Deferred and Non-Prosecution Agreements: A Split Reality for the Individual and Corporate Criminal Defendant


In 2000, Guy Frank was caught stealing two shirts from a Saks Fifth Avenue in New Orleans. Although the total cost of the shirts was under $500, because he had been convicted of theft multiple times, Guy qualified for Louisiana’s Habitual Offender Law, which allowed the judge to impose a 23 years sentence. In April 2021 at 67, Guy was released. While the length of Guy’s sentence may be atypical, his interaction with our criminal system is all too common.

Consider the trajectory of one of the 800,0001 people a year who, like Guy, is arrested for theft in a US state criminal court. After being arrested and booked, this person typically must pay to be released on bail until his court date, or else be held in jail.2 As with Guy, the case will almost certainly resolve via a plea deal, in which the person admits his guilt in exchange for a reduced sentence.3 While the sentence will depend on the jurisdiction and value of the stolen property, the person can expect at the very least to be fined and sentenced to a term of probation, often supervised by a probation officer. Although this initial sentence may not include prison time, the person will often face incarceration if he violates any term of his supervision, including not being able to afford to pay his fees.4

Compare Guy’s path through the US criminal pipeline with that of Massey Energy Co., a corporation responsible for the nation’s worst coal mining disaster in forty years. After 29 miners were killed in an explosion in a Massey-owned mine in West Virginia in 2010, it was revealed that Massey had taken repeated, deliberate steps to circumvent mine safety laws, including the concealment from government inspectors of more than 300 safety law violations.5 Yet, this calculated deception and preventable tragedy of 29 deaths were not enough to initiate criminal charges. Instead, the Department of Justice (DOJ) offered Massey an agreement, allowing the coal producer giant to buy its way out of criminal prosecution by promising to introduce measures to improve safety and fund research into additional safety procedures. Given its $4.29 billion in annual revenues, the $209 million that Massey paid under the agreement in mine-safety improvements, fines, and restitution was a drop in the bucket.

Lenient treatment of corporations that intentionally make profit-maximizing decisions that result in significant harm is widespread. And corporate non-prosecution or deferred prosecution agreements (N/DPAs) — where the state agrees to drop or defer charges, typically in exchange for the corporation paying a fine and implementing policies to encourage future legal compliance — have become the rule in corporate prosecutions.

Part I: The Stakes of the Split Reality

The stark disparity between individual and corporate contact with the carceral state is striking and should deeply unsettle us. This disparity both reflects and reifies the gap in our framing of corporations and individual people who commit crimes. Corporations are styled as legitimate organizations critical to our national economic stability that require only internal monitoring for rehabilitation. Harsh treatment of culpable corporations would be ineffective and merely harm innocent shareholders. Individual people, by contrast, are styled as bad actors for whom harsh treatment is deserved and “soft” treatment would encourage future wrongdoing.

The widescale perception that average people face significant prison sentences but that corporate executives who commit criminal offenses will never serve a single day behind bars generates two significant, negative impacts. First, and perhaps most obviously, this patent double standard under-deters corporations, inviting criminal conduct whenever it is seen as potentially profitable or expedient. Second, and perhaps less obviously, the double standard stalls meaningful criminal justice reform. Public attitudes toward the carceral state and appetites for reform are in part driven by who is subject to its penalties. It is easy for legislatures and those who influence legislation — elite, mainly white subpopulations minimally exposed to the criminal system — to treat our coercive, inhumane, and often counter-productive criminal practices as abstract problems. However, if powerful, privileged, white people shared the possibility of being locked in a cage, the national discussion would surely shift, and the call for reform would become urgent.

For evidence of this deep capture of the national debate, consider the impact of the opioid epidemic on conversations about and punishments for drug users.  The 1970s “war on drugs” and narrative of crack users as “dangerous degenerates” helped to create a moral panic about crack cocaine, ultimately resulting in a heavily punitive approach, particularly to low-income, minority communities. In 1986, three quarters of federal funding to fight crack cocaine was allocated to police and prison rather than treatment and prevention.6 Yet, once higher-income, white people started using, and dying from, opioids, the narrative shifted. The language of war was replaced with the language of compassion, focusing on situational factors outside of a person’s control. Drug-users were “victims” struggling with addiction rather than “predators” who were morally flawed and chose a life of drugs. And in the wake of the more compassionate rhetoric, national policy followed. The funding allocation flipped: in 2017, only 15% of federal funds allocated to fighting opioids was allocated to police and prison; the rest was earmarked for treatment and prevention.7

The stark gap in the response to these two drug epidemics should teach us that the current (usually accurate) perception that privilege places you beyond the reach of the carceral state fundamentally has altered the terms and stakes of the criminal reform debate. While the US prosecutes, convicts, and incarcerates 456,000 people each year on drug charges,8 on the hand, consider the DOJ’s response to the Sackler family, the corporate owners of Purdue Pharmaceutical, on the other. The Sacklers deliberately misrepresented the safety of opioids drugs, leading to mass addiction and to hundreds of thousands of overdoses.9  Yet, they remain immune from criminal prosecution.10 If privileged people such as the Sacklers faced a substantial risk of prison, imagine the acceleration of time and resources that would be devoted to overhauling the criminal system.

