publisher colophon

21Chapter 2

Realms of Change and Obstacles to Citizen Security Reform

Obstacles to citizen security reform appear more quickly than its results. Even the most carefully constructed and broadly supported changes are diluted, undermined, or neglected. Such impediments to change are so potent because they are part of most countries’ politics, institutions, laws, and societies.

Indeed, the impediments to citizen security reform spring from the same processes that create it. Politically, governments are often so pressured by panicked societies to deliver short-term results that they discard ministers and policies at a rate often directly proportionate to their inabilities to face down the entrenched agencies with the most to lose from reform but with the greatest ability to block it. These agencies impede change not only through direct resistance but also through their own deficiencies in oversight, investigation, statistics, budgeting, and management. The law, a central counterweight to such trends, often only fuels those trends. Edicts and internal police regulations, unclear and inconsistent provisions, low-quality legal education, and crime policies unattached to criminology are some of the many legal distortions of law enforcement. Meanwhile, society has a hard time breaking through such conditions because of its own economic struggles, anemic organizations, tensions with state officials, and temptations to use destabilizing alternatives.

Different kinds of such obstacles afflict different countries, but they tend to emerge from conditions that are common in Latin America and other regions. Chapters 2 and 3 discuss two levels of such conditions. Chapter 2 examines the surface level of short-term reform, describing eight realms of change and the specific impediments that lie within them to specific proposals. Chapter 3 then delves into the roots of these impediments in the broader and more fixed traits of power, society, and geography in Latin American history and democracy. Together, these chapters show how changes in citizen security, many of them intertwined with deeper patterns of governance, have set the stage for a struggle between the status quo and attempts to alter it.

1. Opinion and Politics

On citizen security more than on most issues, the public and politicians push each other beyond what either intends. Since the mid-1990s, crime has been the first or second biggest concern almost everywhere in Latin America (IUDOP 1998; Consorcio Iberoamericano de Investigaciones de Mercados y Asesoramiento 2010). In some polls crime jumped from fifth place in 2003 to second place in 2005 as citizens’ biggest worry,1 and it is statistically tied with unemployment in the regional Latinobarómetro survey, more than doubling between 2001 and 2007. The survey asked, “What do you consider to be the most important problem in your country?”2 The public’s fear of crime often rises at an even faster clip than the crime rate itself—and sometimes continues to go up even when crime starts to go down. People simply doubt that any decline will last or attribute it to external conditions such as an improving economy rather than to crime policies. The reasonable and often correct consensus is that crime will return to its regular rise and that the government will have no plan for dealing with it when it does. Although crime rates began declining in Buenos Aires after 2003, record increases in the 1990s and highly publicized crimes gave society the impression of uncontrolled criminality that continues to buoy fear. After a series of robberies in 2004, for example, the percentage of city residents fearing crime climbed to 57 percent—a rate similar to that in much more violent cities such as São Paulo in Brazil.3 Like the panic attacks that people often suffer when fearing the repeat of a trauma, crime has joined inflation and military unrest as an automatic trigger for instability in Argentina.

Reflecting this fear, polls around the region indicate strong public approval for more police and harsher punishments. One poll in Panama, for example, revealed 80 percent support for an iron fist or mano dura.4 But surveys often present these options without suggesting concrete alternatives such as preventive social programs. In many cases, such as the Panamanian poll, surveys are prompted by growing public rancor, so that complaints about public security system overwhelm ideas for its salvation. When possible solutions are left out, problems seem even more intractable. Broader and more balanced polls, such as Chile’s National Survey of Public Safety (Encuesta Nacional de Seguridad Ciudadana), include items such as opinions on specific programs. But most other surveys that go into depth focus instead on questions of personal victimization and “sensation of insecurity.” Even when the objective of such polls is to achieve a detailed picture of crime patterns and concerns, such as in the 2008 door-to-door citizen security survey in Santa Fe, Argentina, expectations are raised that people’s responses will be reflected in policy.5

However, polling is usually drowned out in the media by sensationalism, which ramps up public panic even more directly, often by fixating on a particular crime and judging insecurity by it. For example, the 2008 abduction of a businessman’s son in Mexico prompted a national review of crime policy. Such a focus can paint officials into policy and political corners, provoking them to take concerted actions that bring the rate down. As described in chapter 4 on Honduras, for example, many governments react to a spike in kidnappings—a crime with high levels of media exposure—by quickly forming kidnapping units that fail to address long-term causes that prompt kidnappings to rise once attention has moved on. Emboldened reporters enjoying greater press freedoms are able to cover crime scenes more easily and with less interference, exposing the police force’s fallibilities. But the police, assuming a defensive posture, rarely provide enough internal training in press and public relations, which only confirms officers’ views of this area as tangential to policing and citizens’ views of the police as inept. One police officer in Buenos Aires said that when there is a crime in his district, the press tend to report it under headlines blaring “insecurity.” He then approaches the journalist on a “not authorized” basis to explain the specifics of the case, conducting what is by that time little more than damage control. Moreover, the less public relations training received by officers and the fewer of them working in it, the more likely it is that the information they provide to citizens and policymakers alike will be not be clear and reassuring. Positive programs or developments will go unreported, leaving only the negative ones—of which there is never a lack.

As these cycles accelerate, citizens begin to lose trust in officials who frenetically sacrifice long-term structural change for immediate results. Even without a sense of urgency, most governments lack the political and institutional support to enact long-term plans. Proposals tarred as being soft on crime, or that tackle structural problems head-on, are often dead on arrival. In some cities such as New York and Bogotá, mayors enjoying strong popular momentum for change have initiated a security overhaul based on new models. In most cases, however, the national executive has enough citizen backing to “do something,” but not enough institutional backing to “do something” long-term. Despite executive dominance on citizen security policy, many inside and outside of government around the region comment on the lack of executive interest in its details, often rooted in a longstanding but frequently unstated agreement between government and police not to open each other’s Pandora’s boxes. By the time they do become concerned, problems may be too entrenched to root out. Trapped in these currents of public opinion and political realities, which they help to stir up, most governments paddle through their tenures in office with piecemeal and often contradictory responses.

One of the biggest casualties of such politics is the minister in charge of the police—the interior minister in most Latin American countries. The tenure of any cabinet minister rides on the ups and downs of the issues they oversee, but because of the volatility of citizen security, interior ministers and their staffs cycle in and out of office with especially unsettling frequency. New ministers often discard their predecessors’ plans, particularly those that contributed to their predecessors’ downfall, to make way for their own. Other ministers avoid change altogether because they anticipate a short spell in office, because they do not want to antagonize the police, or because they want to avoid association with yet more failure.

Such instability is one symptom of the wider political reality in which officials are always thinking about moving to the next position because the current one is perceived to be—and thus continues to be in fact—essentially powerless. In addition to rotating ministers, power declines when presidents eliminate ministries, downsize their staffs, or move police authority among them, such as from Interior to Justice. Such moves often seem designed to head off change rather than to promote it. But even though such decisions can stop a major reform in its tracks, they can also revive a moribund one. For example, Venezuela’s Police Reform Commission was created in 2006 by Jesse Chacón, who was one of nine ministers of interior and justice during the first eight years of the government of President Hugo Chávez. The commission conducted extensive national surveys and in-depth studies as a basis for comprehensive change. However, Chávez declined to continue this work after firing Chacón in January 2007, and subsequently proposed sweeping laws that would undermine the commission’s detailed recommendations, one of which called for creating an integrated system to coordinate the country’s disparate agencies. As crime rose to record levels the following year, Chávez put his authority behind a renewed effort that led to a new Police Service Law and more than seven resolutions to implement unifying national standards.

More dexterous presidents and governors escape from this vice of public pressure and policy impotence by adopting mano dura anticrime policies, which are popular, easy to formulate and enact, and the best guarantee for quick results during a limited time in office. Often compared unfavorably and unfairly to their authoritarian predecessors and their abilities to keep order, many elected presidents even affix Cold War–era “national security” language to their policies. Such language opens the door to military involvement in policing and waters down legal oversight of both individual rights and executive action.6 Executives rarely hesitate to use appointments, special funds, and party patronage to keep legislative and judicial intrusion at bay. Such attitudes support the view of contemporary Latin American regimes as “delegative democracies” in which presidents see themselves as the embodiment of the national will and thus find the “nuisances” of other institutions an “unnecessary impediment to the full authority” that they have been “delegated to exercise” (O’Donnell 1993, 6).

Legislatures do not do enough to check such power. Although they are the primary counterweight to executives, they usually lack the party discipline, budgetary resources, or political heft to counter executive decrees, promote viable policy alternatives, or oversee police practice. Most legislatures are too controlled by the president’s party to pass bills that implicitly criticize that party, and even when a different party or no party is in the majority, disorder and a lack of discipline hinder deliberation of feasible policy changes. Hamstrung by such political divides, many legislatures are too afraid to antagonize a powerful police agency or too spooked by public demands to take any meaningful action that might be blamed for any subsequent upticks in crime. Instead, legislators often direct their criticism at the social sectors associated with criminality, often with unrealistic and unconnected proposals bordering on the demagogic. Demands to adopt or reinstate capital punishment are one staple proposal masquerading as decisive action and guaranteed to attract attention.

Both legislatures and courts find it difficult to assert themselves. For one thing, legislatures tend to exercise just a fraction of their oversight power over both civil rights and fiscal matters. In addition to their own reluctance to act, financial and other restrictions curb action by both legislators and judges, raising questions about their authority over the state and the rule of law. Are police roundups justified when the legislature does not provide funds for the investigative training needed to avoid them? What about detentions beyond the time limit set by judges when no judge is available to take the case? Or what about curfews that a government approves but whose excesses it ignores? In a democratic regime, elected officials have the constitutional authority to formulate, regulate, and monitor the rules, funding, structures, and actions of security agencies (see Altamira 1963). In most democracies, the legislature has this “police power,” and police agencies are obligated to abide by the laws and decisions that result from it. But the steady erosion of this norm over the course of Latin American history and the failure to reverse it in the current era have blurred the line between the legislature’s police power and the power of the police. On a daily basis, in fact, it is often the executive and police officials who decide how those structures, funds, and rules are put into practice. In the process the police themselves often redefine, reinterpret, and reformulate those same laws to the point to which their actions become de facto police power.

Under such conditions would-be reformers face an uphill struggle. Just as in the analysis of the health sector by Kaufman and Nelson (2004, 24–31), they confront “a diverse array of opponents and indifferent groups” in an environment in which “rewards are delayed and uncertain, but political costs of reforms are prompt and clear.” Because of public panic, security is an issue particularly prone to such treatment, evident, for example, in the ability of the police and elected officials to boil crime policy down to a choice between mano dura and human rights (see Maravall 1999, 157–58). “While advocates of change must win at all stages of the political process—issue-recognition, decision, and implementation of policy—the defender of existing policy must win at only one stage in the process” (Bachrach and Baratz 1970, 58). As a result, the vast majority of citizen security legislation in Latin America originates in the executive branch and, unless formulated by a new government, strengthens rather than reevaluates existing policy. Such outcomes of executive-legislative relations not only reflect power imbalances, but also increase the uncertainty among citizens (see chapter 3).

