The evolution of corporate criminal settlements: an empirical perspective on non-prosecution, deferred prosecution, and plea agreements.

• Author: Alexander, Cindy R.
• Position: Introduction through III. Sample Construction, p. 537-566

1. Introduction

   Over the past decade, two novel approaches to corporate criminal prosecution have emerged, the Deferred Prosecution Agreement ("DPA") and Non Prosecution Agreement ("NPA"). DPAs and NPAs have been used in virtually all areas of corporate criminal wrongdoing including antitrust, fraud, domestic bribery, tax evasion, environmental violations as well as foreign corruption cases. These legal mechanisms bypass the traditional plea-bargaining process and instead involve a negotiated settlement whereby the organization may agree without pleading guilty (or nolo contendere) to a combination of restitution, forfeiture, monetary sanctions, and other legal and structural governance reforms. DPAs and NPAs did not arise from any change in the federal statutes but through an innovation in criminal enforcement practices and related coordination between criminal and civil enforcement authorities.

   All U.S. federal criminal cases are settled at the discretion of the United States Department of Justice (DOJ), operating through prosecutors located in U.S. Attorneys' Offices and divisions. Those prosecutors, along with representatives of the corporation, are the necessary signatories of each agreement. The government and corporation are thus its presumed beneficiaries. (1) The Criminal Division of the DOJ has been a signatory to the majority of NPAs and DPAs entered into by companies, although other divisions of the DOJ have entered many NPAs, DPAs, and plea agreements, as have its affiliated Offices of the U.S. Attorneys. The Department of Justice has supported the exercise of prosecutorial discretion in these matters through the issuance of a series of memoranda, starting with the Holder Memorandum in 1999, with the most recent occurring in 2008. (2)

   The standards and guidance that prosecutors face in deciding whether to seek a DPA, NPA or traditional plea agreement, and in negotiating sanctions under DPAs and NPAs have evolved over the past decade. The Thompson Memorandum in 2003 marked the start of a "DPA era" by encouraging prosecutors to substitute NPAs and DPAs for plea agreements in those instances where companies voluntarily disclosed and/or cooperated with the investigation of wrongdoing. (3) Two other sources of guidance on charging practices and sanctions for corporations under federal criminal law remain relatively unchanged. The first are the Organizational Sentencing Guidelines. (4) The second, the federal case law, contains only limited discussion of issues relating specifically to NPA or DPA settlements. (5)

   Generally, criminal settlement agreements for corporations--NPAs, DPAs and plea agreements--contain the following four elements: an admission of facts, an agreement of cooperation, a specified duration for the agreement, and an agreement to monetary and non-monetary sanctions. Common sanctions include restitution, fines, probation, appointment of monitors, and termination of responsible individuals. (6) While structurally similar, the legal and policy implications of NPAs, DPAs and plea agreements differ in significant respects.

   While popular among federal prosecutors, the use of DPAs and NPAs has become controversial in both the legal and policy arenas. One noted legal scholar claimed that their use "erodes the most elementary protections of the criminal law, by turning the prosecutor into judge and jury, thus undermining our principles of separation of powers." (7) Another scholar remarked that DPA "agreements typically do not provide for judicial review of implementation or of any alleged breach, and they often require the organization's permanent future cooperation." (8)

   In addition to the concerns raised by legal theorists, there is little agreement about the practical efficacy of DPAs and NPAs. Questions have emerged about the magnitude and potential collateral consequences of both monetary and nonmonetary sanctions under NPAs and DPAs relative to traditional plea agreements, as well as the efficiency and ultimate deterrent effect of these alternative settlement vehicles. Some of these questions are similar to those encountered in the plea-bargaining debate. (9) Yet answers in this new context involve consideration of a wider array of settlement outcomes than tend to arise in the use of corporate plea agreements. (10)

   Some of the debate over NPAs and DPAs has focused on their effectiveness relative to a baseline of criminally charging and prosecuting corporations. On the one hand, the view is that, by avoiding a criminal plea and resulting judgment, offending companies may avoid the costly collateral consequences, such as debarment from government contracts. Indeed, DPAs and NPAs are sometimes regarded as enabling some companies to avoid insolvency. This is reflected in the claim that DPAs and NPAs "rebuild shareholder and public confidence in corporations while allowing companies to avoid certain death by indictment." (11)

