Who Shouldn’t Prosecute the Police

I.     Introduction

The prosecutor’s role in modern criminal law is among the most heavily theorized. Do prosecutors have too much power?1 What is the appropriate balance between plea bargaining and trials?2 What is a prosecutor’s motivation in a given case or set of cases?3 Do prosecutors bring too harsh charges? Too few charges?4 These are just a few of the questions about prosecutors that scholars contemplate with rigor and complexity. Yet, prosecutorial conflicts are largely absent from this analysis.5 In the following pages, I use conflict-of-interest law to conduct a novel examination of a local prosecutor’s conflicting values when she must charge and lead cases against the police in her own jurisdiction.

The nonindictments after police officers choked Eric Garner to death and fatally shot Michael Brown have forced many to once again confront questions about police violence and criminality. As part of that conversation, politicians are starting to focus on the dysfunction of our local, adversarial justice system as it relates to police killings of civilians.6

A number of scholars have also commented on whether local prosecutors should bring charges and lead cases against their closest professional allies.7 These commentaries have called the police–prosecutor relationship too close, describing it as a threat to prosecutorial legitimacy, and stated that prosecutors who work hand-in-hand with police officers cannot “honestly be expected to be impartial and aggressive.”8 Such statements closely mirror the justifications for disqualification or recusal of lawyers and judges in conflict-of-interest cases,9 yet there has been no rigorous analysis how conflict-of-interest law relates to the local prosecutions of police officers.10 

This Article argues that conflict-of-interest law provides the most coherent framework for examining local police prosecutions. A conflict-of-interest-law lens presents a compelling justification for the logical but passing conclusion drawn by many scholars and policymakers that there is something structurally problematic about the role of a local prosecutor in police–defendant cases.11 Conflict-of-interest law—in particular, the Supreme Court’s Fourteenth and Sixth Amendment rulings about other actors in the criminal justice system—asks courts and lawmakers to look at the appearance of fairness, as well as the personal and professional entanglements that may affect a local prosecutor’s ability to fairly review evidence and pursue cases against police. Applying conflict-of-interest law to local prosecutions of police reveals a disturbing picture of a prosecutor, accustomed to playing on the same team as law enforcement, who must switch roles when police become suspects and defendants.

Part II of this Article addresses the dearth of scholarship focused on conflicts between prosecutors and police–defendants.12 This gap is then contrasted with the rich legal and theoretical material that analyzes other conflict-of-interest law situations. The conflicts of judges and attorneys are the subject of many ethical canons, statutes, and Supreme Court rulings.13 A robust scholarly discussion has taken place about, among other things, which personal and professional biases necessitate recusal or disqualification, whether attorneys and judges are able to police their own conflicts, and the centrality of the appearance of justice to maintaining public confidence in the courts.14 While one of these conflicts is often enough to mandate judicial or attorney recusal,15 the convergence of all of these conflicts in one actor or case rattles the very foundation of the adversary process.

Part III examines the many points at which prosecutors who are tasked with leading cases against the police face conflicts, both actual and perceived. It addresses the issues of professional reliance, personal relationships, democratic legitimacy, and perceived bias that arise when local district attorneys use their discretion to decline to bring charges against police, often in secret,16 or to charge police and pursue those cases.

Part IV maps out a more rigorous legal underpinning for the descriptive analysis described in Part III and shows that the theory of conflict-of-interest law mandates the removal of local prosecutors from cases involving
police–defendants. Finally, Part V suggests several other actors who could prosecute the police and addresses the benefits and costs to each proposed solution.

The Article concludes that when local prosecutors face
police–defendants, the convergence of unacceptable appearance concerns and undeniable conflicting professional interests, impacts the very legitimacy of the prosecutor’s role in our criminal justice system, and necessitates an overhaul of the status quo. So long as we continue to use the criminal justice system to discipline police who commit crimes, we must ensure that they are prosecuted by an unconflicted actor.

II.     A Theory of Conflict-of-Interest Law

As lawmakers and scholars look for ways to ensure actual and perceived fairness in the prosecution of criminal police activity, the question of prosecutorial bias is at the forefront.17 But, while we may call for recusals in high-profile cases, few theoretical or legal underpinnings exist to bolster these demands. Conflict-of-interest law can fill this vacuum. In the following Subparts, I will review the thin treatment that scholars and courts tend to give to prosecutorial conflicts and suggest reasons why, despite the logical connection, conflicts with police–defendants have not been raised. The rest of the Part will supply the theory, drawn from conflict-of-interest law and scholarship, which can and should be marshaled to provide support for the calls to disqualify local prosecutors from handling cases against law enforcement defendants. I tease out a number of categories, including the appearance of justice, the potential for financial and personal conflicts of interest, and the legal actor’s inability to perceive her own conflict, that combine to create a structural and unwaivable conflict when local prosecutors face police–defendants.

A.     Undertheorized Prosecutorial Conflicts

While prosecutors have the well-known duty to “seek justice,”18 how that relates to cases where they may have a conflict of interest has not been explored. The Supreme Court has not directly looked at the conflicts prosecutors face,19 despite the many conflict cases it has ruled on regarding other legal actors.20 It has addressed prosecutorial conflicts in dicta, where it has made inconsistent comments regarding the prosecutor’s duty to ensure impartiality and the systemic appearance of justice. The Court has stated prosecutors have an “obligation to govern impartially [that] 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 The Court has cautioned that, “[t]he rigid [conflict] requirements . . . designed for officials performing judicial or quasi-judicial functions, are not applicable to those acting in a prosecutorial or plaintiff-like capacity.”22 Yet, if a conflict exists for a prosecutor, no other factor can ameliorate the structural flaw: “[a] concern for actual prejudice in such circumstances misses the point, for what is at stake is the public perception of the integrity of our criminal justice system.”23 Some lower courts have acknowledged that a prosecutor’s role is that of a “quasi-judicial officer,” which implicates the same due process appearance concerns as that of a conflicted judge.24 Yet in most cases, lower courts have maintained that the appearance of impropriety is not enough to disqualify a prosecutor.25 These statements and rulings do not articulate a clear set of values for the prosecutor’s role in ensuring that the system does justice.

The Supreme Court has not looked at a prosecutor’s personal and professional conflicts either. A federal statute,26 some state statutes,27 and the ABA standards28 allow for the disqualification of a prosecutor when she is related, personally or professionally, to a defendant in a case. But while these laws and rules have been addressed by some lower court rulings,29 they are rarely the subject of criminal law scholars. Nor do any of these scholars address the important issue of a prosecutor’s conflicting interests when the police are defendants. 30

One reason for this lack of scrutiny is that a defendant usually raises the issue of a conflict with a judge or attorney. Because prosecutors do not have a specific client, their conflicts are not scrutinized as closely, and decisions about such conflicts are often left entirely in the hands of the prosecuting attorneys themselves.31 Moreover, any claim by a defendant that a prosecutor has a conflict will be in the posture that she has been overzealous in the prosecution of a case, which is not likely to be an issue in police cases where the entangled relationship will tend to lead to leniency rather than harshness.32 Still, these barriers do not explain why no scholar has sought to connect prosecutorial bias in favor of the police with conflict-of-interest law. This lack of scrutiny is particularly problematic given the attention now placed on the seeming under-enforcement of the criminal law when applied to police. Below are several recurring and important themes in conflict-of-interest law applied to other actors that are particularly germane to the problem of local police prosecutions.

B.     Appearance of Justice

The core of much conflict law and theory is based on the notion that the legal system must appear just. The maxim that “justice must satisfy the appearance of justice” is central to the Supreme Court’s due process rulings on judicial disqualifications and, to a lesser extent, its Sixth Amendment rulings on attorney conflicts of interest.33 While, “at common law, the presumption of [judicial] impartiality was irrebuttable,”34 as the law and concepts of judicial neutrality35 developed, this presumption was largely subsumed under the question of whether a judge’s conflicts or actions led her to appear impartial.36 The Supreme Court has made clear that, even in cases where a judge is accused of an actual conflict, the appearance of bias is of utmost concern.37 Thus, the Court has said that “[d]ue process ‘may sometimes bar trial by judges who have no actual bias and who would do their very best to weigh the scales of justice equally between contending parties.’”38

Federal law requires that a judge recuse herself among other reasons “in any proceeding in which h[er] impartiality might reasonably be questioned.”39 The appearance of impartiality is an essential ingredient of many states’ judicial disqualification statutes,40 and also figures heavily into the ethical canons and statutes that govern judicial recusal or disqualification.41 Lower courts have disqualified judges when they have not found, nor even looked, for actual partiality.42 Recently, Judge Kozinski began a dissent from an en-banc decision of the Ninth Circuit by forcefully stating that he understood the judicial oath of office “to mean that we must not merely be impartial, but must appear to be impartial to a disinterested observer.”43 His dissent remonstrated his colleagues for failing this test when the court upheld the conviction of a defendant who, from assessing the trial record, “[would] have had a fairer shake in a tribunal run by marsupials.”44

Many scholars have also emphasized that judges must appear impartial. Martin Reddish and Laurence Marshall have noted “if there exists any reasonable doubt about the adjudicator’s impartiality at the outset of a case, provision of the most elaborate procedural safeguards will not avail to create this appearance of justice.”45 In other words, no amount of process-oriented protections will ensure the legitimacy of the judicial system if the judge does not appear impartial. This is so both on a micro-level—a jury must not perceive that a judge favors one side or another,46 and on a macro-level—the very functioning of the court system relies on the public’s belief that it has access to impartial tribunals.47

So too is the appearance of justice an important feature in cases and theories behind attorney disqualification or recusal. Of course, it is less a factor in these cases because, unlike judges, there is no corresponding responsibility for a lawyer to appear impartial.48 In order to maintain confidence in the court system, however, lawyers must appear to be unconflicted in their zealous representation of a client. To this end, the Supreme Court has prioritized the appearance of a fair trial, even over other constitutional requirements. In Wheat v. United States, the Court upheld a conviction when the trial court had disqualified defense counsel for a potential conflict of interest, even though the defendant had explicitly waived the conflict.49 In doing so, it emphasized, among other things, that “courts have an independent interest in ensuring that criminal trials are conducted within the ethical standards of the profession and that legal proceedings appear fair to all who observe them.”50 Thus, the Court prioritized the right to conflict-free counsel and to a trial that satisfies the appearance of justice above the Sixth Amendment right to defense counsel of one’s choosing.51

Thus, the appearance of justice is a bedrock principle of constitutional, statutory and common law conflict rulings. While the appearance-of-justice standard has historically been applied to judges,52 the principle applies with equal force to prosecutors who perform an already acknowledged quasi-judicial function. It is particularly applicable because the vast majority of criminal defendants’ cases are adjudicated via plea bargain, where both the charges pled to and the sentencing decision are largely determined by prosecutors with little judicial review.53

C.     Actual and Possible Bias

Another set of themes that emerge from conflict-of-interest law and scholarship are that judges and lawyers may be disqualified based on actual or possible bias. These biases reveal themselves in many forms and in thousands of cases.

For judges, the Due Process Clause requires, at least, an arbiter “with no actual bias against the defendant or interest in the outcome of his particular case.”54 At common law, “the only accepted ground for disqualifying a judge was pecuniary interest.”55 Thus, most due process conflict-of-interest law arises out of cases where judges have a financial interest rather than personal or otherwise.56 But the Court has found disqualification constitutionally necessary in cases where the financial interest was somewhat attenuated.57 For instance in Aetna Life Insurance Co. v. Lavoie, because an Alabama Supreme Court judge was making common law about an area where he had a direct interest, due process required his disqualification.58 Yet, while the judge did have an interest in cases related to the one at hand, there was no direct financial gain for him in deciding that particular case.59 As Justice Brennan wrote in his concurrence: “[A]s this case demonstrates, an interest is sufficiently ‘direct’ if the outcome of the challenged proceeding substantially advances the judge’s opportunity to attain some desired goal even if that goal is not actually attained in that proceeding.”60 Justice Brennan’s reading of due process requirements is also reflected in federal and state statutory law.61

Many scholars have questioned why a financial motive should be the main focus of judicial disqualification. Redish and Marshall, for example, question why “[t]he Court [has not] explain[ed] why a ‘possibility’ of a judge being swayed by financial self-interest is a constitutional matter, while the fact that a judge harbors either a personal prejudice against or a predisposition toward a litigant is not.”62 To this end, federal law requires judicial recusal in situations that do not involve a financial interest.63

Similarly, lawyers have recused themselves or have been disqualified in a host of situations where personal bias has been alleged. In one opinion, Justice Scalia noted critically that lower courts had reversed convictions in cases stemming not just from the classic multiple-representation conflict but also:

[W]hen . . . there is a conflict rooted in counsel’s obligations to former clients, [and] when representation of the defendant somehow implicates counsel’s personal or financial interests, including a book deal, . . . the teaching of classes to Internal Revenue Service agents, a romantic “entanglement” with the prosecutor, or fear of antagonizing the trial judge.64

While Scalia may have been lamenting the reversal of convictions for what he considered attenuated conflicts, his list serves another purpose here: it shows how seriously the courts take both actual—but also possible—conflicts, and in how many varying situations and degrees of remove from a particular controversy such conflict rules are applied.

D.     An Actor’s Inability to Determine Her Own Conflict

The Supreme Court and scholars agree that it is very difficult for a judge or an attorney to determine her own conflicts of interest. In particular, it is difficult for any legal actor faced with a potential conflict to determine how much it will impact her judgment or the quality of her representation. The Court has called attorney conflicts “notoriously hard to predict.”65 It has held that “the Due Process Clause [must be] implemented by objective standards that do not require proof of actual bias” in judicial disqualification cases because of “[t]he difficulties of inquiring into [one’s own] actual bias, and the fact that the inquiry is often a private one.”66

Scholars who have examined judicial and attorney refusals to recuse themselves have also found that an actor’s own assessment of her partiality is not reliable for a number of reasons having nothing to do with her conscious motives.67 As Tigran Eldred explained, behavioral economics, which is applied by scholars to numerous decisional situations, also tells us something about the cognitive problems a judge or attorney may have when faced with a possible conflict.68

Eldred identifies three biases, drawn from behavioral economics, which result in what he terms “bounded ethicality” in conflict determinations.69 One is described as the “self as moral” bias or “illusion of objectivity,” where a person has a “tendency to believe oneself as more ‘honest, trustworthy, ethical, and fair than others.’”70 The second is the “self as competent” bias, wherein a person sees herself, falsely in many circumstances, “as being better than others in possessing a series of desirable attributes.”71 The final bias is the “self as deserving,” bias, where “people allocate more responsibility to themselves for contributions to an outcome than they actually deserve.”72 These biases are made more complicated by the fact that they have been found to be “stubborn”: the person reviewing her own conflict, “not aware of [these biases’] existence, will tend to believe that he or she acted ethically, even in the face of evidence to the contrary.”73

Such unconscious biases infect a forward-looking decision about a conflict, as well as a backward-looking justification for a refusal to recuse oneself from a case or representation.74 For instance, cases of successive representation—where a current client’s interests may be at odds with arguments made on behalf of a former client—may appear to only affect an attorney’s representation of a present client if it does not hurt the former client’s interests, but such a representation may also have a subconscious impact on the current attorney-client relationship. There may be unconscious loyalty to the former client that affects the current attorney-client relationship, or the attorney may believe that her past representation achieved the correct result but must argue for the opposing result in the current case. Another scholar has noted, relatedly, that issues of “loyalty and gratitude” to the political entities that helped appoint and confirm a federal judge may factor into her decisions more than she realizes and require closer scrutiny of recusal decisions.75

Conflict-of-interest law is applied rigorously to many system actors and is the subject of much scholarly debate. The themes addressed above—appearance of justice, personal conflict, and the difficulty in determining one’s own conflict—can also be deployed to elucidate the structural conflict of interest that occurs when local prosecutors face law enforcement defendants.

III.     Describing the Conflict When Local District Attorneys Must Investigate the Police

In this Part, I will explain in detail the many points at which prosecutors and law enforcement are entangled both on a daily basis, and on larger systemic and political levels.

At the outset, it is worth noting a number of other legal barriers preventing the prosecution of law enforcement. A combination of special procedural protections for officers,76 officer-favorable state statutory self-defense laws,77 the heightened protections given to law enforcement by two Supreme Court cases,78 and the natural bias jurors may have in favor of law enforcement79 shed some light on why so few officers are indicted, let alone convicted, of criminal acts under state or federal law, absent any potential structural prosecutorial bias in favor of the police.

Given the already-existing statutory and credibility barriers in the criminal justice system, there is no reason to add a conflicted prosecutor to the parade of obstacles to equitable treatment of police under the criminal law. The following Subparts will describe this conflict in detail. The first Subpart will lay out the reliance of prosecutors on the police for obtaining convictions. Next, this Part will look at how that reliance creates a structural conflict when the police become defendants in a criminal case. Finally, it will address the potential conflicts created by both the real and perceived inequalities inherent in local prosecutions of police–defendants, and the democratic legitimacy issues that may occur as a result.

A.     Inherent Conflicts When Local District Attorneys Must Prosecute Police Officers

This Subpart will address prosecutorial reliance on the police in cases against civilian defendants in terms of arrests, evidence collection, and testimony. Such reliance on the police leads to a conflict of interest when it is an officer who must be prosecuted.

