Against Adversary Prosecution

I.     Introduction

In the book Just Mercy, attorney Bryan Stevenson recounts the true story of his long struggle to exonerate Walter McMillian, a man sentenced to death for a murder he did not commit.1 In one scene of the book, Stevenson meets with the newly elected district attorney for Monroe County, Alabama, where McMillian had been convicted.2 Hoping to convince the prosecutor to agree to reopen the case, Stevenson describes the (rather overwhelming) evidence he has gathered to show his client’s innocence. After some discussion, the prosecutor angrily replies: “[M]y job is to defend this conviction.”3

Implicit in this prosecutor’s statement is a particular view of his role: that he is a partisan lawyer tasked with obtaining and preserving convictions. The literature on American prosecutors suggests that this view is commonly held.4 Prosecutors, as well as the general public, see the criminal justice system as an adversarial contest between two sides.5 The government wins this contest if the defendant is convicted and punished. The government loses if the defendant is acquitted or the conviction is later vacated. Consequently, the prosecutor is expected to advocate zealously for conviction and punishment. This means they must make strategic decisions that will maximize the likelihood of victory, just as a private lawyer would act strategically on behalf of their client.

 Of course, prosecutors are not supposed to be mere adversaries in our system—they also have a special duty to “seek justice.” The American Bar Association’s (“ABA”) Model Rules of Professional Conduct instruct that “[a] prosecutor has the responsibility of a minister of justice and not simply that of an advocate,” and that this responsibility entails ensuring “that the defendant is accorded procedural justice.”6 Prosecutors are thus described as having a “dual role.”7 They must seek convictions and punishments in an adversary contest, while also working to ensure the fairness of that contest.

But these two roles are in conflict with one another. Adversarial lawyering is essentially amoral—it requires acting strategically to win litigation, however the client defines victory, regardless of the lawyer’s own ethical commitments. Seeking justice, however, requires choosing what substantive values to pursue. And so these two roles push prosecutors in different directions. For any particular choice a prosecutor makes in a case—e.g., adding a count, offering plea-bargain terms, revealing evidence to the defendant, fighting a post-conviction innocence claim—they must decide whether they will strategically maximize their likelihood of victory (which might be defined as securing or preserving a conviction, obtaining a significant punishment, or some other outcome), or instead make room for values like mercy, due process, and proportionality.

This tension, combined with the powerful cultural and professional forces that push prosecutors to seek convictions and harsh punishments, causes adversarialism to dominate American prosecution.8 This adversary zeal is not just an academic concern—it has significant consequences for the administration of criminal justice. Prosecutors who see their professional goal as strategically maximizing convictions, or punishments, are more likely to use their charging and plea-bargaining discretion to secure guilty pleas and excessive sentences. Indeed, recent empirical scholarship suggests that prosecutorial charging decisions have been a driving force behind the growth of the American prison population over the last 20 years.9 Adversarial prosecutors are also more likely to ignore evidence of innocence and to fight meritorious post-conviction innocence claims.10 Studies of exonerations suggest that prosecutorial overzealousness plays a major role in wrongful convictions.11 How we define the prosecutor’s role is thus one of the most important design questions in our criminal justice system.

This Article argues that the role ethics of American prosecutors should be reconceived as non-adversarial. Analogous to prosecutors in countries with inquisitorial criminal justice systems, like France and Germany, American prosecutors should not think that their goal is to win a case by securing conviction and punishment.12 Indeed, the argument for non-adversarial prosecution is stronger in the American system than it is in inquisitorial systems. This is because, in America, the prosecutor effectively wields more power than a judge—he or she dictates case outcomes by making charging decisions, controlling pre-trial investigation, and extending plea-bargain offers. By contrast, in inquisitorial systems the judge has power to investigate cases and control the presentation of facts, and prosecutors’ charging discretion and plea-bargaining power are sharply constrained.13 This makes it ironic that American prosecutors are adversarial, while inquisitorial prosecutors are not. Adversary prosecution is uniquely dangerous in a system where the prosecutor is the most important decision-maker.

However, rejecting adversary prosecution, and thereby embracing the pursuit of justice as American prosecutors’ sole mandate, is only the first step. This proposed non-adversarial role must itself be unpacked, because there are many values that prosecution could serve in the name of justice. In particular, one crucial question is whether prosecutors themselves will dictate the system’s value choices, or whether they will implement policies set by other actors. For example, again by analogy to inquisitorial systems, American prosecutors might think of themselves as neutral conduits for the law. This would require outside institutions like courts, legislatures, and sentencing commissions to impose rules that govern prosecutors’ decision-making, and for prosecutors to fit the facts of each case into these externally established rules. Such an ethic might be described as “positivist”—it divests prosecutors of moral agency and turns them into neutral law enforcers. However, as a practical matter, the American system gives prosecutors too much discretion for them to act as morally neutral bureaucrats at every stage of the criminal justice process. American prosecutors have nearly unconstrained discretion to decide on charges and set plea terms. In making such decisions prosecutors must determine, either through established office policies or individual choices, how they are going to weigh the different values that these decisions implicate. This Article thus proposes that “seeking justice” can itself be viewed as containing a dual role for American prosecutors. Where established rules and standards constrain prosecutors’ discretion, they should neutrally apply those rules and standards. Where they exercise discretion, they should make value choices for which they are held morally accountable.14

Part II of this Article explores the structure of prosecutors’ professional ethics in the United States. It begins by describing the current “dual role” model as articulated by courts, the academic literature, bar associations, and prosecutors themselves. First, it analyzes how the dual role is constructed in theory, both by canonical statements of prosecutorial ethics and by legal thinkers. Next, it discusses how the dual role plays out in practice, with several features of the American system causing an adversarial ethic to prevail. It then recounts the transition of American prosecution from a system of private litigation in which prosecution was driven by victims, to a quasi-public system in which prosecution still relied heavily on private funding, and finally to the current system in which the government monopolizes the prosecution function. The duality of American prosecutors’ role ethics is a product of this unique history, reflecting a long-standing mixture of private and public models of prosecution. Part II closes by contrasting the American system with inquisitorial systems in continental Europe, where prosecutors are supposed to be neutral, judge-like figures who do not take an adversarial posture towards the defendant.

Part III makes the case that American prosecutors should not be viewed as adversary attorneys. It does so by developing one argument against adversarial prosecution, and then critically evaluating three possible arguments for adversarial prosecution. First, prosecutors in the American system wield an immense amount of power. They have nearly unilateral control over charging decisions, they determine the evidence that will be used against the defendant, and they set the terms of plea-bargain negotiations.15 Given this power, it is dangerous for prosecutors to think of their job as “winning” cases. Second, American prosecutors do not have “clients” in any normal sense. While victims are sometimes given a voice in the criminal justice process, the prosecutor is not their lawyer. And while it is conventionally said that the prosecutor represents the state (or “the people”), that does not logically entail an ethic of adversarial conviction-seeking, and if anything lends support to a non-adversarial role.16 Thus, to the extent that adversary lawyering can be justified by reference to the lawyer’s relationship with their client, such a justification is unavailable to prosecutors. Third, American prosecutors have nearly unlimited discretion to bring charges or not bring charges according to their moral views about what conduct is worthy of punishment. Thus, to the extent that adversarialism is justified by rule-of-law values because it removes moral discretion from law enforcement, this justification is also unavailable to American prosecutors. Finally, the strongest argument in favor of adversarial prosecution is that the litigation contest helps to generate facts and legal arguments. Unlike in inquisitorial systems, American courts rely primarily on the parties to present evidence and brief legal issues. However, this argument contains two important caveats. The first is that non-adversarial prosecutors can still generate and present the relevant evidence and arguments, though perhaps not with the same competitive zeal. The second is that only a relatively small number of criminal cases are resolved after an adversary trial (much less one with an effective defense), and so it would be perverse to let prosecutors’ role in trial practice justify a general ethic of adversary combat.

Part IV explores the abstract forms of three possible role logics for American prosecutors. This trichotomy describes three different ways that prosecutors can approach their decisions. First is a purely adversarial approach, in which prosecutors wield their discretion strategically to convict the defendant and to punish them. Prosecutors adopting an adversarial approach may use all the lawful tools at their disposal—charging all possible counts, forcing decision on a plea offer before providing discovery, seeking the maximum sentence—to achieve their goal. Second is a model of positivism, where the prosecutor’s task is viewed as a form of value-neutral adjudication. In such a model the prosecutor seeks out evidence of the crime, and then proceeds to help judge and jury resolve the case. This approach requires prosecutors to remain disinterested in the outcome of the case, and to seek only to make (or help other actors make) legally and factually correct decisions. Third is an approach in which prosecutors wield their discretion by deciding between different moral values. In making decisions they must weigh the importance of principles like punishing the guilty, preventing crime, preserving due process, and ensuring equal treatment of defendants. Unlike normal lawyers, prosecutors who adopt such a value-weighing approach are morally answerable for the positions they take. Part IV closes by synthesizing these ideal types into an alternative dual role for prosecutors. It argues that the ethics of positivism and value weighing should be combined, and that the ethic of adversarialism should be excluded from all aspects of prosecution (even trial).

Part V proposes some broad ideas for a program of institutional reform in light of this role ethic. First, it looks at how courts, legislatures, and state bar associations could exercise their power to help determine the values that prosecutors pursue. Second, it considers the structural factors that define the institutional prosecutor’s office—the size, the resources available, the degree of hierarchical control, the way cases appear—and how these factors make value weighing more or less feasible in particular kinds of offices. Part V closes by examining the various incentives that can push prosecutors towards overzealousness, including electoral politics, bureaucratic pressure to meet one’s numbers, and an office culture that values victories, suggesting ways that these incentives could be diminished or overcome.

The thesis of this Article is in one sense a significant departure from the current system. It calls for us to reorient prosecutors’ role morality such that they do not think of convictions and punishments as victories, and acquittals as defeats. In another sense, this reorientation is already implicit in the way the legal profession describes prosecutors’ professional ideal: as “seeking justice,” at least in part. But the continued legitimacy of adversarial prosecution interferes with the search for justice. When a prosecutor goes too far in their partisan zeal—trying to preserve an incorrect conviction or adding charges against a defendant to secure a harsher plea—this is thought of as simply an extension of the prosecutor’s legitimate desire to win. It should instead be considered a betrayal of their role. Describing prosecutors as having a dual role, however, legitimizes the partisan half of that duality, and restricts criticism of prosecutors to cases where they break the rules. The duty to seek justice should mean more than avoiding misconduct. Bringing theoretical clarity to prosecutors’ professional role can help us to build a system of ethics that contains a more robust conception of the duty to seek justice.

II.     The Role of Ethics of American Prosecutors

The role ethics of American prosecution occupy a middle position between two conflicting professional ideals. One ideal is justice-focused, non-partisan, and drawn from government decision-making. The other ideal is victory-focused, partisan, and drawn from private litigation. This Part explores the tension between these two professional ideals and describes how the adversarial ideal dominates in practice. It further shows how this conflict is connected to America’s historical transition from a system of private prosecution to a system of public prosecution. It ends by contrasting the American system with those of France and Germany, where prosecutors are expected to play a neutral role.

A.     The Dual Role in Theory

American prosecutors are officially entreated to seek not merely convictions, but also justice. There is a recognized list of canonical statements to this effect. The most commonly cited is Justice George Sutherland’s opinion in the 1935 case Berger v. United States, in which he wrote that the government’s interest “in a criminal prosecution is not that it shall win a case, but that justice shall be done.”17 The ABA’s Model Rules, which have been adopted by nearly all state bar associations, describe the prosecutor as having “the responsibility of a minister of justice and not simply that of an advocate.”18 The National District Attorneys Association advises that “[t]he primary responsibility of a prosecutor is to seek justice.”19 An inscription on the rotunda of the Department of Justice building reads: “The United States wins its point whenever justice is done its citizens in the courts.”20 Prominent prosecutors sometimes assert in public statements that their job is to be just and fair, not merely to put people away.21 Judges often invoke the “seek justice” norm when chiding prosecutors who have failed to live up to it.22

At the same time, American prosecutors are also understood to be part of an adversarial criminal justice system. In that context, they are expected to be partisan advocates for the conviction and punishment of defendants. Justice Potter Stewart wrote in Herring v. New York that “[t]he very premise of our adversary system of criminal justice is that partisan advocacy on both sides of a case will best promote the ultimate objective that the guilty be convicted and the innocent go free.”23 While the ABA’s Model Rules do not explicitly state that prosecutors must act as partisan adversaries, they also do not excuse prosecutors from that requirement.24 The ABA’s Standards for Criminal Justice summarize prosecutors’ obligations by stating that “[t]he prosecutor is an administrator of justice, a zealous advocate, and an officer of the court.”25 Professional ethics and prosecution casebooks consistently refer to prosecutors as adversary advocates for the conviction of defendants.26 The academic literature on prosecutors reflects this conventional wisdom that they act as partisan lawyers.27

There is a profound tension between these two roles. One role—“seeking justice”—requires attentiveness to systemic concerns, including the rights of defendants. It calls on prosecutors to step outside narrow partisanship and consider how a just system should operate. The other role—adversarialism—puts a premium on winning cases, thus requiring a certain lawyerly amorality. Adversary lawyers are the agents of a cause, and their personal views about the justice of that cause are incidental. So long as they stay within the rules, they should bend all their decisions toward the strategic goal of winning the case. For instance, a private civil lawyer arguing a contract case is not generally supposed to seek “justice” unless doing so helps their client’s goals.28 In the prosecutorial context, this means an adversary prosecutor tries, for instance, to limit defendants’ ability to take advantage of their procedural rights.29 And for certain basic decisions a prosecutor must make—e.g., how to use their charging discretion, how much evidence to disclose, whether to proceed against ineffective defense counsel—these two roles pull them in opposite directions.

Legal thinkers have dedicated much effort to exploring how these two role ethics might coexist in theory. They have adopted three basic approaches. The first is to call for a general ethic of qualified adversarialism, in which the two roles are combined throughout the criminal justice process. The second is to enact specific rules that prosecutors must follow to preserve justice, which act as exceptions to the general rule of adversarialism. The third is to carve out certain stages of the criminal process where prosecutors must attend to the needs of justice, while letting an ethic of adversarialism prevail elsewhere.

The first approach seeks to incorporate both adversarialism and justice-seeking into a single coherent and undifferentiated role morality. It is commonly expressed as a call for criminal prosecutions to be “fair fights,” analogous to holding back while playing basketball against a younger sibling, or to the ethic of fair sportsmanship adopted by hunters.30 The prosecutor knows that the playing field is unequal, so they have a general obligation not to exploit their advantages in a way that would make the contest unfair. They are both competitor and referee. Attorney General Robert Jackson articulated this principle in a famous 1940 speech on the role of prosecutors, stating: “The qualities of a good prosecutor are as elusive and as impossible to define as those which mark a gentleman. . . . A sensitiveness to fair play and sportsmanship is perhaps the best protection against the abuse of power . . . .”31 Several legal scholars have tried to provide a bit more substantive content to this idea of “sportsmanlike” prosecution. In doing so, they have mostly described the duty to seek justice as only a modest impediment to the general rule of adversary prosecution. Fred Zacharias argues that once a decision to prosecute has been made, the duty to seek justice encompasses only a narrow obligation to ensure that the basic elements of adversarialism exist.32 This means, for example, ensuring the tribunal is not overly biased and not taking advantage of ineffective defense counsel.33 But it does not otherwise interfere with prosecutors’ duty to aggressively seek convictions. Kevin McMunigal suggests that prosecutors’ obligation to seek justice is essentially no different from private attorneys’ duties as officers of the court.34 And Bennett Gershman proposes that, while prosecutors may behave as adversary advocates, they also have duties to seek truthful testimony and verdicts, and to refrain from conduct that will impede the search for the truth.35 These and similar theories leave the basic logic of adversary prosecution intact, but incorporate some notions of fair play and rights protection so as to limit abuses.36

The second approach is to create rules that suspend adversarialism for certain specific decisions that prosecutors must make. This means not trying to combine adversarialism and justice-seeking throughout the prosecutor’s job, but instead formally distinguishing them so that a prosecutor knows when to switch modes. This can be done by imposing concrete ethical obligations on prosecutors requiring them to help defendants in certain ways. The ABA’s Model Rules impose several such obligations, requiring that prosecutors make timely disclosure of all exculpatory evidence, ensure that defendants have the opportunity to obtain counsel, and seek to remedy convictions they believe are incorrect.37 Various legal scholars have argued for a number of similar ethical obligations, including requirements that prosecutors critically evaluate evidence from cooperating witnesses,38 counteract systemic racial biases,39 decline to use admissible evidence if it was unconstitutionally obtained,40 waive peremptory juror challenges,41 and assist defendants who have ineffective or conflicted counsel,42 among other proposals.43 Such ad hoc rules provide concrete content to the prosecutor’s duty to seek justice, but they do not override the baseline principle of adversary prosecution.

The third approach legal scholars advocate is to subdivide the prosecution process into different stages, and to have prosecutors act as adversaries only during certain stages, while giving them a non-adversarial role in others.44 Distinguishing between these stages turns on whether the premises of the adversary system are satisfied. If there is effective advocacy on both sides and a neutral decision-maker controlling the proceeding, then prosecutors are free to behave like adversary lawyers. However, if adversarial safeguards are not present, either because the prosecutor has unilateral control (as in charging decisions) or because there is no effective supervision by a neutral party (as in plea-bargain negotiations), then the prosecutor should make efforts to preserve the defendant’s rights. Some scholars have even called for prosecutor’s offices to be subdivided into different groups of prosecutors, with one group engaging in adversary trial advocacy and another performing more managerial functions like charging decisions and investigations.45 This stage-based approach to defining prosecutors’ role formalizes the “dual role” framework by turning prosecution into two different jobs—advocate and adjudicator. It thus embraces prosecutors’ adversary role, while at the same time confining adversarialism to only those stages of the criminal justice process where prosecutors exercise the least unilateral power.

B.     The Dual Role in Practice

The “dual role” in prosecutors’ ethics is dangerous in practice, because the adversarial half of that role creates a permission structure for prosecutors to be harsh and inflexible. This permission structure is especially harmful because prosecutors face powerful professional incentives to exclusively prioritize conviction and punishment.46 Adversarial role morality amplifies these incentives. It is useful here to distinguish between three different concepts of adversarialism—adversarialism in the canonical statements of legal ethics, in individual prosecutors’ understanding of their role, and as a sociological reality for prosecution decisions. Adversarialism in legal ethics—our collective understanding that prosecutors can act as partisan lawyers, which is informed by judicial statements, official texts, and pronouncements of the organized bar47—fuels the other two types of adversarialism. It fuels individual adversarialism because it helps prosecutors cultivate an internal sense that amoral partisanship is ok. This gives them moral license to assume guilt, treat defense claims as presumptively untrue, minimize equities, and ignore doubts. And this in turn fuels sociological adversarialism, rendering our justice system harsher and less fair.

Prosecutors face many incentives to focus on punishment and conviction to the exclusion of all else.48 Adversary role morality strengthens these incentives by giving prosecutors an excuse to succumb to them. Some of these incentives are professional. Prosecutor’s offices promote attorneys who gain a reputation for winning cases, and keep track of conviction rates.49 Certain offices even give prosecutors conviction bonuses, have them compete over the number of convictions they secure, shame them for losing cases, and perform rituals to celebrate trial victories.50 There is also a culture of competitiveness in many prosecutor’s offices, which is fed by the expectations of police, victims, and other prosecutors, as well as the countervailing adversarial behavior of defense attorneys.51 These adversary incentives have an especially strong effect on young and risk-averse prosecutors.52 Further, in my own experience as a public defender, I have observed that individual prosecutors’ adversarialism is reinforced by the day-to-day reality of their work. Most cases first appear as a file that the police hand off detailing what the police claim the defendant has done. The prosecutor conducts little independent investigation in most cases, and is conditioned to be skeptical of new information that the defense attorney provides. They are thus often in a position of uncertainty regarding key facts in the case, and in light of this uncertainty they default to a presumption that the defendant is culpable. This default is reinforced by the conveyor-belt nature of prosecutors’ job. They have many cases at once, they see the same charges over and over, many of the fact patterns are similar, and the local criminal justice system usually sets a market rate for guilty pleas to common offenses.53 Prosecutors also must often get approval from their bosses for more lenient dispositions, which they may consequently be reluctant to seek. This daily reality creates an adversary case-processing system where prosecutors receive a file, assume it is correct, resolve uncertainties in favor of guilt, and treat discussions with the defense lawyer as a strategic game to secure a guilty plea at around the market rate.

As this discussion makes clear, prosecutors are not adversarial solely because the canonical statements of legal ethics permit them to be. In a world where prosecutor’s official job was only to “seek justice,” there would still likely be prosecutors who prioritize conviction and punishment above all else. But these canonical statements still matter. The adversarial role ethic reinforces prosecutors’ adversary incentives, and excuses the harm they cause. Scrapping the “dual role” model altogether, and replacing it with a pure “seek justice” model, will help to undermine adversary case processing. This is preferable because the “dual role” model causes cognitive dissonance that permits adversarialism to dominate in practice.54 Each of the previously discussed approaches to defining the dual role leaves the adversary half of the duality intact. When adversarialism and justice-seeking are combined in an undifferentiated fashion, prosecutors facing pressure to convict can narrowly interpret the mandate to seek justice.55 When specific rules are imposed that require prosecutors to respect defendants’ rights, these can be treated as legalistic restrictions to be followed to the letter but no further.56 When the prosecution process is divided into adversary and non-adversary stages, the pressure to win cases can overwhelm prosecutors’ adjudicatory role.57 The “dual role” model is unstable. In practice, adversarial win-seeking tends to dominate.

One additional question concerning adversarialism in practice is what, exactly, prosecutors are trying to “win.” For a defense lawyer with a human client, victory is securing whatever the client values—spending less time in prison, preserving immigration status, avoiding a probation term, or whatever else. But a prosecutor does not have a human client, and so at some level they must define victory for themselves. One possible metric of victory is convictions, and much of the literature on prosecution assumes that this is what they seek to maximize.58 Another possible metric is severity of punishment.59 These two goals are in some conflict—for example, one would expect a conviction-focused office to offer generous plea deals, and a punishment-focused office to be stingier.60 Further, accounts of misdemeanor court suggest that prosecutors of low-level crime seek to secure control of defendants over time, through mechanisms like diversion and probation, rather than achieving convictions or specific sentences.61 “Winning” for prosecutors necessarily involves limiting the defendant’s liberty in some way. However more specific definitions of victory can vary between prosecutors, between offices, and between cases.62

 C.     The Shift from Private to Public Prosecution

The duality of American prosecutors’ role is a product of our criminal justice system’s unique history. It reflects our long transition from a system of private prosecutions to a system of government prosecutions. As the government gradually came to dominate the prosecution function, the idea that prosecutors must be more than just partisan lawyers gained political traction. Yet adversarialism still persisted as a legitimate prosecutorial role.

