CHAPTER 1
THE INTERNATIONAL CONTEXT – EXPERIMENTS IN POLICING SINCE THE 1970s
Sector Policing on the West Rand
Three Case Studies
JONNY STEINBERG
Back to basics – does policing reduce crime?
It appears that the idea of breaking police station jurisdictions into sectors and having dedicated sector-based teams do policing was first conceived in the United States in the early 1970s.2 The timing is no coincidence. America was experiencing the steepest and most prolonged escalation of violent crime in its history. Between 1963 and 1973, the ratio of homicides to the general population more than doubled from 4.7 to 10.2 per 100,000.3 Long periods of escalating crime inevitably get police forces asking whether the way they police is working and whether new methods ought to be tried. The early 1970s was a time for soul searching. Sector policing was one among a slew of experiments.
The problem in early 1970s was particularly severe. For the previous two decades, police forces in large U.S. cities had generally experimented either with placing more emphasis on random patrols or on rapid response.4 Neither was working. This was evidenced not only by the fact that violent crime levels continued to rise, but in a number of research experiments. The most famous of these was conducted in Kansas City, Missouri in the early 1970s. The experiment spanned three sectors of the city. In the first, the number of random patrol cars was doubled. In the second, the number of patrol cars was left unchanged. In the third, patrols were removed entirely: police only entered the sector to respond to a call. Several months later, crime levels in all three sectors were much the same as they were before the control experiment. The numbers of police made no difference.5
This experiment and similar ones showed fairly conclusively that random patrols and rapid response do very little either to prevent crimes or to catch criminals; that what the police had been doing during the previous two decades was next to useless. Policing theorist Carl Klockars captured the mood of police scholarship at the time when he said: “It makes about as much sense to have police patrol routinely in cars to fight crimes as it does to have firemen patrol routinely in firetrucks to fight fires.”6
So, the policing experiments of the early 1970s emerged in a particular context: one in which police leaders and scholars had gone right back to basics and asked what it is that the police do to reduce crime. The endeavour to answer this question has produced a host of policing innovations in the last 30 years. These innovations can be divided into four categories: 1) hotspot or targeted patrolling; 2) controlling risk factors; 3) problem-oriented policing (POP); and 4) community policing (COP). All four innovations begin from the premise that random patrols and rapid response have a minimal effect on crime reduction. All attempt to deploy visible police more creatively and thoughtfully in order to prevent crime.
You will notice that sector policing makes no appearance on this list. Sector policing is, really, an eclectic composite of some of these innovations. Targeted patrolling and controlling risk factors are not core, definitional components of sector policing, but sector policing can, and usually does, incorporate them. POP and COP are core definitional components of sector policing; without them, the very idea of sector policing loses any meaningful definition.
Targeted patrolling and risk factor control are discussed first, followed by the two innovations at the heart of sector policing – COP and POP.
Hotspots or targeted patrols
Crime is never spread evenly across an entire city. Nor indeed is it ever spread evenly across a neighbourhood or even a single street. A large study conducted in the U.S. in the 1980s, for instance, found that all crime comes from less than three percent of addresses in a city; and even there, crimes are bunched around particular days of the week and times of day.7 This has led Lawrence Sherman to argue that giving each part of the city its “fair share” of policing “may be as useful as giving everybody his or her fair share of penicillin – regardless of whether the person is sick.”8
At the simplest level, the idea of hotspots or targeted patrols is to put visible police patrols where and when crime happens – according to geographic location, time of day and even time of year. Several experiments conducted in the 1980s show that targeted patrols can bring the crime rate down considerably. But they can also backfire and cause crime to increase. To work, they must be designed effectively and executed with intelligence and skill. First, there is the question of police numbers and of the duration of hotspot patrols: the police presence in the hotspot is most effective when it is sudden, large and of limited duration – not shorter than five minutes, but no longer than ten minutes. An experiment conducted in Minneapolis in 1988-1989 found that “the longer police stayed, the longer the hotspot was crime-free after the police departed – but only up to a point. Five minutes of police presence was more effective than one minute, and ten minutes better than five. But much more than ten minutes of police presence produced diminishing returns.”9
Second, there is the question of what the police do for the five or ten minutes they are in the hotspot. Merely remaining in patrol cars or watching street life passively appears to be less effective than policing aggressively. As Sherman argues: “Arrests for minor public infractions by pedestrians and motorists, for example, shows a clear connection to robbery: the more police enforce traffic and disorderly conduct laws, the less robbery there is. High rates of enforcement in public places may convey a sense of control that generally deters street crime, especially among strangers.”10
Yet the line between policing aggressively and policing provocatively is a thin one. Crossing that line can have disastrous effects, pitting the police against entire sections of the population, and triggering widespread public disorder. Indeed, some have argued that targeted patrolling is a blunt instrument: it is simply not subtle enough to respect the line between aggressive and provocative policing. The collateral damage caused by the very successes of targeted policing, the argument goes, means that it is always short lived; it always provokes a backlash and is replaced by gentler but less effective forms of policing.11
Controlling risk factors
Strictly speaking, this is about directing police to look out for things which have been identified as causing elevated risks of public harm. Three of the most obvious risk factors in regard to South African crime are guns, alcohol and ex-convicts. So, risk-controlling strategies may range from efforts to keep guns out of shebeens, to preventing parolees from habituating places where they are most likely to re-offend. Examples of risk-controlling strategies vary from the subtle and the finely-honed, to the blunt and diffuse.
