CHAPTER TEN

POLICE: THE THICK BLUE LINE

Freedom is about authority. Freedom is about the willingness of every single human being to cede to lawful authority a great deal of discretion about what you do.

Rudy Giuliani (1994)¹

In January 2011, police in Framingham, Massachusetts, smashed Eurie Stamps’s front door, tossed in a flash grenade, and stormed Stamps’s home. Police had been tipped off that Stamps’s stepson might be dealing drugs from “around” the house and obtained a search warrant, but Stamps himself was not a target, had no history of violence, and police had been specifically informed prior to the raid that Stamps posed no threat. Nevertheless, police ordered Stamps to the floor with his hands in plain view, and though Stamps complied, one officer thought it would be prudent to train his M4 assault rifle (safety flicked off) to keep the 68-year-old black grandfather under control. At some point during the chaos of the eleven-officer raid, Stamps’s guard tripped or fell, unable to manage his cumbersome, military-grade body armor. The officer’s M4 discharged, and Stamps, who had been contentedly watching television in his pajamas moments earlier, was dead. It was a tragedy made more senseless given that the prime target of the raid, Stamps’s stepson, had been arrested outside the house before police broke down Stamps’s door. While police did find drugs on the stepson, and arrested the raid’s second target inside Stamps’s house, Stamps had no drugs, offered no resistance, and had no real connection to the crime other than the misfortune of having the wrong stepchild.

Shocking as the details of Stamps’s saga were, nothing about a SWAT team armed with military equipment to execute a routine search warrant was uncommon. Nor was it uncommon that the district attorney exonerated the officer, or that eight of ten claims in the Stamps family’s subsequent lawsuit were initially dismissed. The only unusual feature of this “unfortunate matter” (to use the words of the town’s lawyer) was that Framingham’s insurance company agreed to settle the litigation, paying Stamps’s family $3.75 million.² Most victims get nothing.

Militarized raids have become a regular part of American policing, with predictably tragic consequences. Special Weapons and Tactics teams were originally designed for extraordinary situations like hostage-taking, but they have become disturbingly ordinary. SWAT teams, unlike hostage takers, are everywhere: by the mid-2000s, 90 percent of large cities had SWAT teams, as did about 80 percent of small towns.³ The police have become omnipresent, miniature armies kitted out with bomb suits, battering rams, night-vision goggles, armored personnel carriers, helicopters, drones, assault rifles, flashbang grenades, .50 caliber machine guns, and other military paraphernalia.⁴ Police receive much of their stockpile from the Defense Department’s 1033 Program, partially subsidized by the Department of Homeland Security. These gifts come with few meaningful restrictions or guidelines, save one: under standard agreements, any gear transferred to local police must be used within one year or returned.⁵ Rather than give up their deadly baubles, police find reasons to deploy SWAT teams for even the most mundane work. Using statistical sampling, the ACLU concluded that at least 79 percent of SWAT deployments execute mere searches (mostly for drugs), which is about as practical as using a Black Hawk helicopter to pick up the kids from soccer practice.⁶ Predictably, innocent people—a category that has included children, the pregnant, and the elderly—are injured or die, victims of literal overkill. Because of immunity doctrines, victims and their families have little redress. Nor do other police or public prosecutors care to investigate the particulars of most of these cases.

Didn’t the Founding Fathers complain about English troops breaking down doors to enforce tax and customs laws, infringing on sacred liberty? Yes, they did—at exhausting length. And what, precisely, is the practical difference between the militarized policing of 2019 and Redcoat raids in 1776? In conceptual terms, both are objectionable; in practical terms, modern policing is a lot more deadly and invasive. And the whole of contemporary policing now drinks regularly at the well of militarized absurdity. The Los Angeles School District’s in-house police acquired three grenade launchers and a “small tank” because, one supposes, recess gets rowdy.⁷ In 2010, Phoenix sheriffs drove a tank into a man’s house, though the crime in question was cockfighting. It was perverse enough that the outcome of this animal protection mission resulted in the extermination of all the chickens (plus a dog). But then there was the pilot of said tank—truly an epic day for self-government—actor Steven Seagal, who had been, like many others, deputized (many of them as part of a benefits-for-badges program).⁸

Militarization and pay-to-play are two aspects of a much larger problem with policing: America has a lot of cops with a lot of weapons, and no firm consensus about what the police mission is or what oversight is required. The police carry some—but only some—of the blame. Society calls upon cops to be social workers, assault teams, regulatory wardens, detectives, and civic ombudsmen—a clearly unviable combination. But rather than clarify police missions, or subject departments to accountability, government resorts to the tactics of the absentee parent, offering goodies like tanks and immunity. As for the consequences of this odd parenting strategy, it seems government would prefer not to know; government collects data through sporadic and incomplete censuses. Fragmented as the data are, a mosaic can be assembled, and the results do not encourage: police seem to find it difficult to fulfill basic missions like solving crime, while resorting to violence at rates other nations would deem scarcely credible. This raises all sorts of questions, starting with: How did a nation that rebelled against military enforcement of paper and tea taxes end up with a tank-driving Steven Seagal conducting a raid to break up chicken fights?

The Origins of Policing

By law’s long standards, cops are new. For millennia, what public forces existed were special units specially tasked with suppressing insurrection, apprehending those who betrayed the state, and protecting the property of ruling elites. (Cynics would say plus ça change.) Even Imperial Rome, the most bureaucratic state in the West until the Enlightenment, hewed to this model, and the cohortes urbanae were essentially paramilitary forces designed to control riots and fight organized gangs, episodic events where one quasi-military force countered another.⁹ “Policing” in its modern sense was largely unknown and seemingly unwanted well into the eighteenth century.

Therefore, unlike most issues about law and governance, which have been debated for millennia, there’s been comparatively little time for societies to develop deeply held convictions about civilian policing. Even the ancient Greeks, normally inexhaustible in their opinions about governance, and whose polis serves as a root word of “police,” had little to say, almost none of it good.¹⁰ When Greeks needed to control unruly crowds or hold defendants awaiting trial, they delegated those tasks to slaves; no self-respecting Greek would want to exercise dominion over a fellow citizen. And the pillars of modern policing, routine patrol and intrusive investigations, were out of the question, at least where full citizens were concerned.¹¹

The general absence of police did not mean criminals were left alone, only that citizens expected to handle policing themselves, just as they managed private prosecutions. Private parties could make citizen’s arrests (a practice that remains legal in Britain and some American states such as California).¹² In England, a smattering of sheriffs handled the few crimes in which the state took a particular interest, e.g., insurrection, organized banditry, and poaching on aristocratic property. Sheriffs became somewhat more active in the thirteenth century, at the behest of reforming kings including, unavoidably, Edward I.¹³ But even mild institutionalization filled Englishmen with apprehensions about remote and arbitrary authority, and it’s no coincidence that the Robin Hood myth unfolds in the same century that Edward I granted more influence to sheriffs, including the very real Sheriffs of Nottingham.¹⁴ The English, and their Colonial descendants, were perpetually chary about despots trampling on ancient liberties, fears confirmed when the tyrannical French inaugurated large-scale policing in the seventeenth century. Nevertheless, as populations grew and urbanized, the clutter of semiprofessional enforcers—sheriffs, constables, night watchmen (the latter recruited from the uninspiring category of “useful invalids”)—seemed increasingly anachronistic and inadequate, even to passionate libertarians.