Venezuela, where no fewer than ten police reform proposals were under way in 2006, is a good example of both legislative abandonment and general inconsistency in security policy. After the 1958 transition to democracy, two political parties, Democratic Action (Acción Democrática) and the Christian Party (Comité de Organización Política Electoral Independiente), ruled in a pact and rarely enacted the laws needed on police structure, training, and budgets. A lack of oversight also led to widespread abuses, including death squads linked to the police.7 The election of Chávez in 1998, amid the disintegration of the two parties, promised to break this pattern. Since then, however, the legislature has reflected rather than responded to the politicization of policing and other issues as it has been pulled into the confrontation between the president and the opposition, both of whom are trumpeting abuses by police forces controlled by the other side.8 After December 2005, when Chávez-allied candidates took full control of the legislature in the wake of elections boycotted by the opposition, constructive criticism of national crime policy all but ceased for several years.

Under some conditions this sorry state of executive-legislative relations can be remedied. One is when a weak president allows the legislature to be more powerful. In the late 1990s, for example, Uruguay passed reforms under the embattled government of Jorge Battle (2000–5), and Venezuela’s legislature enacted a comprehensive penal process code during the waning government of Rafael Caldera (1994–99). In other countries, such as Bolivia under Gonzalo Sánchez de Lozada in the 1990s, stability allowed the passage of sweeping change. The absence of such reform around the region, however, highlights the fact that such cases are exceptions to the rule of inaction.

Push for reform can also come from below—often when the victims of violent crimes that are committed, ignored, or bungled by the police are from the middle or upper classes. Demands to bring the police to heel are particularly vigorous when a highly publicized act of violence or corruption prompts citizens to realize that unchecked excesses—of corruption, violence, or other misdeeds—distract officers from fighting crime. Some of the biggest attempted crime system overhauls in Latin America, such as in Peru and Mexico City, have stemmed from such a widespread belief. In these cases, though, the resulting pressures either dissipated or were channeled into impractical, contradictory, or even mano dura proposals. For example, the 200,000-strong march in Buenos Aires in April 2004 after the kidnapping and murder of a middle-class student led to a raft of laws to stiffen criminal punishment. Because many such attempts at reform are made during times of political change, such as in Brazil and Argentina in the early 1990s, support for them often dries up once those temporary political conditions pass. With reform policies rarely outliving the sponsoring governments, the police are emboldened in their claims that the policies would hamper crime fighting or create too much uncertainty, such as the Chilean Carabineros’ resistance to structural and personnel changes by a popular democratic regime in the 1990s.

Together, these conditions make it difficult to enact legislation in all five areas of reform described in chapter 1, particularly those, such as control mechanisms and community policing, that require both more funding and oversight. But even when such difficulties stymie change in the short run, they lay the groundwork for it in the long run. Most often they introduce ideas into the public debate that lead to legislation scaled down to fit the limitations. When a security crisis leads to or accompanies an economic collapse or political rupture, the rejection of state elites can open up opportunities for new, more citizen-centered approaches. In fact, in Argentina, Bolivia, Ecuador, and Venezuela the breaking point in the contemporary collapse of constitutional governments was the inability of security forces to handle the citizen unrest that grew out of unmet demands for change.

2. The Law: Criminology and Zero Tolerance

The law, by applying constitutional norms to state action, is supposed to steer and steady such change. Instead, the law is often swept up in it. Criminal laws in Latin America frequently have too much overlap, historical baggage, and inflexible detail to cope with the ever-changing issues of security. As this section discusses, the sheer extent of criminal law and its partiality for zero tolerance policing leave little space for the kinds of citizen participation and transparency that appear contrary to the certainty to which the law strives. Thus throughout Latin America those moving toward problem-oriented policing find it difficult not to succumb to the gravitational pull of this heavy legal structure.

The legal base of Latin American policing comprises penal codes, penal process codes, police codes, ordenanzas (ordinances—the part of administrative penal law used primarily to regulate public order); contravenciones (misdemeanors—both ordenanzas and contravenciones are also called faltas, and most edicts are a category of faltas), reglamentos (regulations—these are part of disciplinary penal law and used mostly for internal police matters such as discipline); prohibitions; police orders; permisos (permissions—these are the exemptions to regulations); special permits or authorizations; and customary laws made up of unlegislated but long-held practices.9 Most laws, ordenanzas, and reglamentos can be replaced or revoked by officials with police power and, in some cases, by the executive agencies that oversee their execution. Although most of these laws are in distinct legal codes covering specific areas, overlap and confusion are common, especially in the areas of policing in which discretion gives judicial interpretation particular weight. Differing views on the use of force, collection of evidence, treatment of at-risk youth, and other key security issues can be and often are supported by different laws, particularly in the absence of judicial clarification. But new national, provincial, and local laws continue to be enacted far more quickly than they are interpreted. The government “changes laws like socks,” as one police chief griped, making it hard for him to know which ones apply and how they do.

Commenting on this rush, government and police officials refer to how poorly new laws are adjusted to institutional capacities. For example, efforts to collect private arms often fail because trust is not built with the targeted community. This problem was evident in 2002, when, in response to a 58 percent increase in firearms killings, Venezuela enacted a disarmament law.10 The law provided economic incentives to turn in private arms—estimated by the police to number 600,000 in the Federal District alone11—but it did not take into account the cooperation needed with community groups to actually collect those weapons. General crime policy is also affected by new laws, such as when court delays slow down detainee processing. For example, the increase in speed that new penal process codes bring to many criminal justice processes is negated by the greater number of arrests that result from wider police power. Because of their ambition, in fact, new penal process codes expose, more than other reforms, the wide gap between the promise of legal change and the reality of legal practice in Latin America. In nearly every country, a combination of inadequate funds, poor training, bureaucratic inertia, and resistance by police agencies resentful of losing investigative power has impeded the application of these codes. Institutional disarray, expected during the initial phase of a democratic transition but continuing in many countries well past it, can thus sideline even the most practical and well-designed legal reform.

The vacuum in criminal justice that results is often filled by zero tolerance policing, a more straightforward approach thought able to cut through the legal thicket. That line of action stems, at least rhetorically, from the “broken windows” theory developed by James Q. Wilson and George L. Kelling in 1982, which asserts that antisocial behavior such as intimidation and harassment, along with the physical deterioration of property, scare off law-abiding citizens and allow crime to take root (Wilson and Kelling 1982). A broken window in a building is a signal that no one cares about the property, encouraging delinquents to gather in front of the building and from there to begin dominating the neighborhood. As residents then minimize their time in the street, the sense of public order slips out of their control even without actual criminal activity. The only effective way to prevent these conditions from setting in, the broken windows theory posits, is by giving the police greater authority to detain people for misdemeanors and antisocial behavior. Proponents of these policies reason that legalized clamp-downs on such actions not only prevent potential criminals from being emboldened to undertake more serious crime but also reveal illegal arms, suspects wanted for outstanding crimes, and other illegal possessions or activities that would otherwise go unchecked.

In New York City this approach appeared to yield an unprecedented drop in the homicide rate from a high of over thirty per 100,000 residents in 1990 to under ten in 1998, with continuing decreases through 2009.12 Because such results gave the policy a high level of credence and support early on, it was adopted in the 1990s by police forces throughout the United States. Other countries soon followed, most formally Indonesia, the Netherlands, South Africa, and the United Kingdom. Many other areas adopted different elements of the approach, leading to innumerable permutations and invariably more heated debates. Understanding whether the broken windows theory works in each of these countries requires a connection between theory and tactics under their particular circumstances—that is, what specific approaches do police develop out of the broken windows theory, and how do those approaches serve its goals? In Latin America, the region that probably has adopted its own forms of zero tolerance more extensively than any other, harsh social conditions and weak states make such questions particularly important to understanding the applications and impacts of broken windows theory in new scenarios.

In any country a first connection between theory and tactics is a definition of disorder. A study in the Netherlands that set up and observed behaviors such as littering, trespassing, and stealing in six locations is perhaps the only controlled experiment that tested the broken window theory itself (Keizer, Lindenberg, and Steg 2008). In the experiment nearly twice as many people violated clearly posted laws in areas with graffiti and garbage as in the same areas without such contamination. This experiment thus supported the contention that the broken windows theory can best be used to reduce disorder when it defines the types of disorder it is targeting.

In practice, however, disorder is not defined and instead is employed as a general term. Indeed, definitions may be politically sensitive and could box the police into particular approaches while holding them to goals they cannot meet. Many practices associated with broken windows are not well defined, and, as a result, they often provoke controversy. In the United States, for example, the police practices arising from broken windows and zero tolerance include racial profiling, harassment, and deployments “to reconstitute” public spaces for the private gain of local business (McArdle 2001, 4). In Latin America, as this section and the following chapters discuss, the theory has been used to justify roundups, curfews, torture, and extra-judicial killings.

A second and closely related connection between theory and tactics is extent and balance. What are measured and proportionate actions against disorderly behavior, and how far should they go? As discussed in chapter 7, the quality of life enforcement first used in New York City’s subways was based on the power to stop people for suspicion of committing a range of minor offenses. Suspects who had identification and no outstanding criminal charges would usually be given a desk appearance ticket (DAT), leading to a fine or community service. Although this approach reduced disorder, its early effectiveness eventually led police to overly depend on it, thereby generating citizen antagonism, police resentment, neglect of larger patterns, and court delays. Early critics also legitimately saw zero tolerance as a form of control and discipline of marginalized people without attention to or rehabilitation of their conditions. However, several elements of that approach continued to be beneficial. One was handgun checks in which a bulge in someone’s clothing allowed the police to ask them to empty their pockets, and, if deemed necessary, to subject them to a pat-down. Zero tolerance has even shifted trends in drug use since 1998, when the New York Police Department (NYPD) began to arrest and book individuals for smoking blunts, marijuana in a cigar wrapping (see Johnson, Golub, and Dunlap 2006). Although the use of blunts is a positive trend in the sense that it has replaced heroin and cocaine as the drug of choice among youth, public smoking is considered a public nuisance and a source of disorder. In smaller cities throughout Latin America, many police officers say their biggest problem is drug gangs, which are often groups of pot-smoking youth. But does their detention go hand-in-hand with the use of social services to prevent recurrence? Would driving illicit youth activity behind doors be acceptable? Zero tolerance of such behavior depends on the community definition of disorder and whether the overall impact is beneficial.

A third important connection between theory and tactics is correlation (also see chapter 3). Many of the most comprehensive studies, most notably by the U.S. National Research Council (Skogan and Frydl 2004) and Sampson and Raudenbush (1999), present mixed or inconclusive results for zero tolerance policing. Some reports on New York City show only a tenuous relationship between zero tolerance policy and the drop in petty crimes (Harcourt 1998; Eck and Maguire 2006; Harcourt and Ludwig 2006). Although the broken windows theory surfaced at the onset of the rapid decline in New York City’s crime rate, it was in fact only one of the catalysts of the decrease. Sustained economic growth during the decade, the end of the crack epidemic, and demographic changes (particularly a relative decline in teenage residents) were also contributing factors (see Karmen 2000). Zero tolerance was also connected to the broad neoliberal economic approaches adopted by governments after the 1980s, which took forms such as welfare reform, public service privatization, and business improvement districts (BIDs). So zero tolerance was part of these larger socioeconomic and policy trends, which are associated with both decreases in crime as well as negative patterns such as societal exclusion (see Punch 2007) and limited public space, as discussed in the next chapter. But the particular contribution of zero tolerance was to help policy be better attuned to changes in economic and demographic conditions by highlighting the need for social services to address those changes, such as for drug treatment, in ways that address the causes of crime.

