publisher colophon

298Chapter 7

Overcoming Obstacles to Reform

Using the theoretical framework of the first three chapters and the empirical experience of the following three, this chapter outlines five strategic approaches that could lay a concrete and visible groundwork for more participatory and effective citizen security. Although the issues to which they are most applicable differ, each of these approaches attempts to connect specific proposals to the three underlying characteristics emphasized in this book: (1) the nature of daily policing, (2) relations among institutions, and (3) drawn-out policymaking. Through their recognition of these traits, such approaches help reformers maintain the innovation and flexibility of reforms through the stream of expected and unexpected obstacles. The approaches described in detail in this chapter, drawing on the experiences of different countries, can be summarized as follows: (1) identifying the points of tension in criminal justice, the alleviation of which can help unblock slow processes; (2) taking into account the discretion of police officers, which both traditional and problem-oriented proponents often misinterpret and manipulate; (3) enhancing the role of evaluation, which should be integrated into reform rather than tacked on at the end; (4) seeking structured citizen engagement, a key to community policing; and (5) focusing on the internal management of security bodies.

Points of Tension

Policymakers are likely to become stuck at the crossroads between traditional and problem-oriented policing when they see reform as a set of unenviable choices they must foist on uncooperative agencies. But those choices might be simpler and more palatable when reform is centered on stages of security actions rather than just on security institutions. This view can be applied to each of the five areas of reform defined in chapter 1. Control mechanisms, for example, often become stuck at particularly sensitive points in investigations of police abuse. They apply more readily to legal changes, which have brought about changes in the entire criminal justice process. Nearly every component of criminal justice is a series of steps, providing openings for action as the process progresses. Because most conflicts, bottlenecks, and various forms of politicized agitation arise in the move from one step to the next, untangling these knots of tension can often be an effective reform strategy.

The description of criminal investigation in Honduras in chapter 3 reveals how many points of tension, and thus of potential change, can be identified in an otherwise chaotic process. In most countries a predetermined series of actions kicks in once a crime is committed (see figure 7.1). Tensions usually flare up between the first and second stages, when the public prosecutor (fiscal) or a representative of the Dirección General de Investigación Criminal (DGIC) arrives at a crime scene that is mishandled by the National Police (PN). Such tensions continue between the third and fourth stages because of incomplete and inconsistent dossiers (expedientes) and between the fourth and fifth stages, when extended delays, mainly stemming from officials’ poor preparation for oral trials, push detentions beyond the thirty-day limit.

Although persistent, each point of tension suggests possible responses. It is usually not feasible to assign more personnel and resources to relieve tensions during an investigation. An alternative being adopted by many countries is joint criminal code training programs for judges, prosecutors, and police. By exposing the conflicts among these officials, such training helps to minimize them during actual cases. Supporting such programs at this stage should be more training on writing expedientes, an internal affairs evaluation of evidence collection practices, and stronger programs for victim assistance and witness protection. Legal assistance units within the police can provide some of these services. Costa Rica’s police lawyers (alpha limas) advise officers on the law, particularly on the legality of specific actions. Because an estimated 65 percent of officers leave the academy without adequate legal training (Bermúdez Coward 2004), the alpha limas help them get past the points of tension in their everyday work. In Bolivia, where about 70 percent of criminal charges are for nonviolent offenses, the greater use of alternative sentencing at the beginning of the process would likewise help decongest criminal justice down the line.

Figure 7.1. Criminal Investigation Flow

Source: The author.

The points of tension can also be identified in other criminal justice processes, from help for at-risk youths to released offenders. Such an approach can be most fruitful for the process by which people report abuse, because it involves the range of citizen concerns. Most of those processes consist of the following steps: receipt of complaints, evaluation of their veracity, investigation of the accusations, adjudication of the charges, and recommendation of dismissal or punishment. The stage that is the subject of the most complaints by citizens can then be evaluated for improvement. Physically locating an accountability agency in a police station, for example, is intimidating and leads to low rates of abuse reporting in many countries (see Lemos-Nelson 2002, 197). For others, it is the police chief’s veto over agencies’ recommendations that most dissuades involvement. The UN initiative International Commission against Impunity in Guatemala, prompted by serial killings of bus drivers and designed to loosen organized crime’s grip on criminal justice, is an example of how an issue or incident can lead to a focus on a problematic stage in criminal justice that undermines the rest.

Policing itself, even when regarded as efficient, also involves many points of tension that can be openings for change. A central example is a clamp-down on antisocial misdemeanors that is given a legal and policy basis. However, because of the broad support it enjoys in most places, this form of zero tolerance policing often lacks the scrutiny that such a hands-on approach needs, even when embedded in the legal structure. Many such provisions are modeled after New York’s Quality of Life summons, which began in 1989 in the New York City Transit Authority. This summons could be issued for twenty-five transgressions, from panhandling to property abandonment. Anyone in the city who received a summons but did not appear in court was subject to a warrant for his or her arrest. The threat of arrest increased both compliance and prevention, because nearly half of those who showed up had a previous record, sometimes for felony charges. Management and human resources were additional motives for this approach. Citing the fact that 10 percent of police were making 90 percent of arrests, during his 1994–96 tenure as commissioner of the New York City Police Department (NYPD) William J. Bratton pushed officers to detain more people. He also introduced CompStat, an organization management tool that required each of the NYPD’s seventy-six precinct commanders to explain the crime-fighting strategies being applied in his or her area. Use of this tool also increased the pressure on officers to tally up arrests and Quality of Life summonses.

And yet as the popularity of and dependence on Quality of Life summonses grew, they became another standardized approach lacking corrective assessments. As summonses and related measures became an integral part of policing, they were no longer regarded as extra work requiring backup, and street officers under Chief Howard Safir (1996–2000) felt that they unfairly bore the brunt of the blame for the inevitable conflict that arose in implementation of the summonses. In addition, the growing number of detentions and summonses clogged the criminal courts, and prosecutors began to dismiss thousands of “flawed” arrests.1 A “policy of aggressive misdemeanor arrests,” as Harcourt (1998, 332) concluded, thus became the “primary engine” of community policing in New York. Such use of zero tolerance even led Kelling, one of the developers of the broken windows theory (see chapter 2) to label it “the bastard child” of broken windows because it lost track of the “extent to which we wrestled with legal, constitutional and moral issues” (Law Enforcement News 1990). Blunt applications of zero tolerance in Latin America on issues such as drugs, from small-time use in Argentine cities to large-scale eradication in Bolivia, also reflected the counterproductive effects of the practice on society, the criminal justice system, and the police itself.