By the 1820s, British reformers decided to rationalize the system by founding a proper police force.¹⁵ First, Britain’s government needed to explain what the police would do, and how they would do it without becoming arbitrary tyrants. That was, and remains, tricky. Even the most basic duty, of enforcing law, raises difficulties. Citizens want laws enforced, but not always and not everywhere; people want to be able to roll a stop or jaywalk on occasion. Anyway, no police department could practice universal enforcement, so law enforcement necessarily becomes the exercise of coercive, discretionary power. That’s precisely what worried Englishmen then, and worries some Americans now: with discretion comes the possibility for abuse.

So, when Sir Robert Peel unveiled London’s Metropolitan Police in 1829, the first modern police department in the English-speaking world, his government tried to gain citizens’ trust by carefully defining the goals and values of the new police. London furnished each officer with “General Instructions,” stating that the principal police function was to “prevent crime and disorder.” But success required the “willing co-operation of the public,” and to reassure a skeptical public, Peel placed special emphasis on what police should not do. They should abstain from violence, deploying the “minimum degree of physical force which is necessary” to achieve a specific objective and “only when” other peaceful means had failed; refrain from “even seeming to usurp the powers of the judiciary”; not involve themselves in policy (the proper concern of legislators); and avoid “judging guilt.”¹⁶ And Peel’s government, quite presciently, forbade police from engaging in a pantomime of arrests or violent showmanship as substitutes for actual prevention. The aim was to assure Londoners that police would not be alien or repressive. London reminded its officers that they served at public sufferance and should always exhibit “courtesy” and “friendly good humour” to the public of which they were a part, a point Peel drove home by dyeing police uniforms blue, in contrast to the army’s red uniforms.¹⁷ It was a shockingly progressive agenda, and London police eventually succeeded in part because they took Peel’s admonitions seriously.

When American police departments cohered from the 1830s to the 1870s, many adopted Peel’s principles. But policing was a harder sell to Americans, given Revolutionary discontents about standing armies conducting invasive searches under “general warrants” (blanket licenses to search and seize).¹⁸ Accordingly, many American departments did not initially provide officers with uniforms or arms; Brooklyn’s police, as just one example, did not carry guns until the 1880s. Nor did police sit entirely comfortably in constitutional schemas. Most state constitutions, and the federal Bill of Rights (subsequently “incorporated” against the states), place significant restrictions on the powers of search and seizure, and most police work is in direct tension with guarantees about the privacy of one’s person and home, requiring a social consensus about balance which has not yet evolved.

Although many early police departments were incompetent and corrupt, they at least posed few major risks to life or liberty, a function of their smallness and political imperatives to leave the people alone. There was, however, one crucial exception: black Americans, specifically slaves. As legal property, slaves (and blacks generally) were excluded from the calculus of liberty. Capturing slaves and repressing their revolts were core functions for many early police departments. Indeed, Southern police vividly alleged the threats fugitive slaves posed to citizens’ lives and property in order to drum up popular support for arming officers.¹⁹ Northern police were also stained by slavery; they often cooperated with Southern colleagues to return slaves. Racial bias has been an original sin from which there has been no easy escape.

Over time, American police embraced the protection of all citizens (with varying degrees of sincerity), but this progress was offset by the development of a warrior mentality, where police stood separate not just from blacks but the public generally. It was the sort of internal army that Peel wanted to avoid, and it eventually employed the invasive practices against which the Founders rebelled: warrantless searches, arbitrary seizures, and paramilitary coercion.^{* 20} The public’s expectations also changed, and the original mission of crime prevention expanded into something more amorphous, with police called upon to fulfill roles that lay well outside the scope of their original mission. But even that mission, in many places, developed in strange ways. Protecting citizens is just one goal to be weighed against other less licit imperatives like revenue collection, protecting the institution of policing, and selective repression. These developments erode legitimacy and divorce police from the public, problems the legal system amplifies by shielding police from accountability.

Structures and Staffing

American policing is highly decentralized, a product of historical assumptions that police bound to their communities would be more responsive and accountable than a central authority. This is an idea with merit, though localization must be balanced against the need for transparency and accountability. As a useful compromise, many nations conduct localized policing but centralize certain operations so that someone, even a distant someone, can collect data, implement policy, and ensure compliance. Although semi-centralization seems to produce higher satisfaction and better outcomes for both public and police, American policing remains resolutely fragmented. Some 15,388 state and local agencies employed about 724,690 sworn officials as of 2013, the latest year for which data is available (iffy data collection being a hallmark of highly decentralized institutions).²¹ By contrast, just forty-three forces police England and Wales, and though localities have substantial influence over police, Parliament supplies two-thirds of policing budgets and exercises substantially greater oversight than Washington does over American departments.²²

Supplementing local police forces are federal and private entities. Federal police comprise only a small part of all law enforcement, with just 120,000 full-time officers, ~80 percent of whom work at just two places, the Department of Homeland Security and the Department of Justice (which houses the FBI).²³ Private police forces, which are something of a throwback to the days of citizen self-help, are more numerous. Reasonable estimates suggest that 90,000 private organizations employ at least 1.1 million people, making private actors the largest component of American policing.²⁴ Most private policing involves the mundane protection of private property—mall cops, really. However, a small but growing fraction of private police wield powers similar to those of normal police, though without the same legal and political restrictions. In some states, private cops can obtain government permission to carry weapons, display badges, make arrests, and exercise other police-like powers.²⁵ And then, of course, there’s Deputy Seagal with his tank, plus other weekend warriors including super-donor/sometime-cop Robert Mercer who, despite his base on the East Coast, served as a police volunteer in New Mexico. Under federal law, Mercer (and certain other “deputies”) could carry concealed weapons, one of the privileges of leasing a badge.²⁶

Save for federal police, most departments operate on geographic lines. The smallest geographic units are metropolitan forces, though urbanization means that cities have the numerically largest forces; New York City, for example, has by far the largest police department in America. The next level up are county forces, usually called sheriffs’ offices. Sheriffs are elected (city police chiefs are appointed) and often have non-patrol duties, like managing jails and providing security for county government. In some cases, sheriffs also provide full police services to towns and unincorporated areas that lack their own departments. State governments also sponsor their own forces to patrol highways, perform investigative work, coordinate across county lines, and fulfill special functions as permitted by state laws—the Texas Rangers being a famous example of the last. Federal agents operate nationwide, but are limited to dealing with crimes that cross state lines or otherwise implicate federal interests. In theory, localization promotes accountability, though many cities are too large for this to be fully plausible. The value of localization is also complicated by federalization, now that Washington can invoke jurisdiction whenever it pleases, thanks to the rise in interstate commerce and the enormous expansion of federal criminal law during the wars on drugs, terror, and, most recently, immigration.