The proliferation of corporate N/DPAs have secured a split reality for corporations and people. But where did these preferential corporate contracts come from, and why have they persisted? Part II of this essay explains the roots of corporate N/DPAs in corporate law and, ironically, in “diversion” programs for young people convicted of non-violent crimes. Part III describes and unpacks two dominant arguments that are marshaled in defense of corporate N/DPAs, both of which reflect and reify the greater “legitimacy” of corporations. Part IV explores the role of capture and collusion in preserving this split reality.

Part II: Origins of Corporate N/DPAs

Corporate Criminal Liability: Robust in theory, weak in practice.

In 1909, the Supreme Court announced a strict rule for corporate criminal liability: corporations are responsible for acts committed by their employees acting in the “scope” (or the course) of their employment that were done at least in part to benefit the corporation.11 This principle of vicarious liability permits prosecution of the corporation as well as those individuals within the corporation found to have committed the underlying offense. In theory, this law supports robust criminal liability for corporations. Corporations are not absolved when the misconduct occurs at low levels within the corporation.12 Nor is it a defense to argue that the offending employee acted contrary to corporate policy.13 In addition, even when there is insufficient evidence to find a single employee criminally liable, it is theoretically still possible to prosecute corporations since a prosecutor need not show that any individual employee satisfies the intent element. Under the “collective knowledge” doctrine, the prosecutor can instead aggregate the knowledge of many corporate employees to establish intent.

Despite the fact that these facially stringent rules have been relatively unaltered in the last century, the DOJ has largely chosen not to enforce these corporate liability doctrines, and corporate criminal prosecutions and convictions have precipitously declined. Since the 1990s, the DOJ has implemented a regime of corporate fines rather than incarceration of culpable corporate executives, often providing generous discounts for “cooperation” or voluntary reporting (in order to encourage “corporate compliance”). This relatively indulgent treatment has become still more friendly to corporations in recent years. While federal prosecutions of drug, immigration, and firearm offenses reached all-time highs in 2018,14 federal corporate criminal prosecutions reached an all-time low in January 2020, falling by 25% since 201515. This trend is a direct result of the proliferation of corporate non-prosecution or deferred prosecution agreements (N/DPAs). For instance, in 2000 the DOJ entered into two corporate N/DPAs but by the 2010s the number reached approximately 40 per year.16

Consider the structure of a standard DPA such as the HSBC agreement in 2012, which settled charges of money laundering violations. In exchange for HSBC admitting to, among other financial violations, processing drug money via its Mexico branch, paying a fine of $1.9 billion, accepting an overhaul of executive leadership and executive bonus performance requirements, and increasing internal monitoring to ensure future compliance with the slew of compliance reforms, HSBC was diverted from the criminal process. The Judge overseeing the DPA remarked, “taking into account the fact that a company cannot be imprisoned, it appears to me that much of what might have been accomplished by a criminal conviction has been agreed to in the DPA.”  This statement captures the default view that such agreements are victories for the justice system. Getting the corporation to pay through the nose to rehabilitate itself is seen as a successful means to deter, or at least mitigate, the risk of future wrongdoing. Yet, such deterrence may be more a pipedream than a reality. Large corporations often pay low fines, if any, under their N/DPAs. Forty-seven percent pay no fine at all, and, typically, the fines are at the very bottom of or below the range prescribed by the Organizational Sentencing Guidelines, which sets forth a uniform sentencing structure for organizations convicted of federal crimes.17

Even assuming that the fines levied on corporations were large enough to deter future criminal acts, such agreements are not victories for the justice system. Notably lacking from the HSBC agreement was a criminal indictment, conviction, or prison sentence, which is typical of N/DPAs. Few corporate criminal settlements involve any prosecution of culpable individuals.18 Yet this absence is not merely symbolic; it actually matters. The state’s choice to stamp individual criminal defendants with criminal convictions while absolving corporations as noncriminal entities serves an expressive function, signaling where the moral boundary between right and wrong falls and which actors society should consider to be worthy of moral condemnation. Indeed, the Justice Department’s routine practice of dropping or deferring criminal charges for corporations communicates that these beneficiaries have not transgressed and ought to retain their stamp of legitimacy.

