3 | Police Violence

On a Thursday afternoon in late May 1957, police in Cleveland, Ohio, broke into the home of a young African American woman named Dollree Mapp. The officers thought Mapp might be running a numbers racket. They also had spotted the car of a man named Virgil Ogletree outside Mapp’s house, and they suspected that Ogletree had been involved in the recent bombing of the home of another numbers operator—Don King, the future boxing promoter. Mapp refused to let the police in without a warrant. When they returned with what they said was a warrant, she did not come to the door. So the officers tried kicking in her back door, and when that didn’t work they smashed a window and let themselves in. They met Mapp in her hallway and showed her the “warrant.” She grabbed the document and stashed it inside her dress. The police wrestled the document back from her, twisting her hand until she cried out in pain, and then handcuffed her and forced her to accompany them while they searched the house. They found some pornographic books and pictures, for which Mapp was subsequently prosecuted. In a basement apartment that Mapp rented out, they also found Virgil Ogletree.

Mapp was convicted and sentenced to prison for possession of obscene materials, but the United States Supreme Court threw the conviction out. It took the Court some time to figure out what bothered it most about Mapp’s treatment. Initially the Justices took the case to examine whether the prosecution infringed unconstitutionally on free expression, but ultimately they reversed Mapp’s conviction on a different ground: searching Mapp’s house without a warrant had violated her Fourth Amendment right against “unreasonable searches and seizures.” No warrant for the search was ever produced in court, and the trial judge questioned whether it had ever existed. Law students today still read the Supreme Court’s decision in Mapp v. Ohio, because it established the Fourth Amendment “exclusionary rule” for state criminal prosecutions—the rule that evidence obtained through an unconstitutional search or seizure generally cannot be introduced at trial against a criminal defendant.¹

Equally significant, though, is what the Supreme Court’s decision in Mapp v. Ohio is not about. It is not about violence. It is not about the police breaking into Mapp’s house, grabbing her, wrestling with her, twisting her hand, and handcuffing her. The Justices took note of all of that, but ultimately it was irrelevant to their decision. At first they thought the case was about free expression, and then they thought it was about warrants. They never thought it was about violence—despite the fact that the Ohio Supreme Court had concluded that the evidence found in the search was admissible because the police hadn’t used “brutal or offensive physical force.” Instead of rejecting that conclusion, the United States Supreme Court simply ignored it. Brutality and “offensive physical force” were beside the point.²

In this respect Mapp v. Ohio is emblematic. There is a long history of concern with police violence—typically called “police brutality” when it is thought unjustified, and “use of force” when it is thought legitimate. Complaints about police violence are as old as the nightstick.³ Recently the issue of police violence has received new attention, largely due to the Black Lives Matter movement and, in particular, the wave of nationwide protests following the death of George Floyd at the hands of the Minneapolis police in May 2020. But whereas violence became increasingly significant in substantive criminal law during the second half of the twentieth century and the first two decades of the twenty-first century, the story was very different in criminal procedure—the rules governing what the police can and cannot do. There, over decades, violence became progressively less consequential.

The moment of inflection in both stories was roughly the same: the early 1960s. The Supreme Court decided Mapp in 1961, one year before the appearance of the Model Penal Code, the first major codification of criminal prohibitions to treat violence as a significant dividing line. Just at the point when violence was beginning to emerge as the criminal law’s most important category for sorting serious from forgivable offenses, the law of criminal procedure, along with the instincts about policing it embodies, was moving in the opposite direction.

The two developments were related. One reason criminal procedure focuses so little on violence is that concerns about violent crime have made police violence seem more necessary and less troubling. As Figure 4 suggests, popular discussion of police violence and “police brutality” rose sharply in the 1960s. In the 1970s and early 1980s, however, when concerns about violent crime skyrocketed, discussions of police violence and police brutality waned. This happened even as discussions of other kinds of controversial police practices—most notably, warrantless searches and seizures—continued to rise.⁴ Part of the explanation is that when violent crime seems out of control, physical force by the police seems more legitimate and is less likely to be called “brutality.” It is a version of the tit-for-tat logic that was discussed in Chapter 1, the sentiment that President Trump invoked when he urged police officers to stop protecting suspects’ heads when putting them into patrol cars: “Like when you guys put somebody in the car and you’re protecting their head … and they’ve just killed somebody.… [Y]ou can take the hand away, okay?” Trump doubled down on this kind of rhetoric during the nationwide wave of protests and lootings in the spring of 2020. Beyond threatening that “when the looting starts, the shooting starts”—the message that prompted a warning label from Twitter—Trump boasted about the force that Secret Service agents quickly used against protesters near the White House who “got too frisky or out of line”: “They didn’t know what hit them.” He repeatedly urged state and local law enforcement agencies to be “tougher” on looters and vandals, to “fight back,” to “do retribution,” and not to be “too careful.”⁵ Trump hardly invented this way of thinking. Here is Manhattan prosecutor Thomas Kane in the 1930s, for example, defending the use of “third degree” to get witnesses to talk: “What are we to do—give our baby killers ice cream?”⁶

But the low salience of violence in the modern law of criminal procedure is not entirely a by-product of the high salience of violence in criminal law, nor does it reflect an indifference to how the police use coercive force. The “third degree,” a pervasive feature of policing in the 1930s, became rare by the late twentieth century, and it has stayed rare. It was eradicated primarily by a wave of public outrage and elite condemnation.⁷ Moreover, concerns about police brutality never disappeared in the 1970s and 1980s, and by the end of the 1980s discussions of the topic were once again growing more common. This happened while the focus on violent crime in popular discourse continued its steep upward climb, and while the line between violent and nonviolent offenses grew more consequential than ever in criminal sentencing schemes and criminal justice reform proposals. The last decade and a half of the twentieth century, remember, was the era of Three Strikes laws; it was also the era of diversion programs narrowly targeted at “nonviolent” offenders. It was not an era, though, in which violence became significantly more important in the legal rules governing the police. That is partly because by the late 1980s the Supreme Court had become far less interested in placing new legal constraints on the police, but it is also partly because, as we will see, the framework the Supreme Court constructed in the mid-twentieth century for regulation of the police—the framework erected in cases like Mapp v. Ohio—focused much less on violence than on violations of privacy.

FIG. 4 Frequency of references to “police brutality” and “police violence” in books published in the United States, 1900–2000, compared with references to “violent crime” and to “warrantless” or “without a warrant.” The vertical axis measures the occurrences of each phrase as a proportion of all sequences with the same number of words. The lines show seven-year trailing averages. (Data source: Google Books Ngram Viewer.)

The relative unimportance of violence in criminal procedure, therefore, is not simply the flip side of the heavy emphasis placed on violence in criminal law; it has its own history and its own drivers. One of those drivers has been the very idea of the police: our underlying notions about what the police are for and our expectations about how they should operate. There is a long-standing sense, particularly strong in the United States, that the fundamental point of the police is to enforce the law and maintain order through the actual or threatened use of physical force. The police themselves have embraced this idea for most of their history. Line officers have long prized their ability to “take care” of themselves on the street, and that is the quality that, more than any other, has gained them respect among their fellow officers.⁸ The notion that violence is inherent to the very idea of the police was a large part of the reason that many of the participants in the nationwide protests in 2020 called, not for the reform of the police, but for their defunding or abolition.