1.     Prosecutorial Reliance on Law Enforcement in Civilian Defendant Cases

Prosecutors rely heavily on police cooperation for the success of their cases.80 Almost no criminal case exists without the police as the first contact point.81 Police officers investigate and arrest suspects, often without any input from the prosecutors who will eventually try the case.82

Take a typical, uncomplicated drug prosecution for example. The police will either see a drug transaction occur while patrolling or will participate in what is known as a “buy and bust.” In a “buy and bust,” an undercover officer, posing as a drug buyer, will approach a suspect, arrange a transaction, and, once the drug is purchased, call in a nearby team to make the arrest. In such cases, the patrolman or undercover officer will be the only witness to the identity of the defendant, the amount and type of drug purchased, the area where the drug was purchased, and a number of other factors that may impact the level of charges brought and the strength of the case.83 In some cases the police may decide whether a case is even worth pursuing.84

After an arrest, the police interview the suspect. During these interactions a number of legal issues can arise that may impact the case, such as whether the search and seizure of the suspect comply with the Fourth Amendment,85 whether the suspect is made aware of her rights, treated fairly at the police station, and gives an admissible confession in compliance with the Fifth Amendment,86 and whether the suspect is given an attorney if one is requested, as is required by the Sixth Amendment.87 These constitutionally significant interactions often occur without any participation from a prosecutor.88 In fact, in most cases, prosecutors do not lay eyes on a potential suspect or her case for many hours or even days after contact is made between the suspect and the police.89 Daniel Richman notes that a prosecutor “generally will not even know that a crime has been committed until [the police] inform[] [her].”90

Moreover, the police create and control the facts of most criminal cases. “The police have, at a most fundamental level, the ability to select facts, to reject facts, to not seek facts, to evaluate facts and to generate facts.”91 Officers are responsible for getting statements from often reluctant witnesses,92 for interviewing and supporting victims and families, and for gathering physical evidence that may appear unimportant but could have enormous impact later on in the process. “Facts, in this sense, are not objective entities which exist independently of the [police] but are created by them.”93

As the recent litigation in New York regarding stop-and-frisk has made clear, however, the police do not have the same level of incentive to ultimately convict defendants as prosecutors do.94 Their job is to investigate crimes and make arrests. The incentives that motivate police to perform these tasks are often not the same as those that ensure an arrest or evidence gathering is constitutional.95 Thus, respect for their prosecutorial coworkers plays a critical role in ensuring that an arrest turns into a conviction.

The need for a good working relationship between prosecutors and police does not stop once a prosecutor is assigned to the case. Taking the above example of a typical drug arrest, it is the police officers who must testify to the grand jury for any charge that requires an indictment.96 While grand juries rarely fail to indict if a prosecutor wants them to,97 one author noted that almost every prosecutor he interviewed “described at least one or two cases that were rejected by the grand jury because of the attitude or the incompetence of the primary police witness.”98 It is also the evidence gathered by the police and their statements that will determine the prosecutor’s strength in plea bargaining, which is the way 90–95% of cases are resolved.99

Police testimony is doubly important to this Article, because it illustrates both the absolute reliance that prosecutors have on the police to achieve convictions, and a major area where local prosecutors have a conflict in charging and prosecuting the police for crimes. While cases of police brutality are unlikely to go completely unnoticed, issues of “testilying”100 by the police, as well as evidence planting, tampering, or withholding, and illegal intimidation tactics will never emerge unless another officer or a prosecutor addresses them.101

For the moment, however, I will focus on the issue of prosecutorial reliance on the police to clear their cases. In order to ensure that police are available to testify, and testify legally and persuasively, prosecutors must work very closely with officers. Moreover, these prosecutors often have little control over when and how long an officer must spend in court on a particular case.

For instance, on many occasions, an officer may be scheduled to testify on a Monday at 9 AM. She will be at court, waiting to testify, and then told that, because of the judge’s commitments, or a problem with a juror, she will not be able to testify that day. She may be called back several times before she is actually able to testify. She will have to travel to court, an inconvenience, and then wait there instead of performing what she likely considers to be her real job—patrolling or investigating crimes. Then, when she finally testifies, she must do so without letting these inconveniences and time lapses affect her demeanor or her memory of the case.102

It is the prosecutor’s job to orchestrate this testimony, to ensure that the officer is in court, that she is made to feel that her time is being used appropriately, and to ask questions while the officer is on the stand. Indeed, former prosecutor Paul Butler has argued that a prosecutor’s main function in many trials is to ensure that the judge and jury believe the police officer’s testimony.103

A prosecutor’s examination of a testifying officer creates another potential source of ill will. Sometimes a prosecutor will have to ask quite confrontational questions, a tactic known as “pulling the sting.”104 For instance if an officer has an admissible disciplinary record, or if there was something problematic, though not unconstitutional, about the way a defendant was treated during her arrest, a prosecutor will likely address such issues on direct examination. If the prosecutor does not pull the sting, a defense attorney may raise these issues during cross-examination and imply that the state or the officer has something to hide. Even though police officers are repeat players in court, and aware of this tactic, it still takes finesse and a good working relationship to ensure that the officer answers the questions without getting angry, becoming defensive, or lying.

To foster such professional reliance, prosecutors must have a smooth working relationship with the police. This relationship naturally carries over outside of work. As Richman puts it, “one ought not underestimate the unifying influence of a shared commitment to ‘getting the bad guys,’ hardened by the adversarial process, nurtured by mutual respect and need, and on occasion lubricated by alcohol.”105 And the criminal justice system relies upon this shared commitment, as any functioning system with multiple actors does. If, as “[o]rganizational theory teaches[,] . . . effective coordination always depends, at least in part, on the development of informal norms and conventions through group interaction, socialization, and experimentation,” we should not look to hinder the “social relationships” between prosecutors and police that “provide a solid foundation for trust.”106

In short, the cooperation between prosecutors and law enforcement is perhaps the most important facet of any criminal case. Maintaining a good relationship with individual officers and the good will of a police department is essential to a prosecutor’s success in obtaining convictions, and thus to her professional life.107

2.     Prosecuting the Allies

As the last Subpart illustrated, a local prosecutor relies heavily on law enforcement’s support and good will to do her job. Yet the same skills and relationships that effectuate a functioning system when prosecuting civilian defendants make local prosecutors the least objective adversaries when it comes to prosecuting law enforcement. When prosecuting an officer, the prosecutor must switch from her reliance on the police as allies to the position of an adversary, questioning the credibility and judgment of a police officer, charging the officer with crimes that often carry stiff prison sentences, and pursuing the case against the officer as vigorously as she would against a civilian defendant.

In many cases, the conflict will be personal; the officer will be someone the prosecutor knows and may be friendly with.108 This is particularly true in small jurisdictions with only a few dozen officers and even fewer prosecutors. In large metropolitan areas, the relationship may be slightly more arm’s length, although it is still likely that the prosecutor has worked with the officer or those who know her. Yet the way the system currently works, we ask local prosecutors to use their discretion to decide whether to bring charges against an officer and then to prosecute the case. Even ignoring for the moment the set of obstacles put in the prosecutor’s way by law enforcement and its unions, it is absurd to assume that a prosecutor can simply switch roles from ally to adversary the moment an officer is accused of criminal wrongdoing.

In addition, while many instances of police brutality are brought to a prosecutor’s attention by a civilian complaint, other instances of police criminality, like perjury or evidence tampering, will only come to light during or after the prosecution of a case currently being pursued by the same office. Recall the example of the drug deal. Imagine that evidence comes to light that an officer perjured herself during a suppression hearing, hiding that she performed an illegal stop or seizure. If the prosecutor acknowledges the perjury, she not only jeopardizes the current case by ensuring that whatever drugs were seized are suppressed, but she is also reporting an officer for illegal conduct. Then, whichever prosecutor is tasked with prosecuting that case of perjury must call the first prosecutor as a witness in the perjury case. The situation is fraught to say the least, and it is not a stretch of the imagination to assume that these instances are rarely “caught,” let alone prosecuted.109

The fact that so few police crimes are reported makes it even more important that vigorous prosecutions of those crimes that do come to light are ensured. But as anecdotes and statistics show, this is simply not the case.110

3.     Prosecuting the Police May Affect an Assistant District Attorney’s Career

Although a prosecutor’s mandate is to “do justice” rather than be a zealous advocate, a widely known and logically salient connection exists between the number of convictions a prosecutor secures and her likelihood of promotion. Paul Butler recalls that “[l]ocking people up” was practically a prosecutor’s job description.111 He recounts that “Eric Holder . . . [at the time the U.S. Attorney for the District of Columbia] asked prospective prosecutors during interviews, ‘How would you feel about sending so many black men to jail?’ Anyone who had a big problem with that presumably was not hired.”112 While anecdotal, this shows the pressure prosecutors are under to obtain convictions.

The pressure to obtain convictions does not stop once a prosecutor is hired. Although not official policy, it is widely known that promotions within offices are often made on the basis of successful conviction rates. As Angela J. Davis writes, to be promoted, a prosecutor must “stay in favor with [her] boss,”113 who is usually elected based on “tough on crime” promises.114 Thus an assistant prosecutor who did not secure convictions “would not be promoted or otherwise advance in that office.”115 As discussed above, assistant district attorneys rely on the police for successful convictions, and therefore, must have a good working relationship with the police for professional advancement. A prosecutor who reports police crimes or advocates zealous prosecution of the police will necessarily run afoul of law enforcement’s good graces, which may impact conviction rates and therefore her career advancement.

Some offices may avoid this conflict by having a special prosecution unit for police crimes.116 Ostensibly, these prosecutors would be immune from pressure to avoid charging and prosecuting police because it is their job description. Still, even if they are insulated from police pressure, their elected bosses are unable to avoid it and may well feel pressure to instruct their employees to decline to bring charges in cases where the crimes are not high profile or in the public’s view.117

B.     Systemic Conflicts, Bias, Secrecy, and Accountability

Even if local prosecutors are able to overcome the personal and professional hurdles they face and switch roles from ally to adversary when dealing with police suspects, any decision they make to decline prosecution will be looked at through a haze of bias. For instance, some may see the decision by Robert McCulloch, the District Attorney in Ferguson, to bring evidence to a grand jury rather than to simply decline to charge Darren Wilson as a way around the bias he would be accused of if he simply used his discretion not to prosecute.118 Yet a whole other parade of biases, real or perceived, was on display.

As we can see from released documents, the Ferguson prosecutors put on hours of conflicting testimony and gave the grand jury reams of paper evidence to “help” it decide whether or not to indict Wilson.119 Yet as those who work in the criminal justice system know, this is not how a normal case is presented to the grand jury. The adage that the grand jury will “indict a ham sandwich” exists partially because of the way prosecutors present information to the jury.120 At a usual grand jury presentation, the prosecutor presents her theory of the case, examines a few witnesses who support that theory, and displays evidence that coheres.121 The presentation lasts anywhere from a couple of minutes to a few hours, and the grand jury votes on whether there is probable cause for an indictment.122

A civilian defendant is usually not permitted to be present at grand jury proceedings.123 If the defendant is allowed and chooses to testify to the grand jury, she has to waive rights to immunity, meaning that any statement she makes can be used against her at plea bargaining or trial. The defense attorney is permitted in the grand jury room if the defendant testifies. The prosecutor will likely question not only the defendant’s version of events but also her credibility, including all prior accusations of crimes, dismissed or not. The prosecutor’s job is to undermine the defendant. 124 That is how the system works for civilian defendants.

On the other hand, as many commentators have noted, the Wilson Grand Jury functioned more as a trial jury, with all the evidence produced, and with the prosecutors acting as much as defense attorneys as prosecutors.125 Moreover, McCulloch’s long, televised statement, describing the evidence in terms that can only be described as favorable to Wilson’s account, in an apparent effort to quell community unrest, had the opposite effect.126 He appeared more like a defense attorney explaining why his client should be absolved than a prosecutor lamenting the fact that no charges were filed against a criminal suspect.127

Whether or not we take as true McCulloch’s subjective claims that he was unbiased,128 the way the prosecutors used the grand jury system resulting in a nonindictment for Wilson raised many relevant and important questions about the equality and fairness of the criminal justice system. Yet in one important way, the Ferguson case was far better than most prosecutorial charging decisions—it was transparent.129

Many decisions not to prosecute officers for crimes against civilians are made by a prosecutor’s office in secret and never even reach a grand jury. As Paul Butler has written: “The head of a prosecution office is the most unregulated actor in the entire legal system. Basically, there are no rules. . . . The lead prosecutor . . . can make whatever decision he wants about whether to prosecute and no judge or politician can overturn it.”130 Because most declinations are made in secret, the public may never discover that a police suspect has not been charged. Even if the public is cognizant of a declination, it can only speculate whether such determinations are made based on an objective evidence review or an entangled relationship with the police.

While the public rarely sees the decision-making process that goes into charging or declining to charge an officer, police unions will always be aware that an investigation is occurring, because they are responsible for representing police suspects.131 Pressure from police unions to refrain from any legal action against officers is visible and intense. When the New York Medical Examiner’s office ruled Eric Garner’s choking death a homicide (the default ruling in deaths which do not occur from natural causes or self-inflicted injury), the Police Benevolent Association (“PBA”) responded immediately, calling the examiner’s ruling “political” and vowing to get its own medical examiners to look at the autopsy.132 This denunciation was particularly telling given that the New York Police Department (“NYPD”) relies upon the very same medical examiner’s office in every single homicide committed in New York. More recently, the shooting of Laquan McDonald illustrated that police unions are not above making negligent, if not intentionally false, statements to protect their members. Days after the teenager was shot by a Chicago police officer, a representative for the Chicago Fraternal Order of Police asserted that the officer only shot the 17-year-old after he lunged at the officer with a knife.133 Police dashcam video footage, however, shows that the teenager never lunged at the officer, and was, in fact, walking away when he was shot, multiple times.134

The pressure police unions put on prosecutors and anyone willing to testify against law enforcement is intensified when an officer is charged. For example, in 2011, after a three-year investigation into a ticket-fixing scheme, a number of NYPD officers were indicted on over 1600 counts.135 At their arraignment, “hundreds of off-duty” officers “taunted” prosecutors, physically stopped the media from filming the defendants, and even called then-police commissioner Raymond Kelly a “hypocrite.”136 Even more nefarious activity takes place behind the scenes. For instance, one former Baltimore police officer, who reported his sergeant and another detective for assaulting a handcuffed suspect, walked out of his house one morning to find a dead rat on his family’s car.137 This was compounded by threats, refusals from other officers to answer his calls while on duty, and a work environment so hostile that he eventually quit the force.138

Police unions assert their power not only in individual cases, but also at election time. More than 95% of local district attorneys are elected.139 As Angela Davis notes, the secrecy with which declinations are made shields elected prosecutors from the accountability we would expect of an elected official because “[t]he public cannot hold the prosecutors accountable for behavior of which they are unaware.”140 Because of this, and many other factors, most citizens have little information or interest in district attorney elections.141 The police, however, are well-informed and interested in the outcome of elections, and their support matters. “[P]olice unions enjoy remarkable political influence” over politicians of all stripes, but particularly over district attorney elections.142 For instance, “[w]hen a powerful police union charges that a politician is ‘soft on crime,’ that candidate’s chances for election or reelection can be dramatically reduced.”143 Given the lengths that the unions will go to intimidate systemic actors who diverge from their agenda, it is likely these unions would bring their numbers out to defeat a district attorney who vigorously prosecutes police–defendants.

Thus a district attorney, who can decline to prosecute police in secret, with no judicial review and little public scrutiny, must contend with the real possibility that she will not be reelected if she crosses the powerful police unions. It is hard to fathom that even the most well-intentioned politician is able to completely separate herself from such unfettered pressure from a highly mobilized and well-informed electorate in her jurisdiction.

IV.     Local District Attorneys Have an Unwaivable Conflict of Interest When Investigating and Prosecuting Police

In the previous two Parts, this Article has laid out a theory of conflict-of-interest law and described the many points at which a prosecutor encounters personal, professional, political, and systemic entanglement with the police. This Part will apply conflict-of-interest law to prosecuting the police, showing why such prosecutions must be removed from the hands of local district attorneys.

Two interrelated reasons emerge to explain why prosecutors have, as a rule, not been asked to recuse themselves from cases against local police, despite the seemingly clear conflict of interest. The first is the posture in which conflicts are usually raised, and the second is the virtually limitless power and responsibility given to district attorneys.

A prosecutorial conflict of interest in a police case, however, is unlikely to be raised by either the police or the prosecutor. First, as has already been suggested in this Article, the prosecutor does not technically have a client. She represents an amorphous public, referred to as “the people.”144 That body is unable to raise the issue of a prosecutorial conflict of interest in any particular case or type of cases because the people are not a party with standing to intervene in a criminal case.145 Similarly, because the prosecutor’s conflict of interest when police are the defendants will likely lead to leniency, the defendant will not raise the issue either. Further, the asymmetrical client relationship in criminal prosecutions of police makes it very unlikely that a prosecutor will be conflicted out of a case.