Textbook law and economics deterrence theory can be invoked to justify the DOJ’s regime of fines and monitoring for corporate bad actors and prison for individual criminal defendants. The classic, stylized deterrence model suggests that monetary fines are the preferred instrument to achieve optimal criminal deterrence since incarceration imposes social costs whereas fines are socially costless transfers. Therefore, the logic runs, prison should be considered a last-ditch solution to deter only those people whose wealth is too low to permit using the threat of monetary sanctions as a deterrent.19 Put simply, loss of liberty should be reserved for shallow pockets — which happens almost always to include individual criminal defendants and almost never corporations. This theory, however, assumes that all non-monetary punishments have a monetary equivalent for deterrence purposes. For large enough corporations and wealthy enough corporate executives, this assumption seems implausible at best. To deter corporate crime successfully, the threat of prison may be necessary, and even more effective than for individual defendants. Unlike monetary fines, loss of liberty cannot be so cavalierly dismissed as a mere “cost of doing business.”

The Prototype for DPAs: Diversion Programs

Deferred prosecution agreements for corporations are modeled on diversion programs for individual people convicted of crimes. Diversion programs aim to extend “second chances” to young people charged with relatively minor offenses by insulating them from the stigma and collateral consequences of a criminal conviction. The modal diverted person is a young, non-violent person with no criminal record. As can readily be seen from the above discussion of Massey’s NPA, HSBC’s DPA, and the seemingly indiscriminate use of such agreements for corporate crimes, the drift in the implementation standards for corporate deferrals is mammoth. Unlike the non-corporate diversion reserved for first-time, non-violent people, N/DPAs remain on the table for severe and repeated corporate criminal conduct. Indeed, the unmooring of the corporate DPA from its origins captures the disparate treatment of corporations and individual people: leniency for corporations and severity for individuals.

The double standard becomes still starker after considering post-deferral prosecution when the person or the corporation violates a term of the deferral agreement. Federal prosecutors often look the other way when corporations break the law during the period when they are still bound by a DPA when the re-offense would trigger sentencing enhancements.20 Individual criminal defendants, by contrast, are less fortunate. If a diverted person fails his diversion program or a person on probation violates a term of his supervision, prosecutors can — and often do — re-initiate charges and “revoke” (i.e. reinstate) the suspended sentence.

Part III: Narratives in Defense of Corporate N/DPAs

Defenders of corporate N/DPAs predictably enlist one of two arguments to rationalize the system’s corporate indulgences. Corporate N/DPAs are celebrated either as “efficient” solutions to keep corporations on the straight and narrow or as essential deals to protect “innocent” bystanders (i.e., corporate shareholders) and forestall economic collapse.

The “Efficiency” Rationale: Corporate Compliance and Rehabilitation

Perhaps the most common justification for corporate N/DPAs is the promise of future corporate compliance. Rather than incarcerating culpable executives, which is socially costly, this efficiency defense supports N/DPAs as facilitating creative, corporation-specific rehabilitation schemes that induce voluntary, internal corporate reforms. In 2005, former US Attorney Chris Christie captured this view when he characterized DPAs as achieving “remedies beyond the scope of what a court could achieve after a criminal conviction” and the specific DPA with Bristol-Myers Squibb as a “push down the road of good corporate citizenship” for BMS through a series of internal reforms.21

However, the trends in corporate crime do not bear out such optimistic claims about incentives and rehabilitation. As discussed above, corporate recipients of DPAs often re-offend during the period of their agreement. Indeed, the list of corporations that were extended DPAs, recidivated during the period of their DPA, and yet did not receive harsher treatment is a long one and includes BP, ExxonMobil, Pfizer, GlaxoSmithKline, AIG, Barclays, HSBC, JPMorgan, and UBS.22 Put simply, agreements to enhance corporate compliance effectively become empty promises if beneficiaries of N/DPAs know that the corporation won’t be held in breach of the agreement if it re-offends.

Contrast the perceived rehabilitative potential of a corporate N/DPA with that of a deferral agreement for an individual defendant charged with theft. It is difficult to imagine legislators or defenders of corporate N/DPAs condoning a deal that dropped all charges against a thief simply in exchange for his promise to implement “personal growth policies” to ensure future lawful behavior. The likely reaction to such a proposed deal would be distrust of the individual who has proven himself to be dishonest and immoral, and for whom such a deal would provide no incentive to change his behavior. Moreover, it would be even more far-fetched to imagine support for a proposal to drop or defer charges against a person with a record of multiple property crimes. After all, in our criminal system, a critical determinant of sentencing outcomes for individual criminal defendants is prior criminal convictions; and virtually all states have enacted some form of “habitual felon” or “three-strikes” statute that mandates sentencing enhancements for defendants with multiple prior convictions. While the goal of rehabilitation has been all but abandoned in individual prosecutions, rehabilitation seems to have become all the rage when it comes to corporate crime.

