During the late 1970s, my wife and I visited New York City’s Washington Square Park. Despite the presence of several police officers, we were solicited three times by drug dealers before we were more than fifty feet into the park. The officers paid no attention. Perplexed, we moved to a position to observe both dealers and police. The dealers were openly selling drugs and the officers were doing nothing about it.

Was this a case of police corruption? Very likely not. In fact, the phenomenon my wife and I observed was almost certainly the result of police practices developed since the turn of the century for the very purpose, among others, of preventing corruption. Understanding and reconsidering these practices is vitally important to New York and other cities that are moving toward community policing.

Community policing entails both a shift in the function of police, from a narrow focus on law enforcement to the provision of a broad range of services, and in tactics, from primarily reactive activities (responding after a crime has occurred) to preventive activities (maintaining order and solving problems before they escalate into violence). But changing the functions and tactics of an organization also requires fundamental changes in the way it is structured and managed. Community policing will fail if police departments remain the highly centralized bureaucratic organizations they have been since the early decades of the twentieth century.

Police often describe their organizations as rubber bands: they can be stretched and twisted, but once the pressure is relaxed, they snap back to their original shape. Some of us remember the attempts at team policing in New York City and elsewhere during the 1960s and 1970s. Team policing had many of the same approaches as community policing: community involvement, permanent assignment of officers to neighborhoods, determination of police priorities on the basis of citizen input, and a focus on neighborhood problems. Team policing vanished, despite its successes, because issues of organization and management were ignored. Unless such issues are addressed, community policing is likely to suffer the same fate.

Today, most departments are organized functionally. That is, each major function—patrol, detectives, and various special units—is organized into a separate bureau run by an upper-level administrator (depending on the department, an assistant or deputy chief, inspector, major, or similar rank). Each of these bureaus runs independently, has citywide jurisdiction, and is overseen by the department’s chief of operations.

In most cities, a captain or commander, accountable to the head of patrol, runs a precinct or district. Assisted by patrol officers, sergeants, and lieutenants, he provides the basic patrol services with which most citizens are familiar. A precinct commander may also have a few detectives, community relations officers, and special unit personnel under his command. But criminal investigation (detectives) and special units to deal with special problems (drugs, intelligence, and vice, for example) are for the most part the province of separate centralized commands. Police leaders have considered this organizational structure so basic to policing that, until very recently, the Law Enforcement Accreditation Commission would not certify a police department unless it was organized by function.

This structure grew out of police reform efforts early in the century. Reformers’ primary goal was to control officers and departments so as to achieve independence from political meddling and reduce police corruption and abuse. In quieter days—the 1930s through the 1950s—this strategy may have been adequate to deal with urban problems as well; it has, however, proven inadequate since then. Today, a new generation of reformers is recognizing the need for a new organizational approach, one in which the basic function of all other personnel and units is to support patrol, rather than the reverse.

Reform, Then and Now

Early police departments, subject to local political control, provided something akin to community policing. They were structured not by function but geographically, with police precincts or districts overlaying political wards. Ward leaders appointed precinct commanders and held them accountable. Because precinct commanders had to please local leaders, who had a vested interest in pleasing their constituents, it was clear that solving local problems was police business.

The strong links between local politicians and police, however, also bred corruption. Sociologist Egon Bittner drolly describes the police of that era: “of all the institutions of city government in late-nineteenth-century America, none was as unanimously denounced as the urban police. According to every available account, they were, in every aspect of their existence, an unmixed, unmitigated, and unpardonable scandal.” Reformers identified three forms of corruption which they linked to political meddling in policing: financial (payoffs for underenforcement of laws), political (“getting out the vote” for local politicians), and inequitable (for example, keeping “undesirables” out of neighborhoods).

Initial calls for reform originated outside policing during the 1890s—in New York City most notably from the Reverend Charles Parkhurst. By the 1910s, police officials began to join the movement for reform. The formula reformers devised to end political and financial corruption included several important elements:

* Creation of civil service procedures to ensure that officers would be hired on merit rather than through patronage;

* Provision of tenure to police chiefs to ensure their autonomy;

* Measurement of police success by “scientific” standards such as crime rates and arrests;

* Creation of police districts and beats on the basis of crime rates (and later the number of calls for service); and,

* Narrowing the police function to the enforcement of laws, especially laws against serious crime.

Centralization of police authority was the mainstay of all reform efforts. It took two forms: consolidation of city police into higher political units (metropolitan areas and states) and centralization of police management authority in the office of the chief.