The only other person who might have the ability or motivation to determine that local prosecutors are conflicted out of prosecuting the police is the elected head of the district attorney’s office.146 Relatedly, the head district attorney has virtually limitless power to decide whom to charge, what to charge, and how to direct her employee–prosecutors to proceed in a given case. District attorneys have little incentive to cede power over police prosecutions.147 Nor do they have the objectivity needed to assess conflicts of interest in their own office—particularly when many of those conflicts are systemic.148 Moreover, given that local district attorneys have the jurisdiction to prosecute police, giving up that responsibility might look to the electing public like dodging politically charged responsibility, no matter how sound the decision may be.149

Conflict-of-interest law illustrates why local prosecutors have a structural and unwaivable conflict of interest when investigating and prosecuting the police. This Part will show both that due process concerns about the appearance of justice and personal conflicts of interest pollute every investigation and prosecution of police crime. This conflict convergence inherent in local prosecutions of police–defendants is so serious, and so unlikely to be raised in any individual case, that local district attorney offices should be automatically conflicted out of every case involving a police crime.150

A.     Local Prosecutions of Police–Defendants Do Not Satisfy the Appearance of Justice

As discussed in Part I, among the most important themes in conflict-of-interest law when applied to both judges and attorneys is that their behavior must satisfy the appearance of justice. This concept should apply with at least as much, if not more, force to prosecutors. A prosecutor’s ethical duty, unlike a defense attorney or a civil attorney, is to “seek justice.”151 Thus, she is not tasked with advocating zealously to the exclusion of other considerations but must take other concerns, such as equity and fairness, into account when determining whether to charge a given defendant, what charges to bring against a suspect, whether to plea bargain, and dozens of other decisions.152

Despite the Supreme Court dicta and other suggestions that the appearance of justice should not apply to prosecutors,153 it is hard to understand why a prosecutorial conflict of interest should not be scrutinized as closely as that of a judge. Because prosecutors have so much discretion when it comes to charging, and because so many criminal cases are resolved by plea bargaining, prosecutors are more critical than judges when it comes to the appearance of justice in the criminal system.154 A prosecutor’s duty not to appear biased in favor of a defendant is particularly important to the public’s trust in a system that may also compel them to give up their right to liberty.155

The killing of unarmed men in Ferguson and Staten Island has drawn much public scrutiny to other unindicted police killings. And the numbers have made the justice system seem wildly favorable to police officers. While police officer shootings and assaults are underreported,156 the raw numbers available suggest that police are prosecuted at an alarmingly low rate compared to civilians accused of violent crime. For instance, out of 179 police killings in New York, only three led to an indictment.157 In Utah, where death by police shooting is the second most common form of homicide, all but one of the 45 killings since 2010 went uncharged by prosecutors.158 It is almost impossible to know whether favoritism or fair evidence review went into the decisions not to prosecute the officers in these cases, but the publicity surrounding these numbers has largely focused on local prosecutors’ favoritism toward police.159 Moreover, the fact that many of these declinations were made in secret heightens the sense that something nefarious is going on behind a decision not to charge police for killing civilians.160

The perception problem is compounded by the basic and unavoidable role that race plays in police brutality in America. A near consensus in American political and legal thought acknowledges that the criminal justice system is disproportionately harsh toward African-Americans and Hispanics, both as victims and defendants. African-American men are 21 times as likely to be killed by the police as their white counterparts.161 African-Americans are incarcerated at six times the rate of whites and together with Hispanics make up 58% of our prison population despite being 25% of the population combined. Yet, because crime is so often intra-racial,162 and because both victims and defendants are so often part of a poor and overlooked minority population,163 the public is generally not focused on their systemic treatment.164 But when white police officers brutalize unarmed African-American men and are not criminally charged, the racial injustice of our criminal law crystalizes in the imagery of white officers escaping judgment for the killing of black victims.165

While minorities are overrepresented in prison and as victims of police shootings, they are woefully underrepresented in law enforcement, in the legal profession, and in prosecutors’ offices. In 2007, a Bureau of Justice Statistical Report showed roughly that whites make up between 75 and 88% of local law enforcement throughout the country, African-Americans between 6 and 12%, and Hispanics between 3 and 10%.166 In other words an overwhelmingly white police force is involved in the incarceration of an overwhelmingly minority population.167 The numbers at prosecutors’ offices are likely no better. Of elected prosecutors, 95% are white, and many states have no African-Americans elected to such positions.168 While it is less clear what the racial makeup is of all prosecutor offices,169 nonwhites make up only 10% of the legal profession, and it is fair to assume a smaller percentage of the minority bar is employed by a majority of prosecutors’ offices.170

Thus local district attorneys, responsible for putting thousands of minorities in prison, are the very same representatives who have often made the decision not to prosecute local police when they brutalize minority victims. Although over-incarceration of minorities is not causally linked to racially disparate police brutality, the fact remains that the optics are ugly. After the nonindictment in Ferguson, McCulloch was not only criticized for favoring police, but also for being an inveterate racist. He was called the “face” of America’s “race problem” for his decision to present reams of exculpatory evidence to the Wilson Grand Jury.171

While in any individual case, a prosecutor’s decision not to prosecute an officer for an on-duty shooting may well be justified, because most of these decisions are made behind closed doors, neither the public nor other system actors have any way of knowing what led to such declinations. This secrecy and the chasm between police killings and indictments lend credibility to those who believe that cronyism and racism are the cause of police nonindictments. And, given the numbers and secrecy, it is not surprising that the public takes a dim view of the justice system when it comes to faith that police who kill citizens will be held accountable. A poll conducted in August 2014, found that only 37% of people believed that the justice system could be trusted to deal with police–defendants.172

The public perception that white prosecutors are holding white police officers above the law certainly justifies questions about a prosecutor’s role as a minister of justice. And in the case of prosecutors and the police, the Supreme Court’s statement about judges—that “what matters is not the reality of bias or prejudice but its appearance”—is equally apt.173

B.     Local Prosecutors Have an Actual Conflict of Interest When Police Are Defendants

Elected district attorneys, who are subject to intense pressure from police unions, and their line attorneys who must rely on law enforcement for the success of every case they try, have a clear conflict of interest when the tables are turned and they must decide whether to bring charges and lead cases against police–defendants. Thus, police–defendant cases are the rare instance where an entire district attorney’s office must be conflicted out of leading the prosecution.174 As the Supreme Court noted in Young:

Between the private life of the citizen and the public glare of criminal accusation stands the prosecutor. That state official has the power to employ the full machinery of the state in scrutinizing any given individual. . . . For this reason, we must have assurance that those who would wield this power will be guided solely by their sense of public responsibility for the attainment of justice.175

It is difficult to see how a local prosecutor can uphold this high standard of impartiality in cases where their lives and careers may be impacted. Yet a police officer–defendant will never move to disqualify the prosecutor in her case unless she believes the prosecutor is biased against her.176 The most germane example of a prosecutor disqualifying herself and her office from a case is when a lawyer in the same office is accused of a crime. In such cases, offices routinely make their own decision to conflict the case out to another office.177 Still, even in these cases, it is not clear whether this policy is an acknowledgement that an actual conflict exists or it is a policy designed to preserve the appearance that the district attorney office is being fair.

In many ways, the conflict between an assistant district attorney and law enforcement is even more problematic than between two prosecutors in the same office. As described in Part II, prosecutors and police work as closely together as any two actors in the legal system. Most prosecutors try their own cases; they do not rely on other lawyers in the office for successful case resolution. On the other hand, there is almost never a criminal case where the police are not involved. It stands to reason, then, that if prosecutors routinely conflict themselves out of cases when someone from their office is involved, they should do so when a police officer in their jurisdiction is the defendant.

Justice Scalia’s comments about the types of cases where defense attorneys are conflicted out of cases because of possible personal bias or financial interest is also instructive. As he noted in Mickens v. Taylor, lawyers are disqualified or recuse themselves when they have a book deal, have taught a class to the IRS, have a romance with a prosecutor, or when there is some concern that they may antagonize the trial judge.178 Without arguing that prosecutors be held to anywhere close the same standard that defense attorneys must adhere to, the above examples show that far too little attention is paid to the way a conflict of interest with the police will impact an assistant district attorney.179

Moreover there is no way to insulate a prosecutor’s office from the conflict it faces when investigating and charging police. At the micro-level, any prosecutor’s reliance on the police will necessarily impact her decisions when confronted with investigating and prosecuting a member of law enforcement. Even if an office has a special unit designated to deal with police crimes, these prosecutors will likely have played or will later play another role in the office.180 And, just as is the case with police informants, these prosecutors will become virtual pariahs with local police unions. More importantly, their decisions are still subject to approval by the head of the office, who will know she must either decline charges against police or potentially face recrimination from these unions in future elections.181 These personal and professional entanglements at every level of a district attorney office compel the conclusion that there is the potential for bias in any prosecution of the police by a local district attorney’s office.

C.     A District Attorney Will Have Difficulty Determining Her Conflict in Police Cases

As has been shown in the last two Subparts, it is almost impossible to conjure a scenario where a local district attorney will not have an inherent conflict when asked to investigate and charge police suspects. To make matters worse, both political pressure and cognitive biases make it extremely difficult for the head of the office to recognize the conflict and recuse her entire office from police cases. First, the district attorney is often elected with the support of local law enforcement unions.182 Thus, the district attorney is already captured by the gratitude and loyalty she feels towards the officers that supported her election. Then, there is the immense reliance, if not obligation, that these officials owe to the police for the successful resolutions of the vast majority of their cases.183

Moreover, we cannot expect these elected officials to be cognizant of their conflicted interests when tasked with investigating and prosecuting police suspects. As this Article previously noted,184 anyone faced with a conflict will likely be impacted by “bounded ethicality,” envisioning herself as more “honest, trustworthy, ethical, and fair than others”; “better than others in possessing a series of desirable attributes”; and “more responsibl[e] . . . for contributions to an outcome than they actually deserve,” all the while unable to see that these biases exist and believing that she has “acted ethically, even in the face of evidence to the contrary.”185 While not putting it in these terms, courts acknowledge such biases when ruling that a judge who has refused to recuse herself must be disqualified.186

Such cognitive biases may explain the unprofessional response of St. Louis District Attorney Robert McCulloch to calls for his recusal from the charging decision in the Michael Brown killing.187 McCulloch was not only challenged for his too-close professional relationship with local law enforcement but for his very personal connection to the police—his own father had been a police officer, killed in the line of duty.188 This personal connection became the subject of many of the recusal calls when the decision about how to proceed against Darren Wilson arose. McCulloch summarily dismissed these legitimate concerns about his impartiality; he even told the state’s Governor to “man up,” after he suggested that McCulloch could recuse himself from the case.189 His office’s presentation to the grand jury has now been the subject of much speculation, including whether he allowed knowingly perjured exculpatory evidence to be presented.190 Whether or not the case was actually handled fairly, McCulloch’s inability to even acknowledge the legitimacy of the concerns over his impartiality presents perhaps the most systemically problematic issue to arise. It is also an illustration of the self-confident biases that confound a district attorney’s ability to judge his own impartiality in a given case. McCulloch’s inability to wrestle with the issue, let alone allow another prosecutor to take over, suggests that we cannot leave recusal decisions in police cases to the district attorneys themselves. Such action must come from another state actor.

When a police officer is accused of a crime, the local district attorney’s office is likely to be unable to view the case with the impartiality that conflict-of-interest law demands. This Part has shown that such prosecutions appear unjust and are rife with personal conflict that goes far above the conflicts that are routinely used to conflict judges and attorneys out of other cases. Because of the cognitive biases and political pressures that exist, we cannot trust elected district attorneys to recognize such conflicts and remove themselves, despite often clear evidence that they must do so. Local district attorneys are unable to determine their own conflicts and remove themselves out of prosecuting local law enforcement. Thus, there should be laws or rules that automatically remove such cases to other prosecutors or attorneys. The next part addresses several potential offices or groups that could fill in when local law enforcement are suspected of crimes.

V.     Potential Solutions

If local prosecutors have an inherent conflict that precludes them from prosecuting the police, to whom should the responsibility fall? This Part will propose several solutions, any of which would be more structurally legitimate than the status quo. Yet, as with any proposed shift in an entrenched system, there will be costs to any different actor taking on the role of prosecuting the police. These costs include not only the financial outlay of instituting a new mechanism for investigating and prosecuting police suspects, but also knowledge costs and, to a lesser extent, accountability costs. These costs increase the farther the system actor is from a local prosecutor. At the same time, however, with each step away from local prosecutors, the chance for an impartial review of a police suspect increases.

Here, this Article presents a sliding scale of most efficient but least impartial, to most impartial but least efficient potential solutions. These suggestions range from simply conflicting police cases out to a district attorney from a neighboring jurisdiction, to appointing the Attorney General or federal prosecutors in police cases. At the other end, this Article proposes instituting a civilian complaint review board vested with the power to review cases. Such a board could recommend charges to an independent prosecutor appointed from a list of qualified attorneys.

A.     State and Federal Prosecutor Alternatives to Local District Attorneys

The first and second levels of removal from local district attorneys’ offices are to appoint either a district attorney from a different county in the same state or the state Attorney General’s office to investigate and, when necessary, prosecute the police. These solutions have several efficiency and accountability advantages. In either case the attorney investigating police crimes will be a citizen of the state, will answer to communities within the state, and will likely know the criminal law of the state to a greater degree. Moreover, many jurisdictions already have mechanisms in place to substitute one of these players for a district attorney in the rare instance where her office is conflicted out of a case. On the other hand, these in-state elected officials may be subject to similar capture problems from unions, may also have to work with the police in the jurisdiction in which they must review accusations against them, and may be more subject to pressure from local district attorneys to do their will.

1.     Removal to a Different County’s District Attorney

The least costly solution would be to automatically remove a police case to a district attorney’s office from a different jurisdiction in the state. This is what happens in some jurisdictions when a district attorney or assistant district attorney is arrested. In those cases, it is policy for the office to recuse itself, so there is no issue as to how to ensure removal. In police cases, unless district attorneys are willing to routinely conflict themselves out, which is quite doubtful,191 the legislature or governor will have to write a statute or issue a rule about how to proceed when allegations of police crimes arise. There are several positive factors in favor of this solution. First, another district attorney in the state will be completely familiar with the state’s criminal laws. No extra training or self-study would be required for them to lead police prosecutions. Second, beyond a rule requiring the practice, no new legislation or procedure would have to be in place to make this solution a reality. These two efficiencies might make such a solution the most politically palatable.

On the other hand, there are several potential concerns related to appearance and impartiality that would hinder the success of this proposal. Depending on the size of the state counties or the distance between counties, these prosecutors may also have to work with the same police they would be prosecuting. These prosecutors may also have developed the same favoritism to law enforcement generally even if they do not rely on the same police force. Similarly, any local district attorney is part of the system that over-incarcerates minority defendants. Thus, when she decides not to prosecute an officer—particularly a white officer accused of a crime against a minority victim—the prosecutor may be perceived as part of the criminal justice system’s race problem.

Second, counterintuitively, these district attorneys would be less accountable to the community than their state prosecutor counterparts because the community in which they would prosecute the police would not be the same as the community that elects them. Finally, any state district attorney will be subject to pressure from state law enforcement unions, who will use the same pressure and obstruction tactics they do for any district attorney who prosecutes one of their members. Moreover, without a central state agency collecting and disseminating data, it might be difficult for the public to know whether or not these prosecutors were actually making fair and impartial decisions.

2.     Removing All Police Cases to the Attorney General’s Office

There are a number of reasons that removing police cases to state attorney general offices might be the best combination of efficiency and fairness. First, as Rachel Barkow writes, most states already have rules that allow an attorney general to step in and investigate and prosecute cases in certain substantive areas of criminal law and where local district attorneys have a conflict.192 Although most states have some provision allowing attorney generals to conduct local criminal cases, Barkow found that, even when the statutory grant to these offices is broad, the state agency rarely uses its power to take over local investigations and prosecutions.193 Especially in conflict cases, attorney general offices tend to take such cases only when a local district attorney decides to recuse herself and refer the case to their office.194 Given the wide public support for removing local prosecutors in police cases, however, attorneys general may be more willing to use their power.195 For instance, Governor Andrew Cuomo has signed an Executive Order removing all civilian deaths at the hands of police to the State Attorney General’s office.196 This came months after Eric Schneiderman, the New York Attorney General, called on the governor to appoint his office in all cases where police must be investigated.197 New York is one of the states with the narrowest grant of authority to its state prosecutor,198 but in other states, the attorney general could simply use her extant authority to effect the removal of such cases.199 Of course, this removal would have budgetary implications for the state agency, so funding from state legislatures would be very helpful, although perhaps not necessary.

Another benefit of this solution is that each state would have a centralized office that investigated and prosecuted police crimes. State troopers could also be employed instead of local law enforcement to conduct investigations. Furthermore, unlike prosecutors from other counties, federal prosecutors, or private citizens, state attorneys general are politically accountable to the same communities affected by police violence as the local district attorney.200 Finally, because state attorneys general do not prosecute drug or violent crimes, they are removed from the racial pathologies that plague local prosecutions and at least some of the appearance problems that flow from these pathologies.201

On the other hand, there are a number of costs to both the efficiency and impartiality of a statewide police prosecution unit. As state attorneys general tend to prosecute only a limited number and types of cases, the office may be less familiar with the state’s criminal law than a local counterpart. Furthermore, while not nearly as captured as a local district attorney, state attorney generals may also face pressure and intimidation from state police unions. After all, they too are reliant on such unions and their members come election time. That said, because this issue is so much less serious at the state level, it should be a low-priority concern for anyone wishing to reform a state’s police prosecution policy. This solution would require a culture shift to the extent that attorney generals are used to deferring to their local counterparts in most criminal matters, and the success of such a program might depend heavily on the public’s support for a measure and the legislature’s willingness to fund it.

3.     Appoint Federal Prosecutors to Investigate Law Enforcement Crimes

Another possible solution is to appoint federal prosecutors to investigate and prosecute allegations of criminal activity by law enforcement. Currently, federal prosecutors investigate and prosecute certain local police cases after state charges are declined, or if state charges do not result in a conviction. This system, while constitutional, is impractical and unfair. Under federal law, prosecutors must prove that the officer intended to deprive a suspect of her constitutional rights, a higher bar than that set by state laws.202 Thus, only in a case where there has been major mismanagement or some other problem will federal prosecutors secure convictions when their state counterparts have failed.203 Moreover, these successive prosecutions do nothing to resolve the inherent conflict problem this Article describes.

Instead, federal prosecutors should automatically handle cases where police officers are suspects. While it is unusual for federal prosecutors to bring charges in state court under state law, it is not impossible so long as state governments give them permission to do so.204 There are a number of immediately tangible benefits to such a solution. Assigning police cases to federal prosecutors would go a long way toward resolving any conflict inherent in state prosecutions of police—these attorneys are generally not beholden to the local police,205 and federal prosecutors are insulated from the political pressures that might plague state actors.206 Moreover, these prosecutors are intimately familiar with the criminal justice process. They are a closer analogue to a local district attorney than the attorney general, at least in terms of the substantive crime they prosecute.207 And, although they might have to learn state criminal law and evidentiary rules, they are familiar with the investigative, charging, plea bargaining, and trial process in general.