Together, these three connections help reveal the differences between traditional and problem-oriented policing, as well as the ways in which they could support each other. In particular, traditional approaches can be employed in the service of a problem orientation by defining the sources of disorder and marshaling state power to reduce them. For example, the hierarchical structure of the police can be effective at coordinating officers’ actions, because consistency and clarity are needed to avoid discrepancies in police action. The reactive responses at the core of traditional policing can also work to handle quickly the most urgent security concerns, giving problem-oriented approaches more time to address their deeper sources. On another level, the executive controls associated with traditional policing can best compel involvement by other state agencies. As it incorporates those other actors, the traditional approach can also minimize the politicization of citizen security by keeping intact the authority of officials, while helping them recognize the need for greater institutional and societal involvement.

To hold broken windows policing true to its original intentions—that is, to keep tactics as close as possible to theory—the police and criminal justice system must have certain baseline capacities. For example, because the police come in closer contact than do other state officials with the causes of crime such as drug addiction, all officers must undergo legal training. Coordination with social services, functioning courts to process detainees, and the oversight of government agencies, courts, NGOs, and the media are also required. In other words, in order to work, policies based on the broken windows theory must combine the co-production part of its approach, in which different agencies work together, with the incivilities component, in which antisocial behavior is curtailed. To take one example, handing out summons for chronic drunkenness works only if the courts and social services are able to stop the revolving door of nights spent in jail. In reality, though, co-production is usually outpaced by the focus on incivilities (Taylor 2006, 106–8). In many areas of Latin America and other regions, zero tolerance is often applied with little outside support and few controls, turning it into a mano dura approach that is little more than a continuation of police practices from the predemocratic period that provide legal cover for the police to control society’s “dangerous” elements. Zero tolerance can thus give the illusion of improvement without necessarily incorporating the structural changes that make it self-sustaining. It reflects a lack of continuity, coordination, and confidence in the criminal justice system.

Although zero tolerance policies are relatively recent, gaining predominance in the 1990s, they have a strong legal basis. Helping to institutionalize and drive this approach are the edicts and other internal regulations that most Latin American police forces have been acquiring since the colonial era. These edicts and regulations empower the police to detain citizens for an expanding range of subjectively defined activities and behaviors, from “vagrancy” to “suspicion of criminal intent.” For example, the provision in Chile’s 1901 penal code allowing police to make arrests based on physical appearance was not eliminated until 1998. Although the majority of these provisions are public, the citizens most affected by them tend to be the least aware of their rights when detained under the provisions. And though technically legal, most edicts often include exemptions from regular judicial procedures—such as required notification of a judge within a certain time period—and contain vague wording that expands police power beyond the intended limits of constitutional law. Because of the upper hand that edicts and other internal rules give officers, actions and detentions based on them take up a large part of officers’ time. In Guatemala, for example, an estimated 80 percent of detentions and arrests are for alleged misdemeanors (Scheye 2005, 2). As police operations become centered on edict detention, statistics are also affected, with the number of such detentions serving as a police force’s self-referential measure of success. Thus once in place, a mano dura approach starts to legitimize itself as arrests begin to replace conviction and recidivism rates as the barometers of policy success.

In nearly every country in which such regulations have been struck down for being unconstitutional or abusive, governments have tried to reinstate them in new forms or by increasing use of other regulations such as checking a person’s identification or police record. For example, Venezuela’s 1939 Law of Vagabonds and Crooks allowed the police to detain anyone deemed “suspicious” and was one of the main bases for police detentions until it was finally declared unconstitutional in 1997. It was quickly replaced, however, by stepped-up checks for identification or a criminal record, which usually involved detention in a police station. Edicts have also been substituted or supplemented by newer “social control” laws enacted since the democratic transition. Often introduced as part of zero tolerance policing (in some cases as measures within special police operations), these laws span the range of police functions and increase the unregulated control of the police over citizens even as the temporary conditions or operations end. At other times the repressive laws and measures applied to guerrilla groups are used against criminals, such as in Venezuela in the 1970s and Peru in the 1990s, supported by extra roundup powers and special courts that are allowed to continue.

The main tactics of zero tolerance policing, particularly detention for lacking identification or for “antisocial” behavior, demonstrate how laws can be counterproductive to law enforcement. They distract from a focus on serious crime and the causes of crime, distend judicial backlogs, raise tensions with citizens, and complicate any moves toward problem-oriented policing. Without follow-up or coordination with social services or the courts, such actions do not serve the broken windows purpose of stopping more violent crime. In Costa Rica, for example, nearly fifty thousand misdemeanor violations are reported each year—a figure that drew attention after the Supreme Court prohibited detentions for them.13 But as many criminal justice officials admit, the real problem is that these detentions involved no civilian assessment or even police reports that could inform crime policy. The region’s new penal codes have raised the bar on the evidence necessary for detention, but instead of scaling back their actions to adjust to the new codes, most police officers have only accelerated their arrests of suspects, most of whom are quickly released. Using edicts and other powers to resist the new codes and other reforms, the police then thicken the line being drawn between “public order,” associated with a strong state, and “human rights,” associated with delinquents. As discussed in chapter 3, human rights groups then often find themselves on the defensive over abuses by the police. Though such abuses are among the most extensive form of rights violations in the region, they are skillfully portrayed as a necessary price for crime control. Indeed, one of the five principal problems that Ecuador’s police defined is that “Human Rights Organisms have carried out campaigns to discredit the Police Institution, which has affected the morale of its members” (Policía Nacional de Ecuador 2006).

3. Police Career: From Recruitment to Advanced Training

In Latin America the regulations that detail each stage of an officer’s job are mocked by the irregularities that most affect it. Such weaknesses complicate agency restructuring, community policing, and any other reform that entails some change in an officer’s responsibilities or evaluation. The contradictions between police career laws and practices begin with recruitment, which in most countries falls short on entrance requirements and incorporating a representative cross section of the population. El Salvador’s rush to deploy its postwar police allowed many youth gang members into the force (Call 2003b), and in Guatemala’s rush to fill its police ranks after the civil war, nearly one-third of its officers did not fulfill the minimum education requirements and nearly half received less than three months of training (Sequén-Monchéz 2003, 148). Many NGOs and reformers point out the short training periods and advocate lengthening them. The average length of training for new recruits in Latin America is six months, which is short but arguably adequate if the expectations and standards of the training are high. Basic training in the New York Police Department’s academy is also six months, but recruits must meet stringent standards. One wrong answer on the use-of-arms exam, for example, results in automatic dismissal.

A lack of clarity in expectations also tends to attract applicants for reasons not associated with the professionalism or reality of the job. Many come from police families or seek steady employment, but others, in part because recruiters fail to disavow such ideas, have exaggerated visions of their law enforcement prowess, or, like some cadets in Peru’s central police academy, admit their attraction to the unofficial financial benefits.14 Limited and uneven training then continues into the force. In the Dominican Republic, where about 80 percent of police personnel did not graduate from the police academy, only 20 percent are promoted based on their official levels of training and preparation (Bobea 2003, 217). Miami, Florida, has also learned that the drawbacks of fast recruitment often outweigh the advantages of recruiting more officers over a short time period (see Dorschner 1993). Without quality control, this approach can produce more slain crime suspects, a declining quality in written police reports, internal conflicts over issues such as promotion, and more administrative demands that draw energy away from daily law enforcement.

Throughout Latin America, academies for both high- and low-ranking officers have gradually responded to their weaknesses. Most have expanded their curricula to incorporate subjects such as human rights and the perspectives of civilian specialists. In addition to collaboration with local universities (discussed in chapter 7), the growth of security studies by private universities has also widened the educational base of the police. In some countries and provinces the focus of reform on the academy has paid off. In Uruguay the Escuela Nacional de Policía (National School of Police) has some of the region’s best facilities and extensive curricula, both of which have contributed to the higher morale among officers and citizens’ trust of them. This impact has even allowed the academy, according to its director, to expand in a time of financial cutbacks.15 But in most countries police academies have not fundamentally altered the basic emphasis on physical preparation and formalistic knowledge of laws rather than critical analysis, open discussion, and application to societal realities. In Venezuela the police themselves rate their basic training as poor—particularly on social conditions, firearm use, and human rights.16 Because the police have depended historically on the military for training and intelligence, the military’s “logic of arbitrary repression” (Kruijt and Koonings 1999, 10) and “indiscriminate stigmatization of adversaries” (Koonings 2003) have been ingrained into the preparation of Latin American police officers. As discussed in chapters 4–6, many police academy directors and teachers around the region feel too constrained by expectations to experiment with new approaches. Moreover, the great majority of academy directors seem to be either long-serving veterans whose embrace of new approaches is limited or outsiders whose ambitious plans dissolve amid internal intransigence. In other words, the first skirmish between traditional and problem-oriented policing often takes place in the police academies.

In most countries such educational deficits are being addressed through programs sponsored by NGOs and universities. But such efforts tend to be geared toward particular specializations without subsequent coordination with human resource planning. In fact, as discussed in this book, a perennial problem is that few officers actually work in the areas in which they were trained. For example, fewer than fifteen of the thousands of police trained as investigators in Guatemala are actually working as investigators (Scheye 2005, 2), while in Venezuela just 52 percent of officers carry out patrols, investigations, or other activities related to citizen security (El Achkar, forthcoming). This problem extends from the lower-level officers used as guards and drivers to the higher-level officials who are never able to use the specialized training they received in areas such as financial or other white-collar crime. Even in Costa Rica, where the police have more political and financial support than do those in most other countries, officers complain about the paucity of educational opportunities, about promotion procedures that do not take into account educational background, and about the courses they do take that lead to a miserly 1.5 percent increase in salary.17 Higher-ranking officers often stress the lack of coordination between police academy and higher-education courses, because they are taught by different institutions and often in different locations. In the middle ranks, officers sent abroad for foreign training are often marginalized once they return, because the training itself is considered a perk (probably awarded unfairly, in other officers’ eyes) rather than a step toward institutional change. In addition, in response to recent educational improvements, the better-educated officers are held back in the lower and middle ranks by superiors suspicious of their new approaches. When imposed from outside, such educational opportunities are often rejected as patronizing external criticisms of the country’s own police training.

Education can be bolstered through its link to promotion. In nearly every country certain courses are required for moving up. But in the promotion process successful completion of those courses is often outweighed by less objective criteria. Indeed, promotion in most police agencies is not based on a critical evaluation of positive actions or specific skill development. Instead, it depends almost entirely on the officer’s seniority, academy exam score, ranking, and whether he or she has caused any problems. Most forms used for individual promotion are usually one or two pages, with few if any probing questions about the agent’s actual police work. Enough officers jump ranks through extraordinary means to elicit vehement complaints among the lower ranks throughout the region. Such favoritism can result in officers being rejected for advancement without cause or assigned to high-crime districts.

All of these educational shortcomings harm policing by limiting officers’ abilities to detect the causes of crime, to adapt to changing situations, and to convince their superiors to incorporate new ideas about police work in policy. The scant coordination between the police and social services on problems such as drug abuse also limits the ability of the police to develop and promote such changes—for example, categorizing at-risk youth in different neighborhoods. This dearth of new approaches at the top then depresses incentives for new approaches at the bottom, particularly in the high-crime urban areas where they are most needed. Any positive steps that individual officers do manage to take are then often lost by changes in minister, policy, or chief. Automatic rotation, which officially happens every two years in most countries, makes it particularly difficult for officers to follow the professional track so neatly outlined in police career laws. In some countries or within certain ranks or units, and often unofficially, rotation can be even more frequent. In Guatemala, for example, comisarios often rotate within a period of less than one year, which is problematic in such an ethnically diverse country, where it takes time for officers to understand each of the communities they serve.

