The Law on Police Use of Force in the United States

Abstract Recent events in the United States have highlighted the fact that American police resort to force, including deadly force, much more often than in many other Western countries. This Article describes how the current regulatory regime may ignore or even facilitate these aggressive police actions. The law governing police use of force in the United States derives in large part from the Fourth Amendment to the United States Constitution, which prohibits unreasonable searches and seizures. As construed by the United States Supreme Court, the Fourth Amendment provides police wide leeway in using deadly force, making custodial arrests, and stopping and frisking individuals. While state and local police departments can develop more restrictive rules, they often do not. Additionally, the remedies for violations of these rules are weak. The predominant remedy is exclusion of evidence, the impact of which falls primarily on the prosecutor and in any event only has a deterrent effect when evidence is sought. Civil and criminal sanctions have been significantly limited by the Supreme Court, particularly through the doctrine of qualified immunity (applied to individual officers) and the policy or custom defense (applied to municipalities). This minimal regulatory regime is one reason police-citizen encounters in the United States so often result in death or serious bodily harm to citizens, in particular those who are Black. The Article ends with a number of reform proposals.

Rayshard Brooks, killed by police in Atlanta, Georgia; 4 Tamir Rice, killed by an officer in Cleveland, Ohio; 5 and Michael Brown, killed by an officer in Ferguson, Missouri. 6 There are hundreds more such recent incidents, and they occur in every area of the country. 7 Of course, the use of deadly force by police is sometimes justifiable. But when the triggering circumstances are ambiguous, Black people are far more likely to be killed by the police than white people. 8 Indeed, police use of force is a leading cause of death for Black men in the United States. 9 Then there are the incidents in which American police use excessive force short of killing someone, undoubtedly numbering in the tens of thousands every year. 10 Adequate data tracking such incidents does not exist, but the available evidence suggests that officers use some type of force over 900,000 times per year. 11 Manifest racial disparities exist here as well, whether the police conduct involves stopping a person for a traffic violation, questioning and frisking individuals, or arresting an individual. 12 At every level of policing in the United States, one sees a massive amount of policecitizen interaction, visited disproportionately on Black people, other minorities and the poor.
There are many explanations for this situation, including embedded racism, a high violent crime rate, tough-on-crime politics, residential segregation, a macho-gun culture, a "warrior mentality" among police, poor officer training, and inadequate screening of police recruits. 13 This Article will focus on the law's complicity for the current state of policing in the United States, specifically the complicity of American constitutional law. As applied to police use of force, that body of law is anemic. It gives the police tremendous discretion in deciding when to initiate encounters with citizens and what happens when they occur. There is no comprehensive set where there are no circumstances that would render a shooting objectively reasonable for the purposes of the Fourth Amendment, Black suspects are more than twice as likely than other suspects to be killed by police); Li Cohen, It's Been over 3 Months Since George Floyd Was Killed by Police. Police Are Still Killing Black People at Disproportionate Rates., CBS NEWS (Sept. 10, 2020), https://www.cbsnews.com/news/george-floyd-killing-police-black-people-killed-164/ (finding that approximately eighty percent of the White people killed by police were allegedly armed at the time of the incident, whereas approximately seventy percent of Black people killed by police were allegedly armed). This Article also found that White people killed by police were unarmed approximately 8.7% of the time, whereas Black people killed by police were unarmed roughly 12.8% of the time. , https://www.usatoday.com/in-depth/news/ investigations/2019/04/24/usa-today-revealing-misconduct-records-police-cops/3223984002/. This story reported that USA Today discovered over 200,000 allegations of police misconduct-involving over 85,000 officers during a ten-year period -more than 22,000 of which had to do with excessive force. Id. The records were obtained from only 700 of the 18,000 police departments in the country and reflect only those cases in which an official complaint was made. Id. of statutes regulating police conduct like that found in other countries such as Germany. 14 As a result, unless their actions are clearly egregious, American police are pretty much free to do what they want, without accountability in the courts.