If such a proposal were used by a state, the federal office for that jurisdiction could have a unit devoted to prosecuting local law enforcement and could hire and train lawyers to be familiar with the law and procedure of the state court. Then, that federal investigatory and prosecutorial agency would have centralized knowledge over the complaints and prosecutions for local law enforcement. This could help the Civil Rights division of the Department of Justice figure out which police departments need further federal oversight.208

Such factors make the idea of federal prosecution of local law enforcement attractive, but in further considering such a proposal, several downsides emerge. First, the objectivity that comes with a federal prosecutor’s systemic distance from local law enforcement comes with a corresponding distance from the constituencies that local and state prosecutors serve. Unelected federal prosecutors are not answerable in any direct way to local communities. There is much to be said for the “community prosecution” movement supported and described by scholars such as Anthony Alfieri and Anthony Thompson.209 Community prosecution movements advocate for more rather than less localization, to know the communities they serve better.210 There is little incentive for federal prosecutors to integrate into local communities. That said, proper training, and top-down incentives to become familiar with local communities could ameliorate this problem so long as the U.S. Attorney’s office in the given jurisdiction were motivated to do so.211

Additionally, due to the perception problem that is at the root of so much public dissatisfaction with local police, (non)prosecutions might continue to plague federal prosecutors’ decisions about charging and prosecuting police. The federal system is roiled by many of the same race and class problems that plague local criminal justice systems.212 Moreover, federal prosecutors tend to be more elite and insulated, in terms of education and prior careers, than local district attorneys.213 Whether they decided to prosecute police accused of crime or not, these attorneys might be subject to charges of insularity, insensitivity to community rights and values, and race-based decision making.

Such drawbacks might lead a state government to be reluctant to cede police prosecutions to federal prosecutors. But, particularly in states where local and state prosecutors have previously been charged with cronyism and self-dealing, bringing in less politically influenced actors might be a good solution.214

B.     Should Outsiders Prosecute the Police?

A more radical solution would be to take prosecutors out of the investigation and prosecution of law enforcement altogether. While a number of practical obstacles exist to implementing what this Article calls the “outsider prosecution solution” of police, such a solution would serve a number of important democratic and legitimacy problems that infect any of the more system-oriented solutions proposed above.215

The outsider prosecution solution could look something like this: a civilian oversight board, constituted with appointed community members, civil rights attorneys, judges, and retired prosecutors and/‌or former high-ranking police officials, could be charged with the investigation of all allegations of police crimes. This board would look something like the already-instituted civilian complaint review boards in several major cities, who are tasked with overseeing allegations of police misconduct.216 However, the prosecution board would have teeth: subpoena power, access to an investigative arm, and the power to decide whether to charge officers or not.217 Unlike prosecutors’ secret determinations about whether to bring charges, this board’s decisions could be publicly accessible.

If the outsider board decided to bring charges against an officer, the court could appoint an attorney from a list of qualified, private lawyers as a special prosecutor. These appointments and their compensation could be modeled on the federal Criminal Justice Act panels—lists of qualified lawyers who are appointed to represent indigent defendants when the federal defender office is unable to.218 They could also be the domain of the many former prosecutors who become partners at law firms and wish to fulfill pro bono requirements.219

The benefits to this roughly sketched outsider prosecution team are significant. First, the conflict-of-interest and decision-making biases that plague all prosecutors to some extent would be nonexistent in such a committee.220 Outsider review boards and prosecutors would ensure that police prosecutions or declinations to prosecute appear just, and are taken on by those without a professional stake in the matter. Second, as Stephanos Bibas has written, the criminal justice system generally suffers from an insider complex, which, among many other problems, “disempowers victims, defendants, and the public . . . hides the workings of the system, leaving outsiders frustrated and mistrustful.”221 The problems of the criminal justice “machine” are at their height when prosecutors police their closest colleagues who are accused of crimes often against the most politically powerless citizens. Bibas has suggested “citizen review boards to oversee prosecutors’ offices and publicize data.”222 This Article’s suggestion is more radical in its power-grant to outsiders, but far more limited in its scope. Like community policing and prosecution, such boards could ensure that prosecutors and police were more sensitive to the needs of the communities they served.223 Unlike these suggestions, this solution would also reallocate actual power to these communities. Third, such boards could generate important data on which officers and which departments are most problematic.224

Of course, this solution will be grist for much criticism. A likely critique is that an outsider prosecution solution will have the opposite, but not less severe, bias problem suffered by local district attorneys: they may be overzealous. This concern is more problematic in theory than in reality. To the extent that critics may be concerned that a civilian review board would be overzealous, ensuring that judges, former prosecutors, and former police officers are part of such a board should ameliorate these concerns. Moreover, these boards and any appointed special prosecutor would be restricted by the amount of funding they had for cases and would have to make decisions about which cases to pursue based on budgetary limitations, much in the same way prosecutors already do.225 Also, to the extent that police feel above the law, the outsider prosecution solution might well be a better deterrent to illegal activity than traditional prosecutor solutions. Their oversight might even reduce the number of allegations against police. Finally, the substantive criminal law described above226 would continue to protect police from criminal charges resulting from potentially problematic but nonetheless legal activity.

Another concern is that such boards would suffer the same fate as current civilian oversight committees—they would be gutted by lack of funding, particularly if they made decisions that rankled those in power.227 The public has been clear, however, that it wants to see more accountability for police crimes. Instituting outsider prosecutions and seeing the fruits of their work could well be a boon to ambitious or reform-minded politicians. Moreover, the fear that this solution may not succeed is no reason to discount it, particularly given the strikingly problematic status quo and the lack of any more perfect solutions.

VI.     Conclusion

Police serve a critical role in the criminal justice system. But when they commit crimes in the course of their duties, they must be held to the same standards as any person accused of criminal wrongdoing. It is unfair to demand that local prosecutors, who work closely with the police and rely on them for professional and political advancement, investigate and prosecute law enforcement when they are accused of committing crimes. This notion is supported by the law and scholarship about conflicts of interest, a much-theorized area of law that has, heretofore, not been applied to local prosecutions of police. This Article has shown how conflict-of-interest law can underpin the removal of local prosecutors from investigating and prosecuting the police. While police reform and public dialogue may help reduce the number of crimes committed by law enforcement, a more impartial and fair system of prosecution must be employed to ensure that those officers who act above the law are not treated so by the criminal justice system.

  1. [1]. See Rachel E. Barkow, Institutional Design and the Policing of Prosecutors: Lessons from Administrative Law, 61 Stan. L. Rev. 869, 871 (2009) (“One need not be an expert in separation-of-powers theory to know that combining [many] powers in a single actor [(the prosecutor)] can lead to gross abuses.”); Daniel J. Freed, Federal Sentencing in the Wake of Guidelines: Unacceptable Limits on the Discretion of Sentencers, 101 Yale L.J. 1681, 1698 (1992) (noting “the extent to which enhanced prosecutorial power would fill the judicial vacuum created by the [sentencing] guidelines”); William J. Stuntz, The Pathological Politics of Criminal Law, 100 Mich. L. Rev. 505, 594 (2001) (noting one “dangerous prosecutorial power: the power to stack charges, to charge a large number of overlapping crimes for a single course of conduct”).

  2. [2]. Compare George Fisher, Plea Bargaining’s Triumph, 109 Yale L.J. 857, 859 (2000) (critiquing too much plea bargaining), with Robert E. Scott & William J. Stuntz, Plea Bargaining as Contract, 101 Yale L.J. 1909, 1910 (1992) (“Not only is the practice [of plea bargaining] widespread, but participants generally approve of it.”).

  3. [3]. See, e.g., Josh Bowers, Punishing the Innocent, 156 U. Pa. L. Rev. 1117, 1128 (2008) (“[P]rosecutors carry mindsets of ‘nondefeat’—aversions to dismissal that they keep in all cases, but that are most pronounced in cases against recidivists. In this sense, prosecutors consistently function as conviction maximizers even if they only rarely operate as sentence maximizers.” (emphasis omitted) (footnote omitted)); Steven Alan Reiss, Prosecutorial Intent in Constitutional Criminal Procedure, 135 U. Pa. L. Rev. 1365, 1366 (1987) (noting the importance of prosecutorial intent in constitutional claims).

  4. [4]. See, e.g., Stuntz, supra note 1, at 579–82; Ronald Wright & Marc Miller, The Screening/Bargaining Tradeoff, 55 Stan. L. Rev. 29, 33 (2002) (“A particularly noxious form of dishonesty is overcharging by prosecutors—the filing of charges with the expectation that defendants will trade excess charges for a guilty plea.”).

  5. [5]. See infra Part II.A.

  6. [6]. See Exec. Order No. 147, A Special Prosecutor to Investigate and Prosecute All Matters Relating to the Deaths of Civilians Caused by Law Enforcement Officers (N.Y. July 29, 2015); see also Erin Durkin, State Pols Propose Bill to Create Special Prosecutor for Police Killing Cases, N.Y. Daily News (Dec. 4, 2014, 2:55 PM),; Ben Kamisar, Majority in Survey Backs Special Prosecutors in Police Shootings, Hill: Briefing Room (Jan. 19, 2015, 2:47 PM), http://the (“76 percent of Democrats and 56 percent of independents are in favor of requiring a special prosecutor. A plurality of Republicans, 48 percent, support the plan, compared to 23 percent who are against it.”); Mike Lear, MO Rep, Sen Want Special Prosecutors in All Police Shootings, Missourinet (Oct. 13, 2014), But see Statement of Brooklyn District Attorney Ken Thompson on Call by New York State Attorney General to Act as Special Prosecutor in Cases of Fatal Shootings, Brook. District Atty’s Off., (last visited Feb. 24, 2016) (arguing that local DAs should keep such cases).

  7. [7]. See, e.g., Brian Beutler, The NYC Cop Who Strangled Eric Garner to Death Is Free Thanks to a Legal Flaw. Here’s How Voters Can Fix It, New Republic (Dec. 3, 2014),
    120478/eric-garner-police-officer-shouldve-been-investigated-independently (quoting Professor Ronald Wright about the ability to appoint special prosecutors in cases of police brutality); Paul Butler, Opinion, The System Must Counteract Prosecutors’ Natural Sympathies for Cops, N.Y. Times (Apr. 28, 2015, 12:26 PM),; Paul Cassell, Who Prosecutes the Police? Perceptions of Bias in Police Misconduct Investigations and a Possible Remedy, Wash. Post: Volokh Conspiracy (Dec. 5, 2014),; Erwin Chemerinsky, Opinion, Police Dodge Accountability for Deaths, Orange County Reg. (Dec. 7, 2014, 12:00 AM),; Kate Levine, The Ultimate Conflict, Slate (Sept. 11, 2014, 2:50 PM), http://www.; Howard Wasserman, Prosecuting Police—The Role of the Grand-Jury Pool, PrawfsBlawg (Dec. 4, 2014, 5:41 PM),

  8. [8]. Sally Kohn, First Mike Brown, Then Eric Garner: Prosecutors Can’t Be Trusted to Try Cops, Daily Beast (Dec. 3, 2014, 2:55 PM),

  9. [9]. See infra Part II.

  10. [10]. Several scholars have written about problems prosecuting the police but none has analyzed these problems through the lens of conflict-of-interest law. See, e.g., Erwin Chemerinsky, The Role of Prosecutors in Dealing with Police Abuse: The Lessons of Los Angeles, 8 Va. J. Soc. Pol’y & L. 305, 309–13 (2001) (focusing on problems prosecutors have in bringing cases against police and noting close relationships); Alexa P. Freeman, Unscheduled Departures: The Circumvention of Just Sentencing for Police Brutality, 47 Hastings L.J. 677, 719 (1996) (“Local prosecutors who ordinarily work closely with the police face an impossible conflict of interest between their desire to maintain working relationships and their duty to investigate and prosecute police brutality.”); Susan N. Herman, Double Jeopardy All over Again: Dual Sovereignty, Rodney King, and the ACLU, 41 UCLA L. Rev. 609, 630 (1994) (“[I]f multiple prosecutions were prohibited in all cases, unscrupulous state actors could immunize a favored defendant (perhaps a fellow state or city employee, like a sheriff or police officer) from further prosecution by instituting a sham state prosecution.”); John V. Jacobi, Prosecuting Police Misconduct, 2000 Wis. L. Rev. 789, 791 (critiquing lackluster effort of local police prosecutions and suggesting new powers for federal prosecutors based on the creation of the International Criminal Court); Laurie L. Levenson, The Future of State and Federal Civil Rights Prosecutions: The Lessons of the Rodney King Trial, 41 UCLA L. Rev. 509, 511 (1994) [hereinafter Levenson, Lessons of Rodney King] (“The Rodney King beating trials have prompted discussion on a number of issues, including the appropriate forum in which to pursue criminal prosecutions against police officers for violating citizens’ constitutional rights. There has been a renewed call for federal prosecutors to take the lead in these prosecutions.” (footnote omitted)); Laurie L. Levenson, High-Profile Prosecutors & High-Profile Conflicts, 39 Loy. L.A. L. Rev. 1237, 1255–56 (2006) (arguing that prosecutor offices should be disqualified in high profile cases where their impartiality may be questioned); Laurie L. Levenson, Police Corruption and New Models for Reform, 35 Suffolk U. L. Rev. 1, 22 (2001) (“[P]rosecutors often enjoy too close of a relationship with local police and are therefore reluctant to turn against those with whom they have worked.”).

  11. [11]. The term “structural defect” or “structural problem” is used to describe criminal appeals where an error in the process is so fundamental that even if the defendant might have been convicted anyway, a conviction must be reversed. See Sullivan v. Louisiana, 508 U.S. 275, 282 (1993) (Rehnquist, C.J., concurring) (“[S]tructural error... cannot be harmless regardless of how overwhelming the evidence of [a defendant’s] guilt.”).

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

  13. [13]. See infra Part II.B–C. Because American prosecutors perform a quasi-judicial function, conflict-of-interest rulings regarding both judges and prosecutors are relevant to analyzing when prosecutors may have problematic conflicts.

  14. [14]. See infra Part II.B–C.

  15. [15]. See infra Part II.B–C.

  16. [16]. But see Kate Levine, How We Prosecute the Police, 104 Geo. L.J. 746, 762–64 (2016) (discussing policy of certain prosecutors’ offices to publish declinations to prosecute in use of force cases).

  17. [17]. See supra note 7 and accompanying text.

  18. [18]. Standards For Criminal Justice §3-1.1(c) (Am. Bar Ass’n 1986); see also Bennett L. Gershman, The New Prosecutors, 53 U. Pitt. L. Rev. 393, 457 (1992) (“Whether American prosecutors can be... ‘ministers of justice’... or should ‘temper zeal with human kindness,’ as Justice Jackson recommended, are unanswerable questions in a criminal justice model that emphasizes crime control over protecting individual rights.”); Fred C. Zacharias, Structuring the Ethics of Prosecutorial Trial Practice: Can Prosecutors Do Justice?, 44 Vand. L. Rev. 45, 46 (1991) (“In civil litigation, the [professional responsibility] codes presume that good outcomes result when lawyers represent clients aggressively. In criminal cases, the codes do not rely as fully on competitive lawyering. They treat prosecutors as advocates, but also as ‘ministers’ having an ethical duty to ‘do justice.’”(footnote omitted)).

  19. [19]. See Marshall v. Jerrico, Inc., 446 U.S. 238, 250 (1980) (declining to rule on “what limits there may be [of] a financial or personal interest of one who performs a prosecutorial function”).

  20. [20]. See infra Parts II.B–D.

  21. [21]. Berger v. United States, 295 U.S. 78, 88 (1935).

  22. [22]. Marshall, 446 U.S. at 248. “[T]he courts may ‘require a stronger showing for a prosecutor than a judge in order to conclude that a conflict of interest exists.’” United States v. Tierney, 947 F.2d 854, 865 (8th Cir. 1991) (quoting Young v. United States ex rel. Vuitton et Fils S.A., 481 U.S. 787, 811 (1987) (plurality opinion)).

  23. [23]. Young, 481 U.S. at 811.

  24. [24]. The New York Court of Appeals recently upheld the disqualification of an entire county district attorney’s office when the defendant was a sitting judge in the jurisdiction. The Court held that the office’s refusal to plea bargain in the case created “a significant appearance of impropriety.” People v. Adams, 987 N.E.2d 272, 275 (N.Y. 2013); see also United States v. Heldt, 668 F.2d 1238, 1276 (D.C. Cir. 1981) (“Given the need to promote the appearance of justice, a trial court on timely motion should disqualify a prosecutor from participating in a criminal action when he has a personal conflicting interest in a civil case.”); United States v. Miller, 624 F.2d 1198, 1202–03 (3d Cir. 1980) (“Public confidence in the government’s prosecutors is essential, but it may be lost if former prosecutors assume private employment that appears to involve conflicts of interests.”); In re Apr. 1977 Grand Jury Subpoenas, 584 F.2d 1366, 1383 (6th Cir. 1978) (“Society also has an interest in both the reality and the appearance of impartiality by its prosecuting officials: ‘It is essential that the public have absolute confidence in the integrity and impartiality of our system of criminal justice. This requires that public officials not only in fact properly discharge their responsibilities but also that such officials avoid, as much as is possible, the appearance of impropriety.’” (quoting People v. Rhodes, 524 P.2d 363, 367 (Cal. 1974))); State ex rel. Burns v. Richards, 248 S.W.3d 603, 605 (Mo. 2008) (noting in a case where prosecutor had represented defendant in unrelated but similar matter that “the appearance of impropriety, without more, requires disqualification”); State v. Crepeault, 704 A.2d 778, 784 (Vt. 1997) (“Our concern is for the integrity of the legal process, which suffers as much from the appearance as the substance of impropriety.”).

  25. [25]. See, e.g., Tierney, 947 F.2d at 865; People v. Vasquez, 137 P.3d 199, 206 (Cal. 2006) (“[T]he Legislature made clear in Penal Code section 1424 that a conflict of interest, whether actual or apparent, required recusal under our statutory law only if it bore an actual likelihood of leading to unfair treatment.”); State v. Cherry, 83 P.3d 123, 128 (Idaho Ct. App. 2003) (“[A] criminal defendant asserting a prosecutor’s conflict of interest must demonstrate actual prejudice in order to obtain relief.”); Soares v. Herrick, 981 N.E.2d 260, 264 (N.Y. 2012) (“[C]ourts, as a general rule, should remove a public prosecutor only to protect a defendant from actual prejudice arising from a demonstrated conflict of interest or a substantial risk of an abuse of confidence.” (citation omitted)); State v. Camacho, 406 S.E.2d 868, 875 (N.C. 1991) (“[T]he mere appearance of impropriety is not of itself sufficient to warrant disqualification of an entire State’s Attorney’s office, based upon one member’s prior representation of a defendant presently under prosecution.”). The ABA’s standards for prosecutorial conflicts used to state that “[a] prosecutor should avoid the appearance or reality of a conflict of interest with respect to official duties”; however, more recent editions have eschewed appearance-based conflicts of interest. See Standards for Criminal Justice Prosecution Function and Defense Function §3-1.3 (Am. Bar Ass’n, 3d ed. 1993),
    justice_standards/prosecution_defense_function.authcheckdam.pdf (noting a change in the standard from a previous edition).

