," speech by Gary R. Spratling, before ABA Antitrust Section 1998 Spring Meeting (April 1, 1998); "Making Companies An Offer They Shouldn't Refuse," speech by Gary R. Spratling, before Bar Association of the District of Columbia's 35^th Annual Symposium on Associations and Antitrust (February 16, 1999).