11
B. The Sources of Police Regulatory Law Most American states have statutes regulating police use of force, and some have statutes regulating other police-citizen encounters. But until the recent spate of deadly force cases, almost all of them simply mimicked the language of judicial case law. 15 To an unusual extent, compared to the rest of the world, the law governing the police in the United States comes from the country's Constitution, as interpreted by the United States Supreme Court and the lower federal courts. This was not always true. Until the 1960s, the provisions in the U.S. Constitution governing searches and seizures, the right to silence during interrogations, and the right to counsel during interrogations and identification procedures applied only to federal officers. But beginning in the 1960s, the federal judiciary engineered a constitutional criminal procedure revolution. The Supreme Court nationalized rules governing warrants, stop and frisk, interrogation, and lineup procedures, using language in the Fourteenth Amendment prohibiting the states from depriving people of liberty without "due process of law." 16 Since then, the Court has issued hundreds of rulings that regulate these police investigative techniques, applicable to every police department in the country. 17 One reason the Supreme Court decided to become so heavily involved in regulating the police is that, unlike policing in many other countries, policing in the United States is very decentralized. The country has over 18,000 police departments-a holdover from colonial times when fear of European-style authoritarianism led to resistance not only to a federal law enforcement body, but even to state-run police agencies. 18 The federal legislature (Congress) has never presumed to regulate local departments, although it has the authority to do so under the Fourteenth Amendment. 19 Thus, the role of laying out nationally-applicable rules for the police has fallen to the federal courts. As Francis Allen pointed out, the impetus for the judicial activism of the 1960s was derived in large part from the fact that fragmented American polities "have not been ingenious in devising institutions that subject criminal justice functions to scrutiny and test." 20 A related and even more central reason that policing law in the United States is constitutionalized is race. As Herbert Packer noted, "the most powerful propellant" of Supreme Court lawmaking in the policing domain was the gross abuse of power by law enforcement officers in the South, usually aimed at Black citizens. 21 According to Packer, writing in the 1960s: "Police brutality, dragnet arrests, and discriminatory official conduct may be debatable issues in the cities of the , http://www.bjs.gov/content/pub/pdf/nsleed.pdf (estimating that approximately 18,000 police forces cover ninety-eight percent of the U.S. population). Even in Germany, where reaction to Nazism has led to an insistence on separate police forces in each of the states, there is centralization within the sixteen states to an extent that does not exist in the U.S. 19 Section 5 of the Fourteenth Amendment authorizes Congress to pass statutes enforcing the Amendment's other provisions, which include a prohibition on depriving people of life, liberty, or property, without due process of law. U.S. CONST. amend. XIV, § 5. dozen other Southern Communities." 22 With Congress moribund, and with local legislatures and courts silent about-if not sympathetic to-police malfeasance, the Supreme Court of the 1960s felt it had to enter the legal vacuum in the policing realm, as it had done with school segregation policies in 1954. 23 This dominance of constitutional rules handed down from the courts as the means of regulating the police has several implications. First, judicial rules tend to be broad, general, and vague. Judges and courts are not institutionally equipped to produce nuanced, complex formulations of the type typically found in statutes. Second, the rules are very fact-specific and acontextual. Courts only decide the case or controversy in front of them, which not only presents an isolated set of facts-divorced from social, economic, and political factors that a legislature might consider-but also may be unrepresentative of the most typical law enforcement situations, because hard cases are the ones that typically get litigated. Third, the rules are relatively rigid, in the sense that, because of stare decisis, courts rarely overturn them. Legislatures are nimbler in this regard. Relatedly, while a statute can be passed quickly-as has occurred in a few states with respect to police use of force within weeks of George Floyd's death-it takes time for cases to percolate up to the Supreme Court, with the result that many rulings on the constitutionality of police techniques are finally and definitively pronounced by the Court decades after police began using them.
In theory, administrative agencies can issue more detailed regulations than might be found in caselaw or statutes, with expert input and through an ongoing regulatory and enforcement process. For instance, the Federal Bureau of Investigation develops its own guidelines. Further, the Department of Justice (DOJ) may, as described further below, pursue consent decrees against local departments. But no federal regulatory agency oversees police agencies, even at the federal level. Nor do individual states tend to adopt comprehensive-or even modest-oversight regimes for policing, apart from loose accreditation requirements.

C. Police Accountability Mechanisms
In this Section, we provide an overview of three ways in which police are held accountable in United States law: The exclusionary rule; federal civil rights actions; and criminal and administrative sanctions. We describe how each has proved to be a fairly deferential and ineffectual source of either accountability or guidance for police agencies and officers.