4. Police Structure: Hierarchy, Decentralization, and Proliferation

Like other state services, security has been decentralized and outsourced throughout Latin America in the current democratic era. This process has spread along three main dimensions: (1) the decentralization of the police force into geographically defined agencies; (2) a functional breakup of the police into units based on functional specializations (such as preventive, investigative, traffic, penitentiary, and special crimes units); and (3) the growth of private security enterprises that have taken on many police services. Most of these changes, particularly the breakup of powerful national police forces, were overdue. The formation of local police forces allows the municipalities to tailor policy to local conditions, in particular, while the functional separation allows a sharper focus on different crimes and the formation of special units allows the police to adopt cool names such as the “Leopards” and the “Cobras.” As discussed in the sections that follow, however, the complications created by these changes often overwhelm their advantages.

Geographically Defined Agencies

As Schumpeter (1950), Dahl (1998), Oxhorn, Tulchin, and Selee (2004), and others recognize, decentralization can improve many areas of state service, because governments usually respond better to citizens when they are closer to them—particularly in services with heavy societal contact such as policing (see Ostrom 1975). Decentralization can also bolster democracy by allowing minorities to use the smaller political arenas to press their demands (as indigenous groups in Indonesia and Mexico are doing) or through local initiatives for equitable resource reallocation (such as participatory budgeting in many Latin American municipalities).

Like many broad changes in governance, however, decentralization often lacks adequate policy deliberation, political objectivity, or operational justification. This is particularly true of citizen security, where provincial and local police are created too quickly, often because they are the easiest way in which the government can respond to political pressures from below or break up poorly functioning national forces. In Ecuador, for example, municipal governments began enacting security programs because of the lack of a national plan (Pontón 2005). Municipal guards in Chile and Guatemala, a precursor of local policing, were also created through local initiatives (Frühling 2003). But without careful preparation, provincial and municipal forces repeat and often amplify the poor management, accountability, and policies of their national progenitors. Although crime does not get simpler at the local level, the range of agencies to deal with it is narrower, giving local executives a smaller base of expertise and support.

Decentralization also accentuates executive and legislative imbalances. Compared with their national counterparts, provincial legislatures and municipal councils have less of the political leverage, information, and funding needed to regulate police action. In Costa Rica the local police lack regulatory statutes that clearly demarcate their roles from those of the national police. Above all, decentralization highlights financial differences by making local areas more protective of their own security budgets. Throughout Latin America this situation has engrained differences in policing quality by allowing wealthy areas to bankroll their own vastly superior security agencies.

Even the most carefully delineated decentralization efforts also raise tensions among federal, regional, and local governments. Broadly understood as the transfer of fiscal, administrative, and political power to various subnational levels, decentralization falls into three main approaches: (1) deconcentration, in which administrative offices are physically dispersed but policy control remains at the center (see figure 5.1 in chapter 5 on Bolivia’s national police structure); (2) delegation, in which some decision-making authority is transferred to the subnational level; and (3) devolution, in which maximum decision-making authority is handed over to subnational governments. Because security requires national policies, a decentralized security system usually adopts one of the first two approaches, allowing national officials to retain policy and often operational control. In addition, as Willis, Garman, and Haggard (1999) point out, when decentralization is carried out by strong national parties—as in most of Latin America—the federal government retains more control than when decentralization is implemented under multiple parties that leverage regional power. When decentralization comes laden with political conditions—what one officer calls an “all strings attached” relationship—the local autonomy central to success is limited. One form of such continuing central control is the use of the military in policing. As this book discusses, the armed forces are brought into police operations through many routes: special raids (such as against gangs), takeovers of high-crime areas (such as in Buenos Aires Province), the formation of joint military-police units (such as the antidrug units in the Andes), and control over issues such as narco-trafficking. The process of decentralization itself can be destabilizing. Mexico City increased insecurity by creating “gaps in authority and power” that cause greater police abuse and corruption as new and old agencies jockey for power and territory (Davis 2003).

Functional Breakup

A second form of police propagation is a breakup into units based on functional specializations, which is also long overdue. But even the rudimentary separation into preventive agencies (to prevent violations) and judicial agencies (to investigate them) is often plagued by poor coordination, overlapping activities, and spotty oversight. Such problems are amplified when agencies are placed under different ministries. For example, Costa Rica’s eleven different police forces belong to five national ministries as well as to the judiciary and municipal governments. But the more agencies chasing after the same problem, the more likely they are to trip over each other in policy and logistics. For example, discord often arises between the more socially focused local drug policy and the more repressive national militarized approaches, while control over international goods often involves clashes among military, border, and customs units. Moreover, the increasingly international scope of crime contrasts with the limited internal jurisdictions of police forces, and the flow across borders of some of the most lethal forms of violence in Latin America—from gang wars to contraband—flourishes amid unsynchronized policing. And yet steps to coordinate or centralize these bodies can foment intrainstitutional conflict, such as that among Italy’s three national police forces and between city police and the Interior Ministry police in France. By contrast, in the United States geographic and political separation has resulted in a proliferation of police forces with minimal institutional conflict. But Latin America, with its own particular historical legacies and geographic disparities, has the worst of both worlds: institutionally rigid and politically powerful police at the federal level and more unprofessional and unaccountable agencies at the provincial and municipal levels.

At times, proliferation has been more bureaucratic than functional. As one Bolivian planner remarked, the organizational diagram of his police force began to resemble a Christmas tree with new units hung on them like ornaments. One trendy form of such decentralization is “touristization,” in which areas of high visitor traffic get extra deployments or even their own police, such as in Mexico City, Guayaquil, and Caracas. Although such a strategy may make sense by providing extra security in areas of high vulnerability and consumption, it can also perpetuate geographic divisions, as discussed in the next chapter. Creating new units may also be a way to marginalize issues such as human rights or community policing while appearing, for public relations purposes, to prioritize them. And when new units are integrated into daily operations, their growing number slows down decision making and dilutes implementation. Often, old units continue to operate despite having been officially dismantled. Particularly in the areas of internal affairs and antinarcotics, agencies have used their institutional powers to withstand the will of civilian overseers, showing the kind of skillful adaptation to external shocks discussed in chapter 1.

Venezuela demonstrates the downside of decentralization. The country has twenty-four state forces and ninety-nine municipal forces, with 126 police agencies, 105 of which were formed after enactment of the 1989 Decentralization Law (a 363 percent increase between 1990 and 2006).18 In addition to the Body of Scientific, Penal, and Criminal Investigations (Cuerpo de Investigaciones Científicas, Penales y Criminalísticas, CICPC) and the Office of Intelligence and Prevention Services (Dirección Sectoral de los Servicios de Inteligencia y Prevención, DISIP), military police forces are also involved in policing, including the Defense Ministry’s Office of Military Intelligence (Dirección de Inteligencia Militar), the Office of Army Intelligence (Dirección de Inteligencia del Ejército, DIE), and the National Guard (Guardia Nacional, GN). The military police overlap with state and municipal police, often with friction. In 2001, for example, GN personnel were forcibly placed in the stations of the Federal District’s Metropolitan Police (Policía Metropolitana, PM),19 which operates alongside nearly a dozen municipal forces in the Federal District. As in other countries, such Balkanized arrangements place the burden on local governments and cause wide disparities in service. Caracas, which has one of the largest and most varied police networks of any city, has surpassed South Africa’s Cape Town as the world’s most crime-ridden city; according to the Justice Ministry, it was the scene of an astonishing 152 murders a week in 2008.20 The wealthy eastern half of the city has ten times more police officers than the poor western half, which has over 70 percent of the city’s residents and most of its crime. Over two-thirds of all criminal reports originate in the biggest western municipality, Libertador.21 Such differences are repeated throughout Latin America. In Medellín, Colombia, the poorer Zone 1 has a homicide rate of 101 per 100,000 persons, while the rate in the wealthier Zone 5 is just 27 per 100,000 persons (RCMP 2006).

Private Security

Perhaps the biggest form of proliferation in policing has taken place in the private rather than the public sphere. Throughout Latin America private security agencies have grown rapidly in the democratic era. In Venezuela, for example, the number of private security firms has quintupled since the 1990s (Sanjuán 2003, 122). This surge has led, as the next chapter discusses, to a greater number of private officers than public officers in most Latin American cities. This growth clearly represents societal demands to fill the gaps left by poor public police services. But like the expansion of local, functional, and specialized forces, private firms complicate basic policing and coordination. In addition, more than state police forces, they pose a challenge to the nature and legitimacy of the state itself by questioning the government’s ability to fufill one of its most fundamental tasks.

5. Management: From Street Patrols to National Budgets

A complex organization like the police, particularly amid changes such as unit proliferation, must be well managed. But in Latin America police management is weak for a host of reasons. During the first years of the recent democratic transition, most of those reasons were external as sudden democratic expectations and harried decentralization bore down on the police’s fossilized hierarchies and inflexibly militarized structures. Growing awareness of these weaknesses since then has led to more deliberate comprehensive steps to improve management, such as more accountable appointments. But failure to synchronize those steps with internal practices—in particular, relations among ranks and the use of human and financial resources—has limited results.

Poor management affects an entire force, but is most evident and damaging in the police station (comisaría), the heart of crime fighting. The midand low-ranking officers that staff comisarías do not have enough autonomy, are punished for mistakes, and are not encouraged to undertake innovation. Together, professional insecurity and arbitrary discipline discourage them from developing ideas, projects, or other forms of problem-oriented policing. In most areas they usually just monitor a single corner without training in proactive crime prevention.22 Meanwhile, rotation and exhausting work schedules make officials both resentful and reluctant to take on the extra work needed to implement new policies.23 These built-in disincentives not only shape daily policing but also undercut the structural reorganization that tries to make it better.

Instead of embedded management practices, much of the debate over human resources in Latin America is instead over the scarcity of police. Although countries vary greatly, the regional average is one officer for every three to four hundred persons, leading to calls for more personnel. But this rate is not particularly low (New York and other large cities have about one for every two hundred persons), and increases in numbers unaccompanied by structural reform or citizen trust do not bring more security. In fact, a comparison of large U.S. cities shows no correlation between police numbers and crime rates.24 Furthermore, low numbers of active officers usually reflect the poor management of resources more than their scarcity. In the United States roughly 60 percent of officers are on patrol, and because most agencies are divided into shifts, only about 15 percent of a force is out at any given time (Bayley 1998, 29). These percentages are similar to those reported by Latin American police. But regionwide the police presence is thinned further by the inordinate amount of time it takes officers to complete forms (calculated as on-duty time) and the use of street officers for special assignments, for traffic patrol, or as guards for officials, buildings, and government installations.25 So not only does the use of human resources lower the actual police presence below that in official reports, but it also adds to the disgruntlement of officials when they are actually out on patrol. As this book’s case studies describe, however, a facade of resistance covers up the real interest in change within the lower ranks. Street officers are far more open to change than their superiors are willing to recognize (or interested in recognizing) because the ineffectiveness of the ad hoc operations that determine their work is so evident to street officers on a daily basis. However, the blame for this ineffectiveness and the burden of addressing it fall mainly on their superiors. When they remain unaddressed, such fundamental differences in experience and responsibility then deepen divisions within a police force.

