Skip to main content

Municipal Strategies of Crime Prevention

The Latin American Program held a conference to discuss municipal-level efforts to improve citizen security, reduce violent crime, and engage citizens and invigorate civil society participation.

Date & Time

Dec. 10, 2009
7:45am – 2:30pm

Municipal Strategies of Crime Prevention

The Latin American Program held the conference, "Municipal Strategies of Crime Prevention," on December 10, 2009, to discuss municipal-level efforts to improve citizen security, reduce violent crime, and engage citizens and invigorate civil society participation. Speakers included: Claudio Beato, CRISP Minas Gerais (Brazil); Rodrigo Guerrero, Cali, Colombia; Liza Zúñiga, FLACSO-Chile; Carlos Basombrío, Latin American Program Consultant (Peru); Juan Salgado, Centro de Investigación y Docencia Económicas (CIDE) (Mexico); Robinsson Caicedo, Chamber of Commerce, Bogotá, Colombia; Abby Córdova, LAPOP, Vanderbilt University; Silvia Vásquez, Guatemala; Renato Sergio de Lima, Forum Brasileño de Seguridad, São Paulo (Brazil); Carlos Romero, Ciudad Nuestra (Peru); and Ana María Sanjuán, Universidad Central de Venezuela.

Rodrigo Guerrero, former mayor of Cali, Colombia, discussed public health approaches to curbing violent crime. Using a scientific approach, he said, means clearly defining the problem, identifying risk factors, and planning interventions appropriately. It also means systematically evaluating the success of interventions and then reformulating interventions based on new evidence.

Guerrero said it is important to understand the contexts and causes of violent crime, and he pointed to symptoms that determine the prevalence of violence in urban spaces. These contextual factors include the presence of organized crime, inequitable justice systems or prevailing impunity, media images over-saturated with images of violence, and the widespread availability and visibility of firearms.

Guerrero said that in designing successful policy interventions, it is important to define explicitly what violence is and to identify the determinants that principally affect levels of violence in a community. Employing a public health strategy means tracking violence as if it were a disease. Within Cali, for example, there exist concentrated pockets of violence, prompting the need for geographically pinpointed interventions. Guerrero provided data showing trends in violence, such as the days of the week that violence is most prevalent and the concentration of violence in particular neighborhoods.

Guerrero identified two general policy interventions that are based on public health analyses of violent crime. In the first category are those policies that impose prohibitions on certain kinds of behavior, such as temporary temperance laws on the purchase of alcohol or the carrying of firearms. In the second category are policies focused on providing alternatives for sub-populations vulnerable to criminal behavior, such as job and educational opportunities for at-risk youth.

In Cali, such interventions were successful in reducing the number of homicides. Combined, prohibitions on firearms and alcohol led to a 49 percent drop in homicides, he said. Guerrero also pointed to a significant reduction in vehicular homicide in Cali. He showed data indicating that obligatory helmet laws and other legislation lowered the death rate of motorcyclists in Cali between 1993 and 2003.

Other important policy interventions in Cali and Bogotá included the institutionalization of peace and coexistence programs; the strengthening of the police and courts; the control of risk factors, through prohibitions on the sale of alcohol and the carrying of firearms, for example; the recovery of public space; and improvements in education and health services.

Guerrero touched on Bogotá's "Sacred Life" program, the goal of which was to reduce violent deaths from a daily rate of 11 to 7, and reduce by 10 percent other serious crimes such as automobile theft and assault. He pointed to data showing a drop in the homicide rate for Bogotá and for the nation as a whole between 1991 and 2004, which had reached a high of more than 80 deaths per 100,000 inhabitants for both locations in 1993 and which dropped to current levels of about 40 and 20 deaths per 100,000, respectively.

The key ingredients for reducing violent crime, Guerrero said, were: 1) political will; 2) continuity; 3) multifaceted approaches; 4) a strengthening of enforcement measures; 5) the targeting of at-risk populations such as youth; and 6) addressing economic and other inequalities.