  26. [26]. See 28 U.S.C. §528 (2012) (“The Attorney General shall promulgate rules and regulations which require the disqualification of any officer or employee of the Department of Justice, including a United States attorney or a member of such attorney’s staff, from participation in a particular investigation or prosecution if such participation may result in a personal, financial, or political conflict of interest, or the appearance thereof. Such rules and regulations may provide that a willful violation of any provision thereof shall result in removal from office.”); Dep’t of Justice, U.S. Attorneys’ Manual §§3-2.170, 3-2.20 (1997) (setting out the standards and procedures for U.S. Attorneys and AUSAs to recuse themselves or, as necessary, their entire office); see also Morrison v. Olson, 487 U.S. 654, 657 (1988) (upholding law vesting appointment of independent counsel with the judiciary for conflicts of interest when executive branch must investigate one of its own employees).

  27. 27. See, e.g., Cal. Penal Code §1424 (West 2011); Colo. Rev. Stat. §20-1-107 (2015); 71 Pa. Stat. and Cons. Stat. Ann. §732-205 (West 2012); N.Y. County Law §701 (McKinney 2004).

  28. 28. Standards for Criminal Justice Prosecution Function and Defense Function, supra note 25, §3-1.3 (discussing conflicts of interest without mention of law enforcement defendants).

  29. 29. See Carrie Leonetti, When the Emperor Has No Clothes III: Personnel Policies and Conflicts of Interest in Prosecutors’ Offices, 22 Cornell J.L. & Pub. Pol’y 53, 89 (2012). Most cases regarding prosecutorial disqualification have to do with the prosecutor’s prior representation of a defendant or animosity toward a defendant. See, e.g., State v. Hursey, 861 P.2d 615, 618 (Ariz. 1993) (holding that a prosecutor who had represented the defendant in two earlier criminal cases should have disqualified himself from prosecuting the defendant in another criminal case); Sears v. State, 457 N.E.2d 192, 195 (Ind. 1983) (same); State ex rel. Keenan v. Hatcher, 557 S.E.2d 361, 370 (W. Va. 2001) (similar); see also, e.g., State v. Snyder, 237 So.2d 392, 395 (La. 1970) (holding that the district attorney should be disqualified from prosecuting the defendant based on personal animosity having campaigned against the defendant during a mayoral election). Prosecutors’ offices also voluntarily recuse themselves when a member of their office becomes a defendant. See, e.g., People v. Schrager, 74 Misc. 2d 833, 834 (N.Y. Sup. Ct. 1973) (“Defendant is an Assistant District Attorney of Queens County.... The motion is grounded upon the District Attorney’s declaration of disqualification stemming from the professional and personal relationships which this defendant shared with the members of the District Attorney’s staff.”).

  30. 30. See Susan W. Brenner & James Geoffrey Durham, Towards Resolving Prosecutor Conflicts of Interest, 6 Geo. J. Legal Ethics 415, 471–72 (1993) (discussing prosecutor conflicts of interest in a number of instances but never mentioning the issue of prosecuting local law enforcement); Leonetti, supra note 29, at 89 (discussing whether “disqualification is warranted on the basis that internal personnel policies [in prosecutor offices] create actual conflicts of interest”); Laurie L. Levenson, Conflicts over Conflicts: Challenges in Redrafting the ABA Standards for Criminal Justice on Conflicts of Interest, 38 Hastings Const. L.Q. 879, 881–901 (2011) (commenting on redrafting of conflict-of-interest laws for prosecutors but not mentioning prosecuting the police); Michael Edmund O’Neill, Private Vengeance and the Public Good, 12 U. Pa. J. Const. L. 659, 698 (2010) (“[A] conflict of interest is presumed to exist when the prosecuting attorney is compensated by the victim.”).

  31. [31]. Levenson, supra note 30, at 885 (“Although most prosecutors appreciate on some intellectual level that they represent the ‘People’ or ‘Government’ or the community-at-large,on a day-to-day basis, they answer only to themselves or to a supervisor.” (footnote omitted)).

  32. [32]. See infra Part III. A notable divergence from the idea that police defendants will not challenge a prosecutor’s conflict of interest is the case of the six officers charged in the killing of Freddie Gray in Baltimore. These officers alleged in a motion to dismiss that the States’ Attorney, Marilyn Mosby, had several conflicts of interest that led her to be overzealous in charging the officers. Their motions were denied and their cases are in various stages of resolution. See Justin Fenton, Officers in Freddie Gray Case Move to Dismiss Charges, Balt. Sun (May 8, 2015, 10:19 PM), (noting that conflicts of interest were among the allegations in the motion to dismiss).

  33. [33]. Leslie W. Abramson, Appearance of Impropriety: Deciding When a Judge’s Impartiality “Might Reasonably Be Questioned, 14 Geo. J. Legal Ethics 55, 56 (2000) (looking at types of appearance of justice issues raised); Amalia D. Kessler, Our Inquisitorial Tradition: Equity Procedure, Due Process, and the Search for an Alternative to the Adversarial, 90 Cornell L. Rev. 1181, 1255 (2005) (discussing adversarial versus inquisitorial system and noting that “the mere appearance that parties can control the master is corrosive to a legal system committed to values of rule of law and equal justice”).

  34. [34]. Charles Gardner Geyh, Can the Rule of Law Survive Judicial Politics?, 97 Cornell L. Rev. 191, 250 (2012).

  35. [35]. Sherrilyn A. Ifill, Racial Diversity on the Bench: Beyond Role Models and Public Confidence, 57 Wash. & Lee L. Rev. 405, 405 (2000) (noting the lack of racial diversity on U.S. courts); John Leubsdorf, Theories of Judging and Judge Disqualification, 62 N.Y.U. L. Rev. 237, 252 (1987) (taking critical and realist critiques of impartiality “to its limits” (i.e., “assume that the judge’s personal values determine the result in every case”), but arguing that disqualification law is just as important even if this assumption were true); Martha Minow, Foreword, Justice Engendered, 101 Harv. L. Rev. 10, 45–46 (1987) (“This aspiration to impartiality, however, is just that—an aspiration rather than a description—because it may suppress the inevitability of the existence of a perspective and thus make it harder for the observer, or anyone else, to challenge the absence of objectivity.”); Judith Resnik, On the Bias: Feminist Reconsiderations of the Aspirations for Our Judges, 61 S. Cal. L. Rev. 1877, 1944 (1988) (“Feminism rejects the choice between being a blank slate and imposing oneself on another, between having no interest and being corrupted by self-interest.”); Adam M. Samaha, Regulation for the Sake of Appearance, 125 Harv. L. Rev. 1563, 1638 (2012) (“Often we think that reality is insulated from any influence that can be linked to appearance, but sometimes an appearance becomes the basis for conduct that fosters a corresponding reality over time, and sometimes the concepts collapse in the first place.”).

  36. [36]. Richard Delgado et al., Fairness and Formality: Minimizing the Risk of Prejudice in Alternative Dispute Resolution, 1985 Wis. L. Rev. 1359, 1368 (“Both internal and external constraints are designed to keep a judge from exhibiting bias or prejudice. Internal constraints stem from a judge’s professional position.”); Geyh, supra note 34, at 250 (“[I]n the 1970s, federal and state laws were revised to require disqualification whenever a judge was biased or his impartiality might reasonably be questioned.”); Martin H. Redish & Lawrence C. Marshall, Adjudicatory Independence and the Values of Procedural Due Process, 95 Yale L.J. 455, 504 (1986) (“We have been unable to envision even one situation in which the values of due process can be achieved without the participation of an independent adjudicator. Moreover, in defining the term ‘independence,’ even the slightest hint of bias or undue influence must, as a general matter, disqualify a particular decisionmaker. Only when it is all but impossible to rectify bias should a potential lack of independence be tolerated.”).

  37. [37]. Caperton v. A.T. Massey Coal Co., 556 U.S. 868, 886 (2009).

  38. [38]. Id. (quoting In re Murchison, 349 U.S. 133, 136 (1955)). The Court also noted that “[a]lmost every State—West Virginia included—has adopted the American Bar Association’s objective standard: ‘A judge shall avoid impropriety and the appearance of impropriety.’” Id. at 888 (quoting Model Code of Judicial Conduct Canon 2 (Am. Bar Ass’n 2004)); see also Liljeberg v. Health Servs. Acquisition Corp., 486 U.S. 847, 865 (1988) (“The very purpose of [the federal judicial recusal statute] is to promote confidence in the judiciary by avoiding even the appearance of impropriety whenever possible.”).

  39. [39]. 28 U.S.C. §455(a) (2012); see also United States v. Amico, 486 F.3d 764, 767 (2d Cir. 2007) (disqualifying a judge who was accused of having a financial motive in a dispute, the Second Circuit went to pains to explain that “[t]his appeal deals exclusively with the appearance of partiality,” and that “nothing” the Second Circuit said in that case “should be understood to conclude—or to imply—that the district judge engaged in misconduct”).

  40. [40]. Peter David Blanck, The Appearance of Justice Revisited, 86 J. Crim. L. & Criminology 887, 901 (1996) (“Many states provide... grounds for disqualifying a judge when prejudice or bias is alleged or could reasonably be inferred. Such provisions seek to preserve the values embodied in the appearance of justice.”).

  41. [41]. Model Code of Judicial Conduct §1.2 (Am. Bar Ass’n 2011) (“A judge shall act at all times in a manner that promotes public confidence in the... integrity, and impartiality of the judiciary....”); id. §1.2 cmt. 3 (“Conduct that compromises or appears to compromise the independence, integrity, and impartiality of a judge undermines public confidence in the judiciary.”); Samaha, supra note 35, at 1566 (“An especially familiar example [of the need for the appearance of justice] arises in codes of judicial conduct. They obligate judges to recuse themselves when their impartiality can be reasonably questioned, not only when it is rightly questioned.”).

  42. [42]. See, e.g., Bradshaw v. McCotter, 785 F.2d 1327, 1329 (5th Cir. 1986) (suggesting that the judge should have disqualified himself because the public could view the judge’s acts as lacking impartiality); see also Blanck, supra note 40, at 891 (“The appearance of bias alone has served as grounds for reversal or judicial recusal, even when the judge is shown to be completely impartial. Courts have found due process violations sufficient to reverse criminal convictions when a trial judge’s behavior created merely the appearance of partiality. Litigants have the right to argue their case fairly before the decision-maker, and thereby, as Justice Frankfurter stated, ‘generat[e] the feeling, so important to a popular government, that justice has been done.’” (footnote omitted)).

  43. [43]. Alvarez v. Tracy, 773 F.3d 1011, 1024 (9th Cir. 2014) (Kozinski, J., dissenting).

  44. [44]. Id.

  45. [45]. Redish & Marshall, supra note 36, at 484.

  46. [46]. Blanck, supra note 40, at 892 (“[J]uries accord great weight and deference to even the most subtle behaviors of the judge. Appellate courts recognize that the impermissible appearance of judicial bias or unfairness at trial often manifests itself through judges’ subtle nonverbal behavior.” (footnote omitted)).

  47. [47]. Id. at 889–90 (“As recently as 1980, the Supreme Court expanded its conception of the appearance of justice to include not only the possibility of judicial influence, but also the general public’s right to have meaningful access to the workings of the judicial system.”); Michael R. Dimino, Pay No Attention to that Man Behind the Robe: Judicial Elections, the First Amendment, and Judges as Politicians, 21 Yale L. & Pol’y Rev. 301, 332 (2003) (discussing the tension between free speech and the announcement of judge’s political views affecting rights of future litigants); Deborah Goldberg et al., The Best Defense: Why Elected Courts Should Lead Recusal Reform, 46 Washburn L.J. 503, 504 (2007) (“Elected courts must demonstrate their accountability for the decisions they make by more aggressively distancing themselves from situations in which their fairness and impartiality might reasonably be questioned.”).

  48. [48]. See Keith Swisher, The Practice and Theory of Lawyer Disqualification, 27 Geo. J. Legal Ethics 71, 145–51 (2014). Swisher notes that in 16 states “an appearance of impropriety can be sufficient, by itself, to justify disqualification of a lawyer or law firm,” while in 20 others it is a factor to be weighed in the decision. Id. at 145–47. Swisher further posits that the use of an “appearance of impropriety standard” applied to attorneys has the potential to protect a number of principles: “the image of justice, the image of the legal profession (at least to the extent the two images intersect), and the reasonable expectations of clients”). Id. at 154. But see Chemerinsky, supra note 10, at 305 (“The law of professional responsibility is absolutely clear that a prosecutor’s ethical duty is to make sure that justice is done.”); infra Part IV.A (arguing that a prosecutor often functions like a judge in our modern criminal justice system and thus their appearance of impartiality, at least as to whom they are prosecuting, is essential).

  49. [49]. Wheat v. United States, 486 U.S. 153, 164 (1988).

  50. [50]. Id. at 160. Scholars have criticized the decision in Wheat for putting the interests of judicial administration ahead of a defendant’s right to counsel of their choice. See, e.g., Bruce A. Green, “Through a Glass, Darkly”: How the Court Sees Motions to Disqualify Criminal Defense Lawyers, 89 Colum. L. Rev. 1201, 1231 (1989) (“By upholding a trial judge’s discretion to disqualify an attorney when there is ‘a showing of a serious potential for conflict,’ the Court implicitly authorized trial judges to undertake an inquiry that potentially imperils the defendant’s ultimate interest in receiving the effective assistance of counsel.” (quoting Wheat, 468 U.S. at 164)); Patrice McGuire Sabach, Note, Rethinking Unwaivable Conflicts of Interest After United States v. Schwarz and Mickens v. Taylor, 59 N.Y.U. Ann. Surv. Am. L. 89, 99 (2003) (“Wheat received wide criticism. The rejection of the defendant’s choice of counsel after the defendant proffered a waiver of such conflict was inconsistent with other Supreme Court decisions that rejected any paternalistic rule protecting a defendant from his intelligent and voluntary decisions about his own criminal case.”).

  51. [51]. Green, supra note 50, at 1208–09 (“Rejecting the defendant’s arguments premised on the sixth amendment right to counsel, the Court determined that a trial judge has discretion to disqualify a defense attorney who has either an actual conflict of interest or a ‘serious potential for conflict.’” (quoting Wheat, 468 U.S. at 164)).

  52. [52]. See, e.g., Marshall v. Jerrico, Inc., 446 U.S. 238, 248 (1980).

  53. [53]. See infra Part IV.A.

  54. [54]. Bracy v. Gramley, 520 U.S. 899, 905 (1997).

  55. [55]. Richard E. Flamm, The History of Judicial Disqualification in America, 52 Judges’ J., Summer 2013, at 12, 13; Redish & Marshall, supra note 36, at 500–01 (“The Court has been extremely reluctant to disqualify a judge when no direct financial interest is involved, finding a due process violation only in cases where the judge and one of the litigants or attorneys are embroiled in a heated personal dispute.”). It was so for lawyers too. See Mark Andrew Grannis, Note, Safeguarding the Litigant’s Constitutional Right to a Fair and Impartial Forum: A Due Process Approach to Improprieties Arising from Judicial Campaign Contributions from Lawyers, 86 Mich. L. Rev. 382, 387 (1987) (suggesting that the same ought to apply for attorneys).

  56. [56]. See, e.g., Caperton v. A.T. Massey Coal Co., 556 U.S. 868, 884 (2009) (holding that disqualification was required when a judge’s financial interest posed “a serious risk of actual bias”). A number of articles have critiqued the Court for ruling too narrowly in Caperton. See, e.g., Penny J. White, Comment, Relinquished Responsibilities, 123 Harv. L. Rev. 120, 124 (2009) (“[B]y repeatedly focusing on the egregious facts of the case, the majority overlooked the broader implications that financial and political influence have for all judicial elections.”); see also David E. Pozen, The Irony of Judicial Elections, 108 Colum. L. Rev. 265, 303 (2008) (“There are good theoretical reasons to think that judicial disqualification is both underused and underenforced... in almost every state,and there is growing empirical evidence to suggest that campaign contributions influence judges’ decisions.” (footnote omitted)).

  57. [57]. There is also a robust scholarly discussion about defense attorney conflicts. See, e.g., Albert W. Alschuler, The Defense Attorney’s Role in Plea Bargaining, 84 Yale L.J. 1179, 1214 (1975) (“Often, however, courts have seemed blind to the basic conflicts of interest that arise when a lawyer represents two or more defendants in a single case.”); Stephanos Bibas, Plea Bargaining Outside the Shadow of Trial, 117 Harv. L. Rev. 2464, 2478 (2004) (noting that in the context of plea bargaining, financial conflicts of interest may lead defense attorneys to accept less favorable pleas for their clients); Jay Sterling Silver, Truth, Justice, and the American Way: The Case Against the Client Perjury Rules, 47 Vand. L. Rev. 339, 377 (1994) (“Numerous incentives exist for a criminal defense attorney to curry favor with a trial judge before whom she regularly appears, thereby representing, in effect, an additional conflict of interest for counsel with respect to her duty to effectively assist the accused.”).

  58. [58]. Aetna Life Ins. Co. v. Lavoie, 475 U.S. 813, 820 (1986).

  59. [59]. Id. at 823–24.

  60. [60]. Id. at 830 (Brennan, J., concurring).

  61. [61]. See, e.g., 28 U.S.C. §455(b)(4) (2012) (requiring that a judge recuse himself if he “individually or as a fiduciary,” or his spouse or minor child has a “financial interest in the subject matter in controversy or in a party to the proceeding”); Cal. Civ. Proc. Code §170.1 (West 2011); Ga. Code Ann. §15-1-8 (2015); Haw. Rev. Stat. §601-7 (2012); N.J. Stat. Ann. §2A:15-49 (West 2000); N.Y. Jud. Law. §14 (McKinney 2002); Tex. R. Civ. P. 18b.