On TV, federal agents usually arrive to save a case from the incompetence or corruption of local police, and while this is a trope, it does reflect real quality variations between policing institutions, much of which begins with recruitment. The FBI, despite ups and downs, has generally been regarded as law enforcement’s elite and, by selection criteria, that’s accurate. FBI special agents must have at least a college degree and three years of work experience, be in excellent physical condition, and pass comprehensive background checks and skills tests. They must also train for twenty-one weeks at the FBI Academy, perform well during a long probationary period, and be willing to accept postings far from home. The whole process, selective and drawn out, deters lesser candidates, helping ensure that FBI agents are both qualified and committed.²⁷

Requirements for local and private policing are vastly more forgiving. While the FBI warns recruits of the rigors of application, the San Francisco Police Department’s website chirpily notes that “Applying Is Easy!” and many forces advertise that jobs can be quickly had.²⁸ Although blue-ribbon commissions have been agitating for higher standards since the 1960s, including college-educated forces, minimum qualifications remain low. About 84 percent of local departments will accept recruits with a high school degree or equivalent. The Bureau of Justice Statistics (BJS) estimated that another 15 percent of departments had “some type of college requirement,” though most departments are satisfied by coursework not leading to a degree, and others accept an associate’s degree—only 1 percent of departments require a full four-year college degree.²⁹ Departments are always free to recruit above the minimum, and many do. Though data are incomplete and many surveys confuse attending college with obtaining a degree, it appears that 30 to 55 percent of newer recruits have some type of college degree (with considerable regional variation). Younger, better-educated recruits have raised departmental averages otherwise depressed by older cops hired when recruiters put less stock in college. Nevertheless, college is merely a preference, and departments can and do hire down to their minimums (educational and otherwise), especially when pressed. This has led to catastrophes, infamously in the case of the District of Columbia Police Department of the 1990s, whose recruitment binge led to bribe-taking, back-scratching, and sundry other corruption.³⁰

While there’s a strange level of controversy over the degree to which higher education benefits policing, it would be remarkable if schooling had no salutary effects. Indeed, several studies have found that better-educated officers provoke fewer citizen complaints and are less prone to use force.³¹ Education may be the explanatory variable, though it’s hard to disentangle associated factors like intelligence, motivation, social position, and higher average age at recruitment. Certainly, police should not be noticeably less educated than the general population, of which 32.5 percent have a four-year college degree (36 percent for those ages 25–44).³² A chronically undereducated force would feed impressions that police are less capable than the citizens they serve, breeding mutual incomprehension and resentment.

After educational screening, departments test for aptitude and temperament. The aptitude tests don’t seem particularly challenging; judging by sample questions in commercial study guides, someone with an eighth-to-tenth–grade education could pass with minimal preparation. The true hurdle is the background check, as many aspiring cops have criminal pasts. To satisfy hiring needs, 80 percent of larger police departments are willing to overlook minor misdemeanors; many departments will forgive DUIs, soft drug use, credit problems, and other indicators of potentially poor judgment.³³ The actual prevalence of waivers depends on department, but it’s odd that a society busily drug-testing Taco Bell employees overlooks criminality in its police. At least, however, departments can screen criminal history. Determining if a potential hire has been terminated from another police job is much more difficult; there’s (surprise, surprise) no truly universal database of officers fired from prior police jobs for misconduct, and even checking the patchwork databases that do exist may not be revealing, as some departments (governed by fear of suit and union rules) fire abusive cops but fail to disclose the misconduct in termination reports.³⁴ Various departments have found themselves saddled with bad cops fleeing other departments, only to fire them without disclosure, perpetuating the cycle.

After recruitment (assuming the hiree has not been an officer before) comes the police academy. The programs are surprisingly short, anywhere from a few weeks to six months, though this perhaps doesn’t much matter, as curricula fall well short of the ideal. Instead of emphasizing seemingly critical subjects including methods of detection, criminal codes, civil rights, and the social functions of policing, the menu boils down to two subjects: guns and paperwork. For example, California’s model curriculum requires a minimum of forty-eight hours on paperwork procedures, seventy-two hours on “firearms and chemical agents,” but just two hours of training on the criminal justice system generally—i.e., barely enough time to screen one of the (six) Police Academy movies.³⁵ Standards for private police run even lower, in some cases verging on the nonexistent.

American police departments are, on the whole, less selective than those of peer nations. Europe’s richer nations require more education for recruitment and promotion, have training programs that are more comprehensive and less obsessed with force, and impose more uniform standards.³⁶ For example, German recruits complete two or more years of specialized training in addition to holding either a college degree or a certificate from a policing-oriented high school, and senior police officials need a master’s degree or better. British police have more stringent background checks, longer and more comprehensive training, and tougher screening tests than their American cousins.³⁷ Foreign departments have noticeably higher rates of solving crime, with fewer allegations of abuse, than American police. Education and training play roles, of course, but so do cultural expectations about what police should do.

Crime-fighting and Guardianship

In 2015, a presidential commission identified two competing models for policing: the “warrior” and the “guardian.”³⁸ The warrior enforces law and fights crime, while the guardian has more general obligations of protection and service. The commission dredged up the term “guardian” from Plato’s Republic, perhaps not realizing that Plato described proto-fascist liars who told people what to do at spear-point—or as Giuliani put it, “[f]reedom is about authority.” (Giuliani had forgotten his Orwell; the correct quote is “freedom is slavery,” but as Big Brother noted, “ignorance is strength.”) Whatever the literary subtext, opinion polls suggest that the public views policing as tilting more toward the warrior/enforcer model, a view not always discouraged by police.³⁹

In fact, warrior-style crime-fighting is not a major component of police work as a fraction of hours worked. Police spend less than 5 percent of their time dealing with violent crimes against persons (murder, kidnapping, battery), and another 15 to 20 percent on property crimes (burglary, trespass, vandalism); all the rest is a mix of traffic enforcement, social services, routine patrols, and paperwork.⁴⁰ The relative infrequency of violent crime as a direct component of police work is important, because fighting violent crime is a frequent justification for aggressive policing.

Perhaps police suppress serious crime by nipping minor crimes in the bud, and so a much larger chunk of police resources is devoted to vigorous crimefighting in some attenuated way. This was the premise of “Broken Windows” policing (popularized by Giuliani), which cracked down on small-scale disorder, on the theory that a broken window would signal that the community would tolerate graver civic decay. Thus, the NYPD cracked down on petty crimes such as graffiti and turnstile-jumping (the criminal equivalent of the broken window) to prevent future and more violent crimes. But dealing with minor crimes is low-risk and requires little force, and it’s hard to justify paramilitary policing because a fare-jumper might turn into the Zodiac killer.

Broken Windows was, in a way, a gussied-up version of the old thinking that more cops on the street would reduce crime. Politicians constantly promise more “feet on the beat,” and while this may be effective as a political strategy, it’s not clear that it’s an effective policing strategy. Since the 1970s, empirical research has cast doubt on the utility of simply beefing up patrols. In 1972 and 1973, Kansas City tried to assess the impact of varying patrol intensity. The city established three study areas: the first as a control; the second, with preventive patrols eliminated entirely; the third, with policing radically increased.⁴¹ Although patrol intensity varied from −100 percent to +300 percent, researchers found no significant difference in crime levels or citizens’ fear of crime or satisfaction with police.⁴² Subsequent studies in other cities came to similar conclusions.⁴³ More sophisticated research, which targeted small and especially violent areas (the police equivalent to the Army’s “surge” strategy), did associate enhanced patrols with falling crime, but improvements eventually subsided.⁴⁴ Thus, politicians’ chatter about beefing up patrols is mostly vacant, as should have surprised no one. In any event, most crimes are not detected by police on patrol. Rather, the public reports criminal activity—and when police solve crimes, they usually do so by building cases against suspects supplied by victims.