“Police brutality” is often described as the illegal exercise of violence by law enforcement officers, but the line between authorized and unauthorized uses of force by the police has always been hazy. The police have frequently understood their role to include the use of unofficial, “extralegal” force; they have assumed, usually with justification, that this is what the public expected of them. The New York Police Department, with the support of local newspapers, encouraged “nightstick rule” in the early twentieth century, and in the 1930s the department’s commissioner, Lewis Valentine, explicitly called on officers to “muss up” gangsters. As late as 1947, then-commissioner Arthur Wallander instructed NYPD detectives to give “the proper treatment” to “hoodlums,” “loafers,” and “easy money men.” In the early 1950s the pioneering police ethnographer William Westley found that officers in Gary, Indiana, understood that local residents wanted them to give “extremely rough treatment” to homosexuals and other “deviants,” but to do so unofficially and out of view. Complying with those expectations, Westley argued, accustomed the police “to use violence as a general resource.” And then there is Trump.⁹

None of this is to suggest that attitudes about police violence have stayed constant, either among the public or among officers themselves. The third degree was once widely accepted; by the closing decades of the twentieth century it was not. Police violence in response to “contempt of cop” remains widespread, but departments are less likely today than they once were to wink at street justice meted out to idlers or suspected sexual offenders. In the 1960s few police departments had written policies regarding the use of force; by the end of the century most did. It became common in the 1990s and early 2000s for police officers to study “verbal judo” and other techniques of de-escalation. Brawling skills became less central to an officer’s reputation within a department. Many departments began providing officers with “crisis intervention” training, to help them avoid violence in encounters with people suffering from flare-ups of mental illness. Finally—and most consequentially—cellphone videos and the Black Lives Matter movement brought new scrutiny to police violence. The nationwide protests in the wake of the 2020 killing of George Floyd were direct outgrowths of that scrutiny.¹⁰

Those protests also highlighted, though, the degree to which the licit and illicit use of force in law enforcement remained widespread, even routine. The protests responded not just to Floyd’s death but to a string of other police killings of African Americans. To a scandalous extent, police violence remained lightly regulated, a peripheral topic in the law of criminal procedure. This chapter will explore why that was, and whether and how it should change. The legal rules governing police conduct, like the rules of substantive criminal law, reflect ideas about the significance of violence, the definition of violence, and the workings of violence. The relative unimportance of violence in the rules of criminal procedure provides a telling contrast with the prominent role violence plays in modern sentencing codes and punishment schemes.

Because criminal procedure doctrines place little emphasis on violence, the definition of violence—or of “brutality” or “force”—matters much less than in substantive criminal law. But in ways that this chapter will explore, language and definitions matter here as well. The term “police brutality,” for example, has important and unfortunate overtones, suggesting that police violence is a form of animalism, driven by individual or group character rather than training, protocols, and workplace culture. It can lead to understandings of police violence that overemphasize the significance of “bad apples” or the otherness of police officers as a whole.

The terminology we use to describe police violence therefore reflects and helps to shape our ideas about how violence operates in policing. And just as in substantive criminal law, ideas about how violence operates matter a great deal in regulation of the police. It matters whether violence is understood to be characterological or situational—and, if it is situational, which aspects of an officer’s situation are thought to be most important in leading the officer to use force. It matters, too, whether violence is thought to be self-limiting or self-reinforcing: whether tit for tat is understood to be the best way to keep violence in check, or a dynamic that is apt to spin out of control. Police violence has been defended since the nineteenth century, by officers and others, as a necessary corrective to private violence. But there is also a long tradition of worrying that police violence breeds further violence: that official lawlessness encourages private lawlessness, and that violent policing produces more, not less, violent crime.¹¹

In criminal procedure, as in substantive criminal law, it is impossible to unravel these various ideas about violence without taking account of race, gender, and class. Race plays a particularly large role here, as in discussions of substantive criminal law. The Black Lives Matter movement, and the nationwide protests in May and June 2020, were responses to the dramatically disproportionate rate at which people of color, particularly young men of color, are victims of police violence. Because of that disparity, the problem of police violence has long been understood differently in minority communities than in white communities. Gender is important, too: excessive violence by police officers has often been blamed on the heavily masculine culture of law enforcement, and—on the other side—tough tactics by the police have been defended and celebrated as a form of manliness. Class figures also, not just because the victims of police violence are disproportionately poor or working class, but because the police themselves are viewed as working class, by officers themselves and by the public. The association of the police with the working class has influenced how police violence is understood and how the law approaches it; it has lent credence, for example, to the notion that police are violent because of the kind of people they are, not because of how they are trained or deployed.

The Insignificance of Violence in Criminal Procedure

To get some perspective on the role that violence now plays in substantive criminal law, we turned to history. We traced how ideas about the relative gravity of criminal offenses have evolved since the eighteenth century. It is harder to do with that criminal procedure, the set of rules governing the police, because those rules are largely a twentieth-century invention. Modern police forces themselves date back only to the middle of the nineteenth century, and constitutional criminal procedure did not develop in earnest until the middle of the twentieth century, when federal law enforcement agencies began to assume significant responsibilities and the Supreme Court began to regulate state law enforcement officers. There is no real analog to our current law of criminal procedure before the 1930s.

Still, criminal procedure has a kind of prehistory: the ragtag collection of statutory restrictions, judicial pronouncements, and received understandings about criminal investigations that served as a background to, and helped to motivate, the Fourth, Fifth, and Sixth Amendments to the United States Constitution. Calling this miscellany “common law,” as the Supreme Court often does, suggests that it had more cohesion and consistency than it actually did.¹² Nonetheless, some generalizations are possible, and they are instructive. Eighteenth- and early nineteenth-century limitations on law enforcement did not categorize official conduct based on whether or not it was violent, but it did focus particular attention on conduct that was violent—similar to the way substantive criminal law focused particular attention on violent offenses like murder, rape, mayhem, and assault. That focus was sharper with regard to police conduct, though, than it was for private behavior.

The common-law crime of burglary, for example, did require a kind of violence—a “breaking” of a house. But the breaking could consist simply of opening a door or window; any use of force sufficed. From a very early date, moreover, common-law authorities recognized “constructive” breaking of houses, involving no use of force whatsoever: when entry was obtained by fraud, for example, or by persuasion of a child or other innocent agent. The restrictions on criminal investigations, however, focused more heavily on actions that could be characterized as violent, such as arrests of suspected criminals, which typically involved physical capture, and forcible entries into homes. Searches of homes, for offenders or for evidence, received significantly greater scrutiny when they involved breaking doors open. The influential early nineteenth-century treatise by the English lawyer Joseph Chitty, for example, stressed that although officers seeking to take an offender into custody could break into a house without a warrant, this was “so violent, obnoxious, and dangerous a proceeding” that “it should be adopted only in extreme cases, where an immediate arrest is requisite.” Even with an arrest warrant, doors could be broken open only “if admittance cannot otherwise be obtained,” and probably only in cases of “treason, felony, or breach of the peace.” “In all cases whatever,” Chitty emphasized, “it is absolutely necessary that a demand of admittance should be made, and be refused, before outer doors can be broken.”¹³

Two decades ago the US Supreme Court concluded that this “knock-and-announce principle” was such a basic part of the common-law restrictions on law enforcement that violating it could make a search or seizure “unreasonable” and hence unconstitutional under the Fourth Amendment. The Supreme Court explained that the rule served three purposes: it avoided unnecessary damage to property, it safeguarded the “privacy and dignity” of residents, and, perhaps most important, it protected “human life and limb”—“because an unannounced entry may provoke violence in supposed self-defense by the surprised resident.” Because the rule was not aimed at helping to hide evidence, though, the Court concluded that the exclusionary rule would not be triggered by its violation. And the Supreme Court made it clear that there would be no violation to begin with if the officers had a good reason not to wait before entering: if, for example, the officers “reasonably believed that a prior announcement would [place] them in peril,” or if they had “reason to believe that evidence would likely be destroyed if advance notice were given.”¹⁴

