Under state law, municipal codes, and collective bargaining agreements, police officers in many jurisdictions benefit from a set of heightened procedural protections. These frequently include provisions restricting the timing and manner by which investigators interview or interrogate police, which we call “interrogation buffers.” One specific buffer is a mandatory period of delay between a use-of-force incident and ensuing investigation or interrogation. Such “delay privileges” have the predictable effect of obstructing investigations, diminishing the likelihood that culpable officers are subject to effective internal investigation, and correspondingly increasing the probability that officers disposed to use excessive force will continue to work the streets without reprimand or supplemental supervision. This Article demonstrates that federal constitutional tort law and state contract law—respectively, municipal liability under 42 U.S.C. § 1983 and the public policy doctrine of contract invalidity—can be leveraged to improve the efficacy of post-incident investigations by challenging delay privileges. Although such suits will not always succeed, there are powerful reasons to use them to raise the fiscal and reputational cost of maintaining a widespread policing practice that serves largely to promote unlawful police violence.


I. Introduction

On July 19, 2015, a University of Cincinnati police officer named Ray Tensing shot and killed an unarmed African American motorist called Samuel Dubose.1 Explaining his decision to use deadly force, Tensing reported “being dragged by [Dubose’s] vehicle and [having] to fire his weapon,” an account corroborated by at least one other officer.2 Tensing, however, was wearing a body camera that documented a different set of events. As the camera showed, Dubose neither threatened nor harmed the officer. No one was dragged by his car.3 Eleven days after the shooting, the Hamilton County prosecutor indicted Tensing on murder charges, explaining that Tensing’s initial account had been fabricated.4 The officer who corroborated Tensing’s false account was not indicted; he apparently remains on active duty.5 The Dubose shooting is exemplary of recent instances in which video evidence reveals a schism between officers’ ex post reports of how force was employed, and the events in question.6 In particular, the Dubose case is illustrative insofar as it involved concurring false exculpatory testimony from plural officers.7

Officers do not always shield their colleagues. In other instances, officers witnessing their colleagues’ use of excessive force choose to become whistleblowers. In 2012, for example, a young Baltimore detective named Joe Crystal reported on two officers who he witnessed assault a drug suspect. As Crystal recounts the story, the beating was particularly deliberate and gratuitous.8 The suspect had attempted to flee by kicking in the door of a random row house apartment and running inside. The two officers removed the suspect, placed him in their vehicle, and drove away. They then apparently received a call from a fourth Baltimore police officer, Anthony Williams, who communicated that his girlfriend lived in the apartment into which the suspect had intruded. Crystal said he had observed the two arresting officers drive back to the scene and take the suspect back into the same apartment. When Officer Williams arrived, he beat the suspect. Despite being warned by his Sergeant not to report these events, Crystal approached the State’s Attorney. As a result, Crystal recounts, he became a target for taunts and harassment while on the job. His peer officers repeatedly failed to respond to his calls for backup; he received threats of perjury prosecution in the criminal matter that his whistleblowing prompted; and a dead rat was left on his vehicle’s windshield.9 Prior to the beating incident, Crystal had been commended for his leadership and promoted quickly. After two years’ retaliation for whistleblowing, he resigned from the force.10

A more complex and compelling example of both police cover-up and whistleblowing occurred in Chicago, in what subsequently became a nationally famous police shooting. On October 20, 2014, a Chicago Police Department officer shot and killed a seventeen-year-old African American named Laquan McDonald.11 Speaking to media at the scene of the shooting, former Fraternal Order of Police spokesman Pat Camden explained that “[t]he boy lunged at police, and one of the officers opened fire.”12 Five officers, including the officer who discharged his firearm, gave statements consistent with this account.13 The next day, the local media offered a brief report on the case, based largely on the description of a police spokesman, who said that McDonald had been slashing tires with a knife, ignored police requests to drop the knife, and approached or “lunged” at officers putting them in fear for their life.14 The report also said simply that an officer shot McDonald in the chest and he was pronounced dead at the hospital. A “preliminary statement” from the police reiterated that McDonald “continued to approach the officers[,]” and that “the officer discharged his weapon, striking the offender.”15 So described, the shooting seemed justified—just one of the many deadly uses of force by Chicago police each year.16

Yet in December of 2014, our colleague Craig Futterman, along with Jamie Kalven of the Invisible Institute, announced the existence of police dashboard video of the shooting, which they said would put the events in a different light. They also claimed to have found a civilian witness who said that other such witnesses were “shooed away” from the scene without having their statements taken.17 In February 2015, Kalven obtained the autopsy report, which showed that McDonald had been shot sixteen times—a fact difficult to square with a simple claim of self-defense.18 In March, the City approved a payment of $5 million to the McDonald family to settle legal claims, even though the McDonalds had filed no suit.19 In April, the Chicago Tribune revealed the shooter’s identity, Officer Jason Van Dyke, and the fact that he had previously been the target of seventeen citizen complaints since 2006, several for the use of excessive force, including one that resulted in a jury award of $350,000.20 In May, local news reported that the manager of a Burger King franchise in the vicinity of the shooting said that, immediately after the shooting, police officers sought access to, and deleted, video from multiple security cameras for the time period around the shooting.21 Finally, journalist Brandon Smith won a state Freedom of Information Act request for the video and a state court ordered its release in November 2015.22

That video did not show McDonald “lunging” towards the officers, nor even approaching them, but walking in a direction angled away, with the knife down by his side in the hand farthest from the officers. Squad cars blocked McDonald’s possible exits. No civilians were within his reach. Van Dyke had recently exited his squad car and taken a step towards McDonald when he opened fire. The first shot hit McDonald, who “immediately” fell to the ground.23 Van Dyke then unloaded his clip, firing a total of sixteen bullets into McDonald’s body.24 Immediately before the video’s release, the State’s Attorney announced that she was charging Van Dyke with murder.25

The response to McDonald’s shooting illustrates what Chicago mayor Rahm Emmanuel has called a powerful “code of silence” among the Chicago Police Department.26 The many officers present at the scene publicly acquiesced in the official, false narrative, one that made the shooting seem a straightforward case of self-defense. Others apparently instructed witnesses to leave the area without taking their statement. Apparently, at least one officer deleted Burger King surveillance videos of the incident. But for the actions of Futterman, Kalven, and Smith, the case might have faded away without criminal, civil, or even professional repercussions for Officer Van Dyke.27 And yet despite the police cover-up, there was also a whistleblower. Kalven and Futterman first asked for an autopsy only because a whistleblower alerted them to the existence of a dashcam video showing “horrific” events.28 The whistleblower remains anonymous. But it is reasonable to assume it was a police officer who wanted accountability for an unjustified killing, but who did not want to suffer Officer Crystal’s fate.

It is not possible to ascertain precisely how often events of the sort depicted in these anecdotes occur. We are discovering more incidents in which officers who have used force against citizens are caught in misrepresentations exposed by dashboard or body cameras. Misrepresentations in police reports are akin to the courtroom misrepresentations academic literature labels as “testilying.”29 A tally of criminal prosecutions or convictions would yield a substantial undercount. Prosecutors are normally reluctant to bring charges against members of the police force, which is, after all, a cooperating state agency’s personnel.30 And jurors may often be unwilling to disbelieve officers.31 Civil actions do not supply a reliable count either, because we do not know what fraction of unlawful force cases produce litigation. Despite the absence of reliable statistics, there is evidence that the rate of improper police uses of force is, by any measure, high. For example, a 2002 Bureau of Justice national survey estimated that police used force against individuals on 664,460 instances annually, and that approximately 587,000 of those usages were perceived as excessive.32 The gap between the perceived incidence of inappropriate force and the exceedingly low rates of discipline raises an obvious question of whether officers always truthfully report use-of-force events to superiors and internal investigators.33 Contemporary examples and history give reason to think not. In its final report on corruption and abuse in New York City police, for example, the 1994 Mollen Commission found that “perjury and falsifications are serious problems in law enforcement that, though not condoned, are ignored.”34 We think this likely remains so.

Police officers’ tendency to use unlawful or unconstitutional force is a practice embedded within a complex assemblage of institutional practices, including cultural norms within police departments, state, and federal employment regulation (including due process protections for state employees), and constitutional rules.35 No one reform or rule will likely prove a panacea. Instead, changes in police practice are likely to emerge from concatenated factors. These likely include shifting public attitudes toward police and the objects of iterative police scrutiny; changes in police culture and leadership; and reinvigorated or reimagined forms of judicial supervision.

Our aim in this Article is to pick out one element of this array of barriers to reform. Having homed in upon a particular element of police practices, we examine how federal and state tort and contract law—more specifically, municipal liability pursuant to 42 U.S.C. § 1983 and the state public policy doctrine of contract invalidity—might be leveraged to improve the efficacy of post-incident investigations, and thereby reduce the rate at which excessive force is used. We focus on one quite specific element of the law. Pursuant to state law or to a collective bargaining agreement (CBA) between a police union and a municipality, individual officers often benefit from provisions that restrict the timing and manner by which investigators interview or interrogate police, which we call “interrogation buffers.” One specific buffer is a mandatory period of delay between a use-of-force incident and the ensuing investigation or interrogation, what we call a “delay privilege.”36 By mandating a delay of either hours or days between the notice the internal investigator first gives to an officer of the existence of the investigation and the need for an interview (or sometimes the occurrence of the incident giving rise to the investigation) and the actual first interview, such provisions predictably obstruct investigations, diminishing the likelihood that culpable officers are subject to effective internal investigation, and correspondingly increase the probability that officers disposed to use excessive force will continue to work the streets without reprimand or supplemental supervision.

Because there is reason to believe that a small minority of officers are responsible for a disproportionate percentage of excessive and unlawful uses of force—i.e., they are repeat players—interrogation buffers have the predictable effect of enabling future incidents of unconstitutional police force. Indeed, part of our argument is that their sole effect and sole plausible justification is to raise the cost of identifying and sanctioning violent cops. Delay privileges, as a result, predictably yield future violations of citizens’ constitutional and legal rights. The core of our argument is that civil litigation in state and federal court can be employed to challenge at least those delay privileges contained in CBAs. These judicial challenges could succeed at invalidating such provisions. Even if they fall short of that goal, they would raise popular awareness of the existence and adverse effect of these provisions, creating political pressure for their elimination. To be clear, we are under no illusion that the civil suits envisaged here would single-handedly solve this problem. Our more modest goal is to illuminate a contribution that ordinary civil law—and those willing to invoke it—can make.

The two main Parts of the essay advance complementary descriptive and legal claims. To begin, Part II documents the existence of interrogation buffers and, specifically, delay privileges under state law and CBAs, and then scrutinizes their potential normative justifications—an analysis that flows inexorably into a tally of their likely consequences. Part III then turns to potential state and federal law litigation responses available when a delay privilege is contained in a CBA.

II.    Delay Privileges in Law and in Practice

We start by describing the emergence and content of delay privileges. We then explore their potential justifications and likely consequences. We focus our discussion on the relationship of delay privileges with the unconstitutional or unlawful use of force37 —which we take as the paradigmatic form of unlawful police action that should be identified, sanctioned, and stopped.

A.    The Law of Interrogation Buffers and Delay Privileges

Over the past half century, police organizations across the United States have acquired generous procedural privileges for disciplinary investigations into the misconduct of individual officers, frequently under the rubric of a “Law Enforcement Officers’ Bill of Rights.” These procedural rules concern civil investigations that can result in employment sanctions, such as suspension, demotion, or dismissal.38 Most of these rules specify restraints on interviewing or interrogating police. Examples include an officer’s right to certain information before any interview; a right to representation by counsel or union official during the interview; the right to take periodic breaks and limit the interview session to a reasonable time; and the right to limit questioning to a single interviewer, who must not be “abusive” of the interviewee for any session. All of these rights serve to create a “buffer” of protection for the officer, to make the examination less threatening to the officer’s interests.39

Yet another type of procedural privilege is the primary focus of this Article: rules favoring investigative delay. For many police departments, the applicable state or local law or a CBA mandates a delay before the interview or interrogation of an officer (as witness or accused) about an incident that could lead to employment discipline of that or another officer. Sometimes the delay is “mandatory” because it does not appear the officer is empowered to waive the period of delay. In these cases, the procedure specifies how many days must pass after the incident being investigated or after the investigator provides notice to the officer of the demand for an interview. Alternatively, sometimes the applicable law or CBA grants the officer an option of delaying the interview or interrogation, usually for the purpose of obtaining representation. Of all interrogation buffers, we focus on delay privileges because we believe it to be the most socially harmful, the least defensible, and so the most vulnerable to legal attack.40

One source of these police privileges is the Law Enforcement Officers Bill of Rights (LEOBR) codified in a state statute. According to Kevin Kennan and Samuel Walker’s leading historical account, Maryland and Florida were the first to enact such a statute in 1974.41 By our count, twenty states now have LEOBRs or other statutes that regulate how administrative investigators can interview or interrogate police officers (as targets or witnesses) in a disciplinary investigation into police misconduct42 such as a police shooting or another potentially unlawful use of force.43 Of the twenty, nine states (Delaware, Illinois, Kentucky, Louisiana, Maryland, Minnesota, Nevada, Rhode Island, and Texas) include a period of delay, ranging from an open-ended “reasonable time” to a specified amount, ranging from forty-eight hours to thirty days, to obtain representation.44

But delay privileges are pervasive beyond these nine states, although it is difficult to determine their precise scope, which turns on local law and contracts. In the other forty-one states, many municipal ordinances provide interrogation buffers, including delay privileges. Consider New York City’s regulation:

A member of the Police Department who is the subject of a complaint shall be given two business days notice prior to the date of an interview, to obtain and consult with counsel. A member of the Police Department who is a witness in an investigation of a complaint shall be given . . . up to two business days, to confer with counsel.45

In other cases, the rules exist within the administrative regulations of the police department. For example, Boston’s department rules allow a delay, to obtain representation, until 10:00 a.m. the day following notice of interrogation.46

Local legislative or administrative rules matter even where a municipality resides in a state with a LEOBR statute because the local rule can supplement the statute. For example, the Illinois statute only says that the investigators must provide the target officer a “reasonable time” to obtain representation.47 In Chicago, by contrast, the superintendent of police has issued a “general order” that defines this delay privilege. For investigations by the Independent Police Review Authority (IPRA) into police shootings in particular: “The shooting member(s) will be required to give their statement . . . no earlier than 24 hours after the shooting incident.”48 Because the interviews are to occur between 6:00 a.m. and 6:00 p.m., the general order notes that the delay may be longer than twenty-four hours.49 Further, the order provides:

When a shooting member advances a claim that they are unable to provide a statement within the time period specified . . . IPRA will handle these claims on an individual basis, making a reasonable inquiry into the reasons for the member’s claim, and accepting at face value all good faith claims of a member’s inability to provide a statement.50

This is not, in other words, a scheme likely to elicit timely disclosures about the use of deadly police force.