This double standard with regard to “rehabilitation” is further revealed in the types of evidence considered probative of intent and guilt in the criminal prosecution of people, on the one hand, and corporations, on the other. The DOJ’s “Principles of Federal Prosecution of Business Organizations” instructs prosecutors to consider post‐offense corporate conduct in their moral culpability analysis, including such factors as post-offense disclosure, remedial actions such as restitution, and even collateral consequences of prosecution.23 Although corporate remedial actions after the discovery of a crime almost certainly do not reflect corporate intent at the time of the offense, the extent of these remedial actions nevertheless partially determines whether any criminal liability exists.

Contrast this DOJ policy for corporations with the unforgiving treatment of post-offense remedial actions in individual prosecutions. For an individual defendant, the legal inquiry for guilt turns on the elements of the crime at the time of the offense. Consideration of whether, for instance, the victim received restitution would not factor in an intent analysis since restitution is not an effective “indicium” of the person’s state of mind.

The divergence in the system’s approach to individual and corporate rehabilitation may partially be grounded in the gulf between the perceived legitimacy of corporate executives and of that of people who commit crimes. Although the law’s sweeping criminal liability standard for corporations appears to strictly enforce liability, in practice it leaves all the discretion to determine a corporation’s degree of blameworthiness and ultimate indictment status in the hands of the prosecutor.  And it is undoubtedly more natural for a federal prosecutor to give powerful corporate executives — the captains of US industry — the benefit of the doubt. Indeed, the promise to implement monitoring policies that ensure future corporate compliance has a sheen of legitimacy that is absent for the typical individual. While the corporate promise is sanitized, the individual promise is tainted. While the corporation has the potential to reform itself, the individual person has proven himself to be a bad actor who cannot be saved. This gap in legitimacy inevitably transforms prosecutorial discretion — a pillar of our criminal system — into an institution of “voluntary deference to directives of legitimate authorities.”24

The “Collateral Consequences” Rationale: Protecting Shareholders and the Economy

Another standard justification for corporate N/DPAs is that they reduce collateral consequences for corporate stakeholders who are not responsible for the criminal offense and so are undeserving of punishment. Indeed, in 2003 the DOJ explicitly adopted a policy to consider a corporate conviction’s collateral consequences in order to reduce the harms to those dependent on the corporation but uninvolved in its misconduct.25 Since shielding shareholders from financial loss comes at the expense of corporate criminal deterrence, this “innocent shareholder” defense of N/DPAs implies that shareholder profit ought to trump the public good. And, here again, the impulse to frame corporate N/DPAs as protecting shareholders rather than disincentivizing social responsibility both reflects and reifies the greater legitimacy of corporations. Ironically, recent work suggests that DPAs may actually increase collateral consequences relative to prosecutions: shareholders of corporations subject to DPAs experience wealth losses relative to prosecuted firms.26 However, it is doubtful that this evidence will effectively sway support for lenient corporate prosecutions.

Perhaps most troubling about the innocent bystander justification is the inconsistency with which it is applied to individual and corporate prosecutions. Collateral consequences of individual convictions are well documented, not only for people who are convicted but also for their families and, in particular, for their children.27 Collateral consequences for those convicted of felonies include deportation, loss of public benefits, and loss of work, all of which may generate lasting consequences for their families and children. It is difficult to imagine anyone arguing that the children of people convicted of felonies ever deserve to be punished — indeed, it would be difficult to construct a better example of an innocent bystander than the child of an incarcerated person — and yet the system seems comparatively unfazed by the staggering impacts of criminal convictions on families and children.

