Theories of Prosecution

Theories of Prosecution

For decades, legal commentators sounded the alarm about the tremendous power wielded by prosecutors. Scholars went so far as to identify uncurbed prosecutorial discretion as the primary source of the criminal justice system’s many flaws. Over the past two years, however, the conversation shifted. With the emergence of a new wave of “progressive prosecutors,” scholars increasingly hail broad prosecutorial discretion as a promising mechanism for criminal justice reform.

The abrupt shift from decrying to embracing prosecutorial power highlights a curious void at the center of criminal justice thought. There is no widely accepted normative theory of the prosecutorial role. As a result, prosecutors are viewed as the criminal justice system’s free agents, deploying the powers of their offices as they see fit to serve constituents, public safety, or, most broadly, the cause of justice.

This Article uses the rapidly shifting views about prosecutors to explore normative theories of prosecution: What should prosecutors be doing? It highlights the emptiness of the current “do justice” model and proposes an alternative “servant-of-the-law” theory of prosecutorial behavior that could place real constraints on prosecutorial excess. It also explores ways in which a servant-of-the-law model could, perhaps counterintuitively, contribute much-needed theoretical grounding to the progressive prosecution movement.




Scholars view prosecutors as “the most powerful officials in the criminal justice system,”[1] and blame them for “[m]uch of what is wrong with American criminal justice.”[2] But there is typically something missing from the accounts. An abundant literature highlights the failings of the nation’s prosecutors.[3] Yet when it comes to setting out principles to govern how prosecutors should act, the commentary offers only platitudes.[4] A 1935 Supreme Court opinion famously reminds prosecutors that it is their special duty to ensure “that justice shall be done.”[5] After that, the consensus crumbles. Despite all the attention paid to prosecutors in recent years, the primary guidance on the prosecutorial function remains a timeworn Rorschach test.[6]

Commentators traditionally viewed the absence of concrete rules governing prosecutorial discretion as a blight on the criminal justice landscape, contributing to overly punitive and often racist criminal justice outcomes.[7] Yet in recent years, with the ascension of unapologetically “progressive prosecutors,” commentary on prosecutorial power developed a new dimension.[8] In 2017, for example, newly elected Orlando District Attorney Aramis Ayala announced that her office would no longer seek the death penalty.[9] Over one hundred prominent law professors and public officials signed an open letter of support. Efforts to override Ayala’s decision, they wrote, threatened prosecutors’ “extremely broad discretion to decide when, and how, to prosecute” and “compromise[d] the prosecutorial independence upon which the criminal justice system depends.”[10] Around the same time, Baltimore’s chief prosecutor, Marilyn Mosby, became a “proxy for a nation reeling with outrage and disbelief” when she swiftly initiated ultimately unsuccessful criminal charges against six police officers involved in the death of Freddie Gray.[11] In a “clip that would echo across the country,” Mosby explained to supporters, “I have heard your calls for ‘no justice, no peace!’”[12] Most recently, Mosby, along with other big-city prosecutors like District Attorneys Larry Krasner (Philadelphia) and Cyrus Vance (Manhattan) announced that their offices would no longer enforce marijuana possession laws.[13] Krasner explained the decision as “the right thing to do”; Vance’s stated goal was “to reduce inequality and unnecessary interactions with the criminal justice system.”[14] Mosby highlighted the laws’ ineffectiveness: “When I ask myself: Is the enforcement and prosecution of marijuana possession making us safer as a city? [T]he answer is emphatically ‘no.’”[15]

These prosecutors are not aberrations. They are prominent representatives of a national movement to leverage prosecutorial power to achieve criminal justice reform.[16] A recent New York Times editorial captures the excitement, embracing the new wave of “state and local prosecutors who are open to rethinking how they do their enormously influential jobs.”[17] In light of legislative obstacles, the Times editorialized, this “wiser generation” of prosecutors is “the best chance for continued reform.”[18] These developments bring renewed urgency to a long-unanswered question: What is the role of the American prosecutor? Given the centrality of prosecutors to academic writing of the past decade, the lack of an answer is striking.

The curious absence of a normative theory of prosecutorial behavior is best explained by the runaway success of the vacuous ideal that reigns in its place. The conventional view of the prosecutorial role derives from an iconic passage in Berger v. United States.[19] There, a unanimous Supreme Court rebuked a prosecutor who made “improper insinuations and assertions calculated to mislead the jury.”[20] A prosecutor, the Court explained, represents “a sovereignty whose obligation to govern impartially is as compelling as its obligation to govern at all; and whose interest, therefore, in a criminal prosecution is not that it shall win a case, but that justice shall be done.”[21]

Although earlier statements to similar effect can be found,[22] jurists, attorneys, and commentators seize upon the Berger passage as “the most authoritative and eloquent description in U.S. law of the role of the prosecutor in administering criminal justice.”[23] The American Bar Association (ABA) adopts this formulation (and little else) in its Model Rules of Professional Conduct, instructing prosecutors to act as “minister[s] of justice.”[24] The largest association of state prosecutors, the National District Attorneys Association (NDAA), similarly describes the prosecutor as “an independent administrator of justice.”[25] Following these organizations’ lead, the “legal profession has left much of a prosecutor’s day-to-day decisionmaking unregulated, in favor of this catch-all ‘seek justice’ admonition.”[26]

It is hard to object to Berger’s iconic description of prosecutors. No one can deny the appeal of justice, Merriam-Webster’s 2018 “word of the year.”[27] Federal prosecutors work in the “Department of Justice.”[28] Superheroes operate out of the “Hall of Justice.”[29] School children pledge allegiance to a nation dedicated to “justice for all.”[30] It’s a great slogan. The problem is that philosophers have been trying to define “justice” for thousands of years and report little progress.[31] John Stuart Mill considered justice to be a placeholder for other considerations, which explained why “so many things appear either just or unjust, according to the light in which they are regarded.”[32] This observation becomes particularly salient in the grey landscapes of the criminal law, where justice has little uncontested content. Reformers urge prosecutors to seek justice by dispensing mercy,[33] while their opponents wield the same slogan to endorse severity.[34] As Abbe Smith points out: “The concept could not be more ambiguous and subject to multiple interpretations.”[35] Other commentators pile on: “When the ABA advises prosecutors to act as ‘ministers of justice’ or ‘administrators of justice,’ it is using juris-babble that is practically meaningless to prosecutors and to the ABA itself.”[36]

Scholars traditionally sought to temper prosecutors’ broad discretion to “do justice” through calls for regulation. In fact, the scarcity of rules governing prosecutorial decisions is a central criminal justice lament.[37] Commentators often cite a lack of “political will” to regulate these powerful actors,[38] but disagreement about what prosecutorial regulations should say is an equally compelling explanation. As David Sklansky wryly points out, “[u]nless they are co-authors, no two scholars ever propose the same metric for prosecutorial effectiveness.”[39]

Most entries in the academic literature about prosecutors fit comfortably within a familiar debate between rules and standards. Some commentators seek to rein in prosecutorial discretion with strict rules.[40] Others disagree, contending that the multifaceted and fact-bound prosecutorial function requires malleable standards.[41] There is a lively derivative discussion about whether new rules (or standards) should be operationalized through judicial oversight, legislative action, ethics boards, or internal guidelines.[42] These are important questions. A central contention of this Article, however, is that all sides of the debate skip an essential step. Whether we think prosecutors should be governed by rules or standards, and regardless of where we think those commands should originate, we must begin with a normative principle. That principle could either generate standards to govern prosecutorial discretion, or it could provide a starting point from which to derive enforceable rules. Right now, the starting point (“justice”) is an analytical dead end. It offers neither a meaningful standard to govern prosecutors, nor a useful guideline for generating specific rules. This core theoretical failing, more than any other factor, explains why academics, judges, and practitioners have made so little progress articulating concrete guidance for prosecutorial behavior.[43]

The problems go beyond accountability. Without a clear consensus about what to expect of prosecutors, legislators struggle to maintain the appropriate balance in the criminal law. Prosecutor offices are populated by a wide variety of attorneys, over twenty-five thousand of them,[44] all doing their own justice thing.[45] Even individual prosecutors may oscillate in their assessment of what approach best serves justice, depending on the suspected crime or characteristics of the offender. If each prosecutor pursues justice as they see it, the same law will be too lenient in one county and too severe in another.[46] Capital punishment presents the most extreme example, with the likelihood of a death sentence depending as much on the prosecutor overseeing the case as the characteristics of the crime.[47]

There is, then, a strong case for developing a normative theory of the prosecutorial role. But what? There are numerous justice themes that prosecutors invoke in the course of their duties: public safety (“tough justice”), serving constituents (“popular justice”), and, most recently, ending mass incarceration (“social justice”).[48] But these rhetorical appeals are typically too indeterminate to generate concrete guidance for prosecutors deciding whether to initiate and how to pursue specific cases.[49] The law is the most likely place to look for this kind of guidance—and prosecutors sometimes invoke their adherence to the law to explain unpopular decisions.[50] Yet scholars vigorously resist such an approach, urging prosecutors to exercise their discretion when necessary to sidestep the law’s commands.[51] Closely tying prosecutorial discretion to substantive law, they contend, will result in wooden formalism and needless severity.[52] These are all good points. But they undergird a hard bargain. Our long experiment with justice as the prosecutorial touchstone has not produced an abundance of leniency or, in the eyes of many commentators, justice. And given the widely recognized dangers of unbridled prosecutorial discretion, it seems worthwhile to explore alternatives.

This Article explores the possibility of using existing legal rules to construct a coherent normative theory of prosecution—and primarily state and local prosecution, the locus of American criminal law.[53] It crafts a “servant-of-the-law” model of prosecution that takes its name from an undeveloped intuition expressed in the same Supreme Court opinion that tasks prosecutors with ensuring “that justice shall be done.” There, the Court also emphasized that the prosecutor is “in a peculiar and very definite sense the servant of the law.”[54] Parallels can be drawn to the perception of prosecutors in continental Europe, although those legal systems are so distinct that isolated comparisons of the prosecutorial function tend to obscure rather than clarify.[55] The Article retains a domestic focus, explaining how a servant-of-the-law orientation would change American prosecutorial behavior. In so doing, it highlights potential benefits for our criminal justice system of shifting the dominant conception of the prosecutorial role from an “advocate for justice”[56] to a “servant of the law.” Perhaps the greatest benefit is determinacy. By stripping away any pretense that prosecutors possess sweeping responsibility for justice, a servant-of-the-law model lays the groundwork for both overarching principles and concrete guidance for how prosecutors best serve the law in specific situations.[57]

Serving the law is a clearer assignment than seeking justice. But it is by no means easy. Prosecutors will always be torn by competing tensions, as is the law. This tension could lead to an equanimity of purpose rather than the adversarial mindset at the root of modern prosecutorial excess.[58] The law criminalizes certain conduct and sets penalty ranges upon conviction. But the law also dictates a process by which convictions must be obtained, including a heavy burden of proof.[59] The law guarantees criminal defendants a right to counsel and asks neutral factfinders to resolve disputes.[60] A prosecutor viewing this whole fabric as “the law” that must be served would be indifferent to wins and losses. Such a prosecutor succeeds when a jury convicts; succeeds when the jury acquits; and succeeds when dismissing a case due to insufficient evidence or an unlawful search or interrogation. As the Supreme Court explained in another context: “The criminal goes free, if he must, but it is the law that sets him free.”[61] A prosecutor with a professional identity as a servant of the law would find satisfaction in that result.

This understanding of the prosecutor as a servant of the law, rather than an advocate for justice, would clarify and, in some areas, transform the traditional conception of the prosecutor’s role. Importantly, this reconceptualization might be accepted (or spurned) across the political spectrum because it does not fit neatly into either “side” of the criminal justice debate. It will nudge prosecutors toward severity when the background law is severe and towards leniency when the background law is lenient. It could make prosecutors less adversarial and more cooperative with defense counsel and judges. At its core, this reconceptualization narrows the rhetorical scope of the prosecutorial function, placing the focus on other actors, and ultimately the system itself, to ensure that justice is done.

Importantly, a prosecutor who embraces the servant-of-the-law model would not robotically enforce every criminal statute in every case. Most obviously, the prosecutor would decline to prosecute cases with insufficient evidence to prove the defendant’s guilt.[62] This would include cases that depend on police officers with credibility problems, jailhouse informants, coerced confessions, flawed identification procedures, or questionable forensic science.[63] The servant-of-the-law prosecutor would also preference defendant-protective state and federal constitutional provisions over the mechanical enforcement of criminal statutes. Thus, the prosecutor would decline to trigger bail conditions, charges, or sentencing enhancements that would violate constitutional provisions such as the Eighth Amendment’s proportionality principle, or its prohibition of excessive bail or fines.[64] The prosecutor would also rework plea practices to ensure that defendants are not coerced to waive their Sixth Amendment right to a jury trial.[65]

Perhaps most significantly, in light of the weighty due process and discovery requirements that attach to even the most petty criminal case, servant-of-the-law prosecutors would freely dismiss minor cases in response to resource constraints.[66] For example, in an age of police body cameras,[67] even the smallest case requires a prosecutor to gather and review the body camera footage from officers who responded to the incident and interacted with the defendant or victim.[68] Prosecutors who cannot satisfy the legal obligations attendant to every case on their docket must dismiss cases until they can.[69] Given the breadth of American criminal laws, all prosecutors decline to pursue some offenses to preserve resources for more consequential prosecutions. Servant-of-the-law prosecutors, keenly aware of the legal obligations that inhere in every prosecution, would be no exception.

Finally, in some cases, the law itself dictates that prosecutors make choices without guidance, such as California’s “wobbler” offenses: identical offenses that can be charged as felonies or misdemeanors at the prosecutors’ discretion.[70] A prosecutor cannot faithfully carry out a legislative command that has no substance. Consequently, servant-of-the-law prosecutors would default to the less severe option. This prosecutorial-discretion version of the rule of lenity would incentivize legislators to eschew laws that rely on prosecutorial intuition rather than statutory guidance.[71]

A servant-of-the-law model will not resolve all prosecutorial choices, but it would provide a default position for the American prosecutor, a much-needed starting point for crafting nuanced rules and guidance. In light of the complexity of American criminal justice, deviations from the default would be expected. However, these deviations would be recognizable and thus more likely to be accompanied by transparent and consistent explanations.

With the beginnings of a concrete normative theory of prosecutorial behavior in hand, the question becomes whether such a theory is preferable to the malleability of the “do justice” status quo. A shift in orientation may prove unpalatable to both prosecutors who prefer the freedom to do “what is right” and reformers inspired by prosecutors’ potential power to unilaterally disarm the criminal justice system. But over the long term, more narrowly channeling the power of prosecutors could be preferable to continued expansion. And many (if not most) of the goals of the prosecutor-driven criminal justice reform movement could be advanced by reorienting prosecutors as less adversarial servants of the law.

The Article proceeds in four Parts. Part I fleshes out the problems of the “do justice” status quo. Part II takes on the practicality question: Is it possible to craft a normative theory of the prosecutor that places concrete limits on prosecutorial discretion? It constructs a servant-of-the-law model of prosecution and illustrates the model’s application to the most important prosecutorial decisions. Part III addresses desirability: Is such a theory preferable to the “do justice” status quo? Finally, Part IV considers the degree to which a servant-of-the-law theory could complement rather than clash with the progressive prosecution movement.

I. What’s Wrong with “Doing Justice”?

The popularity of justice as a guiding prosecutorial principle is no fluke. Like freedom, democracy, and apple pie, justice resonates with the American public. Not only does everybody love justice, but the term is sufficiently flexible to capture every criminal law related intuition. Justice serves as a ready placeholder for a multitude of other values (fairness, proportionality, retribution, mercy), providing a rhetorical roadmap to support whatever action a prosecutor selects. In the hands of the right prosecutor, this freedom can be used for beneficial purposes, particularly when other actors, such as police, legislators, juries, and judges, fail to do justice.

Prosecutors don’t spend a lot of time defining “justice,”[72] but their few efforts to do so hint at the flaws in relying on the term as the prosecutorial touchstone. It covers everything and therefore demands nothing. The NDAA’s National Prosecution Standards explain that the responsibility to seek justice “includes, but is not limited to, ensuring that the guilty are held accountable, that the innocent are protected from unwarranted harm, and that the rights of all participants, particularly victims of crime, are respected.”[73] Reflecting a similar perspective, one prominent prosecutor writes, “[w]hen we decide to prosecute, it is because we decide that this is how justice will be done.”[74] If the defendant resists conviction, this prosecutor explains: “[T]he prosecutor becomes the most zealous champion of justice you can imagine. He is then a full-fledged fighting advocate; and he should be . . . . His job is now to fight fairly and firmly with all his might to see that truth and justice prevail.”[75] This ambiguous zealotry is both admirable and a recipe for disaster. The villainous prosecutors who haunt legal commentary push the boundaries of criminal law, discount exculpatory information, and reflexively champion harsh punishment as a response to crime.[76] Yet, these prosecutors likely embrace the “do justice” mantra just as warmly as their many critics. The two sides merely disagree about what justice means.[77] As a result, continually shouting “do justice” at prosecutors is more than an exercise in futility. It exacerbates the problems it is supposed to address.

Prosecutors’ critics embrace the “do justice” command because it provides a rhetorical benchmark against which they can always find prosecutors’ decisions lacking.[78] Judges, too, frequently invoke the slogan to criticize prosecutorial failings of various stripes.[79] Legislatures point to the prosecutors’ duty to “do justice” to ease worries about the enforcement of overbroad laws.[80] In all of these applications, the malleability of the “do justice” command is its greatest strength.

In practice, prosecutors rely on the freedom that “justice” provides to leapfrog among distinct rationales for challenged behavior. For example, prosecutors frequently invoke the need to ensure public safety.[81] This “tough justice” rhetoric can support severely punishing particular defendants to advance instrumental crime-prevention goals. Most famously, federal prosecutors sent Al Capone, Chicago’s “leading mobster,” to prison on tax charges when they could not win convictions for his more notorious conduct.[82] In modern times, federal prosecutors tout their efforts to obtain severe sentences for gun possession as part of an effort to reduce violence in specific regions[83] and nationwide.[84] New York City prosecutors use an automated “arrest alert system” to “incapacitate high priority offenders with higher bail or more severe incarceration sentences.”[85] This results, they say, “in increased prosecutorial effectiveness and enhanced public safety.”[86]

The progressive prosecution movement centers new rhetorical tools, including a “populist justice” approach that attempts to mold discretionary decisions to accord with constituent preferences.[87] For example, one prominent reform campaign includes the exhortation: “What the public wants to have happen is what the District Attorney should be doing.”[88] The populist prosecutor can justify certain discretionary decisions, such as declining to invoke locally unpopular laws, as channeling the will of the voters. Another progressive variant seeks social justice, by counteracting a system perceived as overly punitive and racist, and vigorously prosecuting police officers who violate the law.[89]

All of the above theories of prosecution fit neatly under a justice rubric. In fact, the same prosecutor can invoke each of the above-described theories in different (or the same) circumstances. For example, when Philadelphia District Attorney Krasner announced the dismissal of marijuana cases, he invoked both social justice—“I felt it was the right thing to do”—and public safety: “We could use those resources to solve homicides.”[90] Widely hailed progressive prosecutors like Kim Foxx similarly center the importance of public safety to their role.[91] Foxx instructs “her prosecutors to focus less on narrower legal points” and more on the big picture: “[D]etermining the fairest outcome for each case.”[92] But Foxx responds to critics who say: “You’ve politicized [the prosecutor’s office] by taking on these social issues,” by emphasizing that she “always” situates change “in the framework of safety.”[93] Baltimore’s chief prosecutor, Marilyn Mosby, invoked the public outcry for justice to defend the controversial prosecution of police officers responsible for the death of Freddie Gray.[94] Saint Louis Prosecutor Robert McCulloch invoked the same principle to defend the controversial decision not to prosecute the Ferguson police officer who killed Michael Brown.[95] This is the allure of a justice benchmark. Freed of any specific guidance, prosecutors, along with their champions and detractors, can mix and match justice-based theories of prosecution to support almost any policy or action.