The Exclusionary Rule
A significant consequence of the judicial management of policing is the rule that illegally seized evidence should be excluded. In 1961, the Supreme Court held that other remedies for unconstitutional searches and seizures were inadequate, in no small part because their scope was dependent on often-recalcitrant legislatures or state courts. 24 The "exclusionary rule," in contrast, was a weapon federal courts controlled, and that they could use to enforce their interpretations of the Constitution. Numerous countries countenance judicial exclusion of evidence under various circumstances, especially of confessions obtained illegally. 25 But American courts have been, by far, the most vigorous proponents of the idea that evidence seized during an illegal search must be 22 Id. at 241. 23 Brown suppressed. While the Supreme Court has cut back significantly on the scope of the exclusionary rule in the past four decades, the rule still requires dismissal of a non-negligible number of cases. 26 The preeminence of the exclusionary rule as a means of holding police accountable has several significant consequences for the law of policing in the United States. First, because the rule only operates in the context of a criminal prosecution, courts generally only consider constitutional claims brought by presumptively guilty people who are trying to suppress probative evidence of their guilt. Thus, in the media, the rule is often characterized as a "technicality" that frustrates the public will. More importantly for Fourth Amendment doctrine, judicial decisions about the scope of constitutional protections are affected both by hindsight bias-to wit, "there must have been probable cause, because the police found evidence"-and extraneous political considerations, in particular the consequences of dismissing charges against a clearly guilty person. The pressure to adopt conservative police-and prosecution-oriented decisions in such a setting is significant.
Second, even in those cases where exclusion occurs, the effect on police and police departments may be minimal. The force of exclusion falls mainly on the prosecutor rather than on the police, who may never even know exclusion has occurred. While the advent of the exclusionary rule in the early 1960s correlated with an increase in police training programs about the law, research indicates that, because the only effect of the rule is to exclude evidence, it is often ineffective as a direct police accountability measure. Police knowledge of the relevant law is poor, 27 police violations of the rules are very common, 28 and-according to at least one study-police prefer the rule not only to a damages regime but even to more training, 29 suggesting that the rule has very little impact on their dayto-day conduct.
Third, the exclusionary rule is simply an inapt remedy in many situations. This is particularly so in cases involving claims of excessive force, such as those mentioned at the beginning of this Article. In such cases, either the target is unprosecutable because of death or innocence, or exclusion of evidence-assuming there is any-is simply not commensurate with the harm done. Further, the Supreme Court has declared that the rule does not apply in proceedings other than criminal trials-such as civil, immigration, and parole revocation proceedings 30 -despite the large number of police investigations focused on those types of violations. Nor does it apply when police violate rules that are not considered core protections, such as the knock and announcement requirement, 31 or when police act in good faith reliance on a warrant. 32

Federal Civil Rights Lawsuits
The central federal civil rights statute, 42 U.S.C. § 1983, enacted after the Civil War, and the Supreme Court's decision in Bivens v. Six Unknown Named Agents of Federal Bureau of Narcotics, 33 allow people to sue state and federal officials who violate their constitutional rights. While tending to provide more meaningful remedies for victims of constitutional violations than exclusion, this body of law-as implemented by the courts-makes successful suits against either the officer or his or her employer very difficult. In an odd type of shell game, cities or counties can usually blame the officer, while the officer can usually claim immunity, with the result that no defendant is liable.