The knock-and-announce principle thus survives in modern search-and-seizure law, but only vestigially. And the rest of search-and-seizure law places little weight on violence in sorting permissible from impermissible police conduct. That wasn’t always the case. When the Supreme Court first began to regulate evidence gathering by state law enforcement officers, in 1952, violence was very much on its mind. The case was Rochin v. California, and it involved three Los Angeles deputy sheriffs who broke into the bedroom of a suspected drug dealer. The suspect, Antonio Rochin, swallowed what the deputies assumed were illegal pills, so they jumped on him, “squeeze[ed] his throat,” and stuck their fingers in his mouth in an effort to recover the contraband. When those efforts failed, the deputies handcuffed Rochin, took him to an emergency room, strapped him to an operating table, and had his stomach pumped against his will. The capsules were recovered; they turned out to contain morphine. Rochin was convicted of drug possession, but the Supreme Court reversed. The Justices reasoned that the force used against Rochin was “so brutal and so offensive to human dignity” as to violate the constitutional guarantee of due process. It was “too close to the rack and the screw.” Sanctioning this kind of police violence, the Court said, would “brutalize the temper of [our] society.”¹⁵

The Supreme Court analogized the forcible pumping of Rochin’s stomach to coercing a confession, an investigative tactic the Justices had earlier found to violate due process. That conclusion, too, was first reached in a case involving police violence. The violence in Brown v. Mississippi—the first case ever in which the Supreme Court found that the Constitution had been violated by state law enforcement officers—was horrifyingly worse than in Rochin v. California. Ed Brown and his two codefendants were Black tenant farmers convicted in 1934 of murdering a white planter. They were convicted based entirely on confessions that had been obtained from them by torture: all three defendants had been whipped, and one had been strung up from a tree by his neck. The deputy sheriffs who had carried out the torture freely admitted it in court. When asked how badly one of the defendants had been whipped, one of the deputies testified, “Not too much for a negro; not as much as I would have done if it were left to me.” The Supreme Court had never before reversed a state criminal conviction because of the actions taken by law enforcement personnel, but it unanimously concluded in 1936 that “brutal treatment” of Brown and his codefendants, and the use of the resulting confessions in court, violated due process. “It would be difficult,” the Justices said, “to conceive of methods more revolting to the sense of justice.”¹⁶

At its origins, then, the Supreme Court’s regulation of state law enforcement tactics focused explicitly on police “brutality.” This was true when the Supreme Court first applied the Constitution to state interrogation practices in 1936, and it was true in 1952 when the Justices began to review searches and seizures by state law enforcement officers. The attention the Justices paid to police violence in Brown v. Mississippi and Rochin v. California reflected attention the issue received outside of the courts in the first half of the twentieth century. Most famously, the Wickersham Commission on Law Observance and Enforcement—a blue-ribbon panel assembled by President Herbert Hoover, which conducted the first-ever federal review of policing and prosecution in the United States—strongly condemned the use of the “third degree” in interrogation rooms, a practice the commission made clear was common throughout the country. The commission’s report, “Lawlessness in Law Enforcement,” issued in 1931, helped solidify public opposition to coerced confessions in the years leading up to the Supreme Court’s 1936 decision in Brown v. Mississippi. The commission paid less attention to police violence outside of the interrogation room, but it didn’t ignore that problem, either. In fact, the commission blamed gun violence by police officers for much of the unpopularity of Prohibition, which the commission concluded had gotten off to a “bad start” in part because “high-handed methods, shootings and killings, even when justified, alienated thoughtful citizens, believers in law and order.” The Wickersham Commission itself was the product of growing concerns about law enforcement in the early twentieth century, concerns that were amplified by Prohibition and that often centered around police violence. Even before the Supreme Court overturned the convictions in Brown v. Mississippi, state courts had repeatedly complained about the “third degree,” although they still allowed prosecutors to use any confessions that seemed reliable, no matter how they had been obtained.¹⁷

The focus on violence in Brown and Rochin didn’t last. The Supreme Court repeatedly reaffirmed the central holding of Brown, that prosecutors could not rely on coerced confessions, but it did so in cases involving haranguing rather than physical torture. The central question was whether the defendant’s “will was overborne by official pressure.” When the Court decided Miranda v. Arizona in 1966, it replaced that amorphous standard with a set of bright-line rules requiring police to advise suspects of their rights, offer to provide them with lawyers, and stop questioning whenever the suspect requested. By then the focus had shifted entirely away from violence in interrogation rooms to the “inherently compelling pressures” of “a police dominated atmosphere.” Although the “third degree” had not been eradicated, the Court stressed that it was now the exception rather than the norm, and interrogation practices had become “psychologically rather than physically oriented.” The Miranda rules were intended to counteract “the compulsion inherent in custodial surroundings.”¹⁸

The shift in focus in interrogation cases, away from violence and toward psychological pressure, partly reflects the striking success the Wickersham Commission and other early twentieth-century reformers achieved in combatting use of the third degree. But the focus has also shifted away from violence in the law of search and seizure, where there is much less reason to believe that the underlying problem of violence has been tamed. In interpreting and applying the Fourth Amendment to the US Constitution, which bans “unreasonable searches and seizures,” the Supreme Court has concentrated heavily on protecting informational privacy, the ability to keep secrets from the police. That is why the Supreme Court ultimately ignored the violence in Mapp v. Ohio and hinged its decision on the fact that the police had entered Dollree Mapp’s house without a warrant. Most of Fourth Amendment law is about where the police can look and what they can monitor. The late legal scholar William Stuntz complained with justification that “we have a large and detailed body of law to tell police when they may open paper bags or the trunks of cars,” but “the law speaks softly (or not at all) when it come to the level of force that may be used in making an arrest or conducting a search.”¹⁹

Stuntz traced this emphasis on privacy to Boyd v. United States, a late nineteenth-century decision by the Supreme Court, which ruled that a company could not constitutionally be forced to turn over documents showing its failure to pay import duties. The Court later abandoned that holding, along with much of the reasoning in Boyd. Boyd relied both on the Fourth Amendment and on the Fifth Amendment privilege against compelled self-incrimination, but the Supreme Court has since ruled that businesses don’t have a Fifth Amendment privilege, and that a subpoena for documents is not equivalent to a “search” under the Fourth Amendment. Another part of the reasoning in Boyd proved ahead of its time, though. The essence of an unconstitutional search, the Supreme Court suggested in Boyd, was “not the breaking of … doors and the rummaging of … drawers,” but violating “the sanctity of a man’s home and the privacies of life.”²⁰

Justice Brandeis leaned heavily on Boyd when he dissented from the Supreme Court’s decision in 1928 that wiretapping was not a “search” or “seizure” regulated by the Constitution. Brandeis argued that the Fourth Amendment should be read to prohibit “every unjustifiable intrusion by the government upon the privacy of the individual.” Four decades later a majority of the Supreme Court took essentially the same view. In Katz v. United States, decided in 1967, the Supreme Court reversed itself and concluded that the Fourth Amendment did in fact regulate electronic eavesdropping on telephone conversations. Although the Court in Katz continued to deny that the Fourth Amendment created any “general constitutional ‘right to privacy,’ ” it stressed that what an individual “seeks to preserve as private, even in an area accessible to the public, may be constitutionally protected.” In subsequent decisions the Court clarified that under Katz the Fourth Amendment regulated any government infringement of “reasonable expectations of privacy.”²¹

Ever since Katz the jurisprudence of the Fourth Amendment has focused overwhelmingly on protecting privacy. And in this respect Katz was prefigured by a series of decisions earlier in the 1960s—including, as we have seen, the case of Dollree Mapp. The facts of Mapp v. Ohio “seemed to cry out for limitation on police force,” points out the legal historian Wesley Oliver, but the Supreme Court reversed Mapp’s conviction only because the police appeared to have lacked a warrant to enter her house.²²

It is no accident that the defendants in Brown v. Mississippi, Rochin v. California, and Mapp v. Ohio all were nonwhite. Police violence has always been directed overwhelmingly at marginalized groups—especially people of color, but also poor whites, political dissidents, and violators of prevailing sexual norms. The restrictions the Supreme Court imposed on state and local law enforcement in the twentieth century were plainly motivated in large part by concerns about how the police treated marginalized groups, particularly African Americans in the states of the former Confederacy. But the Court almost never placed legal significance on the race of the defendants in criminal procedure cases, and it rarely dwelled on the nexus between race and police brutality. Brown, Rochin, and Mapp were typical in this regard.