Carlos Romero of Ciudad Nuestra discussed the efficiency and fairness of responses to crime in Lima, Peru. He pointed to advances in public security and made policy recommendations for establishing a metropolitan-wide citizen security system. He credited civilian constabularies (serenazgos muncipales) created throughout Lima with stemming high rates crime. Nevertheless, he noted warning signs that positive trends were being reversed. These signs included a recent and swift rise in homicides, increasing violence by youth, and the wider availability of firearms.

Among policy recommendations, Romero cited the need to: improve the effectiveness of municipal governments in multiple spheres, create a metropolitan-wide citizens' council to monitor public security, and professionalize the civilian constabularies. He discussed the need for mechanisms of economic compensation for poor municipalities, as well as the need for metropolitan-wide initiatives of prevention, rehabilitation/reinsertion, and assistance to victims and families.

Romero gave an overview of public security in Lima, describing the city's political, socioeconomic, and demographic characteristics. He noted high overpopulation: the number of Lima's inhabitants is nearly five times what it was 50 years ago, with the city's population going from 1.85 million in 1961 to 8.5 million today. Politically and administratively, Lima consists of a metropolitan government and 42 district governments.

Lima's homicide rate is relatively low—about 9 per 100,000 in 2007—Romero said, though this compares with a much lower rate of around 3 per 100,000 annually over the period 1999-2004. However, Romero said that the crime victimization rate was high, citing data from a survey series from 2007, 2008, and 2009. These data showed that approximately 30 percent of Limeños in each of the three years had been victims of a crime in the past year. Most victims had suffered property crimes, and other data indicated that large majorities in the three years were victims of petty theft and/or burglaries. (Street muggings accounted for another large but separate category.) Assailants rarely were armed, and no more than 15 percent of victims reported being assaulted in an armed attack in the three years.

The perception that crime is rampant is very high generally, as well as across all socioeconomic and age bands, according to another survey series from 2008 and 2009. Three-quarters of respondents of this survey cited citizen insecurity and crime as the leading problem of Lima, and around 95 percent of respondents qualified the city as mostly unsafe or not at all safe in both years. Chief among causes of crime were gangs and drug consumption, as cited by respondents in both survey series. In his analysis, Romero noted that majorities and near-majorities of respondents reported adverse environmental factors in their neighborhoods, such as drunks in the streets, dirtiness, drug dealing, street fighting, vandalism, disturbances of the peace in proximity to bars and discotheques, and prostitution. Yet in spite of the perception that crime is rampant, insecurity has not exploded in Lima. This is because of local efforts to promote citizen security, such as the formation of the serenazgos municipales, the creation of neighborhood watch groups, the greater use of private security resources, and law enforcement control of organized crime.

Romero discussed local responses to crime, calling attention to the civilian constabularies in operation in municipalities throughout the Lima metropolitan area. The constabularies arose in the late 1980s in the wealthiest districts as a response to ineffective regular policing. They have now spread to 88 percent of all metropolitan-area districts, numbering a total of 7,627 "watchmen" in the metropolitan area. Constables are unarmed and lack the authority of the regular police. They patrol neighborhoods, provide rapid response to emergencies, and complement regular police. Constabularies make use of other resources, such as call centers, local watchdog and citizen monitoring groups, and security cameras. Large majorities in a University of Lima survey series over 2004-2009 qualified the work of the citizen constables as either average, good, or very good. The percentage of respondents calling the constables "average" rose more than 10 percentage points, going from 42.8 to 53.2 percent, while the percentage qualifying the constables as "bad" or "very bad" dropped 11 percentage points, from 34.7 to 23.6 over the same time period. Additionally, the same survey results showed that respondents cited the constables as "more efficient" than the regular police in the past two years of the series, 2008 and 2009, though the police were considered marginally more efficient in 2007. In spite of the success of the civilian constabularies, their relationship with the regular police in some cases has been conflictive, and Romero identified the need for greater cooperation between the two entities.

Romero discussed neighborhood watch groups as another local response to fighting crime. Groups consist of volunteers, many of whom are women, who work closely with the regular police and civilian constabularies, have a conspicuous presence in working-class and newly-settled areas, and offer support in times of emergency. Yet these groups may not be able to provide lasting continuity to efforts. Romero noted problems associated with the neighborhood watch groups, including their exploitation by political forces.