   On the other hand, relative to "criminal prosecution" by plea agreement, there is the concern that "deferred prosecution and non-prosecution agreements limit the punitive and deterrent value of the government's law enforcement efforts and extinguish the societal condemnation that should accompany criminal prosecution." (12) One commentator expressed concern:

   "The rise of these agreements has undermined the general deterrent and adverse publicity impact that results from corporate crime prosecutions and conviction.... It could very well be that the rise of these deferred and nonprosecution agreement deals represents a victory for the forces of big business who for decades have been seeking to weaken or eliminate corporate criminal liability." (13)

   Still other views of NPAs and DPAs arise from comparisons against a baseline of no criminal action, such as may occur through administrative or civil settlement for the alleged misconduct. There is concern that widespread use of DPAs and NPAs may empower prosecutors to extract concessions beyond what some companies would face if prosecutors were instead left with no alternative to a plea agreement or trial other than non-action. By making it easier for corporations to reach a criminal settlement without a plea or trial, the introduction of DPAs and NPAs may well have increased the reach of the criminal law enforcement into areas that would otherwise be subject only to administrative compliance actions or private litigation.

   The objective of this article is to shed new light on the dramatic transformation in corporate criminal law enforcement that has occurred, pre-versus post-2003 by introducing evidence on the related evolution of corporate criminal settlements over the period. While a few prior studies have documented the increased use of NPAs and DPAs and the provisions that are contained in them, (14) there remains a dearth of empirical evidence on how NPAs and DPAs differ from traditional plea agreements and how the reach of United States corporate criminal enforcement has evolved in their presence. (15) Little is known about either the extent to which NPAs and DPAs are used as substitutes versus complements for each other; or whether NPAs and DPAs have tended to supplant or supplement traditional criminal pleas. Little evidence exists about whether DPAs and NPAs systematically differ from plea agreements (or from one another) on important dimensions such as monetary sanctions, legal process-related provisions, and mandated governance reforms.

   To fill this gap in the empirical literature, this Article examines evidence from a large-scale research project expressly designed to study these settlement agreements. Since no comprehensive data had previously been collected, the study began with an exhaustive effort to identify and catalog all NPAs, DPAs and plea agreements entered into by public corporations between 1997 and 2011, a period that frames the 2003 Thompson Memorandum date and allows us to make comparisons over time, pre-versus post-Thompson, and within eras. The data generated through this effort permit the first systematic characterization and comparison of all known agreements between federal prosecutors and public companies over the period. Of the 486 agreements identified, 157 are NPAs and DPAs and 329 are plea agreements.

   This Article presents initial evidence from analysis of these new settlement data, focusing on three overarching questions. The first is whether the increased use of NPAs and DPAs has been accompanied by a decline (or increase) in the use of plea agreements. The second question focuses on what factors are associated with the choice to settle a criminal investigation with an NPA, DPA or plea agreement. Finally, we examine some of the differences in the settlement outcomes that have been achieved using NPAs, DPAs and plea agreements.

   As to the first question, the data reveal an increase in the use of plea agreements starting in 2007, alongside a marked increase in the use of NPAs and DPAs since 2003. In addition, while foreign and domestic bribery cases make up the largest component of NPAs and DPAs, the post-2003 timing of the enforcement action may be more important than offense type in explaining the use of an NPA or DPA. In particular, the rise of DPAs and NPAs is not explained by the dramatic rise that has occurred in the bribery area, foreign or domestic. Bribery cases account for less than half of DPAs and NPAs since 2003. (16)

   Second, we find that the use of a DPA or NPA rather than a plea agreement is associated with several factors that include: corporation's culpability for the offense, (17) whether or not the offense involved business dealings with the government, (18) and whether a parent company or subsidiary was the defendant. Subsidiaries do not enter NPAs and DPAs as frequently as do their affiliated parent companies. (19)

   Finally, with respect to the differences in settlement outcomes that prosecutors and companies reach through NPAs and DPAs versus plea...