Even the firmest citizen security policy with the tightest rules has a soft core of discretion, which can be defined as the characteristic of a realm of work in which specific action is determined by the actor’s own selection from a range of possible responses. Because of the inherent complexity and daily stress of policing and the impossibility of fully monitoring it, individual discretion is an inescapable part of daily policing. It is also central to reform. Positive impacts from professional support and community policing programs, in particular, often rely on what officers do with their new but often vaguely worded powers. If discretion is defined as the right or power to decide or act according to one’s own judgment, then its role in policing is clear. In each of the three stages of even the most basic police action—contact, process, and exit—officers must choose from a wide range of options in a compressed amount of time. For example, in the contact phase they can listen, question, refrain, counsel, threaten with detention, separate persons, or refer the situation to another agency. While doing so, they must provide a “presence” that brings calm and control through a combination of persuasion, mediation, and force, while identifying “signals” in order to assess quickly how multiple factors might escalate.

Needlessly adding to such inherent complexities in discretion is politics. Corresponding to the conflict between traditional and problem-oriented policing in Latin America, discretion has fallen into two clashing and politicized stereotypes: a professional prerogative strangled by regulation, on the one hand, and a personal license for abuse, on the other. These views often coexist uneasily when the legal controls over discretion can be loosely followed in practice. But any incident can easily throw off such a balancing act, triggering counterproductive patterns in which officers, untrained by their academies and unguided by their superiors, are unsure how to apply the discretion they do have. Even when street officers are motivated to do their jobs, their superiors often do not “trust them to do it” (Bratton and Knobler 1998, x). Those trying to avoid trouble then try to match their actions to official guidelines and objectives, which only reveals the dissociation between daily policing and rules to which it is supposed to adhere. For example, officers using neighborhood contacts to tamp down intergang tensions in San Pedro Sula, Honduras, often cited the Organic Law (LOPN) provisions authorizing actions for public order threats, even though no specific threat existed. To prevent their discretionary action from leading to entanglements with or accusations from citizens altogether, officers often give them pat excuses about the unbendable rules imposed from above that they are compelled to follow. All too typical is the fear by an El Alto, Bolivia, police officer of residents who might be allied with local politicians, which he said led him to ignore calls to intervene in violence between neighbors.

The tendency toward such responses is increased by officers’ awareness that the internal affairs office could be used to either scapegoat them or to shield them if they run afoul of well-connected citizens. But citizens see police agents’ pretensions of being technical as cynically self-serving and devoid of basic concern. When such conflicts and discrepancies over discretion arise and lead to public airings, outside officials push to tighten oversight. But police chiefs often fight such efforts as a zero-sum game of control. Caught up in such struggles, they miss a chance to incorporate into any new rules the informal ways in which street officers address insecurity. To avoid more political problems, they also tend to micromanage even more, which is one of the biggest and most common “failures” in policing, in part because it aggravates the problems discussed in chapter 2 (Rogers 2002, 15). Street officers then continue to be professionally unprepared but politically vulnerable. As they also continue to doubt their actions and cover up their mistakes, they dull their individual judgment, diminish public confidence, and inject criminal policy with false information.

As discussed in chapter 2, the importance of the connection between theory and tactics requires that broken windows and related approaches be linked, under careful leadership and management in the comisaría, to community norms and crime reduction goals. It is about surrounding flexibility with standards. As demonstrated by the contrasting examples of La Ceiba and Rivera Hernández in Honduras, the discretion of the local chief can be pivotal. Youths gathering on a corner are often harassed by police because they fit the stereotype of disorder, are assumed to be a nuisance to the community, and constitute a vulnerable and easily targeted population. And yet often they are not subject to complaints by neighbors, who might recognize the need for such gatherings. A local chief with political connections but without the fear of reprimand might be able to bring those factors into play. When such gatherings become regular, such as the use of a space for late-night loud parties, the chief is also in the position to use zoning regulations or other laws to end or scale down the gatherings. Making such local and specific distinctions, which is part of the nature of discretion, is also needed to accommodate any policy changes from higher up in the government.

To turn discretion from a vice into an advantage, governments should take concrete steps in police-related education, promotion, and policy. First, in training cadets academies should incorporate scenarios that evoke a range of human reactions in order to maximize innovation and criticism. Because every situation is different, as any officer will agree, sessions conducted by veterans help bring out how chronic uncertainty and subjectivity can overshadow written guidelines and set objectives. Along with concrete examples of daily temptations and the consequences for giving in to them, such an approach can limit officers’ inclinations to hide honest errors and help police managers harness the innovative power within their ranks. In contrast to mute passivity during lectures, such real-life experiences would bring police academy classes alive in Bolivia, Honduras, and Argentina. Instead of trying to scrap and replace an existing curriculum, which invariably makes academy directors defensive, this pedagogical approach is more likely to be welcomed by the higher-ups, who can see for themselves the differences in student receptivity.

Such approaches can continue as cadets enter the police force through strategies such as pairing new and senior officers in high-crime areas. For example, in Brooklyn, New York, Operation Impact has helped bring down crime rates amid salary and personnel cuts. Specialized courses on various societal conditions are useful as well. One example is mental illness, which afflicts a high percentage of those who come in contact with the police. Most officers, though, lack the skills to know whether the erratic behaviors they are encountering are temporary, permanent, benign, or dangerous, much less what services are available to treat them. Training on and responses to mental illness differ widely in Latin America. Some agencies give cadets about twenty hours of training, often using outdated videos and dry lectures, which, like some community policing and human rights courses, serve to marginalize the issue. Agencies more cognizant of their shortfalls in this area hire mental health professionals or send in social worker teams. Like the limited courses, however, this approach deprives the majority of street officers of the kind of preparation that enables effective discretion.