Finally, outside of state and local laws or administrative regulations, there are CBAs. These too exist inside and outside of the nine states that create delay privileges by statute. Michigan, for example, has no LEOBR statute, but the CBA between Detroit and the police union provides for forty-eight hours written notice prior to an interview “except in cases of emergency.”51 We have found many such agreements outside the nine states. We are not able to estimate the number of police covered by each of these regimes. But we are certain that a substantial number of all American police are entitled to a specific delay before being interviewed about their use of unlawful force against civilians.52 Quite a few other agreements create a presumptive delay subject to override for some form of necessity, but we do not know how strong the presumption works in practice.53 For that matter, even when there is no formal right of delay in any law or CBA, we do not know that there is not an informal norm of delay.

These legislative, municipal, and negotiated rules, it is important to note, serve a purpose notwithstanding the fact that the Fifth Amendment’s right against self-incrimination allows officers to resist interrogation. Correlatively, eliminating these limitations would have a significant effect even if Fifth Amendment protections continue to apply. Officers benefit from two rights derived from the Fifth Amendment. First, they are entitled to the suite of Miranda warnings required during a custodial interrogation, and can object at a criminal trial if those warnings are not delivered.54 We are, however, concerned here with the full gamut of potential corrective mechanisms, including internal discipline. Even if criminal prosecution is off the table,55 internal investigations remain available as a way of detecting officers inclined to use excessive force.

Second, in the course of routine administrative proceedings, officers can be required, under threat of discipline including job-loss, to answer questions that investigators pose.56 Once statements are compelled, however, the Court in Garrity v. New Jersey57 held that neither such evidence nor its fruit can be employed in a criminal prosecution absent a grant of immunity to the officer.58 Garrity poses a large hurdle to criminal prosecution, and may be vulnerable to criticism on that basis. But it does not prevent administrative investigators from identifying officers who are inclined to use violence, and either disciplining them, retraining them, or reassigning them. The fact that criminal prosecution has been impeded, in short, does not also mean that administrative measures to reduce the expected level of future unconstitutional force are also unavailable. The Fifth Amendment, in short, does not render delay privileges nugatory—as the police unions and officers who lobbied hard for their adoption surely understood.

B.    Potential Justifications for Delay Privileges

At first blush, delay privileges present a puzzle: What might justify the installation of delays between incident and investigation? We consider here the possible functional explanations for these provisions, and conclude that none of them has much merit. The absence of functional underpinnings, though, implicates a further question: If they have no justification, why do they exist?

Start with a simple fact: To the best of our knowledge, police officers are never routinely encouraged to delay or suspend the acquisition of information related to a potential crime in any other circumstance. To the contrary, the conventional wisdom in law enforcement is to move promptly to interview witnesses. The truism for homicide investigations is that the first forty-eight hours are crucial because the odds of finding a lead fall dramatically after that time, but the police view timeliness as imperative for all investigations. For example, 1999 and 2003 guides to the collection and use of eyewitness evidence prepared by the Department of Justice, under different Attorneys General, advise: “Plan to conduct the interview as soon as the witness is physically and emotionally capable.”59 The more recent guide explains: “Once the witness is capable, any delay in conducting the interview should be minimized as there will be less detailed information available as time goes on.”60 The Justice Department also expresses the concern that witnesses may be exposed to media reports or other witnesses, either of which might contaminate their recollections.61 The National Forensic Science Technology Center similarly advises that “[t]he timely interviewing of witnesses is crucial to the solution of a crime. Witnesses to crimes must be identified, secured, [and] questioned at the scene.”62

This common-sense advice is consistent with the science of eyewitness evidence. A 2003 summary of the empirical literature on eyewitness testimony identified several elements of investigative design that could influence accuracy, but did not include the length of time between an incident and subsequent recollections.63 Laboratory investigations of strategies to elicit accurate information likewise do not suggest a need for delay. Instead, they counsel for the use of varied strings of questions and for approaching the incident from diverse perspectives, developed with careful attention to the specific characteristics of the interviewee.64 In short, neither the general practice nor the social science of eyewitness testimony justifies delay privileges.

What then warrants the differential treatment of police? Some advocates for delay privileges contend that the impact on stress upon memory retention justifies a delay of up to forty-eight hours after an incident before engaging in questioning.65 The empirical foundations of this claim have been challenged.66 We agree with the critics: there is little or no evidence to support the underlying claim and no rationale for the differential treatment of police and non-police witnesses.

On the one hand, there is some evidence that stress degrades the formation of memories among eyewitnesses.67 This evidence, however, does not support delay privileges absent evidence that memory somehow improves over time—i.e., that somehow information initially not retained in the memory returns over time.68 Further, this evidence does not explain why police officers should be treated differently from other witnesses.69 Instead, to the extent that police are more likely than other suspects to be trained to manage stress, and more likely to have encountered violent situations before, it would seem that the current dispensation in which police alone benefit from a delay has matters backwards.

On the other hand, evidence of a relationship between sleep and memory formation is “fragmentary.”70 Some studies suggest that certain aspects of a memory are recalled better after a period of sleep.71 But these studies’ implications are equivocal. They show that memory can either get better or worse depending on the precise nature of the information.72 Even if empirical studies gave delay privileges firmer evidentiary footing, it would still be a puzzle why those most likely to be trained to attend to detail should benefit from an asymmetrical timing rule.

Perhaps, however, the explanation for interrogation barriers generally and delay privileges in particular does not sound in psychology or law, but in the social role of police. One justification for interrogation barriers, identified by Keenan and Walker, turned on “the unique role of the police in American society, a role that includes their authority to use coercive force, both lethal and non-lethal, and their capacity to deprive people of their liberty.”73 The original demand for such bespoke procedural protections was supported by pointing to then-recent changes in criminal procedure law that (it was argued) had demoralized officers.74 Most sympathetically interpreted,75 these original justifications home in upon the distinctively central role of force in the police role: the frequency of needful force, and the fact that police must interact disproportionately with a more violent, disputatious selection of the public, means that police are uniquely vulnerable to ex post criticism based on the ordinary performance of their duties. Given their role, and the means typically used to accomplish it, police are more likely than other state employees to be subject to false or unreasonable civilian complaints. On this view, the chilling of police force that results from effectual post hoc investigations has social costs in terms of officer safety and crime control. Police accordingly require an extra margin of appreciation to perform their tasks adequately.

For three reasons, this argument cannot do load-bearing work. First, the constitutional criminal procedure rules that precipitated initial calls for additional procedural protections for officers already reflect a measure of deference to officers.76 The remedial avenues available to subjects of arguably unconstitutional force, including tort remedies against individual officers and the suppression of evidence secured in violation of the Fourth Amendment, also typically turn on a showing that a state actor has not only violated the Constitution, but also violated an established and especially clear constitutional rule previously announced by a federal court.77 If both the substantive constitutional law of policing and the remedial apparatus available for constitutional rights-holders build in large tolerance for officer error, there is no clear justification for a third layer of protection in the form of investigatory delay.

Second, measures like delay privileges, unlike the margins of appreciation already built into constitutional rights and remedies, do not distinguish between plainly unconstitutional uses of force and other incidents. Instead, they indiscriminately raise the cost of identifying—and, correlatively, deterring—all wrongful conduct. Even absent other sources of discretion for front-line officers, institutional mechanisms that equally protect the vicious and the well intentioned seem unwise. In addition, delay privileges diminish the availability of information to police managers about the front-line performance of their officers.78 Contemporary police forces typically lack institutional mechanisms to gather information about the distribution and frequency of illegal force.79 In short, a measure that restricts the flow of information to police managers about whether that discretion has been wisely, or unlawfully, used has large negative externalities.

Finally, we once more find it implausible to conclude that police officers are more vulnerable, and so more inclined to offer false inculpatory statements in the course of station-house interrogations, than the modal criminal suspect.80 Although we think this point relatively self-evident, we find Miranda a useful lens for thinking about whether the police-suspect case presents particular difficulties. Miranda does not distinguish between the sophisticated party attuned to the law and police practices, and the vulnerable naïf possessing only inchoate awareness of his or her rights, and how best to vindicate them. One justification for this conclusive presumption of compulsion in custodial interrogations across sophisticated and naïve suspects is a persistent epistemic asymmetry of information between both kinds of suspects and the police: However knowledgeable a suspect, he or she “needs to know whether the police are prepared to respect [his or her] rights.”81 But if there is a discrete subset of suspects for whom this particular informational asymmetry is less likely to arise, it is members of the police. The latter are more likely to know whether officers in a given police force are likely to be rule-followers or are likely to “lose their tempers” in ways that lead to discomfort, pain, or abuse.82 The justifications for prophylactic measures in the interrogation context, in short, are weaker when the suspect is a member of the police than in other instances. So if there is a justification for delay privileges from intimidation or coercion-based concerns, it suggests the current dispensation is exactly backward.

Because we are lawyers, we cannot resist making a point about the law here. More specifically, constitutional doctrine, just like common sense intuition, lists against delay privileges. Two elements of constitutional criminal procedure in effect subsidize expeditious inquiry while penalizing delays in interrogation. First, the Sixth Amendment’s Confrontation Clause ordinarily bars the use of testimonial statements uttered outside the trial and without the opportunity for cross-examination.83 When police respond to “an ongoing emergency” by eliciting information, however, that information can be used at trial.84 In the Court’s recent treatment of scene-of-the-crime investigations, Michigan v. Bryant,85 Detroit police investigating a shooting found the victim lying in a gas-station parking lot.86 By asking him “what happened,” they elicited information about where the shooting had happened and the identity of the perpetrator.87 The Court declined to treat the statements as testimonial, and therefore covered by the Sixth Amendment, on the basis of an “[o]bjectiv[e]” assessment of “the primary purpose of the interrogation by examining the statements and actions of all participants.”88 With Bryant in pocket, police have an incentive to exploit the ‘exigent’ character of a situation for epistemic gain, safe in the knowledge that the Court’s objective standard precludes inquiry into their felt intentions.

Second, the Fifth Amendment’s right against compelled confessions requires a set of four warnings before custodial interrogation of a suspect.89 The Miranda warnings are perceived as a burden by law enforcement, who have developed strategies for avoiding them or minimizing their effect.90 Most relevant here, the Miranda regime is more likely to apply after a prolonged sequence of police-suspect interactions than at the inception of an encounter. In the first few moments when police confront a suspect, they can often (if not always) elicit information without prefatory warnings for use at trial or for later interrogations. For example, in Oregon v. Elstad,91 police sought out a teenage suspect in his parents’ home to arrest him pursuant to a warrant.92 Before executing the warrant, one officer engaged in conversation that elicited inculpatory information that was used in a subsequent interrogation immediately preceded by Miranda warnings.93 A divided Supreme Court held that information from the second interrogation could be introduced at trial.94 Quite apart from its treatment of the second interrogation, Elstad reflects the fact that interactions with police in the moments after an encounter begins will often rank as noncustodial, such that Miranda does not apply.95 In addition, even when a suspect has been arrested, and therefore is unquestionably in custody for Fifth Amendment purposes, the Miranda regime is suspended in the teeth of exigency such as imminent risk to public safety from a nearby concealed weapon.96 We think that many drug-related arrests will implicate a risk of proximate concealed weapons, such that some questioning at the scene of a seizure can proceed absent Miranda warnings.

The Fifth and Sixth Amendments, in short, operate as a “tax” in delay in interrogations.97 Like any tax, the resulting rules have behavioral effects on the regulated parties. They elicit a shift from later to earlier modes of interrogation. To be sure, this (perhaps inadvertent) element of pre-trial criminal procedure may be inadvisable. The Supreme Court may be the final word on constitutional criminal procedure, but finality is not the same as wisdom. The law’s inclination nevertheless raises question about the function of bespoke delay privileges for police suspects that press in the other temporal direction. However much we see police and their political allies resisting it, there is no evidence that police want the Miranda regime inverted such that it applies to early but not later interrogations on the ground that the latter are more desirable or trustworthy.98 Indeed, we have not been able to locate any criticism of current doctrine on the ground that its temporal logic is perverse or undesirable. The absence of criticism or concern about the law’s tax on early interrogations is yet one more reason to doubt that delay privileges are intrinsically desirable.