Consider first the liability of individual officers. Here, the Supreme Court, concerned that federal lawsuits seeking damages remedies may over-deter and burden the police, has erected a qualified immunity defense, which allows individual officers to avoid liability unless they have broken a "clearly established" rule that was in place at the time of their conduct. 34 Because, for reasons already noted, there are few clear rules, and any given situation may involve different facts, this is a powerful defense. Furthermore, the Court has held that such suits may be dismissed once qualified immunity is found, without having to establish a "clear" governing rule that could apply in future cases. 35 If, instead, suit is brought directly against the police agency, other hurdles must be cleared. State-level agencies-as opposed to police departments run by municipalities-are completely immune from § 1983 actions, for historical reasons. 36 Federal agencies are not immune from suit but may have a good faith defense. 37 While local municipalities cannot rely on that defense, they are only liable if the plaintiff can show that the police conduct was authorized by a policy or endorse a custom that is unconstitutional. 38 There is no respondeat superior liability for municipalities under the federal civil rights statute. Thus, while proof that police acted under instructions from a city official is a sufficient basis for finding municipal liability, 39 a city cannot be found liable for a single incident of "unusually excessive force" unless the plaintiff can show that the city has a policy that on its face authorizes inadequate training, or that the city is in some other way "deliberately indifferent to the rights of persons with whom the police come into contact." 40 Finally, neither federal nor city departments can be sued for punitive damages, a factor that greatly reduces the incentive for plaintiffs' lawyers to file and litigate these complex and charged cases. 41 Similar limitations are placed on attempts to enjoin a police department from engaging in illegal actions under § 1983. In City of Los Angeles v. Lyons, 42 the Court concluded that the plaintiff's allegations-that he had been rendered unconscious when police applied a chokehold to his neck and that police routinely applied such chokeholds to others-fell "far short" of what is necessary to state a claim for an injunction. Rather, Lyons: [W]ould have had not only to allege that he would have another encounter with the police but also to make the 'incredible' assertion either, (1) that all police officers in Los Angeles always choke any citizen with whom they happen to have an encounter, whether for purposes of arrest, issuing a citation or for questioning, or, (2) that the city ordered or authorized police officers to act in such a manner. 43 While this restrictive language is not inconsistent with the persistence requirement found in other injunction settings, it is also explained in part by the Court's reluctance to authorize federal court oversight of state entities. In response to the Lyons ruling, Congress adopted 42 U.S.C. § 14141, which permits the DOJ to challenge unconstitutional policing patterns and practices. The result has been dozens of consent decrees that require police departments to change use of force practices, prohibit racial profiling, and establish training programs. 45 Yet, however important these civil rights settlements have been in achieving more systemic reforms, the DOJ's resources do not permit widespread use of this tool, and some Administrations have been reluctant to wield it at all. 46

Other Accountability Mechanisms
Criminal prosecution of miscreant police officers is also possible and-increasingly, in high profile cases-prosecutors have at least considered whether to bring criminal charges against officers who use deadly force. 47 Even so, the imposition of criminal sanctions on police officers remains rare. First, under the criminal analogue to § 1983 and similar state statutes, the prosecutor must prove beyond a reasonable doubt that the officer had the specific intent to deprive the alleged victim of his or her constitutional rights. 48 Moreover, many states allow a defense if the officer's actions would be permissible under the Fourth Amendment's reasonableness standard. Thus, to take one high-profile example, after a Cleveland police officer shot and killed Tamir Rice, the Cuyahoga County prosecutor concluded no charges should be presented to a grand jury, emphasizing the deference due police under the Fourth Amendment standard 49 and relying on the Supreme Court's stance on qualified immunity in civil cases. 50 Although the constitutional/ non-constitutional and criminal/non-criminal lines are totally distinct, they have often been "very casually" imported from one context to the other. 51 For similar reasons, research suggests that meaningful internal discipline for illegal searches and seizures or interrogations-even those that result in exclusion-is rare, and that civilian review boards, which are often reliant on the cooperation of the police, have had uneven success at monitoring them. 52 Police union contracts, which often incorporate a Law Enforcement Officer Bill of Rights (LEOBR), can also make administrative accountability difficult. Under the typical LEOBR, and sometimes by statute, records of previous misconduct are often protected from disclosure to the public and the courts. 53 Further, interrogation of officers under investigation must be delayed until several days after an incident and be conducted under formal conditions (in contrast to the typical police interrogation of suspects), 54 and suspensions or dismissals are subject to arbitration and often reversed. 55  they inhibit the civil government from intruding into their management, 56 which creates significant tension between the labor and police reform goals of activists. 57 In short, American accountability mechanisms are fairly weak in the policing domain. The exclusionary rule has little impact on policing on the streets, individual officers benefit from immunity doctrines, and the law makes it very difficult to sue a municipality or police agency. That weakness combines with a constitutional regulatory regime that affords police broad discretion in their ability to use force, arrest, and stop citizens.