Helping to reinforce more discretion-centered training would be a promotion structure in which advancement is tied to skills and achievements on matters such as community relations, with extra incentives for specialization in functional and geographic areas. Ecuador offers promotion and salary incentives for relocation to undesirable locations (Policía Nacional de Ecuador 2003), and Costa Rica provides them for high-risk assignments. Awards and other forms of recognition, such as NYPD’s “cop of the month,” cost little but mean a lot to serving officers, and so should also be part of the police career. Other approaches, such as a public scoreboard of actions carried out by each officer or the 2005 proposal in Guatemala to assign entire academy classes to comisarías, can also help turn innovative power into hierarchical power (Scheye 2005, 14). Many such actions can be taken using administrative means that minimize the political entanglements. In Venezuela the municipality of San Francisco in the city of Maracaibo has adopted more citizen channels for complaints, opened an Autonomous Police Institute, and created a Community Affairs section. Those changes have facilitated activities and initiatives that have proven effective, said one district chief, such as better use-of-force discretion, a requirement for serving officers to return to the academy for ongoing education, and evaluations of officers by residents. Together, these measures have helped reduce both criminal and police violence, which have long been at crisis levels in the Maracaibo metropolitan area.2 By contrast, the shame or bullying that often pervades informal interpolice relationships should not be used, as some agencies have tried, in the name of evaluation. Such disincentives, such as Thailand’s practice of requiring errant officers to wear pink “Hello, Kitty” cartoon armbands, may create resentments that offset their short-term impacts and embolden the kinds of repressive subcultures discussed earlier.

Discretion is also strengthened through exposure to different perspectives and options from which police officers can draw. That said, law enforcement officers justifiably suspect many ideas coming from government, international organizations, and other outside sources with less law enforcement experience than they have. Universities can be an exception, because they are usually more knowledgeable, professionally stable, and politically neutral than other outside entities. Although police-academic collaboration is vulnerable to politics and distrust, in Latin America it has been able to withstand stereotyping as leftist by proving to be constructive when focused on a particular issue. The police in nine Brazilian states have developed educational programs with local universities (Leeds 2007, 27). Most notable is the productive collaboration between the police of Minas Gerais and the Universidad Federal of Minas Gerais. Ecuador’s San Francisco de Quito University trained six hundred security officers in 2005, and Chile’s Carabineros developed their Plan Cuadrante with the Catholic University.

The high rate of shootings by police in Latin America, reaching over 40 percent of incidents of civilian resistance to police in Venezuela,3 demonstrates the need for a special focus on the use of force through the lens of discretion. Because whether to use force is the most serious and high-pressure decision a police officer makes, guidelines and training on force must be as clear and thorough as possible. They also must be supported by better management; several studies, underscoring the importance of evaluation, indicate that excessive bureaucratization more than other variables increases the likelihood of the use of force (Worden 1996, 44). Thus, although improved management is a long-term effort, it can begin with steps such as ending the common practice of having officers throw their guns back in the cache at the end of their shifts without any notation of the use of those weapons during their shifts. But as the next section discusses, such steps will have limited results unless police officials and officers understand why and how the problematic practices were part of daily policing in the first place. Discretion-centered training on the graduated use of nonlethal force, by separating out and analyzing the different life-threatening scenarios police encounter, can help create that awareness about the use of force while gradually reducing it. In fact, many academies have been adopting a scale for the application of nonlethal methods for a range of violent situations.4 Because of the high number of guns and crimes committed with them—an estimated two million arms, legal and illegal, are circulating in Central America alone—greater discretion on weapons searches can also cut crime rates while improving police professionalism. Officers discover most arms in the course of other activities rather than in specific searches for them. So they could be given broader authorization with clearer laws to search people for weapons on the condition that any discovered arms would be destroyed without detention or criminal charge. Making new legal controls more acceptable to police and allowing more individual discretion would support reform in general by signaling more confidence in officers.

Finally, discretion can help reform along by being a key term of reference in a terminology of reform tweaked to make it more politically palatable. For example, many kinds of “reform” can be introduced instead as a “strengthening,” and problem-oriented policing can be couched in terms of coordination rather than of control. Repeated references to discretion are particularly important, because such references signal to the rank-and-file that they will be allowed at least some room to adjust new approaches to their daily challenges and limitations. The proactive approach by the CMS officers in the Honduran city of La Ceiba, described in chapter 4, illustrates how such approaches can be put into practice, even in high-crime areas with few resources. Although they never discarded or dismissed traditional tactics, that city’s officers focused on social issues in terms of their importance to criminology, using family visits, youth sports, and other forms of “soft” policing in cracking down on gangs and drug trafficking.

Information-Based Evaluation

Information that is reliably collected and objectively analyzed can help make headway toward problem-oriented policing. Criminal justice, community policing, and accountability mechanisms are often immersed in reams of information, only some of which is relevant. The key indicators of crime and policing are collected in most countries, but much of it remains adrift in an unfiltered flow of reports, statistics, and numbers.

To break out of such informational inertia without triggering too much resistance, a good approach is to initiate separate evaluations of officers, units, and practices. Such evaluations can help identify weak areas, sharpen officer discretion, improve station chiefs’ planning, make personnel distribution more equitable, and prevent policy from congealing around conditions that change quickly. Such an approach is important in each of the five main areas of reform described in chapter 1. A restructuring that eliminates and creates ranks and bodies must be evaluated to ensure the viability of new agencies amid resistance by powerful internal competitors.

Evaluating Officers, Units, and Practices

Such a restructuring should be anchored in an evaluation of criminal justice officials at all levels, with an eye toward boosting incentives for effectiveness. Quality control teams could determine how street officers spend their time. Knowing which repeated problems take up an unnecessarily large part of a police officer’s day—such as chronic vandalism of a basketball court, calls from particular homes, or robberies on a commercial block—would allow the police to better direct officers for rapid follow-up and better refer people to the appropriate services. Evaluation of the types and numbers of arrests would be useful as well. Criminology seeks quantifiable measures, and a broader evaluation of arrest patterns in the context of human resource planning may expose problems such as an overreliance on detention quotas that are met through spurious detentions, or a dependence on the minority of officers (sometimes just 10 percent) who make most of the arrests in many cities (Bayley 1998, 122).