The interchangeability of various, often inconsistent, justice-based prosecutorial approaches is not the only problem. There is also indeterminacy. Many of the above-referenced approaches (public safety, social justice, populism, fairness) provide only slightly more guidance with respect to specific prosecutorial decisions than the overarching “do justice” command itself. Yet, prosecutors must make hard decisions about real cases. For example, if the evidence to support a prosecution is close to the boundary, what guidance does it add to say that the prosecutor should seek “public safety” or “social justice”? Would the public safety prosecutor decrease the requisite evidentiary bar, and the social justice prosecutor increase it? Does it depend on the case, the victim, the defendant, and by how much? At best, concepts like “social justice” and “public safety” hint at the direction of a possible prosecutorial decision, but they give individual prosecutors (and their critics) little guidance about where to draw any particular line.

None of this means that justice should be abandoned as the guiding prosecutorial principle. It may well be the best of a flawed set of alternatives. In the current climate, the freedom of action provided by the absence of a normative theory allows prosecutors to blunt the sharp edges of the system that brought us mass incarceration. But it is worth keeping in mind that mass incarceration arose under the “do justice” status quo.[96] And the lack of a concrete normative principle hobbles efforts to control prosecutorial excesses of all kinds, including unnecessary severity. Granting prosecutors a roaming mandate to “do justice” comes at a cost. The question becomes: Is there a better alternative?

II. An Alternate Theory of the Prosecutor

Justice as a job description works well for superheroes. When an evil villain attempts to destroy the world, justice provides a ready answer for whether to respond (yes!) and how (whatever it takes!). For the typical local prosecutor attempting to emerge from under a pile of paperwork, however, it is a nebulous command. The challenge, then, is identifying an alternative that would offer real guidance in the day-to-day tasks that occupy prosecutors. This Section sketches an alternative “servant-of-the-law” theory to govern those decisions—at least as a default framework for thinking about the prosecutor’s role. It begins with charging—often the first and most important decision that a prosecutor makes—before moving to other aspects of the prosecutorial role like plea bargaining and case processing.[97] Along the way, it highlights the inadequacy of “doing justice” as an answer to these questions.

A. Assessing Evidentiary Sufficiency

Any theory about prosecutorial behavior should begin with charging.[98] Charging at the state and local level typically has two components. First, a prosecutor must determine whether to accept a case initiated by the police. Second, the prosecutor must determine the precise charge (or charges).

For many cases, the charge will be obvious (e.g., residential burglary) and the critical question will be whether there is sufficient evidence that (a) the crime occurred, and (b) the defendant, as opposed to some other person, is responsible. To consistently screen these cases, prosecutors must identify some quantum of evidence to serve as a prerequisite to prosecution. The law sets a low floor: probable cause.[99] Absent probable cause to believe the defendant guilty of an offense, the prosecutor cannot legally maintain a charge.[100] Ethics codes echo this standard.[101] Unfortunately, beyond this minimal guidance, the “codes are deficient regarding the [requisite] degree of confidence and how it should be achieved.”[102]

Those who take justice as the prosecutor’s benchmark struggle to find consensus as to how a prosecutor should assess the requisite proof of guilt ex ante.[103] Among the commonly referenced possibilities are (1) deferring close cases to the jury,[104] or (2) proceeding only in cases where the prosecutor personally believes the defendant to be guilty.[105] Both choices are objectionable, although they fit comfortably under a justice umbrella. Most cases never get to trial, and most convictions result from guilty pleas.[106] Consequently, if a prosecution is unwarranted, that decision cannot be deferred to a jury verdict that may never come. Similarly, relying on a prosecutor’s personal (and potentially idiosyncratic) belief in a particular defendant’s guilt or culpability is unappealing for obvious reasons. Instead, as discussed below, the prosecutor should be tasked with assessing a defendant’s substantive guilt in light of the evidence that would be introduced at trial—i.e., the result most consistent with the law.

A servant-of-the-law approach to charging tracks Department of Justice (DOJ) guidance that instructs federal prosecutors to pursue “readily provable” cases, “cases that the government should win if there were a trial.”[107] The DOJ explains: a charge is not “readily provable” if “the prosecutor has a good faith doubt . . . for legal or evidentiary reasons . . . as to the government’s ability readily to prove a charge.”[108] Note that these formulations subtly discount the prosecutor’s subjective assessment of the defendant’s guilt. They instead focus on the prosecutor’s professional assessment of a jury’s reaction (“should win”) to the expected evidence.[109] “Should,” in this context, is best read to reflect a jury following the judicial instructions that describe its role. Although juries have the raw power to acquit for any reason, courts instruct juries that they “must” (or sometimes “should”) convict when the evidence proves the defendant’s guilt beyond a reasonable doubt.[110] Juries in the post-Civil War South reflexively acquitted white “defendants who committed crimes against African-Americans and white Republicans,” yet the prosecution of such cases is easily defended.[111] In modern times, the government could properly maintain a strong case against a police officer even though it expected a police-friendly jury to acquit. Or the prosecutor could pursue a case against an alleged date rapist even if the prosecutor believed the jury would acquit based on disagreement with the law.[112] The same principle applies even when the prosecutor (as opposed to the jury) disagrees with the law’s application. For example, a servant-of-the-law prosecutor with philosophical objections to drug or gun possession laws would not dismiss a drug or gun case based on their personal opinion that the jury “should” not convict.

Importantly, in making the charging decision, the servant-of-the-law prosecutor—interested only in whether the defendant should, legally speaking, be convicted—would be keenly attuned to detecting faulty prosecution evidence. This would include cases that depend on police officers with credibility problems, jailhouse informants, coerced confessions, flawed identification procedures, or questionable forensic science.[113] Even if unsophisticated jurors might be swayed by such evidence, a servant-of-the-law prosecutor would decline to offer any evidence that a reasonable juror should reject.

Putting the above-described standards together reveals a concrete evidentiary charging standard: a prosecutor should only charge a case when the prosecutor expects that the evidence introduced at trial will prove the defendant’s guilt beyond a reasonable doubt. Importantly, this standard is not merely descriptive. It does not turn on a prediction of whether the jury will convict. Instead, it looks to whether, given the admissible evidence, the jury should (legally speaking) convict. In other words, the standard tracks the law. Servant-of-the-law prosecutors would not select charges based on their subjective perception of guilt or justice, or to protect the public, or by deferring to the result the community prefers. Rather, these prosecutors would seek to obtain results that mirror the applicable laws.[114]

B. Selecting a Charge

Once the prosecutor is satisfied there is sufficient evidence of guilt to warrant prosecution, the next question is how stringently to charge. Even if the defendant’s guilt is clear, a prosecutor has the power to dismiss the case entirely, or to file more serious charges than the facts support.

Application of the servant-of-the-law model is clearest with respect to charges with concrete elements and delineations. For example, Pennsylvania law creates six grades of theft that track “the market value of the property at the time and place of the crime.”[115] The framework (simplified slightly) is as follows:

First-degree felony over $500,000
Second-degree felony $100,000 to $500,000
Third-degree felony $2000 to $100,000
First-degree misdemeanor $200 to $2000
Second-degree misdemeanor $50 to $200
Third-degree misdemeanor up to $50

Imagine that the defendant stole a two-year-old laptop computer left momentarily unattended at the library. The computer cost $2,500 new, but is likely worth significantly less on the resale market. If justice is the measuring stick, it is unclear how the prosecutor should proceed. Some prosecutors may seek justice by giving the defendant a break, while others may view justice as requiring the most serious plausible charge (a felony). The NDAA’s National Prosecution Standards illustrate the problem. The Standards direct prosecutors to “only file those charges that are consistent with the interests of justice.”[116] To help prosecutors implement this command, the Standards enumerate seventeen factors that “may be considered” in determining whether to bring any charge, plus thirteen factors that “may be relevant” to selecting the precise charge.[117] The Standards emphasize that prosecutors can still consider unenumerated factors.[118] Some of the listed factors (“other aggravating or mitigating circumstances”) actually broaden, rather than restrict, the calculus.[119] That’s the “do justice” alternative. Justice sounds like a great standard, but it boils down to a broad license for prosecutors to do whatever they think is right. Apart from obvious consistency concerns, there is a legitimacy problem as well. When prosecutors manipulate charges in the interest of justice, they are effectively amending the laws enacted by the legislature. This may be unavoidable where the law itself provides amorphous and subjective benchmarks. But it is problematic where, as for many of the most commonly charged offenses, the statutory elements are clear.[120]

A potentially preferable alternative directs the prosecutor to “serve the law” by charging the offense for which a jury should (legally speaking) convict—here, the first-degree misdemeanor given the current value of the computer. This standard echoes the discussion from the preceding section and provides both a charging floor and a ceiling. It prevents the prosecutor from giving the defendant a break by downgrading the offense, but it also limits the prosecutor’s ability to “overcharge.”

The concept of overcharging, while ubiquitous in criticism of prosecutorial behavior, is difficult to define using a justice benchmark and so impossible to restrain.[121] The ABA Standards for Criminal Justice illustrate the difficulty:

The line separating overcharging from the sound exercise of prosecutorial discretion is necessarily a subjective one, but the key consideration is the prosecutor’s commitment to the interests of justice, fairly bringing those charges he or she believes are supported by the facts without “piling on” charges in order to unduly leverage an accused to forgo his or her right to trial.[122]

Again we see the familiar role played by generic appeals to “the interest of justice,” and the centrality of vague qualifiers (“fairly,” “unduly”) and colloquial metaphors (“piling on”). A more serviceable definition of overcharging would be: charging the defendant with an offense or enhancement that is not “readily provable,” or an offense for which the jury should not convict.[123] Here the ABA’s illustration resonates with one provided in the federal charging guidance referenced in Part II.A: “charges should not be filed simply to exert leverage to induce a plea.”[124] This principle fits comfortably within the servant-of-the-law model. It also parallels the rhetoric of progressive prosecutors who seek to alter charging practices to eliminate overcharging. In a recent profile, a leading voice of the progressive prosecution movement, Philadelphia District Attorney Larry Krasner, embraced a charging philosophy that parallels the standard described above: “We are not going to overcharge. . . . We are not going to try to coerce defendants. We are going to proceed on charges that are supported by the facts in the case, period. The era of trying to get away with the highest charge regardless of the facts is over.”[125] Interestingly, this notion of structured charging does not fit neatly under a broad appeal to justice. For example, could a prosecutor who is confident of the defendant’s guilt but seeks to spare a child-victim the trauma of trial, “pile on” charges to coerce a plea? An appeal to “do justice,” “public safety,” “social justice,” or “popular will” provides no answer. Perhaps it is no coincidence, then, that the NDAA’s justice-focused National Prosecution Standards do not include any restriction on filing charges to obtain plea bargaining leverage.[126]

The servant-of-the-law charging strategy described above requires additional nuance in scenarios where the law itself provides choices but little guidance. There are three iterations. First, the law may provide only amorphous and subjective guidance. For example, Virginia’s aggravated manslaughter statute applies when “the conduct of the defendant was so gross, wanton and culpable as to show a reckless disregard for human life.”[127] Second, a statute may provide choices without guidance, leaving the charge determination entirely up to the prosecutor. This is the case with California’s “wobbler” offenses—single offenses that can be charged as felonies or misdemeanors at the discretion of the prosecutor.[128] Third, the law may provide a facially clear standard, but signal a desire that the prosecutor invoke the standard sparingly. Examples of this last iteration typically involve draconian enhancements to standard punishments tempered by an extra step beyond charging that the prosecutor can decline to exercise.

The first iteration (crimes with inherently subjective elements) fits into the existing model. The servant-of-the-law prosecutor must make an objective determination as to whether the admissible evidence will establish proof of the ambiguous element (“gross, wanton and culpable”) beyond a reasonable doubt. The exercise will be more difficult than in the earlier Pennsylvania theft example, but it is the same exercise we ask of jurors and judges, and an unavoidable aspect of certain common law crimes. True, prosecutors will reach different conclusions across jurisdictions and over time. But here, the legislature knowingly invites inconsistency by defining a criminal offense by reference to ambiguous terms. A servant-of-the-law prosecutor will have to accept that inconsistency as part of the law.

The second and third iterations described above are anathema to a servant-of-the-law model. The servant-of-the-law prosecutor can and should reject discretionary choices devoid of legal guidance. The most effective way to do this would be to apply a prosecutorial discretion “rule of lenity” that defaults to the less severe option when the legislature tries to dictate a standardless choice. Typically applied by courts in the statutory construction context, the “rule of lenity requires ambiguous criminal laws to be interpreted in favor of the defendants subjected to them.”[129] Among other things, this pushes “Congress to speak more clearly.”[130] It is a small leap to apply this same principle to prosecutorial discretion in a servant-of-the-law regime. Justice Antonin Scalia illustrated the thinking in a related context as follows: “[O]nly the legislature may define crimes and fix punishments. Congress cannot, through ambiguity, effectively leave that function to the courts—much less to the administrative bureaucracy.”[131]

This default to leniency is also appropriate when a standard exists for a sentencing enhancement or other provision, but the legislature expresses a desire for prosecutors to decline to invoke the provision even when that standard is met. For example, federal law authorizes draconian sentences for repeat drug offenders, but in a separate statute neutralizes these enhancements unless the United States Attorney takes an affirmative step to authorize their imposition.[132] Prosecutors got the message. A recent United States Sentencing Commission study found that federal prosecutors triggered these enhancements in only 12.3 percent of cases where they applied.[133] This is an unsettling state of affairs. The law does not impose these life-changing enhancements on eligible offenders. It offers them to federal prosecutors to use as they wish. Declining to accept this offer is consistent with, and best serves, the law. Stated another way, when the legislature provides punishments without standards, there is no law to serve.

Even with the caveats described above, commentators may object to making charging decisions by plugging provable facts into a statutory framework because of the perceived severity of that framework in many jurisdictions.[134] For example, unlike the federal provision, California’s “three strikes” law signals a legislative intent for broad application.[135] A California prosecutor serving the law would charge cases with qualifying “strikes,” placing the defendant at risk for the extreme sentences apparently sought by the legislature. This introduces the issue of prosecutorial nullification.[136] A key limitation of a servant-of-the-law approach is that it disfavors prosecutorial nullification. Justice, by contrast, embraces nullification when it is the right thing to do. The social justice prosecutor can nullify the law to correct perceived legislative and law enforcement inequity. The public safety prosecutor might use the same power to dismiss calls to prosecute police officers or vigilantes. The populist prosecutor can embrace nullification as an opportunity to channel the (local) popular will. The possibilities are endless. Thus, further discussion of prosecutorial nullification must be postponed to Part IV. For now, it is enough to note that even if we accept the appeal of nullification, similar results can often be obtained through structured legal processes, such as drug courts or diversion programs. Here, we are concerned only with charging. From a servant-of-the-law perspective, our hypothetical laptop-stealing defendant could be charged with the offense that best fits the provable facts: first degree misdemeanor theft. (The servant-of-the-law prosecutor could not charge a more severe offense as a negotiating tactic or to try to coerce a plea.) From there, the case can proceed to adjudication or be steered into a diversion program where the prosecutor or judge would dismiss the charge and expunge any record subject to various conditions (e.g., restitution). Diversion programs are discussed further in Part III.B.

It is also worth noting that the servant-of-the-law prosecutor must serve constitutional law that may complicate the application of the most draconian criminal statutes. For example, a servant-of-the-law prosecutor would avoid charging cases that prevent defendants from exercising their right to trial. This scenario arises in at least two ways: (1) when, in a congested court system, defendants are detained prior to trial and an inevitable plea deal will provide a more attractive alternative to a post-trial acquittal,[137] and (2) when a mandatory sentencing provision magnifies the likely punishment to such a degree that even a reasonable innocent defendant would accept a guilty plea (for example, California’s “three strikes” law as applied to a relatively minor offense).[138] In these scenarios, the prosecutor need not ignore this context in selecting a charge. And where the charge itself jeopardizes the defendant’s constitutional rights, a servant-of-the-law prosecutor would preference the Constitution over the robotic application of a criminal statute.

C. Prioritization and Plea Bargaining

Any principle governing prosecutorial behavior must account for resource constraints, particularly at the state level. Police will often present prosecutors with more readily provable cases than they can responsibly prosecute.[139] Issa Kohler-Hausmann documented New York City misdemeanor prosecutors handling caseloads that “often number between 100 and 200.”[140] The volume extends to more serious cases as well. Adam Gershowitz and Laura Killinger, reporting on astronomical state prosecutor caseloads across the country, highlighted Harris County, Texas (Houston), “where some prosecutors are handling upwards of 1500 felonies per year and over 500 felonies at any one time.”[141]

In light of these caseload realities, many prosecutors need to dismiss some cases to competently handle the rest. This requires prioritization. Prosecutors “doing justice” have a free hand in deciding how to prioritize. A servant-of-the-law prosecutor’s decisions would be guided by formal legal strictures. A prosecutor can prioritize cases the law considers most serious, using cues like sentencing ranges and legislative labels (misdemeanor, felony, first degree, second degree, etc.).[142] One Virginia prosecutors’ office, for example, recently responded to caseload pressure by declining to handle misdemeanor cases at all.[143]

The other prosecutorial prioritization tool is plea bargaining. Observers identify plea bargaining as the heart of prosecutorial power, given that 95 percent of criminal convictions result from guilty pleas.[144] Complicating matters, there are two competing theoretical narratives of plea bargaining, labeled here the “Godfather paradigm” and the “shadow-of-trial paradigm.” A servant-of-the-law prosecutor would engage in plea bargaining under the latter, but not the former paradigm.

In the “Godfather paradigm,”[145] the prosecutor unilaterally controls plea bargaining and guides the process to a desired result. For disfavored defendants, the prosecutor inflates charges to a point where the defendant cannot risk trial. The defendant then reluctantly agrees to plead guilty to avoid an excessive trial penalty. Stingy discovery regimes keep defendants ignorant of their trial chances, and draconian sentencing laws push trial out of reach. For favored defendants, the process works in reverse. Prosecutors offer plea terms so lenient that the defendant would be foolish to refuse. These deals can be hammered out without input from victims and police. And the parties can use charge and fact bargaining, and stipulated sentences to minimize judicial second-guessing.[146]

The competing “shadow-of-trial paradigm” posits prosecutors as less like Mafiosos and more like bureaucrats. These prosecutors process more provable cases than they (or the courts) can try. This gives prosecutors an incentive to compromise with defendants (and their attorneys) in much the same way that civil parties reach settlements. The prosecutor and defense attorney assess the likely outcome of the case—in light of the assigned judge and local juries—and craft a resolution that discounts that outcome in exchange for the defendant’s waiver of the right to a trial. For the most frequent state cases (e.g., drugs, common law crimes, DUIs, domestic violence), such deals become standardized. Discovery practices, the defendant’s own knowledge, and counsel’s experience permit sufficient pre-plea case assessments. And the trial penalty, while still present, is modest enough that defendants can still exercise their trial right.