In the early criminal law of England, prosecution was considered a private-law matter. Victims or their families pursued criminal charges against the accused, and they sometimes retained private attorneys for this purpose.63 As Professor John Langbein has shown, the English system of criminal law evolved over the 18th and 19th centuries from a lawyer-free confrontation between accusers and accused into an adversarial, lawyer-dominated system with prosecutors and defense counsel.64 The American colonies adopted similar practices in their early criminal justice systems, relying mostly on private prosecutions that victims organized and funded.65 In the late colonial and early republican periods, states created public prosecution offices with the power to initiate criminal cases, and during the Jacksonian period, these prosecutors became elected officials.66 However private prosecution remained the dominant model in the United States through the first half of the 19th century, with public prosecutors playing only a relatively minor role.67 Even as the administrative prosecution bureaucracy grew in capacity and importance in the late 19th century, public prosecutors were still mostly paid like privately retained attorneys, with their fees contingent on the number of cases that they brought or prevailed in.68 Government prosecutor’s offices were also under-resourced late into the 19th century, and consequently depended on private lawyers financed by victims and their friends to bring many cases.69 Only fairly recently has prosecution come to exclusively rely on a professionalized corps of salaried government attorneys.70 As one scholar has put it, American criminal justice gradually transformed from “a small scale, rurally based system of privately initiated prosecutions with minimal formality” into “a sprawling, large-scale, urban bureaucracy in which criminal prosecutions are initiated on behalf of the state by full-time, publicly paid professionals.”71

This shift from private to public prosecution was met with a parallel shift in the official conception of prosecutors’ role. At the beginning, the system was one of private litigation, with victims personally prosecuting the accused. However, as the public prosecution bureaucracy grew in stature and importance, so did the idea that prosecuting attorneys have duties to act neutrally and avoid miscarriages of justice.72 In 1854, Judge George Sharswood penned an influential essay on the professional ethics of lawyers, asserting that a private lawyer should never participate in prosecuting a defendant if they believe the case is without merit, and further that the prosecutor is “an officer who stands as impartial as a judge.”73 In the late 1800s, several state supreme courts articulated similar visions of the prosecutor’s role, mostly in cases involving challenges to the use of private lawyers as prosecutors.74 The Michigan Supreme Court declared in 1875 that the prosecutor’s job “is one involving a duty of impartiality not altogether unlike that of the judge himself,”75 while the Wisconsin Supreme Court wrote in 1888 that the prosecutor has a duty “to proceed with all fairness in presenting the cause of the state to the jury, and in prosecuting the whole case, even though parts of the case as presented should make in favor of the innocence of the accused.”76 On the other hand, in 1896 the Florida Supreme Court full-throatedly endorsed adversarial prosecution, writing:

The public prosecutor is necessarily a partisan in the case. If he[/she] were compelled to proceed with the same circumspection as the judge and the jury, there would be an end to the conviction of criminals. Zeal in the prosecution of criminal cases is therefore to be commended, and not condemned.77

The California Supreme Court embraced both of these ideas, noting that “[w]e make due allowance for the zeal which is the natural result of such a legal battle as this,” but that the prosecutor has a “sworn duty to see that the defendant has a fair and impartial trial.”78 The emerging dominance of public prosecution thus brought with it an emphasis on prosecutors as neutral figures seeking justice. However it did not occasion a full rejection of prosecutors’ adversarial role. The canonical statement of this developing dual-role model came in the Supreme Court’s 1935 decision Berger v. United States, with its admonitions that while the prosecutor “may strike hard blows, he[/she] is not at liberty to strike foul ones,” and that the government’s interest “is not that it shall win a case, but that justice shall be done.”79

Throughout the 20th century, the dual-role model was repeatedly codified by the ABA. In 1908, the ABA promulgated the first nationwide code of legal ethics, called the “Canons of Ethics.”80 Nearly every state bar association adopted the Canons.81 Canon Five addressed the duties of prosecutors, and stated simply: “The primary duty of a lawyer engaged in public prosecution is not to convict, but to see that justice is done. The suppression of facts or the secreting of witnesses capable of establishing the innocence of the accused is highly reprehensible.”82 This document remained largely unchanged for six decades, until the 1970s and ‘80s saw a renaissance of legal ethics codifications. In that period the ABA enacted several documents that describe in detail the ethical obligations of prosecutors and other lawyers—the Model Code of Professional Responsibility,83 the Standards Relating to the Administration of Criminal Justice,84 and the Model Rules of Professional Conduct (which supplanted the Canons).85 These documents discuss prosecutors’ duties to act impartially and to preserve defendants’ rights, as well as their role as partisan advocates in an adversarial criminal justice system.86 Around the same time period the National District Attorneys Association issued its “National Prosecution Standards,” giving advice to state prosecutors on how to exercise their discretion;87 and the United States Department of Justice promulgated the Principles of Federal Prosecution, providing guidance to federal prosecutors.88 Most of these documents have been revised in subsequent years, and together they codify the official understanding of prosecutors’ ethical responsibilities.

American prosecution has thus undergone an incomplete transition from the model of private partisan lawyering to the model of justice-oriented government lawyering. The same story, in broad outlines, also holds true for other countries that inherited the English system. Australia, New Zealand, Canada, and England have each seen analogous transformations from private to public prosecution, and have each adopted dualist role ethics for prosecutors.89 The story of the United States is unique among these nations, however, because of recent developments in our criminal justice system. In the decades between 1970 and today, the United States has become the most punitive country in the world, and recent empirical work suggests that prosecutors have helped drive this shift.90 American prosecutors appear to have become more aggressive and conviction-focused in the last few decades, just as our professional ethics codes were being updated to emphasize prosecutors’ duties to act neutrally and seek justice.91 Thus partisan prosecution has returned with a vengeance, only this time brought by government prosecutors. And so the pendulum swings.

D.     Contrasting the Inquisitorial Model

In countries with inquisitorial criminal justice systems, such as France and Germany, prosecutors take on a less partisan role than they do in the United States.92 Their professional ideology and incentives emphasize uncovering the truth rather than achieving victories. French and German prosecutors are not supposed to behave as advocates, but instead as neutral agents helping the court to reach the correct factual and legal outcome.93 In service of this goal inquisitorial prosecutors seek out evidence of both innocence and guilt, and sometimes even bring appeals on behalf of convicted defendants.94 In these countries, the criminal process is framed not as a contest between parties, but as a coordinated effort by the government to reach the correct result.95 Inquisitorial prosecutors thus play a role more akin to that of a second judge rather than an advocate arguing for conviction. Indeed, in France prosecutors belong to the judicial branch of the government,96 and the German prosecution service was originally created to counterbalance the perceived overzealousness of judges.97 German and French prosecutors also undergo similar educational training to judges, and the two jobs are seen as comparable career paths.98 This commitment to neutrality provides a sharp contrast to the American prosecutor’s role.

It is important to note, however, that German and French prosecutors operate in a very different institutional context from that in the United States. Significantly, they exercise less discretion in the litigation system than do American prosecutors. Most of inquisitorial prosecutors’ power comes not from their litigation decisions, but from their control over the pretrial investigation process, during which they are supposed to act as neutral truth-seekers.99 In Germany, the principle of mandatory prosecution constrains charging discretion—if sufficient evidence exists to support a charge the prosecutor must bring it, and there are mechanisms for victims to appeal a no-charge decision.100 This rule of mandatory prosecution is designed to make prosecutors into neutral law enforcers, and to prevent them from interposing their own judgments about what cases are worth prosecuting.101 The French and German systems also restrict the use of plea-bargain agreements for serious crimes, with the consequence that a higher percentage of cases go to trial than in the United States.102 This again reduces prosecutors’ discretion because it limits their ability to resolve criminal cases through informal negotiations with the defendant. Furthermore, judges in inquisitorial systems play a much more active role in the presentation of facts at trial than do American judges. Inquisitorial judges question defendants, prosecutors, and witnesses, and order the production and examination of evidence.103 Inquisitorial prosecutors are largely reactive during trial, unlike American prosecutors.104 In sum, while the American adversary system makes prosecutors the decisive legal decision-makers in most cases, the inquisitorial system limits their power.

American scholars have long looked to the inquisitorial model as a source of inspiration for rethinking American criminal justice. The inquisitorial model has provided ammunition to critics of several American practices, including plea bargaining, prosecutorial discretion, and party control over fact-finding.105 Comparisons with France and Germany could also furnish resources for a critique and reinvention of American prosecutors’ role ethics. Indeed, the fact that American prosecutors are more powerful than continental prosecutors would seem to argue in favor of prosecutorial neutrality here rather than there. With only limited charging discretion, plea-bargaining ability, and control over the presentation of evidence, there is only so much harm an adversarial French or German prosecutor can do in the litigation process.106 By contrast, the discretion wielded by American prosecutors makes adversary prosecution here downright dangerous.107

On the other hand, that very discretion makes an ethic of neutrality difficult to realize in the current institutional framework of American prosecution. American prosecutors decide what cases they will pursue, what charges they will bring, and what punishments they will seek.108 It is puzzling to think how a prosecutor could be neutral in the face of such freedom. To adopt a professional ideology that focused on neutrality rather than advocacy, we would need to limit the discretion of American prosecutors and give them guiding principles around which to orient their neutrality.109

 III.     The Case Against Adversary Prosecution

So far this Article has shown that the American system prescribes two different roles for prosecutors, and that the adversarial half of that duality tends to dominate in practice. This Part will build a case against the adversary model of prosecution. It begins by describing the biggest problem with partisan prosecution—that prosecutors are extremely powerful in the United States, and instructing them to behave as adversaries causes much harm. It then turns to three different types of arguments that theorists of legal ethics have developed to justify adversarial legal practice.110 One approach focuses on the relationship of the lawyer to their client, and justifies partisan advocacy on the grounds that it empowers and dignifies the client. This poses obvious problems for prosecutors, who do not have conventional clients, and who make all the major litigation decisions themselves. To defend partisan prosecution, then, is to defend partisanship unmoored from meaningful client control. Two further approaches might furnish such a defense. One is a positivist argument invoking rule-of-law values—the idea that lawyers should not substitute their own moral beliefs for the positive law. The other is an “invisible hand” argument, contending that partisan litigators trying to win cases will incidentally produce essential goods for the legal system, like better evidence and sharper legal analysis.

As will be evident, the argument here is not meant to extend to defense attorneys. This Article argues against prosecutorial adversarialism, not against the adversary criminal justice system in general. Indeed, the theoretical defenses of adversarialism described herein (and critiqued with respect to prosecutors) are more persuasive with respect to defense lawyers.111 This has the consequence that adversarialism should be uneven in our system—defense lawyers should behave as partisans for their clients, while prosecutors should perform a different kind of role.112 The precise contours of that role will be explored in Part IV.

A.     The Prosecutor Wields Immense Power

The standard model of legal ethics in the United States relies on what David Luban has termed the “adversary system excuse.”113 Lawyers are empowered, even professionally obligated, to advance their side in a legal controversy by imposing harms upon others.114 When one objects that a zealous lawyer is acting immorally, the lawyer may reply that their job within the adversary system is to further a specific cause, not to see that justice is done in the aggregate. Whatever one might think of the adversary system excuse in general, it is dangerous in the hands of prosecutors. Prosecutors do not stand in the position of normal attorneys. They are the most powerful actors in the criminal justice system, capable of visiting great harm upon defendants through discretionary choices.115 Further, adversarialism licenses prosecutors to use this power aggressively. If a prosecutor wishes to make a decision that unfairly harms a criminal defendant, they can justify it to themselves and to others by invoking their adversary role. This built-in excuse is dangerous because our justice system is so unbalanced—we have adversary prosecution without adequate adversarial checks.

American criminal justice is essentially a system of negotiated dispositions administered by prosecutors.116 Trials happen only occasionally, in less than ten percent of cases, and judges normally play a limited role in plea bargaining.117 In this system, the prosecutor is the decisive figure. They decide whether to bring a case at all, what crimes to charge, whether to dismiss charges, and how strongly to negotiate for a harsh sentence. During the negotiating process, the prosecutor has unequal bargaining power because they are able to expose the defendant to significantly more punishment if the defendant rejects a deal.118 The prosecutor can also use plea deals to induce defendants to cooperate against one another. And in cases involving lower-level misdemeanor crimes, the prosecutor exercises even more unilateral power because the costs imposed by the process of going through criminal adjudication are commonly more severe than the eventual punishment.119 This is not to say that the prosecutor is omnipotent, or that they can induce all defendants to plead guilty. But for the majority of cases the prosecutor is more like an adjudicator than a party—their discretionary choices decide the outcome by setting the terms of negotiation with the defendant and defense attorney.120 In this context, it is easy to see how prosecutors who take an adversarial approach to their job can cause grave injustices. Prosecutors are not so much adversary advocates as they are adversary adjudicators.

Nor does the prosecutor’s unequal power end with the plea-bargaining process. The trial process is also imbalanced. While defense attorneys provide a check on prosecutors during trial, there are important limitations on their effectiveness. Public-defender services and court-appointed counsel in the United States are not adequately funded, and lack the advantage of a police force to gather their evidence.121 While well-off defendants like O.J. Simpson can use their own wealth to effectively contest their cases at trial, the majority of defendants depend on government-paid lawyers.122 And prosecutors’ power does not fade even after guilt is determined. There are also significant limitations on defendants’ ability to bring post-conviction ineffective-assistance-of-counsel claims, meaning that a prosecutor who takes advantage of an ineffective defense lawyer is unlikely to see the conviction overturned.123 At the sentencing phase, there are a number of tools that empower prosecutors to unilaterally affect the defendant’s punishment, such as departures for substantial assistance in the federal system.124 In post-conviction innocence claims, the prosecutor often has decisive power over the outcome because they control access to potentially exonerating evidence.125 A prosecutor who adopts a winner-take-all approach to their job can thus cause serious harm even in the trial and post-trial phases.

In short, prosecutors’ aggressiveness has severe consequences. They can use their discretion to drive up the number of convictions and the severity of sentences, thereby making the criminal justice system more punitive.126 When prosecutors adopt a conviction mentality they are more likely to hide exculpatory material, ignore contrary evidence, and cause the conviction of innocent people.127 They are also more likely to cause innocent people to remain in prison by denying them access to potentially exonerating evidence.128 Further, adversarial prosecution has distributional consequences. It makes the criminal justice system more unequal by putting a premium on defendants’ ability to hire lawyers who can counteract the prosecution’s aggressive tactics through clever procedural arguments.129 These harms justify rejecting the adversary model of prosecution altogether.

B.     The Prosecutor Has an Abstract Client

One important type of justification for lawyerly adversarialism focuses on lawyers’ ability to empower and dignify their clients by pursuing their clients’ interests as faithful agents. Contemporary theorists of legal ethics have framed this idea in a few ways. Charles Fried has analogized the lawyer’s role to that of a friend.130 In Fried’s formulation, the lawyer’s partisanship to their client’s cause is justified for the same reasons that people are expected to be partial to the interests of their friends and family over the interests of strangers. Abbe Smith and Monroe H. Freedman have constructed an account that focuses on the autonomy and dignity of the client, which is served by their ability to access lawyers who will argue their case zealously.131 Daniel Markovits has developed an account of lawyerly adversarialism that focuses on how it legitimates the legal system to the client.132 According to Markovits, the professional virtues of moral detachment and fidelity to the client’s interests allow a lawyer to translate their client’s claims into a language that the legal system will acknowledge.133 This in turn legitimates the process of adjudication to the client by giving them access to the machinery of law. The common thread in these scholars’ disparate justifications for adversarialism is the value they place on the relationship between lawyer and client, which can be weighed against the harm caused by partisan advocacy.

The problem with applying these kinds of arguments to prosecutors is that prosecutors do not have clients, at least not in any normal sense. The victim and the police play important roles in the criminal process, but they do not have an attorney–client relationship with the prosecutor.134 Indeed, no other actor controls the prosecutor’s decisions in the way a client would, and the prosecutor does not owe an advocate’s duty of loyalty to any person. The prosecutor is effectively a principal in the case—they make the major decisions that would ordinarily be made by an attorney’s client, like what claims to pursue and whether or not to settle.135 Therefore, the arguments for lawyerly adversarialism that rely on dignifying, empowering, or befriending clients all lack force in the prosecution context.

There is, however, one important addendum to this point. It is commonly said that the state (or the sovereign, or the “people”) is the prosecutor’s client. This notion is reflected in case captions—“United States v. Anderson,” “People of the State of California v. Simpson,” and so forth. If one takes this idea seriously, it might perhaps provide a client-based justification for prosecutorial adversarialism.136 The state cannot provide meaningful client control, since it is a legal fiction and not a living, talking entity that can sit next to you in court.137 But if the prosecutor could attribute certain specific goals to “the state,” then perhaps it would be coherent to pursue those goals as an adversary lawyer. There are contexts where such an approach seems at least plausible. Take for example a government lawyer defending a state against a tort lawsuit, or a lawyer representing the United States in a dispute before the World Trade Organization. In these kinds of cases the state is trying to limit its liability or improve its market position, and the government lawyer could pursue litigation with that goal in mind. This approach to representing the state is controversial, but it is also coherent, analogous to representing corporations or other institutional clients with clear goals.138

One could similarly try to argue that the state has an interest in the enforcement of its criminal prohibitions, and that this interest warrants aggressive and adversarial prosecution of accused lawbreakers. However, this is not a necessary or even the most obvious interpretation of the state’s desires in criminal law. Surely, for example, the state does not want prosecution to be so adversarial that innocent people are regularly imprisoned. A canonical principle of American criminal law—the Blackstone principle—holds that conviction of the innocent is worse than acquittal of the guilty.139 And the Constitution itself commits us to a robust set of legal protections for defendants.140 In light of these and other officially endorsed value commitments, it seems dubious to claim that the state as an abstract entity is best served by prosecutors who seek to maximize convictions and punishments.

Indeed, Bruce Green has defended the reverse argument—that the state’s interest is that justice shall be done in a case, not that the criminal law shall be maximally enforced; therefore the prosecutor’s duty to “seek justice” is a product of their client’s goals.141 Various courts and professional-ethics codes have also endorsed versions of this idea.142 The thought is appealing, but it requires some unpacking. How do we know that the state is committed to justice? Certainly it does not follow from the abstract concept of a state—history furnishes many examples of states thoroughly committed to unjust criminal adjudication processes.143 The prosecutor could adopt a normative political theory for the telos of the state, or for that of a “just” or “legitimate” state, by choosing their favorite political philosopher.144 But this would be the lawyer imposing values on the client, not vice versa.145 Probably the most promising way to wring a useful set of values out of the state is to pursue the question in a Dworkinian fashion—to take the existing history of legal and political decisions and infer a set of values from them.146 Criminal statutes are one important source of such values, but there are many others, including court decisions, procedural rules, and abstract maxims such as the Blackstone principle. And the most canonical source for discerning the state’s values is the Constitution, which prosecutors take an oath to protect and defend, and which happens to contain a code of criminal procedure.147 Under this approach the state is an abstraction that stands in for a set of values that have been adopted and memorialized across time, and the prosecutor’s job is to discern these values through moralized interpretation and then to represent them in court.148 Of course, this is no substitute for actual client control—the prosecutor is left with a great deal of discretion to decide what version of justice the state is committed to and how this cashes out in practice. But perhaps it is something short of full delegation.

In any event, theories valorizing the attorney–client relationship cannot justify an adversarial approach to prosecution. While the theory that the prosecutor represents the state may have some purchase, it cannot justify the prosecutor acting as an inverse defense lawyer—maximizing the likelihood of conviction and the severity of punishment while minimizing the scope of procedural rights. Consequently, if we are to justify prosecutors’ adversary role, we must justify it as adversarialism unmoored from client control.

C.     The Prosecutor Makes Moral Choices

A different defense of adversary advocacy focuses not on the lawyer’s relationship to their client, but on the lawyer’s role in upholding the rule of law in a society with pluralistic values. According to this line of argument, the purpose of the law is to settle moral questions. People within a society have disagreements about how that society ought to be governed, and the law is where people hash out these disagreements and come to an answer that settles them.149 The law cannot serve this settling function if the actors who enforce it—lawyers and judges—exercise moral discretion in deciding whether and how to give effect to the law. Therefore, lawyers should not interpose their own moral beliefs between the law and its objects of regulation, because the law itself is supposed to settle moral questions. Doing so reopens the very conflicts that the law was supposed to resolve. Norman Spaulding invokes this positivist idea of law’s purpose to defend the conventional account of lawyers as partisan advocates, and to argue against lawyers substituting their own consciences for those of their clients.150 Spaulding’s argument also applies to prosecutors implementing their own vision of justice.151 If prosecutors decide for themselves what justice requires in a particular case, rather than simply prosecuting that case to the fullest extent, then they substitute their own values for the values enshrined in the law. Doing so replaces the rule of law with the rule of prosecutors. And so the law’s positivist, disagreement-resolving function is served when prosecutors behave like adversary lawyers, abstracting away from their own moral views and enforcing the law vigorously without discretion.

This type of argument provides a motivating ideal for Germany’s principle of mandatory prosecution.152 Mandatory prosecution divests the government of discretion, ensuring that the law is enforced to its letter. The German criminal justice system is designed to accommodate this principle.153 By contrast, the American system is built around the fact of discretionary underenforcement.154 This makes a positivist vision of prosecutors’ role difficult to realize in the United States without significant reforms to the justice system. American criminal laws are written broadly, because the legislators who enact them know that prosecutors will not enforce them to the hilt.155 Indeed, if American prosecutors did adopt a principle of mandatory prosecution, there would be an unsustainable number of criminal cases.156 Further, certain laws in the American system seem designed to invite underenforcement.157 Adopting a positivist vision of the prosecution function would therefore require major changes to the way American criminal law functions. The system would have to enshrine the principle of mandatory prosecution and change the criminal code to make this principle workable, or else adopt as positive law some second-order rule of enforcement that dictates what criminal cases will and will not be pursued.

Even if the system were to make such structural changes to reduce prosecutors’ enforcement discretion, adversarial prosecution is not the only way to pursue positivist rule-of-law values. The positivist critique merely calls for prosecutors to adopt a professional ethic that denies them moral discretion, and there is more than one way to divest prosecutors of such discretion. In Germany, for example, the principle of mandatory prosecution coexists with a professional commitment to neutral prosecution.158 German prosecutors are not supposed to try to “win” cases, but instead to evenhandedly assist the court in finding the legally correct outcome.159 Similarly, W. Bradley Wendel has articulated a vision of positivist legal ethics that eschews adversarial zeal and orients lawyers’ professional role around the principle of “fidelity to law.”160 According to Wendel’s view, private attorneys should not blindly defer to their clients’ desires, but should instead interpret the relevant legal rules and norms in good faith and apply them to the client’s position.161 Wendel’s positivism thus relies not on lawyers’ unreflective commitment to a case, but instead on their commitment to the law itself. He develops a subtle account of lawyers’ duties—they are agents of their clients, but they are not hired guns.