Even with public help, police are notably unsuccessful at bringing criminals to account. The conventional metric for “solving” crime is the “clearance rate,” defined as taking a suspect into custody and remanding to judicial process.⁴⁵ Roughly 22 to 23 percent of known offenses are cleared by arrest, which is not impressive, especially as this overstates the true rate of crime resolution.⁴⁶ First, estimates suggest that fewer than 47 percent of violent victimizations and 35 percent of property victimizations are reported, so true clearance percentages might be much lower (there obviously being no way to fully know).⁴⁷ Second, police “unfound” (their term) many reported crimes, ostensibly because there is not enough evidence to support a charge, a decision about which police have unaudited and legally unaccountable discretion—discretion which has been abused at several departments.⁴⁸ The NYPD had a long practice of manipulating statistics to achieve more pleasing results, and gaming the numbers became particularly intense after New York launched its COMPSTAT program, which provided data reports that became potentially embarrassing media fodder. Other manipulations include reducing the severity of a reported crime (e.g., categorizing a rape as an assault, or a grand larceny as simple theft), or invoking the “exceptional instances” clause of the clearance definition to terminate a case—a procedure originally designed to close out cases where circumstances beyond police control made normal resolution impossible, as when a prime suspect or key witness died. Some departments seem to blatantly abuse these exceptions: between 2006 and 2008, more than three-quarters of crimes cleared by Sheriff Joe Arpaio’s department in Maricopa County were cleared by exception, vastly higher than rates in nearby departments such as Phoenix (0.08 percent), Mesa (6.0 percent), and Glendale (11.0 percent).⁴⁹

Accounting for all these factors, very few crimes result in arrest, likely less than 10 percent. Resolution of property crimes is especially low, perhaps under 5 percent, because police rely on victims to provide a suspect and most theft victims cannot do so. Police also view these matters as low-priority, as property owners who have suffered burglaries know—it’s customary for police to warn theft victims that the case will never be solved (the implication frequently being that no attempt will be made), so filing a police report becomes a mere formality in the insurance claims process, not a step toward justice. Perhaps this is a reasonable posture for police when it comes to minor crimes, and justifies low resolution rates. But serious crimes also have fairly low resolution rates. Murders, among the most thoroughly reported and investigated crimes (dead bodies being harder to ignore than a stolen lawnmower), have a gross resolution rate of 61.5 percent nationally, which is fine but not outstanding. Some departments fall well below the already so-so average; Chicago has cleared fewer than 40 percent of its murders in recent years, and in 2017 the rate dipped under 20 percent, meaning that Chicago’s murder clearance rate is lower than the nominal national average resolution rate for violent and property crimes.⁵⁰ By contrast, Canadian police solve 67 to 74 percent of murders, while German police “clear up” (their term) more than 90 percent of murders/manslaughters.⁵¹

Discretion’s Uses and Abuses

Police have considerable leeway when confronted with criminal conduct. It is largely up to an officer whether to overlook a crime, make an arrest, or use force against a suspect. This is not how the system supposedly works, though when an officer exercises good judgment, discretion can be for the best. An unwitting trespasser might willingly move along if warned by police, or a motorist might be chastened into fastening a seat belt after a warning. When leniency satisfies goals of safety at low cost to enforcement, discretion does net good. However, police sometimes let infractions slide for dubious reasons; cops are especially loath to arrest other cops, for example, and some highway patrols were notoriously forgiving of speeders whose bumper stickers featured police charities. Discretion has occasionally slid into serious derelictions of duty, as happened in Chicago in the 1970s, where police refused to patrol the city’s notorious housing projects, either out of genuine fear or perhaps motivated by the belief that the lower-class, largely minority residents in public housing did not deserve protection. As departments do not gather relevant data, it’s impossible to quantitatively evaluate whether discretion brings net good or ill, though proxy data in some areas, like racial bias in arrest rates, suggest the widespread abuse of discretion.

Victims exercise their own discretion, which police never know about; we all do this when we fail to call in stolen Amazon packages and so on, and this can be socially efficient since the chances of apprehension are near zero, while the personal and public costs of filing a report usually exceed the value of merchandise. Victims also sometimes decline to assist with investigations, either out of fear or indifference, in which case officers usually drop a case. This can be immaterial, though in certain matters, especially those involving domestic violence, it can amount to making social policy on the fly. Some states have addressed this by requiring police to make an arrest whenever they have probable cause to believe domestic abuse has occurred, regardless of the victim’s nominal preference. In general, though, police retain autonomy over arrest decisions.

Few legal officials have discretion quite as wide as the police. This is why Sir Robert Peel hoped laws would be enforced uniformly, and why some critics have called for mandatory arrest policies for all types of crimes, to ensure equal protection. (These proposals are unrealistic, given finite resources, and possibly a case where the medicine would be as bad as the ailment.) A thoughtful decision to arrest, or threaten arrest, balances a range of social interests: the likelihood of conviction, the seriousness of the crime, the opportunity costs of making a particular collar, victims’ preferences, and suspects’ rights. The calculation is subtle, requiring education many officers do not have and experience junior officers (the mostly likely to be assigned beat duties) necessarily lack. It also requires knowledge of law that is unhelpfully vague. Unscrupulous or merely ignorant police can invoke ambiguous laws regarding panhandling, public disturbance and assembly, and so on, discouraging conduct that they incorrectly deem illegal (or find subjectively wrong). The possibility for abuse is not trivial, amplified by “contempt of cop” arrests.⁵² Nor is there much people can do, because most states functionally prohibit “interfering” with police work or resisting arrest (even when the arrest itself is improper). Given the time, expense, and vocational consequences of arrest, citizens may simply comply with an officer’s order, proper or not, building resentment toward the law.

Officers have also used discretion to further criminal activity, sometimes in the name of the greater good, and sometimes not. In limited cases, police allow informants or even their own officers to participate in criminal activity as part of an undercover investigation.⁵³ Freedom of Information Act requests show that the FBI allowed “otherwise illegal activity” at least 5,658 times in 2011. Precise details were withheld, though they may not exist, as the FBI’s internal controls are often lacking. The Bureau informed USA TODAY journalists that it couldn’t so much as find files for prior years (though law requires meticulous record-keeping) and an Inspector General’s report from 2005 suggested that some authorized criminality was not properly reported to supervisors.⁵⁴ These missing data are troubling, because FBI guidelines contemplate that agents might engage in practices such as money laundering and trading in controlled substances.⁵⁵ (For its part, the DEA doesn’t seem to keep track of authorized criminality.) Given the seriousness of these activities, and the known lapses in internal control, the potential for disaster is not small. Local police departments also permit their own to participate in otherwise illegal activities. High-speed pursuits are an undeniable example, and whatever goes on in undercover vice operations is a fertile ground for speculation.

Another difficulty is that improper motives can be hidden behind the banner of discretion, and trading leniency for favors has been a traditional weakness in policing. The free cup of coffee presents no serious problems; indeed, it’s often happily given to local cops, just as many donate to police-affiliated charities out of general gratitude.^* However, anything beyond a small and isolated freebie usually comes with an expectation, and many departments lack clear rules about lower-level gratuities, so there’s ambiguity about how many free pastries and coffees constitute too many. Departments also do not have strong guidelines about freelance work, which many officers perform. Potential conflicts of interest can be substantial. How will officers respond when dealing with a private enterprise, like a bar, which employs a fellow officer as a part-time bouncer? The murk can be considerable; worse, some police have been caught taking bribes, and directly participating in drug running, prostitution, perjury, and bank robbery. Major departments have suffered episodes of severe corruption; in the 1970s, the NYPD was one of the biggest drug dealers in the city, and in the 1990s, there were few types of crime that LAPD’s Rampart Division did not practice, becoming notorious enough to inspire The Shield, FX’s crooked-cops drama.