An example of a classic risk-controlling initiative comes from Bogotá and Cali in Colombia and concerns the control of guns. Colombia is notorious for having the highest rate of firearm murders in the world. From 1983 to 1993, the annual homicide rate in Colombia increased 366% from 24 to 88 per 100,000 of the population. More than 80% of these homicides involved firearms. In the mid-1990s, a thorough docket analysis conducted in the cities of Bogotá and Cali found that the majority of gun-related homicides in public spaces took place on public holidays, election days and on the weekends immediately following payday. Government banned the carrying of concealed weapons on those days, and the ban was policed with enthusiasm; patrol cops were instructed to stop dozens of young men on the assigned days and search them for firearms. Within a year, Bogotá’s homicide rate had declined by 13% and Cali’s by 14%; young men stopped carrying weapons on the days they were most likely to use them.12 Here, the risk factors identified were guns, and the days of the year on which they were most used. The risk-controlling strategy took the form of a well-publicised amendment to the criminal law, and a city-wide reorientation of patrol tactics.
Yet risk-control strategies are vulnerable to the same danger that confronts hotspot policing; the risk factors must be policed aggressively, and the line between aggressive and provocative policing remains a thin one. How do you target those of Bogotá’s young men who carry guns without targeting all of Bogotá’s young men? There is a danger that policing will be reduced to a campaign of serial harassment against entire segments of society. Indeed, New York, which policed for risk factors aggressively throughout the 1990s, witnessed a backlash against provocative policing in the late 1990s; the trigger was the apparent wrongful shooting of a young black man, which led to days of civil unrest in parts of the city.
Sector policing
Sector policing, as indicated earlier, is a composite of two innovations in policing: COP and POP. Before defining these concepts, we would do well to heed the advice of American policing scholar David Bayley. “Wherever I go,” Bayley writes, “I am enormously impressed with the strategic creativity of American police. But what they are doing is so diverse it is hard to describe. For example, Edward Maguire and his colleagues needed 31 separate categories to capture activities that are associated with community policing … COP and POP have been wonderful philosophic sticks for encouraging the police to re-examine customary strategies, but they are awkward descriptive terms for what has been taking place.”13
Sector policing is not a rigid model or a finite set of rules. It is an eclectic composite of open-ended ideas; the practices associated with it are diffuse and myriad. The definition that follows tries to show just how wide an array of different practices the idea of sector policing sponsors.
Problem-oriented policing (POP)
This is perhaps the most innovative development in crime prevention in the last 30 years. Its underbelly, though, is that it requires radical departures from orthodox police methods and presents serious, some argue insurmountable, challenges to police organisation.
POP essentially takes the logic and methods of public health and epidemiology and applies them to policing.14 For instance, tuberculosis was largely eliminated in the developed world when public health analysts identified the social practices that stimulate its transmission and eliminated them through public campaigns and social work programmes. POP, to use a very straightforward example, virtually eliminated robberies of bus drivers in New York in the 1970s by introducing exact change fares and a barrier to cash availability.