If the Court placed little weight on race in criminal procedure, it had nothing at all to say about the policing of sexuality. When William Westley carried out his pioneering ethnography of policing in Middle America, he found that the officers thought it was part of their job to beat up “deviants,” a category in which he lumped together homosexuals, exhibitionists, peeping toms, and rapists. Westley wrote in the early 1950s, in the middle of a nationwide panic over “perverts” and “sexual psychopaths,” labels that in practice often wound up serving as code for homosexual. The “extralegal” violence that police visited on homosexuals was an open secret in the middle decades of the twentieth century, but it was rarely viewed as an important part of the broader problem of police brutality. Westley was unusual in suggesting that the experience of the police in sex cases helped teach them to view violence as a “general resource.” The Supreme Court, in particular, never talked this way. When the Court talked about police violence, it rarely touched on the connection with race, and it never mentioned the connection with sexuality. And as the constitutional rules regulating law enforcement proliferated in the second half of the twentieth century, few of them focused on violence. They focused on privacy. The Katz test for what counts as a “search” was a particularly important example, but far from unique.²³

In recent years the Supreme Court has slightly modified the Katz rule regarding the scope of the Fourth Amendment. The amendment applies, the Court now says, if there is either an invasion of a “reasonable expectation of privacy”—the test derived from Katz—or a trespass on a suspect’s person or his house, papers, or effects. By its terms, the Fourth Amendment bars only “unreasonable searches and seizures” of “persons, houses, papers, and effects,” and when the Supreme Court decided in 1928 that wiretapping wasn’t regulated by the Fourth Amendment, it did so largely because there had been no physical trespass on anything falling within these four categories. For decades Katz was understood to have supplanted the so-called trespass test for whether there had been a search or seizure, but the Court now says that Katz simply “added to” the earlier test, rather than replacing it. As a consequence, Fourth Amendment law focuses somewhat less relentlessly on privacy than it did until recently. The focus on privacy has been supplemented with attention to trespasses, and violence is of course a kind of trespass on a person. But privacy remains by far the dominant focus. Moreover, the trespasses that have caught the Supreme Court’s eye have been trespasses on property—the undercarriage of a car, and the area around a house—not trespasses on persons. Violence remains a peripheral topic in the law of search and seizure.²⁴

Part of the reason may be precisely that police violence is heavily targeted at people on the margins: poor people, nonconformists, and people of color. Invasions of privacy are also experienced disproportionately by people on the margins. All kinds of policing are. But the invasions of privacy associated with policing are at least somewhat more widely shared. Everyone gets pulled over for traffic violations; everyone has their bags searched at sporting events. But most upper- or middle-class white Americans have never been struck by a police officer, and they may not know anyone who has. The lack of firsthand experience may explain, for example, the insignificance of violence in the legal treatment of “stop and frisk,” a mainstay of urban policing since the 1960s.²⁵

Stop and Frisk, Police Shootings, and the Continuum of Force

Here is how the legal scholar Paul Butler describes the stop-and-frisk tactic, as experienced by young African American men: Police jump out of a car with their guns drawn and order you to face the wall with your hands up, then “they put their hands roughly all over your body.” Or “they kick your feet to spread your legs wider” and then “pat you up and down” and “touch your private parts.” Here are a few examples collected by law professor Kami Chavis: an officer shoves a suspect against a wall when he reaches for identification; an officer pulls a suspect’s arm behind his back and threatens to punch him in the face; an officer smashes a suspect’s head into a wall, leaving a gash that requires stitches. These are extreme examples, but they are not outliers. Chavis notes that nearly half of young people surveyed in an intensively patrolled area of New York City said a police officer had used force against them, and the New York Police Department’s own statistics show that a decade ago force was being used in tens of thousands of stop-and-frisk encounters every year. Since then, police in New York City have dramatically reduced their use of stop and frisk, largely in response to complaints about, and court challenges to, its disproportionate use against young men of color. But the tactic still is used hundreds of times a day.²⁶

The tactic is inherently violent, as Chavis points out: it isn’t like having a wand passed over your body at the airport. Butler likens stop and frisk to sexual assault. He calls it a low-grade, sexualized form of police brutality. Butler is neither the first nor the last scholar to notice the sexual overtones of stop and frisk. Seth Stoughton, once a police officer and now a law professor, had a standard phrase he used during a stop and frisk: “I don’t mean to feel you up or nothing, but I’m about to.” He explains, “I may have had to grope people, but I didn’t want to be unprofessional about it.”²⁷

Very little of this makes it into judicial discussions of stop and frisk, and it doesn’t shape how the law treats the tactic. Courts treat “frisks”—like the “stops” that accompany them—as “brief,” “slight,” “narrowly circumscribed intrusions.” It is true that the Supreme Court has required more justification for a frisk than for a stop: a stop requires “reasonable articulable suspicion” that a suspect has committed or is about to commit a crime, whereas a frisk requires both a lawful stop and “reasonable articulable suspicion” that the suspect is “armed and dangerous.” Terry v. Ohio, the 1967 Supreme Court decision that is the source for these requirements, stressed that a frisk “is a serious intrusion upon the sanctity of the person, which may inflict great indignity and arouse strong resentment.” Three years earlier, New York’s highest court had called a frisk “a minor inconvenience and petty indignity,” which police were justified in carrying out whenever they lawfully stopped someone for questioning. But Chief Justice Earl Warren, who wrote the majority opinion for the Supreme Court in Terry, called that suggestion “simply fantastic.” He said the very term “stop and frisk” was a euphemism.²⁸

Still, what Warren emphasized about “frisks” in Terry was the toll they could take on dignity: the demeaning, disrespectful nature of “a careful exploration of the outer surfaces of a person’s clothing all over his or her body in an attempt to find weapons.” He didn’t describe the procedure as violent, or discuss how it could be violent. John Terry, the lead defendant in the case, hadn’t just been patted down; the officer had “grabbed” him and “spun him around.” But Warren made nothing of that; it was mentioned but then ignored, like the police “running roughshod” over Dollree Mapp. Nor did he address the ways in which frisks could involve considerably more force than what Terry had experienced.²⁹

Ever since Terry, courts have largely read violence out of the stop and frisk. More precisely, they have read violence out of the execution of a stop and frisk. In judicial parlance, “frisk” is more or less a synonym for “pat-down,” and that is how the procedure is typically envisioned. On the other hand, violence is very much part of the justification for stop and frisk. The deadly violence that threatens law enforcement officers, the danger that “the answer to the question propounded by the policeman may be a bullet,” is the reason the frisk is permitted, and the reason the Supreme Court conditioned its legality on an officer’s articulable basis for worrying that a suspect is armed and dangerous. But the complete absence of violence from the other side of the ledger may be part of the reason that in practice this legal standard has proven undemanding—why anything beyond a bare intuition of danger, or an uncorroborated anonymous tip that a suspect has a gun, usually suffices to justify a frisk.³⁰