Fairness and Force

Two major categories of abused discretion deserve particular attention: discriminatory policing and widespread use of unnecessary force. Unlike bribery and extortion, these practices are semi-institutionalized, and they are protected both informally (by union power and the “wall of silence”) and formally (by doctrines of legal immunity and “police officer bills of rights” codified in state law).⁵⁶ As a result, where the layperson sees bias or excessive force, the law frequently sees nothing, a divergence created by law’s excessive deference to officer discretion and also by assumptions about police work that are not entirely supported.

Episodes of biased policing have been well publicized, especially after 2013, as the Black Lives Matter movement chronicled the stories of unarmed black suspect-victims like Trayvon Martin, Eric Garner, and Michael Brown. The existence of bias in individual officers and departments is beyond doubt, though overall prevalence is harder to quantify. As usual, direct nationwide data are missing, and what data are recorded sometimes contain errors that seem to be deliberate attempts to reduce racial skew, as when SFPD officers classified some Hispanic arrestees as “white.”⁵⁷ While we can only infer from proxy indicators, it’s undeniable that bias is a real problem. Minorities, especially black persons, are arrested at significantly higher rates than whites. Blacks account for 26.9 percent of total arrests (against 13.4 percent of the general population), the majority of arrests in murders/manslaughters (52.6 percent) and robberies (54.5 percent), and a hefty chunk of loitering violations (40.7 percent).⁵⁸ Blacks also comprise at least 33 percent of the incarcerated population generally, and in some states, much more; blacks constitute 68 percent of Maryland’s incarcerated.⁵⁹ Social factors—such as the historical poverty and discrimination to which blacks are disproportionately subject—explain only some of the disproportion (disadvantaged people of all ethnicities being more likely to be arrested for many offenses). The rest of the difference requires explanation, which biased enforcement largely supplies.

The Constitution guarantees equal protection and forbids racial discrimination, so biased enforcement goes to the heart of law. Many departments, even in highly progressive cities, have fallen under the shadow of prejudice. For example, San Francisco police stop black motorists at approaching triple the frequency that population distribution would predict, and once stopped, the cars of black motorists are subject to nonconsensual searches at more than five times the rate for whites. Police do find contraband in blacks’ cars, but only one-third of the time—notably, the lowest “hit” rate of any ethnic group.⁶⁰ In New York, the stop-and-frisk program, in full force from 2002 to 2013, captured a grossly disproportionate number of blacks and Latinos; the NYPD conducted more than 570,000 stop-and-frisks of minorities in 2011 alone.⁶¹ Most were innocent (or innocent enough that the government declined to take further action; indeed, this was true for 88 percent of all ethnicities stopped).⁶² A federal judge subsequently ruled that New York’s existing stop-and-frisk program was unconstitutional, amounting to racial profiling.⁶³ (Nevertheless, versions of it continue.⁶⁴) Following 9/11, local and federal police also targeted Muslim communities for enhanced surveillance, though these intrusions produced few (if any) significant arrests. Discriminatory policing both perpetuates prejudice and wastes police resources. But while yield rates in San Francisco, New York, and other cities were low, the sheer scale of searches ensured a steady flow of minorities into prisons. As Warren Buffett noted, “[i]f a cop follows you for 500 miles, you’re going to get a ticket,” and this is true, though a day of tailing a geriatric industrialist to write a $200 citation is not a terrific return on social assets, and the same holds for endless frisks to round up a couple of possession charges.⁶⁵

Some activists explain discriminatory enforcement as an inevitable product of biased hiring policies, and if this is true, discriminatory enforcement is likely to be widespread and persistent. Racial/ethnic minorities comprise almost 40 percent of Americans, yet account for just 27.3 percent of police officers, and while the proportions have improved, progress has not been radical of late.⁶⁶ Departments that do not reflect their communities are often plagued by enforcement bias, as has been the case in Chicago, San Francisco, Los Angeles, and other big cities. Nor should citizens be beguiled by large numbers of minority officers if a department still fails to reflect its community. The Baltimore PD is almost evenly split between white and minority officers (though the city itself is less white than that), but that department’s abuses were so bad that the DOJ intervened.⁶⁷

As it happens, the academic literature is mixed on whether the race of an individual officer makes much difference in daily work, but this result is less important than the possibility that unbalanced hiring reflects institutional biases that will be recapitulated in the field and in legitimacy deficits suffered when departments do not reflect their communities. For these and other reasons, 44 percent of blacks report little or no confidence in their communities’ police, versus just 18 percent of whites.⁶⁸ Minorities’ distrust of police is crucial because minorities interact with police with disproportionate frequency, as both involuntary detainees and voluntary users of police services.⁶⁹ More balanced forces would help, though the standard response is that this cannot be achieved because minorities are unusually hard to recruit or fall below police hiring standards. As a statistical matter, some minority groups do have less education and more significant criminal records than the white majority (in which discrimination plays its own part). But even accepting this, the standard explanation is weak, given the low qualification thresholds for police, including forgiving attitudes toward prior convictions. In any event, it’s possible to achieve near-complete parity with the community.⁷⁰ Atlanta police have done so, and that department is largely unburdened by the widespread bias scandals affecting many other big-city departments.⁷¹

Race is not the only way police can belong to a community. Formerly, officers lived inside the bounds of the cities they patrolled. In recent decades, residency requirements have been falling in police departments, and those that exist are often ignored. Today, 60 percent of officers in the seventy-five largest police forces live outside the limits of the cities that employ them, and in Boston, that included most of the police force’s top brass, notwithstanding city hall’s edict.⁷² Many departments with alien forces experience high levels of discontent, while small towns with resident cops generally do not.

Entwined in the dynamics of alienation and discrimination are questions of excessive force, with minorities receiving disproportionately brutal treatment. Blacks are significantly more likely to be killed by police than whites, and minorities of all backgrounds account for over 60 percent of fatal police shootings when suspects are unarmed (here, in a dismal technicality, the minority becomes the majority).⁷³ Adjusting for the usual variables cited to explain away disparities—poverty, neighborhoods, etc.—still leaves a racial skew. What does narrow the gap considerably is frequency of police interaction, and as we’ve seen, minorities are substantially more likely to interact with police than are whites. Minorities still experience violence at a somewhat higher rate, but this reflects a larger problem, as suspects of any ethnicity face unusual dangers when meeting American police.