Romero touched on other problems affecting citizen security efforts, indicating that the law governing the National Citizen Security System represented an important architecture for inter-institutional collaboration, but it was weak in practice. The law provides for mechanisms of inter-institutional coordination for citizen security across districts, provinces, regions, and nationally, and comprises the regular police, prosecutors, the judiciary, and citizen groups. The emphasis of the system is on inter-institutional coordination and crime prevention. However, in practice, the system suffers from a series of problems. These include the lack of leadership from the central government, which fails to provide technical expertise or the budget necessary to achieve greater coordination.

Romero cited other local responses to crime, including preventive programs directed toward young people, special shelters for youth, and a public defender's office for children and adolescents. He discussed the recovery of public spaces and noted efforts to recover the historic downtown, a reorganization of open-air markets, and the recovery of parks and plazas, as well as an increase in the number of "green" spaces.

Liza Zúñiga of the Security and Citizenship Program, FLACSO-Chile, focused on Chile's efforts to translate national-level public security policy to the local level. FLACSO's Security and Citizenship Program, which conducts country-focused and comparative analysis of violence and citizen security in Latin America; develops training mechanisms in citizen security for public actors and academics, and advises public institutions on the design, monitoring, and systematization of policies, programs, plans, and initiatives regarding security. Zúñiga gave an overview of the components and goals of the program, which is oriented around efforts of prevention (interventions at schools and for at-risk youth, redesigning urban spaces to prevent violence, and preventing child abuse); rehabilitation (of abusive men and the societal reinsertion of ex-convicts after serving prison sentences); and, assistance to victims and families (including assistance to the victims of violent crime, prevention and treatment for female victims of violence, restitution for injuries and trauma caused by child abuse.)

Zúñiga discussed different funding mechanisms in Chile for municipal efforts to fight and prevent crime. She noted that a recent trend has been toward the centralization of community policing programming from the central government downward, something that discourages decentralized or bottom-up programming. Although municipalities may be able to tap into more funding than they did in the past, it comes with bureaucratic and accounting restrictions, and communities have become passive and less willing to implement community policing programming.

Abby Córdova, Latin American Public Opinion Project (LAPOP), Vanderbilt University, discussed results from the most recent edition of the long-running Latin American public opinion project, Barometer of the Americas, as well as from a recent USAID-funded evaluation of violence-prevention programs in the region. "In what ways does insecurity (perceptions of crime) affect democratic governance at the local level?" and "What indicators exist that point to the success of local violence-prevention programs?" Córdova pointed to data from the 2008 Barómetro de las Américas to show that perceptions of crime go hand-in-hand with perceptions of municipal governance. Survey results from El Salvador suggest a relationship between high indices of victimization and insecurity with low levels of trust in municipal leadership and low levels of satisfaction with municipal services. Perceptions of crime also link to perceptions of trust among persons generally, with victims of crime less willing to trust their fellow citizens, the survey results suggested.

Córdova matched symptoms of citizen insecurity with governance deficiencies and corresponding public policy interventions. For example, crime victimization redounds negatively in social cohesion, as the data correlating victimization to lack of neighborly trust showed. High levels of violent crime signal a need for heightened community participation in crime-prevention efforts. Perceptions of insecurity in the community may be lessened by routine community meetings among neighbors. Social disorder, manifest in graffiti, gangs, and petty drug dealing, demonstrate a lack of trust in the institutions of the state, namely the police and local government.

To explore how evaluate local violence-prevention programs, Córdova pointed to survey data that asked respondents to evaluate how safe they felt in their neighborhoods. The data pointed to a positive relationship between perceptions of safety and community mobilization efforts. For example, Costa Rica rated second among six countries in the percentage of population involved in anti-crime community mobilization efforts (18.7 percent), while that country scored last in percentage of respondents calling their neighborhoods "unsafe." At the other end of the spectrum, El Salvador scored highest in victimization (19 percent) and highest in perceptions of unsafe neighborhoods (39 percent). In all the countries—Panama, Costa Rica, El Salvador, Honduras, Guatemala, and Nicaragua—city residents reported being crime victims at rates higher than the national average, and perceptions of crime were higher in metropolitan areas than in non-metropolitan areas.