The push for state or regional control largely failed. But the effort to centralize authority in the office of the chief (sometimes called the commissioner or superintendent) was a resounding success. Police executives, supported by Progressives and other advocates of urban reform, gained more and more independence from city hall, ward bosses, and other political figures. In Los Angeles, for example, chiefs in protected civil service positions had virtual autonomy in running their department. Boston created a strong commissioner model in which appointments were not coterminous with mayoral terms. (In practice, however, mayors have been able to pressure commissioners into resigning.) A similar form of administration developed in New York City, although, as Thomas Reppetto of the Citizens Crime Commission has pointed out, the tradition of each mayor appointing his own commissioner limits the power of the office.

In all these cases, reform efforts substantially reduced the influence of political leaders on police departments. So successful were reformers in gaining independence that the University of Wisconsin’s Herman Goldstein has described contemporary police as the most autonomous and least accountable agency in government. (This autonomy created problems in its own right, exemplified by the post-Rodney King developments in Los Angeles.)

Reformers turned to the latest in organizational thinking of the time—the work of Frederick Taylor—to restructure both their organizations and the work of police officers. Taylor invented the factory model of organization we now know: centralized bureaucracy, extensive rules and regulations, lines of authority, layers of command with expertise lodged in middle management, span of control, unity of command, and the routinization and specialization of line functions.

The primary business of police was to be law enforcement—”crime fighting.” As law enforcers, police would need the endorsement of neither politicians nor neighborhoods. So as not to be too intrusive, they would concentrate on broadly condemned common-law crimes—murder, rape, assault, burglary, robbery—and withdraw from order maintenance activities that were inherently ambiguous. Police effectiveness would be measured by newly developed and “scientific” Uniform Crime Reports and arrest rates. Police would pass social, medical, family, and other problems to churches and other community institutions.

Tactics would be primarily reactive. Police would patrol in automobiles, intervene in law-breaking that came under their view, and respond to calls for service. Rules and regulations forbade “idle conversation” with citizens; police were to be remote and professional crime fighters who sought “just the facts, ma’am.” Detectives would investigate crimes once they had occurred.

Consistent with Taylor’s ideas, police work would be simple, routine, and specialized. Patrol officers would operate with little discretion—if someone broke the law, officers would intervene; if citizens were law-abiding, they had nothing to fear from the police. The organization would be structured by function. Special units would deal with special problems—juveniles, vice, drugs, organized crime—as they developed, and patrol officers would stay out of these matters. This explains the experience my wife and I had in Washington Square Park.

The most important organizational and managerial goal was to break the grip that precinct commanders had over their areas. All would come under central control; their links to local politicians and neighborhoods would be severed. To ensure that precinct commanders could not “build kingdoms,” reformers rotated commanders from assignment to assignment—a model developed by J. Edgar Hoover when he took over the corrupt Bureau of Investigation during the 1920s. Extensive rules and regulations would govern all aspects of police performance and organizational decorum. Centrally based middle managers—captains and lieutenants—would design the work, supervisors would direct it, and patrol officers would carry it out. Line officers would do, not think.

In many respects, the reform strategy was remarkably successful. Police departments became relatively well-run organizations. Corruption and abuse lessened throughout the twentieth century. Police embraced the ideology of “crime fighting”—it communicated action, power, and an unambiguous purpose.

The trouble is that although police generated enormous numbers of arrests, the incidence of serious crime continued to rise and street conditions to deteriorate. By the 1960s, the “crime problem” got so bad that it became a major political issue. Yet today, most of us would happily return to that era—at least as far as predatory crime and disorder are concerned. Moreover, the emphasis on “crime fighting” bred a warrior mentality in police: huge areas of cities became battle zones and large numbers of people became enemies—all a far cry from the admonition of Sir Robert Peel, the father of Anglo-Saxon policing, to “keep peace by peaceful means.” Ironically, then, the “crime fighting” strategy failed on its own terms—reducing or preventing crime.

The centralized structure of police departments made it difficult to introduce innovations similar to community policing, which might have effectively addressed these problems. This is best illustrated by what happened to team policing in Cincinnati during the 1970s—the most closely studied team policing experiment. After team policing was tried for two years in areas of Cincinnati, almost everyone involved—citizens, police, and evaluators—agreed that it was a success. The department planned to take team policing citywide. But it was understood only as a good tactic; little thought was given to the department’s organizational structure, which remained typically bureaucratic.