By global standards, American policing is violent. Data are again spotty and prone to faulty self-reporting, though journalists (especially at the Guardian, Washington Post, and New York Times) have done extraordinary work in assembling more comprehensive and disturbing accounts than those on official offer.⁷⁴ As researchers have noted, American police killed “more people in the first 24 days of 2015 than English and Welsh police have killed in the last 24 years.”^{* 75} Black Americans are at high risk, but on a per-capita basis, even white Americans were fatally shot at twenty-six times the rate of German counterparts. American police also shot citizens at an estimated twelve times greater rate than Australian police and four times the rate of Canadian police.⁷⁶ Adjusting for factors like America’s higher crime rate explains only part of the difference, and the unavoidable conclusion is that American policing is abnormally fatal. Besides guns, American police also have greater propensities to use clubs, fists, and chemical irritants. American police also use Tasers, a weapon barred or severely restricted in many other nations, given the dangers of unanticipated trauma and possible death from electric shock, though in America’s unique case, Tasers represent distinct improvements over gunshots.⁷⁷ Even without Steven Seagal’s action-hero antics, American police use force so intensely that the United Nations singled out America for violating the UN’s general torture guidelines, which also cover police brutality.⁷⁸ As a party to some of the relevant international treaties, America had to issue a response, and the State Department furnished a subdued memo, which noted that while America was not in “full agreement” with the UN’s high-minded musings (an understatement), some improvements had been made, including the settlement of a Chicago police torture suit sweetened by “a formal apology to torture victims.”⁷⁹ Of course, this concedes that torture took place.

One reason for force disparities is America’s unusual practice of physically detaining those suspected of minor crimes. In Western Europe, police generally issue tickets/court summonses for lesser infractions, which reduces stakes for both sides and makes violence less likely. For example, selling merchandise without a license is a trivial offense in most countries, and European cops confronted with a street peddler would simply write a ticket and move along. The same crime in America often provokes arrest—this is what happened with Eric Garner, whose sale of loose cigarettes resulted in a fatal chokehold. Both Garner and the arresting officer had records of misconduct, complicating matters, but the triviality of Garner’s offense, combined with video showing that Garner had, at most, displayed some pert exasperation and minor resistance at being harassed for the umpteenth time by local police, suggest that the officer used force grossly disproportionate to any reasonable need. Indeed, force was almost necessarily excessive, as NYPD guidelines prohibited chokeholds as unacceptably dangerous. In other nations, Garner would have been simply ticketed, and in the unlikely event of death, the officer would have faced immediate sanction. But New York is not Frankfurt. Garner died, and the grand jury declined to indict the officer, a normal outcome for police (though not, as we’ve seen, for other citizens).⁸⁰

As international comparisons suggest, “excessive force” is socially determined, and in America, the effective determiners are largely cops and judges. American law lacks a firm, universal definition, which is a problem in itself, though force is generally considered excessive when it exceeds what the officer “reasonably” believes necessary given the circumstances.⁸¹ As a result, it’s hard to calculate prevalence, though citizens, who have their own definitions, report excessive force at a rate of 6.6 complaints per 100 officers, of which only 8 percent were determined to justify disciplinary action.⁸² The BJS has received complaints of data quality issues, and is looking into the matter, though in 2013, it spent all of $36,167 to commission a study.⁸³ We can safely say BJS overspent. Since 1994, the attorney general has been required by Congress to collect and publish various use-of-force data, but that exercise has not been properly undertaken, as the attorney general effectively conceded when DOJ announced, in October 2016, that it would finally implement a full data collection program.⁸⁴ So it takes $0 to conclude that DOJ fell down on the job, and one can venture at a further cost of $0 that it will continue to do so, because after October 2016, priorities changed. Jeff Sessions, the incoming attorney general, seemed deeply uneager to pursue the inquiry. During his confirmation testimony, Sessions fretted that federal nosing around in local law enforcement could undermine “respect” and “morale,” an opinion delivered in the unfortunate context of discussing federal oversight of police in Baltimore (home of the Freddie Gray disaster), Ferguson (Michael Brown), Maricopa County (Sheriff Arpaio’s tent-city internment camp for the un-American), and other problem departments.⁸⁵ In some technical way, Sessions wasn’t wrong. After all, Sheriff Joe did seem distraught about his criminal contempt conviction for defying a court order to stop racial profiling, until the illimitable clemency of POTUS 45 lanced the boil.

It’s clear that police deploy “excessive force,” in the phrase’s vernacular sense, with alarming frequency. But in the term’s legal sense, police hardly ever use excessive force, or more germanely, they’re rarely held accountable. As we’ve seen, law enforcement declines to prosecute itself, so civil suits must suffice. Although citizens have had a federal right to sue officials via a special waiver created during Reconstruction in the early 1870s, it wasn’t until 1961 that the Supreme Court held that citizens could sue police officers personally for federal rights violations, ultimately including use of excessive force (somewhat counterintuitively, the right to sue departments is more limited—due to sovereign immunity). Defeat quickly followed victory, and in 1967 the Court ginned up what it called “qualified immunity,” an offshoot of sovereign immunity.⁸⁶ With qualified immunity, government officers may escape liability for certain actions. The “qualified” bit is that immunity will attach only when an officer’s action is “official,” i.e., not in violation of the law. Should an official violate the law, he may be subjected to legal consequences.⁸⁷ This doesn’t sound so bad, but it is.

First, though, the theory behind qualified immunity. Law wants officers to be able to exercise discretion without fear of lawsuits, which is important because much conduct will exist in a gray area between the clearly lawful and clearly unlawful. Fair enough, but the Court had initially imposed a difficult standard of proof, looking into an officer’s “good faith,” which was hard enough to gainsay (not least because the Court’s language was somewhat inscrutable).⁸⁸ The Court then shifted to an even more stifling test, which granted officers immunity unless their conduct violated a “clearly established statutory or constitutional right of which a reasonable person would have known.”⁸⁹ This is a different and more troublesome animal: the Court is not protecting shades of gray, so much as forbidding liability for any conduct that falls short of a very dark sort of black. This is because—as should be apparent by now—almost nothing in the law is “clearly established,” much less in ways that “reasonable people” would have known. This is particularly so for Fourth Amendment jurisprudence (the usual grounds for suit), a highly confused and fluctuating area of law. What if 10 percent of judges, swimming against the tide, rule that shooting rattle-armed-and-highly-suspicious toddlers presents no Fourth Amendment problems, while 90 percent of judges go the other way? Is the toddler’s right not to be shot while shaking a rattle “clearly established”? If the answer is “no,” it’s open season on toddlers.

Qualified immunity became even harder to overcome during the crime wave and drug hysteria of the 1980s and 1990s and the siege mentality post 9/11, with the Court ever more forgiving of abusive conduct. The current test is whether a victim had both a “clear” right and “every reasonable official would have understood that what he is doing violates that right.”⁹⁰ (The Court, via Scalia, cited prior precedent for that formulation, implying that the “every” official qualifier had been part of the doctrine for twenty-four years; it wasn’t—Scalia slipped it in.⁹¹) Not only must courts agree, perhaps so must cops (and unanimously, at that). Given that several million Americans suspect the Earth is flat, disagreement would seem to create almost insuperable barriers to recovery.⁹² Justice Sotomayor described the present state of qualified immunity as an “absolute shield” that “tells officers that they can shoot first and think later, and it tells the public that palpably unreasonable conduct will go unpunished.”⁹³ Fortunately, Sotomayor said this in dissent, otherwise it would be the one legal standard about which “every” cop would assuredly know.

Victims do sometimes receive compensation for police brutality, but regrettably, where victories are had, the law is usually not the prime mover. Only in special cases—a sympathetic victim, a media-savvy lawyer, and massive public outrage combined with body camera footage—do victims have a decent chance of prevailing. Most of these victories come by way of settlement, and are driven as much or more by fear of losing in the court of public opinion as in a court of law. Settlements do happen, and in 2014 alone, America’s ten largest police departments spent $248.7 million paying for misconduct.⁹⁴ This is limited consolation. Much of that $248.7 million went to blockbuster settlements, often spread across many victims and years. But the average victim, assuming his suit survives pretrial motions, doesn’t receive a giant windfall. Police understand that they are increasingly beyond account, and as the public begins to appreciate this, police legitimacy will fall and citizens will go out of their way to avoid police or simply submit, out of fear, to unreasonable requests that should be refused.