  62. [62]. Redish & Marshall, supra note 36, at 500–01; see also Debra Lyn Bassett, Judicial Disqualification in the Federal Appellate Courts, 87 Iowa L. Rev. 1213, 1242 (2002) (“Requiring recusal for a financial interest however small’ while simultaneously denying a more comprehensive approach for bias or prejudice, places an undue emphasis on a judge’s potential financial interest in a pending case.”); Leubsdorf, supra note 35, at 243–44 (“[W]hen a party claims that the judge’s known passions and opinions will prevent her from deciding according to law—and, in our era, such a claim raises more troubling issues, and risks deeper insult to the sense of justice, than a suit against the judge’s brother....”); Judith Resnik, Managerial Judges, 96 Harv. L. Rev. 374, 428 (1982) (“[C]urrent practices [incorrectly] assume that trial judges can compartmentalize their minds, disregard inappropriate evidence, and reconsider past decisions in light of new information.”).

  63. [63]. 28 U.S.C. §455(b)(1), (4)–(5). The federal statute requires a judge to step aside when he has a “personal bias or prejudice concerning a party,” “knowledge of disputed evidentiary facts,” “or any other interest that could be substantially affected by the outcome of the proceeding,” or “[h]e or his spouse, or a person within the third degree of relationship to either of them, or the spouse of such a person is... likely to be a material witness.” Id.

  64. [64]. Mickens v. Taylor, 535 U.S. 162, 174–75 (2002) (citations omitted).

  65. [65]. Wheat v. United States, 486 U.S. 153, 162 (1988).

  66. [66]. Caperton v. A.T. Massey Coal Co., 556 U.S. 868, 883 (2009).

  67. [67]. See, e.g., Debra Lyn Bassett & Rex R. Perschbacher, The Elusive Goal of Impartiality, 97 Iowa L. Rev. 181, 205 (2011) (“A major roadblock in seeking a more effective recusal process is the human tendency to see oneself as unbiased or able to disregard any possible bias or other improper influence.”); Resnik, supra note 35, at 1888 (“Under what theory of disengagement, disinterest, or lack of involvement might one believe that a judge is the appropriate person to assess his or her own possibly impermissible bias? How could Congress require disqualification whenever a judge has ‘personal knowledge of disputed evidentiary facts,’ yet permit judges to decide both the facts and the law of their own relationship to a case?”); White, supra note 56, at 126 (“A judge’s promise of fairness and neutrality, even after a probing, personal inquiry, is insufficient to satisfy the due process standard.... The inquiry, which includes an appraisal of ‘psychological tendencies and human weakness,’ as well as unconscious judgments, is by its very nature imprecise.” (quoting Caperton, 556 U.S. at 883)).

  68. [68]. Tigran W. Eldred, The Psychology of Conflicts of Interest in Criminal Cases, 58 U. Kan. L. Rev. 43, 66 (2009).

  69. [69]. Id.

  70. [70]. Id. (quoting Dolly Chugh et al., Bounded Ethicality as a Psychological Barrier to Recognizing Conflicts of Interest, in Conflicts of Interest 74, 81 (Don A. Moore et al. eds., 2005)).

  71. [71]. Id. at 67 (quoting Chugh et al., supra note 70, at 84).

  72. [72]. Id. (quoting Chugh et al., supra note 70, at 84).

  73. [73]. Id. (quoting Chugh et al., supra note 70, at 80). This was clearly the case in Caperton where the West Virginia Supreme Court judge wrote several opinions explaining why he was able to remain impartial, despite the substantial contribution made by a party to litigation to his election campaign. Caperton v. A.T. Massey Coal Co., 556 U.S. 868, 886 (2009).

  74. [74]. Eldred, supra note 68, at 69 (“[W]hile the decision-maker will believe that the decision comes from rational deliberation where all competing concerns are considered and weighed, in actuality the automatic bias toward self-interest will often create an error in judgment that favors self-interest, ‘automatically and without conscious awareness.’ In other words, the decision-maker will rationalize behavior as consistent with ethical norms, even when in actuality the decision preferences self-interest.” (footnote omitted)); see also Keith Swisher, Prosecutorial Conflicts of Interest in Post-Conviction Practice, 41 Hofstra L. Rev. 181, 188 (2012) (“The prosecutor has an interest, however subconscious or short-sighted, in not attacking her previous work product.”).

  75. [75]. Laura E. Little, Loyalty, Gratitude, and the Federal Judiciary, 44 Am. U. L. Rev. 699, 754 (1995) (“[G]ratitude and loyalty can have a powerful influence for a federal judge undertaking to decide a case. The problem is complex because loyalty and gratitude pose a greater potential problem for some judges than for others. This complexity emerges to a great degree from the process of nomination and confirmation, which often generates, or at least reinforces, a judge’s sense of loyalty and gratitude to her benefactors.”).

  76. [76]. See Kate Levine, Police Suspects, 116 Colum. L. Rev. (forthcoming 2016) (describing special interrogation protections available only to law enforcement officers).

  77. [77]. Many states have self-defense laws that apply specially to police. See, e.g., Mo. Ann. Stat. §563.046 (West 2012 & Supp. 2015) (effective Jan. 1, 2017). Moreover, it is even harder to meet the burden of proof under federal law. See 18 U.S.C. §242 (2012) (stating that the government must prove beyond a reasonable doubt that the police officer acted with the specific intent to deprive the victim of a constitutional right).

  78. [78]. In Tennessee v. Garner, the Court held that, while

    [a] police officer may not seize an unarmed, nondangerous suspect by shooting him dead[,] . . . [w]here the officer has probable cause to believe that the suspect poses a threat of serious physical harm, either to the officer or to others, it is not constitutionally unreasonable to prevent escape by using deadly force.

    Tennessee v. Garner, 471 U.S. 1, 11 (1985). Later, in Graham v. Connor, the Court ruled that excessive force cases must be analyzed under the Fourth Amendment’s objective reasonableness standard, “judged from the perspective of a reasonable officer on the scene, rather than with the 20/20 vision of hindsight.” Graham v. Connor, 490 U.S. 386, 396 (1989).

  79. [79]. As William Bermeister, former head of New York’s anti-corruption prosecution unit, said, police prosecutions have a double credibility problem where “jurors give officers the benefit of a doubt,” while at the same time “if you don’t have an ‘innocent’ victim, jurors don’t care.” See Asit S. Panwala, The Failure of Local and Federal Prosecutors to Curb Police Brutality, 30 Fordham Urb. L.J. 639, 644 (2003); see also Anthony G. Amsterdam, The Supreme Court and the Rights of Suspects in Criminal Cases, 45 N.Y.U. L. Rev. 785, 792 (1970) (noting that fact finders generally find police testimony credible).

  80. [80]. Daniel Richman, Prosecutors and Their Agents, Agents and Their Prosecutors, 103 Colum. L. Rev. 749, 758 (2003). Richman terms the relationship between federal prosecutors and investigative bodies a “bilateral monopoly” where “[p]rosecutors are the exclusive gatekeepers over [the] court, but they need agents to gather evidence. Agencies control investigative resources, but they are not free to retain separate counsel.” Id. This description applies equally to state and local prosecutors and law enforcement, as does Richman’s point that prosecutors “labor under an informational disadvantage even in those systems where they formally have hierarchical power over police forces.” Id. at 813.

  81. [81]. In some states, a citizen can go to court and file a criminal complaint herself. See, e.g., Ohio Rev. Code Ann. §2935.09–.10 (West 1997).

  82. [82]. Stephanos Bibas, The Machinery of Criminal Justice 32 (2012) (“Police decide whom, where, and what to investigate; whether and whom to arrest or issue citations; and whether and which charges to file. Sometimes they even decide whether to refer a case to federal or state prosecutors.”).

  83. [83]. See id. at 41 (“[P]rosecutors... can often choose from a variety of possible felonies and misdemeanors.”).

  84. [84]. See, e.g., Issa Kohler-Hausmann, Managerial Justice and Mass Misdemeanors, 66 Stan. L. Rev. 611, 630 (2014) (“We can never directly interpret arrest rates as an index of underlying criminal behavior because reporting and police practices mediate criminal events and arrests. This is especially true of misdemeanors. The police can find as many instances of marijuana or drug possession, petit larceny, unlicensed vending, misdemeanor physical altercations, public alcohol consumption, turnstile jumping, prostitution, and disorderly conduct as they devote the time and resources to find.”).

  85. [85]. U.S. Const. amend. IV. See generally Terry v. Ohio, 392 U.S. 1 (1968).

  86. [86]. U.S. Const. amend. V. See generally Miranda v. Arizona, 384 U.S. 436 (1966); Brown v. Mississippi, 297 U.S. 278 (1936).

  87. [87]. U.S. Const. amend. VI. See generally Gideon v. Wainwright, 372 U.S. 335 (1963).

  88. [88]. Bibas, supra note 82, at 32.

  89. [89]. Richman, supra note 80, at 767 (“Maybe some... prosecutors leave the comfort of their offices to pound the pavement investigating cases. But this generally happens only in the movies—which don’t have to worry about niceties like the rule precluding a lawyer from acting as both an advocate and sworn or unsworn witness—if at all.” (footnote omitted)). But see Anthony C. Thompson, It Takes a Community to Prosecute, 77 Notre Dame L. Rev. 321, 346 (2002) (“Some prosecutors have taken the... step of placing prosecutors’ offices within the community itself in storefronts, police precincts, and housing projects.”).

  90. [90]. Richman, supra note 80, at 768; see also Kohler-Hausmann, supra note 84, at 638 (discussing a type of arrest in New York known as a “Desk Appearance Ticket” where the arrestee is released from the police precinct with a ticket notifying her of a court date).

  91. [91]. Mike McConville et al., The Case for the Prosecution 56 (1991); see also Alafair S. Burke, Improving Prosecutorial Decision Making: Some Lessons of Cognitive Science, 47 Wm. & Mary L. Rev. 1587, 1615 (2006) (“Because police agencies act independently of prosecutors’ offices in most jurisdictions, prosecutors have no guarantee that police will give them the information they need to make a fully informed evaluation of a case.”); Richman, supra note 80, at 767–68 (“[Law enforcement officers] have the expertise, the manpower, the technical resources, and, perhaps most importantly, the informational networks that no [prosecutor] possesses, and without which few cases could be brought.”).

  92. [92]. See Paul Butler, Let’s Get Free: A Hip Hop Theory of Justice 113 (2009) (“Witnesses are sometimes reluctant to testify in criminal cases; they don’t want to get involved because of mistrust of the police, fear of the defendant, or not wanting to be perceived as a snitch.”).

  93. [93]. McConville et al., supra note 91, at 56.

  94. [94]. Anil Kalhan, Stop and Frisk, Judicial Independence, and the Ironies of Improper Appearances, 27 Geo. J. Legal Ethics 1043, 1054–55 (2014) (noting that the plaintiffs in Floyd v. City of New York claimed that “based on data received from the City... the NYPD had adopted a de facto policy of unconstitutional racial profiling in violation of the written policy it had adopted pursuant to the settlement.”); Kay L. Levine & Ronald F. Wright, Prosecution in 3-D, 102 J. Crim. L. & Criminology 1119, 1176 (2012) (“[A prosecutor] often has to assert authority in a face-to-face conversation with the arresting officer [when declining to press charges].”).

  95. [95]. See Jennifer E. Koepke, Note, The Failure to Breach the Blue Wall of Silence: The Circling of the Wagons to Protect Police Perjury, 39 Washburn L.J. 211, 221 (2000).

  96. [96]. About half of the states require felonies to be presented to a grand jury; in the other half, and for most misdemeanors, a prosecutor can file charges without an indictment. See Jeffrey Fagan & Bernard E. Harcourt, Professors Fagan and Harcourt Provide Facts on Grand Jury Practice in Light of Ferguson Decision, Colum. L. Sch. (Dec. 5, 2014, 12:00 PM),

  97. [97]. Angela J. Davis, Arbitrary Justice: The Power of the American Prosecutor 25 (2007) (“[B]ecause the prosecutor maintains unilateral control over the grand jury, in most cases [it] is simply a tool of the prosecutor....”). see also Levine, supra note 16, at 761–62 (discussing prosecutorial control over grand juries).

  98. [98]. Ric Simmons, Re-Examining the Grand Jury: Is There Room for Democracy in the Criminal Justice System?, 82 B.U. L. Rev. 1, 63 (2002).

  99. [99]. See Lindsey Devers, Bureau of Justice Assistance, U.S. Dep’t. of Justice, Research Summary: Plea and Charge Bargaining 1 (Jan. 24, 2011),

  100. [100]. See Russell M. Gold, Beyond the Judicial Fourth Amendment: The Prosecutor’s Role, 47 U.C. Davis L. Rev. 1591, 1657 (2014); Christopher Slobogin, Testilying: Police Perjury and What to Do About It, 67 U. Colo. L. Rev. 1037, 1041 (1996).

  101. [101]. Koepke, supra note 95, at 211 (“Officers are more likely to get struck by lightning than prosecuted for perjury.” (quoting Ruben Castaneda, Police Officer Perjury Not Rare, Observers Say, Wash. Post (Feb. 17, 1999),

  102. [102]. United States v. Taylor, 279 F. Supp. 2d 242, 244 n.2 (S.D.N.Y. 2003) (“The Court, however, discredits... the police officers’ testimony regarding Taylor’s alleged hand-and-arm gestures, which the Court, observing the demeanor of the police witnesses as they testified on this subject, found doubtful in the extreme.”).

  103. [103]. Butler, supra note 92, at 102 (“One of your primary functions as a prosecutor is to make the judge and jury believe the police.”).

  104. [104]. See, e.g., United States v. LeFevour, 798 F.2d 977, 983 (7th Cir. 1986) (“May the government pull the sting of cross-examination by asking the question on direct examination? We have twice upheld the propriety of this practice....”); State v. Baines, No. COA11-279, 2011 WL 4357365, at *2 (N.C. Ct. App. Sept. 20, 2011) (“Revealing damaging information on direct examination instead of waiting for it to be revealed on cross examination is a strategy known as ‘pulling the sting’ or ‘drawing the sting.’”).

  105. [105]. Richman, supra note 80, at 792 (footnote omitted).

  106. [106]. Id. (quoting Donald Chisolm, Coordination Without Hierarchy: Informal Structures in Multiorganizational Systems 85 (1989)).

  107. [107]. See infra Part III.A.3; see also Daniel S. Medwed, The Zeal Deal: Prosecutorial Resistance to Post-Conviction Claims of Innocence, 84 B.U. L. Rev. 125, 134 (2004) (“An individual prosecutor’s conviction rate may provide a quantifiable method for superiors in the office to measure that prosecutor’s success in an occupation where job performance, aside from anecdotal evidence, is otherwise difficult to gauge.”); Stuntz, supra note 1, at 534 (“[P]rosecutors have a substantial incentive to win the cases they bring. One piece of evidence for this fairly obvious proposition is the frequency with which elected prosecutors cite conviction rates in their campaigns. This political need is no doubt reinforced by a kind of consumption preference—all litigators prefer winning to losing, and one must assume prosecutors share that preference.”); Zacharias, supra note 18, at 109 n.264 (“[A] prosecuting office might adjust its emphasis on convictions in evaluating individual prosecutors for promotions and other benefits. To the extent a prosecutor’s conviction rate is all that counts, the institutional incentives point toward minimizing the responsibility to ‘do justice.’”).

  108. [108]. See Richman, supra note 80, at 792.

  109. [109]. Davis, supra note 97, at 40 (“[S]ome prosecutors don’t even question police about [whether their practices in a given case are lawful]. It’s easier to simply go forward with the prosecution than engage in the thorny exercise of confronting the very police officers on whom they rely to successfully prosecute their cases.”).

  110. [110]. See, e.g., Panwala, supra note 79, at 647 (“It should not come as a surprise then that the Criminal Section of the Civil Rights Division of the United States Department of Justice reports a higher success rate for all other prosecutions than for official misconduct cases. For example, the Criminal Section’s overall success rate compared to its rate of success in law enforcement cases for the years 1990 to 1994 were 94.4 percent to 77.8 percent (1990), 89.3 percent to 80.6 percent (1991), 85 percent to 62.2 percent (1992), 73.6 percent to 58.7 percent (1993), and 90.2 percent to 78.7 percent (1994).”); Marshall Miller, Note, Police Brutality, 17 Yale L. & Pol’y Rev. 149, 154 (1998) (“[P]rosecutions of police officers occur remarkably infrequently. Between 1981 and 1991 in Los Angeles, the District Attorney brought excessive force prosecutions in forty-three cases—less than one-quarter of one percent of alleged acts of excessive force. Federal prosecutors were even less active. The Department of Justice initiated only three prosecutions against police officers in Los Angeles during the same ten-year period. The import of these statistics is clear: the criminal justice system punishes officers engaging in misconduct so rarely that it could not be expected to deter potential future offenders.” (footnotes omitted)).

  111. [111]. Butler, supra note 92, at 101.

  112. [112]. Id.; see also Thompson, supra note 89, at 331 (“[P]rosecutors use sensible measures to gauge their effectiveness at fulfilling th[eir crime-reduction] mandate. First, they often focus on conviction rates.”).

  113. [113]. Davis, supra note 97, at 34.

  114. [114]. Id. at 47.

  115. [115]. Id. at 34.

  116. [116]. See, e.g., Levenson, Lessons of Rodney King, supra note 10, at 558 (“[T]he [Special Investigation Division (“SID”)] of the Los Angeles County District Attorney’s Office prosecute[s] police misconduct cases. Seventeen lawyers assigned to the Los Angeles SID are responsible for prosecuting police officers and public officials. These individuals are experienced prosecutors who have garnered an average of ten years of experience before they enter the unit.” (footnote omitted)).

  117. [117]. This is certainly true regarding obtaining convictions in nonpolice prosecutions. See Bibas supra note 82, at 43 (“Because [district attorneys] face electoral pressure to maximize convictions, they push their unelected subordinates to increase conviction rates.”).