Over time mid-managers moved to reestablish their authority. In the name of ensuring good personnel management, they reclaimed the right to allocate personnel to teams. In the name of uniform dress codes (rules and regulations), they withdrew team leaders’ authority to assign officers to dress in plain clothes for surveillance or investigation. Other subtle means of control were exerted. But the killer was something called management by objectives (MBO), which the department adopted in the middle of the experiment.

Top management installed MBO to ensure that officers were conducting measurable activities (arrests and the like) department-wide. Teams submitted MBO plans that were based on neighborhood priorities. Management rejected them as not in accord with overall departmental objectives. When neighborhood priorities were resubmitted, management again turned them down. Team members finally got the message that Cincinnati was returning to business as usual. The rubber band snapped back: police work was riding in cars, responding to calls for service, and getting the “stats.” The Cincinnati Police Department failed to recognize that team policing, like community policing, was not merely a tactic; it was a strategic shift in the function and tactics of policing that could not be maintained within the organizational structure imposed by turn-of-the-century police reformers.

Reform failed in another respect as well. Turn-of-the-century police innovations had their origins in a desire to control police organizations and line personnel. Reformers were serious about crime control, but if push came to shove, their concern with corruption led them to make control of the police the top priority: don’t let line officers combat local drug dealing; create special units to deal with vice; limit the discretion of precinct commanders. This focus on control gives rise to two central ironies. First, line officers actually perform complex tasks with little direction or supervision—despite the fervency of the desire to control them, the general view that officers are under close control, and the existence of voluminous rules and regulations and an elaborate command-and-control superstructure. Second, the organizational measures developed to eliminate the autonomy of the precinct commanders have artificially fragmented police work, alienated citizens from police, and diffused accountability.

No Control of Patrol

Community policing depends on a recognition of the subtlety and complexity of policing. But the reality of police work—what officers actually do—stands in stark contrast with the official representation of policing.

In the official view, an extensive administrative structure oversees all aspects of an officer’s work, with great concern given to issues such as span of control (how many officers a sergeant can oversee) and unity of command (each officer takes orders from only one person). Orders flow downward from commanders to officers. Information flows upward for use by planners. Tactics, developed by middle and upper management, specify when, where, and how officers are to patrol. Extensive rules and regulations prescribe organizational manners and mores in extensive detail, down to minutiae about clothing and other forms of decorum.

Elaborate 911, computer, and radio systems keep officers in constant contact with, and under surveillance of, centralized command and control. Formal norms and informal conventions guide how long an officer should spend dealing with each type of call for service. Organizational pressure keeps officers “in service”—riding in cars waiting for the next call—rather than “out of service” and dealing directly with citizens. Reports keep detailed records of officers’ law enforcement activities: arrests, summonses, and tickets.

This top-down perspective gives a coherent view of the officer in the police organization: under tight administrative control, making decisions under the supervision of a sergeant or mid-manager, “fighting crime” by enforcing laws, going from call to call with little free time, guided by extensive rules and regulations, and operating with little discretion. But the view from the bottom is quite different.

First, police work is incredibly complex. Reform administrators were wrong to believe that police work consists primarily of enforcing common criminal laws in ways that could be carried out by people with what Egon Bittner called “manly virtues”—men who are big, heroic, honest, and smart enough to write out simple reports. In fact, most police work consists of managing community and neighborhood conflicts and problems that take an infinite variety of forms. Less than 20 percent of police work, regardless of city or neighborhood, deals with common crimes. And dealing with crimes is itself often complex: there is nothing so simple about domestic assault, for example, that suggests that “manly virtues”—as opposed to intelligence and skill—are what is needed to deal with it.

Second, police work is inherently ambiguous. When an officer arrives on the scene of an incident, it usually not clear who is doing what to whom, who is right and who is wrong—indeed, even what has happened and what is happening. Natural distortions of memory, prevarication, cheerleading and bullbaiting by onlookers, the emotions of victim and perpetrator, and a host of other factors mean that police are often asked to make sense out of confusing and ambiguous circumstances.

Third, despite the total number of calls that come to police departments, officers have considerable free time. How much varies by shift, time, and area, but even during relatively busy shifts, most officers have time that could be used to do things other than ride around in cars. The primary problem police have in scheduling work is its unpredictability rather than its quantity. The result, however, is that officers control their discretionary time.