Latent in qualified immunity doctrine are assumptions about the special dangers of policing; in case after case, judges (often conservative) worry that cops should not have to consider liability when they face a potential threat. But this is a red herring. Certainly, cops shouldn’t have to ponder lawsuits during a chase, but they should always consider Constitutional rights as one of the many variables they calculate, and not killing unarmed suspects in conditions of ambiguity is probably not too much to occupy the police mind, even during a pursuit. Anyway, policing is not unremittingly perilous. Conservative judges sometimes say so, and 93 percent of police believe that their jobs are getting more dangerous but, in fact, policing has become meaningfully less dangerous over the long haul, and being an officer carries lower risk of fatality than being a garbage collector or roofer (or, for God’s sake, a landscaper), and is 90 percent less lethal than commercial logging.⁹⁵ We do not permit garbagemen to shoot guard dogs or loggers to put trespassers in chokeholds. Yet, we grant police something like this leeway, and it results in disproportionate danger to the public. For 2016, the FBI reported that 118 officers from any type of department were killed in the line of duty, of whom 66 died in violent circumstances and the rest by accident; meanwhile, officers fatally shot 963 people, a ratio of 8:1, in the most forgiving calculation.⁹⁶

Death is not the only danger to police, and police reasonably fear violence, though statistically, they’re not likely to be harmed even if directly threatened. In 2016, 57,000–68,000 officers were assaulted in the line of duty, but in only 4.2 percent of those assaults were guns used, while 78 percent of assaults involved no lethal weapon at all, and in 71.1 percent of cases, officers escaped without injury.⁹⁷ Even counting all assaults with weapons of any kind—guns, tire irons, the various cell phones and other shiny objects police have claimed as justification for shooting “armed and dangerous” suspects—produces some 16,500–20,000 officers actually injured. By comparison, between 2002 and 2011, the FBI estimated police threatened or used force on an average of 715,500 adults annually, with 535,300 citizens claiming excessive force.⁹⁸ Even subtracting a large number of serious criminals who required equally serious force to subdue, there’s a large imbalance. The old saw used to be that it was better to let ten guilty men go free than imprison one innocent man; the contemporary version seems to be: better to shoot eight citizens or rough up thirty, than risk one officer.

Pandemic or Pandemonium?

Judging by public opinion, America has experienced an unrelenting increase in crime since 1989, when Gallup began polling the issue.⁹⁹ Curiously, August 2000 to October 2001 was a rare stretch when most Americans believed crime was not on the rise, though that period coincided with one of the few truly significant and well-publicized crime spikes in recent decades: the 2001 accounting scandals (relating to the dot-com crash, and the collapses of Enron and several other firms, some of the most costly financial crimes in history) and the 9/11 terrorist attacks, which murdered 2,977 people and will contribute to the early deaths of thousands of others (against an annual national homicide rate of roughly 15,000).¹⁰⁰ Leaving that aside, since 1993, on average, almost 65 percent of Americans in any given year believed that crime had risen over the past year; more recently, 57 percent believe crime has gotten worse compared to a baseline set in 2008.¹⁰¹ Both Democratic and Republican politicians cater to this, spending heavily on police and pushing harsh sentencing laws and trimming civil liberties, policies pushed by, among others, Clinton, Bush II, and Trump. Courts, as we’ve seen, have also become more deferential to, and protective of, police.

But crime has not been rising, it’s been falling, and by a lot. From a peak of 5,856 crimes per 100,000 people in 1991, crime fell 51.6 percent to 2,837 per 100,000 by 2016.¹⁰² What’s important is that crime began falling before many of the new cops, tougher sentences, or enhanced police discretion were deployed, and while crime continued to drop thereafter, much of the fall occurred before any crime initiatives could take hold. But policing expanded anyway. Since 1992, the nation added almost 200,000 new cops and the police/population ratio grew by over 7 percent, but if anything, crime declines decelerated as policing ratios peaked in 2008.¹⁰³ Since 2014, crime has been roughly flat, hovering near twenty-five-year lows. When Trump lamented American “carnage” in his campaign, crime rates were basically steady, with lower property crime rates offset by a modest increase in violent crime, and most of that increase was driven by just a few cities such as Los Angeles, Charlotte, San Antonio, and Chicago.¹⁰⁴ Chicago, whose civic structures have long been undermined by corruption and financial problems, by itself accounted for half the growth in the murder rate. A few cities are experiencing crime waves, but the nation is not.

The relationship between police and crime suppression is more complicated than just the number of officers. It may be that police have grown more effective, using methods like zero-tolerance policing and statistically driven management. But again, these policies were enacted after crime started dropping—sometimes long after. New York’s crime rate started falling around 1991 or 1992, in line with national trends; at most, pilot experiments in cleaning up the subways had been underway for about a year and Rudy Giuliani, who championed the program, did not take office until 1994. Crime in New York was already waning, as it was in other cities that adopted New York–like models in the 1990s, and even in places like Los Angeles, whose police department barely functioned then, much less innovated. The research on these models is accordingly mixed; new initiatives might have contributed to the crime decline, but did not inaugurate the trend. What’s also overlooked is that programs like stop-and-frisk themselves increased lawlessness, if for no other reason than that several hundred thousand unlawful detentions occurred during those programs and citizen complaints about police spiked (none of which, of course, were recorded as “crimes” in NYPD databases).

Many forces drove crime lower, of which police were merely one and perhaps not the most important. The economy improved in the early 1990s, but this explanation is incomplete: economic improvement in the 1980s did not suppress crime, nor did the Great Recession and its aftermath (2008–2014) spark a crime wave.¹⁰⁵ Adding up rising incomes, new policing tactics, higher incarceration rates, slightly reduced alcohol consumption, and marijuana decriminalization/nonenforcement still leaves a mystery. The proportion of younger men, who have a traditionally higher propensity for crime, did subside in the 1990s, though the giant Millennial generation has not triggered a criminal resurgence. The causes require investigation beyond our scope. (There may have been something unique about Boomers in their youths, an idea I’ve explored elsewhere.) What matters here is that crime fell notwithstanding police policies or procedures, which cautions against blind increases in police spending, as well as against tolerance for brutality, civil liberties violations, stop-and-frisk, and other extraordinary measures. The apparatus of emergency is in place, but the only state of emergency that exists is in law itself.

Asking Too Much

Police duties, and the law governing them, are amorphous—and police are essentially the victims here. Cops spend endless hours dealing with matters that are really the province of social services, handling tasks in which they are not specialists and which legislatures should assign to others, especially caring for the mentally ill. Police are also tasked with generating revenue for governments, a chore that’s distracting, but not civically disastrous so long as fines and forfeitures are by-products of genuine law enforcement and not substitutes for normal taxation.