This approach can be extended to prosecutors. Rather than pursuing cases as adversaries seeking victories, prosecutors might instead eliminate their moral discretion by committing themselves to neutrally enforcing the existing legal rules. However, this approach will work only if there is actual law to govern prosecutors’ choices, and not boundless discretion. If the rule of law abhors prosecutorial discretion, it is best to fill in that discretion with decision rules taken from the law itself. It is unnecessary, and indeed harmful, to fill it in with strategic maximization of criminal penalties.

D.     The Prosecutor and the Invisible Hand

In seeking to defend adversarial prosecution, one might also invoke “invisible hand” arguments. Such arguments justify adversarial role morality by focusing on extrinsic benefits that the competitors incidentally produce. In the litigation context, the basic claim of an invisible hand argument is that attorneys who seek to win cases will produce systemic goods that would not be produced by more neutral attorneys.162 Many such arguments can be applied to defense lawyers. For example, a defense attorney who treats their job as trying to achieve the best outcome for their client (as opposed to trying to achieve the most accurate outcome, or the most just outcome from a third-party perspective) will protect the client’s interests more effectively. They will seek to take advantage of the client’s constitutional and other procedural rights to the greatest extent possible. Similarly, defense attorneys’ zeal helps preserve the principle that the government must be put to its burden of proof. If defense attorneys refused to represent clients they believed to be guilty, or declined to make plausible defense arguments, this would undermine values that are important to our legal system.163

Similar invisible-hand arguments might be applied to prosecutors to justify their adoption of partisan role ethics. One might argue that adversary prosecution is a good thing because prosecutors focused on winning cases will be more likely to convict and punish defendants. If one takes the view (as I do) that our system currently produces too much criminal punishment, then this is not a benefit but a harm. On the other hand, if one takes the view that the production of criminal punishment should be subsidized through professional norms, then adversarial prosecution is one way to do so. An interesting corollary to this point is that an adversarial focus specifically on achieving convictions (as opposed to severe sentences) may skew prosecutors towards bringing to trial only cases where the evidence is strong, and offering steep plea discounts or dismissals in cases where the evidence is weaker.164 This might point prosecutors away from hard cases, cases that might have been pursued by prosecutors who saw their role as seeking “justice” rather than victories.165 One might also see such loss-aversion as a harm. For example, there are cases where the victim has trouble testifying—e.g., a child victim or a victim who is mentally ill—and it could be harmful to disincentivize prosecutors from pursuing such cases.166

Another invisible-hand argument connects adversary advocacy to the production of facts and arguments in litigation. That is, to generate the best version of the truth in court you must motivate prosecutors with something other than the search for truth, namely the desire to win their case.167 When both the prosecution and the defense are driven by competitive energies, they will produce evidence and arguments that confirm their respective accounts and undermine their opponents’. Out of this conflict, the idea goes, the most accurate version of the truth will emerge because the various issues will be well framed for adjudication. This argument is given extra force if one assumes that the defense attorney will act as an adversary lawyer regardless of the prosecutor’s role morality.168 If the defense attorney approaches their job as an adversary, while at the same time the prosecutor behaves in a neutral fashion, or seeks to pursue “justice” rather than convictions, then perhaps the defense will walk all over the prosecution.

Admittedly, there is at least something to this argument. When prosecutors are more motivated to vigorously contest defense arguments, it makes sense that the quality of their advocacy would be higher. In at least some cases, this likely facilitates the process of litigation. But two important caveats are in order. First, the desire to win also motivates lawyers to use a variety of tricks to undermine the quest for truth, such as causing strategic delays, undermining credible witnesses, and failing to search for or disclose inconvenient evidence.169 Adversary role ethics thus create two conflicting effects—an impulse to win by seeking the truth, and an impulse to win by undermining the truth.170 The benefits of adversarialism depend on which of these strategies predominates, which is an empirical question that is probably impossible to answer definitively.171 Second, the truth-generating benefit of adverarialism is confined to the litigation context, where each side is represented by a lawyer and there is a neutral judge or jury making the decision as to guilt. Only a small number of cases actually go to trial in the American system, and of these an even smaller number are contested by effective defense counsel. It makes little sense to define prosecutors’ role based on the few cases where the adversary process works effectively, as opposed to the great majority of cases where the prosecutor is the key decision-maker. Still, the argument that adversarialism provides benefits in adjudication has at least some purchase. In considering whether to eschew adversarial prosecution altogether, as this Article advocates, one must weigh the harm it causes against this relative diminution of motivation to generate evidence and arguments. To do so intelligently, one must first define the alternative to adversarialism. To that project we now turn.

IV.     Alternative Models of Prosecution

This Part steps back from critiquing the present system of prosecutors’ ethics, and analyzes the pure forms of several different roles that might be ascribed to prosecutors. First is adversarialism, which instructs the prosecutor to behave like the inverse of a defense attorney, strategically maximizing convictions and punishments. Second is positivism, which instructs the prosecutor to determine the correct application of the law through analogical reasoning and apply it to the facts of the case. Third is value weighing, which instructs the prosecutor to decide through moral deliberations how to weigh conflicting harms and goods in the criminal justice process. These three possible role moralities represent three different ways for prosecutors to make decisions. This Part will show how each of these role moralities approaches the three major types of questions that a prosecutor faces—questions of proof (how much evidence is needed to condemn the accused), process (how effectively the accused can fight back), and punishment (how much harm the guilty should suffer).

This Part then proposes a different dual role for prosecutors—a combination of positivism and value weighing. When an outside source of law imposes requirements on prosecutors, such as a statute or a court decision, their role should be to apply the law to the facts without partisan bias. On the other hand, when prosecutors exercise discretion, it is up to the prosecutor’s office to determine what values should govern. This Part thus invokes the familiar distinction between neutrally following the law and exercising moral discretion—a distinction developed at length through debates in jurisprudence—to structure prosecutors’ ethical role.172 However, crucially, prosecutors should not think of themselves as partisan lawyers in any phase of the criminal justice process—not even at trial, where the harms of partisan prosecution are arguably at their lowest. This alternative dual role makes prosecutor’s offices into agencies that administer criminal statutes. They do so through administrative policymaking decisions that govern how the investigative and plea-bargaining systems operate, and through litigation decisions that determine how prosecutors will conduct themselves in court.

A.     Adversarialism: The Prosecutor as Combatant

The purely adversarial prosecutor is already familiar from the preceding discussion. Their role is best understood through analogy to the role of a conventional private attorney.173 They are partisans assigned the opposite role of the defense attorney, and they must act strategically within the bounds of the law to ensure that defendants are convicted and punished.174 This partisan role brings with it an assumption of moral non-accountability. If an adversary lawyer does something that is conventionally immoral to achieve their goal, it is commonly understood as not only acceptable, but proper.175 To unpack how the purely adversary prosecutor approaches their task, it will be helpful to distinguish between three different kinds of questions that prosecutors must face: proof questions, process questions, and punishment questions.

Proof questions concern the likelihood that the defendant is factually guilty of the charged crime, and the relative willingness of the prosecutor to risk an incorrect conviction or an incorrect acquittal. Once a case has been initiated, the purely adversarial prosecutor sees it as their job to make every effort to ensure the defendant’s conviction. From this perspective, the likelihood of error is not a relevant factor to the adversarial prosecutor after the case has been initiated.176 There could be a number of different processes for deciding to pursue a case in the first place. Prosecutors might initiate any case where the police hand off a file and there is probable cause for an arrest, or they might instead engage in more searching analysis of the evidence before deciding whether to go forward.177 Once the decision to prosecute is made, the prosecutor’s job is then to obtain a conviction and not to revisit the question of guilt or innocence.178 How the case is initiated is exogenous to the adversarial ethic. Indeed, the ethic of lawyerly partisanship is generally agnostic to how any kind of lawyer takes a case. For example, a defense lawyer might take on a client for many reasons—because they want the fee, because they are a public defender assigned the case, because the client is family, or because they genuinely believe in the client’s innocence. Regardless of the reason for taking the case, the lawyer is still a partisan advocate. It is the same for the purely adversarial prosecutor. And this is true in the post-conviction phase as well, where the prosecutor opposes motions to reopen the case and to reexamine old evidence.179 Much like a private lawyer, the purely adversarial prosecutor’s job is not to uncover the truth but to win the case.

Process questions concern the ability of the defendant to take advantage of their procedural rights. Such rights include the right to an attorney,180 the right to a speedy trial before an impartial jury,181 the right to be free of unlawful searches,182 and the right to discovery of exculpatory evidence.183 The purely adversarial prosecutor views such procedural rights as impediments to be overcome in trying to achieve convictions and punishments. This is not to say that the adversarial prosecutor intentionally violates these rights, only that they give these rights the narrowest possible scope that is consistent with obtaining a conviction and having it preserved after appeals and habeas corpus challenges. For example, the purely adversarial prosecutor will turn over exculpatory evidence to the defendant as required by statute and constitutional holdings, but will not turn over more evidence than required unless they believe that doing so will facilitate the prosecution. And when the scope of the defendant’s rights is debated in court, the purely adversarial prosecutor will take the position that they should be narrowly interpreted. Or, failing that, the prosecutor will argue that the relevant right has no remedy.184 In the American system, one effective way for an adversarial prosecutor to circumvent the defendant’s rights is to bring a number of serious charges against the defendant, and then to negotiate a plea-bargain agreement that reduces their exposure to prison time.185 This strategy, combined with waivers of the defendant’s rights to appeal or to bring post-conviction habeas petitions, allows the prosecutor to convict defendants while limiting the risk posed by their procedural rights.186

Punishment questions concern the degree of harm that should be visited upon the defendant for the crime they have committed. Prosecutors face punishment questions at several different stages in the process—when they make initial charging decisions, when they engage in plea-bargain negotiations, and when they participate in formal sentencing proceedings before a judge or jury. For the purely adversarial prosecutor, more punishment is better. Of course, the punishment of a single defendant is not the only variable in a given case, and there are important tradeoffs that must be dealt with. For example, the plea-bargaining process creates a tradeoff between convictions and punishments, since offers of lighter punishment make it more attractive for defendants to plead guilty.187 And the use of cooperation agreements creates a tradeoff between punishment for one defendant and punishment for another. There is therefore some ambiguity in what a purely adversarial prosecutor would seek to maximize in any given case, which stems from the lack of a real client to direct the prosecutor’s goals in litigation. Nonetheless, all else equal the adversarial prosecutor seeks to impose punishment.

This is a dystopian vision of prosecutors’ ethics, and it would create a nightmare system because of the vast power prosecutors wield.188 Pure adversarialism requires prosecutors to approach their jobs amorally, and to seek convictions and punishments without regard for whether they are deserved. It relies on an absence of ethical reflection by prosecutors, who depend upon the adversary-system excuse to absolve them of blame for their actions. Adversarialism is not about values, it is about winning. The American system does not embrace purely adversarial prosecution, because we also adopt the principle that a prosecutor has duties to “seek justice.” Indeed, prosecutors themselves would very likely oppose eliminating the duty to “seek justice.”189 As noted previously, however, in practice our system sometimes uncomfortably resembles the one described here.190 This is not a model we should embrace, but one we should define our practice against.191

 B.     Positivism: The Prosecutor as Second Judge

The positivist prosecutor finds their closest analogue in the role of the judge. Judges in our legal system are professionally acculturated to be indifferent to the outcome of a case.192 They are told to interpret the law through textual and analogical reasoning and then apply their interpretation to the facts before them.193 The positivist prosecutor is similarly indifferent to the outcomes of cases, not caring whether a defendant is convicted or acquitted but simply seeking the correct legal result.194 In France and Germany this comparison is made explicit, as prosecutors are officially understood to be playing a judicial or quasi-judicial role in the criminal process.195 Much like adversarialism, positivism divests prosecutors of personal moral accountability for their actions. However, while adversarialism replaces moral reasoning with strategic reasoning, positivism replaces it with legal reasoning. The prosecutor is simply doing what the system requires of them. They are a Weberian bureaucrat, a functionary assigned a specific task whose role is not to deliberate on the morality of that task.196 One famous illustration of this role ethic is Charles-Henri Sanson, the executioner of Paris.197 Between 1778 and 1795, Sanson conducted beheadings on behalf of the Bourbons, the Commune, the Girondins, and the Jacobins.198 Throughout these years, Sanson was a model of professional detachment from the ends that his blade served.199 So it might be with the positivist prosecutor—their job is to follow the rules adopted by the system, not to question those rules.

The positivist prosecutor approaches proof questions wanting to generate the correct answer. Unlike the adversarial prosecutor, they do not adopt a default posture of trying to convince a jury that the defendant is guilty. Rather, the question of guilt or innocence is a problem that the positivist prosecutor is trying to solve. Their approach to this problem will vary depending on the procedural context. During the pre-trial phase—investigating, charging, and plea bargaining—the prosecutor is effectively an inquisitorial judge.200 They will search out evidence of the defendant’s guilt or innocence, and try to dispassionately determine whether the defendant committed the crime. When trying a case before a judge or jury, the positivist prosecutor’s role changes somewhat because they are no longer the main decider on the question of guilt. As in inquisitorial systems, this will mean presenting all of the available evidence to the decision-maker to facilitate their coming to the correct conclusion.201 In the post-conviction context, the positivist prosecutor is willing to reopen cases and will dispassionately entertain the possibility that a person might have been wrongfully convicted. They may even devote a special unit to the reinvestigation of old convictions, as do some American district attorneys’ offices.202 In short, positivist prosecution is an extension of adjudication—its goal is to objectively apply law to facts, even if this means confessing error.

The positivist prosecutor also approaches procedural questions from the perspective of a neutral judge. They do not seek to diminish the scope of procedural rights, but instead to determine through interpretive and analogical reasoning the proper reach of those rights. For example, when deciding whether to use evidence gathered in a potentially unlawful search, the positivist prosecutor will interrogate the police about the legitimacy of their warrants and the legal grounds for the search. If the positivist prosecutor determines that the search violated the Fourth Amendment, they will decline to use its results even if a judge might have admitted them.203 Similarly, when the positivist prosecutor is determining what evidence to disclose to the defendant before trial they will look at all of the relevant rules—statutes, judicial decisions, and professional-ethics opinions—and come to a legalistic conclusion about what must be turned over.204 This commitment to finding the legally correct answer will mean that the prosecutor confesses error freely and frequently, either at trial, on appeal, or during post-conviction habeas proceedings.

The positivist prosecutor approaches punishment questions indifferent to the ultimate fate of the defendant. They seek neither to maximize nor to minimize punishment, but simply to ensure that the degree of punishment is legally correct (to the extent that there is any law to apply). In the charging context, this means the system will have to establish some sort of rule of decision that dictates what crimes to charge in different types of cases. The German principle of mandatory prosecution is one such rule, but not the only possible rule.205 Then when making charging decisions, the prosecutor applies this rule to the facts before them and tries to resolve any indeterminacy through analogical reasoning. They do not go beyond these rules by using equitable charging discretion to help sympathetic defendants or hammer unsympathetic ones.206 Similarly, at the sentencing phase, the positivist prosecutor is indifferent to the normative question of how much punishment the defendant deserves. However, if there is a legal question to resolve, for example a complex sentencing guidelines calculation of the type required by the United States Sentencing Guidelines, the positivist prosecutor will approach this question as a second judge and apply the guidelines to the facts.207 Thus we see that, at a certain level, positivism requires an external source of rules for prosecutors to interpret. If there is no rule governing a decision, then the positivist prosecutor cannot perform their prescribed role.

Prosecutorial positivism furthers rule-of-law values. It eliminates human discretion from the prosecution process, replacing it with transparent bodies of rules. The German prosecutor is the exemplar of this vision, due to both their professional ideology of neutrality and their relative lack of discretion in the criminal adjudication process. The basic problem with adopting an ethic of positivism for American prosecutors, however, is that they wield an enormous amount of discretion.208 In the American system, there is no legally determinate answer to questions like “What charges should be brought in this case?” or “What type of plea deal should we offer this defendant?” Prosecutors have discretion over these and other crucial decisions, and so there is no external source of law to orient an ethic of positivism. In this regard, our system is lawless. And if the law is indeterminate, or if there simply is no law, then an ethic of positivism will not work. To get to a system of prosecutorial positivism, then, we would have to establish prosecution guidelines or some other rule framework that would limit prosecutors’ discretion.

C.     Value Weighing: The Prosecutor as “Minister of Justice”

A third way for prosecutors to approach their decisions is by behaving as morally intentional actors. Prosecutors can engage in deliberation over what values they wish to further in the criminal justice system, and exercise their discretion accordingly. They might thus adopt a role morality similar to what the 19th-century legal thinker David Hoffman advocated for all lawyers—that they be guided by their own consciences, and not by externally imposed goals.209 It is commonly said in the academic literature and in professional-ethics codes that the prosecutor is a “minister of justice.”210 However this role is not precisely defined. One possible way to interpret the phrase is as embodying this value-weighing role. When a prosecutor acts as a minister of justice in this sense, they consider the different moral principles that argue for and against a certain course of action and choose the principle they believe should be followed. This is analogous to the way that judges approach sentencing decisions in discretionary sentencing regimes. At sentencing, a judge does not neutrally apply the law, since there is no law to apply, but instead exercises moral discretion in determining how much punishment a defendant deserves consistent with the larger goals of the criminal justice system.211 Thus, as a minister of justice, a prosecutor is neither an adversary nor a neutral actor. Adversarialism and positivism each allow the prosecutor to avoid moral responsibility for their value choices. Both require a certain kind of inauthenticity from the prosecutor, who is understood not to be ultimately endorsing the decisions they make, but instead to be relying on the adversary system excuse or the rule of law to justify their actions. The minister-of-justice role, by contrast, requires that prosecutor’s offices take value positions for which they can then be made morally and politically answerable. This makes them not quite neutrals, and not quite adversaries. Unlike the positivist prosecutor, the minister of justice is not indifferent to a case’s outcome. But unlike the adversarial prosecutor, the minister of justice seeks a certain punishment because their office believes it is the morally right outcome, not because it represents a victory.

A prosecutor can select from between a number of different possible values for any particular discretionary choice they face. When deciding what cases to prosecute, they might consider the harmfulness of different types of criminal activity, the characteristics of offenders, and the risk of unfair or discriminatory treatment of defendants. When deciding how to conduct an investigation, they can consider the constitutional rights of the defendants as well as the harm caused by those defendants’ activities. The ultimate justification for prosecutors to adopt the role of ministers of justice is by reference to the values that they choose to further in such situations. If prosecutors can be expected to advance good values (from the perspective of the system planner), then their moral discretion is beneficial. If they can be expected to advance the wrong values, then their discretion should be cabined.

On questions concerning proof of guilt or innocence, the prosecutor-as-minister of justice, must make a decision about how to weigh the possibility of incorrect convictions against the possibility of incorrect acquittals. This will help determine the threshold for pursuing a case.212 For example, the United States Department of Justice imposes a requirement on its prosecutors that they may only pursue a case if they believe that the defendant would probably be convicted by an unbiased trier of fact.213 Similarly, the ABA’s “Criminal Justice Standards” call on prosecutors to pursue charges only if they believe the available evidence is enough “to support conviction beyond a reasonable doubt.”214 Concern for the possibility of wrongful conviction can be incorporated into other stages of the prosecution process as well. For example, the prosecutor’s office could require disclosure of exculpatory evidence to a grand jury to enhance the grand jury’s screening function.215 Prosecutors could also reevaluate a case after it has been initiated, for instance by holding periodic formal reviews. And if a prosecution office decided it wanted to remedy wrongful convictions, it could devote resources to reinvestigating old cases through a conviction integrity unit. For example, the District Attorney of Dallas County, Texas created a conviction integrity unit in 2007 that subsequently exonerated several dozen convicts.216

On questions of procedure, a minister of justice must consider how much procedural protection criminal defendants should enjoy, and how much to sacrifice in terms of convictions to preserve those protections. This means weighing outcome values against process values—the importance of convictions versus the importance of letting defendants meaningfully contest their cases. One example of such weighing is the Department of Justice’s internal rules restricting federal prosecutors’ ability to subpoena defense attorneys to only a very limited set of circumstances.217 The United States Attorneys’ Manual states that this restriction exists because such subpoenas undermine the attorney–client relationship.218 Another example is the decision of some state prosecutor’s offices to adopt “open file” discovery policies that give defendants access to evidence prosecutors are not required to turn over.219 Such policies give defendants greater opportunity to put on a defense and more effectively attack the charges against them. And prosecutors can decide to voluntarily concede procedural points in litigation, such as statutory time bars, if doing so will protect important rights that defendants would otherwise lose.220 On the other hand, a value-weighing prosecutor may also act on the belief that the values served by convicting a certain defendant are more important than the values served by preserving a certain procedural principle. Ultimately it is the prosecutor’s office that does the weighing.

On questions of punishment, the prosecutor-as-minister of justice must come to moral decisions about which crimes and defendants are worthy of harshness or leniency. At the extreme, this can mean adopting a policy that the office will not pursue charges for certain offenses. For example, the District Attorney of the Bronx has announced that he will not prosecute people stopped by the police for trespassing on public-housing projects.221 Similarly, the Obama Department of Justice established a policy of not prosecuting federal marijuana offenses in states that legalized and regulated marijuana, so long as the offenders were in compliance with state law.222 It can also mean restricting whether and when prosecutors will take advantage of discretionary tools that they can use to determine sentence lengths. In the federal system, for instance, prosecutors can impose sentence enhancements under 21 U.S.C. § 851 on some defendants with prior convictions. The Obama Department of Justice issued guidance restricting the use of such enhancements, instructing prosecutors to impose them only in cases that called for “severe sanctions.”223 Prosecutors may also have different views about the propriety of certain criminal punishments, such as the death penalty. Indeed, the policies of local district attorneys seem to be the most important factor in determining where in the United States the death penalty is currently used.224

In making these and other choices concerning litigation practice and office policy, prosecutors-as-ministers of justice are engaged in moral decision-making of the kind expected of legislators and other government actors. They choose the values that they will commit to and adopt practices that further those values. This way of constructing prosecutors’ role creates a possible danger. Prosecutors could genuinely decide that it is morally preferable to behave like adversarial lawyers. That is, they might come to the view that the only important moral values are those served by securing convictions and punishments and that values like due process, leniency, and limiting wrongful convictions are immaterial. If a “value weighing” prosecutor was to adopt such values they would indeed resemble an adversarial prosecutor, but with one crucial difference—they would not have recourse to the adversary-system excuse. If prosecutors are going to behave like adversary lawyers, they must endorse the values that that means advancing. As Deborah Rhode has noted, “the critical question is not by what right do lawyers impose their views, but by what right do they evade the responsibility of all individuals to evaluate the normative implications of their acts?”225 A minister of justice who furthers only conviction-focused values is accountable for that choice, both discursively as a moral agent and more concretely as a government actor who is elected or appointed by others.