In summary, we are skeptical that any functional justifications for delay privileges exist in general, and are further skeptical that the rule’s observed tailoring to police suspects alone can be justified on policy terms. This is not to say that the rule lacks tangible effects. Instead, it is to the likely consequences that we now turn.

C.    The Consequences of Delay Privileges

The central consequence of delay privileges, as we already intimated, is to raise the costs of investigating potential police misconduct, including illegal or unconstitutional uses of force that are our focus here. This frictional effect on effectual investigation arises from two elements of the policy.

First, mandatory delays before investigative interviews disproportionately benefit officers who need time to prepare for an interview. All else being equal, we think that officers who committed no wrongdoing are less in need of time to compose a convincing narrative of an event for the simple reason that they do not necessarily need one. In contrast, officers who violated a constitutional or legal rule benefit from the delay not only by developing their own plausible story, but also by coordinating with other officers.99 That is, interrogation buffers directly enable norms of investigative non-cooperation among officers.

It is true that delay may not be necessary in some instances because intra-police norms against inculpation of fellow officers are already powerful. But in the many instances in which officers have to coordinate a story, they must identify a single story to tell. There is an academic urban legend that nicely illustrates our point. Four undergraduates fail to show up for an exam. They later tell the professor they had a flat tire on the way back from a weekend trip. The professor appears to accept the excuse and allows the four students to take a make-up exam, though in separate rooms. They are surprised to discover that one of the questions on the exam is: “which tire?”100 If the students were telling the truth, they would offer the same highly salient answer. But, in a group, a lie requires coordination of the details of the false account.101

To return to one of the cases discussed in our introduction, it would hardly have availed Officer Tensing if his colleagues had claimed that Mr. Dubose had pulled a firearm, or threatened his life, as opposed to dragging him with his vehicle.102 This fact is perfectly typical for the kinds of details that matter in these kinds of cases: Did the suspect throw a punch or display a weapon? Was the displayed weapon a knife or a gun? Was it in his right hand or his left? Or was he disobeying orders and suspiciously reaching for something, in which case, did he move his hand towards the glove compartment of his car or under his seat? Successful testilying and its station-house kindred depend on the kind of collective coordination that requires time to invent, communicate, and agree on a narrative. Without the communication enabled by delay privileges, such coordination may be impossible or substantially more difficult.103

The need for coordination does not end there. Those who lie about violent encounters must also worry about consistency between those accounts and the physical evidence. It will not work for officers to construct and recount the same narrative if that narrative is flatly contradicted by powerful external evidence, such as a video. If you have seen the video, you choose a better lie, one that turns on movements and events just outside the camera coverage.104 More generally, one can better invent a plausible narrative if one already knows what facts the other side can prove. The importance of contradictory physical evidence is one reason that police seek to interrogate criminal subjects quickly, to lock in their story before they or their lawyers can ascertain what facts a prosecutor can prove.105 Yet the privilege of delay gives police time to determine some of the fixed facts around which any false narrative must be adapted.

We do not mean to suggest that every officer is immediately willing to coordinate on a fabrication. To the contrary, we included two introductory anecdotes concerning police whistleblowers, Detective Crystal and whoever alerted journalists to the existence of a dashboard video showing the killing of Laquan McDonald. The fact that there are principled and public-minded police officers, however, is part of the problem with investigatory delay: The latter facilitates the coercion of these officers. A misbehaving officer who engages in excessive force might find that a principled officer witnessing his or her conduct is, at first, willing to break the code of silence and be a whistleblower (or in the term internal to some police culture, a “rat”). If the investigation proceeds quickly enough, there would be no time for the misbehaving officer to influence the witness. Such an officer/witness might also find a rationalization for telling the truth despite the code of silence: without knowing what other officers would say, or what the physical evidence would show, they face personal risks from telling a lie that would be exposed and, by being uncoordinated, such a lie would not help the misbehaving officer. Here, delay offers the misbehaving officer the time not only to circulate a false narrative, but also to bring pressure on those who seem reluctant to go along, to remind them of the consequences of contradicting the false narrative. And having received the official false narrative, the potential whistleblower no longer has the excuse for telling the truth that any lie will be detected.

Furthermore, there is a dynamic effect from delay privileges that reduces the likelihood that unconstitutional or unlawful force will be revealed or sanctioned. Although the magnitude of this effect is likely to be smaller than the above causal mechanisms, it is nonetheless worth explicating. Internal investigations are an important source of information for police management about the rate and distribution of problematic force. Delay privileges suppress that information even within the police department. Absent strong leadership from police management, it is likely that the signal of low rates of problematic police force will be treated as evidence that police comply with relevant limits on the use of force. This undermines the case for allocating investigative and institutional resources to such problems. Over time, a feedback mechanism may arise, wherein a weak signal of unlawful force incidence conduces to weaker investigations, which in turn generate even weaker signals of underlying force-related misconduct, and so on.106

This dynamic effect may be compounded by a further mechanism: Internal investigations tending to deny wrongdoing may hinder, discourage, or raise the costs of tort lawsuits filed against either individual officers, which tend to be indemnified by the relevant municipality,107 or suits filed against the municipality itself.108 If this occurs, it is not just that a municipality economizes on scarce resources for maintaining public order by shrinking its investments in internal accountability mechanisms. It also minimizes exposure to future tort settlements.109 In practice, police departments already “limit rather than promote information availability,” failing to create or publicize information that is necessary for sensible regulation of policing.110 As Rachel Harmon persuasively argues, “significant counter-incentives” often lead police management to “underinvest in research that could improve policy” because a department would not internalize the gains from such a policy.111 Where information might make the police worse off, even while making society better off, there is all the more ground for anticipating underinvestment in data collection and analysis. In short, there is every reason to believe delay privileges not only reduce the rate of successful identification of unlawful or unconstitutional police force when first introduced, but also catalyze a feedback mechanism in which a diminishing proportion of such incidents are identified on a year-by-year basis.

Something of this sort is, indeed, probably the intended effect of delay privileges. According to Keenan and Walker, interrogation buffers for police officers—of which delay privileges are but one element—date back to the early 1970s, and were a consequence of rejuvenated police unions’ mobilization at both federal and local levels.112 As of 2007, thirty-eight percent of local police departments, which employ sixty-six percent of all officers, were unionized.113 From their inception, police unions have protested what they perceive as a “lack of internal civil and constitutional rights for officers being investigated for misfeasance and malfeasance.”114 That effect must be understood in light of the two documented effects that unionization has upon police practices. First, it tends to result in policy changes that increase the scope of rank-and-file officers to use deadly force and authorized weaponry.115 Second, unionized officers also “tend to make a high number of discretionary arrests and to encourage the civilians with whom they interact to invoke the formal criminal justice process rather than take advantage of informal, or formal but non-criminal, alternatives.”116 Although the causal mechanisms behind these effects are not well understood,117 their net effect is clear: unionization is associated with both more frequent contacts between police and citizenry, and also a higher likelihood of force during such contacts.

From the rank-and-file’s perspective, procedural protections may be tightly related to these changes on the ground. In effect, greater procedural protections operate to lower the expected cost of a preferred, high-intensity style of policing. They are a kind of insurance against the inevitable conflicts and complaints that emerge from that preferred policing style. To be sure, like many other forms of insurance that pool together heterogeneous individuals, procedural protections for police are of greater value to those with higher expected future costs. Here, the disproportionate beneficiaries are members of a police department most disposed to use violence at the legal border. But such a cross-subsidy, which seems to us implicit in many union-derived procedural protections, might be deemed acceptable for a majority of rank-and-file given risk-averse preferences and a predicable stream of complaints perceived as false or unfounded.118 Moreover, from a public choice perspective,119 delay privileges may be an especially attractive benefit for police. Unlike wage increases, which are highly visible and which sap municipal budgets, procedural protections are a way for a concentrated, well-organized interest group such as the police to extract a subsidy from a diffused and poorly organized public without attracting public scrutiny and hence political opposition.120

In conclusion, the most likely consequence of delay privileges—indeed, perhaps their only likely effect—is to raise the cost across-the-board of investigations into police uses of force. This immediately makes unlawful force harder to identify. Over time, they obscure the incidence of such force, while corroding accountability mechanisms. We discern no justification for these effects, even given the real hazard of baseless or malicious civilian complaints. Unlike rational responses to the latter problem, delay privileges critically fail to sort between justified and unjustified concerns. With magnificent indifference, they protect the vicious and the good-hearted alike.

III.  Litigating Delay Privileges: State and Federal Options

The interest-group politics of delay privileges means they are unlikely to dissipate without exogenous intervention. As the powerful backlash to the “Black Lives Matter” movement suggests,121 efforts to regulate the police trigger considerable political and popular resistance.122 We therefore turn in this Part to legal strategies under current state and federal law that might be used to chip away at the practice. At the threshold, we underscore that the legal theories explored in this part provide no miraculous remedy. As Samuel Walker has noted, “reforms [have] generally failed to develop mechanisms to ensure their own continuity.”123 Litigation in particular does not necessarily elicit learning by police departments.124 Nevertheless, where an interest group has entrenched a regressive and harmful rule that democratic processes alone are unlikely to dislodge, litigation can influence police behavior on the ground,125 draw attention to a rule among the public, illuminate the weak justifications for the rule, and render the rule impossible to defend. In that chastened spirit, we offer two pathways for challenging delay privileges in courts under state and federal law.

A.    Municipal Liability Under 42 U.S.C. § 1983

Section 1983 of Title 42 provides a federal tort remedy against state actors who violate the Constitution or federal law. Enacted in 1871,126 § 1983 was functionally a dead letter until the Supreme Court’s 1961 ruling in Monroe v. Pape,127 which rejected a definition of the statute limited to acts condoned by state law.128 Monroe, however, construed § 1983 not to extend to municipalities. The Court overruled that holding in its 1978 decision Monell v. New York Department of Social Services of New York.129 Importantly, municipal liability suits differ from individual liability suits not only because relief is limited to compensatory damages,130 but also because qualified immunity is unavailable.131

Monell emphatically rejected vicarious liability theories of municipal liability.132 Instead, a § 1983 plaintiff seeking compensation must show that (1) “some official policy,” (2) “‘cause[d]’ an employee to violate another’s constitutional rights.”133 Much litigation has ensued over precisely what counts as a “policy,” and which species of city officials have authority to promulgate one.134 However, a theory of municipal liability trained on delay privileges need not expend much effort on this question. It is likely that in most cases, the official agreeing to the CBA has the requisite authority to create a “policy” for § 1983 purposes, and that he or she has done so by entering into such an agreement when it regiments the workplace. This requirement of municipal policy-making—which will easily be met in the suits we envisage—means that the theory of liability outlined here will not avail when a delay privilege is adopted as a matter of state law. In such cases, no municipal policy-maker exists to be held liable.

A challenge to a municipality’s delay privilege pursuant to § 1983 would proceed along the following lines. The plaintiff would be a person who had been subjected to unconstitutional police force. This plaintiff would file suit not just against the officer in question but also against the municipality asserting violations of the Fourth Amendment.135 A central factual claim in the complaint would be that the defendant officer had previously employed unconstitutional or illegal force in incidents that had been investigated using a delay privilege under a CBA, and that those incidents had resulted in no disciplinary action, training, or other species of course-correction by senior officers. The incidence of such fact-patterns, of course, is an empirical question. Not every incident of allegedly unconstitutional police force will meet these characteristics. But we think that it is likely that a nontrivial number of cases will track this pattern given the evidence that the illegal force incidents tend to be highly concentrated among a few individuals within a police force.136 This is the conclusion, for example, of our colleague Craig Futterman’s analysis of historical police misconduct allegation data from the Chicago Police Department.137 It was also the case in the McDonald shooting.

Given clear evidence of a policy, an unconstitutional act that resulted in compensable harm, the central question under this theory of municipal liability will be one of causation. The Court’s municipal liability jurisprudence underscores plaintiffs’ need to present clear evidence that a policy was the “moving force” behind the violation.138 For example, in cases alleging municipal liability based on improper training or supervision, liability is established only by showing a constitutional deficiency was “so obvious, and the inadequacy so likely to result in the violation of constitutional rights” that policymakers could be said to be “deliberately indifferent.”139 In cases alleging a failure to adequately screen employees, the Court explained that liability attached only “where adequate scrutiny of the applicant’s background would lead a reasonable policymaker to conclude that the plainly obvious consequence of the decision to hire the applicant would be the deprivation of a third party’s federally protected right.”140 In one recent case, the Court overturned a jury verdict that a district attorney should have perceived a need for additional training based on repeated failures to disclose exculpatory material because it was not “so predictable that failing to train the prosecutors amounted to conscious disregard for defendants’ Brady rights.”141 In short, a plaintiff must demonstrate to a high degree of certainty that continued deployment of an officer would have had results of the kind that in fact occurred.

We think that this standard can be satisfied in a challenge to a delay privilege. The argument would go as follows. As we set forth in Part II.B, the only justification for, and the only expected consequence of, such policies is to raise the cost of investigating police misconduct across-the-board. The predictable effect of such policies is to occlude culpability in an instant case, while also eroding over time the informational basis for managers to make informed decisions about how to train, equip, and sanction their officers. Because these dynamics further the fiscal interests of a department by reducing exposure to tort suits, the long-term effect will be systematic failures of sound management and predictable patterns of iterative illegal force by repeat offenders.