To address the complaint that most detainees do not end up in jail, both detectives (inspectores) and street officers must be thoroughly evaluated. Individual police officers pay little attention to conviction rates because they do not affect their salaries or promotions.5 Indeed, some cities actually have more inspectores than cases solved by them. Results-based oversight, clearer timetables of action, and individual incentives can begin to improve and smooth out police performance. In Mexico’s Federal District, an officer making an arrest that leads to a conviction receives a small bonus. A program in New York City’s Bronx borough that helped officers improve their presentation of the evidence in reports sent to the prosecutor raised the conviction rate. The level of convictions was also included in officer evaluations. A review of differences among regions may also help explain why some areas are more able to solve crimes than others.

Because policing goes beyond individuals, units must also be evaluated. Police stations, in particular, should be assessed, with financial awards and professional recognition for those demonstrating reductions in crime without accompanying increases in abuses or complaints (such an evaluation was initiated in Argentina’s capital). This balance between effectiveness and abuse should be part of the evaluations of specific units as well, especially those given greater discretion. For example, in New York City, the NYPD’s Street Crimes Unit earned widespread recognition for its success with sting operations and arrest rates, and was eventually responsible for 40 percent of the guns the police took off the street. But it also killed a disproportionate number of unarmed civilians, stemming in part from the growing discretion that its successes brought it. Officers on foot patrol should also be more carefully evaluated. Even though they are not more effective than other types of officers at reducing crime, they do reduce citizens’ fears and feelings of insecurity. Thus the ways in which they achieve the latter reductions should be studied and incorporated into discretion training. In most cities in Latin America less than 10 percent of 911 calls are about criminal matters and less than 5 percent of crimes are reported while they are in progress. A rapid response does not therefore significantly increase arrest and conviction rates (Spelman and Brown 1984), and so response times that are predictable and reasonable may mean more overall than those that are quick. Greater awareness of these patterns may help reduce the uncertainty and tension in police-society relations discussed in chapter 3. Indeed, categorizing actual police requests and expectations by different social sectors should be central in evaluating policing.

Practices, too, require greater evaluation. Like internal affairs, practices that draw the most criticism are good candidates for better evaluation. Such practices range from traffic stops, notorious for abuse and problem displacement, to an officer’s reliance on firepower. Even with strong legal constraints on the use of force by police, still missing is any analysis of when and why it is most likely to occur. Such analysis should start by assessing whether force actually works, because police often miss their targets when they use their arms. In 2004 and 2005, for example, New York police officers hit their targets less than 23 percent of the time.6 Such studies should be supplemented by a broader understanding of the use of violence in different scenarios. How often do officers use coercive means against a citizen in situations with no resistance or detention? How often does violence ensue because the officer lacks equipment or is unable to get backup? How often does violence occur after detention? In many countries most police violence occurs during detainee treatment and transport, even though that is when it is the least necessary, expected, and justifiable. A better understanding of such practices can enhance discretion and improve policy. For example, beyond assigning and tracking specific arms to each agent, any officer using fatal force could be required to complete a form describing why, where, and how it occurred. Worded to be more informational than accusatory, such forms could help determine whether an action was necessary and whether it presented any legal or policy problems.

How police practices, crime trends, and public space affect each other should also be evaluated, as is done in many countries through geographic and environmental studies known as Crime Prevention through Environmental Design (CPTED). Some European cities, for example, have used this approach to reduce crime by moving recreational fields away from housing centers. CPTED could also be applied to public spaces such as banks and kiosks; the police could draw up a list of locations with the most frequent crimes and monitor them with stakeout teams. Any analysis of a city’s bars, from La Rioja’s whiskerías to El Alto’s roadside vendors, would certainly identify the minority of establishments that are the scene of a majority of violent incidents. Using follow-up measures such as alcohol license suspension or assignment of an extra officer, the police could assess the costs and benefits of their responses in terms of injuries, detentions, and revenue losses. Because roughly three-quarters of all police contacts with citizens occur in private places, such as in the home in cases of domestic violence, CPTED studies of residential space and social relations—such as how well neighbors communicate—may also help inform and improve crime strategies.

How new laws translate into practice is another dimension of evaluation. In Cali, Colombia, between 1992 and 1994, Mayor Rodrigo Guerrero used an information-based evaluation of specific areas to develop multisectoral policies. One initial study showed that laws restricting alcohol reduced homicides by 25 percent, and that stronger disarmament laws could cut homicides by an additional 14 percent. Likewise, in Colombia’s capital of Bogotá, restrictions on alcohol sales reduced killings by 8 percent (Guerrero 2007, 133). Such results should prompt the police to devote more resources to these problems, particularly when they are also highlighted by citizen surveys and community policing meetings.

Policing of Youth

In the policing of youth, risk assessment can help strike a balance between repression and the prevention that best responds to ever-changing activities and locations through solutions such as concentrating on repeat offenders and finding alternative spaces where otherwise legal activity does not bother the community.7 The well-known “pulling levers” strategy, in which criminal groups such as gangs are told they will all be subject to police action if even one member commits a transgression, is successful in part because of the police’s detailed knowledge of gang memberships, locations, and activities (Kennedy 1998). Such knowledge can also help categorize members according to their level of potential criminality. Using interviews to document risk factors, for example, the Minors Section of Costa Rica’s Organización de Investigación Judicial found that about 12 percent of young detainees were using drugs—far lower than previously assumed—and that nearly one-third had no chronic behavior problems.8 As Chicago and New York have learned, observations and diagnosis by individual officers monitoring targeted gang members are more successful in reducing gang activity than extended detentions, which affect only a percentage of gang members.

Using the “restorative justice” approach, which has been gaining ground in Latin America (Carlucci 2004), international organizations have funded many youth offender–related programs, such as those for postdetention services.9 A good example is Uruguay’s National Rehabilitation Center (Centro de Rehabilitación Nacional), which houses about 140 youth offenders and provides them with a wide range of social and educational services by a full-time staff of psychologists and other professionals. The commitment of the directors, such as eating meals with the youth, contrasts with that found in most other facilities in the region.10 Because of the large amounts of time, money, and effort that go into programs like these, serving relatively few people, they are impractical as large-scale solutions. But they help to identify the causes of violence among youth at the early stages, when those causes are easier and more important to disentangle, thereby providing support and reducing the burden on law enforcement agencies. Unconnected to specific crimes or policies, this more dispassionate diagnosis can also lower the decibel level of Latin America’s ongoing debate over youth crime. As discussed in chapter 3, a focus on questions such as the age of legal responsibility distracts from the larger issues such as socialization and drug availability.