Córdova turned to the question of how to evaluate local violence-prevention programs in the region before, during, and after interventions. LAPOP's project involves three Central American countries and 100 neighborhoods. It seeks to identify the ways in which violence-prevention programs are functioning and where and for whom such programs are operating. Additionally, it seeks to explain whether such programs are functioning as designed, as well as to suggest adjustments in program design. She described current USAID crime- and violence-prevention programs in Latin America, which she said aimed to reduce levels of criminality and to improve citizen security in Central America, mainly through capacity-building at the community level, and the creation of educational and job opportunities for at-risk youth. LAPOP's most recent edition of Barometer of the Americas (Barómetro de las Américas) was published in 2008. It covers 24 countries and consists of more than 40,519 interviews. See

Silvia Vásquez discussed efforts over a decade to improve security in the Guatemalan municipality of Villa Nueva. She explored the intersection of social and security policy in a town whose principal crime problems are gangs, domestic violence and child abuse, and truancy. Her presentation focused on the following municipal strategies to combat and prevent crime: 1) inter-governmental collaboration among the military and municipal and national police forces; 2) community-based prevention and treatment efforts; 3) urban planning as a preventive strategy; and 4) community-based organizations and citizens groups.

Vásquez discussed collaboration among the national civil police, the military, and the municipal police; the deployment of joint patrols; and the presence of military units in areas known for gang activity. Coordination with the military and national police enables situational prevention in areas prone to violent crime without the municipal authority overstepping legal boundaries. Vásquez also spoke of treatment centers for victims of domestic violence, which she said had succeeded in generating data on a crime that is rarely reported. And she mentioned the development of recreational and sports facilities to serve at-risk youth, though it is difficult to quantify the effectiveness of such facilities given a lack of measures and benchmarks. The installment of street lamps represents another preventive measure that, along with new recreational centers, enables the recovery of public spaces. In mentioning community-based organizations and citizens councils, Vásquez also called attention to their drawbacks, including the serious problems of lynching, illegal detention, other restrictions on free movement, and vigilantism (by neighborhood patrols with facemasks and firearms).

Vásquez closed by noting that none of the programs mentioned took into account at their inception the need for benchmarks to measure the effectiveness of their results.

Claudio Beato of CRISP, Minas Gerais, noted that since the 1970s, has experienced a steep rise in the rate of homicides, mainly attributable to the growth of youth gangs in large cities. Brazil's economic growth has ironically been accompanied by an increase in violence, noted Beato, yet he also noted a statistical relationship between the relative affluence of a neighborhood in Brazil and its level of violence. Beato highlighted the importance of good data collection and an appreciation for the local situation in developing a comprehensive understanding of crime, explaining that specific types of crime tend to occur at certain locations and times.

Rio de Janeiro has a specific set of challenges, Beato said, including a high level of organized crime and poor citizen-police relations. Due to the difficult conditions, police in the favelas are "prepared for war," said Beato, something that contributes to a negative image of the police within the community. Beato mentioned São Paulo, one of Brazil's success stories, where homicide rates have been reduced to the levels of thirty years ago.

In discussing potential solutions, Beato focused particularly on the program Fica Vivo (literally, "Stay Alert"), a program aimed at reducing gang-related violence among youths in Minas Gerais. Often, Beato said, there is a divide between those who prefer to focus on the "weeds" versus the "seeds." Police, he said, favor a "weeds" approach and look to gain control over the city. Non-governmental organizations and community groups tend to prefer to focus on social development and the factors that give rise to crime. He argued that the success of Fica Vivo comes from its multifaceted approach incorporating elements from both strategies.

Researcher Juan Salgado of of the Center for Economic Research and Teaching (CIDE) discussed police reform efforts in Mexico City. His presentation examined a reform program of the city's municipal police over 2002-2008, with research separately funded by the Mexican non-governmental organization FUNDAR and the Inter-American Development

Salgado began by discussing the principal challenge facing police professionalization efforts, which he considered to be the militarization of policing in Mexico. By militarization, Salgado meant the use of the military in domestic policing activities and the insertion of military officers in municipal and state police forces. Militarization, he said, has identified the need for improvements along the following lines: 1) the appropriate use of force; 2) accountability of actions; 3) relations between the military and the public during policing operations; and, 4) civil-military relations at times of peace.