Violence isn’t entirely absent from the law of criminal procedure. The Supreme Court ruled in 1985 that the Fourth Amendment imposes special limitations on the use of deadly force: stopping a fleeing suspect by shooting him, or using any other kind of deadly force, is an “unreasonable” and hence unconstitutional seizure unless the police have probable cause to believe the suspect poses a risk of death or serious bodily harm to officers or the public. In addition, some lower courts have established loose requirements for using “severe” or “intermediate” force, like a Taser. One federal court of appeals has said that force of this kind requires an “immediate safety risk”; another had said that it is inappropriate when a suspect has been detained for a “minor infraction” and “clearly poses no risk to the officer or to public safety.” Usually, though, the only constitutional standard governing police violence is the general, open-ended rule that the police must act reasonably, under all circumstances. Recently, in fact, the Supreme Court warned against imposing “rigid preconditions” even on uses of deadly force, beyond the general requirement of “reasonableness.” The vagueness of that standard, combined with judicial reluctance to second-guess the police, has meant that almost any violence that police employ—grabbing, shoving, tackling, or striking people, brandishing weapons, using Tasers, breaking down doors, running cars off the road—is lawful as long as the officers could plausibly think it was necessary.³¹

In practice, judicial assessments of police violence have long been even more deferential than that. Unlike an illegal search, uses of force by law enforcement typically do not produce evidence that can be suppressed in a criminal trial, so the main way to challenge them in court is through a civil lawsuit seeking a monetary award. But the Supreme Court has held that police officers can’t be sued for violating constitutional rights unless it was crystal clear that what they did was forbidden. The Justices have said that this rule of “qualified immunity” is “particularly important in excessive force cases,” precisely because liability for excessive force depends so heavily on the particular facts of each individual case. A civil suit against an officer for using excessive force therefore cannot proceed unless, as the Court puts it, “existing precedent ‘squarely governs’ the specific facts at issue.” Generally that means there must have been a previous court decision condemning very similar police conduct in very similar circumstances.³²

We really do not have rules for police uses of nondeadly force; instead we mostly have highly deferential, after-the-fact, case-by-case review, allowing a finding of liability only if prior decisions in cases with similar facts made it crystal clear that the force used was unconstitutional. The nationwide protests against police violence in 2020 prompted renewed scrutiny of qualified immunity, and the Democratic leadership in the House of Representatives introduced legislation to abolish it. Even if legislation of this kind proves successful, however, the legal standards that govern the use of force by law enforcement officers will remain remarkably minimal. We emphatically do not have, in criminal procedure, what we have in substantive criminal law: an understanding of violent conduct as a category apart, deserving of an especially aggressive legal response.³³

I have been focusing on judicial decisions, and especially on rulings of the Supreme Court, but the insignificance of violence in criminal procedure is a broader phenomenon. It is a feature of the legal system as a whole, an aspect of contemporary legal consciousness. It is reflected and reinforced, for example, in how police departments themselves regulate the use of force by their officers. Police departments do regulate violence by police officers, and they pay a good deal more attention to it than they did, say, in the 1950s and early 1960s. Furthermore, many departments tightened their rules, if only by degrees, in response to the nationwide protests against police violence in 2020.

Most departments approach the issue of police violence through a “use of force continuum.” The details vary, but every use-of-force continuum arranges force along a spectrum, starting with mere presence, and then progressing through verbal instructions and steadily increasing degrees of physical coercion. The continuum ends with deadly force. Officers are supposed to start at the bottom of the continuum and work their way upward, going only as far as necessary to get compliance.³⁴

The most important feature of a use-of-force continuum is not the substitution of “force” for “violence,” although this nomenclature certainly matters. Actually, some departments even avoid the language “use of force,” preferring the still more euphemistic phrase “response to resistance.”³⁵ I will return to questions of terminology later in this chapter. The most important thing about a use-of-force continuum, however, is that it is a continuum. When it comes to police tactics, there is no sharp line dividing violence from nonviolence. There are more forceful tactics and less forceful tactics, but they are arranged on a continuous scale. The whole point of a continuum is that we are dealing with differences of degree, not of kind. It is the opposite of the categorical distinction drawn in substantive criminal law between crimes of violence and nonviolent offenses.

The “continuum” approach is reflected in legal scholarship, as well. Stuntz was unusual in arguing for more rules addressing police violence. Most legal scholarship on policing focuses on what the Supreme Court has focused on: invasions of privacy. In recent years a few scholars have placed the issue of police violence front and center, but even they, often as not, share the central idea of the use-of-force continuum. Rather than treat violence as a special category of police conduct deserving special rules, they stress the need to regulate and restrict low-level interventions by law enforcement, the entry points of the continuum, in order to minimize the opportunities for the use of force to escalate.³⁶

The Ebb and Flow of Public Concern about Police Violence

Part of the reason criminal procedure pays so little attention to violence is that it focuses on privacy instead. Another part of the reason, though, is that the level of public concern about police violence has fluctuated. As Figure 4 suggests, concerns about police violence soared in the 1960s but receded in the 1970s and 1980s. In recent decades the issue of police violence has once again started to receive increased attention, but the trend has been slow and uneven, reflecting the conflicting intuitions Americans have about uses of force by law enforcement.

There is a long history of complaints about police brutality and efforts to rein in violence by the police, but there is also a long history of looking the other way, and of vocal defenses of tough, aggressive police tactics. Progressive reformers of the late nineteenth century and early twentieth century, for example, took aim at police corruption but often saw nothing wrong with encouraging officers to use their nightsticks liberally against criminals and reprobates. Newspapers praised “beneficial clubbing” by the police. Many reformers linked police corruption with toleration of vice; they wanted the police to get tough with lawbreakers. This was notably true of Theodore Roosevelt during his tenure on the New York City Police Commission; “speak softly but carry a big stick” wasn’t just a figure of speech. Roosevelt opposed restrictions on the use of nightsticks, and under his leadership the Police Commission gave police officers new, more powerful weaponry, and plenty of leeway to use it.³⁷

Outside of New York, as well, early twentieth-century Americans often explicitly called for “pugilistic police.” Wesley Oliver notes that reformers of the time tended to champion “appropriate violence” in law enforcement. They took the view “that police violence was not necessarily a bad thing so long as it was directed against the criminal element.” This was a common opinion during Prohibition, as well, when there were explicit calls for the police to “muss up” gangsters. The approach always had critics, and the Wickersham Commission, with its attacks on the “third degree,” reflected and strengthened public opposition to illegal—or “extralegal”—police violence. But support for “strong-arm” tactics by the police never went away. That is why, for example, William Westley found in the early 1950s that police officers in Gary, Indiana, believed that the public expected and wanted them to beat up sex offenders.³⁸

Police violence attracted lots of attention in the 1960s, in part because violence in general attracted lots of attention in the 1960s, and in part because of the role the police played in the era’s distinctive politics. Law enforcement was the face of the state. For student protesters and other left-wing activists, police officers personified the Establishment, and police violence was simply the domestic manifestation of the violence the United States government was carrying out in Indochina. So, too, for African Americans in the impoverished neighborhoods of America’s inner cities, the police often seemed, as James Baldwin put it, like “occupying soldier[s] … at the very center of the revolution now occurring in the world.”³⁹