In essence, POP consists of breaking crime patterns down to micro details, reducing each micro pattern to a particular problem or cluster of problems, and then managing or solving that problem. POP strategies range from the very simple to the complex. An example of a simple POP strategy: after carefully analysing a spate of robberies at 24/7 convenience stores in Gainesville, Florida in mid-1980s, police found that robbers only targeted 24/7s which employed a single cashier in the early hours of the morning. Police recommended that all-night stores employ at least two cashiers at all times; the recommendation was passed into law.15
POP gets considerably more complex than that. In the late 1980s, the Newton Street area of south-central Los Angeles experienced a sustained spate of street shootings. The effects on the neighbourhood were severe. Attendance at the local high school dropped to less than 50% as students were too afraid to walk to school. Adults barricaded themselves in their homes at night and confined themselves to rooms that did not face the street.
Police canvassed residents door-to-door and conducted targeted interviews among drug retailers in the area. They found that the violence stemmed from one identifiable source: a group of drug dealers in the neighbourhood to the south of Newton Street was trying to intimidate incumbent Newtown Street dealers off their turf, primarily by spraying their market place with bullets shot randomly from passing cars. In consultation with several well-attended community forums, local government acted on this information by erecting metal gates at the ends of nine streets to create a network of cul-de-sacs, making the area difficult to drive through for outsiders. The shootings quickly abated and school attendance rates doubled; drug market rivals no longer risked coming into the neighbourhood to shoot in the absence of quick exit opportunities.16
In this example, the origins of POP in public health are clearly visible. Instead of treating the symptoms of Newtown Street’s ‘illness’ – trying to catch people who shoot other people in the streets – the area was ‘vaccinated’ against drive-by shootings emanating from the neighbourhoods to the south.
Despite its elegance, POP is extremely difficult to implement. Among several problems it has encountered, the most severe is the radical transformation in policing practice that it requires. As Lawrence Sherman has pointed out:
There is no underlying theory, like the biochemistry of vaccines, that can guide police from one crime problem to the next. Each specific crime problem tests anew the ingenuity of anyone attempting to solve – or more modestly, to manage – that problem. History predicts that such trial-and-error efforts will fail more often than they succeed… That is just what has happened in many of the police agencies adopting P.O.P. since the early 1980s… Policing is an occupational culturedriven by rapid responses to short-term problems, unaccustomed to judgments about success or failure. The new assignment of long-term problem solving has often caused occupational culture shock.17
Indeed, for POP to exist in more than just name, the police organisation itself, and not just individual beat officers, must significantly revise the way it thinks about and does its work. A beat officer can detect as many problems as he likes, but if his organisation does not give him the time and the institutional support to analyse problems and experiment with solutions, he will stop looking for them. Moreover, POP often entails work which flies in the face of traditional police performance indicators as such arrest rates and response times. Radical departures from police practice often trigger resistance, both from beat officers themselves and politically. Especially at times when the fear of crime is high, it takes a brave politician to abandon or modify traditional police performance criteria.
Community policing (COP)
A host of ideas, practices and goals inform COP, and it is difficult to pin it down to a single, pithy definition. One of these ideas is that a local police force that knows its constituency well will gather the sort of intelligence that will help it to prevent crimes. Another is that a police force that liaises with its constituency can draw civilian bodies into crime prevention initiatives. These ideas manifest themselves in several forms: formalised and regular police-community forums; goal-specific joint programmes with business associations and resident bodies; beat police taking more responsibility for following through on citizen requests for assistance as well as on complaints to crime.
In the SAPS’s Draft National Instruction on sector policing, COP rests on two pillars. First, each sector manager is tasked with establishing a police-community Sector Crime Forum (SCF). Second, he or she is tasked with liaising with and mobilising community-based crime prevention initiatives such as neighbourhood watches.
Community policing has been shown to bring success, but usually in particular circumstances and in regard to the execution of particular goals. The Newton Street, Los Angeles example cited above is a case in point. The problem of drive-by shootings would never have been delineated as finely as it was without extensive police community-interaction. And the solution – the erection of metal street barriers – could not have been implemented without community cooperation.
There are a host of examples of creative community mobilisation. In a particularly innovative partnership in London in the 1980s, an open drug market on a high street was pushed off the streets. Drug retailers would spend their days seated at tables at the front of fast food restaurants. Buyers would signal to them from the street to commence a transaction. Police persuaded restaurant owners to move their smoking sections to the back of their premises, away from the front window. Since the vast majority of drug retailers smoked, they moved to the back of the restaurants. The drug trade did not cease: contact between sellers and buyers moved to cell phones instead of through restaurant windows. But the drug trade was taken off the street, which was the intention of the partnership.