  118. [118]. Bernard E. Harcourt, Three Essays in Criminal Justice: Verdict and Illusion 3–5 (Columbia Law Sch. Pub. Law & Legal Theory Working Paper Grp., Paper No. 14-480, 2015), (discussing how the grand jury decision in the Wilson case was referred to as a “verdict,” suggesting an illusion of process that did not exist); Jeffrey Toobin, How Not to Use a Grand Jury, New Yorker (Nov. 25, 2014), (suggesting that the prosecutor used the grand jury to get the result he wanted); cf. Levine, supra note 16, at 772 (discussing how grand juries in police cases represent a model for the criminal justice system).

  119. [119]. See Documents Released in the Ferguson Case, N.Y. Times (Dec. 15, 2014), http://www.ny

  120. [120]. This statement is also supported by the number of cases in which indictments are handed down by grand juries. For instance, according to the Bureau of Justice Statistics, in more than 162,500 cases prosecuted by federal prosecutors, the grand jury failed to return an indictment in only 11. Mark Motivans, Bureau of Justice Statistics., U.S. Dep’t of Justice, Federal Justice Statistics 2010—Statistical Tables 11–12 (2013),
    content/pub/pdf/fjs10st.pdf. This number was arrived at by taking the total number of cases reported (193,021) and subtracting those that were declined by prosecutors (30,670); out of the remainder of cases presented to a grand jury, only 11 were dismissed. See id.

  121. [121]. The Supreme Court has ruled that it is not necessary for prosecutors to present exculpatory evidence to a grand jury. See United States v. Williams, 504 U.S. 36, 51 (1992) (“[R]equiring the prosecutor to present exculpatory as well as inculpatory evidence would alter the grand jury’s historical role, transforming it from an accusatory to an adjudicatory body.”).

  122. [122]. See Davis, supra note 97, at 26.

  123. [123]. Id. (“Neither the defendant nor the defense attorney is allowed to be present during the process.”).

  124. [124]. Simmons, supra note 98, at 37–38 (“[Testifying before the grand jury] entails real risks. Anything that the defendant says in the grand jury can be used against him or her at trial. Also, by presenting a case so early on, the defendant must devise—and effectively disclose to the prosecution—his or her theory of the case. Furthermore, in some jurisdictions, the District Attorney has a policy of refusing to plea bargain any case in which the defendant testifies before the grand jury.” (footnote omitted)).

  125. [125]. See Fagan & Harcourt, supra note 96 (“The proceedings resembled a trial rather than a grand jury proceeding. For example, the transcripts show that the prosecutors cross-examined potential prosecution witnesses, probing for inconsistencies in their testimony. They were openly skeptical of the testimony of others.There were about 60 witnesses called during almost 75 hours of proceedings, resulting in almost 5,000 pages of transcript.”).

  126. [126]. See, e.g., Monica Davey & Julie Bosman, Protests Flare After Ferguson Police Officer Is Not Indicted, N.Y. Times (Nov. 24, 2014),

  127. [127]. Dana Milbank, Opinion, Bob McCulloch’s Pathetic Prosecution of Darren Wilson, Wash. Post (Nov. 25, 2014), (“[McCulloch’s statement to the media] essentially acknowledged that his team was serving as Wilson’s defense lawyers, noting that prosecutors ‘challenged’ and ‘confronted’ witnesses by pointing out previous statements and evidence that discredited their accounts.”).

  128. [128]. See infra Part IV.C.

  129. [129]. Bibas, supra note 82, at 52 (“[G]rand juries used to publicize prosecutorial declinations and other hidden executive actions, which increased accountability and checked agency costs.”); see also J. David Goodman, Appeals Court Rules for Transcripts in Eric Garner Case to Remain Sealed, N.Y. Times (Nov. 23, 2015),

  130. [130]. Butler, supra note 92, at 106; see also Davis, supra note 97, at 5 (“Prosecutors make... most important... discretionary decisions behind closed doors and answer only to other prosecutors.”).

  131. [131]. See, e.g., Legal Department, S. States PBA,
    Department_125.jsp (last visited Feb. 21, 2016) (noting that the PBA’s legal team represents members when they need legal assistance in relation to their official duties).

  132. [132]. Pervaiz Shallwani et al., Police Unions Blast Mayor in Chokehold Case, Wall Street J. (Aug. 6, 2014, 12:41 AM),; Julia Shu & Jorteh Senah, Police Unions Go on Offensive After Eric Garner’s Death, WNYC News (Aug. 5, 2014)

  133. [133]. Mark Berman, Why Did Authorities Say Laquan McDonald Lunged at Chicago Police Officers?, Wash. Post (Nov. 25, 2015),
    11/25/why-did-authorities-say-laquan-mcdonald-lunged-at-chicago-police-officers; see also Bernard E. Harcourt, Opinion, Cover-Up in Chicago, N.Y. Times (Nov. 30, 2015),
    2015/11/30/opinion/cover-up-in-chicago.html?_r=0 (suggesting that the cover-up of the shooting went all the way up to the mayor’s office and calling for Rahm Emanuel to resign).

  134. [134]. DNAinfo Chicago, Dashcam Video of Officer Jason Van Dyke Shooting Laquan McDonald, YouTube (Nov. 24, 2015), (containing video released by the Chicago Police Department). The video was not released to the public until a judicial order, more than a year after the shooting, required its release. The officer, Jason Van Dyke, has been charged with murder by the Cook County State’s Attorney. See People’s Factual Proffer in Support of Setting Bond at 1, State v. Van Dyke, No. 15-127823 (Ill. Cir. Ct., Nov. 24, 2015).

  135. [135]. N.R. Kleinfield & John Eligon, Officers Jeer at Arraignment of 16 Colleagues in Ticket-Fixing Investigation, N.Y. Times (Oct. 28, 2011),

  136. [136]. Id.

  137. [137]. 5W Public Relations & 5WPR CEO Ronn Torossian, Detective Joe Crystal, Today’s Frank Serpico, on the Reid Report MSNBC on Public Corruption, YouTube, (Jan. 5, 2015), https://www. (interviewing former Detective Joe Crystal).

  138. [138]. Id.; see also Koepke, supra note 95, at 216 (“Perhaps what is most disturbing about the [Abner] Louima [police assault] is that of the 100 officers offered limited immunity, only two would testify as to their knowledge of Louima’s torture in the early stages of the investigation.”).

  139. [139]. Davis, supra note 97, at 166.

  140. [140]. Id. at 167.

  141. [141]. Id. (“Very few people understand the day-to-day responsibilities of prosecutors, nor do they seem to be interested in what prosecutors do.”).

  142. [142]. Peter L. Davis, Rodney King and the Decriminalization of Police Brutality in America: Direct and Judicial Access to the Grand Jury as Remedies for Victims of Police Brutality When the Prosecutor Declines to Prosecute, 53 Md. L. Rev. 271, 281(1994).

  143. [143]. Id. at 282.

  144. [144]. See, e.g., Thompson, supra note 89, at 327 (“Prosecutors certainly initiate prosecutions in the name of ‘the people’ and maintain a trustee’s obligation to safeguard the people’s interest. But the extent to which prosecutors actually serve the people themselves or instead serve the government remains unclear.”).

  145. [145]. See, e.g., Brenner & Durham, supra note 30, at 471–72 (discussing the difficulties in identifying conflicts of interest for prosecutors); Levenson, supra note 30, at 880 (“Frankly, defense lawyers have it easy. They have a duty to a client or clients—flesh and blood people who they have to be able to look at and say, ‘I did my best for you. I put your interests ahead of everyone else’s. I pulled out all the stops and did not compromise your interests for anyone else’s—not for my own interests (financial, professional or personal), not for another client’s interests (past, present, or future), and not for the interests of the witnesses, judge, your family or the public. I was there for you!’”); Deborah L. Rhode, Conflicts of Commitment: Legal Ethics in the Impeachment Context, 52 Stan. L. Rev. 269, 280 (2000) (“When acting in a prosecutorial capacity, attorneys have no client whose interests can be objectively gauged or whose consent can be obtained.”).

  146. [146]. Working Families Party v. Fisher, 15 N.E.3d 1181, 1185 (N.Y. 2014) (per curiam) (“Where there is legitimate doubt as to whether a district attorney and his office may proceed with a case, the district attorney is not barred from resolving that doubt by choosing to step aside.”).

  147. [147]. See, e.g., Statement of Brooklyn District Attorney Ken Thompson, supra note 6 (“[L]ocal prosecutors who are elected to enforce the laws in [their] communities should not be robbed of their ability to [prosecute the police].”).

  148. [148]. See supra Part II.D.

  149. [149]. Cf. Rachel E. Barkow, Federalism and Criminal Law: What the Feds Can Learn from the States, 109 Mich. L. Rev. 519, 545 (2011) (“In almost every state, a conscious choice has been made to defer to local prosecutors. States have centralized authority in a statewide prosecutor in a handful of areas, and there is remarkable overlap among the states in terms of the content of those areas. But outside these contexts, local prosecutors are responsible for the vast bulk of criminal law enforcement within a state. And these local prosecutors are operating in most states with little centralized supervision by a state-level actor.”).

  150. [150]. I do not argue that such a conclusion comes without significant costs. I will address the costs as well as several proposed solutions to the conflict in the final Part of this Article.

  151. [151]. Gershman, supra note 18, at 444; see also Zacharias, supra note 18, at 46 (stating it is prosecutor’s ethical duty to “do justice”).

  152. [152]. Steven A. Drizin & Richard A. Leo, The Problem of False Confessions in the Post-DNA World, 82 N.C. L. Rev. 891, 1006 (2004) (“The fact that a conviction is virtually assured [after a confession] can blind some prosecutors to their ethical obligation to pursue the truth and seek justice.”); Marvin E. Frankel, The Search for Truth: An Umpireal View, 123 U. Pa. L. Rev. 1031, 1038 (1975) (noting that prosecutors’ “enthusiasm for the principle that they must seek justice, not merely convictions” has been mixed).

  153. [153]. See supra Part II.A.

  154. [154]. Bibas, supra note 82, at 50 (“People respect the law more when it is visibly fair and when they have some voice or control over its procedures. Procedural fairness, process control, and trust in [prosecutors’] motives contribute greatly to the criminal justice system’s legitimacy.”); Gershman, supra note 18, at 455 (discussing prosecutors’ dual role as both “an aggressive advocate” and a “quasi-judicial official”).

  155. [155]. Brenner & Durham, supra note 30, at 487 (“[T]he notion of ‘justice’ requires that the public shall have absolute confidence in the integrity and impartiality of the process by which it is administered.”).

  156. [156]. In response to the lack of data, President Obama recently signed into law a bill requiring departments to report fatalities. See Death in Custody Reporting Act of 2013, Pub. L. No.
    113–242, §2, 128 Stat. 2860 (2014) (codified at 42 U.S.C. §13727 (Supp. II 2014)).

  157. [157]. Sarah Ryley et al., In 179 Fatalities Involving On-Duty NYPD Cops in 15 Years, Only 3 Cases Led to Indictments—and Just 1 Conviction, N.Y. Daily News (Dec. 8, 2014, 2:30 AM), http://www.nydaily

  158. [158]. Erin Alberty, Killings by Utah Police Outpacing Gang, Drug, Child-Abuse Homicides, Salt Lake Trib. (Nov. 23, 2014, 10:36 AM),

  159. [159]. See, e.g., Ryan Grim et al., From Daniel Pantaleo to Darren Wilson, Police Are Almost Never Indicted, Huffington Post (Dec. 3, 2014, 5:15 PM),

  160. [160]. Bibas, supra note 82, at 40 (“Outsiders... care about a host of process benefits that come from transparency and participation....[I]nsiders, however, do little to deliver these process goods.”).

  161. [161]. Ryan Gabrielson et al., Deadly Force, in Black and White, ProPublica (Oct. 10, 2014, 10:07 AM),; see also Cody T. Ross, The United States Police-Shooting Database: A Multi-Level Bayesian Analysis of Racial Bias in Police Shootings at the County-Level in the United States, 2011–2014, at 7 (Dec. 6, 2014) (unpublished manuscript), (“Across almost all counties, individuals who were armed and shot by police had a much higher probability of being black or hispanic than being white. Likewise, across almost all counties, individuals who were unarmed and shot by police had a much higher probability of being black or hispanic than being white. Most tragically, across a large proportion of counties, individuals who were shot by police had a higher median probability of being unarmed black individuals than being armed white individuals. While this pattern could theoretically be explained by reduced levels of crime being committed by armed white individuals, it still begs the question as to why there exist[s] such a high rate of police shootings... of unarmed black individuals.”).

  162. [162]. See Erika Harrell, Bureau of Justice Statistics., U.S. Dep’t of Justice, Black Victims of Violent Crime 3, 5 (2007), (“About 93% of black homicide victims and 85% of white victims in single victim and single offender homicides were murdered by someone of their race” and “[a]bout four-fifths of black victims of nonfatal violence perceived the offenders to be black.”).

  163. [163]. Id. at 1 (“While blacks accounted for 13% of the U.S. population in 2005, they were victims in 15% of all nonfatal violent crimes and nearly half of all homicides.”).

  164. [164]. Anthony V. Alfieri, Community Prosecutors, 90 Calif. L. Rev. 1465, 1466 (2002) (“[W]ithin the sphere of criminal justice and legal ethics, race shadows the actions of prosecutors and defenders yet often fails to rouse debate.”).

  165. [165]. For a first-hand description of this perception problem in the context of the Rodney King trial, see Levenson, Lessons of Rodney King, supra note 10, at 562–63 (“Consider, for example, what occurred during the federal King retrial. It was at least awkward, if not disturbing, when two white prosecutors argued against four white defense lawyers to a white judge the meaning of the phrase ‘Mandingo Sexual Encounter.’ Neither the lawyers nor the judge seemed familiar with the racially-charged phrase that many African–Americans undoubtedly could have responded to on the spot. It was also awkward that the young African–American prosecutor assigned to the King case ordinarily sat in back of the trial lawyers and played only a small role during the trial. In a case which had caused many African–Americans to become cynical about the responsiveness of the criminal justice system to their needs, there was little visual assurance during the trial that their interests were vigorously represented.” (footnotes omitted)). This passage, of course, described the federal trial and not a local prosecution, but the anecdote is representative of any trial of white police officers accused of harming a minority suspect with white defense attorneys, white prosecutors, and a white judge, which, given the statistics, is likely to be the case in any such trial.

  166. [166]. Brian A. Reeves, Bureau of Justice Statistics., U.S. Dep’t of Justice, Local Police Departments, 2007, at 14 tbl.9 (2010),

  167. [167]. Seth W. Stoughton, The Incidental Regulation of Policing, 98 Minn. L. Rev. 2179, 2218 (2014) (“[M]ost local law enforcement agencies are not racially representative of the population they serve.”).

  168. [168]. Nicholas Fandos, A Study Documents the Paucity of Black Elected Prosecutors: Zero in Most States, N.Y. Times (July 7, 2015),

  169. [169]. The National District Attorney’s Association does not keep statistics on race or gender in state offices. See E-mail from Richard T. Hanes, Chief of Staff, Nat’l Dist. Attorneys Ass’n, to author (Jan. 7, 2015) (on file with author) (confirming that it does not keep race or gender statistics). The Bureau of Justice Statistics also does not have a breakdown of prosecutor’s offices by race. See Steven W. Perry & Duren Banks, Bureau of Justice Statistics., U.S. Dep’t of Justice, 2007 National Census of State Court Prosecutors: Prosecutors in State Courts, 2007—Statistical Tables (2011), (showing statistics of offices, not including race).

  170. [170]. And, in fact, one former prosecutor has argued because of the systemic racism in the criminal justice system, African-Americans should never become prosecutors. See Butler supra note 92, at 101–22. But see Levenson, Lessons of Rodney King, supra note 10, at 562 (“[In 1994 l]ess than 10% of the federal prosecutors in Los Angeles [we]re African-American or Hispanic. By contrast, 22% of the lawyers in the Los Angeles District Attorney’s Office [we]re minorities.” (footnote omitted)).

  171. [171]. Milbank, supra note 127. Meanwhile the Wilson case was widely contrasted with a criminal accusation against black officer, Dawon Gore. As the grand jury decided Wilson’s case, he remained on paid leave. Meanwhile, Gore was on unpaid leave after charges were filed against him for assaulting a suspect by striking him with his baton while effecting an arrest. See Cassandra Fairbanks, St. Louis Police Officer Blows the Whistle on Rampant Corruption Within the Department, Free Thought Project (Aug. 28, 2014), (“Officer Gore is currently suspended without pay after using non lethal force on a man who attacked him in front of witnesses.”). These two cases struck many as stark examples of the differential treatment of black and white police officers in Ferguson. See, e.g., Miles Klee, Ferguson Prosecutor Indicted a Black Cop Who Hit a Man with His Baton, Daily Dot (Nov. 26, 2014, 1:05 PM), (“A guilty verdict [in the Gore case] might serve as another flashpoint in the national debate over institutionalized racism....”).

  172. [172]. Peter Moore, Poll Results: Police, YouGov (Aug. 14, 2014, 11:35 AM),

  173. [173]. See Liteky v. United States, 510 U.S. 540, 548 (1994).

  174. [174]. While rare, there are a few cases where courts have discussed the necessity of removing an entire office. See, e.g., City & Cty. of S.F. v. Cobra Sols., Inc., 135 P.3d 20, 29–30 (Cal. 2006) (“Individuals who head a government law office occupy a unique position because they are ultimately responsible for making policy decisions that determine how the agency’s resources and efforts will be used. Moreover, the attorneys who serve directly under them cannot be entirely insulated from those policy decisions, nor can they be freed from real or perceived concerns as to what their boss wants.”); State v. Kinkennon, 747 N.W.2d 437, 444 (Neb. 2008) (“We recognize that complete disqualification of a prosecutor’s office may be warranted in cases where the appearance of unfairness or impropriety is so great that the public trust and confidence in our judicial system simply could not be maintained otherwise. Such an extreme case might exist, even where the State has done all in its power to establish an effective screening procedure precluding the individual lawyer’s direct or indirect participation in the prosecution.”).