The “Godfather paradigm” neatly suits prosecutors seeking justice. A prosecutor who equates justice with severe penal consequences can seek to maximize guilty pleas and sentencing severity by manipulating the charges. A prosecutor seeking social justice can steer sympathetic defendants to lenient outcomes by offering generous terms. To avoid legislative and judicial override, prosecutors can work with defense counsel to massage the charges and facts to reach their desired result. In essence, plea agreements provide a ready mechanism for parties to bargain around the law to achieve justice.[147]

On the other hand, the “shadow-of-trial paradigm” neatly suits a servant-of-the-law model of prosecution. The widespread settlement of civil cases is typically viewed as a means of enforcing, not undermining, the law.[148] Similarly, plea bargains can be structured so that agreements track likely post-trial outcomes, even with standardized discounts. By adhering to the “shadow-of-trial paradigm,” prosecutors could engage in plea bargaining while adhering to their role as servants of the law.[149] In fact, empirical evidence suggests that in most circumstances, repeat-player attorneys share an assessment of the likely trial outcome.[150] A defendant with this guidance could rationally trade away a potential acquittal for prosecutorial concessions.[151] The resulting plea bargaining in the “shadow of trial” would uphold, rather than undermine, the law.[152]

Reliance on the “shadow-of-trial paradigm” does not unleash servant-of-the-law prosecutors to pursue subjective conceptions of the “just” result. While servant-of-the-law prosecutors could dismiss a provable charge to obtain a guilty plea, they could not support a guilty plea to an offense for which there was no factual or legal basis. Thus, unlike a prosecutor “doing justice,” a servant-of-the-law prosecutor could not bargain for so-called “fictional pleas.”[153] Just as servant-of-the-law prosecutors cannot inflate charges for plea bargaining leverage, they cannot endorse guilty pleas to lesser offenses that never occurred. Servant-of-the-law prosecutors are limited to charging offenses supported by the provable facts,[154] and this obligation continues through guilty plea proceedings. The servant-of-the-law prosecutor would also remain attentive to defense-protective constitutional law during the plea bargaining process. The primary rule in play will be the defendant’s right to effective assistance of counsel.[155] But, as noted in Part II.B, prosecutors must also be cognizant of the pressure certain plea offers place on a defendant’s Sixth Amendment right to trial.

D. Prosecuting Cases

The preceding discussion fleshes out the prosecutorial calculus with respect to the most important discretionary choices made by typical state and local prosecutors: charging and plea bargaining. This Section discusses other important procedural decisions.

Here, we can revisit the iconic Berger decision to consult the most frequently cited guidance in this context.[156] In Berger, the Supreme Court wrote that a prosecutor should “prosecute with earnestness and vigor”: “But, while he may strike hard blows, he is not at liberty to strike foul ones. It is as much his duty to refrain from improper methods calculated to produce a wrongful conviction as it is to use every legitimate means to bring about a just one.”[157] The imagery of a prosecutor striking “blows,” whether hard or foul, hints at an underlying flaw. The ideal theory of the prosecutor’s role should not conceptualize the prosecutor as a physical combatant. The only thing Berger adds to this unhelpful metaphor is the directive that prosecutors refrain from “improper methods calculated to produce a wrongful conviction.”[158] This is just question begging. No one thinks that prosecutors should use “improper methods” to obtain “wrongful convictions.”[159] The question is what makes a method “improper” and a conviction “wrongful”?

Again, Berger generates a nice billboard. But the famous language does no work in filling out a theory of the prosecutor’s role. Similarly, Berger’s exhortation that prosecutors vigorously strive to obtain “just” convictions does little to answer procedural questions in particular cases. Justice rhetoric becomes particularly problematic to the degree it suggests that prosecutors can deviate from laws with which they disagree. It is one thing to say the prosecutor’s duty to justice overrides fealty to the law in the charging or plea bargaining context. It is quite another to cite justice as a trump of the prosecutor’s procedural obligations. Yet, it is a small step from invoking prosecutorial discretion to undermine a substantive law (e.g., marijuana possession or capital punishment) to doing so to undermine a procedural one (e.g., a requirement to provide pretrial discovery or notice to victims).

The search for concrete guidance again leads to the principle that prosecutors must serve the law. As a servant of the law, the prosecutor’s role extends beyond proving a case while adhering to applicable legal strictures. The special supplemental obligation of this model is that prosecutors would work to help other actors fulfill their legal duties. Berger cautions prosecutors not to kick defense attorneys when they are down (“foul blows”). A servant of the law helps struggling defense attorneys up so they can perform their legally mandated roles.[160]

A non-combative prosecutorial mindset, more than any formal prescription, may be the most valuable thing American legal systems can import from continental Europe. The contrast is striking in one recent study of German prosecutors—perhaps the closest real-world analogues to servant-of-the-law prosecutors. Based on interviews, Shawn Boyne reports: German prosecutors eschew “self-aggrandizing terms” and boasts of “individual achievements”; they do not make “grand pronouncements about keeping society safe or getting criminals off the streets.”[161] “Unlike American prosecutors, German prosecutors do not describe themselves as white knights or avenging angels . . . . A German prosecutor does not ‘do battle’ in the courtroom, put the criminal ‘away,’ or triumph over evil.”[162]

The idea of the prosecutor as a disinterested champion of the law contrasts with the traditional American conception of the prosecutor as a zealous partisan. For example, the ABA Criminal Justice Standards state that “[a] prosecutor is not required to make a disinterested exposition of the law . . . .”[163] But why not? It is unclear what we gain by framing the criminal justice exercise as a game, with prosecutors rooting for a particular team. Theoretically, by mixing an adversarial prosecutor with equally adversarial defense attorneys, we get justice.[164] But the playing field is hardly level.[165] And there are reasons to think that ambiguity about the prosecutor’s role actually enhances their advantage. Prosecutors don’t have to signal when they are acting as advocate and when they are acting as a “minister of justice.” They can leverage this ambiguity with juries, judges, and the public as they see fit.[166]

Servant-of-the-law prosecutors’ unique role comes into play not in the fairness of the blows they deliver to defendants, but rather in careful attention to preventing legal system failures. Thus, the discovery obligations that flow from the servant-of-the-law model are broader than those currently recognized under a “do justice” paradigm. A prosecutor seeking to serve the law will want the defense to have all the information that the prosecutor possesses, i.e., open-file discovery.[167] The substantive laws demand punishment only for persons who are factually guilty. Procedural laws assign attorneys to defendants to seek out and present exculpatory information, and demand neutral factfinders to weigh the evidence.[168] The process dictated by the law only works when prosecutors promptly pass along all case-related information to the defense. After all, a prosecutor whose goal is to serve the law, not seek a conviction (or justice), has little incentive to withhold information even when the specific jurisdiction’s rules do not mandate disclosure.

The servant-of-the-law prosecutor’s unique obligations extend beyond discovery. A defense attorney can celebrate when a judge dismisses a case because the prosecutor forgot to subpoena an important witness. A servant-of-the-law prosecutor faced with an analogous defense failure should suggest a postponement, ensuring that the defendant receives “the effective assistance of counsel” demanded by the Constitution.[169] The prosecutor would internalize the correct legal ruling when a trial judge fails to enforce the evidence rules or defense counsel provides constitutionally ineffective representation, even if doing so strengthens the likelihood of an acquittal. If a case depends on evidence seized unlawfully, the prosecutor would dismiss the case, rather than wait to see if the defense attorney and judge catch the constitutional violation. Although such a dismissal may or may not comport with justice, it serves the law.

Another distinction will manifest at sentencing. Prosecutors on “a quest for justice”[170] can rationalize almost any sentence recommendation. Thus, an article published in 2000 by the then-President of the National District Attorneys Association suggests that a prosecutor should normally effectuate the call to “do justice” by seeking “the most severe sentence allowed by law.”[171] By contrast, Philadelphia’s new District Attorney directs his prosecutors “to seek justice for society as a whole” by asking for lower sentences that are sensitive to the financial costs of incarceration.[172] Servant-of-the-law prosecutors would not seek specific sentences. (This was Department of Justice policy for many years.)[173] They would instead address pertinent mitigating and aggravating factors, and inform the court of applicable laws. As servants of the law, prosecutors would have no professional interest in a particular sentence, other than that the sentence fall within the legal parameters.

A similar calculus applies to bail determinations. The law places pretrial release determinations in the hands of the judge.[174] Just as the prosecutor should respect the determination of legislators as to what conduct is unlawful, the prosecutor would defer to the law’s selection of judges as the arbiter of pretrial release decisions and sentences.

A professional identity as a servant of the law would follow prosecutors outside the courtroom. If conceived as “an independent administrator of justice,”[175] the prosecutor can sensibly seek to influence legislative policy. Both the ABA[176] and the National District Attorneys Association[177] encourage prosecutors to try to influence lawmakers to forward the cause of justice. Rachel Barkow, who identifies prosecutors as “one of the most—if not the most—powerful lobbying groups in criminal law,” makes a compelling case that prosecutors use this power “to lobby for harsher sentences because longer sentences make it easier for them to obtain convictions through plea bargaining.”[178] Others call on prosecutors to push legislators to enact progressive reforms like eliminating harsh sentencing provisions and cash bail.[179] By contrast, prosecutor lobbying does not fit under a servant-of-the-law model. Much like judges, servant-of-the-law prosecutors would adopt a professional norm of only presenting unsolicited views to the legislature on matters directly affecting their positions, such as office funding.[180] The prosecutor who is serving the law would leave advocacy for changes to that law to others.

In sum, as a servant of the law, the prosecutor must give the law force even when other actors fail to do so and even when doing so leads to dismissals and acquittals. This prosecutor would be indifferent to winning a case, and zealous only in ensuring that the laws are followed and the adjudicatory process created by those laws functions properly.

III. Moving from Doing Justice to Serving the Law

The previous Section sketches a servant-of-the-law model of prosecutorial behavior as a potential alternative to “doing justice.” This Section discusses whether the alternative is desirable. For all its flaws, the justice model provides prosecutors with flexibility to do the right thing, as they perceive it. It may be better to preserve this flexibility than to limit prosecutorial discretion with more concrete guidance. To analyze the tradeoffs involved, this Section compares the servant-of-the-law model to the “do justice” status quo in various categories of decision-making: charging serious cases, dismissing minor cases, promoting adherence to innocence-protective procedural rules, and incentivizing other institutional actors to do justice.

A. Charging Serious Cases

With respect to charging and plea bargaining decisions in the most serious cases, the move from a “do justice” to a servant-of-the-law model may not be particularly disruptive. The servant-of-the-law model of charging set out in Part II may even track existing state and local prosecutorial practices. This is because charge selections at the state level are typically straightforward for the crimes that matter most. In 2014, for example, prison populations reflected the prioritization of common law offenses. In decreasing order of frequency, state prisoners had been sentenced for Murder (13%), Robbery (12.8%), Rape (12.4%), Aggravated/Simple Assault (10.2%), and Burglary (10.1%).[181] Non-possession drug offenses added another 12.2%.[182] This short list of crimes accounts for 70.7% of state prisoners. The other crimes that add smaller percentages of prisoners, like Theft (3.6%), Fraud (2.3%), Weapons offenses (3.94%), DWI (2.1%), and Motor vehicle theft (.8%), follow a similar pattern.[183] At least with respect to cases that lead to incarceration, state prosecutors can engage in straightforward charging calculations—matching the provable facts to easily recognized crimes. In this area, they may already act like servants of the law.[184] To the extent prosecutors deviate from this calculus, it is just as likely that they are overcharging as undercharging.[185]

Many readers will resist the notion that the vast majority of American prosecutors can or do exercise fairly routinized charging discretion. A central claim of one of the most-cited criminal justice articles of all time is that it is not legislators and judges, but “prosecutors, who are the criminal justice system’s real lawmakers.”[186] This notion, famously associated with the iconic Bill Stuntz, guides modern academic discourse.[187] Critically, though, when Stuntz turned his lens to state and local prosecutors, the narrative changed. Consistent with the above discussion, Stuntz described most serious crimes that come through state prosecutors’ doors as “politically mandatory” in that there is no choice but to prosecute.[188] And recognizing the resource constraints on state (as opposed to federal) prosecutors, Stuntz suggested that “there are enough of these politically mandatory crimes to occupy all or nearly all of local prosecutors’ time and manpower.”[189]

Stuntz’s description of state prosecutors suggests a servant-of-the-law model might be neither radical nor disruptive. He writes: “For the state law crimes that are punished most consistently—basically, FBI index crimes plus distribution of serious drugs—state criminal law functions as law. . . . [P]rosecutors prosecute these crimes systematically and aggressively, meaning that, at least roughly, the crimes are enforced as written.”[190] Prominent scholars used to make similar points about the practical restrictions on the charging discretion of prosecutors outside the federal context,[191] although these observations are often missing from modern accounts.[192]

The potential overlap between state criminal law on the books and current prosecutorial enforcement practices is important. It suggests that a narrow servant-of-the-law role for prosecutors may be both theoretically attractive and a rough approximation of the status quo for serious offenses.[193] That would mean that the primary benefits (if any) of offering prosecutors a broad license to “do justice” manifest during the prosecution of minor crimes. As discussed below, there may be other ways to replicate these desirable aspects of prosecutorial discretion without ceding so much amorphous discretion to prosecutors.

B. Dismissing Minor Cases and Diversion

While a servant-of-the-law model may track current charging practices for serious cases, it likely deviates from those practices—for many prosecutors—for less serious offenses. Due to societal ambivalence about the significance of crimes such as drug possession, shoplifting, and prostitution, a prosecutor “doing justice” can freely dismiss charges for those crimes, or craft ad hoc alternative frameworks for disposing of those cases without a conviction. A prosecutor serving the law would not enjoy the same freedom.

This is the context where a servant-of-the-law model will meet the most resistance.[194] Prosecutors’ offices increasingly embrace “diversion” programs that reroute typically low-level cases to informal resolutions. Even the ABA Standards for Criminal Justice, which shy away from any controversy, unequivocally endorse diversion.[195] (The ABA’s enthusiasm extends to endorsing the threat of prosecution as a means to push suspects “to enter the military service.”[196]) The NDAA’s National Prosecution Standards similarly endorse diversion programs if the prosecutor determines that “diversion of an offender to a treatment alternative best serves the interests of justice.”[197] The Standards explain that diversion programs can “conserv[e] judicial resources for more serious cases,” and reduce offender recidivism “by providing community-based rehabilitation.”[198]

The servant-of-the-law model is not against diversion. But it eschews lawlessness. Thus, a servant-of-the-law model would require prosecutors to distinguish diversion programs created by legislatures (good) from those crafted on an ad hoc basis by prosecutors (bad).[199] In rough outlines, (bad) prosecutor-created diversion programs work as follows.[200] The police present a case to the prosecutor. The prosecutor determines that the case is not worth prosecuting. But rather than dismiss the case, the prosecutor requires the arrestee to jump through a set of hoops.[201] The hoops might include community service, drug treatment, or counseling.[202] Or, the sole requirement may be that the defendant avoid re-arrest for a set period of time.[203] The prosecutor then postpones the case pending satisfaction of the requests.[204] The programs are attractive to both prosecutors and defendants because successful compliance results in a closed case. The prosecutor conserves resources and the defendant avoids a conviction.

The servant-of-the-law model counsels against prosecutor-controlled diversion programs. Once a prosecutor decides not to prosecute a case, this model demands that the criminal justice system release the defendant from its clutches. Whether the issue is lack of evidence, insufficient resources, or compliance with a de minimis dismissal statute,[205] the proper prosecutorial response would be dismissal. Period. Although many defendants no doubt benefit from services they encounter through diversion programs, prosecutors should not be the gatekeepers to these forms of assistance. And while these programs are often viewed as hallmarks of progressive prosecution,[206] there is no guarantee that they function in this manner. The coercive force of the criminal justice system powers diversion programs. A defendant who enters drug court, for example, may end up serving more jail time than a similarly situated defendant who rejects the program.[207] This is particularly true if prosecutors divert weaker cases as a politically expedient alternative to dismissals.[208] Diversion is always an attractive option. So attractive, in fact, that it is likely that some nontrivial percentage of defendants working their way through diversion programs were not, and would never have been found, guilty.[209]

The same concerns about unguided, justice-based discretion in other contexts resurface with diversion. Indeed, these concerns are not just theoretical. A 2016 New York Times report on 225 prosecution-run diversion programs in 37 states found widespread abuses, including that “in many places, only people with money” can afford to participate.[210] The article explained, “Prosecutors exert almost total control over diversion, deciding who deserves mercy and at what price . . . .”[211]

The inconsistency between the servant-of-the-law prosecutor and prosecutor-driven diversion programs does not mean that the use of diversion must be reduced. There is no particular magic to prosecutor-controlled diversion programs. These programs’ only real selling point is that they are relatively cheap and easy to create. Ad hoc prosecutor-driven programs can be reconstructed, more comprehensively, through legislative action. And unlike prosecutor-controlled programs, statutory diversion can come with appropriations that override financial barriers to participation, and grant other officials, such as judges, de facto gatekeeping power. For example, a District of Columbia law allows trial judges to defer proceedings against a person convicted of a narcotics offense for up to a year, and at the end of the period, dismiss the case without a conviction.[212] The statute then provides for expungement of the official record.[213]

Legislatures can enact more diversion programs and broaden programs already in existence.[214] Servant-of-the-law prosecutors would play an active role in these statutory diversion programs. Their role would be to ensure that qualified defendants are funneled to diversion programs according to statutory entry criteria. To the extent statutes mandate prosecutorial approval as one of the criteria, prosecutors should simply approve all requests for entry so long as the defendant meets the other eligibility criteria.[215] Police departments can also fill the void, diverting arrestees from the adjudicative process even earlier in the process.[216]

Servant-of-the-law prosecutors could also sidestep diversion entirely and dismiss minor cases pursuant to de minimis dismissal statutes, already present in nineteen states.[217] These statutes authorize the dismissal of the types of cases that typically wind up in diversion programs. As Anna Roberts explains, de minimis “statutes are not about the legal or factual merits of a prosecution, or about guilt or innocence, but are about a determination that while the case is permitted in criminal court, it should not be pursued.”[218] In jurisdictions with de minimis statutes, servant-of-the-law prosecutors would regularly dismiss diversion-type cases—either pursuant to statutory direction or in anticipation of judicial dismissals—with no strings attached.