The other innocent bystander that corporate N/DPAs ostensibly serve to protect is the national economy. Many large corporations have amassed so much power that the criminal system perceives them as too large to “efficiently sanction” or “too-big-to-jail.” Deputy Attorney General Larry Thompson captures the spirit of this view in his 2011 keynote speech, “The Reality of Overcriminalization.” Thompson cautions that “there is a real risk of over-deterrence when corporations are convicted of crimes.” Since a conviction typically “sounds the death knell for a corporation,” “[o]ver-deterrence, thus, comes at a price. Costs to shareholders, to the economy, to the community in which the corporation is located, to employees, as entities that formerly contributed to a thriving organization, that may disappear forever.”28 Here too the analogous economic impacts of incarcerating people are relatively overlooked. Abundant evidence from sociological and economic reveals that prison stigmatizes formerly incarcerated individuals, erodes their job skills, and reduces their social capital, all of which adversely impact their own future employment prospects as well as the potential growth of the aggregate U.S. economy.29

By marshaling the language of economic collapse exclusively in the corporate context, the too-big-to-jail argument both reflects and reinforces the greater importance of corporations. When the stakes of corporate criminal deterrence are highest, this economic-fallout defense does the most work. Ironically, the larger and more powerful the corporation, the larger is its potential to harm the public, and the more likely it is to benefit from a DPA.

Part IV: Capture and Collusion

Internal Investigations: Capture from Within

One feature of the criminal process for corporate prosecutions in particular invites a dynamic of capture and collusion between federal prosecutors and corporate defense attorneys. The DOJ relies on a corporate defendant’s internal investigation for its evidence. Although the choice of the internal investigator must be approved by the prosecutor, it is the corporate defendant that is charged with hiring and paying the outside counsel and law firm to head up the internal investigation. It is almost impossible to imagine an analogous arrangement in criminal investigations of individual defendants. Indeed, it would strike us as ludicrous to permit a defendant to choose one of his friends to investigate the extent of his alleged wrongdoing.

Internal corporate investigations generate structural weaknesses in corporate prosecutions. Perhaps most obviously, the outside counsel and firm investigating the corporation have a clear conflict of interest: they are at once counsel for the defendant and the prosecution’s source of evidence. These competing roles create an incentive for the investigator to appear as though she is vigorously investigating wrongdoing while she is actually stopping well short of a full probe of the company — and perhaps even secretly hiding new inculpatory evidence. Firms and outside counsel that are best able to play this game and secure corporate-friendly N/DPAs will build reputations enabling them to secure future corporate clients.

Even more damning to corporate criminal prosecutions, federal prosecutors often are wholly dependent on the internal investigations for the production of case documents, the analysis of firm data, and the questioning of corporate employees. In investigations of non-corporate conspiracies, prosecutors will often question suspects individually, stymieing their ability to coordinate their stories. By contrast, in corporate investigations, the defense counsel conducts employee interviews prior to any questioning from the prosecutor. This timeline permits the defense attorney to coach corporate employees, allowing them to coordinate their answers to potentially tough, incriminating questions.

The Interaction of No Judicial Review and the Revolving-Door Problem

The absence of judicial review of N/DPAs also leaves the process vulnerable to corporate capture. It allows prosecutors to enter into expedient deals that secure the political benefits of a corporate prosecution without risking recriminations from the bench for a half-hearted investigation that fails to secure evidence inculpating corporate management. This absence of judicial scrutiny creates particularly weak incentives to vigorously prosecute corporations in light of the revolving door between the DOJ and the corporations it prosecutes.30 Although it is impossible to prove the extent of this phenomenon, circumstantial evidence of the revolving-door pattern — federal prosecutors who were former corporate attorneys and who ultimately return to corporate defense work after leaving government work — abounds.{{Below are examples of press releases of former AUSAs joining private criminal defense firms:}} For instance, many DOJ leaders during the Obama administration — including Attorney General Eric Holder, Assistant Attorney General Lanny Breuer, and many of Breuer’s deputies — had previously worked at the law firm of Covington & Burling (and returned there immediately after).31 And Quinn Emanuel, a criminal defense firm, boasts on its website over 25 of its partners who have experience as a federal prosecutor.32

Given this revolving door, federal prosecutors who seek future corporate employment have a clear incentive to avoid throwing potential future clients in prison. The absence of judicial review of N/DPAs helps to facilitate this temptation. After all, the lack of oversight from a federal judge means that there is nothing to deter federal prosecutors from extending attractive agreements to those corporations whose ranks they hope to join after their stint in government service. Indeed, it is easy to see how the interaction of the revolving door phenomenon and the absence of robust judicial scrutiny could generate a culture of DOJ deference to corporations that face criminal indictment.


The overhaul of corporate N/DPAs is long overdue. Most obviously, abandoning the DOJ’s default practice of extending attractive deals to corporations would help to restore corporate criminal deterrence. As long as corporations can continue to secure attractive agreements from the DOJ, they have little incentive to avoid criminal conduct whenever it is potentially profitable or expedient. More fundamentally, reversing this corporate indulgence could alter the future trajectory of the carceral state. As soon as powerful, privileged people face the risk of the state stripping them of their liberty, legislatures would shift into gear and real criminal reforms may be realized.