  175. [175]. Young v. United States ex rel. Vuitton et Fils S.A., 481 U.S. 787, 814 (1987).

  176. [176]. There is no instance where such a motion has been reported. However, it is imaginable that in a high profile case where there is political pressure to bring a case but not actual evidence that a crime was committed, an officer might raise the issue.

  177. [177]. See, e.g., People v. Schrager, 74 Misc. 2d 833, 834 (N.Y. Sup. Ct. 1973) (“Defendant is an Assistant District Attorney of Queens County.... The motion is grounded upon the District Attorney’s declaration of disqualification stemming from the professional and personal relationships which this defendant shared with the members of the District Attorney’s staff.”). While few reported cases exist to support this statement, I have had conversations with a number of former assistant district attorneys who confirm that this practice is routine for all district attorney offices.

  178. [178]. See Mickens v. Taylor, 535 U.S. 162, 174 (2002); supra Part II.C.

  179. [179]. The few instances where prosecutors are disqualified amplify this point. Prosecutors have recused themselves or have been disqualified when they have a “significant professional relationship with [the defendants].” State v. Gonzales, 119 P.3d 151, 162 (N.M. 2005); see also State v. Hursey, 861 P.2d 615, 617 (Ariz. 1993) (en banc); People v. Manzanares, 139 P.3d 655, 658–59 (Colo. 2006) (en banc); Gatewood v. State, 880 A.2d 322, 331 (Md. 2005). They have been disqualified when they had a conflicting interest in a civil case. See Sinclair v. State, 363 A.2d 468, 469–70, 475 (Md. 1976). They have also been disqualified when the prosecutor received “a crime victim’s payment of substantial investigative expenses already incurred by the public prosecutor.” People v. Eubanks, 927 P.2d 310, 315, 320 (Cal. 1996), as modified on denial of reh’g (Feb. 26, 1997) (“More to the present point, a prosecutor may have a conflict if institutional arrangements link the prosecutor too closely to a private party, for example a victim, who in turn has a personal interest in the defendant’s prosecution and conviction.”). They have been disqualified when the defendant allegedly made derogatory statements about the prosecutor. See State v. King, 956 So. 2d 562, 562 (La. 2007) (“Because we find that the district attorney’s personal animosity toward defendant stemming from the district attorney’s belief that defendant started or spread a salacious rumor about him and a member of his family was a factor in making certain prosecutorial decisions, we believe his ability to fairly and impartially conduct defendant’s trial was called into question.”). And they have recused themselves when a testifying victim was a cousin of the head of a U.S. Attorney’s office. See United States v. Sigillito, 759 F.3d 913, 928 (8th Cir. 2014).

  180. [180]. See Levenson, Lessons of Rodney King, supra note 10, at 558 (noting that in the special investigation unit of the Los Angeles District Attorney’s office, the prosecutors had an average of 10 years of experience as prosecutors).

  181. [181]. Recently, the incumbent District Attorney in Dallas, Texas, lost his reelection bid after the police union (the “DPA”) mobilized against him because he prosecuted an officer. Tristan Hallman, Dallas Police Association Optimistic About Future with New DA, Dall. Morning News (Jan. 7, 2015, 11:20 PM), (quoting DPA President, Ron Pinkston, as stating that “[o]ur previous DA made a point of being confrontational to law enforcement,” and noting that the DPA’s political action committee had contributed more to an opponent’s successful campaign bid than in any other election after the incumbent district attorney removed police investigation units from his charging decisions about police crimes). In Riverside County, California, the incumbent district attorney was also unseated by a challenger who raised $60,000 from the police union and had strong ties to law enforcement. Riverside County District Attorney Paul Zellerbach Unseated After One Term, NBC L.A. (June 4, 2014, 8:12 AM), http://www.nbc; see also Jeff Horseman, Challenger Out-Raises Incumbent in DA Race, Press Enterprise (Feb. 5, 2014, 6:11 PM), (“Thanks in part to law enforcement unions [including deputy district attorney unions and police unions], Mike Hestrin raised more money in 2013 than incumbent Paul Zellerbach in the race for Riverside County district attorney, campaign finance records show.”); Private Investigators Working for Police Union Arrested for Crimes Against Costa MesaCouncilmen, CBS L.A. (Dec. 11, 2014, 12:26 PM), (“[The arrestees were retained by the] Costa Mesa Police Officers’ Association... to conduct ‘candidate research,’ including surveillance on Costa Mesa city council members, in the months leading up to the November 2012 election, authorities said.”).

  182. [182]. See supra note 181 and accompanying text.

  183. [183]. Little, supra note 75, at 754 (“[G]ratitude and loyalty can have a powerful influence for a [system actor] undertaking to decide a case.”). The same can be said for prosecutors deciding whether or not to pursue a case.

  184. [184]. See supra Part II.D.

  185. [185]. Eldred, supra note 68, at 67. See generally Chugh et al., supra note 70.

  186. [186]. See supra Part II.B–C. At least one state court noted such cognitive dissonance when disqualifying a prosecutor who had refused to recuse himself. State v. King, 956 So. 2d 562, 566 (La. 2007) (stating that, while the court was “satisfied... of the sincerity of [the prosecutor’s] belief that he has impartially discharged the duty of his office,” the conflict required his disqualification (quoting State v. Take, 171 So. 108, 112 (La. 1936))).

  187. [187]. See, e.g., Igor Bobic, Bar Association Calls on Prosecutor to Recuse Himself from Ferguson Investigation, Huffington Post (Aug. 19, 2014, 11:55 AM),

  188. [188]. See, e.g., Kimberly Kindy, Objectivity of Prosecutor in Missouri Shooting of Michael Brown Is Questioned, Wash. Post (Aug. 15, 2014),

  189. [189]. See, e.g., Mayra Cuevas, Prosecutor in Michael Brown Case to Governor: “Man Up” If You Want Me Out, CNN (Aug. 21, 2014, 8:07 AM),

  190. [190]. See, e.g., Peter Holley, Ferguson Prosecutor Says He Knew Some Witnesses Were ‘Clearly Not Telling the Truth.’ They Testified Anyway, Wash. Post (Dec. 20, 2014), http://www.washingtonpost.

  191. [191]. See, e.g., Letter from Frank A. Sedita, III, President Dist. Attorneys Ass’n of the State of N.Y. to Andrew M. Cuomo, Governor (Dec. 16, 2014),
    decisions15/012115letter.pdf (noting that its members “unanimously... oppose[ ]... any per se rule mandating [recusal of the local district attorney in] fatal police–citizen encounters”).

  192. [192]. Barkow, supra note 149, at 549–50.

  193. [193]. Id. at 551–52 (“Most state-level prosecutors who possess broad jurisdictional grants of power are exercising their discretion sparingly and with a focus on specific areas.”).

  194. [194]. See, e.g., 71 Pa. Cons. Stat. §732-205(a)(3) (2012); Tenn. Code Ann. §8-6-112(a)(2) (West 2013).

  195. [195]. See Kamisar, supra note 6. But see David M. Jaros, Preempting the Police, 55 B.C. L. Rev. 1149, 1151 (2014) (“Unfortunately, although egregious cases of police misconduct can temporarily galvanize the public and, for a short time, their representatives, the politics of crime tends to deter politicians from taking an active role in limiting police power.”).

  196. [196]. See Exec. Order No. 147 , supra note 6.

  197. [197]. See Letter from Eric T. Schneiderman, Attorney Gen. of N.Y., to Andrew M. Cuomo, Governor (Dec. 8, 2014),

  198. [198]. Barkow, supra note 149, at 551 (“[T]he [New York] AG’s office cannot bring criminal charges without the approval of the governor and the appropriate district attorney.”).

  199. [199]. Id. at 549 n.140 (citing state statutes that allow attorneys general to step in for conflicts).

  200. [200]. See, e.g., Samuel Walker & Morgan Macdonald, An Alternative Remedy for Police Misconduct: A Model State Pattern or Practice Statute, 19 Geo. Mason U. Civ. Rts. L.J. 479, 539 (2009) (“[S]tate attorneys general are held accountable by the ballot box....”).

  201. [201]. Barkow, supra note 149, at 545 (“States have centralized authority in a statewide prosecutor in a handful of areas [including public corruption, election fraud, benefits fraud, and regulatory crimes] and there is remarkable overlap among the states in terms of the content of those areas. But outside these contexts, local prosecutors are responsible for the vast bulk of criminal law enforcement within a state.”).

  202. [202]. 18 U.S.C. §242 (2012). But see Levenson, Lessons of Rodney King, supra note 10, at
    556–58 (arguing that this higher bar may not actually be so different in practice). However, Levenson’s article draws only on the Rodney King prosecution. Thus, the insight may not be more
    generally applicable.

  203. [203]. This is what happened after the state prosecution failed to secure a conviction against any of the accused officers in the Rodney King case. See generally Levenson, Lessons of Rodney King, supra note 10.

  204. [204]. Susan N. Herman, Reconstructing the Bill of Rights: A Reply to Amar and Marcus’s Triple Play on Double Jeopardy, 95 Colum. L. Rev. 1090, 1097 (1995) (“[S]tatutes, or perhaps even administrative regulations, could authorize federal investigators and prosecutors to cooperate in state criminal investigations and prosecutions.”); Donald H. Zeigler, Twins Separated at Birth: A Comparative History of the Civil and Criminal Arising Under Jurisdiction of the Federal Courts and Some Proposals for Change, 19 Vt. L. Rev. 673, 775 (1995) (“Federal prosecutors could appear in state court without any additional authorization by Congress, although the states might have to give permission for the federal prosecutor also to prosecute the related state charges.” (footnote omitted)).

  205. [205]. But see Wayne A. Logan, Creating a “Hydra in Government”: Federal Recourse to State Law in Crime Fighting, 86 B.U. L. Rev. 65, 69 (2006) (noting that state and federal law enforcement frequently work together in narcotics prosecutions).

  206. [206]. Samuel W. Buell, The Upside of Overbreadth, 83 N.Y.U. L. Rev. 1491, 1517 n.83 (2008) (“Federal prosecutors are largely proactive and insulated from electoral politics.”); Frederick M. Lawrence, The Evolving Federal Role in Bias Crime Law Enforcement and the Hate Crimes Prevention Act of 2007, 19 Stan. L. & Pol’y Rev. 251, 274 (2008) (“Federal prosecutions [are] brought by an appointed United States Attorney who, although not necessarily altogether isolated from the political process, is nonetheless largely immune from politics.”).

  207. [207]. Barkow, supra note 149, at 571 (“Federal gun and drug laws in particular overlap with large swaths of traditionally local categories of crime, and broad statutes like RICOand the Hobbs Act similarly allow federal prosecutors to pursue local crimes like murder or robbery.” (footnotes omitted)).

  208. [208]. See Steven D. Clymer, Compelled Statements from Police Officers and Garrity Immunity, 76 N.Y.U. L. Rev. 1309, 1310–11 n.6 (2001) (“There is no consensus on the frequency with which police engage in corruption or use excessive force.”); Wesley Lowery, How Many Police Shootings a Year? No One Knows, Wash. Post (Sept. 8, 2014),
    08/how-many-police-shootings-a-year-no-one-knows (“[T]he Justice Department keep[s] no comprehensive database or record of police shootings, instead allowing the nation’smore than 17,000 law enforcement agencies to self-report officer-involved shootings as part of the FBI’s annual data on ‘justifiablehomicides’ by law enforcement.”).

  209. [209]. Alfieri, supra note 164, at 1466; Thompson, supra note 89, at 346; see also Barkow, supra note 149, at 536 (“[F]ederal prosecutors do not concern themselves as much with how their selection of cases affects a community. They do not have an obligation to fix local problems, and they are not directly accountable to those communities. The federal government’s enforcement decisions therefore largely ignore the day-to-day realities of local communities.”).

  210. [210]. Thompson, supra note 89, at 324 (“Some... community prosecution programs have begun to forge exciting new working partnerships with communities in preventing and addressing crime and in defining justice.”).

  211. [211]. Id. at 336–38 (discussing training strategies for community prosecutors, some of which could be applied to federal prosecutors tasked with investigating and prosecuting local law enforcement).

  212. [212]. See, e.g., Mary Fan, The Case for Crimmigration Reform, 92 N.C. L. Rev. 75, 110–11 (2013) (“By 1998, Hispanics constituted 37% of all federal defendants while whites constituted 32% of the federal defendant population. In contrast, Hispanics constituted a minority of about 11.7% of the United States population in 1999.” (footnote omitted)); David Patton, Federal Public Defense in an Age of Inquisition, 122 Yale L.J. 2578, 2587 (2013) (“[I]n contrast to 1963, when around seventy percent of federal defendants were white, today the racial ratio has flipped: only twenty-six percent of federal defendants are non-Hispanic whites.” (footnote omitted)). The latest iteration of the U.S. Sentencing Commission’s Final Quarterly Data Report, covering fiscal year 2013, reported that 23.8% of all federal defendants were non-Hispanic whites. U.S.
    Sentencing Comm’n, Final Quarterly Data Report: Fiscal Year 2013, at 44 tbl.23 (2013),

  213. [213]. See, e.g., Fan, supra note 212, at 118 (“[Federal prosecutors]... generally enjoy greater... prestige than their state counterparts.”); Janet C. Hoeffel, Prosecutorial Discretion at the Core: The Good Prosecutor Meets Brady, 109 Pa. St. L. Rev. 1133, 1138 (2005) (“The typical AUSA has graduated from a good law school near the top of his class. He did not simply want to be a prosecutor; he wanted to be a federal prosecutor. Federal prosecutors are an elite group with enormous prestige.”).

  214. [214]. See Akhil Reed Amar & Jonathan L. Marcus, Double Jeopardy Law After Rodney King, 95 Colum. L. Rev. 1, 22–24 (1995). The authors argue against federal intrusions into criminal law except for instances of “government self-dealing: [when] a state is applying criminal law against its own officials. But outside this self-dealing situation, states generally may be presumed trustworthy enforcers of criminal law.” Id. at 24.

  215. [215]. The Wisconsin legislature introduced a bill in 2013 that would create a review board similar to what I have suggested. See Assemb. B. 409, 2013–2014 Leg., Reg. Sess. (Wis. 2014). The bill created a board consisting of a retired judge, a law enforcement officer, an assistant district attorney, a professor, and a former district attorney or district attorney who would be responsible for investigating police killings. Governor Walker signed a revised bill into law on April 23, 2014, which requires investigations by at least two officers not employed by the same agency as the accused officer. The law also says that if the district attorney declines to prosecute, the investigators will release their report. See Wis. Stat. Ann. §175.47 (West Supp. 2015).

  216. [216]. Joanna C. Schwartz, What Police Learn from Lawsuits, 33 Cardozo L. Rev. 841, 872 (2012) (“Approximately twenty percent of large police departments have some form of civilian review....”); David Alan Sklansky, Police and Democracy, 103 Mich. L. Rev. 1699, 1802–03 (2005) (“The vast majority of big-city police departments are now subject to some form of civilian oversight. The institutional structure of that oversight varies widely.” (footnote omitted)). For a good overview of civilian oversight boards, see Sean Hecker, Race and Pretextual Traffic Stops: An Expanded Role for Civilian Review Boards, 28 Colum. Hum. Rts. L. Rev. 551, 594–96 (1997).

  217. [217]. Michael P. Weinbeck, Note, Watching the Watchmen: Lessons for Federal Law Enforcement from America’s Cities, 36 Wm. Mitchell L. Rev. 1306, 1317 (2010) (“[A] corollary flaw of civilian oversight agencies is their inability to require discipline.”). But see Schwartz, supra note 216, at 872 (“[A] quarter of... civilian review boards have independent investigatory authority.”).

  218. [218]. See 18 U.S.C. §3006A(a) (2012).

  219. [219]. Charles D. Weisselberg & Su Li, Big Law’s Sixth Amendment: The Rise of Corporate White-Collar Practices in Large U.S. Law Firms, 53 Ariz. L. Rev. 1221, 1283 (2011) (“[A] substantial number of white-collar partners in large firms have served in leadership positions in U.S. Attorneys’ Offices or in important posts at Main Justice.”).

  220. [220]. I address the possible critique that such boards and special prosecutors may be overzealous below.

  221. [221]. Bibas, supra note 82, at 129.

  222. [222]. Id. at 145.

  223. [223]. See generally Alfieri, supra note 164; Thompson, supra note 89.

  224. [224]. Debra Livingston, Police Discretion and the Quality of Life in Public Places: Courts, Communities, and the New Policing, 97 Colum. L. Rev. 551, 665–66 (1997) (“Complaints are an invaluable source of community feedback and information. Analysis of complaint patterns can be used to identify individual officers who generate a disproportionately large number of citizen grievances, to highlight the need for improved training in some areas, and also to suggest the reconsideration of some police strategies.” (footnote omitted)).

  225. [225]. Cf. Angela J. Davis, The American Prosecutor: Independence, Power, and the Threat of Tyranny, 86 Iowa L. Rev. 393, 444 (2001) (“[P]rosecutors work within a prescribed budget and must allocate their resources accordingly.”).

  226. [226]. See supra Part II.

  227. [227]. Cf. Sklansky, supra note 216, at 1822 (“The history of police reform is littered with promising innovations abandoned when budgets tightened.”).


Acting Assistant Professor of Lawyering, NYU School of Law. 

Many thanks to Miriam Baer, Rachel Barkow, Erin Collins, Samuel Estreicher, Nicholas Frayn, Trevor Gardner, Bennett Gershman, Bernard Harcourt, Roderick Hills, Peter Joy, Paul Pineau, Daniel Richman, Anna Roberts, Ronald Rychlak, Andrew Schaffer, Carol Steiker, Jocelyn Simonson, Anthony Thompson, Howard Wasserman, and the participants in the NYU Lawyering Scholarship Colloquium and the SEALS Junior Scholars Works-in-Progress Workshop for helpful conversations and comments on drafts. Thanks also to the editors of the Iowa Law Review. Micah Doak and Benjamin Mejia provided excellent research assistance.