Technology can only go so far in improving policing under such conditions. Mapping and results-based programs such as CompStat (short for Computer or Comparative Statistics) are being introduced in the region to help police respond more quickly and accurately to crime statistics and patterns. New York and other early adopters found that such programs brought police together with prosecutors, probation and parole officers, and other criminal justice officials. At these meetings police chiefs are questioned on crime patterns in their districts, and response strategies are developed by all participants. However, as this book discusses, CompStat and similar programs in many Latin American cities stop at politically uncomfortable questions about how statistics and human resources have been handled. The numbers produced by CompStat are publicized, but not followed up in the service of problem-oriented policing to identify chronic perpetrators, weapon types, and other information that can prevent future assaults rather than simply respond to prior ones. Unless such an information system is hooked into management, it becomes an electronic white elephant.

In 2000 former New York City police chief William J. Bratton developed a plan for the metropolitan police in the Federal District of Venezuela. Centered on model stations, small kiosks, and CompStat, the plan brought murder rates down from 236 between January and June 2000 to 132 for the same period in 2001, with a 30 percent fall in crimes in other areas.26 But structural weakness ultimately led to the plan’s undoing. Police chiefs in many areas had no control over their personnel, who often did not show up, making implementation difficult. A basic level of trust by residents was also lacking, and even the model stations and kiosks did not provide the hoped-for sense of security; indeed, several people were killed right in front of some them.27 The benefits of such programs come not from their technology, but from channeling updates on the causes of crime up and down the chain of command between policymakers and police stations. But as information is blocked and distorted along these channels, such technology is often adopted as a solution rather than as the instrument it is.

Corruption complicates efforts to tackle such management problems. It is universally cited in the citizen security literature, but its patterns, sources, and connections are understudied. As this book describes, there are three main levels or forms of police corruption. The most common are the innumerable petty acts by individual officers. Among the lower ranks who patrol traffic and the streets, unchecked corruption often supersedes policy and training as the main determinants of their schedules and community relations. A second higher-level form of corruption is that found in police agencies, most often perpetrated by those in charge of fines, issuance of identification cards, customs, or other areas involving goods or fees. As discussed in the following chapters, corrupt agencies affect the functioning of the entire criminal justice system. A third level of corruption is the money in “official” financial transactions that is embezzled in contracts, skimmed from budgets, or sliced out of other “routine” transfers. Such illicitly gained funds allow chiefs to loosen the strings to which the government budget is attached.

Anticorruption efforts to stem these and other harms usually focus on internal accountability. But often overlooked is how the larger political structure has allowed finances to become one of the least transparent but greatest sources of police power. Formally, in nearly every country the legislature and the executive (usually the finance minister) determine and approve the police budget; controllers, ombudsmen, and other units have some accounting oversight. In reality, though, the police propose a budget, and it is either approved or altered by a certain percentage, depending on the overall budget needs and the political power of the police at the time. Very little debate is held over the specific use of different budget items, thereby missing an important opportunity to review police efficacy, management, and internal power relations. As members of the legislatures in several countries acknowledge, scrutiny of police budgets gets lost amid the more urgent or controversial issues that accompany the yearly budget process.

Compounding this neglect is the inattention to the perennially unresolved debates over financial issues such as salaries. Low salaries are the most common explanation for at least the first and second levels of corruption. In most countries the lowest-ranking officers earn $200 a month and the highest-ranking about $750 (El Achkar forthcoming). Thus officers say that the current salaries would have to triple to provide an adequate income. But for several reasons governments are reluctant to raise them. For one thing, they believe (and not unreasonably) that low salaries are more than compensated for by bribery and other sources of funds. However, both scholarship (see Rose-Ackerman 1999) and experience counsel against simple salary increases as a solution. For example, to curb the high levels of corruption between 2004 and 2007, Tijuana, Mexico, raised the salaries of most municipal police officers to a rate higher than that of any of their colleagues around the country, but the increases did not put a dent in organized crime’s hold on the officers.

On another level, salaries already make up the bulk of police budgets in nearly every country, and so legislators are even more reluctant to increase them. However, because of mismanagement, salaries rarely provide a stable standard of living. For example, over 70 percent of Nicaragua’s police budget goes to salaries (Bautista Lara 2006, 71–73), and yet only in 2002 did it cover basic food costs, which inflation then drove up. At the same time police still lack adequate budgets for basic items such as equipment. The Ecuadorian police estimate that the equipment such as vehicles and munitions currently available is only 25 percent of what is needed and that the general budget is only half of what it should be (Policía Nacional de Ecuador 2003). Thus without more careful accounting, increases in salary are unlikely to provide the expected return in either individual well-being or collective effectiveness.

6. Information

Crime statistics are the starting point for citizen security policy, but throughout Latin America they are uneven, incomplete, and unreliable. As this book discusses, the reasons for such inadequacy include incomplete survey data, dismissal of noncriminal information, and the ways in which police and statistics agencies report, or fail to report, crimes. Often there are great discrepancies in reporting by different agencies, and most processes of collection and compilation are not particularly scientific. In smaller stations, attending officers frequently jot down reports in log books with little room for detail and without standardization. In Venezuela about one-third of all crimes are reported to the disparate offices of the judicial police, which lack a mechanism to consolidate and channel them, or even to include the crime reports received, managed, and retained by state and local agencies.

One of the biggest gaps in Latin American crime statistics is the cifra negra (black figure) of crimes not reported at all. In Mexico and Argentina this figure amounts to 70–75 percent of all crimes, between 60 and 80 percent in Venezuela, just over half in Nicaragua, and nearly two-thirds in Colombia, Guatemala, and El Salvador (Bautista Lara 2006, 122).28 However, these numbers may be as unreliable as crime statistics, because only good victimization surveys elicit citizen openness. Such unreliability is revealed by looking comparatively at the reported numbers. In Costa Rica, for example, police estimate a cifra negra of 70 percent, which is among the highest in the region and contrasts with the country’s relatively low rate of crime and good relations with society. Therefore, this figure probably reveals both the underestimates in other countries as well as Costa Rica’s relatively more honest efforts to report this rate accurately.

The crime-reporting problems are on clear display in table 2.1. Like the cifra negra, the more complete and consistent numbers from countries with lower crime rates such as Chile, Costa Rica, and Uruguay reveal better reporting. In Chile numbers vary because some years they include attempted as well as committed crimes. Crime reports in most other countries are characterized by unreliable or unreported statistics for many years since the 1990s. Of the numbers that are reported, there are often improbable jumps or drops from year to year. In Colombia and other countries, such changes are caused by reliance on different sources, with the police reporting lower numbers.

Documentation of particular crimes or criminality in general is only as useful as its reliability. Although crime reporting and tracking are improving, the resulting statistics still provide only the roughest of guides for crime policy. Throughout Latin America the lack of reliable crime statistics bedevils not just the police, but also congresses, ombudsmen, and other agencies trying to monitor and evaluate crime policy. As poor statistics extend into the criminal justice system, the management of victimization, detentions, and trials becomes more difficult.

Of all major crimes, the two reported most reliably are homicide, the most serious crime, and auto theft, reported for insurance purposes. Even though it is a partial and imperfect proxy for crime, homicide is used here to map general statistical reliability across countries—an approach adopted by other studies such as those by Fajnzylber, Lederman, and Loayza (2000) and Camara and Salama (2004). Appendix A lists Latin America’s homicide rates for each year between 1995 and 2009. It includes the lowest and highest reliable estimates provided by national police forces, judiciaries, government statistics agencies, independent observatories, and academic researchers. A quick review of this table reveals two major gaps. The first is the wide gap in numbers in many years; the second is the variations between concurrent years, with statistically improbable oscillations.

Table 2.1 National Crime Rates, 1997–2006 (per 100,000 persons) Country

Aside from chronic underreporting, these gaps reflect at least four discrepancies that expose the weakness of crime statistics in Latin America. First, numbers reported by international agencies often conflict with each other. For example, the Inter-American Development Bank (IADB) reports a homicide rate of 19.7 per 100,000 persons in 2000 in Brazil, whereas the Pan American Health Organization (PAHO) cites a rate of 29.3. In Honduras, Interpol reported a homicide rate of 154.2 per 100,000 persons in 1998, far above the rates of 40–60 reported in both the prior and following years by PAHO and that country’s police.

Second, estimates based on comparing raw figures with national populations, which nongovernmental sources are more likely to do, tend to be higher than those of national statistical bodies, indicating a pattern of official underreporting. In Costa Rica, for example, the press reported a homicide rate of 25.9 in 2001, whereas the state’s Organization of Judicial Investigation (Organización de Investigación Judicial, OIJ) said it was only 6.1. For 2004 the OIJ reported a homicide rate of 212, while the press reported 275.

Third, additional discrepancies arise from the categorization of homicides. Most countries report only homicidio doloso, intentional homicide, while others include forms of unintentional homicide, such as involuntary manslaughter, or the general category of “violent deaths,” which does not specify deaths that are not technically homicides. Some police distinguish between “homicide” and “murder.” Murder refers to the equivalent of first-degree homicide, and homicide refers to all other cases. Although both should be included in homicide rates, reporting only “homicide” lowers the final tally. In Venezuela killings of suspects while “resisting arrest” are not included (and it is a significant number, because thirty-nine suspects are killed for every officer death in “confrontations”). Also excluded are killings in prisons (more than one per day) or other violent deaths that go uncategorized.29 Statistical utility is also diminished when other forms of homicide are not separated out, such as when killings by police or criminal suspects are folded into the final tally—a number that accounts for up to one in five homicides in countries such as the Dominican Republic. The extreme swings and questionable variations in crime reports that result from this methodological sloppiness, such as the Dominican Republic’s report of a 100 percent increase in youth crime (Bobea 2003, 194), demonstrate their lack of reliability.

A fourth discrepancy is among criminal justice agencies. In Brazil the homicide rates reported each year by the civil and military police differ almost continually by two hundred homicides (Ribeiro 2007). In El Salvador until 2004 there were consistent gaps of between 20 and 40 per 100,000 persons in the homicide rates reported by the National Civil Police (Policía Nacional Civil, PNC), the Attorney General of the Republic (Fiscalía General de la República, FGR), and the Institute of Forensic Medicine (Instituto de Medicina Legal, IML)—see Pleitez Chávez (2006, 13) and Savenije and Van der Borgh (2004, 157).30 But in 2005 these three agencies formed the Technical Homicide Board (Mesa Técnica de Homicidios), which, together with outside agencies such as the Observatorio Centroamericano sobre Violencia (OCAVI), has verified and consolidated reports. As a result, differences for 2005 narrowed considerably, from police estimates of 47.731 to OCAVI’s estimates of 55.0 per 100,000 persons.

Although not all countries adopted such official changes in statistics collection, the narrowing gap in reported numbers does indicate improved coordination and accuracy in general. Studies that examine and document trends over time generally report more consistent numbers. PAHO’s estimations of Costa Rican homicides, for example, were in the range of 7.0–8.0 between 1996 and 2005. Statistics also tend to be most reliable in urban areas when collected by groups such as the UN-funded Violence Observatories or by a consortium of agencies (see chapter 7). In Quito, for example, information is compiled by an Interinstitutional Technical Committee composed of the police, prosecutor, National Gender Office, public hospitals, and academic institutions. Overall, the more consistent statistics are from the region’s stronger states, such as Argentina, Chile, Costa Rica, and Uruguay, where homicide rates approximate those of North America. But the high-crime countries are in the majority in the region, and thus that is where clarification is most needed.