Finally, prosecutors of all stripes will decline to prosecute a subset of minor offenses to focus finite court and prosecutorial resources on the most serious crimes.[219]

C. Promoting Adherence to the Rules and Protecting the Innocent

The previous Sections address changes in charging practices that would follow from a transition to a “servant-of-the-law” model. This Section looks at whether retiring Berger’s age-old guidance sacrifices anything outside of the charging context. Two wide-ranging benefits of the “do justice” exhortation are its implicit directions that prosecutors: (1) respect legal rules, and (2) protect the innocent. As explained below, these benefits are only enhanced by a servant-of-the-law model.

Commentators sometimes invoke the prosecutor’s duty to “do justice” as if it were independent authority requiring prosecutors to follow legal rules.[220] But the courts do not actually use the “do justice” command in this manner.[221] They don’t need to. Prosecutors, like everyone else, must comply with the rules of evidence, trial procedure, and discovery.[222] That is the way to understand Berger v. United States itself, where the prosecutor transgressed generally applicable rules of trial procedure.[223] When prosecutors breach a rule, it is that rule, not their obligation to “do justice,” that provides a remedy. State prosecutors, for example, must turn over exculpatory evidence.[224] But this is not because they have a duty to “do justice.” Instead, prosecutors must comply with the requirements of Brady v. Maryland because failure to do so results in “a violation of the Due Process Clause of the Fourteenth Amendment.”[225] If anything, the “do justice” command undermines the clarity of the principle that prosecutors must follow the rules. For example, the NDAA’s National Prosecution Standards assert that there may be circumstances where “a prosecutor chooses to disregard a code or rule because of a belief that his or her duty to seek justice requires the same.”[226] By contrast, asking prosecutors to conceptualize their roles as servants of the law strengthens the directive that they comply with the rules governing their own conduct.

Another potential benefit of the “do justice” command is that it prompts prosecutors to care about a defendant’s potential innocence. While the content of justice is debated, one point of general agreement is that it is unjust to punish people for crimes they did not commit.[227] Again, however, this same principle can be derived from a solemn focus on the law. The criminal laws dictate punishment only for those guilty of violating their terms. If, for example, the law provides a severe punishment for murder, it is inconsistent with that law to impose that penalty on someone who did not commit the offense. In addition, as already discussed in Part II.D, the servant-of-the-law model nudges prosecutors to effectuate procedural law, including many innocence-protective rules. Unlike the adversarial champion of justice, a servant-of-the-law prosecutor must facilitate the defense and judicial functions, enhancing judges and defense attorneys’ ability to protect the innocent.

D. Pushing Other Institutions to Do Justice

One of the clearest costs of the “do justice” model of the prosecutorial role is that it creates ambiguity about what prosecutors should be doing. This ambiguity obscures the capacity of other criminal justice actors to do justice—and prevents voters and other observers from holding government officials accountable for perceived injustices. This Section discusses how the clarity of a servant-of-the-law theory of the prosecutorial role interacts with, and could improve, the actions of other criminal justice actors.[228]

The most significant potential impact of reconceptualizing prosecutors as servants of the law would be on the legislature. With respect to setting forth the criminal law, the prosecutor acting as a servant of the law defers to the legislature.[229] The prosecutor would not override the legislature’s determination of what conduct is unlawful, and how it should be punished. In his groundbreaking study of discretion, Kenneth Culp Davis makes this point in describing the role of the police: “When a legislative body enacts that an act is a crime, . . . police nonenforcement on the ground that the enactment is unwise seems clearly an unlawful assumption of power. When administrations flagrantly violate clear statutory provisions, they reject the central idea of the rule of law.”[230] The usual response to Culp Davis is that mechanical application of law will often be too harsh and therefore unjust. A shift from a “do justice” to a “servant-of-the-law” paradigm, however, does not mean that justice-related concerns would be banished from the criminal justice system. In fact, justice, broadly conceived, might be achieved more consistently as other criminal justice actors emerge from the prosecutor’s shadow.

With the prosecutor’s role clarified, a more natural focus on the legislature and judiciary to do justice would emerge. A German law professor explaining how the German system of “compulsory prosecution”[231] differs from the United States’ adversarial model makes these points:

Citizens are protected against unjust convictions and oppressive punishments by the Penal Code rather than by individual prosecuting attorneys. . . . [T]here is general agreement in Germany that the Penal Code must be amended, rather than the polices of the prosecutor altered, if the administration of the criminal law produces undesired results. . . . German law relies on careful and elaborate judicial interpretation of the substantive law to solve problems that the United States often leaves to the discretion of the prosecutor.[232]

The various marijuana legalization mechanisms spreading across the country are a modern American illustration of this dichotomy. In some jurisdictions, states decriminalize marijuana possession through popular referenda or legislative action.[233] In other jurisdictions, prosecutors seek to achieve a roughly equivalent policy outcome through charging policies.[234] Obviously, state-wide referenda and legislation are more comprehensive and enduring means to change drug policy. Prosecutor policies only apply within the prosecutor’s local jurisdiction, are non-binding, can change at any time, and do not prevent police from continuing to make arrests for drug offenses.[235] In fact, prosecutors’ well-intended efforts could, paradoxically, short circuit broader reforms by relieving political pressure in jurisdictions where opposition to drug laws is most acute.[236]

A servant-of-the-law model will also bring out a greater case-sorting role for, or simply highlight the existing role of, the police. This is because even as the laws improve, lawmakers cannot anticipate every scenario. The “rule of law” necessarily involves discretion. No matter where power to enforce the law resides, it will incorporate a power not to act. That “negative power” is particularly ungovernable when delegated to numerous individuals like police officers or local prosecutors.[237]

Yet just as there will always be discretion, that discretion will always be exercised in the first instance by the police. As Culp Davis recognized—and contrary to modern academic thought—it is local police, not prosecutors, who wield the greatest discretion in the American criminal justice system.[238] Police (not prosecutors) conduct stings, make traffic stops, and respond to 911 calls. Prosecutors, by contrast, have little ability to detect crime. Thus, a speeding motorist can get off with a warning, a public drunk can be sent home in a cab, and a bellicose teenager can be lectured, not jailed. A group of co-workers in a college basketball “March Madness” pool need not be dragged into court to face gambling and tax evasion charges. Yet none of these mercies come about because we give broad discretion to prosecutors. These decisions are typically made by police.[239] This is the case in Germany, perhaps the closest example of something like a servant-of-the-law model. While the German prosecutor possesses some discretionary power to decline to prosecute minor cases, this comes up infrequently: “In practice, the job of exercising discretion in these cases is performed by the police.”[240] This means that in conceptualizing a role for prosecutors, we need not decide whether to create a discretionary screen between technical law violations and formal court action. The question is what we gain (and lose) by telling prosecutors to re-evaluate a question—is a particular infraction worth prosecuting?—already answered by local police.

Even if police were not already making these decisions, they are the more logical official to take on the role. Local police are closer to the context of an offense, its role in society, and the likely impacts of legal intervention on the individuals involved. Unlike prosecutors, police come into regular contact with offenders, victims, and witnesses, developing a deeper sense of which cases and persons warrant formal intervention. This is particularly true for assessing a particular defendant’s culpability or remorse. The American adversarial system maintains an almost sacred separation between the prosecutor and a criminal defendant.[241] Unlike a police officer, the prosecutor is “isolated from those—the defendant, his family and friends, and often, his witnesses—who might arouse the prosecutor’s empathy or stimulate concern for treating him fairly.”[242]

Of course, any discretionary screening decision in the American criminal justice system raises concerns about racial bias and other pernicious factors.[243] But there is no reason to expect prosecutors to be preferable to police in this respect.[244] In fact, with respect to important demographic factors, such as race, police departments are far more diverse than prosecutor offices.[245]

Another reason to push discretionary, community-oriented determinations through a police filter is that the broadest justice-related screen should happen as early as possible. If legal intervention is not warranted, prosecutorial dismissals come too late. It is preferable that persons not be arrested than be arrested, jailed, and later have their case dismissed.[246] Malcolm Feeley famously explained that for minor offenses, “the process itself becomes the punishment”; but that was an indictment of the system, not an endorsement.[247]

None of this means that servant-of-the-law prosecutors must prosecute every case brought to them by the police. But it does narrow the prosecutor’s screening role. The most compelling reason to have the prosecutor re-screen a case selected for prosecution by police is that prosecutors, unlike police, are formally trained in the law. While a police officer might decide that an offense is worth pursuing from a community “protect and serve” standpoint, the prosecutor could detect legal or evidentiary flaws in the case that preclude prosecution. Employing typical tools of statutory construction, including prioritizing legislative intent over text,[248] declining to enforce impermissibly vague laws,[249] and the rule of lenity,[250] a prosecutor might determine that a case that technically fits a criminal statute should not be pursued. For example, the teenager who texts a picture of herself to a friend while handing him a marijuana joint could be charged with distributing both child pornography and drugs. The prosecutor’s decision to drop the charges would not constitute nullification, but application of the intent of the law over the literal text—a mainstream interpretive philosophy[251]—particularly in light of constitutional constraints.[252] One might object that this function overlaps too closely with what judges do. But there are good reasons to ask prosecutors to mirror judges under a servant-of-the-law model. Perhaps the most compelling reason is that, in a system where cases are typically resolved prior to trial,[253] a legal screen that would ideally be applied by judges must be delegated elsewhere. Finally, as already explained in Part III.B, many provable cases will be channeled to statutory diversion programs or dismissed entirely to allow prosecutors serving the law to focus finite resources on the most legally serious cases.

IV. Can Progressive Prosecutors Serve the Law?

The previous Sections introduced the servant-of-the-law model of prosecution and analyzed the implications of moving toward that model and away from “doing justice.” This Section addresses the desirability of a servant-of-the-law approach in the current criminal justice moment. Prosecutors constrained by the law may be attractive in the abstract, but a bolder theory of the prosecutorial role may be needed in a country with broad criminal laws and unprecedented incarceration levels. Why not employ every possible tool, including raw prosecutorial power in the form of “progressive prosecution,” to counteract the criminal justice system’s severity?[254] As this Section explains, there is little inherent tension between a servant-of-the-law model and progressive prosecution. In fact, this model could provide much-needed theoretical grounding for evolving perceptions of the ideal prosecutor.

Progressive prosecution holds great appeal for reformers seeking to undo mass incarceration. A progressive prosecutor can dismiss cases to nullify unjust laws; counteract racial and socioeconomic imbalances in police enforcement or legislative drafting; defang the most severe punishments by declining to trigger them; and aggressively prosecute certain under-prosecuted crimes to promote broad policy goals.[255] This progressive wish-list sounds like a challenge to a servant-of-the-law paradigm. In fact, the most concrete aspects of progressive prosecution fit well with a servant-of-the-law model.

Again, we should begin with charging, the most important prosecutorial task. Discussions of progressive prosecution do not typically include a charging standard. But it is difficult to construct a progressive charging standard that would differ significantly from the proposal in Parts II.A and B. Prosecutors, progressive or otherwise, can discern the appropriate charge by matching the readily provable facts to statutorily defined offenses. Importantly, the proposed charging standard captures the progressive intuition that prosecutors should not charge a more severe offense to obtain plea bargaining leverage. Progressive icon Larry Krasner’s recent explanation of the charging practices in his office mirror this standard: “We are going to proceed on charges that are supported by the facts in the case, period. The era of trying to get away with the highest charge regardless of the facts is over.”[256] Another authority in the progressive prosecution movement provides similar guidance: “Don’t make a plea offer if you can’t prove the charge beyond a reasonable doubt.”[257] These sentiments fit neatly with a servant-of-the-law, not a “do justice,” model.[258] In addition, as explained in Part II.B, progressive prosecutors could embrace the latitude the servant-of-the-law theory offers by defaulting to the less severe alternative whenever legislators provide choices without guidance.[259]

The most likely point of contention between servants of the law and progressive prosecutors will arise when the latter reject certain disfavored laws. But there is a critical difference between viewing certain charges through a more skeptical lens and rejecting them out of hand. For example, Rachael Rollins, the newly elected District Attorney of Suffolk County (Boston), generated buzz among progressives with her “plan to forgo prosecution of 15 offenses, ranging from trespassing to drug possession with intent to distribute.”[260] Similar sentiments appear throughout the progressive prosecution movement.[261] But typically, progressive guidance does not reject offenses outright. As Rollins herself emphasizes, her “list ‘is not a blanket commitment.’” Instead, Rollins’s actual proposal is that prosecutions for certain less serious offenses require supervisory approval[262]—something that is required for any charge in many jurisdictions where supervisors screen all initial charging decisions.[263] Rollins herself stresses that her do-not-prosecute list is not radical: “It’s about memorializing what is already happening in the majority of cases.”[264]

General guidance to avoid prosecutions of minor offenses fits into a tradition of American prosecutors responding to resource constraints by declining cases presented to them by police.[265] Minor crimes that are extremely common, like drug possession, trespassing, and loitering, provide the most obvious opportunity for prosecutors to conserve resources. Limiting prosecution of these crimes also fits neatly into a servant-of-the-law paradigm. These offenses are among the least severely punished by legislatures and judges and yet, due to their volume, have the potential to clog court dockets. Prosecutors faced with inevitable resource constraints serve the law by declining to prosecute these offenses, enabling a tighter focus on more significant offenses. Progressive prosecutors may dismiss more minor offenses than non-progressive prosecutors, but apart from the rhetoric surrounding those decisions, the distinction is one of degree rather than kind.

Other attributes of the servant-of-the-law framework fit even more squarely within the progressive prosecution movement. Providing broad discovery to criminal defendants,[266] applying the law to crimes committed by police officers,[267] dismissing cases that rely on thin or questionable evidence,[268] and taking claims of innocence seriously[269] all fit neatly into a servant-of-the-law paradigm.

Anchoring the evolving ideal of the prosecutorial role in a theoretical framework provides a variety of benefits. By embracing a servant-of-the-law model, progressive prosecutors could ground their reforms in a rule-of-law paradigm that will appeal across the political spectrum, including to line prosecutors, police, and legislators. At a more general level, there are benefits to tying prosecutors, including those leading the progressive prosecution movement, to an overarching theory of prosecution. A rubric that places some theoretical parameters on prosecutorial freedom resists the inexorable extension of “the prosecutor’s empire.”[270] For decades, reformers railed against the evils of unbounded prosecutorial discretion.[271] In one of the more influential treatments, Angela Davis contended that “everyday, legal exercise of prosecutorial discretion is largely responsible for the tremendous injustices in our criminal justice system.”[272] The compelling concerns about prosecutorial discretion expressed in longstanding academic critiques do not vanish when the discretion is (occasionally) used in a favored direction.[273] To the contrary, the absence of a normative theory of the prosecutorial role enhances the dangers of expanding prosecutorial power. There are over twenty-five thousand prosecutors spread across the country.[274] A growing, but ill-defined sense that prosecutors not only can, but should, bend the laws to further individual conceptions of justice will not be limited to certain prosecutors.[275] It will benefit unreasonable prosecutors, corrupt prosecutors, benevolent prosecutors, and everyone in between.

Relatedly, voters might heed the call to pay more attention to prosecutorial discretion, but do so to endorse severity rather than lenience. Even in a time of enhanced attention to mass incarceration, populist reactions often favor increased severity.[276] Prosecutors finely attuned to community sentiment and, more precisely, electoral approval may prove disastrous for vulnerable defendants and victims.[277] The emerging wave of progressive prosecutors, then, generates an opportunity for crafting a theory of prosecution, while simultaneously highlighting the urgent need for some theoretical foundation in which to ground the prosecutorial role.

By tying reforms to a servant-of-the-law paradigm, reform prosecutors also limit the prospect of backlash.[278] A servant-of-the-law orientation folds prosecutors into the underlying mission of the criminal justice system, rather than asking them to override the actions of other powerful actors—even if the ultimate outcome remains the same. One explanation for legislatures’ general indulgence of prosecutorial power over the years is that prosecutors have acted in ways that legislators endorse. If that changes, legislatures may respond. This is a common pattern in other areas, such as gun regulation, where “pro-gun” state legislatures jumped to override “pro-gun-control” localities that strayed from state-wide policy preferences.[279] Another example comes from the not-too-distant past. Just as legislatures sought to rein in lenient judges through sentencing guidelines,[280] legislatures could enact charging or plea bargaining requirements to respond to prosecutors who appear to be flouting the laws,[281] or just ratchet up severity more generally.[282] Or legislators might empower other actors, such as police or victims, to initiate and prosecute cases themselves.[283]

In summary, there are reasons to be wary of enhancing the prosecutor’s freedom to “do justice” as an answer to the system’s failings; that freedom may be part of the problem. An alternative path to reforming prosecutorial excess can achieve many of the goals of the current reform movement, while stressing adherence to the law. A prosecutor zealously serving the law will be less likely to bend or break existing legal rules by overcharging,[284] relying on coercive plea bargains,[285] failing to provide discovery,[286] improperly striking jurors,[287] maintaining faulty cases,[288] engaging in deceptive closing argument,[289] and so on. This model would also support broader cooperation with defense attorneys, such as open-file discovery, and more transparent plea bargaining and charging practices. That cooperation would facilitate the innocence-protective roles of defense attorneys, judges, and juries. Finally, by clarifying lines of accountability, the servant-of-the-law model will shift the reform focus to other actors with more mature capacities to enact lasting changes, like legislators, judges, and police.


While there is abundant criticism of prosecutors, there is no coherent normative theory of the role prosecutors should play in our criminal justice system. We tell prosecutors to “do justice” and hope for the best. The resulting dissatisfaction with prosecutorial behavior should come as no surprise.

A meaningful theory of how prosecutors should behave is particularly important at a moment when the country has finally woken up to mass incarceration.[290] With millions under correctional supervision, dropping crime rates, and a bipartisan desire for change, we stand at the precipice of true criminal justice reform.[291] Many academics and activists look to prosecutors as drivers of that reform. Indeed, some of the harshest critics of prosecutors’ vast discretion seek to repurpose prosecutorial power to revolutionize American justice.[292] But in a nation populated by thousands of prosecutors with widely divergent viewpoints, unbounded prosecutorial discretion to “do justice” will always be a many-edged sword. And as this Article explains, we need not turn over the keys to the justice system to prosecutors to achieve reform. Instead, there are clear benefits to reorienting prosecutors as non-adversarial, servants of the law. By cooperating with other criminal justice actors—particularly defense counsel—to enforce defendant-protective rules, prosecutors could help those actors fulfill their roles in combating injustice. And prosecutors could still dismiss cases and avoid mindless severity in service of the law. Finally, narrowing the role of the prosecutor under a servant-of-the-law model could redirect the focus of reform toward other actors in the criminal justice system, and the system as a whole, to ensure “that justice shall be done.”[293]


Jeffrey Bellin: Professor, William & Mary Law School. Thanks to Dan Epps, Carissa Hessick, Kay Levine, Lauren Ouziel, David Sklansky, and Jenia Turner for comments on an early draft.

           [1].     Angela J. Davis, Arbitrary Justice: The Power of the American Prosecutor 5 (2007); see also Adam M. Gershowitz, Consolidating Local Criminal Justice: Should Prosecutors Control the Jails?, 51 Wake Forest L. Rev. 677, 677–78 (2016) (“No serious observer disputes that prosecutors . . . hold most of the power in the United States criminal justice system.”).