D. The Law on Use of Deadly Force
The use of deadly force has been a particular focus of litigation, for obvious reasons. In Tennessee v. Garner, 58 the Court held the Fourth Amendment-which prohibits "unreasonable searches and seizures"-was violated by an officer who, in responding to a burglary call, shot the suspect in the back of the head as he was trying to jump a fence, even though the officer was "reasonably sure" the individual was unarmed. But Garner was only a small step forward. It permitted the use of deadly force whenever police have probable cause to believe a person is armed and dangerous and the person has failed to heed a warning to desist from resisting or escaping. There is no requirement that police consider alternatives to the use of deadly force or whether the use of such force is proportionate.
Garner's reach is limited even further because, as is true of most of the American regulatory regime, it derives from the Constitution, which places constraints on the scope of deadly force rules. Although the Supreme Court could have regulated use of force through the Due Process Clause of the Fourteenth Amendment, which prohibits depriving people of life or liberty "without due process of law," the Court instead has decided that the Fourth Amendment's prohibition on unreasonable seizures is the more appropriate source of law in this setting. 59 Given the amendment's language requiring a "seizure," numerous lower courts have refused to find that a constitutional violation occurs when police shoot and miss a suspect, do not physically restrain a person, or physically restrain a person but with no intent to do so. 60 Violence, per se, is not regulated as a constitutional matter.
Furthermore, in the years since Garner, the Supreme Court has refused to set out any more protective deadly force rules, even when the Fourth Amendment does apply. In Graham v. Connor, the Court ruled that, in deference to "an officer's need to make split-second judgments," a court should only look at the reasonableness of an officer's decision at the moment of a shooting; 61 there is no need to take into account whether the officer could have taken obvious steps to avoid using deadly force in the first place. Graham is also "indeterminate" and "confused." 62 It provides no guidance on how to judge whether the conduct was "reasonable," nor does it reference acceptable or required training or tactics. Thus, a rash, poorly-trained officer who killed for no reason might still be found to have been reasonable for a split second, even though a well-trained officer in the same situation might have resorted to voice commands or de-escalation techniques, or spent more time assessing the situation. Police can go with their gut, contrary to training and policy, and not be held accountable. For instance, the Supreme Court has refused to find liability under the Fourth Amendment where an officer shot a man in the back as he tried to drive away from her; 63 where officers, engaging in a high-speed chase, rammed the suspect's rear bumper and caused the suspect to crash; 64 and where officers shot a person they knew was mentally ill who had threatened them with a knife. 65 In each of these cases, the Court held that the officers had acted reasonably in light of established law at the time, and thus had qualified immunity. In one of these cases, the Court stated that there is no "magical on/off switch that triggers rigid preconditions" on the use of deadly force by police. 66 In another case, the Court noted that disregarding training and engaging in "imprudent, inappropriate, or even reckless" conduct leading up to the incident is not of constitutional relevance. 67 The Court continues to emphasize an approach that lacks guidance or principles and adopts the view that deference is good for its own sake.
This use of force doctrine allows police to benefit from several layers of "reasonableness" deference, or "reasonableness on top of reasonableness": [D]ue to the doctrine of qualified immunity, a court may ask at the summary judgment stage whether a (1) reasonable jury could conclude that the police officer conducted an (2) unreasonable search or seizure or even, if so, whether the officer acted objectively (3) reasonably given the circumstances, and in (4) reasonable reliance on clearly established law. 68 As Professors Sam Kamin and Justin Marceau add, this means that: "It is now possible to speak of that famous conundrum of reasonable unreasonable searches [and seizures]-those . . . that are sufficiently unreasonable that they deprive the defendant of his Fourth Amendment right, but not so unreasonable that any remedy will be forthcoming." 69 Some police departments have adopted policies that are more restrictive than required by the Court's decisions, often as part of consent decrees litigated by the DOJ. These policies reflect an improved understanding of the tactics that can prevent interactions from escalating into violence. 70 But departments are not required to take these steps under the U.S. Constitution, and many simply emphasize that officers must act reasonably, taking a cue from the U.S. Supreme Court. 71 Thus, a California statute that would have permitted deadly force only when it was "necessary to prevent imminent death or serious bodily injury to the officer or to another person" was successfully resisted by the California Police Officers' Association, because it "raised the legal use of force standard" above what was required by the Supreme Court. 72 The Court's rulings on the use of force, probably intentionally kept vague because of their national applicability and the 63 Court's inability to anticipate every possible context, have become the default standard governing the use of force.