Table 2.2 Average Annual Gap between Lower and Upper Homicide Rate Estimates, 1995–2009


Average annual gap











Costa Rica


Dominican Republic




El Salvador




















Source: See appendix A.

Reflecting these discrepancies, table 2.2 uses the numbers detailed in appendix A to provide the average yearly gap in homicide statistics between 1995 and 2009. Such averages, of course, are a very crude reflection of statistical gaps. Many reports do not cite the source or methodology of their numbers, and yet they are repeated in the media and other studies. As one government ministry official in Bolivia said, his ministry tends to accept any reasonable number because it lacks the ability to evaluate them independently. When the gap is particularly high, as it was for many years in Venezuela, the higher numbers easily become ammunition for those proposing crackdowns. So even when a reported rate is almost certainly too high, it can still be wielded politically. In addition, the average gap in many countries is artificially narrowed in years in which only one reliable statistic is reported, and is exaggerated by outlying numbers far larger than in other reports. But, overall, these averages reveal inherent flaws in crime compilation and reporting in Latin America. In many cases, the average gap exceeds the lower estimate of homicides, affirming that crime statistics should be approached with suspicion.

Because of these problems, countries depend less and less on statistics as a basis for policy, depriving it of an important foundation. Even in Costa Rica’s relatively well-supported security structure, the lack of a modern system of information prevents the formation of consistent policies.32 For example, the 2003–7 plan of the Ecuadorian national police reported a 5.79 percent increase in crime reports but a 5.71 percent decrease in detentions without explaining what accounted for the contrary pattern or how to address it (Policía Nacional de Ecuador 2003).33 Civilian policymakers, who are just as uncertain about how to critically apply crime statistics to policy, if not more so, often selectively use numbers when approving of stronger policing. Ironically, though, giving the police more power after crime rates rise may reduce their incentives to be more effective crime fighters. Furthermore, if a crime statistic as central and basic as the national homicide rate is unreliable, then so are the abilities of reformers to gather their arguments for change, of legislators to know which areas need emphasis, and of officials to measure progress. Without consistently dependable criminal justice numbers, public officials can find it difficult to know why reforms in community policing and oversight reforms are needed, what they need to address, and whether they are reaching their goals.

7. Criminal Justice: From Crime Scenes to Prisons

Much of the information needed by the judiciary is contained in the evidence and testimony of witnesses uncovered during a criminal investigation. Investigation is the most important task carried out by the police because it is the primary way in which they track down criminals, secure convictions, identify patterns of violence, and prevent future offenses. But officers throughout Latin America are inadequately trained in investigation, and so they haphazardly carry out the steps needed to secure evidence and witnesses such as crime scene protection, coordinated forensics, and cooperative relations with the public. To make up for their lack of preparation, funding, and societal trust, police officers resort to blunt and ineffective tactics such as mass roundups (razzias) in poor neighborhoods (barrios), witness intimidation, indiscriminate detentions, and armed “confrontations” that end in the suspects’ deaths. For example, although Venezuelan law clearly states that the police may take only an “informative” declaration from detainees and that fiscales must monitor detainee rights in police stations, in practice the police take many declarations in station basements without observation.34 Even though such tactics may “resolve” or rustle up enough material for a particular case, each time they sacrifice a bit of future effectiveness.

Coming at the beginning of the investigative process, such tactics also taint the rest of it. Although reform packages are placing a greater emphasis on technical training in criminal investigation, nearly every country still allots far more material and human resources to preventive policing than to investigative policing. In fact, no judicial police budget receives more than 5–10 percent of the total security allocation. A case in point is Venezuela’s CICPC, whose budget has barely risen since 1999. Its Homicide Division has just a single ballistic analysis team, and the agency says it needs to triple its ten thousand officials.35 In the United States about 15 percent of police officers carry out investigations—a low number, but far higher than that in most Latin American countries, which hovers between 2 and 10 percent (Bayley 1998, 71).36 Not only are the numbers of investigative police low, but they are undermined by competition from the preventive police forces, a lack of forensic and technical support, and chronic judicial delays.

The investigative incapacity of the police is often matched by that of the judiciary—a fact readily used by the police to redirect blame for insecurity and impunity, particularly when the courts manage to fall below the police in public opinion polls. But like the police, criminal justice officials such as judges, prosecutors, and public defenders lack the budget, training, and autonomy to carry out their jobs fairly and efficiently. Meanwhile, their efforts to overcome common institutional constraints or assess important areas such as rehabilitation and alternative sentencing are drowned out by the technical and politicized accusations that ricochet around the criminal justice system. For example, when police grouse about the release of a detainee for “technical” reasons, such “clemency” usually stems less from a judge’s leniency than from other factors such as prison overcrowding or a lack of appropriate treatment programs.

New penal process codes, the cornerstone of criminal justice reform in contemporary Latin America, have the potential to overcome such problems. In many countries the codes have produced impressive results—but primarily at the beginning. In Nicaragua citizen crime reports rose by 4 percent and police detention dropped by 30 percent in the first year after the new code went into effect in 2004. Along with new conditional release laws, Venezuela’s new code led to a nearly 40 percent drop in the prison population between mid-1999 and mid-2000. As criminal justice officials took advantage of alternative sentencing to reduce backlogs, similar results were seen in other countries. But as time has passed and crime has continued its relentless rise, the codes have become increasingly exposed politically. In most countries they have become a convenient target and are blamed for fostering crime and impunity by protecting criminals’ rights (see Sieder 2004; Rocha Gómez 2007). Grappling with the demands of the new codes, many officials are only too happy to encourage that view. By introducing new procedures, guarantees, and courts—particularly oral trials—these new codes, by definition, mean an upgrade of the entire criminal justice system. But they are not supported by the funds or training needed to change practices and convince officials to come on board. In Ecuador the new code came into force in 2001, but the government did not form any coordinating agency to oversee its implementation or provide the funds needed for the greatly expanded prosecutor. In Guatemala budgetary constraints have limited language interpretation, which the new code requires, while numerous pretrial motions violate the provision that the criminal dossiers (sumarios), which are the complete file of relevant information prepared for a trial, be completed “as soon as possible.” In the Dominican Republic the sumario is limited to sixty days, but it can be and often is extended as many times as deemed necessary. In other countries new codes lack guarantees for offenses that are “minor” but nevertheless involve extended incarceration without access to counsel.

More attention to the implementation of these codes by criminal justice officials, as well as more citizen support for them, would help to solve these problems. For the police, new codes mean stricter controls on evidence collection, detention, questioning, and level of proof. Not only are police officers under greater scrutiny—a perceived affront to their long-standing autonomy—but their investigative authorities have been given to fiscales, who, in turn, have had to learn to run criminal investigations. In Venezuela and Bolivia police accuse fiscales of continuing to protect defendants’ rights, as they did under the old codes, instead of prosecuting them, which is their responsibility under the new code.37 Judges appointed in new types of courts face a particularly steep learning curve, but many of them (as police and prosecutors are quick to point out) are ill-prepared for their new responsibilities. Venezuela adopted a new penal code in 1998. The Organic Penal Process Code (Código Orgánico Procesal Penal, COPP) goes further than most other new codes. For example, it allows both judges and citizens (escabinos) to head most criminal trials.38 But good design has not forestalled poor implementation. Saying that “we speak different languages,” police inspectors accuse prosecutors of lacking both the will and the training to carry out the COPP. And although the fiscales have formulated criteria to implement the COPP,39 delays and abuses continue in both the MP and the police. Many Venezuelan police agents continue to withhold information from judges, use false witnesses, and tamper with evidence,40 often justifying such actions with their belief that judges free criminals and that public defenders hide incriminating evidence.

But police continue to blame the COPP rather than their own weaknesses—a position enthusiastically taken up by politicians and the media. In the first year after its approval, Venezuela’s new code was subject to a barrage of alarming articles. “The COPP makes police action difficult,” admonished one headline. “The police do not detain a criminal because of the COPP,” screamed another.41 A year later, the chief of the CICPC said he attributed the rise in crime to “three fundamental factors … the economic crisis, the political transition, and the application of the COPP.”42 And a year after that, the interior minister blamed “high rates of social violence … in large part on the COPP.” The police of Vargas state, which has some of the country’s most violent barrios, complained that most of their detainees were being released because of code protections.43 It was certainly true that far more detainees were being released because of the code’s provisions, such as an emphasis on the presumption of innocence, but studies comparing detention and crime rates demonstrated no impact on crime in Venezuela. And in fact the number of resolved cases jumped considerably, from just over 86,000 in 1999 to nearly 179,000 in 2003 (Chen, Leonte, and Lopez 2008). Nevertheless, around the region blame is commonly placed on the penal process codes. One of the five main security problems in Ecuador, according to its police, is the “implementation of the new Penal Process Code, which permits detainees of police operation to be put in liberty immediately by the judicial authorities” (Policía Nacional de Ecuador 2006).

Execution of the new codes is further restrained by the limited number of criminal justice officials. Throughout Latin America public defenders have far too many cases to provide adequate defense for the great majority of them. Most criminal justice officials—particularly public defenders them-selves—recommend a maximum of 150 cases, especially because most Latin American countries make an adequate legal defense a constitutional right and because public defenders represent about 80 percent of detainees in most countries. By all measures Costa Rica comes closest to that number, but is still far from reaching it.44 In most other countries public defenders estimate that they handle over 320 cases at one time—and those in juvenile justice well over 400. And these caseloads will only grow because the number of penal cases is rising about 25 percent each year.45 Public defenders not only take on far too many cases to deal with most of them effectively, but also are further slowed by a lack of clarity about “obligations and duties,”46 are paid less than most other lawyers, often have little job security, and are subject to little oversight. In El Salvador higher salaries have created a bigger public defender corps, but working conditions remain poor. Argentina hires “adjunct” defenders (defensores) who may weaken due process when up against more experienced fiscales. In Rio de Janeiro an estimated 60–70 percent of the small-time drug dealers in prison would not be there if they had better lawyers,47 and in São Paulo State poor public defense means that only a fraction of offenders benefit from the 1998 alternative sentencing law. Countries are beginning to hire more public defenders, restructure their agencies, and give them more autonomy.48 But they have a long way to go before public defense meets its constitutional obligations.

These deficiencies contribute to a low percentage of crimes investigated and a low percentage of arrests—between 2 and 5 percent—that lead to conviction. As discussed in this and other chapters, if suspects are not identified quickly there is little chance that the crimes will be solved. In El Salvador fewer than one in ten violent crimes are investigated (Programa Centroamericana de la Federación Luterana Mundial 2005).49 In Mexico less than 5 percent of crimes and 7 percent of homicides lead to a conviction.50 In Ecuador only twenty of ninety reported homicides in 2003 led to a detention, and an average of sixty persons were arrested for the 750 crimes that occurred each day (Arroba 2003; Pontón 2005, 369). In Venezuela only 36 percent cases of reported crime were closed, and only 2.7 percent ended in a court sentence (Sanjuán 2003). When crimes are broken down by geography and gender, the rates of crime resolution can fall to especially low levels. In Costa Rica 80 percent of rape cases do not go to trial—even though the victims must often describe the attacks on seven different occasions—and the courts hand down convictions in only half of the 20 percent of cases that they do hear.51 In Guatemala, which has alarming rates of feminicide, only 5 of the 1,897 killings of women between 2001 and 2005 resulted in a court conviction.52 In addition, as discussed in other chapters the rates of homicide solution tend to be lower in rural areas than in urban areas. Although the average rate of solution in the Ecuadorian city of Guayaquil is 30 percent, for example, in many rural provinces it is less than 10 percent.