Police officers in the United States in the 1960s were overwhelmingly white, overwhelmingly male, and overwhelmingly conservative, culturally as well as politically. These demographics heightened the insularity of the police, pushed them toward reactionary politics, and made it easier for minorities and antiwar activists to see the police as the enemy. Furthermore, the police often reacted clumsily and heavy-handedly to political protests and the wave of urban riots in the late 1960s, which reinforced the perception that they were violent and out of control. More and more people were afraid of the police, and what they feared mostly was police brutality: “the gun in the holster, and the swinging club.” When Baldwin wrote that he knew from firsthand experience how it felt to be “at the mercy of the cops,” he meant that in a very tangible, bodily sense: he knew “the thunder and fire of the billy club, the paralyzing shock of spittle in the face, and … what it is to find oneself blinded, on one’s hands and knees, at the bottom of the flight of steps down which one has just been hurled.” When antiwar protesters chanted “the whole world is watching” outside the 1968 Democratic Convention in Chicago, what the world was watching was the protesters getting pummeled and teargassed by the police. There were echoes of that chant fifty-two years later, during the protests over police killings in 2020. Once again, moreover, heavy-handed treatment of protesters by the police reinforced many of the concerns that had motivated the protests in the first place.⁴⁰

Unsurprisingly, then, much of the criticism of the police in the 1960s, and much of the energy of police reformers at the time, focused on limiting and controlling police violence. When civilian oversight boards for law enforcement agencies began to appear in the 1960s and 1970s, they were aimed, above all else, at the problem of police brutality. Scholars who wrote about policing focused on violence, too. The criminologist Jerome Skolnick called his widely influential 1966 study on the police Justice without Trial, placing at the very center of his account the ability of the police to inflict punishment outside of any legal process. William Muir, a political scientist who published a celebrated analysis of policing in 1977, thought the central challenge of law enforcement was to be wise and humane when employing coercive force—that is, when resorting to or threatening the use of violence. Contemporary with Muir’s research, a team of academics led by the criminologist Hans Toch explored the possibilities for involving rank-and-file officers in the design and implementation of police reform; the problem the group chose to attack, naturally enough, was violence between officers and civilians.⁴¹

But public attention to police violence waned in the 1970s and 1980s, for several reasons. First and foremost was the growing concern about violent crime, which made many people tolerant of, or even eager for, “rough tactics” by the police. Just as in earlier eras, there was a growing sentiment that it was time for the police to take their gloves off. Gun violence associated with Prohibition Era gangsters led prosecutors like Thomas Kane to defend the use of the “third degree” during interrogations. The spiraling crime rates of the 1970s and 1980s made a box-office hero of Dirty Harry Callahan, the rogue, gun-happy officer played by Clint Eastwood in five movies between 1971 and 1988. The Dirty Harry movies were part of a whole wave of escapist entertainment in the 1970s and 1980s that celebrated vigilantism, sometimes by rule-bending police officers and sometimes by crime victims who had been pushed too far. These on-screen stories reflected a broad appreciation for fighting fire with fire: using justified violence, whether legal or illegal, to avenge unjustified violence and to bring it under control.⁴²

Rising crime rates were not the only reason concerns about police violence receded in the 1970s and 1980s. The police themselves became less frightening to many Americans. Litigation over hiring and promotion policies forced many metropolitan police forces to become less monolithically white and male during this period. The number of minority officers and female officers rose dramatically. So did the number of minority and female police chiefs, albeit more slowly. Civilian oversight boards, which police departments initially fought tooth and nail, gradually became commonplace. And the community policing movement changed the way police departments all across the country thought about and talked about themselves. Rhetoric about the “thin blue line” gave way to calls for partnership with the community. Police departments grew less insular, less defensive, and more sophisticated about and responsive to the diverse communities they served.⁴³

In the process, the police also became better at controlling crime. But crime rates did not begin to drop until the 1990s, and it was even longer before fear of crime began to recede. In the mid to late 1970s and throughout the 1980s, public attitudes toward the police were shaped both by a sense that crime, and particularly violent crime, was getting worse and worse, and by a sense that the police were getting smarter, fairer, more open, and more diverse. These perceptions were based in reality. Crime, including violent crime, really did spiral upward in the 1970s and 1980s, and there were important respects in which the police really did improve.

By the 1990s and early 2000s, all of this had led in many quarters, particularly outside minority communities, to a certain degree of complacency about the police. There was a sense that law enforcement had been broken but it was now largely fixed, or at least that we knew how to fix it. The problem of police brutality, in particular, dropped off many radar screens. That issue periodically reemerged as a focus of national debate, after events like the beating of an African American motorist named Rodney King by Los Angeles police officers in 1991 and the fatal shooting of a Guinean immigrant named Amadou Diallo by New York City police officers in 1999. Increasingly, though, progressive police executives tended to think that the problem of policing had been solved, that all that remained was implementation. Plenty of people outside policing thought that, too. It was common for scholars of law and public policy, for example, to use the reformed police departments of the 1990s and early 2000s as models for how other governmental agencies should be overhauled. Scholarly criticism of the police never disappeared, of course, but a remarkably broad consensus emerged that, by and large, law enforcement agencies were on the right course.

Today it’s difficult to find scholars, reformers—or, for that matter, police chiefs—who are so sanguine about the state of American policing. The bleaker outlook is largely the result of three controversies that have dogged police departments over the past decade. The first has to do with complaints about stop and frisk, particularly in New York City, complaints that have belatedly brought attention to the violence often associated with “frisks,” particularly when the suspects are young men of color. The second has to with the militarization of policing: the increased use of military-style equipment and tactics by police departments. The third, and most important, is the renewed concern that the Black Lives Matter movement brought to the issue of police killings, the concern that exploded across the country in 2020. What all three of these issues have in common is that they sit at the intersection of concerns about racial bias in policing and concerns about violence in policing. And all three received relatively little notice until the second decade of the new millennium.

This is particularly true of the latter two issues, law enforcement militarization and police killings. The militarization of policing over the past half century began with SWAT—“special weapons and tactics”—teams, which large departments began forming in the 1970s. Initially these units were reserved for rare but highly volatile situations: riots, hostage takings, barricaded suspects, and so on. Over time, though, SWAT teams began to be used heavily in drug searches, and they became something that even smaller departments thought they needed. The “war on drugs” in the 1980s accelerated the expansion and repurposing of SWAT teams, and it gave new plausibility to the idea that police should look and act like warriors. Then the federal government got involved, donating surplus military equipment to police departments and, especially after the terrorist attacks of September 11, 2001, giving them money to buy advanced weaponry and other battlefield equipment. The Obama administration imposed some restrictions on the use of federal funds to supply local police departments with certain kinds of military equipment. The restrictions were mild: they applied largely to categories of equipment that police departments had not been receiving. But the Trump administration rescinded them anyway. Military-style police equipment was prominently employed during the protests and looting in the spring of 2020, including an infamous use of tear gas against protesters near the White House so that the president could pose with a bible in front of a damaged church.⁴⁴

SWAT teams and other militarized forms of policing have always been used disproportionately in communities of color, which is one reason the violence associated with them often went under the radar. This was true, as well, of police killings. In the 1990s and early 2000s, few scholars or reformers concentrated on the use of deadly force by the police. Even Paul Chevigny, a stalwart campaigner against police brutality in New York City since the 1960s, believed by the 1990s that “the principal problem of police violence for the present day” was nondeadly force. Throughout the first decade of the twenty-first century, police shootings received far less press coverage and far less scholarly attention than official executions, even though statistics maintained by the federal government indicated that more than ten times as many people were killed each year by police than by applications of the death penalty. And the official statistics, it turns out, undercounted police killings and downplayed the reasons to worry about them—omitting more than half of the deaths, and classifying all of the rest as “justifiable homicide.” About a thousand people are killed by police every year in the United States, vastly more than in any country to which we would care to compare ourselves. This is true even after adjusting for population: the odds of getting killed by the police are roughly five times higher in the United States than, for example, in Canada, Australia, England, Italy, or Germany.⁴⁵