Further, the impetus for stricter rules that might be provided by consent decrees is diminished by limited resources and by an express reluctance on the part of some Administrations, such as the current one, to regulate local police. 73 That is why many current reform proposals aimed at the police focus on use of force concerns. A popular set of reforms called "Eight that Can't Wait" calls for banning chokeholds and strangleholds, requiring de-escalation tactics, requiring a warning before shooting, exhausting all alternatives before shooting, requiring fellow officers to intervene against an officer who is using excessive force, prohibiting shooting at moving vehicles, adhering to a use of force continuum, and comprehensive reporting of incidents in which force is used. 74 Under current constitutional precedent, in contrast, only the warning before shooting is required, and only when "feasible." 75 Some advocates for the police resist most or all of these reforms on the ground that they will compromise the police's ability to do their job. To be sure, as compared to many countries, the United States only lightly regulates the purchase and possession of weapons-in part because the Supreme Court has interpreted the Second Amendment to provide that individual citizens have a constitutional right to bear arms 76 -and the U.S. also has much higher homicide and police-death rates than other Western countries. 77 That does not excuse the failure of the law to hold police accountable for poor training and rash decision-making.

E. The Law of Arrests
An arrest resulting in custody is the disposition of choice for American police to an extent unimaginable in most European countries. Roughly ninety-five percent of arrestees in the United States are physically detained-that is, handcuffed, taken to the police station, and booked (processed). 78 As a result, the number of U.S. detainees who are confined prior to trial-in the area of twelve million a year and over 400,000 daily 79 -is roughly three times the analogous number in the twenty-seven Member States of the European Union (EU), despite the fact that the EU has 150 million more people than the United States. In Germany, for instance, only three to five percent of arrested individuals are remanded to custody. 80 Alternatives to arrest-such as citations or summonses-are rare in most American jurisdictions, even for minor crimes like public drunkenness, vagrancy, loitering, and putting one's feet up on the subway. 81 Eric Garner was selling "loosies" (single cigarettes) before he was choked to death, and George Floyd was accused of trying to pass a $20 counterfeit bill before the same thing happened to him. Further, as these cases suggest, the power to make custodial arrests is often exercised in a racially disproportionate manner. The DOJ found in its § 14141 investigation following the death of Michael Brown that Ferguson, Missouri's "municipal court and police practices are due, at least in part, to intentional discrimination, as demonstrated by evidence of racial bias and stereotyping of African American residents by certain Ferguson police and municipal court officials." 82 Worse yet, these custodial arrests may be driven in part by fiscal motivations. Fines, fees, bail bonds, and forfeitures can all be a profitable source of income for some communities, 83 and they often have a racially disproportionate impact. 84 Thus, the National Center for State Courts has recommended that: "Courts should acknowledge that their fines, fees, and bail practices may have a disparate impact on the poor and on racial and ethnic minorities and their communities." 85 A principal reason for this state of affairs is the law's willingness to grant the police a large degree of deference in making arrests. Similar to the law of deadly force, arrest law, as pronounced by the Supreme Court, does not concern itself with proportionality. The Court has explicitly declared that, given probable cause, an officer may "make a custodial arrest without balancing costs and benefits or determining whether . . . [the] arrest was in some sense necessary." 86 In one case, the Supreme Court upheld a custodial arrest for violation of a seat belt law, 87 and in another it sanctioned a custodial arrest for driving with a revoked license. 88 Once again, the constitutional provenance of these decisions partly explains them. American courts normally abide by stare decisis and historical practices, particularly when interpreting the Constitution. Thus, centuries-old traditions-in this case, beginning a criminal prosecution with a custodial arrest-will persist in the absence of very good reasons to change them. Further, the constitutional language from which the arrest decisions come only requires that arrests be "reasonable," and implies that a search or seizure is always reasonable if based on probable cause. The result is a law of arrest that is un-nuanced, and a recipe for unnecessary and often denigrating physical confrontations between police and citizens, as the Eric Garner and George Floyd cases illustrate.