One could also address this danger by saying more about the moral discretion prosecutors should exercise when they weigh competing values. Much like adversarialism, the ethic of value weighing only comes into play when prosecutors have discretion. But the scope of moral discretion that value-weighing prosecutors wield could be cabined depending on how one sees their role. At one extreme, one might view prosecutors as totally free to impose their own value choices, so long as those choices are not unlawful. Here an analogy can be drawn to Edmund Burke’s speech to the electors at Bristol, in which he argued that as a legislator he is free to pursue his own vision for the good of the nation.226 At the other extreme, one might see prosecutors as moral interpreters of the existing legal and political system akin to Dworkin’s Judge Hercules.227 On such a view prosecutors are not supposed to exercise total moral freedom, but are instead supposed to see their choices as constrained by the principles that are implicit in our current legal order. The sources for such principles would include the Constitution (on which prosecutors take an oath), statutes, judicial decisions, and canonical maxims of law such as the Blackstone principle.228 The best view of prosecutors’ degree of moral freedom likely lies between these extremes, because the prosecutor stands in two different roles that carry with them two different logics of legitimacy. In the American system, they are almost entirely elected officials (or directly appointed by elected officials), and this argues for a high degree of moral discretion. If prosecutors are held accountable to the public through democratic feedback loops, they can be seen as exercising moral agency on behalf of the public. On the other hand, prosecutors also function as enforcers of criminal laws, adjudicators of specific cases, and representatives of the state. In these roles prosecutors are tasked with implementing the law, not with making it. These features argue for a more interpretive approach to value weighing, one that takes the system’s established value commitments as given rather than as revisable through enforcement decisions. Prosecutors stand somewhere between legislators and judges229—between authors of the system and interpreters of the system.230 They are not fully free to pursue only conviction-oriented values, even if they believe those to be the right values, but are at least partly constrained by their role-based commitment to the Constitution and to other norms of criminal law.

D.     A Different Dual Role: Prosecutors as Agencies

The discussion so far has focused on the pure forms of these three visions of prosecution. The real challenge of professional ethics, however, is not in defining pure forms, but in navigating the conflicting demands of one’s role. Judges, lawyers, doctors, journalists, interpreters, professional wrestlers, academics, architects, priests, and any number of other professionals face clashing imperatives that require mediating the conflict between different ethical obligations. It is no different for prosecutors. In the United States, prosecutors’ professional ethics are currently defined by the line between adversarialism and “seeking justice.”231 They should instead be defined by the line between positivism and value weighing. Where an external rule dictates how a prosecutor should act, they should neutrally implement that rule. Where no such rule exists, and the prosecutor’s office must decide for itself what to do, the office should weigh competing values in coming to an answer. Prosecutors should weigh these values differently in the different phases of the prosecution function—investigation, plea bargaining, trial, and post-trial—because they exercise different degrees of power in each. But they should not approach any aspect of their role, even trial practice, as adversary lawyers.

The ethics of positivism and value weighing complement each other well, because each one’s paradigm case is the other’s limiting case. Positivism functions most readily in situations where an external source of law gives prosecutors a rule that they can dispassionately implement. Value weighing functions most readily in situations where prosecutors face a discretionary choice and must come up with a reason to select one option over the other. For example, prosecutors have an obligation as neutral implementers of the law to turn over any evidence covered by the Supreme Court’s decision in Brady v. Maryland and by criminal discovery statutes.232 But, above the floor set by such external laws, prosecutors have discretion over what evidence to turn over and when. They should exercise this discretion thoughtfully by weighing the importance of procedural fairness against other principles like the protection of witnesses and the values served by conviction. And so prosecutors’ value choices can fill in the interstices where external law has no answer. This dichotomy between rule-boundedness and discretion also has a corollary—the dichotomy between values imposed from the outside and values generated from within. The positivist prosecutor enforces value judgments made by other actors (e.g., courts or legislatures) that those other actors have codified and imposed from without. The value-weighing prosecutor, by contrast, enforces value judgments that have been generated within their own office.233 Indeed, this dichotomy can even be extended to within the prosecution office. If an office is hierarchically managed, the job of the line prosecutor is to implement value judgments made by their bosses. And so the “dual role” of positivism and value weighing can actually be understood as one unified role logic, where the major determining variable is which actor exercises moral discretion—the prosecutor or another actor.

This alternative dual role also creates a useful framework for thinking about when prosecutors should have moral discretion in our system, and when they should instead implement choices made by other actors.234 One variable to consider within this framework is which actors are likely to care about which values. For example, one criticism of prosecutors is that they do not often use equitable discretion for the benefit of less culpable defendants.235 If true, that provides an argument for restricting prosecutors’ discretion over charging and plea bargaining, and for giving other actors power to monitor prosecution decisions.236 The desire to avoid arbitrariness is another potential concern, since enforcement discretion can lead to discriminatory treatment and undermine rule-of-law values.237 And the dominance of prosecutors in our system is also troubling for separation-of-powers reasons—perhaps ethics rules for the grand inquisitor are not adequate protection from tyranny. Such concerns might justify prosecution guidelines promulgated by a legislature or commission, or a system where judges supervise charging, pleas, and dismissals, creating something like a common law of prosecution.238 On the other hand, one possible benefit of giving prosecutors discretion is that they are close to individual cases, and so are well positioned to sand off the hard edges of the criminal law through non-enforcement decisions.239 By contrast, legislatures tend to think about criminal law in abstract terms, which leads them to take symbolic stands and enact laws that are insensitive to the complexity of individual cases.240 And judges, while closer to individual cases, are supposed to adopt a more neutral posture that stands in tension with supervising charging and plea-bargaining decisions.241 Thus, the question of who should decide what values are adopted—whether it be prosecutors or other actors—requires considering a number of different institutional tradeoffs.

Why not add adversarialism into this equation, and make it a triple role ethic? One argument for doing so can be found in the adversary stage theory. According to this theory, prosecutors should establish a division of ethical labor within their offices.242 During the investigation of a case and the pre-trial proceedings, prosecutors should act neutrally. But during the trial phase, they should be purely adversarial.243 The issue here is whether it is better to think of trial prosecutors as adversaries or as value-weighers. The adversary stage theory’s basic point makes intuitive sense—the prosecutor’s role is different at trial, because trials involve presenting evidence to third parties rather than coming to unilateral decisions. In particular, prosecutors can safely be more aggressive at trial because external checks limit the potential for harm.

Nonetheless, it is better to understand this increased aggressiveness within the value-weighing framework, rather than as a shift to adversarialism. There are certain values served at trial by the aggressive presentation of evidence. These include, most significantly, the values connected with ensuring that people who commit crimes are punished. At the same time, it is dangerous for prosecutors to ignore other values that may be implicated in a trial. If the defendant’s lawyer is incompetent, for example, the prosecutor should not take advantage of this fact as readily as a private lawyer would.244 Additionally, if new evidence appears that cuts against the defendant’s guilt, or otherwise helps the defense case, the prosecutor should take this into account, quickly disclose it, and thoughtfully consider whether pursuing the case is still the right decision.245 These kinds of value judgments are unavailable to the purely adversarial prosecutor, who treats trial advocacy as a win-loss proposition. Thus, even in the trial phase, the prosecutor should understand their role as weighing competing values, although the relative importance of those values may differ because of the context.

This distinction between prosecutors’ trial role and their non-trial role invites an analogy to administrative agencies.246 Much like prosecutors, agencies engage in both general policymaking and case-specific adjudication. These two aspects of agencies’ role function as complementary methods for administering the law. Take for example the Internal Revenue Service. In administering the tax code, the IRS creates and administers a set of procedures for people to voluntarily file tax returns, and it audits some fraction of those returns.247 If the IRS informs an audited taxpayer that they owe more money, that taxpayer has recourse to IRS procedures requiring explanation for the assessment of tax liability and establishing an internal agency appeals process.248 Once these internal IRS procedures are exhausted, a tax dispute moves into the court system. Similarly, prosecutors adopting the framework advocated here could adopt different sets of policies for the internal adjudication process and the external litigation process. For the great majority of cases that are resolved informally through plea bargaining, the prosecutor’s office develops a set of internal policies much like the IRS’s voluntary-compliance regime.249 For the small number of cases that go to trial, the prosecutor’s office establishes a different set of procedures that take into account the presence of judges and defense lawyers in formal litigation. The importance that prosecutors place on protecting procedural values should be higher in the plea-bargaining system, simply because in the plea-bargaining system prosecutors have unilateral control over the process, while in the trial system other actors check their power. By analogy to administrative agencies like the IRS, prosecutors could also be made to promulgate public regulations that would be subject to notice-and-comment procedures, or perhaps even to revision by outside actors.250 Doing so would facilitate internal and external deliberation about how prosecutors should administer the criminal justice system, and about what values they should seek to further.

V.     Non-Adversarial Prosecution in Practice

The argument in this Article is directed primarily at the organized bar. I am trying to articulate a vision of prosecutorial ethics that is non-adversarial, and to develop a vocabulary that state bar associations could adopt in embracing that vision. The upshot of the discussion so far has been that we should abandon prosecutors’ adversarial role, and separate out their minister-of-justice role into two distinguishable but connected ethics: positivism and value weighing. This thesis could be implemented by amending state professional-ethics codes to say something like the following: “The prosecutor has two roles, to neutrally enforce the existing law, and to carefully deliberate over what justice requires in situations where the prosecutor exercises discretion.” Rewriting the ethics codes in this way would be a consequential change. These codes comprise our legal system’s official statements about lawyers’ ethical imperatives. They are taught to prosecutors and to law students who will become prosecutors. They are used to criticize prosecutors who fail to live up to their requirements. They also help set the agenda for the profession’s institutional evolution, over decades and over generations.

Nonetheless, this project cannot succeed if it begins and ends with changing state ethics codes.251 It is also vital to consider how a non-adversary professional ideal would interact with real-world prosecution institutions. This Part puts the proposed ethic of non-adversarial prosecution into conversation with the academic literature on the institutional prosecutor’s office. It considers how prosecutor’s offices interact with the outside bodies that impose mandates on them—courts, legislatures, and the organized bar. It shows how these external actors can regulate prosecutors both by directly constraining them and by engaging in dialogic policymaking with them. It then looks at the major factors that are blamed for prosecutorial overzealousness—e.g., elections, conviction rate statistics, and office culture—and considers how these might be mitigated or overcome.

A.     Determining Prosecution Values from the Outside

Prosecutors are constrained by a number of outside institutions. They must follow rules laid down by legislatures, courts, and professional bar associations. Each of these actors regulates prosecutors in distinct ways. Legislatures enact substantive criminal laws that define what prosecutors can charge, as well as procedural laws that govern how the justice system functions. Courts directly supervise prosecutors at trial, and also interpret and enforce constitutional and statutory rules. Professional bar associations enact and enforce ethical rules, including Model Rule 3.8, which effectively contains a code of criminal procedure.252 The framework advocated in this Article requires prosecutors to neutrally implement the rules imposed by these outside bodies. It further calls on prosecutor’s offices to form their own value judgments when these outside bodies leave them with discretion. But there is also a type of prosecutorial decision-making that lies between external constraint and internal choice: Prosecutors can make policies in dialogue with outside institutions. They can do so by adopting value choices made by the legislature, the judiciary, or the state bar, and incorporating these into their own internal policy decisions. Such interbranch dialogue straddles the distinctions laid out in the previous Part between positivism and value selection, and between constraint and discretion.253 It permits moral discretion to be shared across multiple institutions, rather than dividing the criminal justice system into the actors that create rules and the actors that implement them.

One way that external institutions can engage in policy dialogue with prosecutors is by enacting a set of principles and delegating to prosecutor’s offices the task of filling in the details. Such delegation gives prosecutors a mandate to guide their choices but also leaves them with some degree of discretion in deciding upon the specifics. For example, the judiciary could announce a decision rule in a constitutional criminal procedure case that is explicitly underinclusive, and instruct prosecutors to implement the relevant constitutional norm above the court-imposed floor.254 That is to say, a court could announce that it will enforce a constitutional right only up to level X, and then state that prosecutors should (or perhaps must) enforce that right upon themselves beyond level X. For example, consider the Double Jeopardy Clause. The Supreme Court has interpreted the Double Jeopardy Clause as containing a dual-sovereignty exception, so that duplicative state and federal prosecutions are permitted.255 This arguably underenforces the relevant constitutional norm, because it allows duplicative prosecutions. However, the Department of Justice’s “Petite Policy” restricts second prosecutions to only a limited set of cases where there is a substantial federal interest.256 One could imagine a similar internal policy enforcing the penumbra of the Equal Protection Clause, which the judiciary restricted to cover only intentional discrimination in Washington v. Davis, but which could be expanded by prosecutors to combat disparate impact discrimination in the criminal justice system.257

Such guidance can also come from non-judicial bodies. For example, a legislature could enact a statute establishing certain principles for prosecutorial decision-making. This statute could, for instance, require prosecutors to significantly reduce the prosecution of non-violent offenders, or set forth certain criteria for charging juveniles, and then permit prosecutors to promulgate regulations that bring these broad mandates into practice.258 Prosecutors could also use sentencing commissions or other government bodies to create such policies. For example, in Washington, the sentencing commission enacted a set of nonbinding prosecution guidelines.259 The state bar could be another source of institutional support for such policymaking. States could create special prosecutor’s bar associations, or perhaps prosecution guideline commissions, to promulgate regulations or other guidance for prosecutors.260 And prosecution regulations could even be adopted with notice-and-comment procedures involving broad public participation.261 For example, in the 1990s and early 2000s, the head prosecutor of Kitsap County, Washington (a midsized county outside of Seattle) adopted charging-policy guidelines with input from the public in the form of a “citizen’s committee” of local residents and representatives of criminal justice institutions.262

Another, somewhat less idealized mechanism of dialogue exists where a risk of external regulations motivates prosecutors to adopt their own internal regulations. This form of dialogue relies on threats rather than cooperation. An outside institution will threaten to constrain prosecutors directly in the hope of causing policy changes within the office. To take one recent example, in the wake of the Department of Justice’s Brady violations during the prosecution of former Senator Ted Stevens, several members of Congress attempted to enact legislation that would have imposed new disclosure requirements on federal prosecutors.263 The Department of Justice responded to this threat by adopting internal rules that expanded disclosure requirements, as well as by creating new annual Brady trainings and establishing Brady coordinators in every office.264 As another example, a professional association of prosecutors in Florida created guidelines for the application of a habitual-offender law after a legislative report claimed the law was being administered in an arbitrary and racially biased fashion.265 Further, in the 1990s, the New Jersey Supreme Court declared that certain laws granting prosecutors unilateral sentencing discretion were unconstitutional.266 However, rather than striking these laws down, the court ruled that they would be upheld only if the Attorney General of New Jersey created guidelines that determined how prosecutors would use their sentencing discretion.267 The Attorney General promulgated the called-for guidelines, and a few years later the court held that there was too much county-by-county variation and ordered a new set of guidelines.268 The Attorney General complied again, and so the court caused two major shifts in state prosecutors’ internal regulatory structure. This mechanism of dialogue-by-threat relies on prosecutors’ desire to preserve their discretion. Indeed, prosecutors even adopt prophylactic measures to protect their discretion without an immediate external threat. A number of the Department of Justice’s internal self-constraints—for example its rules restricting RICO prosecutions, public-corruption prosecutions, and the subpoenaing of journalists—can be interpreted as prophylactic efforts to head off judicial or legislative rules.269

This idea of interbranch dialogue adds a functional dimension to the ethical framework advocated in Part IV. If prosecutors adopted the non-adversary approach this Article proposes, their actual decisions could fall between the pure forms of “positivism” and “value weighing.” The rules imposed by courts, legislatures, and state bar associations can influence prosecutors’ internal value choices. And conversely, in situations where externally imposed rules contain indeterminate areas, prosecutor’s own value choices can help to fill in the gaps. Outside actors can also respond to prosecutors’ decisions, and vice versa, creating iterative policy feedback loops. Such inter-branch dialogue permits other institutions to engage collaboratively (or conflictually) with prosecution offices in determining how the justice system functions and what values it serves.

B.     Overcoming Adversary Incentives

Prosecutors face a number of different professional incentives to behave like adversary conviction-and-punishment-maximizers. These incentives could interfere with efforts to implement the role ethic endorsed in this Article. Indeed, if prosecutors outwardly adopted an ethic of non-adversarialism but inwardly remained committedly adversarial, this might be uniquely harmful because it could cause outside actors like judges to be inadequately skeptical of prosecutors’ actions.270 Thus, it is important to consider the various incentives that may push prosecutors to prioritize obtaining convictions and punishments over other goals. Three of these shall be discussed here: electoral politics, offices’ focus on conviction rates, and a general culture of competition.

It is conventional wisdom that the fact that most head prosecutors in the United States are elected makes them more punitive and more focused on conviction statistics.271 If this is accurate, then it creates a significant impediment to the vision of prosecutorial role ethics advocated in this Article. Depending on the size of the effect, an electoral incentive to focus on convictions would undermine prosecutors’ ability to focus on other values. And if electorates really do vote prosecutors out of office for being insufficiently punitive, or for losing cases, then district attorneys who announce a commitment to non-adversarial role ethics will frequently be replaced by more conviction-oriented competitors.

Recent empirical scholarship gives some reason to doubt the existence of this electoral incentive. Several studies suggest that prosecutorial elections are low-information affairs in which incumbency is an overwhelming advantage (indeed, many are uncontested), and in which conviction rates and other metrics of adversarial success bear little apparent relationship to electoral success.272 Importantly, this is a different question from whether prosecutors perceive that their conviction statistics matter for reelection. There is some empirical evidence that prosecutors change their behavior in election years, at least at the margins.273 While politics do play an important role in high-profile cases, such as murders that generate significant media coverage,274 such electoral and other political pressures do not generally intrude on specific prosecution decisions, which are usually low-visibility and attract little attention.275 Such pressures can thus likely be managed within a non-adversarial prosecution framework that treats convicting defendants as one goal among many, rather than as the overriding objective. Indeed, the politics of prosecution may not be as one-dimensionally punitive as is often assumed. In the last few election cycles a number of reform-minded district attorneys in cities like Philadelphia, Houston, and Chicago have won high-profile races with promises to liberalize the criminal justice system.276

The structural incentives established by prosecutor’s offices, including the use of conviction rates as a metric for professional advancement, can also interfere with efforts to make prosecution non-adversarial.277 But the use of conviction rate as an evaluative metric is not a necessary or inevitable feature of American prosecution. Head prosecutors committed to a non-adversarial understanding of their role could reform the way they evaluate their line prosecutors. A number of scholars have suggested alternative evaluative metrics that focus on factors like the protection of defendants’ rights, absence of misconduct and error, and the input of judges, defense attorneys, and other criminal justice stakeholders.278

Beyond establishing methods of evaluation, head prosecutors also set the tone for an office’s professional identity by training line prosecutors, determining what lawyers are hired in the first place, and establishing expectations and monitoring procedures.279 Such organizational decisions determine the norms of the prosecutor’s office and consequently also determine how prosecutors relate to one another, to outside actors, and to the people they prosecute.280

Different prosecution offices also have different cultural norms, which can vary from aggressive and adversarial to moderate and managerial.281 There are prosecution offices, in both state and federal systems, with relatively balanced professional cultures that would be more compatible with the non-adversarial approach advocated here.282 However, there are also offices that throw parties after securing death sentences, award conviction bonuses, and establish competitions over which prosecutor can get the most people incarcerated.283 The culture in such offices will likely not change quickly, because our system is so decentralized and the incumbents in those offices hire and train their new attorneys. In thinking about this problem, then, it is probably necessary to approach the shift in prosecutors’ role as a long-term project. If our legal system formally rejects an adversarial role ethic for prosecutors, their new role would be taught to law students, embraced by more moderate prosecutor’s offices, and proselytized by state and federal bar associations. Such official, system-wide changes would hopefully help to spur a larger cultural shift, even in more aggressive offices. There is historical precedent for such a transformation. Over the 18th and 19th centuries, America abandoned private prosecution and built up the institution of the salaried public prosecutor.284 Similarly, in the 21st century we should shift away from adversary prosecution.

VI.     Conclusion

This Article argues that prosecutors should not be allowed to invoke the adversary-system excuse. Prosecutors should behave as morally responsible agents in situations where they exercise discretion. They should implement the law in situations where they lack discretion. But they should never act as partisan advocates. This way of thinking about prosecutors’ role will help to clarify the inquiry into what substantive values prosecutors should actually pursue. Current debates over prosecutorial ethics are mostly limited to situations where prosecutors have broken the rules. This limitation is a consequence of adversarialism’s tendency to devour prosecutors’ other ethical duties. If prosecutors are expected to act strategically within the rules, then they can only really be criticized if they break the rules. Yet, in recent years, some scholarship on prosecutors’ ethics has moved beyond this narrow focus on misconduct. There is a growing literature that suggests several ambitious ways prosecutors can realize their duty to “seek justice.” Scholars have called on prosecutors to combat racial discrimination, protect citizens from abusive searches, expand defendants’ constitutional rights, reduce the punitiveness of the justice system, and more.285 If the framework in this Article were embraced, then this genre of scholarship could no longer be dismissed as addressing only one aspect of prosecutors’ role. Eliminating the current “dual role” framework, and its attendant licensing of amoral partisanship, would help establish a firmer foundation for debate amongst the organized bar, scholars, government institutions, and the broader public over the values that prosecution should advance. Society would see prosecutors more clearly as morally accountable actors, akin to legislators or judges, whose choices cannot be excused by the adversary-litigation system.

  1. [1]. See generally Bryan Stevenson, Just Mercy: A Story of Justice and Redemption (2014).

  2. [2]. Id. at 109–13.

  3. [3]. Id. at 110.

  4. [4]. See, e.g., Mark Baker, D.A.: Prosecutors in Their Own Words 78–79, 82 (1999) (“It really came down ultimately to getting a plea or winning a trial so I could go home that day and say, ‘Okay, I won today. That game is over.’”); Paul Butler, Let’s Get Free: A Hip-Hop Theory of Justice 114–18 (2009) (describing how the adversary system derails idealistic prosecutors and turns them into win-seekers); Milton Heumann, Plea Bargaining: The Experiences of Prosecutors, Judges, and Defense Attorneys 111 (1978) (“What the new prosecutor is taught is that no matter how solid a case he[/she] has, there is always the possibility that he[/she] will lose at trial. And a defeat at trial means total loss ... .”); Nicole Gonzalez Van Cleve, Crook County: Racism and Injustice in America’s Largest Criminal Court 70–71, 136–37 (2016); Stanley Z. Fisher, In Search of the Virtuous Prosecutor: A Conceptual Framework, 15 Am. J. Crim. L. 197, 206–07 (1988); Lara Bazelon, The Innocence Deniers, Slate (Jan. 10, 2018),

  5. [5]. Former Attorney General Richard Thornburgh, for example, has described the prosecutor’s job as “constantly pushing the edge of the envelope out to see if you can get an edge for the prosecution” because “[y]ou’re trying to get every edge you can on those people who are devising increasingly more intricate schemes to rip off the public, hiring the best lawyers, providing the best defenses.” Jim McGee, War on Crime Expands U.S. Prosecutors’ Powers, Wash. Post (Jan. 10, 1993),; see also Object Anyway, Radiolab Presents: More Perfect, WNYC (July 15, 2016),
    story/object-anyway (quoting a prosecutor as stating that “there isn’t a prosecutor in the country” that would avoid making a racially motivated peremptory challenge if they thought it would help win them a conviction). Note also that the media commonly refers to convictions as “wins” or “victories” for the prosecution. See, e.g., Alex Berenson, Prosecutors Score White-Collar Victories, N.Y. Times (Apr. 4, 2004), (“‘The conviction rate in these cases is 85 percent in federal court,’ said Ira Lee Sorkin, a New York defense lawyer and former prosecutor. ‘It’s not hard to win these cases.’”).