The current presidential administration's open-ended use of the military in a war against organized crime has diverted attention away from municipal-level police reform, he said. Yet this kind of police reform is essential, since the municipal police represent Mexico's largest law enforcement force and represent a principal contact between the state and the public. Another major challenge is infiltration and cooptation of the municipal police by organized crime, he added. Mexico City is worthy of focus since it has the country's largest police force, at around 80,000, meaning that approximately 1 out of every 5 municipal police officers nationwide is based there.

Salgado gave an overview of police reform efforts in Mexico at the federal level. The three branches of government in recent years all have promoted reform efforts, some with greater energy and effectiveness than others. Congress has been the most proactive branch, encouraging structural changes, while the executive has implemented some public policy changes. Reforms by the judicial branch were considered the weakest, given that Mexico lacks an activist bench like the U.S. Warren Court (1953-1969), which made decisive and precedent-setting rulings in favor of defendants and against police and prosecutorial misconduct, he said.

Salgado identified five main goals for police reform programs:

• Improvement of recruiting policies
• Improvement of training, including human rights and community policing
• Improvement of accountability mechanisms
• Improvement of on-the-street supervision of patrolmen
• Improvement of internal control measures

Mexico City Police Reform
Salgado discussed a municipal-level police reform program in Mexico City over 2002-2008. The program aimed to "redignify" the much-stigmatized Mexico City preventive police corps by improving working conditions and fighting corruption. The program relied on consultation from Giuliani Associates, the consultancy of the former New York City mayor, and made use of Compstat, a computer modeling system used to map crime. Remarkably, the program has survived three municipal administrations, he added.

Salgado elaborated on the program, which was implemented in about half of Mexico City's precincts. It sought to improve working conditions of officers by shortening their workday to eight hours when previously shifts lasted 12 or 24 continuous hours (followed by like breaks). The program also attempted to put more officers on the streets by lessening their administrative tasks. "This is important given the high level of bureaucracy and the amount of time that this takes from operations," Salgado said. The reform also created a set of criteria to evaluate officers, when previously evaluations were based largely on the discretion of superiors.

The program, however, has shortcomings. It does not enable sanctions against individual officers from being made known publicly and the objective evaluation system has not been fully implemented. Also community policing appears to be downplayed. Salgado said that weekly meetings between police officers and community members took place in only 4 of the 15 units that he directly observed. "In a way, the community relations meetings were rhetorical," he said.

More generally, the program's principal drawback was the incompatibility between a progressive reform project and repressive police enforcement policy. Salgado cited Giuliani Associates' preference for what he considered questionable enforcement tactics, such as "stop-and-frisk." Additionally, the program failed to incentivize police officers to uphold human rights. For example, the set of evaluation criteria omitted inclusion of a human rights score, or of an individual officer's community policing skills. Rather, it concentrated on assessing an officer's swiftness of response to a crime scene and whether he was properly wearing his uniform.

Besides these structural shortcomings, Salgado identified a topical event that occurred during the reform program that questioned its effectiveness. In June 2008, 12 young people died after a police operation gone wrong at a dance club in Mexico City. The young people died after a stampede ensued following a surprise raid by police, who were acting on a tip of illegal alcohol and drug sales. Salgado noted that some of the officers involved in the raid were participants in the reform program. The lesson, he said, spoke to the limits of progressive reform efforts, since a top-down command structure still exists in the corps. Officers told Salgado that they must follow orders, regardless of how absurd they may sound. Another lesson is that police reform is necessarily self-limiting in the absence of reforms to the justice system and of prosecutors, he said. An inherent tension exists between progressive reform and repressive enforcement, he added.

Challenges to Democratic Police Reform in Mexico
Salgado identified the following as challenges to democratic police reform efforts in Mexico:
• Generally, municipal-level reform efforts have short lifespans in Mexico, given the hard term limits on mayors who are restricted to non-consecutive three-year terms.
• The public's demand for hard line enforcement can conflict with the community policing and progressive goals of reform efforts
• The militarization of policing activities in Mexico detract from municipal-level reform efforts and from community policing efforts
• The lack of emphasis on preventive measures both within police departments and outside police departments

According to Robinsson Caicedo of the Chamber of Commerce of Bogotá discussed chamber initiatives to fight and prevent crime and evaluated the public security policies of the past few mayoral administrations.