           [2].     David Alan Sklansky, The Nature and Function of Prosecutorial Power, 106 J. Crim. L. & Criminology 473, 474 (2016) (“Much of what is wrong with American criminal justice—its racial inequity, its excessive severity, its propensity for error—is increasingly blamed on prosecutors.”). For a critique of this critique, see Jeffrey Bellin, The Power of Prosecutors, 94 N.Y.U. L. Rev. 171 (2019).

           [3].     See Shima Baradaran Baughman, Subconstitutional Checks, 92 Notre Dame L. Rev. 1071, 1076 (2017) (coining the phrase “the Prosecutor Problem” to describe “what modern scholars claim is responsible for the astronomical increase in incarceration in America in the last fifty years”); Jeffrey Bellin, Reassessing Prosecutorial Power Through the Lens of Mass Incarceration, 116 Mich. L. Rev. 835, 837 (2018) (book review) (“Prosecutors are the Darth Vader of academic writing: mysterious, powerful and, for the most part, bad.”); Bruce A. Green & Fred C. Zacharias, Prosecutorial Neutrality, 2004 Wis. L. Rev. 837, 845–46 (2004) (“[P]rosecutors have been criticized for bringing cases that are too weak or poorly investigated, for bringing prosecutions that are unduly harsh, and for other purported excesses.”); 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, 1329 (1993) (“The American prosecutor has been under nearly constant attack in the criminal procedure literature . . . .”).

           [4].     See Stanley Z. Fisher, In Search of the Virtuous Prosecutor: A Conceptual Framework, 15 Am. J. Crim. L. 197, 201 (1988) (“Discussions of prosecutorial duty in the literature tend to remain either on the level of broad platitudes or, when addressed to particular practices, to lack a coherent framework.”); Green & Zacharias, supra note 3, at 840 (“Commentators need to be more precise about what they expect of prosecutors . . . .”); infra notes 19–26 and accompanying text.

           [5].     Berger v. United States, 295 U.S. 78, 88 (1935).

           [6].     See infra notes 19–26 and accompanying text.

           [7].     See Davis, supra note 1, at 17 (“[E]veryday, legal exercise of prosecutorial discretion is largely responsible for the tremendous injustices in our criminal justice system.”); Kenneth B. Nunn, The “Darden Dilemma”: Should African Americans Prosecute Crimes?, 68 Fordham L. Rev. 1473, 1492 (2000) (“Arguably, the overrepresentation of African Americans in the criminal justice system is due largely to the actions of prosecutors.”); Steven Alan Reiss, Prosecutorial Intent in Constitutional Criminal Procedure, 135 U. Pa. L. Rev. 1365, 1477 (1987) (“It is, however, precisely because prosecutors wield so much power that the restraints which limit exercise of that power must be more discernable and objective.”); James Vorenberg, Decent Restraint of Prosecutorial Power, 94 Harv. L. Rev. 1521, 1572 (1981) (“[E]nvision[ing] a system in which prosecutors have far less power than they have today.”).

           [8].     Deborah Tuerkheimer, Criminal Justice and the Mattering of Lives, 116 Mich. L. Rev. 1145, 1165 (2018) (book review) (calling on prosecutors in this “new era of prosecution” to “recast the pursuit of criminal justice as an antisubordination mandate” and “transform the very meaning of criminal justice”); Collier Meyerson, Prosecutors Keep Their Jobs by Putting People in Jail. Can They Be Leaders in the Fight for Criminal-Justice Reform?, Nation (Nov. 14, 2017), [] (“Recently, a rash of self-proclaimed progressive prosecutors have made public moves toward reform.”); Editorial, A Wiser Generation of Prosecutors, N.Y. Times (Feb. 6, 2017), [] (hailing rise of “[r]eformist prosecutors”).

           [9].     See Ayala v. Scott, 224 So. 3d 755, 756 (Fla. 2017) (describing controversy); see also Kyle Clark, New DA McCann Says Denver is Done with the Death Penalty, (Jan. 10, 2017), [] (reporting on Denver’s chief prosecutor: “I don’t think that the state should be in the business of killing people.”).

         [10].     Statement from Current or Former Judges, Prosecutors, and Law Professors Requesting Reversal of Decision to Remove Ayala from the Markeith Loyd Prosecution, Nat’l Black Prosecutors Ass’n (Mar. 27, 2017), [].

         [11].     Wil S. Hylton, Baltimore vs. Marilyn Mosby, N.Y. Times Mag. (Sept. 28, 2016), [].

         [12].     Id.; Page Croyder, Police Charges in Freddie Gray Case Are Incompetent at Best, Balt. Sun (May 5, 2015), [] (criticizing Mosby’s decision to charge the officers).

         [13].     Philadelphia District Attorney’s Office, New Policies Announced February 15, 2018, []; Manhattan District Attorney’s Office, District Attorney Vance to End the Prosecution of Marijuana Possession and Smoking Cases, (May 15, 2018), []; Tim Prudente, Baltimore Will Stop Prosecuting Marijuana Possession, Mosby Announces, Balt. Sun (Jan. 30, 2019), [].

         [14].     Manhattan District Attorney’s Office, supra note 13; Brian X. McCrone, Marijuana Criminal Cases Dropped En Masse by Philadelphia District Attorney, NBC Phila. (Feb. 15, 2018), [] (“‘I did it because I felt it was the right thing to do,’ Krasner said. ‘We could use those resources to solve homicides.’”).

         [15].     Prudente, supra note 13.

         [16].     See Emily Bazelon & Miriam Krinsky, There’s a Wave of New Prosecutors. And They Mean Justice, N.Y. Times (Dec. 11, 2018), [] (“In the past two years, a wave of prosecutors promising less incarceration and more fairness have been elected across the country.”); Scott Bland, George Soros’ Quiet Overhaul of the U.S. Justice System, Politico (Aug. 30, 2016), [] (chronicling reform movement); Christopher Connelly, National Advocacy Groups Back Candidates To Challenge Local Prosecutors, NPR (Apr. 10, 2018), []; Henry Gass, Meet a New Breed of Prosecutor, Christian Sci. Monitor (July 17, 2017), []; Jon Schuppe, Criminal Justice Reformers Aim Big by Targeting Local DA Races, NBC (Feb. 14, 2018), [].

         [17].     Editorial, supra note 8.

         [18].     Id.; see also Joseph Neff, How Prosecutor Reform is Shaking Up Small DA Races, The Marshall Project (May 1, 2018), [] (highlighting “a new breed of prosecutors” elected across the country, often with “outside help” from progressive sources).

         [19].     295 U.S. 78 (1935).

         [20].     Id. at 85.

         [21].     Id. at 88.

         [22].     See, e.g., Hurd v. People, 25 Mich. 405, 416 (1872) (“The prosecuting officer represents the public interest . . . . His object, like that of the court, should be simply justice . . . .”); Eric S. Fish, Against Adversary Prosecution, 103 Iowa L. Rev. 1419, 1438 (2018) (collecting other examples).

         [23].     Bennett L. Gershman, “Hard Strikes and Foul Blows:” Berger v. United States 75 Years After, 42 Loy. U. Chi. L.J. 177, 196 (2010); see Fish, supra note 22, at 1426 (describing Berger passage as the “most commonly cited” of the “canonical statements” of prosecutor’s role); Bennett L. Gershman, The Prosecutor’s Duty to Truth, 14 Geo. J. Legal Ethics 309, 317 (2001) (describing Berger as the “seminal case defining the prosecutor’s legal and ethical role as a minister of justice”); Michael A. Simons, Prosecutors as Punishment Theorists: Seeking Sentencing Justice, 16 Geo. Mason L. Rev. 303 (2009) (“It is a truism that prosecutors are called not just to win, not just to zealously represent their clients, but rather to ‘seek justice.’”); Sklansky, supra note 2, at 479 (describing Berger as a “hallowed text[]”). A recent Westlaw search reveals that Berger’s exhortation is quoted in over one thousand judicial opinions and seven hundred secondary sources.

         [24].     Model Rules of Prof’l Conduct r. 3.8 cmt. (Am. Bar Ass’n 2019) (“A prosecutor has the responsibility of a minister of justice . . . .”).

         [25].     Nat’l Dist. Attorneys Ass’n, National Prosecution Standards pt. 1-1.1 (3d ed. 2009), [] (“The primary responsibility of a prosecutor is to seek justice . . . .”); Robert Rubinson, Professional Identity As Advocacy, 31 Miss. C. L. Rev. 7, 30 (2012) (characterizing the NDAA as “the preeminent national organization for state prosecutors”).

         [26].     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, 637 (2006); see also Daniel S. Medwed, The Prosecutor as Minister of Justice: Preaching to the Unconverted from the Post-Conviction Pulpit, 84 Wash. L. Rev. 35, 43 (2009) (“[T]his absence of guidance has left tough ethical questions largely in the hands of trial prosecutors alone.”); Fred C. Zacharias, Structuring the Ethics of Prosecutorial Trial Practice: Can Prosecutors Do Justice?, 44 Vand. L. Rev. 45, 58 (1991) (“The framework that the drafters have chosen consists primarily of the ‘do justice’ rule.”)

         [27].     ‘Justice’ is Merriam-Webster’s 2018 Word Of The Year, NPR (Dec. 17, 2018), []; Merriam-Webster’s Word of the Year 2018 (2018), [] (noting that despite a lack of clarity about its content, “justice has been on the minds of many people in 2018”).

         [28].     See U.S. Dep’t of Justice, [].

         [29].     See Hall of Justice (comics), Wikipedia (Sept. 20, 2019), [] (“The Hall of Justice . . . is a fictional headquarters appearing in American comic books published by DC Comics. The Hall of Justice serves as a headquarters for the Justice League.”).

         [30].     The Pledge of Allegiance, USHistory, [].

         [31].     See generally Otto A. Bird, The Idea of Justice (1967) (collecting commentary).

         [32].     J. S. Mill, Utilitarianism 99 (Roger Crisp ed., 1998).

         [33].     See, e.g., Angela J. Davis, The Prosecutor’s Ethical Duty to End Mass Incarceration, 44 Hofstra L. Rev. 1063, 1081 (2016) (arguing that the prosecutor’s duty to do justice “could be fulfilled by working to reduce mass incarceration through the expanded use of diversion and clemency programs”).

         [34].     See, e.g., Nat’l Dist. Attorney’s Ass’n, Message from the President: A Sense of Who We Are, 34 The Prosecutor 5, 5 (Sept./Oct. 2000) (explaining that a prosecutor’s responsibility “to accomplish justice” is “commonly understood” as a directive “to zealously pursue a conviction and then to seek the most severe sentence allowed by law”).

         [35].     Abbe Smith, Can You Be a Good Person and a Good Prosecutor?, 14 Geo. J. Legal Ethics 355, 377 (2001).

         [36].     Kenneth Bresler, Pretty Phrases: The Prosecutor As Minister of Justice and Administrator of Justice, 9 Geo. J. Legal Ethics 1301, 1301 (1996).

         [37].     See Stephanos Bibas, Prosecutorial Regulation Versus Prosecutorial Accountability, 157 U. Pa. L. Rev. 959, 962 (2009) (“Many, if not most, other government actors enjoy less power yet are subject to far more regulation than prosecutors are.”); Steven Alan Reiss, Prosecutorial Intent in Constitutional Criminal Procedure, 135 U. Pa. L. Rev. 1365, 1365 (1987) (“Few officials can so affect the lives of others as can prosecutors. Yet few operate in a vacuum so devoid of externally enforceable constraints.”).

         [38].     See, e.g., Bruce A. Green, Prosecutorial Ethics in Retrospect, 30 Geo. J. Legal Ethics 461, 462 (2017) (pointing to prosecutors’ “political clout” as an obstacle to reform); Niki Kuckes, The State of Rule 3.8: Prosecutorial Ethics Reform Since Ethics 2000, 22 Geo. J. Legal Ethics 427, 433 (2009) (criticizing “the ABA’s seeming lack of political will to tackle prosecutorial ethics”); Nirej Sekhon, The Pedagogical Prosecutor, 44 Seton Hall L. Rev. 1, 45 (2014) (decrying “absence of political will to regulate prosecutors”).

         [39].     See Sklansky, supra note 2, at 516.

         [40].     See Erik Luna & Marianne Wade, Prosecutors as Judges, 67 Wash. & Lee L. Rev. 1413, 1417 (2010) (“[A]cademic solutions to the problems of prosecutorial discretion came in two forms: the promulgation of internal office guidelines to control prosecutorial decision-making and the development of external limitations through restrictive legislation or heightened judicial review.”); Robert P. Mosteller, The Duke Lacrosse Case, Innocence, and False Identifications: A Fundamental Failure to “Do Justice, 76 Fordham L. Rev. 1337, 1341 (2007) (emphasizing “the need for concrete rules”).

         [41].     See Bresler, supra note 36, at 1305 (suggesting that rule makers should “[s]tick with the mission for prosecutors — to seek justice” because “[i]t probably cannot be described more eloquently”); Cassidy, supra note 26, at 640 (“Any attempt to regulate how prosecutors should ‘act’ in certain highly contextualized and nuanced situations by developing more specific normative rules is unworkable.”); Nat’l Dist. Attorney’s Ass’n, supra note 34, at 6 (“A subjective sense of justice is fundamental to a good prosecutor.”).

         [42].     John F. Pfaff, Locked In: The True Causes of Mass Incarceration—and How to Achieve Real Reform 210-11 (2017) (suggesting legislative plea bargaining guidelines); Bibas, supra note 37 (“While many scholars discuss prosecutorial discretion as a problem, most favor external regulation of prosecutors by other institutions.”); Luna & Wade, supra note 40, at 1417 (summarizing the debate between proponents of external control and supporters of internal guidelines); Medwed, supra note 26, at 43–44 ( “Some commentators press for more stringent rules . . . ; others recommend increased training for prosecutors in rendering ethical decisions; and still others suggest . . . ‘a renewed focus’ on virtuous conduct.”); Fred C. Zacharias & Bruce A. Green, The Duty to Avoid Wrongful Convictions: A Thought Experiment in the Regulation of Prosecutors, 89 B.U. L. Rev. 1, 5 (2009) (proposing stronger ethics regulation).

         [43].     Leslie C. Griffin, The Prudent Prosecutor, 14 Geo. J. Legal Ethics 259, 307 (2001) (“Discussions of prosecutorial discretion . . . show much confusion about what [good judgment] is, and what mix of moral and legal reasoning enhances it.”); M. Elaine Nugent-Borakove, Performance Measures and Accountability, in The Changing Role of the American Prosecutor 91, 93 (John L. Worrall & M. Elaine Nugent-Borakove eds., 2008) (“A central problem in the attempt to define performance measures in prosecution is the lack of a clear definition of prosecutorial goals and objectives.”); Sklansky, supra note 2, at 480 (“[T]he absence of coherent aspirations may go a long way toward explaining the dearth of meaningful rules.”).

         [44].     Steve W. Perry & Duren Banks, Prosecutors in State Courts, 2007 – Statistical Tables, Bureau of Just. Stat. 2 (Dec. 2011), [] (“The nearly 25,000 FTE assistant prosecutors employed in 2007 represented a 7% increase from the number reported in 2001 . . . .”).

         [45].     See Roy B. Flemming, The Political Styles and Organizational Strategies of American Prosecutors: Examples from Nine Courthouse Communities, 12 L. & Pol’y 25 (1990) (chronicling varying approaches of chief prosecutors).

         [46].     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, 1344 (1993) (“[T]wo prosecutorial offices in the same state will treat the possession of a small amount of cocaine, a first time property offense, or drunk driving differently.”).

         [47].     See Katherine Barnes et al., Place Matters (Most): An Empirical Study of Prosecutorial Decision-Making in Death-Eligible Cases, 51 Ariz. L. Rev. 305, 360 (2009) (highlighting “large geographic disparities in the rates of . . . death penalty prosecutions and convictions”); Adam M. Gershowitz, Pay Now, Execute Later: Why Counties Should Be Required to Post a Bond to Seek the Death Penalty, 41 U. Rich. L. Rev. 861, 862–63 (2007) (noting that while most Texas counties hadn’t had a death penalty prosecution in decades, Harris County prosecutors “sought the death penalty more than a dozen times per year,” a pattern repeated in Philadelphia, Cook County, and “throughout the country”).

         [48].     See Connelly, supra note 16 (“social justice”).

         [49].     See infra Part II.

         [50].     See, e.g., Sarah Fearing, ‘Oath to Uphold the Law’: Historic Triangle Prosecutors Plan No Change in Handling Marijuana Cases, WY Daily (Jan. 16, 2019), [] (quoting prosecutors’ explanation for why they would continue to prosecute marijuana cases); Press Release, Cal. Dep’t of Justice, Attorney General Kamala D. Harris Issues Statement on Appeal of Court Ruling on California’s Death Penalty (Aug. 21, 2014), [] (explaining decision to defend death penalty law).

         [51].     See, e.g., Josh Bowers, Legal Guilt, Normative Innocence, and the Equitable Decision Not to Prosecute, 110 Colum. L. Rev. 1655, 1665 (2010) (describing such an approach as suitable only for “an imaginary state of affairs that is almost wholly contrary to the qualities that typify modern American criminal justice”); Fish, supra note 22, at 1423 (“[T]he American system gives prosecutors too much discretion for them to act as morally neutral bureaucrats at every stage of the criminal justice process.”).

         [52].     See, e.g., Barbara McQuade, Smart on Crime Charging Policy Provides Roadmap for More Effective Criminal Justice Strategy, 30 Fed. Sent’g. Rep. 207, 208 (2018) (critiquing policy of “strict application of the statute”); Bowers, supra note 51; Fish, supra note 22, at 1423.

         [53].     See Darryl K. Brown, Prosecutors and Overcriminalization: Thoughts on Political Dynamics and a Doctrinal Response, 6 Ohio St. J. Crim. L. 453, 453 (2009) (“Criminal law is overwhelmingly state law . . . .”); Susan R. Klein & Ingrid B. Grobey, Debunking Claims of Over-Federalization of Criminal Law, 62 Emory L.J. 1, 7 (2012) (“The action in criminal law has always been, and continues to be, at the state and local levels.”).

         [54].     Berger v. United States, 295 U.S. 78, 88 (1935); cf. U.S. Const. art. II, § 3 (setting out the executive’s constitutional duty to ensure the laws are “faithfully executed”); Donnelly v. DeChristoforo, 416 U.S. 637, 648–49 (1974) (Douglas, J., dissenting) (“The function of the prosecutor . . . is to vindicate the right of people as expressed in the laws and give those accused of crime a fair trial.”); Robert H. Jackson, The Federal Prosecutor, 31 J. Crim. Law & Criminology 3, 6 (1940) (“[T]he citizen’s safety lies in the prosecutor who . . . serves the law and not factional purposes . . . .”).

         [55].     See Kenneth Culp Davis, Discretionary Justice: A Preliminary Inquiry 194 (1969) (commenting favorably on the German prosecution model); Sklansky, supra note 2, at 493–94 (discussing traditional differences between American and European prosecutors, but noting modern convergence); Thomas Weigend & Jenia Iontcheva Turner, The Constitutionality of Negotiated Criminal Judgments in Germany, 15 German L.J. 81, 101–02 (2014) (discussing constraints on German prosecutors and implications for plea bargaining).