  6. [6]. Model Rules of Prof’l Conduct r. 3.8 cmt. 1 (Am. Bar Ass’n 2016).

  7. [7]. See, e.g., Alafair Burke, Prosecution (is) Complex, 10 Ohio St. J. Crim. L. 703, 706 (2013) (reviewing Daniel S. Medwed, Prosecution Complex: America’s Race to Convict and its Impact on the Innocent (2012)); R. Michael Cassidy, Character and Context: What Virtue Theory Can Teach Us About a Prosecutor’s Ethical Duty to “Seek Justice, 82 Notre Dame L. Rev. 635, 651 n.125 (2006); Fisher, supra note 4, at 235–41; Fred C. Zacharias, Structuring the Ethics of Prosecutorial Trial Practice: Can Prosecutors Do Justice?, 44 Vand. L. Rev. 45, 72 (1991). See generally Daniel S. Medwed, The Prosecutor as Minister of Justice: Preaching to the Unconverted from the Post-Conviction Pulpit, 84 Wash. L. Rev. 35 (2009) (arguing for greater involvement of prosecutors in investigating wrongful convictions).

  8. [8]. See David A. Harris, Failed Evidence: Why Law Enforcement Resists Science 8–14 (2012); Hadar Aviram, Legally Blind: Hyperadversarialism, Brady Violations, and the Prosecutorial Organizational Culture, 87 St. John’s L. Rev. 1, 25–36 (2013); Alex Kozinski, Criminal Law 2.0, 44 Geo. L.J. Ann. Rev. Crim. Proc. iii, xxvi (2015); Daniel S. Medwed, The Zeal Deal: Prosecutorial Resistance to Post-Conviction Claims of Innocence, 84 B.U. L. Rev. 125, 134–37 (2004).

  9. [9]. See John F. Pfaff, The Micro and Macro Causes of Prison Growth, 28 Ga. St. U. L. Rev. 1239, 1241 (2012) (“On the micro side, data indicate that at least since 1994, prison growth has been driven primarily by prosecutors increasing the rate at which they file charges against arrestees.”).

  10. [10]. See Medwed, supra note 8, at 132–50; Bazelon, supra note 4.

  11. [11]. See, e.g., Jon B. Gould et al., Predicting Erroneous Convictions: A Social Science Approach to Miscarriages of Justice 15–16, 19 (2012); Peter A. Joy, The Relationship Between Prosecutorial Misconduct and Wrongful Convictions: Shaping Remedies for a Broken System, 2006 Wis. L. Rev. 399, 399–400. See generally Barry Scheck et al., Actual Innocence: When Justice Goes Wrong and How to Make It Right (2003) (examining cases of wrongfully convicted individuals whose convictions have been overturned by the Innocence Project).

  12. [12]. See infra Part II.D.

  13. [13]. See infra Part II.D.

  14. [14]. There are other reforms one could make to the criminal justice system that would mitigate the harms of prosecutorial partisanship. One could, for example, establish a system of rules severely constraining prosecutors’ discretion or give judges more supervisory control over charging decisions and plea-bargain negotiations. For a discussion of the tradeoffs involved in such choices, see infra Part IV.D.

  15. [15]. For a discussion of prosecutors’ disproportionate power in plea bargaining, see Stephanos Bibas, Plea Bargaining Outside the Shadow of Trial, 117 Harv. L. Rev. 2463, 2470–72 (2004) (exploring a number of structural and psychological factors that distort the plea-bargaining process), and William J. Stuntz, Plea Bargaining and Criminal Law’s Disappearing Shadow, 117 Harv. L. Rev. 2548, 2553–54 (2004).

  16. [16]. See Bruce A. Green, Why Should Prosecutors “Seek Justice”?, 26 Fordham Urb. L.J. 607, 615 (1999).

  17. [17]. Berger v. United States, 295 U.S. 78, 88 (1935); see Bennett L. Gershman, “Hard Strikes and Foul Blows:” Berger v. United States 75 Years After, 42 Loy. U. Chi. L.J. 177, 178–79 (2010) (“Indeed, courts have cited this decision so often that it has attained a near-iconic status for its description of the prosecutor’s duty to serve justice, play by the rules, and not hit below the belt.”).

  18. [18]. Model Rules of Prof’l Conduct r. 3.8 cmt. 1 (Am. Bar Ass’n 2016).

  19. [19]. Nat’l Prosecution Standards § 1-1.1 (Nat’l Dist. Attorneys Ass’n, 3d ed. 2009).

  20. [20]. Brady v. Maryland, 373 U.S. 83, 87–88 (1963).

  21. [21]. See, e.g., Transcript of Oral Argument, Miranda v. Arizona, 384 U.S. 436 (1966) (No. 759), reprinted in 63 Landmark Briefs and Arguments of the Supreme Court of the United States: Constitutional Law 597, 864 (Philip B. Kurland & Gerhard Casper eds., 1975) (discussing the prosecutor’s duty “to see that justice is done”); Sari Horwitz, New Deputy Attorney General: ‘We’re Not the Department of Prosecutions, Wash. Post (May 17, 2015),
    national-security/new-deputy-attorney-general-were-not-the-department-of-prosecutions/2015/05/17/8404e500-fb0f-11e4-9ef4-1bb7ce3b3fb7_story.html (“‘We’re not the Department of Prosecutions or even the Department of Public Safety,’ [Deputy Attorney General Sally] Yates said. ‘We are the Department of Justice.’”); Jodi Wilgoren, Prosecutors Use DNA Test to Clear Man in ‘85 Rape, N.Y. Times (Nov. 14, 2002), (“The major reason we undertook this review is because of the attack on prosecutors and the criminal justice system lately—I’m afraid that it’s left an impression with the public that all we care about is convictions, and not justice . . . .”); Memorandum from David W. Ogden, Deputy Att’y Gen., to Dep’t of Justice Prosecutors, Guidance for Prosecutors Regarding Criminal Discovery (Jan. 4, 2010), (“First and foremost, however, such compliance will facilitate a fair and just result in every case, which is the Department’s singular goal in pursuing a criminal prosecution.”).

  22. [22]. See, e.g., United States v. Hinson, 585 F.3d 1328, 1337–38 (10th Cir. 2009) (noting abuses by Kansas federal prosecutors and quoting a passage from Berger); State v. Monday, 257 P.3d 551, 556 (Wash. 2011) (en banc) (“The prosecutor owes a duty to defendants to see that their rights to a constitutionally fair trial are not violated.”); Transcript of Record at 5195, 5198, 5202, United States v. Ruehle, No. SACR 08-139-CJC, 2009 U.S. Dist. LEXIS 117895 (C.D. Cal. Dec. 15, 2009) (quoting the passage from Berger while holding that the prosecutor had committed misconduct); Kozinski, supra note 8, at viii (“The Supreme Court has told us in no uncertain terms that a prosecutor’s duty is to do justice, not merely to obtain a conviction... . There is reason to doubt that prosecutors comply with these obligations fully.”). The author also notes that Judge Pierre Leval of the Second Circuit expresses displeasure at federal prosecutors who have abused their power by asking them: “Who do you work for?” (The correct answer being “the Department of Justice,” and not “the Department of Public Prosecutions.”).

  23. [23]. Herring v. New York, 422 U.S. 853, 862 (1975); see also United States v. Nobles, 422 U.S. 225, 230 (1975) (“The dual aim of our criminal justice system is ‘that guilt shall not escape or innocence suffer.’ To this end, we have placed our confidence in the adversary system, entrusting to it the primary responsibility for developing relevant facts on which a determination of guilt or innocence can be made.” (citation omitted) (quoting Berger v. United States, 295 U.S. 78, 88 (1935))); cf. United States v. Cronic, 466 U.S. 648, 656–57 (1984) (“When a true adversarial criminal trial has been conducted—even if defense counsel may have made demonstrable errors—the kind of testing envisioned by the Sixth Amendment has occurred. But if the process loses its character as a confrontation between adversaries, the constitutional guarantee is violated.” (footnotes omitted)).

  24. [24]. See Zacharias, supra note 7, at 109 (“By subjecting prosecutors to the professional codes, the bar imposes on them the obligation to be zealous. Within the adversary system, that turns prosecutors loose as advocates.”).

  25. [25]. Standards of Criminal Justice Relating to the Prosecution Function Standard
    3-1.2(a) (Am. Bar Ass’n 2015).

  26. [26]. See, e.g., Monroe H. Freedman, Lawyers’ Ethics in an Adversary System (1975); Gwladys Gilliéron, Public Prosecutors in the United States and Europe 134 (2014) (“As an advocate for the state, the prosecutor’s aim is to win the case by obtaining convictions.” (footnote omitted)); Geoffrey C. Hazard, Jr. et al., The Law and Ethics of Lawyering 712 (5th ed. 2010) (“However, the characterization of prosecutors as ‘ministers of justice’ describes only one of their roles. They are also expected to be zealous advocates.”); Deborah L. Rhode
    & David Luban, Legal Ethics 355 (5th ed. 2009) (“Codes and commentary on prosecutorial ethics generally build on a shared premise. Prosecutors have a dual role as advocates and officers of justice ... .”); Mortimer D. Schwartz et al., Problems in Legal Ethics 261 (11th ed. 2015) (“But the prosecutor is also an advocate in the adversary system of criminal litigation. Thus the prosecutor is asked to assume a dual role as a partisan advocate and a quasi-judicial officer—a role difficult to achieve in practice.”).

  27. [27]. See, e.g., Marvin E. Frankel, Partisan Justice 89 (1980) (“At the level of ideology, where due process vies with crime control, the essentially adversary premise is not questioned. It is accepted and shared on both—or on all—sides.”); Albert W. Alschuler, The Prosecutor’s Role in Plea Bargaining, 36 U. Chi. L. Rev. 50, 59–60, 67–68 (1968); Daniel Epps, Adversarial Asymmetry in the Criminal Process, 91 N.Y.U. L. Rev. 762, 776 (2016); George T. Felkenes, The Prosecutor: A Look at Reality, 7 Sw. U. L. Rev. 98, 110–14 (1975); Fisher, supra note 4, at 208–11; Bruce A. Green, The Ethical Prosecutor and the Adversary System, 24 Crim. L. Bull. 126, 129–30 (1988); Green, supra note 16, at 615–17; Robert M. Ireland, Privately Funded Prosecution of Crime in the Nineteenth-Century United States, 39 Am. J. Legal Hist. 43, 58 (1995); Kenneth J. Melilli, Prosecutorial Discretion in an Adversary System, 1992 BYU L. Rev. 669, 691–97; Whitney North Seymour, Jr., Why Prosecutors Act Like Prosecutors, 11 Rec. Ass’n B. N.Y.C. 302, 312–13 (1956); Jay Sterling Silver, Equality of Arms and the Adversarial Process: A New Constitutional Right, 1990 Wis. L. Rev. 1007, 1035–37; Abbe Smith, Can You Be a Good Person and a Good Prosecutor?, 14 Geo. J. Legal Ethics 355, 375–76, 388 (2001); H. Richard Uviller, The Neutral Prosecutor: The Obligation of Dispassion in a Passionate Pursuit, 68 Fordham L. Rev. 1695, 1698–99 (2000); James Vorenberg, Decent Restraint of Prosecutorial Power, 94 Harv. L. Rev. 1521, 1557 (1981) (“It is customary to note that while prosecutors act as the government’s representative in the adversary system, they are expected to be more (or is it less?) than an adversary ... .”); Zacharias, supra note 7, at 51–52.

  28. [28]. See Model Rules of Prof’l Conduct, pmbl. and scope, para. 2 (Am. Bar Ass’n 2016).

  29. [29]. See Ken White, Confessions of an Ex-Prosecutor, Reason (June 23, 2016), http://reason.
    com/archives/2016/06/23/confessions-of-an-ex-prosecutor (“A persistent professional obligation to argue that violations of constitutional rights don’t matter can’t help but influence how prosecutors look at rights, and treat them.”).

  30. [30]. See, e.g., Sportsmanlike/Unsportsmanlike Behavior, Wash. Hunter Educ. Course, https://
    0105001_700046830 (last visited Mar. 23, 2018) (“It’s not illegal to shoot ducks on the water or pheasants on the ground. No true sportsman would ever think about taking such a shot!” (emphasis omitted)).

  31. [31]. Robert H. Jackson, The Federal Prosecutor, 31 J. Am. Inst. Crim. L. & Criminology 3, 6 (1940).

  32. [32]. Zacharias, supra note 7, at 60.

  33. [33]. Id. at 66–74, 85–88.

  34. [34]. Kevin C. McMunigal, Are Prosecutorial Ethics Standards Different?, 68 Fordham L. Rev. 1453, 1472 (2000). McMunigal does allow that prosecutors are held to a higher standard in some areas, but holds that these differences are only matters of degree. Id.

  35. [35]. Bennett L. Gershman, The Prosecutor’s Duty to Truth, 14 Geo. J. Legal Ethics 309,
    311–15 (2001).

  36. [36]. For other approaches, see, e.g., Cassidy, supra note 7, at 639; Fisher, supra note 4, at
    201–02; Samuel J. Levine, A Consideration of the Prosecutor’s Ethical Obligation to “Seek Justice” in a Comparative Analytical Framework, 41 Hous. L. Rev. 1337, 1339–40 (2004).

  37. [37]. Model Rules of Prof’l Conduct r. 3.8 (Am. Bar. Ass’n 2016).

  38. [38]. Melanie D. Wilson, Prosecutors “Doing Justice” Through Osmosis—Reminders to Encourage a Culture of Cooperation, 45 Am. Crim. L. Rev. 67, 68–69 (2008).

  39. [39]. Angela J. Davis, Prosecution and Race: The Power and Privilege of Discretion, 67 Fordham L. Rev. 13, 50–53 (1998); Justin Murray, Reimagining Criminal Prosecution: Toward a Color-Conscious Professional Ethic for Prosecutors, 49 Am. Crim. L. Rev. 1541, 1575–77 (2012).

  40. [40]. Russell M. Gold, Beyond the Judicial Fourth Amendment: The Prosecutor’s Role, 47 U.C. Davis L. Rev. 1591, 1632–34 (2014).

  41. [41]. Maureen A. Howard, Taking the High Road: Why Prosecutors Should Voluntarily Waive Peremptory Challenges, 23 Geo. J. Legal Ethics 369, 409–19 (2010).

  42. [42]. Bruce A. Green, Her Brother’s Keeper: The Prosecutor’s Responsibility When Defense Counsel Has a Potential Conflict of Interest, 16 Am. J. Crim. L. 323, 335–36 (1989); Smith, supra note 27, at 394–95.

  43. [43]. See, e.g., R. Michael Cassidy, (Ad)ministering Justice: A Prosecutor’s Ethical Duty to Support Sentencing Reform, 45 Loy. U. Chi. L.J. 981, 992–97 (2014) (discussing the prosecutor’s duty to join the effort for repeal of mandatory minimum sentences in low-level drug and non-violent cases); Angela J. Davis, The Prosecutor’s Ethical Duty to End Mass Incarceration, 44 Hofstra L. Rev. 1063, 1064 (2016) (arguing “that prosecutors have an ethical duty to take action that significantly reduces the incarceration rate in the United States”); Bennett Gershman, The Prosecutor’s Duty of Silence, 79 Alb. L. Rev. 1183, 1213–20 (2015) (explaining a prosecutor’s duty to avoid discussing specifics of cases that are currently being investigated or prosecuted); Judith A. Goldberg & David M. Siegel, The Ethical Obligations of Prosecutors in Cases Involving Postconviction Claims of Innocence, 38 Cal. W. L. Rev. 389, 410–12 (2002) (discussing “response[s] to a defendant’s request for the application of new scientific testing to evidence”); Bruce A. Green, Access to Criminal Justice: Where are the Prosecutors?, 3 Tex. A&M L. Rev. 515, 519–35 (2016) (explaining a prosecutor’s duty to “seek justice” and avoid both individual and systemic injustices).

  44. [44]. See, e.g., Fisher, supra note 4, at 224–26; Roberta K. Flowers, A Code of Their Own: Updating the Ethics Codes to Include the Non-Adversarial Roles of Federal Prosecutors, 37 B.C. L. Rev. 923, 927 (1996); cf. Eric S. Fish, Prosecutorial Constitutionalism, 90 S. Cal. L. Rev. 237, 253–70 (2017) (arguing that prosecutors should be especially attentive to preserving defendants’ constitutional rights in situations where a lack of adversary safeguards undermines those rights). See generally Uviller, supra note 27 (distinguishing between a prosecutor’s adjudicatory, investigatory, and advocacy functions, and concluding that the former two call for neutrality while the third calls for adversarialism).

  45. [45]. See, e.g., Rachel E. Barkow, Institutional Design and the Policing of Prosecutors: Lessons from Administrative Law, 61 Stan. L. Rev. 869, 895–906 (2009); Uviller, supra note 27, at 1716–18.

  46. [46]. See Angela J. Davis, Arbitrary Justice: The Power of the American Prosecutor 3 (2007); see also Daniel S. Medwed, Prosecution Complex: America’s Race to Convict and Its Impact on the Innocent2–4 (2014).

  47. [47]. See supra Part II.A.

  48. [48]. See Medwed, supra note 46, at 2–4.

  49. [49]. See, e.g., Harris, supra note 8, at 104 (“[P]rosecutors’ careers advance according to their conviction rates. The higher the rate, the better they do.”); Alschuler, supra note 27, at 106,
    110–11; Kenneth Bresler, “I Never Lost a Trial”: When Prosecutors Keep Score of Criminal Convictions, 9 Geo. J. Legal Ethics 537, 541–44 (1996); Fisher, supra note 4, at 205–07; Thomas A. Hagemann, Confessions from a Scorekeeper: A Reply to Mr. Bresler, 10 Geo. J. Legal Ethics 151,
    151–53 (1996); Kozinski, supra note 8, at xxvi; Medwed, supra note 8, at 134–37; Daniel C. Richman, Old Chief v. United States: Stipulating Away Prosecutorial Accountability?, 83 Va. L. Rev. 939, 966–69 (1997). On the more general phenomenon of quantification bias in bureaucracies, see Paul H. Rubin, Business Firms and the Common Law: The Evolution of Efficient Rules 116–17 (1983).

  50. [50]. See, e.g., Van Cleve, supra note 4, at 70–71; Peter A. Joy & Kevin C. McMunigal, Contingent Rewards for Prosecutors?, 26 Crim. Just., Fall 2011, at 55, 55–56; Medwed, supra note 8, at 134–37; Catherine Ferguson-Gilbert, Comment, It Is Not Whether You Win or Lose, It Is How You Play the Game: Is the Win-Loss Scorekeeping Mentality Doing Justice for Prosecutors?, 38 Cal. W. L. Rev. 283, 293 (2001); Evan Moore, ‘Win at All Costs’ is Smith County’s Rule, Critics Claim, Hous. Chron. (June 11, 2000, 5:30 AM),; Maurice Possley & Ken Armstrong, The Flip Side of a Fair Trial (pt. 2), Chi. Trib. (Jan. 11, 1999),

  51. [51]. See Fisher, supra note 4, at 206–11 (“The moral and political climate in an agency can foster a ‘conviction psychology’ more powerfully than can any specific policy basing promotions on an assistant’s conviction rate.”); Medwed, supra note 8, at 132–50; Melilli, supra note 27, at 691; Daniel Richman, Prosecutors and Their Agents, Agents and Their Prosecutors, 103 Colum. L. Rev. 749, 792 (2003) (“[O]ne ought not underestimate the unifying influence of a shared commitment to ‘getting the bad guys,’ hardened by the adversarial process, nurtured by mutual respect and need, and on occasion lubricated by alcohol.” (footnote omitted)); Smith, supra note 27, at 388 (“In view of the institutional culture of prosecutor’s offices and the culture of the adversary system generally, it is perhaps inevitable that the overriding interest of prosecutors would be winning.” (footnote omitted)); White, supra note 29 (“[M]y experience showed me that prosecutors are strongly influenced to disregard and minimize rights by the culture that surrounds them. Disciplining or firing miscreants may be necessary, but it’s not enough: It doesn’t address the root causes of fearful culture and bad incentives.”).

  52. [52]. See Ronald F. Wright & Kay L. Levine, The Cure for Young Prosecutors’ Syndrome, 56 Ariz. L. Rev. 1065, 1071–76 (2014).

  53. [53]. For example, in San Francisco misdemeanor court in 2017, the market rate for pleading guilty to a first-time misdemeanor DUI was three years of informal probation, a three-month DUI class, and fines and fees totaling around $1,000.

  54. [54]. See Alafair S. Burke, Improving Prosecutorial Decision Making: Some Lessons of Cognitive Science, 47 Wm. & Mary L. Rev. 1587, 1605–08 (2006); Bennett L. Gershman, The Zealous Prosecutor as Minister of Justice, 48 San Diego L. Rev. 151, 155 (2011) (noting that prosecutors’ current justice-seeking role “is maddeningly vague and frustratingly amorphous”).

  55. [55]. See Cassidy, supra note 7, at 638 (“At worst, it allows prosecutors to rationalize any response to an ethical dilemma by arguing that their chosen conduct increases the likelihood of conviction and incarceration of a guilty person.”); Zacharias, supra note 7, at 53–54. See generally Vanessa Merton, What Do You Do When You Meet a “Walking Violation of the Sixth Amendment” if You’re Trying to Put That Lawyer’s Client in Jail?, 69 Fordham L. Rev. 997 (2000) (describing a case where a prosecutor committed to the principle that prosecutors should protect defendants’ rights nonetheless took advantage of incompetent defense counsel).

  56. [56]. See generally Samuel J. Levine, The Law and the “Spirit of the Law” in Legal Ethics, 2015 J. Prof. Law. 1; Fred C. Zacharias, Specificity in Professional Responsibility Codes: Theory, Practice, and the Paradigm of Prosecutorial Ethics, 69 Notre Dame L. Rev. 223 (1993) (criticizing the professional ethics codes’ focus on specific, rule-based restrictions on prosecutorial conduct rather than more general standards).