Citizen Involvement in Evaluations

Citizen involvement, which boosts objectivity and prevention, should also play a greater role in evaluation of the criminal justice system. One form of involvement is violence observatories such as those in Honduras and Ecuador, which can become their country’s preeminent sources of complete crime statistics, in part by attracting public attention and collaboration through dissemination of the times and places in which violent acts occur. Such entities are more likely to elicit involvement by those who shun more public venues such as community policing forums. Violence observatories can also support police reform. For example, La Rioja’s police were unable to answer even a fraction of the calls to a new telephone line they set up, but they could have used the communications expertise that a violence observatory brings. When citizen-run observatories are not sustainable, more institutionally acceptable may be consortiums of criminal justice agencies that report crime numbers, such as the one in El Salvador described in chapter 2. In addition to sorting out statistical discrepancies, El Salvador’s Mesa Técnica de Homicidios also encourages wider coordination and oversight with interested NGOs.

The combined importance of discretion and information is evident in the ways police determine the relevance of the information they do get. Leads on serious crimes or patterns of crime in general are often overlooked because officers ignore sources that appear irrelevant. In most countries up to 80 percent of citizen contacts with the police are not about crime, say officers; they are mostly grievances about a wide range of problems, most of which are not the police’s direct responsibility. But when the police tune out most of this noncriminal information, they miss the portion of it that could be valuable. In New York City most citizen reports were long dismissed as “kites”—that is, they were considered flimsy and expected to (hopefully) just float away. But in areas such as Washington Heights, officers began to use them to gather evidence about persons and places, enabling them to wrap up drawn-out drug operations with arrests and convictions that effectively ended neighborhood trafficking (see Hartnett and Andrews 1999). Sustainable improvements in criminal justice, in short, need objective collection and evaluation of information.

Prison Conditions

Information may also be the best way to begin dealing with the appalling state of Latin America’s prisons, described in chapter 2. Beginning in the 1990s, international criticism and the consequences of penitentiary conditions, such as prison-based criminal networks, sparked efforts to reduce incarcerated populations. Bail and conditional release laws have often been accompanied by the release or suspension of charges for those who have not been tried, who were arrested for minor crimes, or who were first-time offenders.11 In 1999 Costa Rica began to offer community service as an option for those convicted of crimes punishable by less than three years in prison. The change relieved the system of four thousand inmates, cut crime (the program’s recidivism rate is well under one-fifth of those who serve time), and saved money (prison costs $50 a day per person, whereas the alternative program costs $0.50). Longer-term progress, however, will require addressing systemic practices. Countries have focused so much on violence prevention and daily maintenance that prison officials in nearly every country have lost track of who is being held and for how long. Prisons, then, should be given the computer programs they need to document the number of incarcerated persons, the charges and status of their cases (particularly for drug possession and consumption), the level of legal access, and the length of incarceration (the courts received similar software from international agencies to track cases). This information will help reduce overcrowding, process detainees, meet international laws on detainee separation, and identify those who could benefit most from rehabilitation. It could also document sanitary conditions, access to mental and physical health care, availability of detox services, the quality and quantity of food, and the existence and functioning of work, self-help, personal development, skill-building, and other education and rehabilitation programs. This information will also help prison administrations meet basic needs while reducing violence and recidivism, and it should enable officials to summon some of the will needed to at least address prison conditions incrementally without being exposed to political attacks triggered by reform overhauls.

Structured Citizen Engagement

Democratization literature refers to a “virtuous cycle” during which an expanding number and density of associations in which citizens have overlapping memberships fortify the state-society relations that sustain democracy. On the issue of security, such cycles can be set in motion by citizens identifying and responding to the sources of crime. As councils at the local, regional, and national levels interact and form a societal network, participants begin to cast their visions beyond their own neighborhoods. In his categorization of incentives for participation, Olson (1965) found that selective incentives reflecting individual interests then begin to be replaced by collective incentives focusing on the community good. In Caracas, for example, innovative neighborhood projects and political debates have been pollinated through grassroots publications. The newspaper La Chispa (The Spark) contains articles written by activists who are developing social programs in the city’s geographically disparate barrios, from Petare in the far eastern end of the district to Nuevo Horizonte (New Horizon) in the far west. A local citizen security council in Buenos Aires, responding to residents’ complaints about the insecurity posed by indigent men collecting discarded cardboard and other recyclables, found a solution by leaving cardboard out for collectors but restricting the times they can collect it. This response was taken up by other areas as well, helping to reduce tensions around the city.

When it shows results, such civic activism can be promoted among police officers as an untapped possibility rather than an additional uncertainty in their work. It may then boost legal changes, accountability processes, and the other reforms examined in this book. As many cities have experienced, citizen participation remains low when police step up engagement only during crises. But if residents can initiate actions rather than just receive help, they will usually provide more useful information to the police, which has been the experience in cities like Bogotá, Santiago, Philadelphia, and Baltimore. The police then become a catalyst of the virtuous cycle of participation. Scholars in fact describe how governance is facilitated by relations among state and civic actors that are strong but flexible enough to encourage innovation (Migdal 1989; Putnam, Leonardi, and Nanetti 1994; Hagopian 1994; Grindle 2004, 196–205). Tendler (1997) points out the importance of strong relations among different levels of government institutions and between those institutions and civil society. Meanwhile, many authors stress the “complementarity” of mutually supportive public-private relations heavy on personal ties and light on top–down directives (Evans 1997; Hildebrand and Grindle 1997). In her work on policing, Ostrom (1975, 1997; Ostrom and Whitaker 1973) shows how public service delivery succeeds when governments engage with social groups to help determine a project’s direction but are careful not to destroy local organizing with excessive directives.

But how can such a relationship be built to deal with citizen security, and how can progress toward it be measured? As much as possible, community policing programs must incorporate concepts of citizen accessibility and policy flexibility. In particular, neighborhood forums should be structured to minimize paralysis, politicization, and replication of existing power dynamics. Domination by established groups, more highly educated participants, or mano dura police-resident alliances is always likely, and will reinforce rather than change the kinds of community relations that stifle collectively constructive responses. In addition, because of its loose definitions and high expectations, community policing must incorporate specific objectives and mechanisms that measure progress toward them on an ongoing basis rather than after practices have set in and shut down the opportunities for new approaches. Indeed, many evaluation mechanisms do not evaluate what they should, such as by measuring “success” as the number of meetings or citizen demands logged rather than specific results.