         [56].     United States v. Wilson, 578 F.2d 67, 71 (5th Cir. 1978) (“The prosecutor must be an advocate for Justice.”)

         [57].     See infra Part II.

         [58].     See Wilson, 578 F.2d at 71 (“Caught up in the adversary process and the emotional atmosphere of trial combat, prosecutors too often pursue strategies with a singular determination rather than with a careful deliberation.”); Medwed, supra note 26, at 36 (emphasizing that the “image of the prosecutor as carnivorous aggressor in the adversarial den of the criminal courts is alive and well”); Sklansky, supra note 2; Bellin, supra note 2.

         [59].     See Jackson v. Virginia, 443 U.S. 307, 323 (1979) (describing the “constitutional necessity of proof beyond a reasonable doubt”).

         [60].     See, e.g., U.S. Const. amend. VI (guaranteeing criminal defendants facing incarceration the right to counsel and trial by an impartial jury).

         [61].     Mapp v. Ohio, 367 U.S. 643, 659 (1961).

         [62].     See infra Part I.A.

         [63].     See Brandon L. Garrett, Convicting the Innocent: Where Criminal Prosecutions Go Wrong (2011) (chronicling causes of convictions of innocent defendants); Justin George & Eli Hager, One Way to Deal With Cops Who Lie? Blacklist Them, Some DAs Say, Marshall Project (Jan. 17, 2019), [] (reporting on prosecutors’ decisions to decline cases that depend on police officers with credibility problems).

         [64].     U.S. Const. amend VIII (“Excessive bail shall not be required, nor excessive fines imposed, nor cruel and unusual punishments inflicted.”); Ewing v. California, 538 U.S. 11, 20 (2003) (“The Eighth Amendment, which forbids cruel and unusual punishments, contains a ‘narrow proportionality principle’ . . . .”).

         [65].     U.S. Const. amend VI (protecting the “right to a speedy and public trial, by an impartial jury”).

         [66].     See Brady v. Maryland, 373 U.S. 83, 86 (1963) (prosecutor must disclose all exculpatory information held by government); State v. Crawford, 371 P.3d 381, 388 (Mont. 2016) (explaining that Brady’s rule applies “[i]n all criminal cases”).

         [67].     See Jeffrey Bellin & Shevarma Pemberton, Policing the Admissibility of Body Camera Evidence, 87 Fordham L. Rev. 1425, 1429–32 (2019) (chronicling rapid spread of body cameras).

         [68].     See Brady, 373 U.S. at 86 (requiring disclosure of potentially exculpatory evidence in the possession of the government); Mark Bowes, Chesterfield Prosecutors Plan to Stop Handling Misdemeanor Criminal Cases, Traffic Offenses Starting May 1 Because of Workload, Richmond Times-Dispatch (Feb. 21, 2018), [] (describing a local prosecutor’s decision to stop prosecuting misdemeanor offenses, in part, due to their inability to fulfill an obligation to review police body camera footage); Kimberly Kindy, Some U.S. Police Departments Dump Body-Camera Programs Amid High Costs, Wash. Post (Jan. 21, 2019), [] (“Prosecutors, too, are struggling to keep up with the added workload and cost of body-camera footage.”); Kim Ogg, For Harris County, More Prosecutors Means More Justice, Hous. Chron. (Feb. 9, 2019), [] (explaining that prosecutor’s task now includes “gathering and reviewing a much greater amount of evidence,” including “body camera footage from every officer at a scene”). Prosecutors could also arguably fulfill Brady obligations by disclosing the material without review, although doing so would raise prosecutorial competency and privacy concerns. See Jenia I. Turner, Managing Digital Discovery in Criminal Cases, 109 J. Crim. L. & Criminology 237 (2019).

         [69].     Cf. Josh Bowers, Physician, Heal Thyself: Discretion and the Problem of Excessive Prosecutorial Caseloads, A Response to Adam Gershowitz and Laura Killinger, 106 Nw. U.L. Rev. Colloquy 143, 148 (2011) (“[E]specially in misdemeanor cases, some amount of measured discretion is not just permissible, but essential.”).

         [70].     See Davis v. Mun. Court, 757 P.2d 11, 21 n.9 (Cal. 1988) (discussing judicial authority to reduce “wobbler[s]” charged as felonies to misdemeanors at sentencing); Manduley v. Superior Court, 104 Cal. Rptr. 2d 140, 157 (Cal. Ct. App. 2001) (Nares, J., dissenting) (noting prosecutors’ “unbridled discretion to charge a wobbler as a felony or misdemeanor . . . without any statutory criteria or standards”).

         [71].     See infra Part II.B.

         [72].     Brian Forst, Prosecution Policy and Errors of Justice, in The Changing Role of the American Prosecutor 51, 52 (John L. Worrall & M. Elaine Nugent-Borakove eds., 2008) (“Prosecutors rarely articulate a rationale for determining how these choices are made.”).

         [73].     Nat’l Dist. Attorneys Ass’n, supra note 25, ¶ 1-1.1.

         [74].     Whitney North Seymour, Jr., Why Prosecutors Act Like Prosecutors, 11 Rec. Ass’n B. City N.Y. 302, 312 (1956).

         [75].     Id. at 313; see also Mark Baker, D.A.: Prosecutors in Their Own Words 52 (1999) (“Prosecutors believe they are doing the right thing. . . . You’re completely sure that what you’re doing is the right thing all the time . . . .”); John Cannizzaro, Prosecutorial Ethics: New Insights After 75 Years of Jurisprudence, The Prosecutor Feature, Apr./May/June 2011, at 20, 29 (explaining that prosecutors’ special duty is “to do the right thing”); Amy Weirich, The Changing Role of the District Attorney, Daily Memphian (Dec. 7, 2018), [] (“As I tell our new assistant district attorneys at orientation, our job is to do the right thing every day for the right reason.”).

         [76].     See Mark Godsey, Blind Injustice: A Former Prosecutor Exposes the Psychology and Politics of Wrongful Convictions 21 (2017) (emphasizing the aggressive mindset of prosecutors as a critical failing of criminal justice); Fish, supra note 22, at 1446 (emphasizing that “prosecutors’ aggressiveness has severe consequences”); Bellin, supra note 2; Sklansky, supra note 2.

         [77].     See Jean G. Sturtridge, The Trial Prosecutor and Ethics, in Doing Justice: A Prosecutor’s Guide to Ethics and Civil Liability 83, 83 (Ronald H. Clark ed., 2002) (highlighting the “win-at-any-cost” prosecutors “who zealously represent society’s interest with a deeply rooted conviction that their cause is just” as one of the most problematic in the office); cf. Thompson v. Calderon, 120 F.3d 1045, 1075 (9th Cir. 1997) (Kleinfeld, J., dissenting) (“The majority rightly says that the prosecutor must seek above all to do justice, but seems confused about what justice is.”); United States v. Wilson, 578 F.2d 67, 71–72 (5th Cir. 1978) (“Justice includes the acquittal of the innocent, but Justice also includes the conviction of the guilty.”).

         [78].     See Josie Duffy Rice, Cyrus Vance and the Myth of the Progressive Prosecutor, N.Y. Times (Oct. 16, 2017), [] (criticizing prosecutors who claim to be reformers as insufficiently progressive).

         [79].     State v. Rivera, 99 A.3d 847, 859-60 (N.J. Super. A.D. 2014) (criticizing a prosecutor for getting into the jury box during a witness’ testimony as “at best, a distracting antic inconsistent with the seriousness of the prosecutor’s obligation to do justice”); Watters v. State, 313 P.3d 243, 248 (Nev. 2013) (criticizing a prosecutor’s use of a power point slide that included the defendant’s picture and the word “guilty” as conflicting with a prosecutor’s obligation to “seek justice”).

         [80].     See Rachel E. Barkow, Recharging the Jury: The Criminal Jury’s Constitutional Role in an Era of Mandatory Sentencing, 152 U. Pa. L. Rev. 33, 62 (2003) (“[L]egislators can blame prosecutors if they seek to charge individuals under an overinclusive law.”).

         [81].     See Leon Neyfakh, How to Run Against a Tough-on-Crime DA—and Win, Slate (Nov. 12, 2015), [] (quoting prosecutor Scott Colum: “[E]ven if people don’t know exactly what [prosecutors] do, they know that they’re responsible for public safety”).

         [82].     See Daniel C. Richman & William J. Stuntz, Al Capone’s Revenge: An Essay on the Political Economy of Pretextual Prosecution, 105 Colum. L. Rev. 583, 583 (2005).

         [83].     See Daniel C. Richman, “Project Exile” and the Allocation of Federal Law Enforcement Authority, 43 Ariz. L. Rev. 369, 379 (2001) (summarizing program in Virginia).

         [84].     Ali Watkins, In Fight Against Violent Crime, Justice Dept. Targets Low-Level Gun Offenders, N.Y. Times (May 7, 2018), [].

         [85].     Jennifer A. Tallon et al., The Intelligence-Driven Prosecution Model, Ctr. for Court Innovation 23 (Sept. 2016), [].

         [86].     Id.

         [87].     See, e.g., Hylton, supra note 11 (“I have heard your calls for ‘no justice, no peace!’”).

         [88].     Brooklyn Defender Services, Power of Prosecutors, YouTube (Sep. 10, 2017), [] (“What the public wants to have happen is what the District Attorney should be doing.”); see also Katherine K. Moy et al., Rate My District Attorney: Towards A Scorecard For Prosecutors’ Offices 4 (2018) (proposing a mechanism for rating prosecutor offices that would help reveal “whether a prosecutors’ office has effectively pursued the electorate’s policy priorities”).

         [89].     See Anthony V. Alfieri, Community Prosecutors, 90 Calif. L. Rev. 1465, 1508 (2002) (proposing “to recast the professional community of prosecutors and, in so doing, reinvigorate[] the idea of prosecuting for social justice”); Bazelon & Krinsky, supra note 16 (“In the past two years, a wave of prosecutors promising less incarceration and more fairness have been elected across the country.”)

         [90].     Brian X. McCrone, Marijuana Criminal Cases Dropped En Masse by Philadelphia District Attorney, NBC Phila. (Feb. 15, 2018), [].

         [91].     See Megan Crepeau, After Momentous Week, Prosecutor Kim Foxx Says ‘We Have To Right Wrongs, Chi. Tribune (Nov. 20, 2017), [] (“As a prosecutor’s office, inasmuch as we fight for public safety, we also recognize that we have to right wrongs and be willing to do that.”).

         [92].     Id.

         [93].     Ted Slowik, Foxx Understands Retailers’ Frustrations with Theft Policy; Flossmoor Resident’s Office is prosecuting fewer offenders but says her new approach is addressing the underlying causes of crime; Interview with Cook County State’s Attorney, Daily Southtown (Feb. 25, 2018),

         [94].     See Hylton, supra note 11.

         [95].     Jeremy Kohler, Statement of St. Louis Prosecuting Attorney Robert P. McCulloch, St. Louis Post-Dispatch (Nov. 24, 2014), [] (“It is my sworn duty and that of the grand jury to seek justice, and not simply obtain an indictment or conviction.”).

         [96].     See Aziz Z. Huq, The Difficulties of Democratic Mercy, 103 Calif. L. Rev. 1679, 1697 (2015) (“[T]he rise of prosecutorial discretion since the 1970s coincided with an unprecedented leap in the incarceration rate.”).

         [97].     Lissa Griffin & Ellen Yaroshefsky, Ministers of Justice and Mass Incarceration, 30 Geo. J. Legal Ethics 301, 315 (2017) (“A prosecutor’s charging decision lies at the core of the prosecution function.”).

         [98].     Am. Bar Ass’n, ABA Standards for Criminal Justice Prosecution Function and Defense Function 72 (1993) (Standard 3-3.9 cmt.) (“The charging decision is the heart of the prosecutor function.”). The comment also states: “By its very nature, however, the exercise of discretion cannot be reduced to a formula.” Id.

         [99].     See Bordenkircher v. Hayes, 434 U.S. 357, 364 (1978).

      [100].     Id.

      [101].     See Am. Bar Ass’n, supra note 98, at 70–71 (Standard 3-3.9(a)); Dep’t. of Justice, United States Attorney’s Office Manual ¶ 9.27.200 cmt. (“[F]ailure to meet the minimal requirement of probable cause is an absolute bar to initiating a federal prosecution . . . .”).

      [102].     Bennett L. Gershman, The Prosecutor’s Duty to Truth, 14 Geo. J. Legal Ethics 309, 337 (2001).

      [103].     See Zacharias & Green, supra note 42, at 50 (“The academic literature reflects vigorous disagreement about how convinced of guilt prosecutors should be before bringing or continuing charges.”).

      [104].     See H. Richard Uviller, The Virtuous Prosecutor in Quest of an Ethical Standard: Guidance from the ABA, 71 Mich. L. Rev. 1145, 1159 (1973); see also Maura Ewing, America’s Leading Reform-Minded District Attorney Has Taken His Most Radical Step Yet, Slate (Dec 4, 2018), [] (reporting former Pennsylvania governor and Philadelphia District Attorney Ed Rendell’s disagreement with Krasner about charging homicides, with Rendell arguing “that juries and judges were best situated to determine the facts of a case that might lead to a heightened sentence”).

      [105].     Bennett L. Gershman, A Moral Standard for the Prosecutor’s Exercise of the Charging Discretion, 20 Fordham Urb. L.J. 513, 522 (1993); cf. Am. Bar Ass’n, supra note 98, at 71 (Standard 3-3.9(c)) (“A prosecutor should not be compelled by his or her supervisor to prosecute a case in which he or she has a reasonable doubt about the guilt of the accused.”).

      [106].     See Missouri v. Frye, 566 U.S. 134, 143 (2012) (“Ninety-seven percent of federal convictions and ninety-four percent of state convictions are the result of guilty pleas.”).

      [107].     U.S. Dep’t of Justice, Thornburgh Bluesheet (1989): Plea Policy for Federal Prosecutors, 6 Fed. Sent’g. Rep. 347, 348 (1994) [hereinafter Thornburgh Memorandum]. With respect to the evidentiary threshold, the National District Attorneys Association’s standard states that “[a] prosecutor should file charges” that “she reasonably believes can be substantiated by admissible evidence at trial.” Nat’l Dist. Attorneys Ass’n, supra note 25, ¶ 4-2.2.

      [108].     Thornburgh Memorandum, supra note 107.

      [109].     The Crown Prosecution Service, responsible for charging in England and Wales, requires the prosecutor to find a “realistic prospect of conviction” which “means that an objective, impartial and reasonable jury or bench of magistrates or judge hearing a case alone, properly directed and acting in accordance with the law, is more likely than not to convict the defendant of the charge alleged.” The Code for Crown Prosecutors: The Evidential Stage, The Crown Prosecution Serv. §§ 4.6–­4.7 (Oct. 2018), [].

      [110].     Eric L. Muller, The Hobgoblin of Little Minds? Our Foolish Law of Inconsistent Verdicts, 111 Harv. L. Rev. 771, 785 (1998) (noting that “[j]uries are routinely instructed” that they “must” or “should” convict defendants “whose guilt the government proved beyond a reasonable doubt”).

      [111].     See Albert W. Alschuler & Andrew G. Deiss, A Brief History of Criminal Jury in the United States, 61 U. Chi. L. Rev. 867, 890 (1994) (chronicling the problem of “Southern jury nullification following the Civil War”).

      [112].     But cf. Holly Yan, A Prosecutor Refused to Try an Ex-Frat Leader’s Sex Assault Case. Here are her Eye-Opening Reasons Why, CNN (Dec. 13, 2018) [] (recounting a prosecutor’s explanation for declining to pursue rape charge because “our jurors aren’t ready to blame rapists and not victims”).

      [113].     See Garrett, supra note 63; George & Hager, supra note 63.

      [114].     See, e.g., Ewing, supra note 104 (quoting Larry Krasner on charging: “We are going to proceed on charges that are supported by the facts in the case, period”).

      [115].     Grading of Theft Offenses, 18 Pa. Cons. Stat. § 3903(c)(1) (2017). The statute also singles out certain property theft (e.g., theft of “anhydrous ammonia”) for special treatment regardless of value. Id. § 3903(a)(4).

      [116].     Nat’l Dist. Attorneys Ass’n, supra note 25, ¶ 4-2.4.

      [117].     Id. ¶ 4-1.3 (“Screening: Factors to Consider”), ¶ 4-2.4 (“Charging: Factors to Consider”). The Standards include another twenty factors, and eleven sub-factors, to consider in negotiating a plea agreement. Id. ¶ 5-3.1.

      [118].     Id.

      [119].     Id. ¶ 4-2.4.

      [120].     See infra Part III.A (listing commonly prosecuted offenses).

      [121].     See, e.g., Davis, supra note 1, at 31 (“Prosecutors routinely engage in overcharging, a practice that involves ‘tacking on’ additional charges that they know they cannot prove . . . or that they can technically prove but are inconsistent with the legislative intent or otherwise inappropriate.”); Kyle Graham, Overcharging, 11 Ohio St. J. Crim. L. 701, 702 (2014) (recognizing prevalence of the term, but noting that typically commentators “have failed to explain precisely what they mean by ‘overcharging’”).

      [122].     Am. Bar Ass’n, supra note 98, at 77 (Section 3-3.9 cmt.).

      [123].     See Part II.A. Use of the term “readily” to modify “provable” suggests a limit on the effort that will be spent trying to establish currently unclear offenses. For an example of a prosecutor’s description of her office’s strained interpretation of this limit, see Stephanie M. Rose, Finding Humanity, 30 Fed. Sent’g Rep. 186, 187 (2018).

      [124].     John Ashcroft, Memo Regarding Policy on Charging of Criminal Defendants, Dep’t of Just. (Sept. 22, 2003), (echoing earlier guidance).

      [125].     Ewing, supra note 114.

      [126].     The only limit in the Standards, placed under the header “Improper Leveraging,” is that the prosecutor “should not file charges where the sole purpose is to obtain from the accused a release of potential civil claims.” Nat’l Dist. Attorneys Ass’n, supra note 25, ¶ 4-2.3

      [127].     VA Code Ann. § 18.2-36.1(B) (2004).

      [128].     See Davis v. Mun. Court, 757 P.2d 11 (Cal. 1988); Manduley v. Superior Court, 104 Cal. Rptr. 2d 140 (Cal. Ct. App. 2001). For an example, see Cal. Pen. Code § 473(a) (setting forth penalty for forgery as “imprisonment in a county jail for not more than one year, or by imprisonment” of sixteen months, two years, or three years under Cal. Pen. Code § 1170(h)).

      [129].     United States v. Santos, 553 U.S. 507, 514 (2008) (plurality).

      [130].     Id.

      [131].     Whitman v. United States, 135 S. Ct. 352, 354 (2014) (Scalia, J. writing separately on denial of writ of certiorari) (emphasis removed).

      [132].     21 U.S.C. § 851(a)(1) (“No person . . . shall be sentenced to increased punishment by reason of one or more prior convictions, unless . . . , the United States attorney files an information with the court . . . .”).

      [133].     U.S. Sentencing Comm’n, Application and Impact of 21 U.S.C § 851: Enhanced Penalties for Federal Drug Trafficking Offenders 6 (2018), [].