  57. [57]. See Bruce A. Green & Rebecca Roiphe, Rethinking Prosecutors’ Conflicts of Interest, 58 B.C. L. Rev. 463, 510–11 (2017) (arguing that separation-of-prosecution functions intensifies adversarialism and makes prosecutors less familiar with the facts of their cases); Melilli, supra note 27, at 687–90 (pointing out that separating out charging and litigating functions can lead to a diffusion of responsibility, resulting in less internal scrutiny of a case); cf. Barkow, supra note 45, at 898 n.159 (arguing that investigating prosecutors are poorly placed to be neutral, because they form views about what they are investigating).

  58. [58]. See Alschuler, supra note 27, at 106; see, e.g., Bibas, supra note 15, at 2471; Josh Bowers, Punishing the Innocent, 156 U. Pa. L. Rev. 1117, 1128 (2008); Stephen J. Schulhofer, Plea Bargaining as Disaster, 101 Yale L.J. 1979, 1987–88 (1992); Stuntz, supra note 15, at 2553–54.

  59. [59]. Richard T. Boylan, What Do Prosecutors Maximize? Evidence from the Careers of U.S. Attorneys, 7 Am. L. & Econ. Rev. 379, 379 (2005) (drawing on an empirical study of United States attorneys’ performance and subsequent careers to argue that prosecutors are evaluated by the total sentence lengths they achieve, and not by their conviction rates).

  60. [60]. See Alschuler, supra note 27, at 54–56; Epps, supra note 27, at 777–78; Robert A. Weninger, The Abolition of Plea Bargaining: A Case Study of El Paso County, Texas, 35 UCLA L. Rev. 265, 270 n.16 (1987) (“[El Paso County DA] Simmons’ often quoted remark, ‘We don’t make sweetheart deals with criminals,’ makes him popular among voters who generally believe that plea bargains allow criminals to get off easy in exchange for pleading guilty to lesser charges.”).

  61. [61]. See, e.g., Issa Kohler-Hausmann, Managerial Justice and Mass Misdemeanors, 66 Stan. L. Rev. 611, 614–16 (2014).

  62. [62]. See Stuntz, supra note 15, at 2554 n.6 (“There is as yet no developed social science literature on what prosecutors maximize, probably because the solution is too complex to model effectively.”).

  63. [63]. See Allen Steinberg, The Transformation of Criminal Justice: Philadelphia, 1800–1880, at 5 (Thomas A. Green ed., 1989); John H. Langbein, The Origins of Public Prosecution at Common Law, 17 Am. J. Legal Hist. 313, 317–18 (1973).

  64. [64]. See generally John H. Langbein, The Origins of Adversary Criminal Trial (2003) (detailing the development of the adversarial system in English criminal law).

  65. [65]. See Davis, supra note 46, at 9–11; Roger A. Fairfax, Jr., Delegation of the Criminal Prosecution Function to Private Actors, 43 U.C. Davis L. Rev. 411, 421–24 (2009); Allen Steinberg, From Private Prosecution to Plea Bargaining: Criminal Prosecution, the District Attorney, and American Legal History, 30 Crime & Delinq. 568, 569–72 (1984).

  66. [66]. See John L. Worrall, Prosecution in America: A Historical and Comparative Account, in The Changing Role of the American Prosecutor 3, 5–9 (John L. Worrall & M. Elaine Nugent-Borakove eds., 2008); Michael J. Ellis, Note, The Origins of the Elected Prosecutor, 121 Yale L.J. 1528, 1550–64 (2012).

  67. [67]. See Fairfax, supra note 65, at 413; Carolyn B. Ramsey, The Discretionary Power of “Public” Prosecutors in Historical Perspective, 39 Am. Crim. L. Rev. 1309, 1327–31 (2002); Steinberg, supra note 65, at 570, 580–81.

  68. [68]. See Nicholas R. Parrillo, Against the Profit Motive: The Salary Revolution in American Government, 1780–1840, at 255–94 (2013).

  69. [69]. See Ireland, supra note 27, at 43–45. The practice of using privately funded prosecutors was heavily criticized by defense attorneys and prohibited by some state supreme courts (although most state supreme courts had allowed it by the end of the 19th century). Id. at 47–51.

  70. [70]. See Steinberg, supra note 65, at 586. It is also worth noting that private prosecution has not been totally eliminated—some states still allow for it in certain circumstances. See Ireland, supra note 27, at 55–58; Michael Edmund O’Neill, Private Vengeance and the Public Good, 12 U. Pa. J. Const. L. 659, 683–704 (2010).

  71. [71]. William F. McDonald, The Prosecutor’s Domain, in The Prosecutor 15, 19 (William F. McDonald ed., 1979).

  72. [72]. See Steinberg, supra note 65, at 584–85.

  73. [73]. George Sharswood, An Essay on Professional Ethics 37 (1854). This essay heavily influenced the later development of the ABA’s professional ethics rules. See Green, supra note 16, at 612 & n.10.

  74. [74]. See Green, supra note 16, at 613–14; Ireland, supra note 27, at 49–51.

  75. [75]. Meister v. People, 31 Mich. 99, 104 (1875); see also Hurd v. People, 25 Mich. 405, 416 (1872) (“[The prosecutor’s] object like that of the court, should be simply justice; and he[/she] has no right to sacrifice this to any pride of professional success. And however strong may be his belief of the prisoner’s guilt, he[/she] must remember that, though unfair means may happen to result in doing justice to the prisoner in the particular case, yet, justice so attained, is unjust and dangerous to the whole community.”).

  76. [76]. Biemel v. State, 37 N.W. 244, 247 (Wis. 1888).

  77. [77]. Thalheim v. State, 20 So. 938, 943 (Fla. 1896).

  78. [78]. People v. Lee Chuck, 20 P. 719, 723 (Cal. 1889); see also People v. Cahoon, 50 N.W. 384, 385 (Mich. 1891) (“Zeal in a prosecuting attorney is entitled to the highest commendation, but that zeal must be exercised within proper limits... . In criminal cases the prosecuting attorney is a public officer, acting in a quasi judicial capacity.”).

  79. [79]. Berger v. United States, 295 U.S. 78, 88 (1935); see Charles W. Wolfram, Modern Legal Ethics § 13.10.2 (1986) (referring to Berger as “[t]he locus classicus for discussion of the extraordinary duties of a prosecutor”); Gershman, supra note 17, at 178–79 (noting the “near-iconic status” of Berger).

  80. [80]. See generally Canons of Prof’l Ethics (Am. Bar Ass’n 1908).

  81. [81]. See James M. Altman, Considering the A.B.A.’s 1908 Canons of Ethics, 71 Fordham L. Rev. 2395, 2395–96 (2003).

  82. [82]. Canons of Prof’l Ethics Canon 5.

  83. [83]. Model Code of Prof’l Responsibility (Am. Bar Ass’n 1980).

  84. [84]. Standards for Criminal Justice: Standards Relating to the Administration of Criminal Justice (Am. Bar Ass’n 1974).

  85. [85]. Model Rules of Professional Conduct, Am. B. Ass’n,
    professional_responsibility/publications/model_rules_of_professional_conduct.html (last visited Mar. 23, 2018).

  86. [86]. See Model Rules of Prof’l Conduct r. 3.8 (Am. Bar Ass’n 2016) (imposing a number of specific responsibilities on prosecutors, including not bringing a charge without probable cause and making reasonable attempts to ensure the accused has access to counsel); Standards for Criminal Justice: The Prosecution Function § 3-1.1(b) (Am. Bar Ass’n, 3d ed. 1993) (“The prosecutor is both an administrator of justice and an advocate . . . . Although the prosecutor operates within the adversary system, it is fundamental that the prosecutor’s obligation is to protect the innocent as well as to convict the guilty, to guard the rights of the accused as well as to enforce the rights of the public.”); Model Code of Prof’l Responsibility EC 7-13 (Am. Bar Ass’n 1980) (“The responsibility of a public prosecutor differs from that of the usual advocate; his duty is to seek justice, not merely to convict... . [D]uring trial the prosecutor is not only an advocate but he also may make decisions normally made by an individual client, and those affecting the public interest should be fair to all ... .”); see also Lon L. Fuller & John D. Randall, Report of the Joint Conference on Professional Responsibility of the Association of American Law Schools and American Bar Association 15 (1958) (“The public prosecutor must recall that he[/she] occupies a dual role, being obligated, on the one hand, to furnish that adversary element essential to the informed decision of any controversy, but being possessed, on the other, of important governmental powers that are pledged to the accomplishment of one objective only, that of impartial justice.”).

  87. [87]. See generally Nat’l Prosecution Standards (Nat’l Dist. Attorneys Ass’n, 3d ed. 2009).

  88. [88]. See generally U.S. Dep’t of Justice, Principles of Federal Prosecution (1980).

  89. [89]. See R. v. Cook, [1997] 1 S.C.R. 1113, 1124 (Can.) (“[W]hile it is without question that the Crown performs a special function in ensuring that justice is served and cannot adopt a purely adversarial role towards the defence, it is well recognized that the adversarial process is an important part of our judicial system ... . Nor should it be assumed that the Crown cannot act as a strong advocate within this adversarial process.” (emphasis omitted) (citations omitted)); Pub. Prosecution Serv. of Can., Deskbook 2.2 Duties and Responsibilities of Crown Counsel (2014); David Plater, The Development of the Prosecutor’s Role in England and Australia with Respect to its Duty of Disclosure: Partisan Advocate or Minister of Justice?, 25 U. Tas. L. Rev. 111, 153 (2006). See generally Brian A. Grosman, The Role of the Prosecutor in Canada, 18 Am. J. Comp. L. 498 (1970) (discussing the transformation from private to public prosecution and the role discretion plays in various aspects of criminal trials in Canada); David Plater, The Changing Role of the Modern Prosecutor: Has the Notion of the “Minister of Justice” Outlived Its Usefulness? (April 2011) (unpublished manuscript) (on file with author) (describing the development of prosecution and prosecutorial ethics in Australia and England).

  90. [90]. See Pfaff, supra note 9, at 1241.

  91. [91]. Legal scholars offer a wide variety of explanations for this shift. See, e.g., Michelle Alexander, The New Jim Crow: Mass Incarceration in the Age of Colorblindness 40–57 (2010) (emphasizing political changes beginning in the 1950s, and especially the backlash to the civil-rights movement); Katherine Beckett, Making Crime Pay: Law and Order in Contemporary American Politics 5–6 (1999) (emphasizing the role of self-interested politicians stoking fear of crime); David Garland, The Culture of Control: Crime and Social Order in Contemporary Society 169–92 (2001) (emphasizing broad shifts in cultural expectations about the purpose of government); William J. Stuntz, The Collapse of American Criminal Justice 4–7 (2011) (emphasizing the increased importance of criminal procedure, increases in violent crime and prosecutorial discretion, and the decline of local democracy); Michael Tonry, Malign Neglect: Race, Crime, and Punishment in America 49–80 (1995) (emphasizing racial politics); see also John F. Pfaff, The Empirics of Prison Growth: A Critical Review and Path Forward, 98 J. Crim. L. & Criminology 547, 552–65 (2008) (discussing various theories for the increased harshness of American criminal justice since the 1970s).

  92. [92]. See generally David Alan Sklansky, Anti-Inquisitorialism, 122 Harv. L. Rev. 1634 (2009) (describing how the American system defines its criminal procedure against the European inquisitorial model).

  93. [93]. See Erik Luna & Marianne Wade, Prosecutors as Judges, 67 Wash. & Lee L. Rev. 1413, 1468–74 (2010) (describing the role of prosecutors in continental Europe).

  94. [94]. See Strafprozessordnung [StPO] [Code of Crim. Procedure], § 296(2), translation
    at (Ger.) (“The public prosecution office may also make use of [appellate remedies] for the benefit of the accused.”); Shawn Marie Boyne, Uncertainty and the Search for Truth at Trial: Defining Prosecutorial “Objectivity” in German Sexual Assault Cases, 67 Wash. & Lee L. Rev. 1287, 1289–90 (2010).

  95. [95]. See Boyne, supra note 94, at 1289.

  96. [96]. See Luna & Wade, supra note 93, at 1470.

  97. [97]. See Shawn Marie Boyne, The German Prosecution Service: Guardians of the Law? 25–28 (2014) (describing the history of prosecutors in Germany).

  98. [98]. See id. at 38–42; Luna & Wade, supra note 93, at 1471–74.

  99. [99]. See Boyne, supra note 94, at 1290.

  100. [100]. StPO § 152(2); see Boyne, supra note 97, at 106.

  101. [101]. See Boyne, supra note 97, at 104–07. In recent work, Shawn Marie Boyne has argued that the German system falls short of this rule in practice due to resource constraints. See id. However, she also notes that mandatory prosecution remains a canonical ideal of the German system, and that prosecution chiefs seek to realize it despite the difficulties. Id. at 107–08, 237.

  102. [102]. There are practices in the French and German systems that legal scholars compare to plea bargaining, such as the recent allowance of confession agreements in Germany and the negotiation of charges downward to avoid necessity for trial in France. See Erik Luna & Marianne L. Wade, The Prosecutor in Transnational Perspective 92–94 (2012); Richard S. Frase, Comparative Criminal Justice as a Guide to American Law Reform: How Do the French Do It, How Can We Find Out, and Why Should We Care?, 78 Calif. L. Rev. 539, 628–47 (1990). These practices give prosecutors more discretion than they would have if every case had to go to trial. However, these practices are not nearly as pervasive as American-style plea bargaining, they do not rely on sentence coercion in the way the American system does, they occur after the investigation is complete, and they involve a great deal more judicial supervision and presentation of evidence. See Luna & Wade, supra note 93, at 1462–63; Jenia I. Turner, Plea Bargaining and Disclosure in Germany and the United States: Comparative Lessons, 57 Wm. & Mary L. Rev. 1549, 1572–76 (2016).

  103. [103]. See William T. Pizzi, Trials Without Truth: Why Our System of Criminal Trials Has Become an Expensive Failure and What We Need to Do to Rebuild It 98–99 (1999).

  104. [104]. See Boyne, supra note 94, at 1314.

  105. [105]. See, e.g., Robert A. Kagan, Adversarial Legalism: The American Way of Law 61–81 (2001); Pizzi, supra note 103, at 117–39; Albert W. Alschuler, Introduction: Adding a Comparative Perspective to American Criminal Procedure Classes, 100 W. Va. L. Rev. 765, 767–71 (1998); Craig M. Bradley, Reforming the Criminal Trial, 68 Ind. L.J. 659, 660 (1993); Frase, supra note 102, at
    610–45; Richard S. Frase & Thomas Weigend, German Criminal Justice as a Guide to American Law Reform: Similar Problems, Better Solutions?, 18 B.C. Int’l & Comp. L. Rev. 317, 353–57 (1995); Amalia D. Kessler, Our Inquisitorial Tradition: Equity Procedure, Due Process, and the Search for an Alternative to the Adversarial, 90 Cornell L. Rev. 1181, 1251–60 (2005); John H. Langbein, Land Without Plea Bargaining: How the Germans Do It, 78 Mich. L. Rev. 204, 204 (1979) [hereinafter Langbein, Land Without Plea Bargaining]; Mirjan Damaska, Comment, The Reality of Prosecutorial Discretion: Comments on a German Monograph, 29 Am. J. Comp. L. 119, 127 (1981). See generally John H. Langbein & Lloyd L. Weinreb, Continental Criminal Procedure: “Myth” and Reality, 87 Yale L.J. 1549 (1978); William T. Pizzi & Walter Perron, Crime Victims in German Courtrooms: A Comparative Perspective on American Problems, 32 Stan. J. Int’l L. 37 (1996); Christopher Slobogin, Lessons from Inquisitorialism, 87 S. Cal. L. Rev. 699 (2014); Thomas Weigend, Continental Cures for American Ailments: European Criminal Procedure as a Model for Law Reform, 2 Crime & Just. 381 (1980).

  106. [106]. See supra notes 92–104 and accompanying text.

  107. [107]. See infra Part III.A.

  108. [108]. See infra Part III.A.

  109. [109]. See William T. Pizzi, Understanding Prosecutorial Discretion in the United States: The Limits of Comparative Criminal Procedure as an Instrument of Reform, 54 Ohio St. L.J. 1325, 1330 (1993) (“The idea that we can leave our criminal justice system and our legal tradition substantially intact, but yet achieve meaningful reform of prosecutorial discretion by borrowing mechanisms for controlling such discretion from the civil law tradition is mistaken and unfair to both great traditions.”).

  110. [110]. For a substantively similar taxonomy of arguments for adversarialism, see David Luban, Lawyers and Justice: An Ethical Study 67–93 (1988) (describing three “consequentialist” and three “non-consequentialist” arguments for an adversarial system).

  111. [111]. I will not venture to make a case for defense-lawyer adversarialism in this Article, but note that the debate between David Luban and William Simon on the question is quite illuminating. Compare David Luban, Are Criminal Defenders Different?, 91 Mich. L. Rev. 1729 (1993), with William H. Simon, The Ethics of Criminal Defense, 91 Mich. L. Rev. 1703 (1993) [hereinafter Simon, The Ethics of Criminal Defense], and William H. Simon, Reply: Further Reflections on Libertarian Criminal Defense, 91 Mich. L. Rev. 1767 (1993).

  112. [112]. To a certain degree, uneven adversarialism is already a requirement of the current “dual role” model. See Epps, supra note 27, at 770, 775–77.

  113. [113]. David Luban, Legal Ethics and Human Dignity 19–64 (2007).

  114. [114]. Id.

  115. [115]. See Davis, supra note 46, at 3–5; Stephanos Bibas, Prosecutorial Regulation Versus Prosecutorial Accountability, 157 U. Pa. L. Rev. 959, 960–62 (2009); Green, supra note 16, at
    625–33; Luna & Wade, supra note 93, at 1414–15; Wright & Levine, supra note 52, at 1067.

  116. [116]. See Stephanos Bibas, The Machinery of Criminal Justice 30, 41–43 (2012); George Fisher, Plea Bargaining’s Triumph: A History of Plea Bargaining in America 222–27 (2003); Rachel E. Barkow, Criminal Law as Regulation, 8 N.Y.U. J.L. & Liberty 316, 325–26 (2014); Máximo Langer, Rethinking Plea Bargaining: The Practice and Reform of Prosecutorial Adjudication in American Criminal Procedure, 33 Am. J. Crim. L. 223, 248–51 (2006); Gerard E. Lynch, Our Administrative System of Criminal Justice, 66 Fordham L. Rev. 2117, 2149 (1998) (“[P]rosecutors, in their discretionary charging and plea bargaining decisions, are acting largely as administrative, quasi-judicial decision-makers ... .”).

  117. [117]. See Fed. R. Crim. P. 11(c); Rishi Raj Batra, Judicial Participation in Plea Bargaining: A Dispute Resolution Perspective, 76 Ohio St. L.J. 565, 572–79 (2015) (describing variation amongst state systems, with most states disallowing or discouraging judicial involvement in plea bargaining); Daniel S. McConkie, Judges as Framers of Plea Bargaining, 26 Stan. L. & Pol’y Rev. 61, 66–69 (2015) (describing the lack of judicial involvement in plea bargaining in the federal system). But see generally Nancy J. King & Ronald F. Wright, The Invisible Revolution in Plea Bargaining: Managerial Judging and Judicial Participation in Negotiations, 95 Tex. L. Rev. 325 (2016) (describing a recent trend in some states towards more judicial involvement in plea bargaining, as well as more formally regulated mechanisms for judicial involvement).

  118. [118]. See Stuntz, supra note 15, at 2560; see also Bibas, supra note 15, at 2515–19 (exploring a number of structural and psychological factors that distort the plea-bargaining process).

  119. [119]. See Malcolm M. Feeley, The Process is the Punishment: Handling Cases in a Lower Criminal Court 199–243 (1992); Josh Bowers, Legal Guilt, Normative Innocence, and the Equitable Decision Not to Prosecute, 110 Colum. L. Rev. 1655, 1709 (2010) (“For disposable cases, prosecutors’ initial decisions of what and whether to charge are somewhat dispositive on the question of whether the defendant will ultimately end up with some type of conviction—even if some equitable play remains in the punishment joints.”); Kohler-Hausmann, supra note 61, at 663 (describing the misdemeanor system in New York as one in which people are marked and sorted based on their prior contacts with the system, and in which defendants who cannot make bail simply plead guilty); Alexandra Natapoff, Misdemeanors, 85 S. Cal. L. Rev. 1313, 1336 (2012) (“[P]rosecutors routinely convert the majority of such arrests into cases, and once that happens, most defendants plead guilty.”).

  120. [120]. See Barkow, supra note 45, at 876–83; Robert L. Misner, Recasting Prosecutorial Discretion, 86 J. Crim. L. & Criminology 717, 718 (1996) (“In the past thirty years ... power has increasingly come to rest in the office of the prosecutor. Developments in the areas of charging, plea bargaining, and sentencing have made the prosecutor the preeminent actor in the system.”); David Alan Sklansky, The Nature and Function of Prosecutorial Power, 106 J. Crim. L. & Criminology 473, 499–502, 504–06 (2016) (arguing that it is prosecutors’ ability to mediate between adversarial and inquisitorial models of justice, as well as between law on the books and discretionary enforcement, that makes them such powerful figures).

  121. [121]. See Am. Bar Ass’n Standing Comm. on Legal Aid & Indigent Defendants, Gideon’s Broken Promise: America’s Continuing Quest for Equal Justice 38 (2004), (noting that the underfunding of public defense “places poor persons at constant risk of wrongful conviction”); Mary Sue Backus & Paul Marcus, The Right to Counsel in Criminal Cases, A National Crisis, 57 Hastings L.J. 1031, 1045 (2006) (“By every measure in every report analyzing the U.S. criminal justice system, the defense function for poor people is drastically underfinanced.”). See generally Carol S. Steiker, Gideon at Fifty: A Problem of Political Will, 122 Yale L.J. 2694, 2696 (2013).

  122. [122]. See Caroline Wolf Harlow, Bureau of Justice Statistics, Defense Counsel in Criminal Cases 5 (2000) (“8 in 10 felony defendants in large [s]tate courts used publicly financed attorneys . . . .”).

  123. [123]. See Strickland v. Washington, 466 U.S. 668, 687 (1984) (establishing the high standard to prove an ineffective-assistance-of-counsel claim); Stephanos Bibas, The Psychology of Hindsight and After-the-Fact Review of Ineffective Assistance of Counsel, 2004 Utah L. Rev. 1; Eve Brensike Primus, Procedural Obstacles to Reviewing Ineffective Assistance of Trial Counsel Claims in State and Federal Postconviction Proceedings, 24 Crim. Just., Fall 2009, at 6. See generally Lisa Kern Griffin, Criminal Adjudication, Error Correction, and Hindsight Blind Spots, 73 Wash. & Lee L. Rev. 165 (2016).

  124. [124]. U.S. Sentencing Guidelines Manual § 5K1.1 (U.S. Sentencing Comm’n 2016). See generally Cynthia Kwei Yung Lee, Prosecutorial Discretion, Substantial Assistance, and the Federal Sentencing Guidelines, 42 UCLA L. Rev. 105 (1994).

