Criminal prosecutions of large organizations exhibit a unique power dynamic. The target organizations include goliaths—some of the largest corporations in the United States, including AIG, America Online, Bristol-Myers Squibb Co., Computer Associates, HealthSouth, KPMG, MCI, Merrill Lynch & Co., and Monsanto. A U.S. Attorney’s office with its limited resources may look like a tiny David by comparison. But prosecutors have their slingshot: they wield the threat of an indictment, which results in potentially catastrophic collateral and reputational consequences to a corporation. Yet it is a threat that prosecutors can ill afford to carry out due to those consequences. The détente resulting from the collision of those oversized forces has taken a surprising turn, perhaps because there was nowhere else to turn—from criminal prosecution towards structural reform. By that I mean that prosecutors adopted a strategy to avoid an indictment and a conviction by entering into detailed compliance agreements with organizations. In one example of a demanding structural reform agreement, KPMG International, charged with marketing illegal private tax shelters, agreed to shut down its private tax practice, to cooperate fully in criminal investigations of former employees, and to hire an independent monitor for three years to implement an elaborate compliance program.
Joseph Bernardino, former CEO of Arthur Andersen LLP, testifies before the House Financial Services Committee in February 2002.
In my piece, “Structural Reform Prosecution,”1 I present a picture of why and how federal prosecutors now enter into such agreements supervising the rehabilitation of these goliath organizations. The Article examines the agreements’ origins, goals, terms, and the broader legal and institutional setting, including through empirical analysis of the agreements entered after the Department of Justice (“DOJ”) announced its new approach in January 2003.2 While hue and cry over organizational prosecutions have focused on privilege waiver and employer payment of attorney fees, those two issues just scratch the surface of the complex problems that these massive efforts raise. I hope here to draw attention first to a series of problems raised by how these agreements define compliance and second to the multi-polar context in which these agreements are entered. “Structural Reform Prosecution” concludes by posing questions for future work. I expand on that discussion here by proposing reforms that, from different perspectives, address some of the difficult issues that these agreements raise.
1. DEFINING ORGANIZATIONAL COMPLIANCE
Agreements between large organizations and prosecutors are very different from the typical deferral agreement in which a person agrees to refrain from any additional offenses for a period of time and perhaps to enter into a rehabilitative program. Instead, I have argued that these agreements should be viewed as structural reform efforts, designed to prevent future criminality among the members of a complex entity.
The first respect in which these agreements resemble classic civil structural reform efforts is that they seek to reform entire institutions. Many of these agreements carefully describe institution-wide goals. For example, the KPMG agreement details the changes to be made to KPMG’s tax program and specifies an elaborate compliance program.3 The terms of compliance in such agreements are fairly clear, as is a core understanding of what sort of criminality the agreement is intended to prevent from recurring. The agreements also rely on many of the same tools that commonly appear in civil structural reform cases: independent monitors tasked with overseeing reforms, detailed injunctive provisions, and reporting and auditing requirements to assess progress towards compliance.
As in civil efforts, these agreements also envision a rehabilitative process lasting for some time, and therefore, also as in civil cases, one can imagine their compliance goals shifting over time. There will be some uncertainty in any such ongoing reform effort, given broad terms due to a need for flexibility should the compliance process itself uncover new problems. Unlike civil structural reform efforts (and unlike earlier prosecutorial settlements of labor racketeering cases in the 1980s), these efforts are largely non-public. This poses a difficulty for outsiders in assessing how compliance was defined during the implementation of any particular case. In most cases all we have is the text of the agreement, with no information regarding the subsequent implementation. Perhaps over time more information will emerge regarding the successes and failures in achieving the sought-after compliance.
Cooperation requirements included in many agreements appear to last in perpetuity. How long does the KPMG agreement last? Like many of the others, the KPMG agreement extended for three years and has now terminated (with two additional years of IRS supervision, during which time the DOJ may still terminate). Yet KPMG also agreed to a series of “permanent restrictions and elevated standards for its tax practice.”4 Nor may KPMG ever make statements contradicting the representations in the Agreement, including in civil litigation. Nor may KPMG cease “its continuing cooperation” with the DOJ’s ongoing criminal investigations, “even after the dismissal of the Information,” and with any other prosecution or agency action “relating to or arising out of the conduct.”5 Other agreements have still broader language regarding an obligation to cooperate. For example, Roger Williams Medical Center agreed to “cooperate fully and actively with the [United States Attorney’s Office for the District of Rhode Island (“USAO-RI”)], and with any other government agency designated by the USAO-RI . . . regarding any matter being investigated by the government . . . about which RWMC has knowledge or information.”6 These terms suggest that compliance in the form of cooperation with government investigations may last into the indefinite future regarding a broad array of matters of indefinite scope.