Participatory budgeting, judges of the peace, and community justice forums are good examples of how the power to decide on central community concerns such as public projects and group conflicts attracts a wide range of residents, who acquire skills in debating, mobilizing, and governing through their participation.12 Progress on such issues, which frame the start and endpoints of community policing, can be made through structures that are centered on criteria that lend substance to citizen participation while eliciting useful information. Table 7.1, developed for the Chicago police, shows how this can be done (Chicago Community Policing Evaluation Consortium 2002). In this model, at each community policing meeting, attending police officers, community leaders, and outside observers all note their general ratings in four related categories. The first is problem identification, which is the ability of the meeting to specify the problem under discussion. This area is usually the strongest, because the problem is usually the reason for the meeting. Less certain are the plans to respond, because planning requires compromise and commitment by those attending. If the problem is not clearly identified, the specificity of plans will almost always be at the lower end of the scale. But police support for plans, the third category, might salvage meetings from such uncertainty by giving official backing to certain responses. The most wide-open category, contribution to criminal policy, is usually rated after a policy’s implementation has begun, as the impacts come into focus. After synchronizing their observations and giving plans a specified time for implementation, observers then meet to connect weak or negative results back to the original deliberations and restructure or reframe follow-up ones accordingly.

Table 7.1 Process Framework for Community Policing

table 7.1. For example, in the high-crime East New York section of Brooklyn, neighborhood forums have helped bring out and defuse smoldering sources of tension such as the after-school police stops of students and the absence of patrols in high-rise buildings. Identifying points of tension also helps focus participation and integrate evaluation by giving residents specific targets for improvement. And a sharper focus encourages participation by nonpolice officials. For example, Uruguay’s Citizen Security Program has resource and logistics centers that are actively supported by city mayors and councils. In 1990 Uruguay’s capital, Montevideo, began a decentralization process to encourage citizen participation. The city was divided into eighteen administrative zones, each with neighborhood councils to help bring in services, oversee resource allocation, and allow residents to become links to government officials (Canel 2001). Although both the general and community policing councils encountered resource limits and often unreasonable expectations, state support helped overcome those obstacles, in part by enabling discussion of policy and conditions in general, such as the effects of poverty and inequality on crime.13 In 1995 Bogotá, Colombia, began reducing violence through problem-oriented citizen security schools and regional “fronts.”

In many countries in Latin America such programs are shored up by neighborhood Fiscalía offices as well as by legal aid NGOs (see Brinks 2008). To expand such outside support, the police could create a database of organizations willing to provide resources, or willing to offer the expertise needed to establish programs such as those that alert residents and police to violence-inducing tensions. Umbrella consortiums that bring together businesses in a specific sector are particularly well suited to draft support from its members. An example is the insurance association in Argentina, Asociación Argentina de Compañías de Seguros, which has shown support for police reform. In addition, in that country’s Capital Federal the many local councils have effectively pressured the Policía Federal Argentina to integrate the city’s popular community policing efforts into its daily operations, such as through joint patrols and coordinator office hours in police stations.14

Because demands and expectations are part of such efforts, police structures must be both reliable and reliably flexible to respond to them. In particular, police must be ready to deal with tensions among neighborhoods that can flare up in a community policing project. For example, at a meeting in the middle-class San José neighborhood of Zapote in Costa Rica, residents said that their area was “calm for decades” until the intrusion of drug users from adjacent neighborhoods.15 Some residents then tried to make their citizen patrols more aggressive, which led to their discontinuation for being abusive. Officers’ abilities to handle such volatility are rooted in large part in their own acceptance of the community policing idea. In the 1960s the United States took the first steps away from the professional model in order to address urban unrest and growing officer disenchantment. The goals were better coordination of preventive and investigative policing and a reduction in overspecialization. In the 1970s and 1980s traditional policing was then often replaced by “team policing,” along with geographic reorganization and other approaches later associated with community policing. But the process of internal acceptance and appreciation of community policing was neglected. As a result, many officers at all ranks considered community relations tangential to crime fighting, and most U.S. cities “tolerated” disciplinary review boards as a way to placate restless neighborhoods (Chevigny 2003, 59). In fact, authors such as Mastrofski (2006, 65) conclude that throughout the United States community policing “has not transformed the structure and operation of American policing so much as it has altered its rhetoric.” In many American cities officials involved in community policing continue to be dismissed by the rest of the agency as “social workers” (or, more damning, as “empty holster” agents). Meanwhile, community-centered public campaigns and slogans, such as the omnipresent “Block Watch” programs in the United States, remain more symbolic than substantive.

To avoid such dead ends, police must bend their agencies’ structures to fit the strategies adopted to fight each cause of crime. In other words, from youth truancy to narco-trafficking infiltration, the problem should be the basic unit of policing. To move toward that goal, as discussed in this book, police commissioners should have greater discretion in at least two areas. The first is spending; community policing is not traditional, and so the use of resources and directives do not have to be either. Second, centralized scheduling should be replaced, where possible, by teams responsible for a particular area (such as ten blocks) headed by a leader authorized to alter schedules, strategies, and resource use. For example, Houston’s Directed Area Responsibility Teams (DARTs) combine decentralized administration with “substations” headed by block captains (Brown 1987). Many officers rightly warn that such restructuring may open up more opportunities for corruption—one reason for automatic rotation—but the same kinds of informal controls that allow community policing to reduce crime also allow it to check corruption. These efforts also require state commitment. For example, the state may establish commissions made up of agencies with incentives to contribute to reaching clearly stated goals, such as helping reduce youth truancy or domestic violence by a certain percentage (see Sherman 1995). Through such commissions police could work more closely with education and health professionals to better understand and respond to the causes of crime and to undertake initiatives such as neighborhood cleanup campaigns in which residents and police gather in a high-crime street to remove trash, fix locks on homes, and make needed repairs in public spaces.