The theory of causation in these circumstances would parallel the causal mechanism proved up in cases concerning municipal failures to investigate or discipline. For example, in a 1992 case involving allegations of false arrest and sexual assault by an officer, the Eighth Circuit Court of Appeals found that the North Little Rock, Arkansas, police department had “implemented a policy of avoiding, ignoring, and covering up complaints of physical and sexual abuse.”142 Systematic failures of investigation by civilian complaint review boards have also generated liability.143 Some years before the Eighth Circuit decision, the Second Circuit held that a failure to investigate prior complaints can evince deliberate indifference sufficient to establish causation for Monell purposes.144 In a sense, these cases rest on the same theory of liability as decisions in which police are found liable for abandoning suspects in high-crime areas, who are subsequently harmed by third parties.145 Both kinds of cases concern a decision taken by an official tortfeasor, where the harm that is likely to arise is, in a sense, inflicted by a third-party. In neither line of cases does the fact that a third party (either the officer or a criminal) intervenes to harm the plaintiff break the causal chain.

A recent case involving allegations of unconstitutional force by Denver, Colorado, police, provides a useful example of how causation would be demonstrated in a case involving delay privileges.146 In Ortega v. City and County of Denver,147 four plaintiffs alleged the unconstitutional use of force against them by police officers.148 Among their theories of liability, the plaintiffs alleged that the City of Denver had “failed to discipline officers for the use of excessive force” in the past.149 Based on plaintiffs’ evidence that “Denver had a ‘systematic problem’ of officers not being held accountable for their uses of force,” and a finding that the individual defendant officers had previously been the object of insufficiently investigated complaints, the district court rejected the City’s motion for summary judgment on that claim.150 Of relevance here, the district court also accepted a separate claim that “Denver has a custom of tolerating its police officers’ ‘code of silence.’”151 Under the plaintiffs’ theory of the case, the failure to discipline and the code of silence allegations provided complementary causal mechanisms: both counted as “policies” for Monell purposes. It seems likely that in many cases, as in Ortega, there will be multiple causal pathways linking a municipality to an alleged tort in this fashion. It is possible to imagine a federal court analyzing each of those strands separately, and concluding that each strand failed to satisfy the “moving force” threshold for causation identified by the Supreme Court if treated in isolation. This kind of pixelated approach analysis, which the Ortega Court eschewed, is implausible. If there are several elements of city policy that cause a constitutional tort, and the city is responsible for adopting each one, there is no clear reason why the city should be able to escape responsibility for its deliberate choice to combine those policies by insisting on a seriatim analysis of causation.

At the same time, we do not mean to suggest that all relevant precedent counsels in favor of liability. Other decisions concerning failure to discipline allegations, in contrast, have been dismissed on the ground that “the failure of a police department to discipline in a specific instance . . . [is rarely] an adequate basis for municipal liability under Monell.”152 It is possible, though, that a plaintiff could successfully distinguish the latter decisions on the ground that they were based on a failure to show the causal effect of a policy, a showing that could be made in challenges to delay privileges.

In summary, the theory of municipal liability for delay privileges falls into a legible, long-established strand of municipal liability jurisprudence. We cannot be certain the theory will prevail in any given case. Nevertheless, we think that such litigation, given the right factual record and a judge not disposed against § 1983 litigation, could prevail, and thereby influence municipal practice.

B.    State Law

1.    A Contractual Provision Void as Against Public Policy

We also offer an alternative litigation strategy for state courts—one seeking to invalidate the investigatory delay provisions in police CBAs, although not state or local law (challenges to state or local law will depend on the federal constitutional claims just described). The National Labor Relations Act does not apply to CBAs involving public employees.153 CBA provisions are therefore subject to state law, including state contract law (as modified by the state’s labor law). As contracts, there is a plausible argument that a right to delay investigation is contrary to public policy and therefore void.154 The doctrine of void-as-against-public-policy is old and established in state contract law. The starting point, of course, is that contracts are void if part of the promised performance is illegal, as in a contract to murder, but the doctrine stretches far beyond simple illegality and has been applied in a wide variety of settings.155

One of the most pervasive examples of contractual invalidity across different jurisdictions, based on public policy, involves judicial creation of a whistleblower exception to the default rule of at-will employment.156 In many states with at-will employment, the courts have nonetheless created for employees a contractual right not to be fired for revealing the employer’s legal violations—i.e., whistleblowing. The basic idea is that public policy favors detection of law violations, including the statutory, administrative, or common law violations the employer commits. In many states, the public policy doctrine requires a clear and explicit statement of policy from the legislature. But it is not difficult for courts to find this in the statutes whose violation is the subject of whistleblowing. If the state has a statute prohibiting bribery of public officials or the dumping of toxic waste, then there is a public policy to detect bribery and dumping that is impeded by recognizing the employer’s contractual right to fire employees reporting such behavior. As one court explained, “[i]t is public policy . . . everywhere to encourage the disclosure of criminal activity.”157 In general, then, a contractual provision that allows employees to be fired for the public service of disclosing legal violations is contrary to public policy and void.

The same public policy condemns investigatory delay provisions. The police officer willing to testify to another officer’s use of excessive force is a potential whistleblower. For reasons given above, investigatory delay suppresses whistleblowing of this variety. First, delay gives misbehaving officers time to pressure the potential whistleblower, by describing past retaliations against officers who break the code of silence and “rat” on fellow officers and by promising to harass the current witness in a similar way in the future should he expose them. Second, the time to coordinate on an official false narrative deprives the potential whistleblower of the excuse that he had for implicating a fellow officer by telling the truth—that before there was a coordinated false narrative, he had no other way to avoid being caught in a lie than to tell the truth. Finally, if the potential whistleblower expects multiple other witnesses to swear to the false narrative, he or she may rationally decide that it is pointless to tell the truth when one won’t be believed anyway (or not sufficiently to cause the officer to be disciplined). The public policy of detecting crime is inconsistent with a collective bargaining provision that has no other substantial purpose but to conceal wrongdoing and deter whistleblowing.

Obviously, the argument we are making requires an extension of existing precedent. In the cases we are discussing, the municipal employer does not threaten to fire or discipline the whistleblower for telling the truth, but merely engages in a time-limited omission—the temporary failure to interview police witnesses. Yet this is a conceptual distinction without a policy difference. From the law of constructive discharge, when an employer knowingly tolerates sufficient co-worker harassment to cause resignation, it is tantamount to a dismissal.158 Scholarship on policing amply shows that police officers enforce the code of silence with social ostracism, refusals to answer calls for back-up, denials of promotion, reassignments to less desirable postings, and threats of violence. When such harassment forces an officer to resign, the resignation should be understood as constructive discharge that itself violates whistleblowers protections.159 Yet the well-known reality is that it remains difficult or impossible to prove constructive discharge in any individual case, given the code of silence. If the problem is largely intractable, the least the court should do is not to make it worse by tolerating a contractual provision that makes the constructive discharge of whistleblowers easier (by giving the misbehaving officers time to coordinate on a false narrative and to communicate threats of harassment to potential whistleblowers), thus deterring whistleblowers from coming forward. The protection of police whistleblowers demands that the municipality retain the power of immediate investigation of police wrongdoing, including the prompt interview or interrogation of police suspects or witnesses. Contractual provisions to the contrary are void as against public policy.160

As an analogy, consider that a municipality could not lawfully promise in a CBA with a police union not to discipline any officer for any use of excessive force against citizens. Such a provision would not only be against public policy; it would shock the conscience. When police force is excessive, it is beyond the bounds of a criminal defense and therefore constitutes a crime, either assault or homicide. As a matter of expressed policy, every American government aims to discourage violent crime. It would deeply compromise this policy for an employer to promise not to discipline an employee for violent crimes committed in the course of employment. For that reason, we think that such a contract provision would be void regardless of the employer’s identity. This does not mean the employer is legally obligated to discipline the employee; rather, the employee could not enforce a promise not to discipline him or her for a violent crime committed in the course of employment. But a no-discipline-for-crime provision would be even more adverse to public policy if the employer is a government that cloaks employees with legal authority and empowers them to use physical force against others. The public policy against violent crime demands that governments retain the discretion to discipline or fire those individuals who abuse their official power by committing violent crime in the course of their public duties.

If governments are not permitted to forego entirely the power to discipline or fire officers using excessive force, they should also not be permitted to forego the power substantially but indirectly by hamstringing their own disciplinary investigations. A promise of investigatory delay is obviously not exactly the same as a promise not to discipline for the simple reason that, ultimately, some officers may be disciplined given the former, but not the latter. But the imperfect protection of investigatory delay should not detract from the analogy. A police union could not hope to negotiate an overt shield from discipline given the political reality. But they can hope to achieve a substantial amount of disciplinary immunity though an obscure privilege of investigatory delay, a seemingly innocuous detail within a group of more defensible interrogation buffers. A contractual provision that significantly strengthens the code of silence and impedes the disciplining of officers who engage in excessive force is void as against public policy.161

We have found no case directly on point, but there are cases in parallel settings that provide support for this reasoning. In Illinois State Police v. Fraternal Order of Police Troopers Lodge No. 41,162 two state troopers arbitrated separate complaints that an internal-affairs officer had interviewed them without the benefit of the interrogation buffers provided by the LEOBR provisions of their CBA, such as twenty-four hour notice of “the nature of the complaint and pertinent facts alleged” and the right to have counsel present.163 Internal affairs replied that the interrogations were for purposes of a standard criminal investigation (concerning insurance fraud and sexual abuse) and not investigations concerning job-related discipline. The arbitrator nonetheless ruled in the state troopers’ favor. This ruling suggests that police officers not only benefit from procedural rights for disciplinary investigations related to on-the-job misconduct, but also, by virtue of a CBA, are entitled to special rules for all interrogations by fellow troopers. These include advance notice of the alleged facts and the presence of counsel for all criminal wrongdoing, even entirely unrelated to official duties. For example, an officer alleged to have committed spousal abuse would benefit from the CBA’s protection even in the course of a domestic violence investigation of purely private conduct. Given that defense counsel invariably advise clients to refuse to talk to police, the arbitrator’s construction of the CBA could be understood to confer on troopers a blanket freedom from criminal interrogation by their own police force.

The Appellate Court of Illinois vacated the arbitrator’s award on two grounds. First, it held that the CBA by its terms simply did not apply to criminal interrogations, but only to interrogations directed at employment discipline.164 The second reason—more relevant here—was that impeding a criminal investigation in this manner was void as against public policy.165 The Appellate Court reached this conclusion even though there is a strong presumption in favor of judicial enforcement of arbitral decisions.166 Given this presumption, it is quite possible for a court to find a contractual provision void against public policy, but then to defer nonetheless to an arbitral decision enforcing the provision. But the Appellate Court in Illinois State Police was not willing to defer in this fashion. Instead, it held that there was an “established,” “basic,” and “unquestionabl[e]” public policy in favor of effective law enforcement.167 Further, the court ruled that interrogation buffers “contravene[]” the internal affairs division’s “ability to investigate crimes.”168 The division “presented the undisputed testimony that the element of surprise is very important in conducting criminal interrogation.”169 The court concluded that the arbitrator’s award violated the public policy of effective law enforcement.

Of course, Illinois is one of the states that incorporates a delay privilege into state statutory law, so one cannot credibly claim that CBAs instantiating such rights violate the state’s public policy. Indeed, the defendants in Illinois State Police offered various state statutes of this sort in support of the arbiter’s award and the court distinguished them solely on the ground that the investigation at issue was criminal, not civil. Yet in forty-one states, delay privileges are not embodied in state law. In those jurisdictions, one can ask courts to give the reasoning of Illinois State Police its full logical scope. If interrogation buffers impede the investigation of criminal wrongs for the purpose of criminal prosecution, they also necessarily impede the investigation of criminal wrongs for the purpose of job discipline (excessive police force being a crime—assault or homicide). Even if one stubbornly resisted this logical implication for some interrogation buffers, the logic is overwhelming when applied to delay privileges. Regardless of why police crime is investigated—to impose job-related discipline or criminal sanctions—the investigator’s success depends on proceeding expeditiously, moving before their official narrative takes hold, and retaining “the element of surprise.” A CBA provision delaying the investigation is void.

2.    Standing to Bring Public Policy Arguments Against CBAs

Ordinarily, the only actors with standing to raise the invalidity of a contractual provision are the parties to the contract. The city or state that entered into a collective bargaining agreement with a police union clearly has standing to challenge the provision, should it be motivated to disregard the delay privilege and promptly interview police officers involved in police misconduct, as targets or witnesses. The invalidity would be a defense to any effort to enforce the contractual limit.

We realize, though, that a city or state that signs an agreement containing that provision may have no incentive to challenge it. That presents a serious obstacle to reform. The standing barrier is a matter of state law, so we are only generalizing when we say that citizens, even those with claims against police, will lack standing to challenge the contract on the grounds we set out. It is worth investigating in any particular jurisdiction.