Fourth, rules and regulations have very little to do with the routine work of police officers. Regulations specify proper procedures for arrests (which constitute a relatively small portion of actual police work), tell officers what they cannot do, and manage the relationships between ranks. But they have little to say about most of the conflicts and problems with which officers deal. Most such work, in fact, goes unrecorded. Suppose an officer sees a dispute between neighbors. If the officer intervenes before blows are struck, and all is settled, then officially nothing has happened and, except perhaps for the officer’s personal log, the department has no record of the activity. If, on the other hand, the officer delays, lets a fight start, and then makes an arrest, an official record is kept.

Finally, police are subject to little direct supervision. Most officers conduct the bulk of their critical work, including using force, alone or with a partner. They are not “overseen” in the sense of Taylor’s factory model: no one can look over their shoulders to determine whether they are doing their work properly. Generally, the quality of police work can only be determined after the fact, by reconstructing an incident on the basis of records and personal accounts of the officers, participants, and other observers. In the final analysis, then, police officers’ performance is determined largely by their own values, knowledge, and skills—be they appropriate or not.

The reform strategy put control of officers at its very core; created an elaborate control infrastructure of rules and regulations, layers of command, and monitoring technology; and ostensibly narrowed, simplified, and routinized police work. Yet this strategy not only failed in terms of crime fighting, but in terms of control as well. At its best, this strategy, with its functional structure and Taylorian managerial processes, can keep officers out of trouble by isolating them and preventing them from doing much of anything. It cannot produce quality policing because it does not address the essential work of police officers.

Powerless Precincts

Community policing is also hampered by the centralized structure’s tendency to reduce the power and prestige of precinct or district commanders, which makes police departments less responsive to local concerns.

Specialized units, as we have seen, were created as part of a strategy to achieve centralized control over the police in an entire city. Such units limited the control of precinct commanders and extended centralized control over day-to-day policing in wards and neighborhoods. Specialized, centrally controlled units, moreover, extended the reach of the police chief into precinct personnel, shaping how patrol officers view themselves and their jobs. Special units, akin to detectives, became prestige units in police departments. Working in a special unit often means special hours, increased pay and recognition, exciting work, plain clothes, and access to other good jobs in the future. By offering such benefits, police executives are able to “cream” the most talented officers from precincts.

The result is that precinct commanders get what’s left over—leftover problems and leftover personnel. Precinct personnel do the residual work of responding to calls for service, patrolling in cars, and conducting preliminary investigations, which, regardless of quality, are often replicated by detectives.

This is troublesome in many ways. Two are particularly important to the prospects for community policing. First, precinct commanders cannot be held responsible for their areas. Stripped of resources at every “emergency,” their precincts policed by any number of special units over which they have no control, their priorities determined by centralized planning units, and their officers under control of a centralized 911 system, precinct commanders lack authority and resources and are left with little to do other than follow the dictates of the chief. At the precinct level, initiative and leadership are lost, and commanders are unable to respond to local problems. Precinct commanders might have a few community and special officers to try to reach out to neighborhoods, but “real” police work is controlled and carried out elsewhere. The central bureaucracy commandeers personnel and other resources, establishes priorities, and, through 911 and its requirement that police stay “in service,” sequesters the lion’s share of officers’ time.

Second, the results are equally devastating to patrol officers’ initiative. Special units and detectives come into officers’ beats with impunity. Specialized personnel share information with beat officers on the basis of their “need to know”—which in practice means patrol officers are generally left in the dark about important information regarding their beats. This is particularly problematic in departments that formally or informally reward arrests (with assignment to special units, promotions, court overtime, and the like). Special unit personnel tend to share useless information with patrol officers but to hoard information that may result in arrests. Moreover, it is not unusual for beat officers to know nothing in advance about planned special operations in their areas. Depending on the department, patrol officers feel demeaned, frustrated, and bitter. The idea of being “chiefs of their beats” is a joke to most of them. In the worst case, officers are dispatched to a “gang war” call that is really a shootout between special unit officers and drug dealers—a situation that jeopardizes both patrol officers and special unit personnel.

How to Decentralize

Decentralization—that is, devolution of authority to lower levels of police organizations—is essential in large cities because problems are local. No city has a drug problem; no city has a violence problem; no city has a disorder problem. The substance, shape, and importance of problems vary by neighborhood. The drug problem in one neighborhood is gangs shooting at each other. In another, it is a traffic problem: cars lined up blocking streets. In another, the drug problem is panhandlers blocking store entrances. In yet another, it is youths partying in parks. And so on. Neighborhood police officers need the time and resources to discover, understand, and deal with local problems. They also need the authority to work with citizens in negotiating priorities and devising solutions.