  125. [125]. See Douglas H. Ginsburg & Hyland Hunt, The Prosecutor and Post-Conviction Claims of Innocence: DNA and Beyond?, 7 Ohio St. J. Crim. L. 771, 778 (2010) (“Often the prosecutor is, as a practical matter, the sole arbiter of whether a defendant has access to potentially exculpatory material, including DNA, and the prosecutor’s support or opposition may make or break the defendant’s chance at exoneration through whatever procedure remains available, such as executive clemency.”); Bruce A. Green & Ellen Yaroshefsky, Prosecutorial Discretion and Post-Conviction Evidence of Innocence, 6 Ohio St. J. Crim. L. 467, 502 (2009) (explaining how after conviction “[t]he prosecutor, not the judge or jury, is the key fact finder”); Fred C. Zacharias, The Role of Prosecutors in Serving Justice After Convictions, 58 Vand. L. Rev. 171, 173 (2005); see also Brandon L. Garrett, Judging Innocence, 108 Colum. L. Rev. 55, 118–19 (2008) (noting that prosecutors initiated or consented to DNA testing in a significant subset of cases where a convicted person was exonerated).

  126. [126]. See generally John F. Pfaff, Escaping from the Standard Story: Why the Conventional Wisdom on Prison Growth Is Wrong, and Where We Can Go from Here, 26 Fed. Sent’g Rep. 265 (2014).

  127. [127]. See Bennett L. Gershman, Reflections on Brady v. Maryland, 47 S. Tex. L. Rev. 685, 688 n.18 (2006); Editorial, 152 Innocents, Marked for Death, N.Y. Times (Apr. 13, 2015), https://www. (“[A]s advances in DNA analysis have accelerated the pace of exonerations, it has also become clear that prosecutorial misconduct is at the heart of an alarming number of these cases. In the past year alone, nine people who had been sentenced to death were released—and in all but one case, prosecutors’ wrongdoing played a key role.”); cf. Ken Armstrong & Maurice Possley, The Verdict: Dishonor (pt. 1), Chi.
    Trib. (Jan. 11, 1999),
    html (discussing a study of 11,000 homicide convictions between 1963 and 1999 showing that 381 were reversed for misconduct); Steve Weinburg, Breaking the Rules: Who Suffers When a Prosecutor Is Cited for Misconduct?, Ctr. for Pub. Integrity,
    5517/breaking-rules (last updated May 19, 2014, 12:19 PM) (“In 28 cases, involving 32 separate defendants, misconduct by prosecutors led to the conviction of innocent individuals who were later exonerated, the Center found.”); Mississippi Innocence (The Southern Documentary Project 2011) (documentary film about two men wrongly convicted of murder due to conduct by a prosecutor who used flawed bite mark evidence in both cases and then refused to dismiss one case after DNA evidence implicated someone else); The Thin Blue Line (Miramax Films 1988) (documentary about a man wrongly convicted of murder because of the overzealousness of the Dallas district attorney).

  128. [128]. See Medwed, supra note 8, at 129 (“Empirical proof suggests that prosecutors have consented to DNA tests in less than fifty percent of the cases in which testing later exonerated the inmate. Likewise, qualitative evidence of prosecutorial indifference and, on occasion, hostility to even the most meritorious of post-conviction innocence claims is alarming. Some prosecutors have continued to fight these claims despite clear evidence, including DNA test results, exculpating the defendant ... .” (footnotes omitted)).

  129. [129]. Kagan, supra note 105, at 93–96; cf. William J. Stuntz, The Uneasy Relationship Between Criminal Procedure and Criminal Justice, 107 Yale L.J. 1, 28 (1997) (showing that the increased emphasis on criminal procedure defenses helps wealthier defendants at the expense of poorer ones).

  130. [130]. See generally Charles Fried, The Lawyer as Friend: The Moral Foundations of the Lawyer-Client Relation, 85 Yale L.J. 1060 (1976).

  131. [131]. Monroe H. Freedman & Abbe Smith, Understanding Lawyers’ Ethics 20–22, 50 (5th ed. 2016).

  132. [132]. Daniel Markovits, A Modern Legal Ethics: Adversary Advocacy in a Democratic Age 139–42 (2008); Daniel Markovits, Adversary Advocacy and the Authority of Adjudication, 75 Fordham L. Rev. 1367, 1371 (2006) [hereinafter Markovits, Adversary Advocacy].

  133. [133]. See supra note 132.

  134. [134]. In a system where victim-funded lawyers prosecute the case, which historically was the American practice, there would be a client-focused argument for adversarialism by prosecutors. But today we have a system of public prosecution. See supra Part II.B; see also Nat’l Prosecution Standards § 1-1.2 (Nat’l Dist. Attorneys Ass’n, 3d ed. 2009) (“A prosecutor should zealously protect the rights of individuals, but without representing any individual as a client.”); Paul G. Cassell, Recognizing Victims in the Federal Rules of Criminal Procedure: Proposed Amendments in Light of the Crime Victims’ Rights Act, 2005 BYU L. Rev. 835, 838–50 (explaining the limited role of victims in the criminal justice system, and the history of the victims’ rights movement).

  135. [135]. See Model Code of Prof’l Responsibility EC 7-13 (Am. Bar Ass’n 1980) (“[D]uring trial the prosecutor is not only an advocate but he[/she] also may make decisions normally made by an individual client, and those affecting the public interest should be fair to all ... .”); see also Russell M. Gold, “Clientless” Prosecutors, 51 Ga. L. Rev. 693, 695 (2017).

  136. [136]. Professor Fried seems to endorse such a justification, although he does not develop the argument. See Fried, supra note 130, at 1076. Professor Markovits, on the other hand, rejects the idea. See Markovits, Adversary Advocacy, supra note 132, at 1371 n.18.

  137. [137]. One might try to argue that district-attorney elections are a form of client control, but they do not provide direct control over specific prosecution decisions.

  138. [138]. There is a robust and fascinating debate on the proper role of civil-government lawyers and whether they should act as adversaries or should pursue justice in a broader sense. Compare Michael Asimow & Yoav Dotan, Hired Guns and Ministers of Justice: The Role of Government Attorneys in the United States and Israel, 49 Isr. L. Rev. 3, 3–11 (2016), Steven K. Berenson, Public Lawyers, Private Values: Can, Should, and Will Government Lawyers Serve the Public Interest?, 41 B.C. L. Rev. 789, 790–91 (2000), and James R. Harvey III, Note, Loyalty in Government Litigation: Department of Justice Representation of Agency Clients, 37 Wm. & Mary L. Rev. 1569, 1570–73 (1996), with Catherine J. Lanctot, The Duty of Zealous Advocacy and the Ethics of the Federal Government Lawyer: The Three Hardest Questions, 64 S. Cal. L. Rev. 951, 952–53 (1991), Geoffrey P. Miller, Government Lawyers’ Ethics in a System of Checks and Balances, 54 U. Chi. L. Rev. 1293, 1294–97 (1987), and Peter L. Strauss, The Internal Relations of Government: Cautionary Tales from Inside the Black Box, 61 Law & Contemp. Probs. 155, 156–59 (1998). I do not venture to take a position in this debate here—the argument of this Article is confined to the unique context of prosecutors.

  139. [139]. See Daniel Epps, The Consequences of Error in Criminal Justice, 128 Harv. L. Rev. 1065, 1067–68 (2015) (“‘[B]etter that ten guilty persons escape, than that one innocent suffer’ is perhaps the most revered adage in the criminal law, exalted by judges and scholars alike as ‘a cardinal principle of Anglo-American jurisprudence.’” (alteration in original) (footnotes omitted)); Alexander Volokh, n Guilty Men, 146 U. Pa. L. Rev. 173, 174 (1997) (discussing the “Blackstone ratio”).

  140. [140]. See U.S. Const. amends. IV, V, VI.

  141. [141]. See Green, supra note 16, at 612–13, 634.

  142. [142]. See, e.g., Berger v. United States, 295 U.S. 78, 88 (1935); Handford v. United States, 249 F.2d 295, 296 (5th Cir. 1957); State v. Monday, 257 P.3d 551, 556 (Wash. 2011) (en banc); Model Code of Prof’l Responsibility EC 7-3 (Am. Bar Ass’n 1980).

  143. [143]. See, e.g., Monroe H. Freedman, Our Constitutionalized Adversary System, 1 Chap. L. Rev. 57, 59 (1998) (“As expressed by law professors at the University of Havana, ‘the first job of a revolutionary lawyer is not to argue that his client is innocent, but rather to determine if his client is guilty and, if so, to seek the sanction which will best rehabilitate him.’ Similarly, a Bulgarian attorney began his defense in a treason trial by noting that ‘[i]n a Socialist state there is no division of duty between the judge, prosecutor, and defense counsel ... . The defense must assist the prosecution to find the objective truth in a case.’ In that case, the defense attorney ridiculed his client’s defense, and the client was convicted and executed.” (alterations in original) (footnotes omitted)).

  144. [144]. See, e.g., Robert Nozick, Anarchy, State, and Utopia (1974); John Rawls, A Theory of Justice (1971).

  145. [145]. It also seems a bit strange to require that prosecutors resolve deep questions of political theory.

  146. [146]. See Ronald Dworkin, Law’s Empire 337–54 (1986) (defending an approach to legal interpretation in which the judge first imagines that all of the existing legal structures and prior decisions in the system taken together create a coherent hierarchy of principles, and then chooses the interpretation that best fits with their own normative interpretation of that hierarchy).

  147. [147]. See Richard M. Re, Promising the Constitution, 110 Nw. U. L. Rev. 299, 306–12 (2016) (discussing how the oath of office establishes an official duty to adhere to the Constitution). See generally Fish, supra note 44 (arguing that prosecutors have independent obligations to protect constitutional rights).

  148. [148]. This is one possible approach to what I describe below as a role ethic of value weighing. See infra Part IV.C.

  149. [149]. For an extended defense of this type of positivist account of law’s function, see generally Scott J. Shapiro, Legality (2011).

  150. [150]. Norman W. Spaulding, The Rule of Law in Action: A Defense of Adversary System Values, 93 Cornell L. Rev. 1377, 1392 (2008); accord Epps, supra note 27, at 826–28.

  151. [151]. Indeed, Spaulding notes this fact. See Spaulding, supra note 150, at 1392 (“I find that prospect chilling, just as chilling as I find the exercise of unfettered discretion and moral activism by prosecutors, the one group of lawyers that professional standards enjoin to do ‘justice’ rather than just diligently pursue the interests of private clients.”).

  152. [152]. See infra Part III.D.

  153. [153]. See Langbein, Land Without Plea Bargaining, supra note 105, at 211 (“Obviously, the German rule of compulsory prosecution of serious crime is no happenstance. The statutory standards, limitations, and remedies have been meticulously designed to fit the institutional structure and to serve the larger policies of the German criminal justice system.”).

  154. [154]. See William J. Stuntz, The Pathological Politics of Criminal Law, 100 Mich. L. Rev. 505, 510 (2001) (noting how enforcement discretion causes legislators to err on the side of overcriminalizing through broad liability rules); see also Parrillo, supra note 68, at 258 (claiming that the shift to paying prosecutors fixed salaries was in part intended to give them autonomy to “‘sand off’ the hard edges” of the criminal law through discretion).

  155. [155]. See Stuntz, supra note 154, at 510.

  156. [156]. See Kenneth Culp Davis, Discretionary Justice A Preliminary Inquiry 164 n.4 (1977) (“[R]eform of existing criminal statutes is an obvious prerequisite to substantially full enforcement. Anything approaching full enforcement of present statutes would be unthinkable.”); Roger A. Fairfax, Jr., Grand Jury Discretion and Constitutional Design, 93 Cornell L. Rev. 703, 732 (2008); Russell M. Gold, Promoting Democracy in Prosecution, 86 Wash. L. Rev. 69, 91–93 (2011).

  157. [157]. See Jessica Bulman-Pozen & David E. Pozen, Uncivil Obedience, 115 Colum. L. Rev. 809, 811 (2015); Lynch, supra note 116, at 2136–37.

  158. [158]. See Luna & Wade, supra note 93, at 1432–33.

  159. [159]. Id. at 1468–74.

  160. [160]. See W. Bradley Wendel, Lawyers and Fidelity to Law 83 (2010) [hereinafter Wendel, Fidelity to Law]; W. Bradley Wendel, Civil Obedience, 104 Colum. L. Rev. 363, 412 n.181 (2004); W. Bradley Wendel, Professionalism as Interpretation, 99 Nw. U. L. Rev. 1167 (2005) [hereinafter Wendel, Professionalism as Interpretation].

  161. [161]. See Wendel, Professionalism as Interpretation, supra note 160, at 1168 (“Professionalism is a stance toward the law which accepts that a lawyer is not simply an agent of her client (although the lawyer–client relationship is obviously governed by the law of agency). Rather, in carrying out her client’s lawful instructions, a lawyer has an obligation to apply the law to her client’s situation with due regard to the meaning of legal norms, not merely their formal expression.”).

  162. [162]. See generally Adrian Vermeule, The Invisible Hand in Legal and Political Theory, 96 Va. L. Rev. 1417 (2010) (describing the structure of invisible-hand arguments).

  163. [163]. See Luban, supra note 111, at 1758 (“[T]he injunction to zealous advocacy is at its most demanding in the criminal defense context, where the liberal argument for overprotecting rights against the state gives added heft to the norm.”).

  164. [164]. See Heumann, supra note 4, at 111; Alschuler, supra note 27, at 60–61.

  165. [165]. See Epps, supra note 27, at 832–33 (discussing this as a potential benefit of adversarial prosecution).

  166. [166]. See Tamara Rice Lave, The Prosecutor’s Duty to “Imperfect” Rape Victims, 49 Tex. Tech. L. Rev. 219, 223 (2016) (arguing that prosecutors should bring sexual-assault cases where the victim has credibility problems); see also Alschuler, supra note 27, at 63 (discussing the problem of one-witness cases).

  167. [167]. See Zacharias, supra note 7, at 56 (“To the extent that the adversary system works according to theory, government lawyers promote justice by playing the same role at trial as private advocates. They contribute to truth by defending their own factual hypotheses and contesting those of their opponents. Prosecutors help courts assess defendants’ rights; the claims of defendants’ champions must be contested to determine their validity.”).

  168. [168]. See Berenson, supra note 138, at 805–07; Vermeule, supra note 162, at 1443–44 (“If the premise is that the defense lawyer may be a zealous advocate because a system of competitive production of evidence by parties best promotes truth overall, it is not obvious how one can go on to deny that the other party, namely the prosecutor, should be equally entitled to produce evidence in a competitive and partisan fashion. A system in which prosecutors but not defense lawyers have an obligation to present evidence impartially to the tribunal might be the worst of all possible worlds.”).

  169. [169]. See Frankel, supra note 27, at 87–100; Pizzi, supra note 103, at 117–39; Melilli, supra note 27, at 691–97; Simon, The Ethics of Criminal Defense, supra note 111, at 1704­–05; Slobogin, supra note 105, at 702–10; Gordon Van Kessel, Adversary Excesses in the American Criminal Trial,
    67 Notre Dame L. Rev. 403, 450–51 (1992).

  170. [170]. One potential use of the “dual role” model could be to constrain prosecutors’ ability to undermine truth-seeking in litigation. Indeed, the ABA’s Criminal Justice Standards (which are non-binding) contain a number of provisions designed with this purpose in mind, including a suggestion that prosecutors should not undermine truthful witnesses or use delay as a litigation tactic. See ABA Standards for Criminal Justice: Prosecution & Def. Function §§ 3-2.9(b) to -5.7(b) (Am. Bar Ass’n, 3d ed. 1993). But it is questionable how effective such constraints are, given the powerful incentives that an adversarial role ethic generates to achieve convictions. See infra Part IV.A.

  171. [171]. See Melilli, supra note 27, at 694 (“The fact is that, because we do not know ‘the truth’ in any absolute sense, we cannot measure the truth-revealing capabilities of the adversary system except on the basis of anecdotal accounts and intuition. But the idea that trial by combat, whatever the weapons, will reliably produce truth is both counter-intuitive and contradicted by experience.” (footnotes omitted)).

  172. [172]. Cf. Margaret H. Lemos, Democratic Enforcement? Accountability and Independence for the Litigation State, 102 Cornell L. Rev. 929, 943–64 (2017) (distinguishing between administrative regulation as adjudication and administrative regulation as legislation). See generally Dworkin, supra note 146; Shapiro, supra note 149; Wendel, Fidelity to Law, supra note 160.

  173. [173]. It should be noted that the ethics codes recognize that private attorneys also have duties beyond those owed to their clients, including duties to the court and to the public. See Nancy J. Moore, “In the Interests of Justice”: Balancing Client Loyalty and the Public Good in the Twenty-First Century, 70 Fordham L. Rev. 1775, 1782 n.33 (2002). Thus, while private lawyers are predominantly partisans of their clients in our system, they too perform multiple roles.

  174. [174]. See generally Epps, supra note 27 (providing a partial defense of purely adversarial prosecution).

  175. [175]. See Freedman & Smith, supra note 131, at 87; Wendel, Fidelity to Law, supra note 160, at 6.

  176. [176]. This is certainly not to say that real-world prosecutors are indifferent to the innocence of those they prosecute. Here I am only discussing the pure form of one possible role.

  177. [177]. See generally Ronald Wright & Marc Miller, The Screening/Bargaining Tradeoff, 55 Stan. L. Rev. 29 (2002).

  178. [178]. See Melilli, supra note 27, at 691–97 (discussing prosecutors who take the view that they have no role in deciding guilt or innocence).

  179. [179]. Cf. Medwed, supra note 8, at 132–69 (discussing prosecutors’ motivations for opposing post-conviction innocence claims).

  180. [180]. U.S. Const. amend. VI.

  181. [181]. Id.

  182. [182]. Id. amend. IV.

  183. [183]. Brady v. Maryland, 373 U.S. 83, 87 (1963).

  184. [184]. See generally Carol S. Steiker, Counter-Revolution in Constitutional Criminal Procedure? Two Audiences, Two Answers, 94 Mich. L. Rev. 2466 (1996) (showing that the Supreme Court has, in recent decades, curtailed remedies for violations of criminal procedure rights while leaving the scope of those rights intact).

  185. [185]. See John H. Langbein, Torture and Plea Bargaining, 46 U. Chi. L. Rev. 3, 3–8 (1978) (analogizing the American plea-bargaining system to the medieval practice of confession through torture, in that both systems established burdensome procedural rights but then developed informal workarounds that undermined those rights); John Pfaff, Why Do Prosecutors Go After Innocent People?, Wash. Post (Jan. 21, 2016),
    wp/2016/01/21/why-do-prosecutors-go-after-innocent-people (“In real life, once a prosecutor decides to file felony charges against a defendant, that defendant will almost certainly be convicted ... .”).

  186. [186]. See Alschuler, supra note 27, at 68 (A defense attorney stated: “Settling cases is a game, like basketball. The only difference is that there are no rules in plea bargaining.”).

  187. [187]. See supra note 60 and accompanying text.

  188. [188]. See supra Part III.A.

  189. [189]. See Daniel Markovits, Legal Ethics from the Lawyer’s Point of View, 15 Yale J.L. & Human. 209, 263 (2003) (“[T]he adversary lawyer must still explain her professional activities to herself so as to make them fit (or at least not clash) with her more general first-personal ethical aims and ideals, her first-personal account of the life of which she wishes to claim authorship.”); Smith, supra note 27, at 378–81 (describing prosecutors’ passionate belief that they are doing justice).

  190. [190]. See supra Part II.B.

  191. [191]. Daniel Epps has suggested that more adversarial prosecution may cause political feedback loops that result in fewer, narrower, or less punitive criminal laws. See Epps, supra note 27, at 828–31; cf. Bulman-Pozen & Pozen, supra note 157, at 829–30. In Germany, by analogy, which has a rule of mandatory prosecution, legislative reforms in the 1970s moved several categories of cases out of the criminal law system. See Klaus Sessar, Prosecutorial Discretion in Germany, in The Prosecutor, supra note 71, at 255, 256–60, 272. However, the last several decades of criminal legislation in the United States, and especially legislatures’ appetite for lengthy mandatory sentences, make me skeptical of this suggestion in the American context.

  192. [192]. Margaret H. Marshall, The Promise of Neutrality Reflections on Judicial Independence, 36 Hum. Rts. 3, 4 (2009) (“[W]e expect judges to adjudicate, not advocate. Implicit in our constitutional compact is the guarantee that judges will give us a fair hearing; that they will treat each litigant who comes before them nonpreferentially ... .”).

  193. [193]. The principle of prosecutorial positivism is quite similar to Brad Wendel’s more general idea that the lawyer’s professional role requires engaging in neutral legal interpretation. See Wendel, Professionalism as Interpretation, supra note 160, at 1177.

  194. [194]. See Bruce A. Green & Fred C. Zacharias, Prosecutorial Neutrality, 2004 Wis. L. Rev. 837, 860–70 (exploring several different conceptions of prosecutorial neutrality, including the idea of prosecutorial neutrality as non-partisanship).

  195. [195]. See supra Part III.D.

  196. [196]. See Max Weber, Economy and Society 226 (Guenther Roth & Claus Wittich eds., 1968).

  197. [197]. See Arthur Isak Applbaum, Professional Detachment: The Executioner of Paris, 109 Harv. L. Rev. 458, 458­–60 (1995).

  198. [198]. Id. at 459–60.

  199. [199]. Id. at 460.

  200. [200]. Cf. Lynch, supra note 116, at 2148–51 (arguing that our criminal justice system is effectively an inquisitorial system managed by prosecutors).

  201. [201]. See supra Part III.C.

  202. [202]. See Rachel E. Barkow, Organizational Guidelines for the Prosecutor’s Office, 31 Cardozo L. Rev. 2089, 2116–17 (2010); Green & Yaroshefsky, supra note 125, at 481.

  203. [203]. See generally Bennett Gershman, Prosecution Stories (2017) (discussing a case in which the author, a prosecutor, not only confessed error for a conviction stemming from an unlawful search, but even argued for reversal through two levels of appeal); Gold, supra note 40 (arguing that prosecutors should play this role in the Fourth Amendment context).

  204. [204]. See Fish, supra note 44, at 294–95 (discussing various sources of prosecutors’ disclosure obligations).

  205. [205]. See, e.g., CPS, The Code for Crown Prosecutors 6–10 (2013) (UK) (describing a seven-part test for British prosecutors to decide whether or not a case should go forward); Memorandumfrom Richard Thornburgh, Att’y Gen., to U.S. Attorneys, Plea Policy for Federal Prosecutors (Mar. 13, 1989), reprinted in 1 Fed. Sent’g Rep. 421, 421–22 (1989) (memorandum requiring federal prosecutors to charge the most serious readily provable offense).