Costa Rica’s community policing program, one of the region’s first, is an example of how constructive police and citizen involvement can be when structured in these ways. Begun as the Unit of Community Extension in the police academy in 1994, the program has since expanded to thousands of barrios throughout the country.16 It is headed by the Office of Community Security (Dirección de Seguridad Comunitaria), and has about a hundred permanent staff. Program officers have more flexible schedules and different merits for promotion, and are rotated every four years, which gives them enough time to become fully acquainted with a community and its problems. The program tries to be flexible as well, says its director, in order to learn from early mistakes such as excluding local police in its original development.17 Meanwhile, citizen participation is channeled through the Basic Training Program in Citizen Security (Programa Básico de Capacitación en Seguridad Ciudadana), composed of eleven participatory modules that last about two months each. They cover subjects such as prevention, community organizing, criminology, drugs, and child abuse, as well as areas not traditionally included in community policing such as penal, family, and other areas of criminal law. At the weekly talks (charlas) in different areas, residents are expected to take the lead, mainly by identifying the causes of insecurity and specific responses to them. Designed to cultivate and identify citizen leaders, these charlas typically start out with about fifty residents, dwindling by the end to about fifteen, four of whom remain as permanent contacts heading up projects such as telephone networks.

In addition to bringing out the causes of insecurity and the points of tension, these charlas expose the limits of community policing in general. Residents are primarily concerned about recent incidents and requests for more protection by local police. Attending officers then spend much of the session time responding to and imparting information, which reinforces citizens’ traditional relationship with them. Even at a charla on the judiciary in the working class San José neighborhood of Tibás that was supposed to be about the penal process code, residents still focused on particular crime incidents in their neighborhood.18 Such retreat into traditional relationships not only leaves little time to develop new policies, but also can deprive officers of the opportunity to convey to residents the procedural limits to solving crimes in which prosecutable evidence is lacking. By diminishing unrealistic expectations, such candid forms of communication may also shore up citizens’ commitments to playing a more direct role in their own security.

Police Management: Standards without Standardization

The police could better meet the great demands for its help by paring down the tasks associated with police management. Rotations, paperwork, special assignments, and centralized controls, many commanding officers say, make it impossible to manage their corps. For example, most police commanders in Caracas are given daily the lists of officers under their command, but often they do not have time to find out why many of them do not show up for work. It is unrealistic to expect such problems to be fixed overnight, as this and other books on police management show, but other tangible and politically acceptable alterations in administration and resources can go a long way in freeing up commanders to focus on and achieve results. As described in this section, a strengthening of human resources administration, officer living standards, inspection, accountability, and discipline can all help fortify police management.

First, human resources must be approached as holistically as political relations will allow. At the broadest level, national and regional governments should hash out a distribution of human and material resources based on population and crime rates. Such a plan should include incentives for recruiting in underrepresented regions and assignments to high-crime areas. Police districts should also be reconfigured if necessary. Each one should have no more than 300,000 citizens, and all should be run by officials with more autonomy in implementing national training curriculum and standards.19 For example, arguing that power must “be realigned if the agency and its members are to achieve their full potential,” Henry (2002, 26) describes how the NYPD identified management as a key point for change and strengthened middle-level managers by giving them more disciplinary authority. At the local level in Latin America, most commanders should aim to put a quarter of those officers doing administrative work on the street, replacing them with local civilians who have the pertinent education and background. This change would increase the presence and transparency of the police, thereby improving community relations, and it would strengthen efficacy.

Second, both management and oversight must better accommodate the daily realities of policing. Most officers resent handling complex tasks without the financial compensation or social recognition that other professionals such as lawyers enjoy for working under less dangerous conditions, and they resent the irregular work hours that are harmful to family life. Relatively easy ways in which to help officers while boosting morale include housing loans, transport subsidies, health benefits, and other forms of support contingent on clean records. As public employees, many police already enjoy such benefits, but more attention to the particular needs of officers—especially physical and mental health—is helpful. Internal affairs units can also boost officers’ daily experience. Those units are often regarded by officers as an instrument less of support than of blame, but they can help reverse negative internal practices and pressures (see McArdle 2001). At all levels, officers pay a heavy professional, social, and psychological price for their refusal to participate in illicit acts, and those who stop being corrupt rarely denounce larger abuses for fear of exposing their own or regard the cover-up of corruption as a requisite self-defense of the agency. To chip away at this thick layer of corruption, internal affairs offices should adopt more mediative approaches. For one thing, instead of being immediately pressed for information about illegal activity, officers should first be asked about any problems they may have, such as with family or finances, and how such problems could be resolved. Surveys of officers on their views of the seriousness of and appropriate discipline for specific infractions, from free meals to wide kickback schemes, will also help police officials understand the types of temptations officers face and their reactions to them in daily policing (see Klockars, Ivkovic, and Haberfeld 2004). Interviews of individual officers also will reveal the discretion and practices that underlie problem-oriented policing (Capowich and Andrews 1995), but interviews cannot be conducted if they are viewed as a prelude to punishment.

Officers’ responses in both individual interviews and surveys, along with reports on the use of force and logs of citizen complaints, can be used for an Early Intervention (EI) system for identifying “problem” officers with a high level of incidents and complaints (see Walker 2005). Because police bureaucracy has been identified by both officers and outside studies as a major cause of stress, this approach could also help identify the institutional sources of behavioral problems. To boost EI procedures, the government could have superiors conduct routine line inspections or have outside teams conduct personnel inspections, with each inspection focusing on a specific personnel mater (such as attrition) or organizational matter (such as under-utilization of radios). Supporting such efforts to identify unseen problems could be targeted and random “integrity tests,” in which an officer is given a chance to commit a corrupt act without realizing that he or she is being monitored (Newham 2003). Together, the problems and problem officers identified could be cited by an independent official, usually called an auditor, in investigating and proposing reforms of the structures or policies that might be contributing to such abuses.

When such EI or auditor investigations do expose wrongdoing and lead to disciplinary action, the consequences must be clear. Exactly what happens to officers after the first, second, and third incident of different kinds of abuse must be well publicized. Otherwise, officers will form their own interpretations of who is and is not disciplined and why, thereby diminishing the chances of cooperation. And in cases of mass purges, officers may form an armed body poised against reform in general. In such scenarios, accountability agencies will never reach the third stage of reversal described in chapter 3. Instead, they will just scratch out victories subsequently engulfed by institutional conflict, police resentment, or politics. Most of the reasonable and much-needed recommendations of New York City’s Mollen Commission, for example, were commandeered by interagency disputes and political grandstanding.20 Again, careful meditative approaches can help avoid such fates. One area in which such mediation can help is deciding who participates in disciplinary actions. There are many reasons to restrict that process to police: only they have the experience and perspective to understand their colleagues’ actions, and a ruling by a police official is more acceptable internally. For those reasons, through the 1980s nearly 80 percent of the disciplinary process in the United States included only the police. And most forms of citizen participation that were created were limited to review of either the initial complaint or the final decision.