There is, however, frequently an alternative. The state attorney general would plausibly have parens patrie standing to represent the public’s interest in asserting the invalidity of police contract provisions that violate public policy. The Third Circuit allowed such standing in Pennsylvania v. Porter170 when the state attorney general sued a police officer and members of the city council for violating state law by infringing the constitutional rights of state citizens. The attorney general argued that the police officer had engaged in a pattern or practice of misconduct by mistreating, threatening, harassing, and illegally detaining and arresting citizens in the municipality of Millvale with the acquiescence and ratification of the Mayor and Council.171 Under settled doctrine, a state has standing to sue “only when its sovereign or quasi-sovereign interests are implicated and it is not merely litigating as a volunteer the personal claims of its citizens.”172 The Court held that the state attorney general could bring an action against a municipality and its officers because it had “significant sovereign interests of its own in the prevention of future violations of constitutional rights of its citizens.”173

Similarly, in New York v. Town of Wallkill,174 the court permitted New York’s attorney general to bring a parens patriae claim against a town for knowingly allowing its police to target and stop women and critics of the department in violation of federal and state law.175 The court reasoned that Supreme Court precedent made it difficult for plaintiffs seeking injunctive or declaratory relief against law enforcement officers,176 and that many victims were deterred from reporting police conduct due to fear of retribution.177 Because of the futility of private rights of action against law enforcement, the court concluded that a parens patriae action “is the best and, likely, the only, method for obtaining the sort of systemic, forward-looking relief that the State here seeks.”178 The court applied Pennsylvania v. Porter’s holding that states are entitled to parens patriae standing when “vital” interests, such as “the prevention of a pattern of police lawlessness, under official sanction,” are at stake.179 By the same reasoning, a state attorney general should have standing to challenge a provision in a police collective bargaining agreement that systematically protects officers guilty of the use of excessive force.

IV. Conclusion

We do not think that delay privileges are the most important problem on the police reform agenda at present. Far from it. But we do think that their reform is, in a sense, an example of low-hanging fruit. They provide an especially attractive target for actors outside the police interested in seeking reform. In contrast to highly visible, often textually memorialized delay privileges, many police practices that are causally related to higher incidence rates of unlawful police force are hard to discern from outside the institutional bounds of a department. Training, for example, may be inadequate or even invoke antiquated, stereotypical ways of framing criminality. Criteria for hiring may select officers least likely to take care in using force. And rules regulating promotions may select on the basis of, and then create, incentives for less careful husbanding of coercive authority. In each of these cases, the relevant police practice may not be visible from outside the police department. By contrast, delay privileges are highly visible, and their impact on rates of unlawful force is likely to be both significant and consequential. Their challenge and ouster, we emphasize, is the beginning and not the end of police reform. Much remains to be done to understand and change the institutional circumstances of constitutional violations. The reform of delay privileges should not be an excuse for failing to release more data about coercive practices, hiring and promotion criteria, disciplinary mechanisms, and complaint rates—or for remaining inert when new forms of transparency reveal further causal mechanisms sustaining current rates of unlawful force. Nevertheless, our modest aim in this Article has been to identify one immediate platform for reform that need not await disclosures by police force. We hope, therefore, that delay privileges are challenged in litigation that brings both public attention and financial penalties to a practice that lacks substantial normative justification, and that has only regressive and unjust effects.