The rules governing searches incident to arrest create additional incentives for the police to abuse their discretion. A pedestrian taken into custody is subject to a full search of their person in the field or at the station, regardless of the crime. 89 A driver taken into custody is also subject to such a search; 90 further, the car itself may be searched if there is reason to believe evidence of the crime of arrest will be found, 91 evidence is seen in "plain view," 92 or the driver consents (the latter a frequent occurrence because police do not have to tell drivers they have a right to refuse consent and drivers are often fearful of turning down a police request). 93 82 Arrests associated with traffic laws are particularly prone to abuse. Traffic rules are pervasive; drivers can be arrested when they go over the speed limit, fail to signal, cross a median line, have faulty equipment, or-as in the case noted above-fail to wear a seat belt. Some officers talk of the "three-block rule"-the idea that every driver violates at least one traffic law once every three blocks. 94 That fact, combined with search incident to custodial arrest rules, allows police to stop and search almost any car, a power that can be, and has been, exercised in racially discriminatory ways. 95 The vast bulk of traffic cases brought in many states is astonishing: Over fifty million traffic cases per year. 96 One study found that one in seven drivers in North Carolina had a suspended license due to unpaid or unresolved traffic cases-a fact that provides police with custodial arrest authority over large numbers of drivers. 97 Reformers have suggested moving toward a citation-only system, or using technology rather than police to detect traffic violations and other minor infractions. But those reforms are often resisted on the ground that confrontational policing methods are necessary to maintain order and can sometimes uncover serious crime. The Supreme Court has noted, for instance, that Timothy McVeigh-the man convicted of blowing up the federal building in Oklahoma City-was in custody for a traffic violation when the FBI linked the bombing to him. 98 Another proposed reform is to sanction police use of their arrest power when there is proof that it is used pretextually-that is, as a way to carry out improper or unauthorized agendas. But such proof is hard to come by. In any event, the U.S. Supreme Court has refused to adopt a ban even those arrests that are clearly pretextual, except when they result from racial animus, 99 a mental state that is virtually impossible to demonstrate unless the officer admits to it. 100 The rationale for this stance, once again, is that the Fourth Amendment only requires reasonableness, and an arrest is reasonable if it is based on probable cause.

F. Police Stops
As in many countries, police are permitted to detain people for investigative purposes even when they lack grounds for arrest. This stop, question, and frisk technique was given a constitutional imprimatur by the Supreme Court in Terry v. Ohio, 101 decided in 1968. In Terry and its progeny, the Court held that a person can be subjected to a brief stop and questioned if the police have an "articulable suspicion" that the individual is engaged or will soon engage in crime; they may frisk or pat down the person if, as a result of the questioning, the police suspect the person is armed. The standard that arose out of Terry is called "reasonable suspicion," and is meant to connote something less than probable cause. This lesser justification is sufficient, the Court reasoned, both because-analogous to the German proportionality principle-the intrusion it permits is limited compared to a full-blown arrest, and because the government interest in preventing incipient crime is at stake.
While Terry requires articulable suspicion as a formal matter, this standard is relatively easy to meet as applied. The Court has held or implied that police may stop and frisk a person coming out of a crack house, 102 a person who runs from police in a "high crime" area, 103 and a person who purchases two forty-eight-hour roundtrip tickets from Honolulu to Miami with cash from a large roll of bills and gives the ticket agent a false phone number. 104 More importantly, the police themselves interpret the reasonable suspicion standard to permit them to stop people who seem "out of place" or "look dirty." 105 In some cities, research has shown that less than five percent of stops result in the discovery of weapons or contraband, and that people of color are significantly more likely to be stopped, even when other relevant factors are taken into account. 106 Yet, because exclusion, which is triggered only when there is a prosecution, is the remedy of choice, this low hit rate usually has no negative consequences for the police.
The potential for arbitrary police stops has been magnified by a 2016 Supreme Court decision holding that, even if the police do not have reasonable suspicion when they stop someone, any evidence they find in an ensuing search is admissible if they discover the person has an outstanding warrant, on the ground that the warrant is an "independent legal source" of justification. 107 Since such warrants can be issued for failing to pay parking tickets, traffic violations and other minor infractions, 108 illegal stops can be a tempting ploy for an officer who wants to search someone. Theoretically, the individual could sue for the underlying constitutional violation, but that is unlikely. Further, virtually every state criminalizes the failure to comply with an arrest-even one that is unlawful-so any attempt to resist a warrant-based search because of the arbitrariness of the initial stop will only make matters worse. 109 As a result, police have very little to lose and much to gain from an illegal stop.