This is not to say that there are no citywide problems. But the vast majority of urban police business is neighborhood-based. Moreover, policing is, by its very nature, geographical. Since their inception, police departments have always distributed officers throughout cities rather than clustered them in centralized offices where they would be amenable to direct supervision.

The crucial principle of community policing is that the local precinct should provide total police services—including criminal investigation, drug enforcement, and other services that are now the province of centralized units. A variety of means could be employed toward this end—organizing geographically, flattening police organizations (reducing the number of ranks), “de-squading” departments (reducing the number and size of special units), rethinking how demand comes into police departments (recognizing that 911 calls are only one source of demand), and others. All are required to some extent. New York City, for example, could give “superchiefs” jurisdiction over a borough rather than a function, reduce ranks substantially, and cut back on special units while making them responsive to patrol officers.

Strong leadership would be essential in a decentralized department. The chief (or commissioner) would strive to infuse the department with a vision of policing congruent with democratic governance and constitutional values. The chief would also be in charge of such department-wide matters as recruitment and training, allocation of personnel and resources among precincts, and coordination of activities both among precincts and with other governmental agencies.

Precinct commanders (or captains) would be held accountable for the quality of policing in their neighborhoods, based on both departmental values and local needs. There would inevitably be some tension between the departmental mission and the perceived needs of neighborhood residents; not all neighborhood demands, after all, are legitimate. Precinct performance would be judged on such factors as responsiveness to neighborhood concerns, officer integrity and judiciousness, integration of police activities with those of other problem-solving agencies (private as well as governmental), cooperation with other police precincts, and efficient use of resources. Commanders would structure beats congruent with neighborhoods, assign officers to beats and shifts, and work with those in the community to identify local problems and solutions.

What must be put aside in this mandate for change is the reflexive move to centralize police authority whenever some crisis, especially a corruption crisis, develops. During the mid-1980s, for example, a relatively small number of police officers in a New York City precinct were suspected of financial corruption. An order went out of the commissioner’s office: every patrol officer in the department would be rotated from beat to beat every six months. The assumptions behind this order were that familiarity breeds corruption and that police work is so standardized and routinized that officers can be moved from beat to beat with few consequences. The first assumption is arguably true in some cases, although I will return to this point. The second assumption, as research and experience have shown, is clearly false. Police work is complex, subt1c, and dependent on information. Moreover, citizens do not want strangers policing their neighborhoods. Police must earn the respect and trust of citizens, and that takes time—at six months most officers have just achieved it.

The crucial point is that it is one thing to attempt to organize police departments to prevent corruption; it is something else entirely to organize a department to ensure quality policing. To organize based on the misguided assumption that all officers will cynically exploit opportunities for corruption guarantees ineffective policing. Perhaps many of the draconian anticorruption measures implemented in the New York Police Department after the Knapp Commission report of the early 1970s were necessary because of the systemic nature of corruption then. But these measures—referred to by at least one of their advocates as “management by terror”—have also created an ethos throughout the department of “staying out of trouble.” The surest way to stay out of trouble, of course, is to do nothing.

Controlling corruption will always be difficult. But this problem is not unique to policing. In any organization—a university, a pharmaceutical company, a widget maker, or a police department—some corruption will inevitably exist. Acknowledging this does not imply that corruption is acceptable. Both corruption and abuse still occur far too often in American police departments. But citizens must understand that contemporary corruption and abuse do not begin to approach the “bad old days” of ward bosses or the systematic police abuse of African-Americans during the nineteenth and early twentieth centuries.

In no way should this be read as a justification of corruption and abuse. Far from it. It means taking them seriously, but doing the hard work—the very hard work—of both maintaining integrity and providing quality police services to communities. Settling for the appearance of control at the expense of good policing is the ultimate form of corruption. The business of the organization becomes looking good rather than regaining control over streets, neighborhoods, and public spaces. Community policing can be made to work only by developing police leadership that inspires a democratic vision of policing—police as an extension of the citizenry rather than as warriors—and by establishing a congruence between the realities of police work and the organization and management of departments. Policing ultimately is neighborhood-based and decentralized—that is its very nature. If police departments’ organizational structure and management processes reflected this reality, officers like those in Washington Square Park would be far better able to do what citizens expect of them.


City Journal is a publication of the Manhattan Institute for Policy Research (MI), a leading free-market think tank. Are you interested in supporting the magazine? As a 501(c)(3) nonprofit, donations in support of MI and City Journal are fully tax-deductible as provided by law (EIN #13-2912529).

Further Reading

Up Next