He began by giving an overview of Bogotá, the capital of Colombia, the capital of Cundinamarca department, and home to a large share of the country's total population and economic production—about 20 and 60 percent, respectively. Bogotá's prominence and its ability to collect revenue as Cundinamarca's capital have enabled it to make notable investments in the area of public security, he said.

Caicedo diagnosed Bogotá's crime problems, which, he said, reflected national trends. He cited four problems:
• Small-scale drug dealing is a major crime challenge, he said, as dealers have assumed control of public spaces and local consumption is growing. "If Colombia in the past was, sadly, an exporter of drugs, it is now becoming a consumer of drugs," he said.
• Internal refugees represent another challenge that sharpens economic inequalities in the city. Approximately 600 displaced persons arrive to Bogotá in a given day, escaping armed conflict in rural Colombia. Many of these refugees wind up living in so-called goteras, slums where living conditions are "infrahuman" and children, especially, are among the victims, he said.
• Armed groups have become urbanized. The decrease of armed conflict in rural areas has led some in Colombia to believe the country is living in a "post-conflict" era. But this ignores the fact that armed groups, pushed out of the countryside, are making their way into the city and using similar tactics of violence. Armed groups are "urbanizing," he said, staking out territory and taking over illicit, as well as licit, markets.
• Urban disorder. Caicedo cited the problem of "urban disorder," a continuing challenge for Bogotá but less so since improvements during the mayoralty of Enrique Peñalosa Londoño (1998-2001). Urban disorder can lead to the creation of "hotspots" of violence.

Caicedo discussed the methodology the chamber uses to assess crime in Bogotá. The business council relies on victimization surveys to evaluate levels of personal and property crime. This is because only about 40 percent of all crimes are reported to authorities, because of laziness, distrust of authorities, and the fear of reprisals, he said. The surveys divide victims into two categories: direct victims and indirect victims, i.e. those with a household family member who was a victim. The surveys have been conducted twice a year for the past seven years every January and June. The chamber seeks to standardize statistics and typically chooses the "worse" statistic. For example, in compiling data, the chamber favors public health over police statistics for certain categories. If an assault victim dies a few days after his injury, the crime is recorded in police files as merely an assault; but public health statistics would list it as a homicide.

Caicedo gave results from the most recent victimization survey. A large percentage of Bogotá citizens—39 percent—were victims; 18 percent direct victims and 21 percent indirect victims. The majority of victims are young people under 34 years of age mainly because they are the most exposed segment of the population, the most likely to make use of public space, and the ones who are most economically active (working, studying, consuming). Caicedo said Bogotá's homicide rate was 19 per 100,000, which is average for Latin America and one of the lowest rates in Colombia. Cali, for example, and an area of the country's coffee-growing region have much higher rates at near 80 per 100,000. Where drug trafficking is prominent, homicide rates are higher, he indicated.

Caicedo discussed perceptions of crime in Bogotá, which he qualified as "high." He said that 55 percent of Bogotá residents asserted that crime had increased between the two most recent reporting periods, mainly because citizens lack faith in the policies of the current mayor. Paradoxically, those who feel most insecure, such as old people, are the least likely to be victims of crime, he noted. Older people fear crime more than younger people

Hosted By

Latin American Program

The Wilson Center’s prestigious Latin American Program provides non-partisan expertise to a broad community of decision makers in the United States and Latin America on critical policy issues facing the Hemisphere. The Program provides insightful and actionable research for policymakers, private sector leaders, journalists, and public intellectuals in the United States and Latin America. To bridge the gap between scholarship and policy action, it fosters new inquiry, sponsors high-level public and private meetings among multiple stakeholders, and explores policy options to improve outcomes for citizens throughout the Americas. Drawing on the Wilson Center’s strength as the nation’s key non-partisan forum, the Program serves as a trusted source of analysis and a vital point of contact between the worlds of scholarship and action.  Read more


Documents & Downloads

Event Feedback