Such judicial ineffectiveness is also weighed down by other problems discussed in this chapter such as criminal statistics. In Costa Rica the OIJ, the investigative wing of the police, exposed big gaps in judicial functioning through a criminological study in San José that connected a majority of incidents to a specific downtown area and a particular group of generally young offenders. In the process the OIJ discovered that nearly half of those arrested had been detained previously and that nearly half of the police’s crime reports on these cases had informational inadequacies that led judges to throw them out. In many cases the authorities neglected to take notice of the fact that those who were arrested for one crime were wanted in connection with earlier crimes (Programa Estado de la Nación 2004). Costa Rican officials, who estimate that from one-third to one-half of murders are not solved,53 respond to such conditions by complaining of poor cooperation between the preventive Public Force and OIJ detectives.

Finally, throughout the region the failures of the criminal justice process are most evident in its final stage: prison. Because the police rely on edicts and because of their “abuse of preventive detention,” most people are imprisoned for property- or drug-related crimes.54 In fact, the Inter-American Drug Abuse Control Commission (CICAD) of the Organization of American States (OAS) estimates that up to 80 percent of detentions are drug-related.55 Because suspects are locked up much faster than the courts can process them, a regional average of about 60 percent of detainees in Latin America have not been tried at any given time (see table 2.3), and many wait for trial beyond the two-year limit specified in the regional Pact of San José. Prisons are woefully underfunded, from basic needs such as medicine to rehabilitation services and renovation of dilapidated facilities, and thus the region’s penitentiary systems have become inhumanely violent and overcrowded. With the exception of Chile and Mexico, in Latin America prison violence is responsible for hundreds of prisoner deaths each year, with mass riots almost an annual occurrence. Indeed, Latin America has been the site of the most deadly incidents of prison violence in the modern era.

Prison officials say on the record that such conditions result from a shortage of money, but off the record they point to a shortage of political will. The two are, however, closely linked: officials do not fund or back up policies considered soft on crime or criminals. Even alternative sentencing, such as community service for nonviolent or first-time offenders, is often passed over by judges who opt for imprisonment. Indeed, up to 70 percent of cases in Western Europe and North America end with alternative sentences, but in some Latin American countries such as Brazil less than 5 percent of cases do. Venezuela’s Assistance Centers, which have helped nearly 130,000 persons and kept recidivism below 4 percent, are underfunded and inconsistently applied (Elia de Molina 1992). In most countries many detainees, such as those held for drug violations, are excluded from bail. Meanwhile, lack of personnel and poor resocialization programs prevent the majority of those eligible for release from being freed. In many countries the “principle of opportunity,” in which the prosecutor can offer an alternative to incarceration, is also underused. Thus imprisonment—and all the inefficiencies and abuses that ensue—remains the primary end point of the long criminal justice processes described in this book. Like homo sacer, a subject under Roman law who committed a crime that put him beyond legal protection (Agamben 1998), imprisoned criminals are both controlled by the state but beyond the law, and they are in facilities whose practices counter their officially rehabilitative aims. The prisons have also become incubators of crime, offering detainees an opportunity to become more integrated into organized networks, many of which operate freely from within prisons. Thus although imprisonment in Latin America may be the ultimate blow of the iron fist, it also belies the claim of that approach to reduce crime.

The judiciary’s role extends beyond specific cases. As in most democracies, the courts have “police power” to rule on the constitutionality and application of police laws and actions. On issues ranging from state power to civil rights, a history of executive dominance and police violence makes this role critical. Although nominally nonpolitical, the judiciary is also highly susceptible to political pressures, which in much of Latin America extend from judges’ nominations to constitutional jurisprudence. Backroom party deals in the selection of judges (despite the initial selection of candidates by judicial councils), the use of temporary judges vulnerable to professional reprimand and dismissals, and small judicial budgets (between 2 and 5 percent of the total budget in most countries) also keep the courts off balance and the channels of pressure wide. As for citizen security, court officials acknowledge that the direct and indirect pressure to be tough on crime is most evident in judges’ tendencies to choose incarceration over alternative sentencing in cases of violent crime.

Table 2.3 Prison Populations in Latin America, by Country and Category


Such pressure is also seen in how closely courts hew to a wide interpretation of police power. In contrast to unequivocal court action against abusive authoritarian-era laws are the region’s rulings on demonstrably unconstitutional mano dura laws. In upholding this legislation, judges tap into a long history of jurisprudence. Since its first use in the 1827 U.S. Supreme Court case Brown v. State of Maryland as the authority to limit citizen actions for the good of social order, the term police power has been interpreted along a spectrum ranging from narrow conceptualizations limited to specific actions to broad conceptualizations with an all-encompassing promotion of the general good.56 The broad approach has predominated in Latin America’s history, and it has been used to legalize actions such as police control of civil society groups. In 1922 Argentina’s Supreme Court extended police jurisdiction to the “general good,” including economic matters, and in 1950 it said that the power of the police is “justified by the necessity of the defense and strengthening of morality, health, collective fitness, and the community’s economic interests.”57 While less overarching, such legal leaning continues on both police actions and the power of police agencies. Ecuador’s judiciary has given special police courts the jurisdiction to try police for offenses not committed as part of their duties, giving precedence to the Police Code over the national constitution. In 2001 Venezuela’s Supreme Court ruled that the armed forces can be used to help guarantee public order. In addition, although the “delegation” of legislative powers to the executive is unconstitutional, a “general delegation” of police power to the executive is often allowed in situations such as emergencies or in the absence of needed norms.

In addition to cutting off a critical legal channel, such jurisprudence reveals ambivalence in democracies’ relationships with the law. The state carries out the law but is also bound by it, deriving strength from the consistency and predictability that laws engender, but limited by the oversight and accountability they demand. This balance is most tenuous in times of uncertainty or transition, when state agencies often resist changes in policy, personnel, or government. Even in democratic regimes the daily functions of state agencies are often more suited and accustomed to less constitutional but more efficient internal regulations such as on detentions. But because the law is the only legitimate basis of state action in a democracy, state agencies learn to justify or couch their actions with it. For example, joint police-military raids in poor neighborhoods or searches without a warrant are often retroactively legitimized through “emergency” decrees. On policing, the law may sometimes lead not to a reduction in undemocratic practices but a greater aptitude for hiding them. Instead of simply responding to expansions of police power, judiciaries are supposed to restrict them by delineating the range of acceptable actions by state agencies. For state practices in which monitoring is difficult, that means drawing the tightest legal circle possible around officials’ discretion. As discussed in other chapters, because officers must make so many decisions in the course of their work without oversight, discretion is an integral part of policing. Similar to “the hole in a doughnut,” discretion “does not exist except as an area left open by a surrounding belt of restriction (Dworkin 1977, 31). As Latin America’s courts continue to stick to positivist interpretations that justify law as it is written and overlook its normative intentions, with rulings that rarely venture beyond specific cases to the laws and policies behind them, they loosen that belt of restriction and allow other agencies to expand the discretionary center through exceptions such as curfews and military police actions.

8. Police-Society Relations

It is not surprising that most Latin Americans mistrust their police. In all but four of the region’s countries at least two-thirds of the population has little or no confidence in the police. In the other four countries, only about half the population expresses trust in the police.58 “We just don’t see the police do anything,” said one resident of the Bolivian city of Sucre. “Crime is not as high here, but they seem to use that to excuse their inaction.” Such remarks indicate that the public’s dim view of the police goes beyond catching criminals. It is also about the exercise of power in daily policing and how it affects police-society relations. Many complaints by citizens are based on what they regard as a lack of respect and degrading words. At times, such attitudes extend to specific violations such as the failure to ask permission before entering private property. More generally, people suspect any reforms to be Trojan horses to give the police even more power. Indeed, worldwide community policing is equated with “community spying” (Brogden and Nijhar 2005, 14). So even when policy improves and such improvement is recognized by society, building citizen trust also requires changes in the innumerable types of contacts that comprise daily policing.

This relationship goes both ways, with citizen distrust and resentment usually reciprocated fully by the police. Feeling that the complexity and danger of their work are not appreciated and the strain on their personal lives is not acknowledged, many officers begin to see civilians as unsympathetic, unaware, and even aggressive. When police chiefs continue or adopt militaristic and siege mentalities (see Skolnick and Fyfe 1993), the rank-and-file officers are even more likely to believe, as do many in Venezuela, that society sees them “as an enemy.”59 Much of daily policing operates through this prism. In explaining violent contacts with civilians, for example, many police claim that they detected “suspicious attitudes” or had experienced some form of aggression. Such reactions often come as a surprise to the involved civilians when brought up in arbitration forums, heightening tensions and stereotypes. But ironically most distrust between citizens and police arises from the same two suspicions: protection of criminals and selectivity in responses to insecurity. Both the police and citizens regard each other as collaborating with criminals and only willing to cooperate when it serves their own interests. “Some police are good and help solve problems,” said one shopkeeper in the colonia of San Miguel in Tegucigalpa, “but overall they do not do anything besides protect their own interests.”

Not only are such views more likely for police agencies (many state agencies operate unperturbed by public distrust or even hostility), but they also can severely undercut their work, mainly because the majority of crime suspects and evidence is brought in by the public. Investigators do not locate suspects based on evidence, as commonly believed, but in practice round up suspects in order to gather the evidence needed to prosecute. Thus the need for cooperation from both detainees and the public at large makes it more difficult and more important to move toward problem-oriented policing. Because cooperation is regarded by most people as risky or a form of collusion, police officers are, in turn, less inclined to promote the kind of societal relations that would reduce reliance on abusive information-gathering tactics.

In addition to harming investigation, such conditions deprive community policing programs (the main channel for problem-oriented policing) of the uncensored dialogue and cooperative policy formulation they require. These conditions also weaken the abilities of these programs to avoid being pulled down into local power relations. The Venezuelan city of Maracaibo created hundreds of Neighborhood Security Committees to support the police, but they largely lack the oversight needed to avoid entanglements with that high-crime city’s deep social divisions. Like many state-sponsored reforms, citizen-based security initiatives are often hobbled by the citizens themselves. Neighborhood councils and patrols become co-opted by neighborhood commissioners, by drug traffickers, or by program directors who channel funds to their friends. But more common than domineering neighborhood organizations is the lack of them. Most communities are deficient in the cohesion, finances, and experience needed to form durable groups that can consistently identify the causes of insecurity—much less break down the fear, distrust, and violence that characterize their relationships with the police. As economic trends continue filling poor urban areas with newcomers, even established neighborhoods have difficulty maintaining community organizations, in turn limiting the knowledge and information that reach the police.