The victims of police shootings, moreover, are disproportionately African American. Blacks comprise roughly 12 percent of the American population, but more than a quarter of all people killed by the police. As with the use of military equipment and tactics by the police, it is hard to escape the conclusion that the problem of police killings received relatively little attention throughout the 1990s and early 2000s in part because the problem was experienced disproportionately by people of color. Police killings did not become a subject of national discussion until late 2014, and what brought attention to the issue then, more than anything else, was the Black Lives Matter movement and especially the protests held in the wake of the deaths of Michael Brown, who was shot by a police officer in Ferguson, Missouri, and Eric Garner, who was suffocated by a police officer on Staten Island. The protests turned police killings from a largely forgotten issue into what the Associated Press called “the top news story of 2014.”⁴⁶

The Black Lives Matter movement didn’t come out of nowhere. It was made possible by two technological developments: cellphones with video cameras, and social media. But it also was the product of political and intellectual developments. It was the coming of age of a new generation of racial justice activists, and—more importantly for our purposes—it reflected a reorientation of thinking about race, a renewed focus on the tangible, corporal complaints of people of color, the way that racism operated not just as a system of abstract advantages and disadvantages, not just as a demeaning series of insults to status, but as a history of violence, a history of assaults and trespasses on physical bodies, often at the hands of the police. This is why Paul Butler, for example, argues that the damage done by stop and frisk is “more like police brutality than racial microaggression.” It is why the sociologist Nikki Jones emphasizes how police violence shapes the lived experience and the outlook of minority youth. And it is why Ta-Nehisi Coates stresses the “visceral experience” of racialized policing and of racism more broadly, the way that it “dislodges brains, blocks airways, rips muscle, extracts organs, cracks bones, breaks teeth.”⁴⁷

Force, Brutality, and Theories of Police Violence

Police violence usually isn’t called violence. It is usually called either “use of force” (the phrasing employed by the police and those sympathetic to them) or “police brutality” (the language used by their critics). Both terms are worth unpacking, because they can influence how people think about police violence: how much to worry about it, the pathways that produce it, and the best ways to control it.

The reason the police don’t describe what they do as “violence” is that the term has a pejorative connotation. It usually, although not always, is understood to refer to the illegitimate or unjustified use of force. So describing police conduct as “violent” can suggest that it is necessarily wrongful. As I suggested in Chapter 1, this is the strongest argument for defining violence differently: for keeping the concept of violence value-neutral. Making wrongfulness part of the definition of violence makes it impossible to discuss, for example, whether and when the violence of war is justified—or about the circumstances under which we want the police to be violent.

That difficulty is averted by using the term “use of force” rather than the term “violence” to describe what the police do when they tackle a suspect, break through a door, or fire their weapons. But “use of force” has a euphemistic quality. It sounds measured and methodical, rather than explosive and hard to control. It downplays the elements of danger and destruction, the “thunder and fire.” “Use of force” is not by definition legitimate and justified; no one thinks that “wrongful use of force” is a contradiction in terms. But the language of “force” does tend to make police violence seem less frightening and troubling than it otherwise might be. This is even truer of the phrase that some departments now prefer, “response to resistance.”

“Police brutality,” a term that opponents of police violence have used since the mid-nineteenth century, has different problems. It is unmistakably pejorative: no one talks about “legitimate” or “justified” brutality. In fact, police brutality is usually defined as the use of unlawful violence by law enforcement officers. The bigger problem with the phrase “police brutality” is the suggestion it conveys that violent officers are acting not as human beings but as beasts. It is similar in this way to calling violent criminals “animals” or calling an impoverished urban neighborhood a “jungle.” Historian Marilynn Johnson points out that “in highlighting the animal-like traits of policemen—most of whom came from working-class backgrounds”—Progressive Era critics of “police brutality” tapped into and reinforced elites’ long-standing tendencies to think of laborers as “bestial or subhuman.” She notes that this kind of thinking proved remarkably persistent, “with middle-class radicals of the 1960s casting police as ‘pigs’ and ‘brutes.’ ”⁴⁸

Class bias is a long and troubling feature of criminal procedure law, and it is not restricted to discussions of police violence. Complaints about government searches in the Colonial Era, for example, routinely appealed to class privilege; officers were denounced as “dirty,” “insolent,” “impertinent,” “rude,” and so on.⁴⁹ But the rhetoric of “police brutality,” like the wider practice of comparing police to animals, does not just encode class prejudice. It also may subtly shape how the legal system thinks about the origins of police violence and the best ways to control it.

Describing violence as brutality, or describing police as “brutes” or “pigs,” reinforces a view of police violence as characterological rather than situational: the consequence of individual officers’ dispositions rather than the circumstances in which they find themselves. And that is a deeply entrenched tendency. It has long been common to trace all forms of police illegality, not just unlawful violence, to rogue officers—“bad apples” that spoil the barrel. That intuition helps to explain the Supreme Court’s strong reluctance to penalize officers for “good faith” violations of the law, either with evidence suppression or with money damages. The Justices have steadily expanded the “good faith” exception to the rule calling for suppression of evidence obtained in violation of the Fourth Amendment—the principal means of enforcing constitutional restrictions on the search-and-seizure powers of the police. And, as we have seen, the Court has immunized police officers from civil suits based on constitutional violations, as long as they acted in the reasonable, good-faith belief that their actions were lawful. Part of the rationale for the good-faith exception to the exclusionary rule has been skepticism that “innocent” mistakes by the police can be deterred. But the Court has also reasoned that the exclusionary rule should be reserved for “flagrant” violations, and shouldn’t be something that “conscientious … responsible law enforcement officers” need to worry about.⁵⁰ Similar considerations lie behind the rule that officers cannot be sued for constitutional violations if they acted in reasonable good faith, a rule that the Supreme Court has stressed is “particularly important in excessive force cases.”⁵¹

Police departments and local governments do not enjoy this kind of immunity. They can be sued for constitutional violations by the officers they employ, even when the officers acted in good faith, if the violations can be traced to an official policy or established custom.⁵² Likewise, the federal government can sue local police departments that have a “pattern or practice” of violating constitutional rights.⁵³ That statutory authority—created by Congress in 1994—has been used against excessively violent departments, but only sparingly, and the Trump administration all but renounced it. Moreover, the pervasive understanding of unlawful police violence as “brutality” colors even efforts to hold departments and local governments liable for condoning excessive violence, and it has shaped the kinds of structural reform sought by private plaintiffs and, occasionally, by the federal Department of Justice. Those lawsuits have sometimes faulted how officers are trained regarding the use of force, but they have placed much greater emphasis on systems for identifying and tracking “problem officers.” The goal has been to get police departments to do a better job identifying their bad apples—and then either rehabilitating them or taking them off the streets.

Activists and scholars have often criticized this focus on bad apples. The problem, they have argued, is not isolated officers but the occupational culture of law enforcement—a culture of paranoia, intolerance, and above all toughness. Police brutality, these critics have suggested, is the product not of violent officers but of a violent system, a violent mentality. Sometimes that mentality is blamed on what working as a police officer does to a person; sometimes it is blamed on the kinds of people who become police officers. Either way, blaming the mentality or culture of law enforcement for police brutality is itself a kind of characterological explanation. It replaces rogue officers with rogue forces, or a rogue profession, but it still finds the root of police violence in the police themselves, in their dispositions and proclivities, not in the circumstances in which they find themselves.⁵⁴

Still, characterological understandings of violence have not dominated recent criminal procedure law and discussions of police violence in the way they have increasingly dominated substantive criminal law and discussions of violent crime. Police departments and their defenders have focused very much on circumstances in their discussions of police violence. They have identified the origins of police violence in the dangers and difficulties that officers face. Much of an officer’s day may be filled with tedium, but the stresses of the job are undeniable. And those stresses are greater in the United States than in Europe, or in Britain or Canada or Australia, if only because there are so many more guns in the United States. American officers are much more likely to get shot. And even when suspects do not fire at the police, even when they are unarmed and do not appear threatening, the danger of gunplay lurks in the background of every encounter officers have when patrolling or responding to a call.