      [134].     See Bruce Green & Ellen Yaroshefsky, Prosecutorial Accountability 2.0, 92 Notre Dame L. Rev. 51, 91 (2016) (highlighting criticism of “resulting harsh sentences”).

      [135].     California law states, “[a]ny person convicted of a serious felony who previously has been convicted of a serious felony . . . shall receive” an enhanced sentence. Cal. Pen. Code § 667(a)(1) (2019). See also id. § 667(b) (declaration of legislative intent); id. § 667(g) (“The prosecution shall plead and prove all known prior felony serious or violent convictions and shall not enter into any agreement to strike or seek the dismissal of any prior serious and/or violent felony conviction allegation . . . .”). On the other hand, the law also authorizes the prosecutor to “move to dismiss” a strike “in the furtherance of justice,” which could trigger the default to leniency discussed above. See id. § 667(f)(2).

      [136].     See Roger A. Fairfax, Jr., Prosecutorial Nullification, 52 B.C. L. Rev. 1243, 1245 (2011) (exploring “prosecutorial nullification as a distinct species of prosecutorial discretion that presents a set of intriguing, important, and complex questions”).

      [137].     Laura Sullivan, Inmates Who Can’t Make Bail Face Stark Options, NPR (Jan. 22, 2010), [].

      [138].     See Jeffrey Bellin, Waiting for Justice, Slate (Feb. 7, 2018), []. Note that the Supreme Court has not yet recognized a Sixth Amendment constraint on plea bargaining. Cf. Bordenkircher v. Hayes, 434 U.S. 357, 365 (1978) (rejecting constitutional challenge to prosecutor’s use of draconian enhancement as plea bargaining leverage). But the Court has also not ruled it out, leaving prosecutors some freedom to fill in those contours. Id. (“[B]road though [the prosecutor’s] discretion may be, there are undoubtedly constitutional limits upon its exercise.”).

      [139].     See Nat’l Dist. Attorneys Ass’n, supra note 25, § 5-4 cmt. (“There are few prosecutors who have the resources that would be required to try every case.”).

      [140].     Issa Kohler-Hausmann, Managerial Justice and Mass Misdemeanors, 66 Stan. L. Rev. 611, 664 (2014).

      [141].     Adam M. Gershowitz & Laura R. Killinger, The State (Never) Rests: How Excessive Prosecutorial Caseloads Harm Criminal Defendants, 105 Nw. U. L. Rev. 261, 271 (2011).

      [142].     See infra Part II.D.

      [143].     See Bowes, supra note 68.

      [144].     See Jed S. Rakoff, Why Prosecutors Rule the Criminal Justice System—And What Can Be Done About It, 111 Nw. U. L. Rev. 1429, 1430 (2017) (“The plea bargain is the ultimate source of this ever-increasing prosecutorial power.”).

      [145].     I draw this descriptive term from the movie The Godfather. For the clip of the famous “offer you can’t refuse,” see IsiKermesse, I’m Gonna Make Him an Offer He Can’t Refuse, Youtube (Oct. 18, 2010), [].

      [146].     See Jenia Iontcheva Turner, Judicial Participation in Plea Negotiations: A Comparative View, 54 Am. J. Comp. L. 199, 213 (2006) (discussing various plea bargaining strategies including charge and fact bargaining).

      [147].     Cf. Michael Tonry, Sentencing in America, 1975-2025, 42 Crime & Just. 141, 165–69 (2013) (summarizing empirical studies that showed that some prosecutors adjusted their practices to avoid harsh mandatory penalties).

      [148].     See David A. Sklansky & Stephen C. Yeazell, Comparative Law Without Leaving Home: What Civil Procedure Can Teach Criminal Procedure, and Vice Versa, 94 Geo. L.J. 683, 696 (2006) (noting prevalence of civil settlements, “a generally accepted and widely encouraged practice”).

      [149].     Shawn D. Bushway et al., An Explicit Test of Plea Bargaining in the “Shadow of the Trial,52 Criminology, 723, 741 (2014); cf. Anna Offit, Prosecuting in the Shadow of the Jury, 113 Nw. U. L. Rev. 1071 (2019) (concluding from ethnographic research that federal prosecutors internalize such considerations in making case decisions).

      [150].     Bushway et al., supra note 149, at 741; Mona Lynch, Realigning Research: A Proposed (Partial) Agenda for Sociolegal Scholars, 25 Fed. Sent’g. Rep. 254, 255 (2013) (describing research that “revealed how routine criminal matters were resolved in reference to ‘going rates’ for different offenses”).

      [151].     Davis, supra note 1, at 43 (“Both the prosecutor and the defendant reap benefits from plea bargaining.”); Malcolm M. Feeley, The Process Is the Punishment: Handling Cases in a Lower Criminal Court 13 (1979) (“For [defendants] it is quite simply a rational decision, an alternative which they have the full right to exercise.”).

      [152].     See Stephanos Bibas, Plea Bargaining Outside the Shadow of Trial, 117 Harv. L. Rev. 2464, 2467 (2004) (defining the “shadow-of-trial” model as plea bargains shaped “by the strength of the evidence and the expected punishment after trial” and criticizing current plea bargaining practices for deviating too often from that model).

      [153].     See Thea Johnson, Fictional Pleas, 94 Ind. L.J. 855, 855 (2019) (defining fictional pleas as “one[s] in which a defendant pleads guilty to a crime he has not committed, with the knowledge of the defense attorney, prosecutor, and judge”).

      [154].     See supra Part II.B.

      [155].     Lafler v. Cooper, 566 U.S. 156, 162 (2012) (“During plea negotiations defendants are ‘entitled to the effective assistance of competent counsel.’”).

      [156].     Lara A. Bazelon, Hard Lessons: The Role of Law Schools in Addressing Prosecutorial Misconduct, 16 Berkeley J. Crim. L. 391, 398 (2011) (“We are all familiar with the Supreme Court’s seventy-five-year-old admonition that ‘while [a prosecutor] may strike hard blows, he is not at liberty to strike foul ones.’”).

      [157].     Berger v. United States, 295 U.S. 78, 88 (1935).

      [158].     Id.

      [159].     Cf. Am. Bar Ass’n, supra note 98, at 18 (Standard 3-1.5 cmt.) (“In no event, however, should a prosecutor participate or assist in the commission of illegal activity whatever the direction he or she may have received from a supervisor.”).

      [160].     Cf. Zacharias, supra note 26, at 49 (deriving from obligation to “do justice” that “prosecutors must ensure that the basic elements of the adversary system exist at trial”).

      [161].     Shawn Marie Boyne, The German Prosecution Service: Guardians of the Law? 56 (2014).

      [162].     Id. (internal citations omitted).

      [163].     Am. Bar Ass’n, supra note 98, at 36 (Standard 3-2.8 cmt.).

      [164].     See Gideon v. Wainwright, 372 U.S. 335, 344 (1963) (highlighting the critical role of the defense attorney to “our adversary system of criminal justice”).

      [165].     See Mary Sue Backus & Paul Marcus, The Right to Counsel in Criminal Cases, A National Crisis, 57 Hastings L.J. 1031, 1130 (2006) (documenting “the crisis of the American criminal justice system in terms of providing lawyers for poor people”); Ronald F. Wright, Parity of Resources for Defense Counsel and the Reach of Public Choice Theory, 90 Iowa L. Rev. 219, 222 (2004) (explaining that “[p]arity of resources is not the current reality in criminal justice funding”); Zacharias, supra note 26, at 59 (explaining ways in which “a prosecutor enjoys practical advantages over her adversaries”).

      [166].     See United States v. Eley, 723 F.2d 1522, 1526 (11th Cir. 1984) (reviewing a challenge to a prosecutor’s comment in closing argument that invoked the DOJ motto that “the United States wins whenever justice is done”); Daniel Epps, Adversarial Asymmetry in the Criminal Process, 91 N.Y.U. L. REV. 762, 783 (2016) (critiquing prosecutors’ neutrality); Zacharias, supra note 26, at 59 (“Because she represents the community, she commonly carries more influence with juries than attorneys allied solely with individual clients.”).

      [167].     Cf. Jencks v. United States, 353 U.S. 657, 668–69 (1957) (“Because only the defense is adequately equipped to determine the effective use for purpose of discrediting the Government’s witness and thereby furthering the accused’s defense, the defense must initially be entitled to see them to determine what use may be made of them. Justice requires no less.”); Am. Bar Ass’n, supra note 98, at 81 (Standard 3-3.11(a)) (“A prosecutor should not intentionally fail to make timely disclosure to the defense, at the earliest feasible opportunity, of the existence of all evidence or information which tends to negate the guilt of the accused . . . .”).

      [168].     See U.S. Const. amend. VI (guaranteeing criminal defendants the right to counsel, the right to compel witness testimony, and trial by an impartial jury).

      [169].     See McMann v. Richardson, 397 U.S. 759, 771 n.14 (1970) (“It has long been recognized that the right to counsel is the right to the effective assistance of counsel.”).

      [170].     Nat’l Dist. Attorneys Ass’n, supra note 25, § 2-14 cmt.

      [171].     Nat’l Dist. Attorneys Ass’n, supra note 34.

      [172].     Philadelphia District Attorney’s Office, supra note 13, at 3.

      [173].     See Simons, supra note 23, at 313 (quoting 1997 U.S. Attorneys’ Manual § 9-27.710–730 (1997)) (describing policy).

      [174].     See, e.g., 18 U.S.C. § 3142(a) (describing pretrial release requirements that can be imposed by a “judicial officer”).

      [175].     Nat’l Dist. Attorneys Ass’n, supra note 25, ¶ 1-1.1 (“The primary responsibility of a prosecutor is to seek justice.”).

      [176].     See Criminal Justice Standards for the Prosecution Function, Am. Bar Ass’n (2017) [] (Section 3-1.2(f)).

      [177].     Nat’l Dist. Attorneys Ass’n, supra note 25, § 1-1 cmt. (“[A] prosecutor should take an active role in the legislative process when proposals dealing with the criminal justice system are being considered.”).

      [178].     Rachel E. Barkow, Administering Crime, 52 UCLA L. Rev. 715, 728 (2005) (“No other group comes close to prosecutorial lobbying efforts on crime issues.”).

      [179].     See Chiraag Bains, Looking in the Mirror: The Prosecutor’s Role in Ending Mass Incarceration, 30 Fed. Sent’g. Rep. 197, 200 (2018) (urging prosecutors to lead reform efforts); R. Michael Cassidy, (Ad)ministering Justice: A Prosecutor’s Ethical Duty to Support Sentencing Reform, 45 Loy. U. Chi. L.J. 981, 996–97 (2014) (“Prosecutors should thus feel ethically compelled to lend their considerable expertise and political leadership to the emerging movement to repeal mandatory sentences.”); Griffin & Yaroshefsky, supra note 97, at 314 (arguing that prosecutors should be “advocating for repeal of mandatory minimum sentencing provisions for most drug and non-violent offenses”); Rice, supra note 78 (“We should ask: Are prosecutors opposing new mandatory minimum sentences during legislative debates?”).

      [180].     See J. Jonas Anderson, Judicial Lobbying, 91 Wash. L. Rev. 401, 416 (2016) (quoting Justice Rehnquist’s 1993 Report on the Judiciary: “[W]hat is an appropriate sentence for a particular offense, and similar matters, are questions upon which a judge’s view should carry no more weight than the view of any other citizen.”).

      [181].     E. Ann Carson & Elizabeth Anderson, Prisoners in 2015, 14, tbl. 9, Bureau of Justice Statistics Bulletin 14 (Dec. 2016) [].

      [182].     Id.

      [183].     Id.; see also Thomas H. Cohen & Tracey Kyckelhahn, Felony Defendants in Large Urban Counties, 2006, Bureau of Justice Statistics Bull. 3, tbl. 1 (May 2010) [].

      [184].     See Bruce Frederick & Don Stemen, The Anatomy of Discretion: An Analysis of Prosecutorial Decision Making – Technical Report 60 (2012) [] (reporting that “strength of the evidence” is the primary charging factor identified by prosecutors); cf. Stephen Markman, Substantive Limitations on the Criminal Law: Random Thoughts of a Judicial Conservative, in In The Name of Justice 166 (Timothy Lynch ed., 2009) (“There is little in the everyday work of the typical prosecutor in any one of Michigan’s 83 counties, with the possible exception of drug prosecutions, that would upset most libertarians.”).

      [185].     See supra note 125 and accompanying text.

      [186].     William J. Stuntz, The Pathological Politics of Criminal Law, 100 Mich. L. Rev. 505, 506 (2001). Westlaw reports that the article has been cited over 850 times. It has been cited in “secondary sources,” like law review articles, more often than the Supreme Court’s Berger decision.

      [187].     See Miriam H. Baer, Choosing Punishment, 92 B.U. L. Rev. 577, 632 (2012) (recognizing “William Stuntz’s seminal article on overcriminalization, The Pathological Politics of Criminal Law”); Bellin, supra note 2, at 173–74 n.9 (tracing Stuntz’s influence); Todd Haugh, Overcriminalization’s New Harm Paradigm, 68 Vand. L. Rev. 1191, 1201 (2015) (explaining that “almost every scholar working in the area draws on Stuntz”).

      [188].     Richman & Stuntz, supra note 82, at 600.

      [189].     Id.

      [190].     Id. at 605.

      [191].     Id. (“[P]olitically accountable local district attorneys must spend the bulk of their time enforcing a small number of serious crimes. Those crimes are defined nonstrategically. They must be enforced, roughly, as written.”); Henry M. Hart, Jr., The Aims of the Criminal Law, 23 L. & Contemp. Probs. 401, 428 (1958) (describing the exercise of discretion by local prosecutors as consisting “largely of making specifically professional, and inescapable, judgments”).

      [192].     See Bellin, supra note 2 (chronicling academic overstatements of prosecutorial power).

      [193].     Frederick & Stemen, supra note 184, at iii (reporting after empirical study of prosecutor decisions that “prosecutors often rely on a fairly limited array of legal and quasi-legal factors to make decisions”); John L. Worrall, Prosecution in America: A Historical and Comparative Account, in The Changing Role of the American Prosecutor 3, 14 (John L. Worrall & M. Elaine Nugent-Borakove eds., 2008) (“Most prosecutors remain case processors.”).

      [194].     See Bowers, supra note 51, at 1658 (“[T]he need for equitable discretion tends to rise as crime severity falls.”); cf. Griffin & Yaroshefsky, supra note 97, at 321 (“[P]rosecutors should focus more attention on . . . diversion for certain cases that are prosecuted.”); Erica McWhorter & David LaBahn, Confronting the Elephants in the Courtroom Through Prosecutor Led Diversion Efforts, 79 Alb. L. Rev. 1221, 1244 (2016) (encouraging prosecutors to “take the lead” in transforming misdemeanor case processing through diversion programs).

      [195].     See Am. Bar Ass’n, supra note 98, at 69 (Standard 3-3.8(a)) (“The prosecutor should consider in appropriate cases the availability of noncriminal disposition, formal or informal, in deciding whether to press criminal charges which would otherwise be supported by probable cause . . . .”).

      [196].     Id. at 70 (Standard 3-3.8 cmt.) (“Another technique of long standing is for prosecutors not to prosecute an offender who has agreed to enter the military service . . . .”).

      [197].     Nat’l Dist. Attorneys Ass’n, supra note 25, ¶ 4-3.1.

      [198].     Id. ¶ 4-3 cmt.

      [199].     See Wayne R. LaFave et al., Criminal Procedure 703–04 (4th ed. 2004) (under § 13.6, contrasting diversion programs authorized by rules or statutes with “informal and often haphazard” prosecutorial diversion programs).

      [200].     For examples, see City of Phx., Diversion Programs, [] (last visited Mar. 16, 2020) (noting that eligibility depends on the facts of the case, and that a “prosecutor reviews all cases with diversion eligibility issues”); Third Judicial Dist. Attorney (New Mexico), Pre-Prosecution and Diversion Program, [] (listing requirements for participation, including that subjects abide by any “reasonable conditions as deemed necessary by the District Attorney’s Pre-Prosecution Diversion Program”). For a comprehensive exploration of diversion programs, see Am. Bar Ass’n Comm’n on Effective Criminal Sanctions, Report to the House of Delegates on Alternatives to Incarceration and Conviction, 22 Fed. Sent’g. Rep. 62 (2009).

      [201].     Andrew W. Lehren et al., The Prosecutor’s Deal, the Defendant’s Dilemma, N.Y. Times (Dec. 12, 2016), [] (“Your eligibility will depend almost entirely on your prosecutor.”).

      [202].     See supra notes 196–197 and accompanying text .

      [203].     See Bellin, supra note 2, at 1.

      [204].     Id.

      [205].     See infra notes 217–­218 and accompanying text.

      [206].     See McWhorter & LaBahn, supra note 194, at 1244.

      [207].     Feeley, supra note 151, at 234 (reporting that the New Haven pretrial diversion program “increases rather than decreases the harshness of the criminal process”); Josh Bowers, Contraindicated Drug Courts, 55 UCLA L. Rev. 783, 792 (2008) (explaining that studies of New York drug courts “found that the sentences for failing participants in New York City drug courts were typically two-to-five times longer than the sentences for conventionally adjudicated defendants”).

      [208].     Mae C. Quinn, Whose Team Am I on Anyway? Musings of a Public Defender About Drug Treatment Court Practice, 26 N.Y.U. Rev. L. & Soc. Change 37, 58 (2001) (highlighting “the potential for prosecutors to use treatment court as a place to ‘dump’ weak cases” and noting the difficulties faced by defense attorneys in convincing clients to refuse a referral to drug court in such circumstances).

      [209].     See id.; cf. Bowers, supra note 207, at 788 (“[W]hen drug courts imprison failing participants, they punish them not for their underlying crimes, but for their inability to get with the program.”).

      [210].     Shaila Dewan & Andrew W. Lehren, After a Crime, the Price of a Second Chance, N.Y. Times (Dec. 12, 2016), [].

      [211].     Id.; see also State v. Long Fox, 832 N.W.2d 55, 60 (Konenkamp, J., concurring) (highlighting “the void in our law governing deferred prosecutions” by stating “deferred prosecution agreements in South Dakota have no standards, no guidelines for eligibility, and no formalized procedures authorized by legislation. They operate informally and purely at prosecutorial discretion.”); Michael Gordon, Yes, Rich People Have a Better Chance of Getting Off in Court, Public Defender Says, The Charlotte Observer (Sept. 22, 2017), [] (reporting that offenders were required to pay restitution before entering diversion program).

      [212].     D.C. Code § 48-904.01(e)(1-2) (2019); see also Pernell v. United States, 771 A.2d 992, 999 (D.C. 2001) (Ruiz, J., dissenting) (“[This section’s] application bestows important benefits that go beyond the outcome of the immediate criminal proceeding and can have an impact on a person’s subsequent activities in school, employment, etc.”). Virginia courts have found inherent authority to create de facto diversion programs by postponing judgment and then, if certain conditions are met, dismissing them without ever entering a conviction. See Starrs v. Commonwealth, 752 S.E.2d 812, 819 (Va. 2014).

      [213].     D.C. Code § 48-904.01(e)(1-2) (2019); see Margaret Colgate Love, Alternatives to Conviction: Deferred Adjudication as a Way of Avoiding Collateral Consequences, 22 Fed. Sent’g. Rep. 6, 7 (2009) (describing similar programs in other jurisdictions).