In addition to the important similarities, structural reform takes on a different form in criminal law than in civil law, particularly when implemented at the charging stage, due to the unique power dynamic of an organizational prosecution. These agreements, though often subject to court-approval, are supervised not by a court, but by prosecutors. Prosecutors lack palatable alternatives to structural reform. They cannot easily decline to prosecute the most serious organizational crime cases, but nor can they abide the collateral consequences of seeking convictions in all such cases. Upon entering into settlement negotiations, however, prosecutors will typically have enormous negotiating leverage by threatening the “nuclear” option of an indictment, and it shows in the terms of the agreements. For example, in most of these agreements, the DOJ retains its enormous stick throughout the term of the agreement by retaining the unilateral authority to find a breach and then prosecute. Though few agreements say anything about what the DOJ can count as a breach, courts may not be able to remedy effectively an arbitrary declaration of breach unless they provide pre-indictment relief.7
In some cases, that power dynamic may undermine the goal of achieving compliance. Illustrative of the often serious consequences of an organizational prosecution—even where the organization settles—is a recent nonprosecution agreement (entered in March 2007, after my Article’s study period of January 2003 through January 2007) with the Dallas law firm Jenkins & Gilchrist regarding the same tax shelters at issue in the KPMG prosecutions.8 The agreement was far simpler than the others. It required cooperation in the ongoing investigation, privilege waiver, and noted a fine ($76 million) paid to the IRS. It did not include compliance. Instead, the agreement noted that the firm planned to close its doors.9 Though part of the reason the firm voted to dissolve was due to settlements in civil suits, it was also significant that most of the firm’s attorneys left during the ongoing criminal investigation of the firm’s tax group.10 This example shows how even an investigation ending in a settlement can result in a catastrophic result for the entity, perhaps justifiably to deter future wrongdoing, but also not so different from the result had the firm been indicted and been convicted. An IRS Commissioner commented, “This should be a lesson to all tax professionals that they must not aid or abet tax evasion.”11 That sort of punitive goal was at odds with the rehabilitative purpose of these agreements; if the goal was to teach a lesson, then why not indict? The reason for settling seemed chiefly to secure cooperation in individual prosecutions. The agreement thus highlighted not only the power prosecutors may wield, but also how the structural reform goals that ostensibly animate these agreements can fall by the wayside, resulting not in a structural reform agreement but rather a cooperation agreement.12
2. MULTI-POLAR PROSECUTIONS
A multi-polar dynamic created these prosecution agreements, as in classic public law adjudication. State prosecutors, the Sentencing Commission, regulatory agencies, compliance experts, industry, and federal prosecutors each developed parallel approaches toward organizational compliance. Which structural reformer came first is hard to say. If the DOJ decided to stop entering into these agreements by instructing all of the U.S. Attorneys’ Offices not to defer prosecution in this way, similar agreements would still be entered in large numbers and in many of the same cases by regulators.
First, several regulators may be involved in negotiating any one structural reform agreement. Federal prosecutors often take only egregious cases referred by regulatory agencies, which retain an important role in subsequent negotiations and implementation. This follows where a raft of regulatory agencies have sought to accomplish similar goals through voluntary disclosure regimes or consent decrees for decades. For example, the SEC’s Seaboard Report looks very much like the DOJ’s Thompson and McNulty Memos.13 Nothing could be more standard practice than granting organizations cooperation credit and encouraging self-investigation and self-reporting. Little distinguishes these current deferral and nonprosecution agreements except that they occur in criminal cases where the stakes may be particularly high. These agreements were often investigated in conjunction with agencies, negotiated in conjunction with agency consent decrees, and then supervised by independent monitors jointly appointed by the DOJ and agencies and reporting to both. The nature of this interaction between agency compliance regimes and the DOJ’s emerging regime is an important area for future study.
Second, federal efforts should be considered alongside efforts by state Attorneys General to pursue similar goals using similar methods. Although traditionally states—like the federal government had been—were reluctant to enforce criminal laws against corporations, Elliot Spitzer transformed the practice by leading efforts to pursue structural reform, even crafting industry-wide agreements. Others have followed suit, such as California Attorney General Bill Lockyer and Massachusetts Secretary of State William Galvin.14
Third, the Sentencing Guidelines emphasize compliance goals at the penalty phase. A conviction can result in a similar remedy as that in a settlement, where organizations may be required by courts to create compliance programs as a condition of probation. Further work could investigate how judges now apply those Guidelines.
Fourth, the role of outsiders—in particular, the compliance industry —will be a rich subject for future study. Former prosecutors or regulators have often been appointed as independent monitors, and were frequently active in the burgeoning compliance industry, serving the needs of organizations under investigation or seeking to head off potential scrutiny. Legal scholars should continue to explore the emerging influence and role of compliance experts.
Finally, the compliance industry has eager clients. Organizations have themselves focused—in response to regulators, of course—on compliance-oriented approaches. Many of the entities prosecuted had already made significant structural changes once these entities discovered the malfeasance. Bristol-Myers, for example, had already retained Judge Frederick B. Lacey, later appointed as an independent monitor, to conduct a review of its internal controls. Pursuant to this review, it had made “significant personnel changes” at the highest levels and had adopted a whole series of other auditing and compliance measures.15 In turn, organizations may influence the nature of compliance demands. The political dynamics of such prosecutions are very different in a RICO prosecution than in a prosecution of a firm under the Foreign Corrupt Practices Act. Some cases have industry-wide effects and entire industries may lobby the executive branch or Congress; the DOJ has already changed its approach regarding privilege waiver and employer payment of attorney’s fees in response to pressure.