However, some governments rely on police to furnish large chunks of civic budgets, creating unhealthy incentives for extortion, especially when departments share in the proceeds. Ferguson provides an extreme example, where the DOJ concluded that “law enforcement practices are shaped by the City’s focus on revenue rather than by public safety needs,” leading to overheated policing of the city’s municipal code and aggressive arrests (disproportionately of minorities), the essential purposes of which were drumming up cash.¹⁰⁶ Civil fines and fees have been imposed for seemingly minute violations of the local codes; one black woman in Ferguson who kept her family’s old car as a sentimental possession had it towed on the grounds that it was a “derelict vehicle in the driveway”—she was ultimately assessed $1,200 in fines and never had the car returned.¹⁰⁷ Few cities are as bad as Ferguson, but about 6 percent of cities generate more than one-tenth of their revenue through police collections.¹⁰⁸ This is an abuse of both the public and the police. These levies are arbitrary, regressive, and, given the prevalence of bias, have outsized impact on minorities. These practices also divert officers from real work, and researchers have noted an inverse relationship between revenue generation and crime clearance rates in small towns—the more cops tax, the less they police.¹⁰⁹

These practices also create an odor of corruption around other parts of government, including courts tasked with judging cases of police abuse. For example, the Nevada judiciary’s budget substantially depends on traffic ticket revenue, and when ticketing fell off, Nevada courts had to cut back on judicial services, threatening the administration of justice.¹¹⁰ While the funding crisis affected no rulings by the court on policing procedures, the appearance of bias and conflict is hard to avoid, and who knows what will happen when courts find themselves against the budgetary wall? Well, Ferguson does—its municipal court relied on police fines for its operational budget, and justice in Ferguson is not the force keeping Liberty’s torch upright.

The most troubling version of cops-cum-taxmen involves civil forfeiture, whereby police seize properties based on the suspicion that the items have criminal associations—without needing to charge the owner with a crime. (Technically, the property itself is the “defendant” in these in rem proceedings, which produces strange case names like United States v. Approximately 64,695 Pounds of Shark Fins.) The usual criminal procedural protections don’t apply to civil forfeitures, and the burden of proof rests on the property owner, who must show that seized items are untainted by criminal activity. In other words, you may be presumed innocent, but your bank account, house, and car are not. The particular moral hazard in this practice is that police share in some or all of the proceeds, creating unhealthy incentives while reducing accountability, as police can self-finance without voter consent. This is precisely the sort of arbitrary power that the freedom- and property-minded Founders loathed, and forfeiture traditionally existed only at the margins of American law, in customs enforcement and direct contraband seizures. The wars on drugs and terrorism were used to justify expanding the practice—another in the tiresome litany of extraordinary actions premised on dubious emergencies. Since 1985, when federal forfeiture kicked off, annual receipts increased from $25 million to ~$2 billion in recent years, and up to $4.5 billion in bumper years like 2013–2014; as of 2017, the forfeiture funds’ gross assets stood at over $8 billion (because only ~40 percent of proceeds were disbursed to victims), though of course, it’s hard to tell, as auditors identified significant deficiencies in the funds’ internal controls.¹¹¹

States have also gotten aggressive since 2001, and now reap several hundred million dollars through forfeitures. Many states’ laws have historically disfavored forfeiture, but again, special circumstances provided a loophole—local police could seize assets and turn them over to Washington, whose forfeiture laws were more accommodating, and Washington returned the favor by granting police a cut that local laws would otherwise forbid. (The euphemism here is “equitable sharing,” though the adjective applies only to law enforcement, and the only ones really “sharing” are citizens.) Thus, procedures that once applied to customs runners and drug lords have become part of standard practice, to relieve ordinary citizens, many of them innocent, of their property. Forfeiture represents a considerable source of funding and is extremely flexible, unburdened by legal hurdles (overseen, in any event, by courts who may themselves also benefit) and the tedious process of convincing voters to pay more in taxes. So this is the future, of police transitioning from de facto IRS agents into government-sanctioned shakedown artists.¹¹²

As police find themselves occupied by other tasks, the public has turned to private protection, especially for the property crimes police lack either the time, skill, or inclination to solve. Policing has not gone nearly as far down the road toward privatized justice as arbitration already has, nor is it as unaffordable as private lawyering increasingly is. But private police already outnumber regular police, and their menu of services is growing longer; it’s not unreasonable to believe that many clients will soon pay for services formerly supplied by police, if only because the ultra-rich already do so, and their habits will trickle down to the merely affluent. In litigation, wealthy clients routinely rely on private investigators for both criminal and civil cases, and private police already handle many special crimes like wine counterfeiting and art theft and will, in due time, expand into more prosaic duties. Of course, private policing has a long tradition; the only difference is that before the mid-nineteenth century, private policing was the norm, just as private schooling was. Today, use of private services implicitly comments on the trajectory of public equivalents. Just as Stanford has eclipsed UC Berkeley, so Pinkerton’s descendants may eclipse some public police. There’s nothing wrong with this so long as it’s a choice, not a necessity. However, if and when private police become much better than the public option, the implications for social and legal equality will become serious.

Whenever political discourse routinely invokes the unrelieved glory and nobility of a particular group, and even mild deviations prompt thunderous denunciations, we can be sure that unsavory compromises have been made. The police are a troubling case, because unlike the celebrated American farmer (the customary symbol of self-sufficient virtue, tariffs and subsidies notwithstanding), police have been granted the right to use force against fellow citizens and to deprive them of their liberty. This calls for the highest scrutiny, not unthinking deference.

Although police view themselves as under siege and without adequate support, the public and legal system actually grant officers extraordinary and undue leeway. The general public holds the police in higher regard than almost any institution save small business and the military, higher even than organized religion or medicine.¹¹³ Citizens proffer generous funding; in many cases, police consume almost half of municipal budgets. Police require, and politicians supply, regular and degrading homage. When tribute is not forthcoming, police freely complain about their civilian bosses, sponsor political campaigns, and vent spleen in astounding ways, as when officers turned their backs on New York mayor Bill de Blasio during an officer’s funeral. Certainly, police deserve respect for facing danger. But the troops face equal or greater dangers daily, without the overt politics or public disobedience.

The police describe themselves as the “thin blue line” separating society from disorder. It may perhaps be time to retire that phrase (not least because the case that made the phrase famous sentenced an innocent man to die).¹¹⁴ The line is no longer thin, and it no longer wears blue, with many officers favoring paramilitary black, an apt if dismaying choice. This internal army was never contemplated by the Framers or the Constitution, which hardly expected (and arguably does not permit) armed agents to execute mass searches and seizures. The Framers would be shocked that courts have eroded the Fourth Amendment’s warrant requirements to the point of meaninglessness, so that mere suspicion now allows millions of detentions, arrests, and property seizures to which citizens, who once had a common-law right to resist improper arrest, must now simply submit without complaint. Nor does the Constitution countenance the sort of routinely unequal and lethal enforcement regularly practiced by police, and while the Court may say otherwise, it does so by pawning intellectual credibility in service of paramilitary operators who require no further help from fellow civil servants also clad in black.

Nothing is less compatible with American traditions, or Peel’s principles, than martial law, the suspension of regular order in favor of discretionary rule by militarized forces. Policing may be far from those shores now, but that’s the direction in which it drifts, with sanction by courts and politicians citing crime waves that do not exist and intractable circumstances that somehow do not obtain in other advanced nations. The consequences for liberty are serious, and for evidence of their scope, one need look no further than the enormous prison state that warehouses the products of police zeal. We’ll take up prison shortly. But one subject must be addressed first. Is any of the evidence that bridges the river of doubt, from arrest to conviction, even trustworthy?