      [214].     Love, supra note 213, at 7 (“Deferred adjudication schemes are statutorily authorized in over half the states.”).

      [215].     See supra Part II.B (arguing for a prosecutorial discretion version of the rule of lenity when legislatures dictate choices without guidance).

      [216].     See What is LEAD?, LEAD Nat’l Support Bureau, [].

      [217].     Anna Roberts, Dismissals as Justice, 69 Ala. L. Rev. 327, 330 (2017).

      [218].     Id. at 336

      [219].     See supra Part II.C.

      [220].     See, e.g., Scott E. Sundby, Fallen Superheroes and Constitutional Mirages: The Tale of Brady v. Maryland, 33 McGeorge L. Rev. 643, 644 (2002) (noting the view that the Brady decision requiring disclosure of exculpatory evidence “embodied the prosecutor’s ethical duty to pursue ‘justice’”).

      [221].     Id.; see also United States v. Agurs, 427 U.S. 97, 110 (1976) (“If the suppression of evidence results in constitutional error, it is because of the character of the evidence, not the character of the prosecutor.”).

      [222].     See, e.g., Carter v. State, 738 N.E.2d 665, 672 (Ind. 2000) (discussing prosecutor’s obligation to follow local discovery rules).

      [223].     Berger v. United States, 295 U.S. 78, 89 (1935) (“the misconduct of the prosecuting attorney . . . was pronounced and persistent . . . .”).

      [224].     Brady v. State of Maryland, 373 U.S. 83, 86 (1963).

      [225].     Id.

      [226].     Nat’l Dist. Attorneys Ass’n, supra note 25, Commentary to § 1-1.

      [227].     See Joshua Dressler, Hating Criminals: How Can Something That Feels So Good Be Wrong?, 88 Mich. L. Rev. 1448, 1451 (1990) (“Negative retributivism, accepted by many or perhaps most scholars, holds . . . that it is morally wrong to punish an innocent person even if society might benefit from the action.”).

      [228].     Hart, supra note 191, at 402 (“[E]ach agency of decision must take account always of its own place in the institutional system and of what is necessary to maintain the integrity and workability of the system as a whole.”).

      [229].     Id. at 428 (asserting that prosecuting attorneys “have a lesser role to play, accordingly, in the conscious shaping of the aims of the criminal law”).

      [230].     Culp Davis, supra note 55, at 87.

      [231].     The rigidity of this constraint on German prosecutors historically, and over time, is a subject of debate. See Shawn Boyne, Is the Journey from the In-Box to the Out-Box a Straight Line? The Drive for Efficiency and the Prosecution of Low-Level Criminality in Germany, in The Prosecutor in Transnational Perspective 37, 39 (Erik Luna & Marianne L. Wade eds., 2012) (“This vision of German prosecution practice as a function of the law on the books, rather than organizational incentives, proves to be anachronistic, if not obsolete.”).

      [232].     Joachim Herrmann, The Rule of Compulsory Prosecution and the Scope of Prosecutorial Discretion in Germany, 41 U. Chi. L. Rev. 468, 470–72 (1974).

      [233].     See Marijuana Overview, Nat’l Conf. of State Legislatures (Oct. 17, 2019), [] (describing legalization procedures in eleven states and the District of Columbia, through public votes or legislative action).

      [234].     See supra note 13 (examples in Introduction).

      [235].     See Bellin, supra note 2; Ronald F. Wright, Sentencing Commissions as Provocateurs of Prosecutorial Self-Regulation, 105 Colum. L. Rev. 1010, 1024 n.58 (2005) (discussing resistance to the use of prosecutorial guidelines as the basis of enforceable rights for defendants).

      [236].     The same phenomenon occurs when prosecutors’ diversion programs occupy the space of more comprehensive, legislatively created versions that would limit rather than empower prosecutors.

      [237].     Culp Davis, supra note 55, at 188 (exploring challenge).

      [238].     See Bellin, supra note 2, at 191 (exploring ways in which police possess even more discretion than prosecutors); Culp Davis, supra note 55, at 8 (“Among the most important administrators in America are the police—all 420,000 of them. They make some of our most crucial policies and a large portion of their function is the administration of justice to individual parties.”).

      [239].     See Baker, supra note 75, at 47 (quoting a prosecutor’s observation that “[o]ther agencies have already filtered out the problems with relatively easy solutions” before cases get to prosecutors’ offices).

      [240].     Herrmann, supra note 232, at 484.

      [241].     See Fisher, supra note 4, at 208.

      [242].     Id. See generally Matter of Howes, 940 P.2d 159, 162 (N.M. 1997) (issuing public censure and costs upon prosecutor for violating bar rules by listening to communications from a defendant); Ben Kempinen, The Ethics of Prosecutor Contact with the Unrepresented Defendant, 19 Geo. J. Legal Ethics 1147 (2006). There may also be functional distinctions in the prosecutor and police role that favor police in this context. See Kay L. Levine, The New Prosecution, 40 Wake Forest L. Rev. 1125, 1130 (2005) (highlighting comparative advantages of police over prosecutors at community caretaking functions).

      [243].     See Davis, supra note 1, at 5.

      [244].     See Sunita Sah et al., Blinding Prosecutors to Defendants’ Race: A Policy Proposal to Reduce Unconscious Bias in the Criminal Justice System, 1 Behav. Sci. & Pol’y 69, 70 (2015) (“A recent review of empirical studies examining prosecutorial decision making and race found that most of the studies suggested that the defendants’ ‘race directly or indirectly influenced case outcomes, even when a host of other legal or extra-legal factors are taken into account.’”).

      [245].     See Katherine J. Bies et al., Stanford Crim. Just. Ctr., Stuck In The ‘70s: The Demographics Of California Prosecutors 6 (2015), [] (documenting lack of diversity among prosecutors); David Alan Sklansky, Not Your Father’s Police Department: Making Sense of the New Demographics of Law Enforcement, 96 J. Crim. L. & Criminology 1209, 1210 (2006) [hereinafter Sklansky, Not Your Father’s Police Department] (“The virtually all-white, virtually all-male departments of the 1950s and 1960s have given way to departments with large numbers of female and minority officers, often led by female or minority chiefs.”); David Alan Sklansky, The Progressive Prosecutor’s Handbook, 50 U.C. Davis L. Rev. Online 25, 41 (2017) [hereinafter Sklansky, The Progressive Prosecutor’s Handbook] (“Prosecutors’ offices lag behind police departments in this regard, at least with respect to race and ethnicity.”).

      [246].     See generally Issa Kohler-Hausmann, Misdemeanorland (2018) (documenting enormous burdens placed on defendants whose cases are eventually dismissed in New York City’s misdemeanor courts).

      [247].     Feeley, supra note 151, at 201.

      [248].     See Pub. Citizen v. U.S. Dep’t of Justice, 491 U.S. 440, 454 (1989) (“Where the literal reading of a statutory term would ‘compel an odd result,’ . . . we must search for other evidence of congressional intent to lend the term its proper scope.”) (internal citation omitted).

      [249].     See Johnson v. United States, 135 S. Ct. 2551, 2556–57 (2015) (“The prohibition of vagueness in criminal statutes ‘is a well-recognized requirement, consonant alike with ordinary notions of fair play and the settled rules of law,’ and a statute that flouts it ‘violates the first essential of due process.’”).

      [250].     See United States v. Bass, 404 U.S. 336, 347 (1971) (“[A]mbiguity concerning the ambit of criminal statutes should be resolved in favor of lenity.”) (citing Connally v. Gen. Constr. Co., 269 U.S. 385, 391 (1926)).

      [251].     See John F. Manning, The Absurdity Doctrine, 116 Harv. L. Rev. 2387, 2388 (2003) (“From the earliest days of the Republic, the Supreme Court has subscribed to the idea that judges may deviate from even the clearest statutory text when a given application would otherwise produce ‘absurd’ results.”).

      [252].     See Amy Forliti, Judge Dismisses ‘Absurd’ Case Against Minnesota Teen Who ‘Sexted’ Boy, Associated Press (March 16, 2018), [].

      [253].     See Missouri v. Frye, 566 U.S. 134, 143 (2012) (“Ninety-seven percent of federal convictions and ninety-four percent of state convictions are the result of guilty pleas.”).

      [254].     See Davis, supra note 33, at 1085 (“As ministers of justice, prosecutors have an ethical duty to take action to correct injustices in the criminal justice system.”); Cassidy, supra note 179, at 983 (“Prosecutors have a duty . . . to think about the delivery of criminal justice on a systemic level, promoting criminal justice policies that further broader societal ends.”); Alexes Harris, Justice Shouldn’t Come With a $250 Fine, N.Y. Times (Jan. 3, 2018), [] (“New prosecutors have the power to stop coloring within the lines of our unjust, unfair and unrealistic systems of justice.”).

      [255].     See Bruce Green, Urban Policing and Public Policy — The Prosecutor’s Role, 51 Ga. Law Rev. 1179, 1199 (2017) (arguing that prosecutorial discretion “can serve as a check on bad legislative decisions”); K. Babe Howell, Prosecutorial Discretion and the Duty to Seek Justice in an Overburdened Criminal Justice System, 27 Geo. J. Legal Ethics 285, 288 (2014) (calling on “prosecutors’ offices to respond to and ameliorate the disparate impact and failure of the justice system caused by policing of minor offenses”); David Alan Sklansky, The Changing Political Landscape for Elected Prosecutors, 14 Ohio St. J. Crim. L. 647, 655–56 (2017) (noting that Baltimore District Attorney Marilyn Mosby, who campaigned on a promise “to be more aggressive in seeking charges against police officers,” was criticized for having “moved too quickly and overcharged” the officers involved in death of Freddie Gray, even though the questionable charges “helped to calm Baltimore in the wake of Gray’s death”).

      [256].     Ewing, supra note 114.

      [257].     Fair and Just Prosecution, 21 Principles For the 21st Century Prosecutor 5 (2018), [].

      [258].     See supra Part II.B.

      [259].     See supra Parts II.B (charging and enhancements); II.C (diversion).

      [260].     Maria Cramer, DA Candidate Rachael Rollins Hailed Nationally, But Locally Her Plan Not to Prosecute Petty Crimes Alarms Some, Bos. Globe (Sept. 12, 2018), []; Rachael Rollins, Suffolk Dist. Attorney, Charges to be Declined, [] (listing offenses).

      [261].     See Fair and Just Prosecution, supra note 257, at 10 (“In general, do not charge misdemeanors, such as trespassing or loitering, which are associated with poverty, mental illness, and homelessness.”); id. at 11 (“In general, do not charge sex workers or clients when both parties are over 18 and consent.”).

      [262].     Rollins, supra note 260.

      [263].     See Paul T. Crane, Charging on the Margin, 57 Wm. & Mary L. Rev. 775, 797 (2016) (“One common practice, especially in larger offices, is to designate a group of prosecutors as having primary responsibility for screening incoming cases and making charging decisions.”); Ronald Wright & Marc Miller, The Screening/Bargaining Tradeoff, 55 Stan. L. Rev. 29, 63 (2002) (“All attorneys in the Screening Section served previously (usually a couple of years) in the Trial Section.”).

      [264].     Cramer, supra note 260.

      [265].     See Bellin, supra note 3, at 846­–47 (discussing empirical data on prosecutorial dismissals).

      [266].     Fair and Just Prosecution, supra note 257, at 17; Sklansky, The Progressive Prosecutor’s Handbook, supra note 245, at 33 (“Have a clear, generous, and administrable disclosure policy.”); see also supra Part II.D.

      [267].     See Fair and Just Prosecution, supra note 257, at 19; Sklansky, The Progressive Prosecutor’s Handbook, supra note 245, at 38 (“Investigate police shootings independently and transparently.”); supra Parts II.A & B.

      [268].     Supra Parts II.A & B.

      [269].     See Sklansky, The Progressive Prosecutor’s Handbook, supra note 245, at 32 (encouraging prosecutor’s offices to make space for “objectively reviewing claims of wrongful conviction”).

      [270].     Levine, supra note 242, at 1211.

      [271].     See Davis, supra note 1; Gershowitz, supra note 1; Bibas, supra note 37; Reiss, supra note 37.

      [272].     Davis, supra note 1, at 17.

      [273].     Cf. Norman W. Spaulding, The Rule of Law in Action: A Defense of Adversary System Values, 93 Cornell L. Rev. 1377, 1392 (2008) (describing the prospect of “moral activism by prosecutors” as “chilling”).

      [274].     Perry & Banks, supra note 44 (“The nearly 25,000 FTE assistant prosecutors employed in 2007 represented a 7% increase from the number reported in 2001.”).

      [275].     See Harris, supra note 254 (advocating that prosecutors “stop coloring within the lines of our unjust, unfair and unrealistic systems of justice”).

      [276].     See Rachel Elise Barkow, Prisoners of Politics 5–6 (2019) (chronicling a “pathological political process that caters to the public’s fears and emotions” that allows reform efforts to be “derailed with a single story”); John Ehrett, Public Choice and the Mandatory Minimum Temptation, 35 Yale L. & Pol’y Rev. 603, 604 (2017) (discussing legislative enactment of new mandatory minimum sentences in response to lenient sentence given to Brock Turner); Huq, supra note 96, at 1698 (noting that in America, “[p]ublic punitiveness” typically pushes democratic levers to create “an expansion in criminal liability and harsh, mandatory sentences”).

      [277].     See Sklansky, supra note 255, at 673 (“The danger of politicizing the handling of particular cases is, in fact, a worrisome aspect of the growing attention voters seem to be paying to prosecutorial elections.”).

      [278].     See, e.g., Nicole Malliotakis, Progressive Prosecutors’ Pathetic Retreat, N.Y. Post (Feb. 4, 2018), [] (criticizing NYC prosecutors’ declaration that they would not prosecute turnstile jumpers: “You can now ride the subway for free.”).

      [279].     See Jeffrey Bellin, The Right to Remain Armed, 93 Wash. U.L. Rev. 1, 17 (2015) (describing movement in state legislatures to use state law to “override local gun control preferences” in cities like Philadelphia, Atlanta, Cleveland, and San Francisco).

      [280].     See National Research Council, The Growth of Incarceration in the United States 71-73 (Jeremy Travis et al. eds., 2014) (explaining legislative reaction to perception of lenient sentences, with mandatory minimum sentences, three strikes laws, and truth in sentencing requirements).

      [281].     See, e.g., Cal. Pen. Code § 667(g) (2019) (“Prior serious or violent felony convictions shall not be used in plea bargaining . . . .”); Wash. Rev. Code § 9.94A.421(6) (2010) (explaining that “in no instance may the prosecutor agree not to allege prior convictions” as part of a plea agreement).

      [282].     See Michael M. O’Hear, Plea Bargaining and Victims: From Consultation to Guidelines, 91 Marq. L. Rev. 323, 346 (2007) (“Legislators may be trying to use a predictably tough sentencing system to offset an arbitrarily lenient plea bargaining system.”); Ronald F. Wright, Prosecutorial Guidelines and the New Terrain in New Jersey, 109 Penn St. L. Rev. 1087, 1093 (2005) (describing the evolution of New Jersey plea bargaining guidelines).

      [283].     See, e.g., Alexandra Natapoff, When the Police Become Prosecutors, N.Y. Times (Dec. 26, 2018) [] (“In hundreds of misdemeanor courts in at least 14 states, police officers can file criminal charges and handle court cases, acting as prosecutor as well as witness and negotiator.”); cf. Luke Broadwater, Gov. Hogan Calls on Attorney General Frosh to Prosecute More Baltimore Criminal Cases, Slams City Prosecutors, Balt. Sun (Sep. 18, 2019), [] (reporting on efforts to transfer prosecuting authority from progressive prosecutor to other officials); Akela Lacy & Ryan Grim, Pennsylvania Lawmakers Move to Strip Reformist Prosecutor Larry Krasner of Authority, The Intercept (July 8, 2019), [] (same).

      [284].     See supra Part II.B.

      [285].     See supra Part II.C.

      [286].     See Brady v. Maryland, 373 U.S. 83, 86 (1963) (holding that a prosecutor must disclose material exculpatory information).

      [287].     See Batson v. Kentucky, 476 U.S. 79, 89 (1986) (“[T]he Equal Protection Clause forbids the prosecutor to challenge potential jurors solely on account of their race . . . .”).

      [288].     See supra Part II.A.

      [289].     See Berger v. United States, 295 U.S. 78, 82–88 (1935) (reversing conviction based on improper statements in prosecutor’s closing argument).

      [290].     See Kimberlé W. Crenshaw, From Private Violence to Mass Incarceration: Thinking Intersectionally About Women, Race, and Social Control, 59 UCLA L. Rev. 1418, 1421 (2012) (“Michelle Alexander’s widely discussed The New Jim Crow has provided a popular framing of the problem that has galvanized a critical mass of the population that may finally be willing to listen.”).

      [291].     See Barack Obama, The President’s Role in Advancing Criminal Justice Reform, 130 Harv. L. Rev. 811, 817 (2017) (“There is a growing consensus across the U.S. political spectrum that the extent of incarceration in the United States is not just unnecessary but also unsustainable.”).

      [292].     See, e.g., Davis, supra note 33, at 1070–77 (attributing the problem of mass incarceration, in part, to prosecutorial discretion in charging, plea bargaining, and sentencing); id. at 1085 (“Prosecutors have played a significant role in creating the crisis of mass incarceration, and they have an ethical duty to do whatever they can to fix it.”).

      [293].     Berger, 295 U.S. at 88.

More in this Issue

John Robinson Wilkins and the Resources of the Law: Testing the Limits of Race, Law and Development, and the American Legal Profession

In the fall of 1964, my uncle John Robinson Wilkins joined the Berkeley Law School faculty. He was the first black professor in the school’s illustrious history and only the second in the entire UC Berkeley system. Tragically, my uncle’s time on the Berkeley faculty would be short. In 1970, he was diagnosed with an […]

Shadow Governance

Corporations have something to say about some of the most important social and economic issues of our time—and one way they say it is through shadow governance. This Article spotlights a group of influential corporate policies comprising what we call “shadow governance.” These non-charter, non-bylaw governance documents express a corporation’s commitment to and process on […]

The Title IX Paradox

When Christine Blasey Ford explained to the Senate Judiciary Committee why she had not reported her sexual assault at age fifteen, she captured the struggle of many children who must decide whether to make such reports: “For a very long time, I was too afraid and ashamed to tell anyone the details.” Thousands of sexual […]

The Conscience Defense to Malpractice

This Article presents the first empirical study of state conscience laws that establish explicit procedural protections for medical providers who refuse to participate in providing reproductive health services, including abortion, sterilization, contraception, and emergency contraception. Scholarship and public debate about law’s role in protecting health care providers’ conscience rights typically focus on who should be […]

Fleeing for Their Lives: Domestic Violence Asylum and Matter of A-B-

This Note argues that it is a mistake to classify domestic violence as a primarily “private” crime given its widespread and gendered nature. Further, in some cases, the infliction of domestic violence is ignored—if not condoned—by state actors, casting doubt upon Attorney General Jeff Sessions’s claim in Matter of A-B-that such violence does not involve […]