Another innovative partnership evolved in Sea Point, Cape Town in 2004. Members of a Sea Point anti-crime forum dressed in bright orange bibs, assembled at drug retail points late at night, and stared drug retailers in the face for hours on end. One of the ideas behind the initiative was to shelter the neighbourhood from the dangers of provocative policing. The cold stares of civilians, it was thought, would be as effective at inhibiting drug dealing as aggressive, targeted patrolling. This is a creative example of a common and rudimentary neighbourhood partnership: using civilians as security guards who watch hotspots, repeat offenders and repeat victims, often in radio contact with police patrols.
But these examples of COP worked because police-community liaison was honed to a particular problem and in the search for an elegant solution. When the goals of COP are ill-defined, problems will arise. Indeed, more often than not, COP has arisen at times and in places where police-citizen relations are poor. The thrust of its intervention has been to restore police legitimacy. The result has often been a conflation of goals. Restoring police legitimacy is not necessarily the same as getting the police to reduce crime. Time and again, these two goals have been conflated, not least in South Africa after 1994 with the establishment of Community Police Forums.
COP has also at times been offered as a general palliative to all problems. Yet sometimes, as James Wilson argues, the problem simply isn’t one of communication between police and citizens. “There is genuine conflict,” he argues, “between [for instance] youths, who want to be left alone, and the police, who regard the young (rightly) as the chief source of crime and disorder, and who seek various means, some proper and some improper, to control them – often on behalf of older [citizens] who want ‘better police protection’”.18
Community policing, then, is probably a necessary ingredient for best practice, but it is also often among the cocktail of ills found in worst practice. Much depends on the skill and intelligence with which it is implemented, the manner in which it is received by local police culture, and, above all, the uses to which it is put.
Policing in geographic sectors
This is obviously a prerequisite of any definition of sector policing. The police station jurisdiction is divided into several sectors, and a number of staff members are set aside to devote exclusive attention to each sector. The rationale of policing in sectors is instrumental: it is not an end in itself but a tool for making COP and POP more effective. To have any meaning at all, a degree of responsibility and decision-making, and not just personnel, must also be devolved to the sector level. This devolution has taken a myriad of forms, driving sector policing in several directions. For instance, is sector policing practiced by all uniformed police, or is it assigned to a separate unit? Should sector police patrol in cars, on bicycles, work out of mobile officers or be based at the police station? What precisely do they do – are they responsible for responding to calls, or should they be exempt from responsive policing and given the space to research and analyse problems? Should the detective branch also be devolved into sectors? All of these options have been included at various times and places. The job of ‘the sector police officer’ has spun off in a hundred directions.
Problems with sector policing
The authors of the previous Institute for Security Studies monograph on sector policing argued that the concept had been abandoned in its countries of origin and that South Africa was thus “adopting an idea that has failed or fizzled out” elsewhere.19 The issue is not quite that simple. As the definition above shows, sector policing is 1) an eclectic composite of several ideas and 2) open to a wide array of different interpretations and practices. Declaring that ‘it’ has failed is to assume that ‘it’ is a monolithic model which either stands or falls.
Sector policing is not so much a monolithic concept as a cocktail of many of the ideas that have developed in the field of crime prevention in the last three decades. All of these ideas – policing in smaller units, COP, POP, community mobilisation – are extant in dozens of cities and towns across the northern hemisphere. In cities such as Chicago and Boston, models of policing, which while not formally called ‘sector policing’, have adopted all of its pillars and are alive and flourishing.20 So, rather than talk of its demise, it would perhaps be more fruitful to talk of the problems it has encountered, and there have been many.