The multi-polar nature of these negotiated dispositions has not been sufficiently recognized. Whether the convergence in compliance approaches represents path dependency or emerging best practices in part depends on their efficacy. Future work will hopefully take on the very difficult task of assessing whether these approaches successfully prevent, detect, or remedy organizational crime.A Series of Reform Proposals
Taking as a given that organizational prosecutions will continue for some time and that some kind of settlement option will remain preferable for both prosecutors and for organizations, defining the scope of those settlements and their terms is an important and under-examined project. The DOJ has not yet made policy statements regarding most structural reform aspects of the agreements. Reforms have not even been suggested much less explored, perhaps because these issues were drowned out by the political fray over privilege waiver issues (and now over U.S. Attorney firings). Structural reform prosecutions are so new that any next generation approach, moderated by reforms and adjustments by all sides, may be years away. It will be fascinating to see how future administrations approach such prosecutions and then how regulators, industry, and Congress react. A few possibilities for reform are outlined below, not because they should necessarily be adopted but because they suggest additional ways to think about structural reform prosecutions.
First, severe collateral consequences of indictment in organizational cases could be decreased if the relevant agencies reinterpreted debarment rules or if Congress legislated to modify those regulatory consequences. Regulatory or legislative change to the collateral consequences that organizations face would totally alter the underlying bargaining relationship between corporations and prosecutions. Reputational effects would remain, but the change could, among other things, move the negotiations from the charging stage to the plea bargaining stage. That result would permit adjudication later in a case, with more court involvement and with more information exchanged between the parties. The disadvantages, however, include a reduced deterrent threat of indictment.
A statute permitting a court to enjoin a prosecution if a prosecutor arbitrarily or unjustifiably declares a breach of an agreement would reduce the unilateral power that the terms of these agreements provide prosecutors during their implementation. An arbitrary declaration of a breach would violate the organization’s due process rights, but courts are reluctant to enjoin an indictment. Unless prosecutors no longer insist on such terms, legislation would be necessary to prevent the possible harm of an improper declaration of breach and indictment. Little attention has been paid to the enormous leverage the provisions give prosecutors at the termination stage, perhaps because prosecutors themselves may have been wary of relying on those terms and risking the dire consequences to an organization of finding a breach. Reforms might give prosecutors less catastrophic means to address a partial failure and therefore provide more appropriate tools to obtain the sought after compliance. A discussion about how these agreements terminate would also focus attention on how prosecutors should make the ultimate decision in assessing whether compliance is finally obtained.
More broadly, the DOJ could issue guidelines explaining the remedies it seeks in these agreements. All that the DOJ has done thus far is include in the McNulty memo restrictions on securing privilege waiver and nonpayment of employee legal fees. Though prosecutors have informally imitated each other’s agreements and shared practices (perhaps simply cutting and pasting other agreements, or perhaps with more consultation), there appears to be no formalized assessment within the DOJ regarding what remedies work best and which should be sought. The lack of collaboration regarding organizational prosecution policy spilled into the press recently with accusations that the rank and file within the DOJ were not even consulted during deliberations regarding the important McNulty Memo changes.16 The DOJ has adopted a structural reform mission but appears not to be using its Corporate Crime Task Force to evaluate that mission to assess whether the sought after reforms are being achieved. Nor for that matter have any regulatory agencies pursuing similar structural reform goals issued public remedial guidelines.
Finally, during the implementation of these agreements, independent monitors have not released their compliance reports, apparently for confidentiality reasons. The DOJ does not appear to be sharing this information internally or releasing general information regarding how the compliance process has been conducted. Without such information sharing, one cannot expect effective best practices to evolve internally, nor can one expect the public (or legal academics) to assess whether meaningful structural reform is occurring.
To return to where I began: difficult practical and theoretical problems of remedial design occur in any structural reform enterprise. They surface now in criminal law because the DOJ increasingly confronts Goliath but not by securing an indictment or a conviction. The United States and Goliath instead negotiate and then implement an ongoing project of structural reform. Structural reform prosecutions raise a series of complex questions and possibilities, on which my colleagues who have generously agreed to comment have shed more light. This prosecution approach warrants a sustained effort by the DOJ, organizations, regulators, legislators, courts, and scholars to assess which structural reforms effectively address underlying criminality and to carefully consider their design. Hopefully our discussion and others yet to come can be of some use both to the United States and Goliath.
Preferred citation: Brandon L. Garrett, United States v. Goliath, 93 Va. L. Rev. In Brief 105 (2007), http://www.virginialawreview.org/inbrief/2007/06/18/garrett.pdf.
[Back to this issue]