New York City is a good example of the tensions that can arise among the procedures, actors, and agencies involved in discipline. The average number of complaints of police misconduct filed each year to the city’s Civilian Complaint Review Board (CCRB) rose from two hundred to two thousand during the 1990s,21 and it reached a record of over eight thousand in 2009.22 In both 2008 and 2009, most charges brought to the CCRB came from among the record number of people in New York—an estimated half-million—who were stopped and frisked each year. Despite this surge in complaints, the percentage of them that the NYPD refused to prosecute grew from 2 percent in 2005 to 33 percent by 2008. Although the conviction rate at departmental trials rose from 30 percent in 2004 to 60 percent in 2008, these patterns have led to a flare-up in tensions between citizens and the police. So even though the police’s overall authority over discipline should be acknowledged, the police should use complaints to identify specific points of tension and to address practices. Stopping and frisking is clearly one of those problematic practices, and if the police force does not address it—at the very least by involving citizens in an evaluation—they may lose control over it to officials who take citizen complaints in the political arena.


Those who develop security reform strategies must tap into the inherent powers and limitations of policing and criminal justice. Daily policing and institutional relations, in particular, exemplify both the positive and negative reach of criminal justice in high-crime societies. By starting from recognition of that impact, the five areas discussed in this chapter—points of tension, officer discretion, policy evaluation, structured citizen engagement, and internal management—attempt to utilize the potential of policing to minimize its harm.

1. Clyde Haberman, “Crime Down, but Courts Are Clogged,” New York Times, January 3, 1997.

2. Pedro Luis Tang Urdaneta, Policía de San Francisco, author interview, Puerto la Cruz, November 29, 2005.

3. There was a jump from nine citizen deaths for each police death in 2000 to twenty-eight per officer in 2001, and the 170 extrajudicial killings recorded in 2000 were the most since 1988.

4. The U.S. Supreme Court and those in most Latin American countries limit the use of force to situations in which an officer is threatened with injury or death.

5. In some cities less than 20 percent of police officers are responsible for more than half of all the detentions leading to convictions.

6. Al Baker, “A Hail of Bullets, a Heap of Uncertainty,” New York Times, December 9, 2007, 4.

7. Responding to the fact that a small number of criminals commit a disproportionate amount of crime, the Repeat Offender Project in Washington, D.C., identified two kinds of repeat offenders: those already wanted on one or more warrants and so could be arrested on sight and those believed to be criminally active but not currently wanted.

8. “Inquieta violencia en Menores,” La Nación, March 31, 2000, 6A.

9. For example, a 2004 Inter-American Development Bank citizen security loan of $7.2 million to Nicaragua aims to reduce violence and juvenile delinquency in communities. Similar programs have been developed for Brazil, Peru, and Honduras.

10. Residents and Gabriel Courtoisie, head of rehabilitation, Centro de Rehabilitación Nacional, interviews with author, Montevideo, August 18–19, 2003.

11. Colombia frees inmates who have served 60 percent of their terms unless charged with serious crimes such as drug trafficking. Ecuador’s constitution allows judges to release unconvicted prisoners who have been in jail for over a year. In Bolivia a 1996 law allows the release of prisoners if no charges are brought within eighteen months. Chilean judges may suspend charges for first-time offenders, but try them for two crimes after a repeat offense.

12. As Baiocchi (2003) reports in a study of Porto Alegre, public budgeting led to an increase in public investment from 2 percent in 1989 to 20 percent in 1994, almost tripling the number of schools.

13. Author interviews and observations, national meeting, Programa de Seguridad Ciudadana, Montevideo, August 25–26, 2003.

14. Observation of the community policing program of the city of Morón, Buenos Aires Province, May 27, 2004; Claudio Suárez, head, security program of the government of the city of Buenos Aires, interview by author, May 26, 2004; author meeting with four coordinators of the city of Buenos Aires security program, May 19, 2004.

15. Author interviews and observations, Zapote neighborhood meeting, San José, June 20–21, 2006.

16. Ana Durán Salvatierra, vice minister of governance and police, interview with author, San José, Costa Rica, June 19, 2006.

17. Alberto Li Chan, quartermaster and coordinator, Seguridad Comunitaria, Ministerio Seguridad Pública de Costa Rica, interview with author, San José, Costa Rica, June 20–21, 2006.

18. Law charla in Tibás, San José, Costa Rica, June 22, 2006. About twenty residents attended.

19. In the United States some local law enforcement officials, such as those in Detroit and Portland, have challenged and rejected federal operations on the basis of local ordinances, policies, and programs. See Printz v. United States, 521 U.S. 898 (1997).

20. The Mollen Commission, appointed by Mayor David Dinkins in 1992, exposed networks of police officers dealing drugs, robbing residents, and taking cash from crime scenes. Commenting on the situation, Amnesty International (1996) said that “it is rare for NYPD officers to be criminally prosecuted for on-duty excessive force and rare for convictions to be obtained.” The commission’s recommendations included an independent Police Commission with investigatory and subpoena powers (Selwyn Raab, “New York’s Police Allows Corruption, Mollen Panel Says,” New York Times, December 29, 1993). But in 1994 Mayor Rudolph Giuliani vetoed the City Council’s creation of such a body on the grounds that these two powers belonged exclusively to the mayor. After a 1994 hotel rampage by police, Giuliani adopted many of the Mollen Commission’s proposals, but shifted investigation from the Internal Affairs Bureau to precinct commanders (William Rachbaum, “Getting Back in Line: Rudy Unveils Plan to Stem Cop Corruption,” Newsday. June 15, 1995).

21. “Complaints against the Police Rise,” New York Times, June 11, 1998, A25.

22. Christine Hauser, “Few Results for Reports of Police Misconduct,” New York Times, October 4, 2009.

Previous Chapter

6. Argentina

Next Chapter

8. Conclusion

Additional Information

Related ISBN
MARC Record
Launched on MUSE
Open Access
Creative Commons
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.