  • 1Richard Pérez-Peña, University of Cincinnati Officer Indicted in Shooting Death of Samuel Dubose, N.Y. Times (July 29, 2015), http://www.nytimes.com/2015/07/30/us/university-of-cincinnati-officer-indicted-in-shooting-death-of-motorist.html (https://perma.cc/V96J-GAZD).
  • 2Id.
  • 3Id.
  • 4Id.
  • 5Elisha Fieldstadt, Samuel Dubose Shooting: No Charges for Two Officers Who Responded, NBC News (July 31, 2015), http://www.nbcnews.com/news/us-news/cincinnati-traffic-stop-shooting-no-charges-two-officers-who-responded-n401971 (https://perma.cc/Q2Z5-5QV4).
  • 6See, e.g., Ty Burr, In the Video of Sandra Bland’s Arrest, the Tape Doesn’t Lie, Boston Globe (July 23, 2015), https://www.bostonglobe.com/arts/2015/07/23/video-bland-arrest-tape-doesn-lie/fqhkLb94dREEcjTNb4EcxM/story.html (https://perma.cc/CFC7-25GX] (noting “flagrant contradiction” between the video of an arrest and the officer’s subsequent account); Michael S. Schmidt & Matt Apuzzo, South Carolina Officer Is Charged with Murder of Walter Scott, N.Y. Times (Apr. 7, 2015), http://www.nytimes.com/2015/04/08/us/south-carolina-officer-is-charged-with-murder-in-black-mans-death.html (https://perma.cc/8Y95-RUGT). The problem is hardly a new one. See, e.g., Fox Butterfield, 2 Boston Police Officers Charged with Lying in Killing of Officer, N.Y. Times (Feb. 16, 1990), http://www.nytimes.com/1990/02/16/us/2-boston-police-officers-charged-with-lying-in-killing-of-officer.html (https://perma.cc/45GH-KMNF).
  • 7See, e.g., Mike Lowe, Four Police Officers Charged with Perjury in Glenview Drug Case, WGN-TV (June 8, 2015), http://wgntv.com/2015/06/08/four-police-officers-charged-with-perjury-in-glenview-drug-case/ (https://perma.cc/KN6Q-QKTK).
  • 8See Colin Campbell, Batts Defends Officer Who Claims He Was Intimidated for Testifying, Balt. Sun (Feb. 27, 2014), http://www.baltimoresun.com/news/maryland/crime/blog /bs-md-ci-police-support-crystal-coming-forward-20140227-story.html (https://perma.cc/6DFK-WSRF]; Ian Duncan, Officer Found Rat on Car After Assisting Probe of Colleagues, Balt. Sun (Feb. 19, 2014), http://www.baltimoresun.com/news/maryland/crime/blog /bs-md-ci-officer-witness-retaliation-20140219-story.html (http://perma.cc/CYQ6-A3QQ]; Ed Krayewski, Ex-Baltimore Cop Alleges Retaliation for Reporting Police Brutality, Reason (Dec. 26, 2014), https://reason.com/
    blog /2014/12/26/ex-baltimore-cop-alleges-retaliation-for (https://perma.cc/Y4VB-L5K4).
  • 9Sebastian Murdock, ‘Rat Cop’ Shunned from Baltimore Police Department After Reporting Police Brutality, Huffington Post, (June 16, 2015), http://www.huffingtonpost.com/2015/06/16/baltimore-joe-crystal_n_7582374.html (https://perma.cc/L3Z5-Q76M).
  • 10He has subsequently sued the police force. See Edward Ericson Jr., Whistleblower Cop Joseph Crystal Recalls His Battles with Baltimore’s Blue Wall of Silence, City Paper (Jan. 20, 2015), http://www.citypaper.com/news/mobtownbeat/bcp-whistleblower-cop-joseph-crystal-recalls-his-battles-with-the-baltimore-police-departments-blue-wall-20150120-story.html (https://perma.cc /AV2T-C8YA]; Krayewski, supra note 8. Officer Williams was ultimately convicted of assault and hindering an investigation and, after saying that he did nothing wrong and would change nothing he did that day, received a sentence of forty-five days. Jayne Miller, Baltimore Officer Sentenced in Drug Suspect’s Assault: Officer Anthony Williams Gets 45 Days in Jail for Assaulting Man, WBAL-TV (Apr. 17, 2014), http://www.wbaltv.com/news/baltimore-officer-connected-to-witness-intimidation-case-sentenced/25514064 (https://perma.cc/F469-ZZ2E). The other officer, Sgt. Marinos Gialamas, was convicted and sentenced to probation. See Ericson, supra note 10.
  • 11No Bail for Chicago Officer Jason Van Dyke, N.Y. Times (Nov. 24, 2015), http://www.nytimes.com/interactive/2015/11/24/us/video-chicago-laquan-mcdonald-jason-van-dyke-shooting.html (https://perma.cc/JE2H-UU2X).
  • 12Police Accountability Task Force, Recommendations for Reform: Restoring Trust Between the Chicago Police and the Communities They Serve 2 (Apr. 2016), https://chicagopatf.org /wp-content /uploads/2016/04/PATF_Final_Report_4_13_16-1.pdf (https://perma.cc/4SLF-MV4K] [hereinafter Task Force Report].
  • 13Id. at 2–3.
  • 14Marissa Bailey, Police Shoot, Kill Knife-Wielding Teen on South Side, CBS Chi. (Oct. 21, 2014), http://chicago.cbslocal.com/2014/10/21/police-shoot-kill-knife-wielding-teen-on-south-side/ (https://perma.cc/VT6C-B7BU]; Quinn Ford, Cops: Boy, 17, Fatally Shot by Officer After Refusing to Drop Knife, Chi. Trib. (Oct. 21, 2014), http://www.chicagotribune.com/news/local/breaking /chi-chicago-shootings-violence-20141021-story.html (https://perma.cc/SRF4-A9ZZ] (including references to the report from police spokesman Pat Camden); Teen Shot, Killed by Police Officer on Chicago’s Southwest Side, NBC Chicago (Oct. 21, 2014), http://www.nbcchicago.com/news/local/chicago-shooting-4100-south-karlov-279884562.html#ixzz3sihriX2S (https://perma.cc/6NRU-MCTN).
  • 15Jason Meisner & Jeremy Gorner, Cop Who Shot Teen 16 Times Has History of Citizen Complaints, Chi. Trib. (Apr. 25, 2015), http://www.chicagotribune.com/news/ct-police-shooting-16-shots-04—-met-20150425-story.html#nt=related-content (https://perma.cc/F3FF-Z9YP).
  • 16According to Craig Futterman, Chicago police officers shoot about fifty civilians per year. See Curtis Black, How Chicago Tried to Cover Up a Police Execution, Chi. Rep. (Nov. 24, 2015), http://chicagoreporter.com/how-chicago-tried-to-cover-up-a-police-execution/ (https://perma.cc/B7YM-2TBH).
  • 17Id.
  • 18Id.
  • 19Id.
  • 20Meisner & Gorner, supra note 15.
  • 21See Carol Marin & Don Moseley, Missing Minutes from Security Video Raises Questions, NBC Chi. (May 26, 2015), http://www.nbcchicago.com/investigations/laquan-mcdonald-investigation-305105631.html (https://perma.cc/9EHV-DABW).
  • 22See Jeremy Borden, How a Little-known, Uber-driving Freelancer Brought the Lawsuit that Forced Chicago to Release a Police Shooting Video, Colum. Journalism Rev. (Nov. 25, 2015), http://www.cjr.org /united_states_project /brandon_smith_chicago_police_laquan_mcdonald.php (https://perma.cc/GS4H-CXB7).
  • 23Task Force Report, supra note 12, at 2.
  • 24Id.
  • 25Jason Meisner et al., Chicago Releases Dash-cam Video of Fatal Shooting After Cop Charged with Murder, Chi. Trib. (Nov. 24, 2015), http://www.chicagotribune.com/news/local/breaking /ct-chicago-cop-shooting-video-laquan-mcdonald-charges-20151124-story.html (http://perma.cc/3WNG-CT7N).
  • 26Rahm Decries Police ‘Code Of Silence’ Ahead Of Morning Speech to City Council, Chicagoist (Dec. 9, 2015), http://chicagoist.com/2015/12/09/rahm_gives_preview_of_city_council.php (https://perma.cc/89G9-KGKN).
  • 27State’s Attorney Anita Alvarez claimed that her investigation was proceeding without regard to the public pressure or the video’s disclosure. There was also an FBI investigation. So it is difficult to know what would have happened without the pressure brought by Kalven, Futterman, and Smith. But we think one does not have to be too jaded to think that political will to proceed would have been missing absent their efforts.
  • 28See Black, supra note 16.
  • 29Milton Mollen et al., Report of the Comm’n to Investigate Allegations of Police Corruption and the Anti-Corruption Procedures of the Police Dep’t 36 (July 7, 1994). This term has generated extensive scholarly commentary. See generally Gabriel J. Chin & Scott C. Wells, The “Blue Wall of Silence” as Evidence of Bias and Motive to Lie: A New Approach to Police Perjury, 59 U. Pitt. L. Rev. 233 (1998); Morgan Cloud, Judges, “Testilying,” and the Constitution, 69 S. Cal. L. Rev. 1341 (1996); Christopher Slobogin, Testilying: Police Perjury and What to Do About It, 67, U. Col. L. Rev. 1037 (1996). The misrepresentation we are concerned with here, however, does not necessarily occur under oath.
  • 30See Chin & Wells, supra note 29, at 261–64.
  • 31See Allison Patton, The Endless Cycle of Abuse: Why 42 U.S.C. § 1983 Is Ineffective in Deterring Police Brutality, 44 Hastings L.J. 753, 754 (1993).
  • 32Matthew R. Durose et al., Contacts Between Police and the Public, 2005, Bureau of Just. Statistics, U.S. Dep’t of Just. 8 (Apr. 2007), http://www.bjs.gov/content /pub/pdf/cpp05.pdf (https://perma.cc/32ED-B2GF). The number of police-caused fatalities is subject to considerable dispute, but the best estimates for 2015 puts the figure at greater than 1,000. Jon Swaine & Oliver Laughland, Number of People Killed by U.S. Police at 1000 After Oakland Shooting, Guardian (Nov. 16, 2015), http://www.theguardian.com/us-news/2015/nov/16/the-counted-killed-by-police-1000 (https://perma.cc/8WZS-SNJR).
  • 33Our home town Chicago, for example, had more fatal shootings by police than any other American city between 2010 and 2014, but internal investigations found only two of the 409 shootings during that period unjustified. Ben Austen, Chicago After Laquan McDonald, N.Y. Times (Apr. 24, 2016), http://www.nytimes.com/2016/04/24/magazine/chicago-after-laquan-mcdonald.html?_r=0 (https://perma.cc/QVP7-G4BC). Might there be too little police violence? The argument for this position might be that increasing police activity leads to diminished crime rates, which on balance are socially beneficial. See, e.g., Steven D. Levitt, Understanding Why Crime Fell in the 1990s: Four Factors that Explain the Decline and Six that Do Not, 18 J. Econ. Persp. 163, 176–77 (2004) (finding that an increased number of officers on the street during the 1990s can account for 1/5 to 1/10 of the decline in crime during that period). No study of which we are aware, however, shows that the use of unconstitutional force tends to reduce crime rates. Rather, empirical studies of even marginally less intrusive measures that are more likely to influence crime rates—for instance, stop-and-frisk practices—do not support the conclusion that more police activity (even of a less coercive form) necessarily translates into less crime. See Richard Rosenfeld & Robert Fornango, The Impact of Police Stops on Precinct Robbery and Burglary Rates in New York City, 20032010, 31 Just. Q. 96, 97–98 (2012) (finding no evidence that increased use of stop and frisk reduced burglary and robbery rates). To the contrary, there is evidence that crime control goals can be furthered even as the volume of police activity is reduced. See generally David Weisburd et al., Could Innovations in Policing Have Contributed to the New York City Crime Drop Even in a Period of Declining Police Strength?: The Case of Stop, Question and Frisk as a Hot Spots Policing Strategy, 31 Just. Q. 129 (2014) (arguing that increased use of “hot spot” policing during periods of downsizing in the New York City police department led to reductions in crime rates). 
  • 34Mollen et al., supra note 29, at 41.
  • 35See Rachel A. Harmon, The Problem of Policing, 110 Mich. L. Rev. 761, 763–64 (2012) (pointing out that legal scholars have “[neglected] the full web of federal, state, and local laws that govern the police outside of the context of criminal investigations” in analyzing police regulation).
  • 36In their seminal treatment of the topic, Kevin M. Keenan and Samuel Walker call these “waiting period” provisions. See Kevin M. Keenan & Samuel Walker, An Impediment to Police Accountability? An Analysis of Statutory Law Enforcement Officer’s Bill of Rights, 14 B.U. Pub. Int. L.J. 185, 212–14 (2005).
  • 37See infra note 135, for our discussion about the constitutional law of excessive force under the Fourth Amendment.
  • 38The procedures do not apply to criminal investigations, as by prosecutors or homicide detectives, though we note such investigations of police are exceedingly rare. When a criminal investigation does occur, it frequently begins after a period of delay and is governed by other constitutional rights.
  • 39Some procedural rights are broader than the issue of interrogation. For example, some rules prohibit an investigation unless the citizen complaint is sworn to in an affidavit. In this paper we do not take issue with any of these provisions except for those that grant officers a privilege of investigatory delay. For a suggestion that procedural rights granted to police should be a model for the procedural protections of citizen, see Kate Levine, Police Suspects, 116 Colum. L. Rev. 1197 (2016).
  • 40Another detail we bracket here is the trigger for an internal investigation: If such inquiries occur only after a citizen complaint is lodged, they will be far less effective than if they occur automatically whenever force is known to have been used by officers.
  • 41Keenan & Walker, supra note 36, at 197.
  • 42See Ar. Code. Ann § 14-52-303(3); Az. Stat. § 38-1104(B) (2); Cal. Gov. Code § 3300-3313; 11 Del. Code § 9200; Fl. Stat. Ann. § 112.532(1); Ill. Comp. Stat. § 725; Iowa Code Ann. § 80F.1(5); Ky. Rev. Stat. § 15.520; La. Rev. Stat. Ann. § 40:2531; Md. Code, Public Safety § 3-104; Minn. Stat. Ann. § 626.89; Nev. Rev. Stat. § 289.010-289.120; N.M. Stat. Ann. § 29-14-4 to 29-14-5; Or. Rev. Stat. § 236.360; R.I. Gen. Laws § 42-28.6.2; Tn. Code Ann. § 38-8-302; Tx. Local Govt. § 143.123(f); Va. Code. Ann. § 9.1-500-507; W. Va. Code § 8-14A-2; Wis. Stat. § 164.02.
  • 43Our discussion focuses on police shootings because this is an important category of state coercion warranting regulation. We do not mean to suggest that other uses of police force, including physical restraints or Tasers, do not raise the same concerns.
  • 44See Del. Code § 9200(c) (9) (proceedings shall be suspended “for a period of time” if an officer under questioning desires representation “until such time as the officer can obtain the representative requested if reasonably available”); Ill. Comp. Stat. § 725/3.9 (if an officer under investigation requests representation, before or during the interrogation, “no interrogation shall proceed until reasonable time and opportunity are provided the officer to obtain counsel”); Ky. Rev. Stat. § 15.520(1) (c) (“No police officer shall be subjected to interrogation in a departmental matter involving alleged misconduct on his or her part, until forty-eight (48) hours have expired from the time the request for interrogation is made to the accused officer, in writing.”); La. Rev. Stat. Ann. § 40:2531(B) (4) (b) (an officer as target or witness “shall be granted up to thirty days to secure . . . representation, during which time all questioning shall be suspended.”); Md. Code, Public Safety § 3-104(j) (2) (i)&(ii) (stating that, if an officer being interrogated desires representation, the interrogation “shall be suspended for a period not exceeding 10 days until representation is obtained,” which may be further extended by the chief “for good cause”); Minn. Stat. Ann. § 626.89(9) (if any officer giving a formal statement requests representation, before or during an interrogation, “no formal statement may be taken until a reasonable opportunity is provided for the officer to obtain the presence of the attorney or the union representative.”); Nev. Rev. Stat. § 289.060(1) (as a general rule, “a law enforcement agency shall, not later than 48 hours before any interrogation or hearing is held . . . provide a written notice to the peace officer who is the subject of the investigation” and any “officer who serves as a witness during an interview must be allowed a reasonable opportunity to arrange for” representation); R.I. Gen. Laws § 42-28.6-2(9) (if the officer “under investigation” desires representation, “[t]he interrogation shall be suspended for a reasonable time until representation can be obtained.”); Tx. Local Govt. § 143.123(f ) (“Not later than the 48th hour before the hour on which an investigator begins to interrogate a fire fighter or police officer regarding an allegation based on a complaint, affidavit, or statement, the investigator shall give the fire fighter or police officer a copy of the affidavit, complaint, or statement.”).
  • 45Rules of the City of New York Title 34-A, Civilian Complaint Review Board, New York City § 1-24(b), http://rules.cityofnewyork.us/content /section-1-24-conduct-interviews (https://perma.cc/F8VY-WLCL). As with state laws, some municipal regulations create procedural privileges that do not include investigatory delay. See, e.g., Hurricane, West Virginia, Code of Ordinances § 133.07(b).