  206. [206]. See Stephanos Bibas, The Need for Prosecutorial Discretion, 19 Temp. Pol. & C.R. L. Rev. 369 (2010) (arguing that it is necessary for prosecutors to exercise equitable discretion); Bowers, supra note 119, at 1660 (arguing that prosecutors are bad at exercising equitable discretion).

  207. [207]. See generally U.S. Sentencing Guidelines Manual (U.S. Sentencing Comm’n 2016).

  208. [208]. A number of scholars have explicitly rejected the idea of positivist neutrality for American prosecutors. See, e.g., Green, supra note 27, at 129–30 (“A prosecutor is not supposed to be neutral and detached. It is not the prosecutor’s duty to present both sides of a criminal case. Nor is it the prosecutor’s duty to urge the jury to draw inferences in favor of the defendant.”); Pizzi, supra note 109, at 1349–55; Zacharias, supra note 7, at 51–52.

  209. [209]. See David Hoffman, Fifty Resolutions in Regard to Professional Deportment, in 2 David Hoffman, A Course of Legal Study 752–75 (William S. Hein & Co. 1968) (1936); Fred C. Zacharias & Bruce A. Green, Reconceptualizing Advocacy Ethics, 74 Geo. Wash. L. Rev. 1, 2 (2005).

  210. [210]. See, e.g., Model Rules of Prof’l Conduct r. 3.8 cmt. 1 (Am. Bar Ass’n 2016) (“A prosecutor has the responsibility if a minister of justice and not simply that of an advocate.”); supra notes 17–22 and accompanying text.

  211. [211]. See Kate Stith & José A. Cabranes, Judging Under the Federal Sentencing Guidelines, 91 Nw. U. L. Rev. 1247, 1252–54 (1997).

  212. [212]. See generally Wright & Miller, supra note 177 (discussing the tradeoff between using resources to screen out cases and mechanically pushing cases forward to the plea-bargaining stage).

  213. [213]. U.S. Dep’t of Justice, United States Attorneys’ Manual: Principles of Federal Prosecution § 9-27.220 (2015) [hereinafter United States Attorneys’ Manual].

  214. [214]. Standards of Criminal Justice Relating to the Prosecution Function Standard 3-4.3 (Am. Bar Ass’n 2015).

  215. [215]. See, e.g., United States Attorneys’ Manual, supra note 213, § 9-11.233 (requiring disclosure to the grand jury of any “substantial evidence that directly negates the guilt of a subject of the investigation”).

  216. [216]. See Barkow, supra note 202, at 2116.

  217. [217]. See United States Attorneys’ Manual, supra note 213, § 9-13.410; Fish, supra note 44, at 298–99.

  218. [218]. United States Attorneys’ Manual, supra note 213, § 9-13.410.

  219. [219]. See, e.g., Medwed, supra note 46, at 47 (noting that Milwaukee District Attorney John Chisholm implemented open file discovery around 2010); Editorial, Justice and Open Files, N.Y. Times (Feb. 26, 2012),

  220. [220]. See, e.g., Leah M. Litman & Luke C. Beasley, Jurisdiction and Resentencing: How Prosecutorial Waiver Can Offer Remedies Congress Has Denied, 101 Cornell L. Rev. Online 91, 103–12 (2016) (arguing that the Department of Justice can waive restrictions imposed by the Antiterrorism and Effective Death Penalty Act on successive petitions).

  221. [221]. See Joseph Goldstein, Prosecutor Deals Blow to Stop-and-Frisk Tactic, N.Y. Times (Sept. 25, 2012), (explaining how the Bronx district attorney’s office is attempting to diminish the number of people arrested for trespassing).

  222. [222]. See Memorandum from James M. Cole, Deputy Att’y Gen., to U.S. Attorneys, Guidance Regarding Marijuana Enforcement 2 (Aug. 29, 2013),

  223. [223]. Memorandum from Eric Holder, Att’y Gen., to Dep’t of Justice Attorneys, Guidance Regarding § 851 Enhancements in Plea Negotiations (Sept. 24, 2014),

  224. [224]. See Robert J. Smith, The Geography of the Death Penalty and its Ramifications, 92 B.U. L. Rev. 227, 232–33 (2012).

  225. [225]. Deborah L. Rhode, Ethical Perspectives on Legal Practice, 37 Stan. L. Rev. 589, 623 (1985).

  226. [226]. Edmund Burke, Speech to the Electors of Bristol (Nov. 3, 1774), reprinted in 2 The Works of Edmund Burke 7 (1839).

  227. [227]. See Dworkin, supra note 146, at 354. For a cogent argument that government attorneys should not act like Dworkin’s Judge Hercules, see W. Bradley Wendel, Sally Yates, Ronald Dworkin, and the Best View of the Law, 115 Mich. L. Rev. Online 78, 79–80 (2017). I do not believe that the logic of Wendel’s argument extends to prosecutors, for the reasons discussed supra Part III.B–C.

  228. [228]. Prosecutors adopting this logic would act very differently in different legal systems. For example, in communist Bulgaria it seems defendants were given very few procedural rights. See Freedman, supra note 143, at 59. On a smaller scale, one might imagine prosecutors in different American states discerning different moral commitments in their respective systems, based on differences in statutory, constitutional, and judge-made law.

  229. [229]. Of course, judges are elected in many states as well. But there is a deep and intractable conflict between the logic of electoral politics and the role we ascribe to judges. See David E. Pozen, Judicial Elections as Popular Constitutionalism, 110 Colum. L. Rev. 2047, 2099–104 (2010).

  230. [230]. One might also break this distinction down into different aspects of the prosecutorial role. For instance, when determining what types of crimes to prioritize, prosecutors might have wide discretion, but when determining the proper scope of procedural rights, they might approach their role more like judges. See Richman, supra note 49, at 960–65 (arguing that prosecutors’ decisions about the allocation of enforcement resources should reflect the community’s preferences); cf. Lemos, supra note 172, at 933–34 (arguing that administrative agencies should be more politically accountable when making general priority-setting decisions, and more politically independent when making concrete decisions about particular cases).

  231. [231]. See supra notes 17–28 and accompanying text.

  232. [232]. Brady v. Maryland, 373 U.S. 83, 87 (1963).

  233. [233]. Indeed, judges in the American system have this same basic duality of role. When there is an external source of law that supplies a legal rule they take on an ethic of neutrality (e.g., when applying a statute), and when they exercise discretion they take on an ethic of value selection (e.g., when sentencing a defendant without guidelines).

  234. [234]. Recall that in inquisitorial systems like those in France and Germany, judges exercise much more discretion, and prosecutors much less, than in our own system. This is ironic, because French and German prosecutors embrace neutrality as their dominant role ethic. See supra Part II.D.

  235. [235]. See Bowers, supra note 119, at 1658–59.

  236. [236]. See Susan R. Klein, Enhancing the Judicial Role in Criminal Plea and Sentence Bargaining, 84 Tex. L. Rev. 2023, 2029 (2006).

  237. [237]. See, e.g., Davis, supra note 39, at 20–21.

  238. [238]. See Abraham S. Goldstein, The Passive Judiciary: Prosecutorial Discretion and the Guilty Plea 52–75 (1981) (arguing for a common law of prosecution, with active judicial review of charges, dismissals, and pleas).

  239. [239]. See Parrillo, supra note 68, at 258; Bibas, supra note 206, at 370 (“Even in a world of unlimited resources and sane criminal codes, discretion would be essential to doing justice. Justice requires not only rules but also fine-grained moral evaluations and distinctions.”).

  240. [240]. See Eric S. Fish, Sentencing and Interbranch Dialogue, 105 J. Crim. L. & Criminology 549, 569–73 (2015) (discussing how legislatures impose mandatory-minimum laws and binding sentencing guidelines that limit judicial discretion); Stuntz, supra note 154, at 530 (discussing legislators’ tendency to think of criminal laws in terms of symbolic stands).

  241. [241]. See Inmates of Attica Corr. Facility v. Rockefeller, 477 F.2d 375, 380 (2d Cir. 1973) (“[T]he problems inherent in the task of supervising prosecutorial decisions do not lend themselves to resolution by the judiciary. The reviewing courts would be placed in the undesirable and injudicious posture of becoming ‘superprosecutors.’”). Though it is notable that a number of states do permit judicial involvement in plea bargaining. See King & Wright, supra note 117, at 335. There does seem to be something of a direct tradeoff between the moral discretion of prosecutors and the moral discretion of judges. See, e.g., Kate Stith, The Arc of the Pendulum: Judges, Prosecutors, and the Exercise of Discretion, 117 Yale L.J. 1420, 1493–94 (2008) (discussing how federal sentencing law has shifted over the years from empowering judges to empowering prosecutors to once more empowering judges, depending on the prevailing framework).

  242. [242]. See Barkow, supra note 45, at 911–13; Uviller, supra note 27, at 1696–97.

  243. [243]. One criticism of this approach is that it diffuses responsibility within the office, and so actually undermines prosecutors’ case screening function. See Green & Roiphe, supra note 57, at 510–11; Melilli, supra note 27, at 687–90 (“In fact, at least in prosecutors’ offices burdened with large caseloads, detailed investigation and preparation of cases may not take place until trial is imminent. As noted, however, by that time the earlier screening stages may be accorded an unjustified deference interfering with objective evaluation.” (footnote omitted)).

  244. [244]. See supra notes 33–36 and accompanying text.

  245. [245]. Cf. Miriam H. Baer, Timing Brady, 115 Colum. L. Rev. 1, 10, 34–38 (2015) (arguing that prosecutors’ heightened adversarial incentives immediately before and during trials make Brady violations more likely than at earlier stages of a case).

  246. [246]. One important distinction is that, unlike prosecutors, agency bureaucrats in the United States do not have a highly articulated role morality. This is a quirky feature of American legal culture—lawyers have well-defined ethics, but non-lawyer government functionaries generally do not. This is not so in a number of other countries. For example, in the United Kingdom “Her Majesty’s Civil Service” has a deep professional commitment to neutral administration. See James Morrison, Essential Public Affairs for Journalists 103–04 (4th ed. 2015); Vanessa MacDonnell, The Civil Servant’s Role in the Implementation of Constitutional Rights, 13 Int’l J. Const. L. 383, 392 & n.72 (2015).

  247. [247]. See James K. Wilkens & Thomas A. Matthews, A Survey of Federal Tax Collection Procedure: Rights and Remedies of Taxpayers and the Internal Revenue Service, 3 Alaska L. Rev. 269, 270, 272 (1986).

  248. [248]. See Steve R. Johnson, Reasoned Explanation and IRS Adjudication, 63 Duke L.J. 1771, 1793–99 (2014).

  249. [249]. See Langer, supra note 116, at 291–98. For example, the Department of Justice’s internal policies with respect to the prosecution of business organizations are highly complex, and reflect a number of value judgments about the goals of a prosecution and the factors that should be weighed in going after a company and accepting a plea agreement. See United States Attorneys’ Manual, supra note 213, § 9-28.000.

  250. [250]. See Barry Friedman & Maria Ponomarenko, Democratic Policing, 90 N.Y.U. L. Rev. 1827, 1858 (2015) (making a similar proposal for police forces); cf. Richard A. Bierschbach & Stephanos Bibas, Notice-and-Comment Sentencing, 97 Minn. L. Rev. 1, 4–5 (2012) (proposing a notice-and-comment sentencing approach “for soliciting an array of information and viewpoints that bear on the public interest in sentencing”); Dan M. Kahan, Is Chevron Relevant to Federal Criminal Law?, 110 Harv. L. Rev. 469, 519 (1996) (“Courts should therefore promote formality by reserving deference for Justice Department readings that appear in publicly accessible formats.”).

  251. [251]. If the ethics codes changed but all else about prosecution remained the same, they would function as an ironic comment on the gap between ethical theory and reality. Cf. Christine M. Korsgaard, The Right to Lie: Kant on Dealing with Evil, 15 Phil. & Pub. Aff. 325, 325–26 (1986) (interpreting Immanuel Kant’s moral philosophy as an uncompromising form of ideal theory that makes it difficult to deal with real-world evils).

  252. [252]. Model Rules of Prof’l Conduct r. 3.8 (Am. Bar Ass’n 2016) (imposing obligations on prosecutors to only pursue charges if they believe there is probable cause, to inform the defendant of their right to counsel, to disclose evidence that a conviction was incorrect, and so on).

  253. [253]. Cf. Fish, supra note 240, at 554 (arguing that judges and sentencing commissions should engage in iterative dialogue over the policy choices contained in sentencing guidelines).

  254. [254]. See generally Lawrence Gene Sager, Fair Measure: The Legal Status of Underenforced Constitutional Norms, 91 Harv. L. Rev. 1212 (1978) (arguing that underenforced constitutional norms are legally binding on non-judicial actors, notwithstanding judges’ underenforcement of them).

  255. [255]. Abbate v. United States, 359 U.S. 187, 195 (1959); Bartkus v. Illinois, 359 U.S. 121, 138–39 (1959).

  256. [256]. United States Attorneys’ Manual, supra note 213, § 9-2.031; see Fish, supra note 44, at 239.

  257. [257]. Washington v. Davis, 426 U.S. 229, 246–48 (1976); cf. Sophia Z. Lee, Race, Sex, and Rulemaking: Administrative Constitutionalism and the Workplace, 1960 to the Present, 96 Va. L. Rev. 799, 875–77 (2010) (describing how the Federal Communications Commission enforced constitutional antidiscrimination principles that were broader than judicial doctrine); David A. Sklansky, Quasi-Affirmative Rights in Constitutional Criminal Procedure, 88 Va. L. Rev. 1229, 1292–99 (2002) (arguing that courts should expand constitutional criminal procedure rules through rules that are reversible by political branches, since that approach would foster interbranch dialogue).

  258. [258]. See Fish, supra note 240, at 571 (describing how legislatures enact “‘purposes of sentencing’ statutes” that influence judicial sentencing without dictating results in specific cases); Green & Zacharias, supra note 194, at 874 n.128 (“A rare example of a statute that does set forth standards of prosecutorial decision-making is the Juvenile Justice and Delinquency Prevention Act of 1974 ... .”).

  259. [259]. See Kate Stith, Principles, Pragmatism, and Politics: The Evolution of Washington State’s Sentencing Guidelines, 76 Law & Contemp. Probs. 105, 124–25 (2013); Ronald F. Wright, Sentencing Commissions as Provocateurs of Prosecutorial Self-Regulation, 105 Colum. L. Rev. 1010, 1031–33 (2005).

  260. [260]. I got this idea from Bennett Gershman, who presented it at a criminal justice conference in June of 2016.

  261. [261]. See Friedman & Ponomarenko, supra note 250, at 1834.

  262. [262]. See Medwed, supra note 46, at 25–26; see also Travis Baker, Prosecutor Invites Public Review of Charging Policies, Kitsap Sun (Jan. 30, 2003), (“Those invited to review the standards include defense lawyers, police officers, a crime victim, the family of an inmate, a drug therapist, a criminal justice educator, and representatives of business, the military, each local city, corrections, mental health, minorities and juvenile justice.”). I owe this example to a conversation with Stephanos Bibas.

  263. [263]. Brad Heath, Rules to Keep Federal Prosecutors in Line Revealed, USA Today (Mar. 30, 2015, 5:44 PM), It is not too difficult to perceive a bit of institutional self-interest on the part of Congress in this saga as well. See Craig S. Lerner, Legislators as the “American Criminal Class”: Why Congress (Sometimes) Protects the Rights of Defendants, 2004 U. Ill. L. Rev. 599, 635–40 (wonderfully cynical article pointing out that major pro-defendant legal reforms, to the extent they are ever enacted, frequently seem to happen when prosecutors go after politicians).

  264. [264]. See Susan R. Klein, Monitoring the Plea Process, 51 Duq. L. Rev. 559, 589–91 (2013).

  265. [265]. See Marc L. Miller & Ronald F. Wright, The Black Box, 94 Iowa L. Rev. 125, 192–93 (2008). This seems to be another example of dialogue-by-threat.

  266. [266]. State v. Lagares, 601 A.2d 698, 704 (N.J. 1992); see also State v. Vasquez, 609 A.2d 29, 32–33 (N.J. 1992) (applying Lagares).

  267. [267]. Lagares, 601 A.2d at 704.

  268. [268]. State v. Brimage, 706 A.2d 1096, 1107 (N.J. 1998); see Ronald F. Wright, Prosecutorial Guidelines and the New Terrain in New Jersey, 109 Penn. St. L. Rev. 1087, 1101–02 (2005); Wright, supra note 259, at 1031–33. Notably, the New Jersey Attorney General made major revisions to the Brimage guidelines in 2004 after consulting with judges, defense attorneys, and local prosecutors. Id. at 1032–33. This provides an example of threat-based dialogue creating policies that were subsequently updated through a more collaborative form of dialogue.

  269. [269]. See, e.g., David E. Pozen, The Leaky Leviathan: Why the Government Condemns and Condones Unlawful Disclosures of Information, 127 Harv. L. Rev. 512, 538–39 (2013) (discussing how the Department of Justice uses internal restrictions on subpoenaing reporters to try to prevent external bodies from imposing a reporter’s privilege).

  270. [270]. Cf. Meir Dan-Cohen, Decision Rules and Conduct Rules: On Acoustic Separation in Criminal Law, 97 Harv. L. Rev. 625, 635 (1984). This danger is especially pronounced in the current “dual role” structure, where part of prosecutor’s official role is to seek justice but their adversarial role dominates in practice. See supra Part II.A. See generally Lauren M. Ouziel, Legitimacy and Federal Criminal Enforcement Power, 123 Yale L.J. 2236 (2014) (arguing that the punitiveness of the federal system is connected to judges’ and juries’ perception of federal law enforcement as objective).

  271. [271]. See Baker, supra note 4, at 79; Medwed, supra note 8, at 153–56; Richman, supra note 49, at 967; Stuntz, supra note 154, at 533–39; Van Kessel, supra note 169, at 442 n.164; Wright
    & Miller, supra note 177, at 35 n.13 (quoting news reports in which prosecutors cite their conviction rates in seeking reelection). For an illustration of a prosecutor publicly promoting his conviction record, see Elizabeth Nolan Brown, Indiana Prosecutor Bradley Cooper Is ‘Proudly Over-Crowding our Prisons,’ Reason (Mar. 1, 2016, 11:30 AM),

  272. [272]. See Ronald F. Wright, How Prosecutor Elections Fail Us, 6 Ohio St. J. Crim. L. 581,
    601–05 (2009); David Schleicher & Elina Treyger, The Case Against the DA (unpublished manuscript) (on file with author); see also Juleyka Lantigua-Williams, Are Prosecutors the Key to Justice Reform?, Atlantic (May 18, 2016),
    are-prosecutors-the-key-to-justice-reform/483252 (“In Oregon, the local chapter of the American Civil Liberties Union took a look at the decade between 2004 and 2014, and determined that
    78 percent of district attorneys were elected in uncontested races.”).

  273. [273]. See generally Siddhartha Bandyopadhyay & Bryan C. McCannon, The Effect of the Election of Prosecutors on Criminal Trials, 161 Pub. Choice 141 (2014) (claiming that reelection pressure on prosecutors causes an increase in the number of jury trials and a decrease in average sentence in both trials and pleas); Sanford C. Gordon & Gregory A. Huber, The Effect of Electoral Competitiveness on Incumbent Behavior, 2 Q.J. Pol. Sci. 107 (2007); Gregory A. Huber & Sanford C. Gordon, Accountability and Coercion: Is Justice Blind when It Runs for Office?, 48 Am. J. Pol. Sci. 247 (2004) (using a study of judicial sentencing practices to argue that electoral pressure causes judges to give higher sentences).

  274. [274]. Two recent cases illustrate this point. In Bridgeport, Connecticut, a shooting at a public-housing project swung a mayoral election and consequently created a great deal of political pressure to bring the killers to justice. See Daniel Tepfer, Trial Begins in Gang Shooting Spree Case, Conn. Post (July 6, 2016, 5:02 PM), Similarly, the recent murder of an imam in New York has forced the Queens district attorney to respond to public outrage without compromising his strategic position in the case. See Alan Feuer, Detached but Sympathetic: Prosecutor’s Dual Response to Imam’s Killing, N.Y. Times (Aug. 19, 2016),

  275. [275]. See Richman, supra note 49, at 965.

  276. [276]. See Nina Agrawal, Philadelphia’s New District Attorney Isn’t Who You’d Expect. Is His
    Election a Sign of More Change to Come?
    , L.A. Times (Dec. 31, 2017, 5:00 AM), http:// (“Krasner joins a growing list of district attorneys around the country—including Eric Gonzalez in Brooklyn, Kim Foxx in Chicago and Kim Ogg in Houston—who have declared that their role isn’t simply to prosecute, but to protect defendants from the excesses of the criminal justice system.”).

  277. [277]. See supra note 45 and accompanying text.

  278. [278]. See, e.g., Bibas, supra note 115, at 986–87. See generally Tracey L. Meares, Rewards for Good Behavior: Influencing Prosecutorial Discretion and Conduct with Financial Incentives, 64 Fordham L. Rev. 851 (1995).

  279. [279]. See Bibas, supra note 115, at 1007–11; Miller & Wright, supra note 265, at 177–78 (“[I]n the radically decentralized prosecutorial services of the United States, it is the elected district attorney who contributes most powerfully to the norms that prosecutors pursue.” (footnote omitted)). A prosecutor once told me an anecdote that captures one small way that a head prosecutor can establish non-adversarial expectations for their line prosecutors. At a certain office, the top prosecutor would make congratulatory calls after a line prosecutor had finished presenting their evidence at trial, but before the jury had come back with a verdict. This was meant to underscore that the prosecutor’s job was not just about getting a conviction.

  280. [280]. See Kay L. Levine & Ronald F. Wright, Prosecution in 3-D, 102 J. Crim. L. & Criminology 1119, 1128–29, 1135–41, 1146 (2012) (reporting a qualitative study of several prosecutor’s offices and showing that different organizational choices, such as whether an office is hierarchical or whether it hires inexperienced prosecutors, cause the office’s prosecutors to be more team-oriented or more independent).

  281. [281]. See Pamela J. Utz, Two Models of Prosecutorial Professionalism, in The Prosecutor, supra note 71, at 99, 103–14 (contrasting the adversary culture of San Diego County’s DA office with the managerial culture of Alameda County’s).

  282. [282]. See, e.g., Green, supra note 16, at 607–10 (discussing the culture of the U.S. Attorneys’ Office for the Southern District of New York); Miller & Wright, supra note 265, at 162–64 (discussing offices in Milwaukee, San Diego, and Charlotte that are working with the Vera Institute to try to identify racially disproportionate charging practices).

  283. [283]. See, e.g., Radley Balko, In Louisiana Prosecutor Offices, a Toxic Culture of Death and Invincibility, Wash. Post (Apr. 6, 2015),
    =.c299d849a8a1; supra note 46 and accompanying text.

  284. [284]. See supra Part II.C.

  285. [285]. See supra notes 30–43.


Law Professor, University of California at Davis.

This Article is not meant to represent the views of any particular office or institution, only those of the author.