Conclusion and a Look Ahead

Interactions among the realms of change outlined in this chapter deepen Latin America’s citizen security crisis by multiplying the difficulties of reform. For example, a politicized promotion process helps perpetuate bad management, and weak internal affairs units enable police stonewalling. More broadly, societal unrest or political uncertainty can be used by police to drive a wedge between reformers and executive officials giving priority to short-term stability. Even community policing, designed to avoid all of these traps, can become snared in them because of a lack of agency restructuring and training. With no shortage of police, politicians, and journalists ready to pounce on the weakness in any reform, the failure of a specific program can result in repudiation of the entire problem-oriented approach.

Moreover, as discussed in the next chapter, these realms of change are part of the larger picture of contemporary democracy in Latin America and other regions. Patterns of political power, geographic differences, and societal fragmentation have deep roots in history, many of which democratization accentuates. As a result, problematic political, institutional, legal, and societal conditions are not as amenable to change as reformers often anticipate. But by laying out the connections and joint consequences of these obstacles to change, this chapter and the next outline a framework for understanding how to overcome them. On this basis, the case studies and concluding chapters that follow show how the citizen security crisis can be more precisely addressed in both theory and policy.

1. Jorge Sapoznikow, Inter-American Development Bank, “Expertos debaten prevención de violencia juvenil en América Central, Panamá y República Dominicana,” press release, October 6, 2006.

2. The Santiago-based Latinobarómetro survey polled 20,212 people in eighteen countries between September 7 and October 8, 2007. The average margin of error was 3 percent.

3. Gabriel Picciano, secretary of administrative management, Ministry of Justice, Security, and Human Rights (Ministerio de Justicia, Seguridad y Derechos Humanos, MJSDH), interview with author, Buenos Aires, May 20, 2004.

4. Betty Brannan Jaén, “Mano dura, una mentalidad medieval,” La Prensa (Panama City), August 1, 2004.

5. “Realizarán encuestas para armar mapa del delito en Santa Fe,” El Santafesino, May 25, 2008.

6. The term national security has become so common that many police officials now criticize its use in training.

7. “Sistema de drogas y reciclaje de químicos eran los funciones del cuerpo parapolicial,” El Nacional, February 13, 1988, D-22. The involvement of the Defense Ministry’s Office of Military Intelligence (Dirección de Inteligencia Militar, DIM) and the Technical and Judicial Police (Policia Tecnica y Judicial, PTJ) was suspected.

8. The opposition complains of the repression by police controlled by Chávez allies and by the pro-Chávez Bolivarian Circles (Círculos Bolivarianos), while government allies criticize the Metropolitan Police (Policía Metropolitana, PM) and other opposition-controlled police agencies. One legislator accused the PM of being “criminal and repressive” and accused the police of Chacao and of Baruta of protecting the opposition but killing pro-Chávez activists. William Tarek Saab, member of legislature, Movimiento Quinta República, interview by author, Caracas, February 27, 2003.

9. Many areas of law deal with the police. Disciplinary law delineates police responsibility, administrative law defines sanction authorities and processes, and penal law regulates constitutional order and internal state security.

10. “Asesinadas 30 personas en el mes de Octubre,” Últimas Noticias, November 7, 2002, 20.

11. El Universal, April 20, 2002, I-4.

12. U.S. Federal Bureau of Investigation, Uniform Crime Reports,

13. Most of these misdemeanors carry a fine, and the ruling was based on the argument that because a fine is a form of debt to the state and because the constitution prohibits incarceration for debts, detentions for misdemeanors are illegal.

14. Escuela de Altos Estudios Policiales, Policía Nacional de Perú (Advanced School of Police Studies, National Police of Peru), interview by author, Lima, June 18, 2001.

15. Celso Rodríguez, director, Escuela Nacional de Policía, interview by author, Montevideo, August 20, 2003.

16. Inspectors and subinspectors receive up to four years of education, but most Metropolitan Police officers receive only three months.

17. Police officials, interviews by author, San José, Costa Rica, June 18–30, 2006.

18. Venezuela has an average of 457 officers per 100,000 inhabitants, but in seventeen of twenty-four states the number of officers is below the standard of 350 per 100,000 inhabitants (El Achkar forthcoming).

19. In early 2001 six hundred GN officers were deployed in key parts of the capital, in part to take over many of the areas not being covered by striking local police. Using checkpoints at key intersections—and seizing massive amounts of narcotics and large numbers of vehicles, crime suspects, and weapons—they lowered the crime rate in the first six months of 2001 by 16 percent. “Hubo sólo 5 homicidios en áreas cuidadas por el GN,” El Nacional, July 18, 2001.

20. Associated Press, “Hundreds Protest Violent Crime in Venezuela,” June 21, 2009.

21. “El avance fué muy escaso,” El Universal, July 23, 2001, 4–1. By investing 33 percent of municipal income in security during the 1990s, the mayor of Baruta municipality was able to reduce its crime rate by nearly a third.

22. Commissioner, subcommissioner, and two subcommissioned officers, PFA Comisaría 32, interviews by author, Buenos Aires, May 25, 2004; Mittrany (2008).

23. With their average seventy-hour workweek, most Venezuelan police officers work 40 percent more than other state employees. Rafael Rivero M., “Cuánto y qué gana un policía,” Diario de Caracas, June 14, 1990, 20.

24. Sewell Chan, “Counting Heads along the Thin Blue Line,” New York Times, March 26, 2006.

25. In Guatemala such use is blamed for the police force’s scarce coverage of national territory (Sequén-Monchéz 2003, 147).

26. William J. Bratton and William Andrews. “Driving Out the Crime Wave: The Police Methods that Worked in New York City Can Work in Latin America,” Time, July 23, 2001.

27. “Roban y matan a taxista frente a módulo policia,” El Universal, July 4, 2001.

28. In Venezuela the figure of 80 percent is based on a poll of ten thousand Venezuelans in which only 11 percent of those who said they were crime victims reported the crimes (Sanjuán 2003, 120). The estimate of 60 percent is from the Venezuelan Violence Observatory (Observatorio Venezolano de Violencia 2007).

29. “Deadly Message,” The Economist, July 19, 2008, 47.

30. In 2004, for example, the police reported 2,768 homicides and the fiscal reported 3,897 (OCAVI 2009).

31. Reported in Observatorio de la Violencia (2007).

32. Carlos Mora, “Costa Rica atemorizada por el hampa,” La Nación, April 17, 2006.

33. Policía Nacional de Ecuador, Plan estratégico para el Quinquenio 2003–2007, Quito.

34. Carolina Oliva, liaison with legislature, Office of the Attorney General, interview by author, Caracas, April 6, 1995.

35. Jerssen Mojica, inspector, CICPC, interview by author, Caracas, February 26, 2003.

36. About 2 percent of all Venezuelan police officials work in investigation, and the rest carry out patrols or administrative tasks. Victor Amram Lazes, commissioner, PTJ, interview by author, Caracas, July 2, 1998.

37. Mojica, interview.

38. Luis Enrique Oberto, president, Comissión Legislativa, interview by author, Caracas, June 29, 1998; IADB (1998, 5–15).

39. Dilia Parra, director, Instituto de Estudios Superiores, Ministerio Público (Institute of Superior Studies of the Attorney General), interview by author, Caracas, February 25, 2003; Omar Jiménez, prosecutor 18, Guarenas, interview by author, Caracas, February 26, 2003.

40. “La Ley Antidroga es una Patente de Corso de Jueces y Policías,” El Nacional, September 9, 1988, 12; “PTJ: Un cuerpo que vive entre la enfermedad y la depuración,” Diario de Caracas, June 21, 1989, 18.

41. First headline, El Universal, July 3, 1999; second headline, El Universal, July 4, 1999.

42. El Universal, February 2, 2000,

43. “Asesinadas 30 personas en el mes de Octubre,” Últimas Noticias, November 7, 2002, 20.

44. In 1979 each Venezuelan defender (defensor) handled an average of under seventy cases; by 1995 the average had shot up to over three hundred (Dirección de Planificación, Consejo de la Judicatura, 1994). Rarely can government and private legal aid centers make up for such inadequacies. Free legal clinics are underfunded, have high turnover rates, and are concentrated in capital cities.

45. Montserrat Solano C., “Causas asfixian a la defensa pública,” El País, San José de Costa Rica, November 17, 2001.

46. María Antioneta Acuñade, president, Asociación de Defensores Públicos de Venezuela (Public Defenders Association of Venezuela), interview by author, Caracas, May 22, 1995. The institutional affiliations of public defense agencies vary; they may be in the judiciary, in the executive, or on their own.

47. Julita Lemgruber, ex-chief, state prisons, interview by author, Rio de Janeiro, February 3, 2005; Fraser (2001),

48. Argentina’s Act 24.946 of 1998 gave defensores more organizational autonomy, created a professional career law, and established mobile defenders (defensores volantes) to coordinate the many levels of defenders’ work. María López Puleio, attorney coordinator, Defensoría General de la Nación, interview by author, Buenos Aires, December 17, 1996.

49. The crimes covered are homicide and robbery.

50. Centro de Investigación para el Desarrollo, “El subsistema de seguridad ciudadana y justicia penal, y la procuración,” CIDAC, June 9, 2003,; “Critical Threat,” The Economist, June 15, 2002, 36.

51. Carlos Arguedas, “80 percent de denuncias por el delito de violación no llegan a juicio,” La Nación, July 3, 2005.

52. “Impunity Rules,” The Economist, November 18, 2006, 40–41.

53. The number of unsolved homicides grew from 28 of 184 in 1995 to 78 of 212 in 2004 (or of 275, depending on the source of homicide statistics). Estimates of 37 percent are given by Álvaro Sánchez Córdoba, “Homicidios son más complicados,” Al Día, April 12, 2005, 4; and reports of up to 50 percent by Otto Vargas M., La Nación, December 28, 2004, 10.

54. Ingeniero Freixas, ombudsman, Argentina’s federal prisons, interview by author, Buenos Aires, November 3, 1994.

55. Inter-American Drug Abuse Control Commission, Organization of American States, interviews by author, Washington, DC, February 17, 2005.

56. Brown v. Maryland, 12 Wheaton 419 (1827).

57. Ercolano c/Lanteri de Renshaw (Jurisprudencia Argentina, 1922, 136:170). The broad approach was dominant in the United States until 1877. See Aron Rabionovich (Jurisprudencia Argentina, 1950, 217:469). By contrast, most Argentine scholars support the narrow interpretation, citing the dissenting opinion in Renshaw. See Anglo c/Gobierno Nacional (Jurisprudencia Argentina, 1934, 171:366) as an example of wide police powers.

58. Latinobarómetro 2003. The results indicate the percentage of valid responses of “little or no trust” to the question “Would you say that you trust the police a lot, a little or not at all?” The countries in which about half of the people trust the police are Chile, 51 percent; Colombia and Costa Rica, 55 percent; and Uruguay, 49 percent. Other country results: Argentina, 77 percent; Brazil, 63 percent; Costa Rica, 55 percent; El Salvador, 64 percent; Guatemala, 78 percent; Honduras, 67 percent; Mexico, 83 percent; Nicaragua, 62 percent; Panama, 63 percent; Paraguay, 76 percent; Peru, 77 percent; and Venezuela, 78 percent. In Peru 53.1 percent of respondents in a 2007 Lima-area poll disapproved of the police (Instituto de Defensa Legal 2007), and less than one-third of respondents in other polls say they have a lot or some confidence in the police (Mendoza, Hugo, and Neild 2007).

59. Valmore Leegos, CICPC detective, interview by author, Caracas, February 26, 2003.

Previous Chapter

1. Introduction

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.