  • 46This delay rule is the least favorable to police that we have found, creating a limit rather than a right: “the interrogation may not be postponed for purpose of counsel past 10 a.m. of the day following the notification of interrogation.” See Rule 109: Discipline Procedure, Amended, Boston Police Dep’t § 51(i) (Apr. 12, 1983), http://static1.squarespace.com/static /5086f19ce4b0ad16ff15598d/t /52af5e69e4b0dbce9d22a6f9/1387224681015/Rule+109.pdf (https://perma.cc/
  • 47See Ill. Comp. Stat. § 725/3.9.
  • 48Chicago Police Department General Order G08-01-01, Department Member’s Bill of Rights, Chi. Police Dep’t § 3-C (Mar. 11, 2013), http://www.chicagopolice.org /2013MayDirectives/data/a7a57be2-12cc274e-6a512-cc28-0757e267c9e275a8.html (https://perma.cc/P57E-LVFJ).
  • 49Id. at § 3-D.
  • 50Id. at § 3-H (emphasis added).
  • 51See Master Agreement Between the City of Detroit and the Detroit Police Officers’ Association, City of Detroit § 16(J) (3) (Oct. 1, 2014), http://www.detroitpoa.com/Uploads/2014___2019_Contract.pdf (https://perma.cc/SN3S-NNXE).
  • 52See, e.g., Agreement by and Between the City of Seattle and Seattle Police Officers’ Guild, Dep’t of Labor § 3.6(F ) (2) (2005), http://www.dol.gov/Olms/regs/compliance/cba/olms/public /810635.pdf (https://perma.cc/85DA-Z39S] (providing that the officer under investigation be notified of an interview at least five days before it occurs, though the Chief of Police may reduce this to a one-day notice for “exigent circumstances.”); Agreement Between Fraternal Order of Police Montgomery County Lodge 35, Inc. and Montgomery County Government, Dep’t of Labor § 43(D) (1) (b) (2004), http://www.dol.gov/Olms/regs/compliance/cba/olms/public /810571.pdf (https://perma.cc/4Y5H-7D2Y] (providing an officer twenty-four hours to secure representation before an interview); Agreement Between the City of Omaha, Nebraska and the Omaha Police Union Local No. 101 § 18a(c) (2001), http://www.irle.berkeley.edu/library/pdf/0042.pdf (https://perma.cc/M2S7-X8AX] (providing an officer with no less than twenty-four hours notice before the interview, except if the complaint alleges on-duty intoxication or drug use). Some CBAs are with state-wide unions. See, e.g., Agreement: The State of Florida and International Union of Police Associations: Law Enforcement Bargaining Unit, Dep’t of Labor Art. 7 § 1(c) (2003), http://www.dol.gov/Olms/regs/compliance/cba/olms/public /800315.pdf (https://perma.cc/TQ42-YDCM] (providing that an officer may upon request have twenty-four hours to obtain and consult with counsel before an interview); Commonwealth of Massachusetts and the State Police Association of Massachusetts (S.P.A.M.), Dep’t of Labor Art. 27 § 1(8) (Oct. 1999), http://www.dol.gov/Olms/regs/compliance/cba/pdf/cbrp1161.pdf (https://perma.cc/48DF-HT2G] (providing that an officer may delay the interview for the purpose of conferring with counsel until 10:00 a.m. the day following the day the notice is given); Labor Contract Between the State of Nebraska and the Law Enforcement Bargaining Unit Represented by the State Law Enforcement Bargaining Council (SLEBC) § (May 18, 2001), http://www.irle.berkeley.edu/library/pdf/0365.pdf (https://perma.cc/23Y9-3GG9] (providing an officer the opportunity to review the case file at least forty-eight hours before an interview).
  • 53See, e.g., Collective Bargaining Agreement Between Anchorage Police Department Employees’ Association and Municipality of Anchorage, Dep’t of Labor Art. 5 § 1(G) (4) (Dec. 16, 2008), http://www.dol.gov/Olms/regs/compliance/cba/public /6040_pub.pdf (https://perma.cc/5EMH-Y2KA] (providing an officer twenty-four hours notice of an interview, unless the Chief of Police believes notice would jeopardize the investigation); Memorandum of Understanding Between City of San Diego and San Diego Police Officers Association, Dep’t of Labor Art. 41 § 4(A) (July 1, 2003) http://www.dol.gov/Olms/regs/compliance/cba/olms/public /820238.pdf (https://perma.cc/U9WH-WFNE] (providing that an officer “shall receive at least three days notice prior to an interrogation except where such delay will hamper the gathering of evidence”); Memorandum of Understanding Between County of Orange and Orange County Employees Association, 2001–2004, Dep’t of Labor Art. 8 § 8(B) (Oct. 4, 2001), http://www.dol.gov/Olms/regs/compliance/cba/pdf/cbrp1584.pdf [perma.cc/G86P-6FU3] (“All investigatory meetings shall be scheduled to allow an employee a reasonable opportunity to obtain representation. Whenever practicable, such notice shall be given at least three (3) working days prior to the meeting.”); Michigan, State of and Michigan State Police Troopers Association, Inc., Dep’t of Labor Art. 7 § 5(b) (2003), http://www.dol.gov/Olms/regs/compliance/cba/olms/public /800411.pdf (https://perma.cc/XG5H-LQCJ] (“Wherever practicable, the employee shall be given 48 hours advance notice of the questioning.”).
  • 54Miranda v. Arizona, 384 U.S. 436, 447–48 (1966).
  • 55There is no bar to using statements obtained in violation of Miranda for internal investigations. Indeed, there is some authority for the proposition that no Fifth Amendment violation occurs until a statement is introduced at trial. See Chavez v. Martinez, 538 U.S. 760, 768 (2003) (plurality opinion) (“We fail to see how Martinez was any more ‘compelled in any criminal case to be a witness against himself ’ than an immunized witness forced to testify on pain of contempt.”).
  • 56Steven D. Clymer, Compelled Statements from Police Officers and Garrity Immunity, 76 N.Y.U. L. Rev. 1309, 1315 (2001).
  • 57385 U.S. 493 (1967).
  • 58Id. at 499; see also Lefkowitz v. Turley, 414 U.S. 70, 82 (1973) (stating that “Garrity specifically prohibited [an effort] to compel testimony that had not been immunized.”). Garrity immunity encompasses use and derivative use immunity. See Clymer, supra note 56, at 1319.
  • 59See, e.g., David L. Carter & Jeremy G. Carter, Effective Police Homicide Investigations: Evidence from Seven Cities with High Clearance Rates, Ind. U.-Purdue U. Indianapolis 19 (2015), https://scholarworks.iupui.edu/bitstream/handle/1805/6160/Carter_2015_effective.pdf ?sequence=1 (https://perma.cc/KV6M-SLHK] (“It is generally recognized that the critical time interval for identifying suspects, witnesses and evidence is the first 48 hours after a homicide is reported.”); Dominic Casciani, When the Murder Trail Goes Cold, BBC News (May 6, 2010), http://news.bbc.co.uk/2/hi/uk_news/magazine/8662635.stm (https://perma.cc/J5KV-BNP9] (quoting retired detective chief inspector Will O’Reilly as stating: “‘Time is of the essence at the beginning or [sic] any investigation,’ he says. ‘We talk about a golden hour where there is an opportunity to make progress—but that can often be pushed further to the first 24 hours and then the first 48.’”).
  • 60U.S. Dep’t of Justice, Nat’l Inst. of Just., Eyewitness Evidence: A Trainer’s Manual for Law Enforcement 13 (2003) [hereinafter The 2003 Manual]; U.S. Dep’t of Just., Off. of Just. Programs, Eyewitness Evidence: A Guide for Law Enforcement 21 (1999); see also The 2003 Manual, supra at 21 (“During or as soon as reasonably possible after the interview, the investigator should . . . [d]ocument the witness’s statements . . . and ask the witness if there is anything he/she wishes to change.”) (emphasis added); accord Thomas F. Adams et al., Police Field Operations 301 (8th ed. 2014); Jill Leovy, Ghettoside: A True Story of Murder in America 134 (2015) (reporting a belief that the “first forty-eight hours” are crucial in a murder investigation).
  • 61See The 2003 Manual, supra note 60, at 12.
  • 62Crime Scene Investigation: A Guide for Law Enforcement, Nat’l Forensic Sci. Tech. Ctr. 8 (2013), http://www.nist.gov/forensics/upload/Crime-Scene-Investigation.pdf (https://perma.cc/8LC9-QRP5).
  • 63Gary L. Wells & Elizabeth A. Olson, Eyewitness Testimony, 54 Ann. Rev. Psychol. 277, 285 (2003).
  • 64R. Edward Geiselman & Ronald P. Fisher, The Cognitive Interview: An Innovative Technique for Questioning Witnesses of Crime, J. Pol. & Crim. Psychol. 2, 3–4 (1988).
  • 65See Samuel Walker, Police Union Contract “Waiting Periods” for Misconduct Investigations Not Supported by Scientific Evidence, 4 (2015), http://samuelwalker.net /wp-content /uploads/2015/06/48HourSciencepdf.pdf (https://perma.cc/ZGQ4-H839).
  • 66Id. at 3–4; see also Geoffrey P. Alpert et al., Working Toward the Truth in Officer-Involved Shootings, FBI Law Enforcement Bull., https://leb.fbi.gov/2012/may/working-towards-the-trust-in-officer-involved-shootings-memory-stress-and-time (https://perma.cc/L2QR-Y86P] (suggesting officer recall is better with immediate, rather than delayed, questioning).
  • 67See Kenneth A. Deffenbacher et al., A Meta-Analytic Review of the Effects of High Stress on Eyewitness Memory, 28 L. & Hum. Behav. 687, 687 (2004).
  • 68For example, advocates of the police delay privilege have cited Jeffrey M. Ellenbogen et al., Interfering with Theories of Sleep and Memory: Sleep, Declarative Memory, and Associative Interference, 16 Current Biology 1290–94 (2006). The experiments reported provide some evidence that, given a twelve-hour delay between learning and memory testing (where the memory task is made more difficult by “associative interference”), memory is improved by the subject having slept. But that is not evidence that delay (and sleep) improves memory compared to a situation of no delay.
  • 69Walker, supra note 65, at 5.
  • 70Pierre Maquet, The Role of Sleep in Learning and Memory, 294 Science 1048, 1048 (2001) (note that this study is about the effect of sleep on memory generally, without regard to stress).
  • 71See generally Björn Rasch & Jan Born, About Sleep’s Role in Memory, 93 Physiological Rev. 681 (2013); Jeffrey M. Ellenbogen, et al., The Role of Sleep in Declarative Memory Consolidation: Passive, Permissive, Active or None?, 16 Current Opinion in Neurobiology 716 (2006).
  • 72See, e.g., Jessica D. Payne et al., Sleep Preferentially Enhances Memory for Emotional Components of Scenes, 19 Psychol. Sci. 781, 781–88 (2008) (finding that sleep leads to better memory of “negative objects,” but decaying recall of background details). Further, eyewitnesses can err not only if they “forget both important and unimportant details” but also “through commission, incorporating new information that may seemingly sharpen the experience or shape the narrative for the witness in ways that, even if factually correct, become no longer a veridical report of that eyewitness’s original experience.” Nancy K. Steblay, Scientific Advances in Eyewitness Identification Evidence, 41 Wm. Mitchell L. Rev. 1090, 1107 (2015). Indeed, one (somewhat dated) metastudy found that the rates of both positive and false confirmatory identifications slope upward with time. Peter N. Shapiro & Steven Penrod, Meta-Analysis of Facial Identification Studies, 100 Psychol. Bull. 139, 151 (1986).
  • 73Keenan & Walker, supra note 36, at 186.
  • 74Kami Chavis Simmons, The Politics of Policing: Ensuring Stakeholder Collaboration in the Federal Reform of Local Law Enforcement Agencies, 98 J. Crim. L. & Criminology 489, 522–23 (2008) (reporting police unions’ arguments that the “lack of due process rights has led to a loss of officer confidence in the disciplinary process and a loss of morale”).
  • 75To be clear, we do not necessarily endorse this sympathetic reading. The shifting procedural landscape of the 1960s and 1970s reflected racial dynamics in ways that may well have induced negative emotions among police. But even setting aside issues of race, an explanation of police anxieties about new procedural protections might reflect their anxiety about “relative position[s]” rather than any actual concern about material consequences. Richard H. McAdams, Relative Preferences, 102 Yale L.J. 1, 3 (1992). Early sociological accounts of police found that “the job they shared isolated them from the community and threatened their collective sense of status,” and motivated the creation of “norms [endorsing] the selective use of illegal violence against suspects, for example, and [forbidding] officers from testifying against each other.” David A. Sklansky, Police and Democracy, 103 Mich. L. Rev. 1699, 1732 (2005). Legal constraints on the use of force constitute a material loss of status for police, on this view—and law enforcement bills of rights were mechanism for clawing back a modicum of this prestige.
  • 76Hence, the Court often explicitly provides ample breathing room for “officer safety” in the design of constitutional rules. See, e.g., Pennsylvania v. Mimms, 434 U.S. 106, 109–11 (1977) (per curiam).
  • 77The relevant law is summarized, and critiqued, in Aziz Z. Huq, Judicial Independence and the Rationing of Constitutional Remedies, 65 Duke L.J. 1 (2015) (describing qualifying immunity in constitutional tort law and the fault rule that has emerged in the exclusionary rule context). Note that although most public employees benefit from qualified immunity, it is only police who benefit from leeway built into constitutional criminal procedure rules and the current exclusionary rule.
  • 78Note that this is not the case for all forms of procedural protections to which officers might be entitled. See Garrity v. New Jersey, 385 U.S. 493, 500 (1967) (holding that police “like teachers and lawyers” are entitled to due process protections in the employment context). Due process rights provide notice and an opportunity to present one’s case expand the information available to decision-makers, unlike delay privileges.
  • 79Joanna C. Schwartz, Myths and Mechanics of Deterrence: The Role of Lawsuits in Law Enforcement Decisionmaking, 57 UCLA L. Rev. 1023, 1028–30 (2010) (finding that law enforcement agencies rarely gather and analyze information from lawsuits brought against them and their officers).
  • 80Relatedly, we think it is unlikely that police officers require more time to identify a lawyer than the modal suspect. To the contrary, we suspect that police unions shrink search costs on this score. Cf. Harmon, supra note 35, at 799–801 (noting that police unions fund appeals from discipline determinations).
  • 81But cf. Stephen J. Schulhofer, Reconsidering Miranda, 54 U. Chi. L. Rev. 435, 447–48 (1987) (“Even the sophisticated law professor or professional investigator, if he found himself suspected of crime, would be under considerable pressure to cooperate with the police, to try to get them on his side by telling what he knew or what he thought he could safely disclose, rather than standing confidently on his right to remain silent.”).
  • 82Id. at 447.
  • 83Crawford v. Washington, 541 U.S. 36, 50 (2004).
  • 84See Michigan v. Bryant, 562 U.S. 344, 376 (2011).
  • 85562 U.S. 344 (2011).
  • 86Id. at 349.
  • 87Id. at 349–50, 374–75
  • 88Id. at 370.
  • 89Miranda v. Arizona, 384 U.S. 436, 479 (1966); see also Rhode Island v. Innis, 446 U.S. 291, 301 (1980) (defining interrogation as “any words or actions . . . that the police should know are reasonably likely to elicit an incriminating response from the suspect”).
  • 90See Charles D. Weisselberg, Mourning Miranda, 96 Cal. L. Rev. 1519, 1525 (2008) (explaining that “police training has undermined the effectiveness of a system of warnings and waivers”).
  • 91470 U.S. 298 (1985).
  • 92Id. at 301.
  • 93Id. at 301–02.
  • 94Id. at 309. A different result obtains, however, if police deliberately use an unwarned interrogation in violation of Miranda in a second, warned interrogation. See Missouri v. Seibert, 542 U.S. 600, 611 (2004) (plurality opinion).
  • 95Hence, Miranda does not apply where a suspect is stopped for a traffic offense and ordered out of his or her vehicle. See Berkemer v. McCarty, 468 U.S. 420, 437–40 (1984). Nor does it apply when a suspect voluntarily accompanies an officer. See generally California v. Beheler, 463 U.S. 1121 (1983).
  • 96New York v. Quarles, 467 U.S. 649, 656 (1984).
  • 97William J. Stuntz, The Distribution of Fourth Amendment Privacy, 67 Geo. Wash. L. Rev. 1265, 1274–75 (1999).
  • 98To the contrary, several of the coping or circumvention measures identified by Weisselberg, involve “softening up” a suspect before custodial interrogation begins. Weisselberg, supra note 90, at 1555–56. This suggests police still view the period immediately following an arrest and prior to “custody” as significant and useful.