Political problems
The London Metropolitan Police adopted the idea of sector policing with great enthusiasm in the early 1990s only to abandon it in the mid-1990s. Whatever the problems with sector policing on the ground, it can be said with certainty that the concept ultimately whittled as a result of political initiatives taken in national government. As Dixon and Rauch point out, in 1996 then UK Prime Minister John Major launched a Citizen’s Charter initiative along with a “fresh wave of targets and performance standards”:
These standards tended to emphasise performance against traditional ‘hard’ crime targets and were enforced with scarcely diminished rigour when the Labour Party came to power in 1997. While the long term problem-solving work of sector policing might eventually reduce demand for police services, statistical targets for the time taken to respond to emergency calls and the judicial disposal of criminal cases had to be met in the here and now. The inevitable result was that available resources became increasingly focused on dealing with the traditional priorities of crime fighting and incidence response.21
Indeed, all models of policing are prone to the vagaries of politics, perhaps more so than any other sphere of public policy. This is so for two reasons. First, the debate about which forms of policing – if any – reduce crime is undecidable. Some argue that we are no closer to knowing which models are more efficacious than we were when the famous Kansas City experiment was conducted in 1973. The result is that defining best practice will always be subject to heated debate.22 Fashions come and go and then come again. Second, political decisions in regard to policing practice tend to follow well-worn cycles. When the issue of police legitimacy, rather than crime levels, is prevalent in public discourse, ‘softer’ innovations like COP and POP win political space. When the public mood swings and the fear of crime predominates, the pressure to produce short-term results, in the form of increased arrest rates, quicker response times, high visibility saturation campaigns, and so forth, tends to drain COP and POP of political support. The result is that COP and POP programmes become radically redefined to accommodate hard, law and order policing, or they are abandoned completely, in both name and substance. In South Africa, where the governing party has placed great emphasis on traditional performance measures in recent years, this a point worth bearing in mind.
The pains of community involvement
In a famous article titled “The Asshole”, John Van Maanen described police culture by way of officers’ view of the public. He argued that police divided people into three categories: suspicious persons (believed to have committed crimes), assholes (people who incessantly challenge police legitimacy), and know-nothings (ordinary citizens).23 Following this bleak and jaundiced typology, sector forums are attended primarily by assholes, partly by know-nothings, and sometimes even by suspicious persons.
Beneath the frivolity of these terms lie some serious problems community involvement has encountered. Dozens of studies from around the globe have come across similar community forum failures and limitations, such as 1) ethnically diverse neighbourhoods are seldom evenly represented in community crime forums: the elderly, the white and the middle class predominate, 2) residents tend to bring quality of life matters rather than crime matters to forums, matters police officers believe are either beyond their jurisdiction or demeaning, 3) in diverse and socially divided jurisdictions, police are at times in danger of being co-opted by one social faction in its internecine conflicts against another. In this context, community involvement can do great damage to police legitimacy.
As James Q. Wilson has dryly commented, reflecting a common police complaint with COP: “police recognise that since the vast majority of citizens commit no serious crimes and know no serious criminals, they have little information to offer.”24
The point is that the goals of community involvement need to be carefully delineated and the terms of community engagement thoughtfully considered. Failure can, ironically, lead to estrangement between police and residents.
Undefined goals, police resistance, organisational failure
In his seminal article on POP, Herman Goldstein quoted from a newspaper article he had just read. “Complaints from passengers wishing to use the Bagnall to Greenfields bus service that the drivers were speeding past queues of up to 30 people with a smile and a wave of the hand have been met by a statement pointing out that ‘it is impossible for the drivers to keep their timetable if they have to stop for passengers’.”25 Organisational change, Goldstein was saying, is always in danger of losing sight of its goals.
Sector policing initiatives are particularly prone to this syndrome. A host of potential goals are implicit in the concept. Choosing them can become a muddled affair, however. Police officers are told to abandon the way they have always done their work, to operate according to new performance criteria and under stricter managerial supervision, but without a crisp purpose. Sector policing initiatives can thus trigger passive resistance in the organisation. Police officers are demoralised; many find ways to skirt new responsibilities and cling to old styles of policing; managers lose their authority.
Even when goals are clearly spelled out, there is a danger that the organisation simply doesn’t have the capacity to help police officers to implement them. Many of ideas that animate sector policing are complex and making them work requires a great deal of managerial skill.
POP in particular is managerially ambitious. It is labour-intensive, plodding and prone to error; it requires subtlety and ingenuity; it also requires the entire police organisation to reorient itself. A police officer tasked with solving problems, but who isn’t given the institutional support to do so, finds that he is asked to produce a form of policing at which he simply cannot succeed. In this regard, the history of POP is not a happy one. In a wide survey of POP, Ronald Clarke recently reported that despite its enthusiastic formal adoption across the US, “few projects consistent with Goldstein’s original vision of problem-oriented policing have been reported,” not because they haven’t been tried, but because they have failed.26
In South Africa, where the skill-base the SAPS inherited from apartheid is poor, the danger of asking too much of policing is a grave one.