Policing in the United States is less dangerous than it used to be. The death rate of officers on the job fell by roughly two-thirds in the last quarter of the twentieth century, partly because of the widespread adoption of Kevlar. But the odds of suffering a lethal assault on the job still are more than ten times higher for police officers in this country than in, for example, England or Germany.⁵⁵

Here, as elsewhere, rival theories of violence capture separate, partial truths. It is a question of what you want to emphasize, and for what purpose. Some officers are more violent than others. Some departments are more violent than others. Programs targeting problem officers, and problem departments, are sometimes effective in reducing violence. So there is in fact a characterological side to police violence. But there is a situational side, too. That is one reason why even officers caught on video beating up or shooting a suspect can often defend themselves successfully against charges of brutality: the circumstances in which the police operate do help to explain at least some police violence, regardless of whether they excuse it.

There has been a missing piece, though, in many discussions of police violence. Critics of police brutality have emphasized the characterological explanations of violence; the police and their defenders have traced police violence to the circumstances that officers encounter. The missing piece has to do with the responsibility of the police for the situations in which they find themselves: the tactics that place officers in positions where they feel called upon to use force. Good police tactics help officers avoid those positions and give them time to reflect. Good police departments train their personnel in those tactics and require their use. But many departments have not trained their officers in these tactics, and even more departments have lacked policies requiring their use.⁵⁶

A good deal of police violence can be attributed in part to the training that officers receive, and the tactics they employ in the hours, minutes, and seconds leading up to the use of force. But courts—including the US Supreme Court—generally do not hold officers or their departments accountable for the tactics that put officers in situations where they are called upon to use force. Fourth Amendment law, as interpreted by the Supreme Court and as applied by lower courts, focuses instead almost exclusively on the moment in which an individual officer employs violence, and courts routinely declare their reluctance to second-guess the “split-second” decisions officers make “in circumstances that are tense, uncertain, and rapidly evolving.”⁵⁷

When courts assess the “reasonableness” of police violence, why do they place so little weight on the training officers receive, and on the tactics an officer employs early in an encounter? Part of the reason is a series of procedural rules the Supreme Court has established for civil rights lawsuits, which make it easier for plaintiffs to focus their arguments on the actions of individual officers rather than on a department’s policies or general practices, not just in cases alleging brutality, but in all cases challenging the constitutionality of police conduct. Liability can be based on poor training, for example, only if the defendant can be shown to have been “deliberately indifferent” to constitutional rights, and courts typically will not even take up the issue of departmental or municipal liability until an underlying constitutional violation by an individual officer has been shown. Another part of the reason, though, is the ingrained habit of looking only at the moment in which force is used: what law professors Brandon Garrett and Seth Stoughton call the “ ‘split second’ theory of policing.” And that theory, in turn, can be traced partly to the nature of the debate over the causes of police violence. One camp has emphasized characterological explanations, focusing on either occupational culture or problem officers; the other camp has emphasized the circumstances that lead officers to use force. Too often what has been crowded out is attention to how police get themselves into those circumstances, and the training and tactics that could help officers avoid them—or, as a last resort, navigate them with less risk to life or limb.⁵⁸

Here is an example of the kind of issue that gets crowded out. Most of the roughly one thousand people shot to death by the police in the United States each year are carrying guns, but roughly 15 percent—about 150 a year—are armed only with knives. Police officers are almost never killed with knives. Moreover, police in England or Europe almost never shoot anyone armed only with a knife. The reason so many people armed only with knives are killed by the police in the United States comes down to training and tactics. Police overseas are trained to “contain and negotiate” when facing suspects armed with knives. In the United States, in contrast, police officers have been told for years about a “21-foot rule”: once a suspect with a knife is within 21 feet of you, you won’t have time to draw and fire your gun before you are stabbed.

There is zero science behind this “rule.” It was devised by a police trainer who timed how long it took officers to unholster and shoot their weapons (one and a half seconds, he concluded), and how fast a typical man can run in that time (21 feet, he found). He published his results in a trade magazine, and they quickly became part of the lore of American policing. Criminologists and police executives have repeatedly dismissed the 21-foot rule as a poor guide for officers in the field. Even the trainer who came up with the rule has disavowed it as anything more than a rough heuristic. Nonetheless, belief in the rule remains common among officers, and the rule is still often taught in police academies. Moreover, no police department has been found liable for failing to disabuse its officers of the rule, partly because of the high bar set by the “deliberate indifference” standard, partly because of the requirement that plaintiffs first prove a constitutional violation by an individual officer, and partly because the very fact that the rule is so widely relied upon makes it seem “reasonable” to many courts.⁵⁹

Comparing Ideas about Violence in Criminal Law and Criminal Procedure

Substantive criminal law and criminal procedure law have both been shaped by ideas about what causes violence. As we saw in Chapter 2, modern criminal law takes a strongly characterological view of violence: it reflects the assumption that violence is driven more by individual proclivities than by circumstances, that violent offenses are committed mostly by violent people. That is why a person needs to commit so few violent offenses—often one will suffice, and two or three is almost always enough—to be branded as a violent offender for purposes of sentencing enhancements, restrictions on parole eligibility, and exclusion from drug courts, veterans courts, and similar diversion programs. The strongly characterological focus of modern criminal statutes is something of a break with tradition. Two signal features of Anglo-American criminal law—the actus reus requirement, and the rule against character evidence—reflect a long-standing wariness about characterological explanations of criminality. There are few traces of that wariness in modern criminal statutes targeting violent crime.

Modern criminal procedure law is also shaped by ideas about the nature of violence, but the commitment to characterological explanations here is far less complete—partly, perhaps, because the image of police brutality is less racialized than the image of violent crime. In the public mind, the face of violent crime is the face of a young man of color, and offending by members of racial minorities is more readily attributed to character rather than circumstances. But most police officers are white. Law enforcement is far more integrated than it was half a century ago, but minorities remain underrepresented.⁶⁰ Class prejudice does shape the public perception of police officers—and, even more so, of violent police officers. That kind of prejudice also lends credence to characterological explanations of violence, but it doesn’t operate as strongly as racial bias.

That may be one reason there is a powerful counter-narrative to accounts that tie police brutality to the character of individual officers, to the culture of particular departments, or to the occupational mind-set of law enforcement. The counter-narrative traces police violence to the circumstances in which officers find themselves. Because this counter-narrative is generally advanced by the police themselves or by their defenders, it tends to gloss over the responsibility of the police themselves for the circumstances in which they find themselves—which is to say, it downplays the importance of police training and tactics. For their part, campaigners against police brutality also often slight those factors, because their attention is focused on matters of individual character, departmental culture, and occupational mind-set.

That is not the sharpest contrast, however, between how ideas about violence operate today in criminal law and in criminal procedure. The sharpest contrast has to do with the significance of violence. Until the early 1960s, criminal statutes placed little weight on the distinction between violent and nonviolent offenses. Since then, though, that distinction has become criminal law’s master divide, the most common proxy for identifying the worst offenses and the defendants least deserving of mercy. Nothing like that has happened in criminal procedure, or in our thinking more generally about the police. If anything, violence matters less in criminal procedure today than it used to. Some of that can be traced to the way in which privacy has crowded out other values in criminal procedure, some of it to the idea that the police need to be violent to combat violent crime, and some of it, perhaps, to the habit of referring to police violence simply as “force” or “response to resistance.” It is a mistake to make too much of violence, but it may be worse to fail to even name it.