  • 99See Task Force Report, supra note 12, at 14 (“[T]he CBAs make it easy for officers to lie if they are so inclined—they can wait 24 hours before providing a statement after a shooting, allowing them to confer with other officers, and they can amend statements after viewing video or audio evidence.”).
  • 100See Tire Sum Excuse, Snopes (1995), http://www.snopes.com/college/exam/flattire.asp (https://perma.cc/93YV-LM93).
  • 101The Chicago Task Force assembled in the wake of the McDonald shooting noted the same problem. Task Force Report, supra note 12, at 71 (“Critics contend that the waiting period provides officers time to protect themselves by agreeing on a false story they will all tell when they have their official interviews with investigators.”).
  • 102See supra Introduction.
  • 103One of us has explored in detail elsewhere the relationship between communication and the aligned expectations necessary for collective action. Richard H. McAdams, A Focal Point Theory of Expressive Law, 86 Va. L. Rev. 1649, 1658 (2000) (“What can produce the aligned expectations? One possibility is communication.”).
  • 104The CBA in effect in Chicago contains a provision that prevents officers from being charged with making a false statement on the basis of testimony that is contradicted by video or audio evidence unless the officer has had both a chance to review the recording before giving testimony and a chance to amend their statement subsequently. Task Force Report, supra note 12, at 72. We agree with the Task Force’s conclusion that this rule “does not seem to serve any valid purpose.” Id.
  • 105See supra text accompanying note 62.
  • 106There is no research that directly addresses this relationship. One empirical study of patterns of police misconduct, however, identified the increasing size of a department, along with the presence of an internal affairs unit, as predictors of higher rates of misconduct. David Eitle et al., The Effect of Organizational and Environmental Factors on Police Misconduct, 17 Police Q. 103, 113 (2014).
  • 107Study after study finds pervasive indemnification of individual officers. See, e.g., Lant B. Davis et al., Suing the Police in Federal Court, 88 Yale L.J. 781, 810–12 (1979) (reporting government defense and indemnification of police officers in Connecticut); Theodore Eisenberg & Stewart Schwab, The Reality of Constitutional Tort Litigation, 72 Cornell L. Rev. 641, 686 (1987); Joanna C. Schwartz, Police Indemnification, 89 N.Y.U. L. Rev. 885, 890 (2014).
  • 108We discuss such liability extensively in Part III.B.
  • 109And even if this effect is small at the margin, it means at the very least that municipalities have no intrinsic reason to scale up internal investigations absent external pressure. For suggestive anecdotal evidence of the pressure faced by municipal attorneys to suppress tort liability, consider the case of City of Chicago attorney Jordan Marsh, who allegedly lied about evidence in a tort action brought by the family of a police-shooting victim. Sarah Kaplan, Judge: Emanuel Administration Lawyer Withheld Evidence About a Fatal Police Shooting in Chicago, Wash. Post (Jan. 5, 2016), https://www.washingtonpost.com/news/morning-mix/wp/2016/01/05/judge-emanuel-administration-lawyer-intentionally-withheld-evidence-in-case-about-a-fatal-shooting-by-chicago-police/ (https://perma.cc/2J9J-3VFD).
  • 110Rachel Harmon, Why Do We (Still) Lack Data on Policing?, 96 Marquette L. Rev. 1119, 1129 (2013).
  • 111Id. at 1131.
  • 112Keenan & Walker, supra note 36, at 196–97; see also Samuel Walker, The Neglect of Police Unions: Exploring One of the Most Important Areas of American Policing, 9 Police Practice & Res. 95 (2008) (discussing police unions).
  • 113Brian A. Reaves, Local Police Departments, 2007, U.S. Dep’t of Just. 13 (Dec. 2010), http://bjs.gov/content/pub/pdf/lpd07.pdf (https://perma.cc/9BWS-XL37). Only Arizona, Georgia, North Carolina, Mississippi, South Carolina, and Virginia prohibit collective bargaining. See Seth W. Stoughton, The Incidental Regulation of Policing, 98 Minn. L. Rev. 2179, 2207 (2014).
  • 114Hervey Juris & Peter Feuille, Police Unionism 20–21 (1973). Thirty-four states require government employers to engage in collective bargaining with public-sector employees, and another nine states permit it. Richard B. Freeman & Eunice S. Han, The War Against Public Sector Collective Bargaining in the US, 54 J. Indus. Rel. 386 (2012).
  • 115J. M. Magenau & R.G. Hunt, Police Unions and the Police Role, 49 Human Rel. 1315, 1330–31 (1996); J. M. Magenau & R.G. Hunt, Sociopolitical Networks for Police Role-Making, 42 Human Rel. 547 (1989).
  • 116Stoughton, supra note 113, at 2211.
  • 117Id.
  • 118Also, there are many other contexts in which a profession’s high self-regard occludes their perception of “bad apples.” Consider priests’ or investment bankers’ responses to recent scandals in their ranks.
  • 119See Aziz Z. Huq, Standing for the Structural Constitution, 99 Va. L. Rev. 1435, 1502–03 (2013) (summarizing public-choice theory of legislative production).
  • 120The asymmetry in transaction costs is accentuated here because those harmed by delay privileges are unknown at the time the latter are negotiated. Hence, potential victims have no way of organizing at the necessary moment. Accord Brian Galle, Hidden Taxes, 87 Wash. U. L. Rev. 59, 71 (2009) (“Prices are also difficult for consumers to identify where costs depend in part on future events.”). For the argument that those attracted to policing would feel rewarded by the power to impose extra-legal harm on those they believe to deserve punishment, see Richard H. McAdams et al., The Law of Police, 82 U. Chi. Law Rev. 135 (2015); Dhammika Dharmapala et al., Punitive Police? Agency Costs, Law Enforcement, and Criminal Procedure, 45 J. Legal Stud. 105 (2016).
  • 121See Editorial, The Truth of ‘Black Lives Matter’, N.Y. Times (Sept. 3, 2015), http://www.nytimes.com/2015/09/04/opinion/the-truth-of-black-lives-matter.html (https://perma.cc/UX49-9PB6).
  • 122This resistance will include the unions responsible for pressing for delay privileges in the first instance. See Wesley G. Skogan, Why Reforms Fail, in Police Reform from the Bottom Up: Officers and Their Unions as Agents of Change, in 149 Police Reform from the Bottom Up: Officers and Their Unions as Agents of Change (Monique Marks & David Sklansky eds., Routledge 2012).
  • 123Samuel Walker, Institutionalizing Police Accountability Reforms: The Problem of Making Police Reforms Endure, 32 St. Louis U. Pub. L. Rev. 57, 59 (2012).
  • 124Joanna C. Schwartz, What Police Learn from Lawsuits, 33 Cardozo L. Rev. 841 (2012).
  • 125Carol A. Archibold & Edward R. Maguire, Studying Civil Suits Against the Police: A Serendipitous Finding of Sample Selection Bias, 5 Pol. Q. 222, 228 (2002) (summarizing previous empirical litigation showing some police sensitivity to the prospect of lawsuits). On the other hand, a survey of police chiefs in Texas found that two-thirds believed that “the possibility of lawsuits only mildly affected or had no effect at all on their departments.” Michael S. Vaughn et al., Assessing Legal Liabilities in Law Enforcement: Police Chiefs’ Views, 47 Crime & Delinq. 3, 6–7 (2001).
  • 126Ku Klux Klan Act of April 20, 1871, ch. 22, § 1, 17 Stat. 13 (current version at 42 U.S.C. § 1983 (2000)). In relevant part, the provision states: “Every person who, under color of any statute, ordinance, regulation, custom, or usage, of any State . . . subjects, or causes to be subjected, any citizen of the United States or other person . . . to the deprivation of any rights, privileges, or immunities secured by the Constitution and laws, shall be liable to the party injured . . . .”
  • 127365 U.S. 167 (1961) (overruled by Monell v. Department of Social Services of New York, 436 U.S. 658 (1978)).
  • 128Id. at 172 (rejecting the description of state action offered in Barney v. City of New York, 193 U.S. 430 (1904)) (abrogated by Estate of Conner by Conner v. Ambrose, 990 F. Supp. 606 (1997)).
  • 129436 U.S. 658, 690 (1978).
  • 130City of Newport v. Fact Concerts, Inc., 453 U.S. 247, 265–67 (1981).
  • 131Owen v. City of Independence, Mo., 445 U.S. 622, 624–25 (1980). One of us has argued elsewhere that certain elements of municipal liability doctrine echo qualified immunity. See Huq, supra note 77, at 25–26.
  • 132Monell, 436 U.S. at 691–92. Monell also stated that a city could be liable based on a custom among its employees. Id.; Adickes v. S. H. Kress & Co., 398 U.S. 144, 166–67 (1970) (emphasizing the difference between a custom among state employees and a custom among the general polity).
  • 133Monell, 436 U.S. at 691–92 (quoting 42 U.S.C. § 1983).
  • 134See, e.g., City of St. Louis v. Praprotnik, 485 U.S. 112, 124 (1988) (holding that “the identification of policymaking officials is a question of state law”); Pembaur v. City of Cincinnati, 475 U.S. 469, 480 (1986) (“With this understanding, it is plain that municipal liability may be imposed for a single decision by municipal policymakers under appropriate circumstances.”).
  • 135Tennessee v. Garner, 471 U.S. 1, 7–8 (1985) (establishing a totality of the circumstances test for Fourth Amendment challenges to police force); Graham v. Connor, 490 U.S. 386, 386 (1989) (same). This standard has been persuasively criticized as “indeterminate and undertheorized.” Rachel A. Harmon, When Is Police Violence Justified?, 102 Nw. U. L. Rev. 1119, 1120 (2008).
  • 136See, e.g., Ben Brumfield, Before Tackling James Blake, Police Officer Had Complaints of Brutality, CNN (Sept. 14, 2015), http://www.cnn.com/2015/09/14/us/james-blake-tackled-officer/ (https://perma.cc/PK32-E8AN]; see also Myriam E. Gilles, Breaking the Code of Silence: Rediscovering “Custom” in Section 1983 Municipal Liability, 80 B. U. L. Rev. 17, 67 (2000) (recounting other cases in which officers using unconstitutional force had histories of such conduct).
  • 137Personal communication of authors with Prof. Craig Futterman.
  • 138Bd. of Cnty. Commissioners of Bryan County, Okl. v. Brown, 520 U.S. 397, 401 (1997); id. at 415 (“Where a court fails to adhere to rigorous requirements of culpability and causation, municipal liability collapses into respondeat superior liability.”).
  • 139City of Canton, Ohio v. Harris, 489 U.S. 378, 390 (1989); see also City of St. Louis v. Praprotnik, 485 U.S. 112, 124–27 (1988) (establishing a narrow definition of official policymakers for § 1983 purposes).
  • 140Brown, 520 U.S. at 411.
  • 141Connick v. Thompson, 563 U.S. 51, 71 (2011) (emphasis in original) (citing Brown, 520 U.S. at 409); Huq, supra note 77, at 26 (criticizing the logic of Connick). Earlier cases took the plausible view that “[c]ircumstances can alter cases,” such that a historical pattern of violations can alert policy-makers to the actual effects of a policy sufficiently to establish Monell liability. Pena v. Leombruni, 200 F.3d 1031, 1033–34 (7th Cir. 1999) (Posner, J).
  • 142Parrish v. Luckie, 963 F.2d 201, 204 (8th Cir. 1992).
  • 143Cox v. District of Columbia, 40 F.3d 475 (D.C. Cir. 1994); see also Hogan v. Franco, 896 F. Supp. 1313, 1316 (N.D.N.Y. 1995).
  • 144Fiacco v. City of Rensselaer, N.Y, 783 F.2d 319, 326 (2d Cir. 1986) (finding that “indifference was demonstrated by the failure of the City defendants to exercise reasonable care in investigating claims of police brutality”); see also Hazel Glenn Beh, Municipal Liability for Failure to Investigate Citizen Complaints Against Police, 25 Fordham Urb. L.J. 209 (1998) (discussing those cases).
  • 145See Michael S. Vaughn, Police Civil Liability for Abandonment in High Crime Areas and Other High Risk Situations, 22 J. Crim. Just. 407 (1994) (summarizing relevant case law).
  • 146Ortega v. City and Cnty of Denver, 944 F. Supp. 2d 1033 (D. Colo. 2013).
  • 147944 F. Supp. 2d 1033 (D. Colo. 2013).
  • 148Id. at 1036.
  • 149Id. at 1039.
  • 150Id. at 1039–40 (internal citation omitted).
  • 151Ortega, 944 F. Supp. 2d at 1040–41 (internal citation omitted).
  • 152Schneider v. City of Grand Junction Police Dep’t, 717 F.3d 760, 777 (10th Cir. 2013) (citing Butler v. City of Norman, 992 F.2d 1053, 1056 (10th Cir. 1993)).
  • 153See 29 U.S.C. § 152 (defining the term “employer” under the National Labor Relations Act to exclude “the United States or any wholly owned Government corporation, or any Federal Reserve Bank, or any State or political subdivision thereof.”).
  • 154See, e.g., 51 C.J.S. Labor Relations § 328 (2016) (“As with contracts generally, a collective-bargaining agreement that is contrary to public policy is unenforceable.”); Am. Fed’n of State, County and Mun. Employees, AFL-CIO v. Dep’t of Cent. Management Services, 173 Ill.2d 299, 307 (Ill. 1996) (“As with any contract, a court will not enforce a collective-bargaining agreement that is repugnant to established norms of public policy.”). Regarding the public policy doctrine generally, see Restatement (Second) of Contracts § 178 (1981).
  • 155See, e.g., 5 Samuel Williston and Richard A. Lord, A Treatise on the Law of Contracts § 12:2 (4th ed. 1990) (“[A]t least some courts are not the least bit reluctant to rely on public policy in declaring bargains they perceive to be inappropriate invalid per se.”).
  • 156See, e.g., Elletta Sangrey Callahan & Terry Morehead Dworkin, The State of Whistleblower Protection, 38 Am. Bus. L.J. 99 (2000).
  • 157Lachman v. Sperry-Sun Well Surveying Co., 457 F.2d 850, 853 (10th Cir. 1972).
  • 158See, e.g., Smith v. Delaware State University, 47 A.3d 472, 477 (Del. 2012) (“We hold that an employee who is constructively discharged can pursue a claim under the Whistleblower Act, on the same basis as if she had been formally discharged.”); Univ. of Texas Medical Branch at Galveston v. Hohman, 6 S.W.3d 767, 773 (Tex. App. 1999) (same).
  • 159See, e.g., Ballinger v. Delaware River Port Auth., 172 N.J. 586, 606–07 (N.J. 2002) (firing police officer for reporting suspected criminal behavior outside the chain of command violates public policy).
  • 160Cf. Shadis v. Beal, 685 F.2d 824, 832 (3d Cir. 1982) (holding that a contractual provision by which association partly funded by the state government waives its rights to attorney fees in civil rights litigation against the state is void).
  • 161Cf. Jones v. Chevalier, 579 So.2d 1217 (La. Ct. App. 1991) (void to contract to warn a person of police investigation).
  • 162751 N.E.2d 1261 (Ill. App. Ct. 2001); see also City of New York v. Uniformed Fire Officers Ass’n, Local 854, 263 A.D.2d 3 (N.Y. App. Div. 1999); Prince George’s County, MD. ex rel. Prince George’s County Police Dep’t v. Prince George’s County Police Civilian Employees Ass’n., 98 A.3d 1094 (Md. Ct. of Special Appeals 2014). But see City of Highland Park v. Teamster Local Union, 828 N.E.2d 311 (Ill. App. Ct. 2005).
  • 163751 N.E.2d at 1263.
  • 164Id. at 1265–66.
  • 165Id. at 1266–67.
  • 166See United Paperworkers v. Misco, 484 U.S. 29 (1987). Nonetheless, there are some exceptional cases that have held CBA provisions upheld by an arbitrator to be void as against public policy. See Am. Fed’n of State, County and Mun. Employees, AFL-CIO v. Dep’t of Cent. Management Services, 173 Ill.2d 299 (Ill. 1996).
  • 167751 N.E.2d at 1266 (citing Palmateer v. Int’l Harvester Co., 421 N.E.2d 876, 880
    (Ill. 1981)).
  • 168Id. at 1267.
  • 169Id. (emphasis added).
  • 170659 F.2d 306 (3d Cir. 1981).
  • 171Id. at 310.
  • 172Id. at 329 (citing Pennsylvania v. New Jersey, 426 U.S. 660, 665 (1976)).
  • 173Id. at 316.
  • 174No. 01-Civ-0364, 2001 US Dist. LEXIS 13364 (S.D.N.Y. Mar. 16, 2001).
  • 175See generally id.
  • 176See Deshawn E. v. Safir, 156 F.3d 340, 344 (2d Cir. 1998) (citing City of Los Angeles v. Lyons, 461 U.S. 95, 10–06 (1983)) (“A plaintiff seeking injunctive or declaratory relief cannot rely on past injury to satisfy the [Art. III] injury requirement but must show a likelihood that he or she will be injured in the future.”).
  • 177See New York v. Town of Wallkill, No. 01-Civ-0362, 2001 U.S. Dist. LEXIS 13364, at *20 (S.D.N.Y. Mar. 16, 2001).
  • 178Id. at *21.
  • 179Id. at *10.