This state of affairs not only increases the chance of an unnecessary confrontation with the police, it also exacerbates tensions with the community and, given the racial imbalance of such stops, is particularly toxic in communities of color. Many Black and Latino men who live in poorer urban neighborhoods are routinely stopped, questioned, and frisked. 110 The damage to policecitizen relations, and to race relations more generally, are a primary motivation for the Black Lives Matter movement, the Defund the Police initiative, and other reform efforts.

G. Toward Policing Reforms
In a report entitled Changing the Law to Change Policing: First Steps, the authors and several other legal academics have proposed a number of changes to the law of policing. 111 The report briefly sets forth twenty-eight possible reforms, divided between proposals that should be adopted by Congress and those that are more appropriately implemented at the state level. The ultimate aim of this agenda, however, is to bring about change at the local level.
Not surprisingly, given the fragmented nature of American policing, most of the proposals are aimed at federal entities, in an effort to standardize regulatory efforts. We propose the elimination of qualified immunity (albeit still allowing indemnification for individual officers), removing the policy and custom defense for municipalities, and improving the federal statute (including § 14141) that permits suits to be brought against illegal police and prosecutor patterns and practices. Further, we propose enhanced prosecutions of police officers who willfully violate First Amendment rights (to wit, the freedoms of speech, assembly, and press), federalization of a use of force standard that recognizes proportionality, and reinforcing the constitutional standard that police always announce their presence, even when circumstances justify not knocking.
Additionally the report proposes that the federal government maintain a national database identifying officers who have been de-certified or otherwise officially found to have engaged in misconduct, that states be required to collect certain types of data about de-certified officers and other aspects of policing, and that national standards be established for data that should be collected and maintained. Finally, we propose greater coordination between the many disparate units within the federal DOJ that deal with policing, a stronger focus on public safety-broadly defined-in national funding of research grants and policing programs, and a national accreditation body that can use funds as leverage to assure state and local departments meet minimum requirements.
At the state level, we propose specific reforms regarding the use of force, militarized tactical teams, the use of invasive surveillance technologies, and stop, search, and arrest practices. In the latter category, we suggest, for instance, that stops based on reasonable suspicion be limited to certain categories of crimes, that people be informed they have a right to refuse consent, that consent searches be banned entirely in connection with traffic stops unless reasonable suspicion exists, and that custodial arrests be reserved primarily for serious offenses. We also propose taping all interrogations and implementing explicit rules about when police body camera footage must be made available to the public.
At the institutional level, we propose that state and local departments establish mechanisms for independent review of critical incidents, de-certify wayward police and report such decertifications to the national authorities, reexamine union contracts and LEOBRs with an eye to removing unreasonable limits on police accountability, and collect and publish various types of information about their policies and practices. Finally, we propose that states review their criminal codes and consider whether steps should be taken to decriminalize certain offenses, remove certain custodial penalties, and discourage overcharging, all with the goal of curbing excessive use of the criminal law to authorize stops and arrests. The continuing role of activists, civil society, and non-profits in documenting police misconduct, pushing for transparency, and advocating reform will be crucial to ensure that further progress is made.

H. Conclusion
Policing in the United States is in urgent need of reform, especially given its racially disparate impact. The governing law should be changed with respect to every stage of the street policing process: Pedestrian and traffic stops; custodial arrests; police use of force, including deadly force; and the means of holding police accountable. The Black Lives Matter movement, joined by advocates and policymakers from all points of the political spectrum, have increasingly called for systemic changes. Many of the constitutional rules that we have described set out only a minimum floor and can be supplemented through new statutes and new police practices. Policing needs to move away from its constitutional orientation and toward best practices designed to prevent loss of life, racial discrimination, and abuse; it also needs to foster an inclusive and broader vision of public safety.
We hope that the tragic lessons of decades of racially charged deaths at the hands of police, wrongful arrests, humiliating stop and frisks, manufactured criminal debt, and so many other systematic consequences visited chiefly upon poor and minority citizens will finally result in meaningful change. We are gratified to see activists, lawmakers, and even some courts engage with these issues in a more sustained way in recent years. We hope that the time is ripe for serious proposals that result in deep-seated reform to policing in the United States.