In the area of public security and policing, Mexico faces a painful dilemma, one which appears not to admit of a solution. Its police forces utterly lack public trust. Citizens betray a palpable fear that contact with the police will more likely cost them some of their rights (or goods) than preserve and protect them. In short, reform of the police is needed, urgently.

At the same time, the nation finds itself in the midst of the most troubling period of criminality in its modem history. The common perception that crime is more common, better organized, more dangerous and generally more out of control than before only exacerbates the enormity of the challenge to meet the multiple challenges involved in vetting, retraining, managing, and radically upgrading police forces at every level and across the nation. For Mexicans, this is attempting to repair or replace a faulty parachute while free falling. Faced with dual emergencies, politicians run a great risk of seizing upon an approach that is too fast and not carefully considered, as a means of placating a public clamoring for action. That police reform is a new endeavor and local police experts scarce only enhances the potential for hazard.

This document attempts to offer a lens through which to view this challenge, and a few recommendations for public officials to consider. Their focus is as much on initiating the processes for change as the specific reforms themselves. Much of these draw on the lessons of experiences elsewhere, because they are relevant and because Mexico's own lack of experience with democratic, responsive policing obligates its leaders to create a new tradition, supported by ideas adapted to local realities.

Police reform: a proposed view of the underlying challenge

As has been noted by many observers, policing is set apart from many other aspects of the criminal justice process (and indeed many other occupations) by two central characteristics. Analysis of the potential for police abuse (or inefficiency or corruption) typically begins with the wide discretion that officers continually exercise, and the low visibility of their daily decisions. These are distinct from the often mentioned police monopoly on the application of force, but concern over the combination of great discretion and little visibility is obviously amplified by the high stakes that are ever present when state agents wield the right to detain, subdue, even kill.

For comparison's sake, consider the work of a criminal court judge. Working within a normative framework which governs substantive matters and indeed, constitutes a formal process for assimilating facts and applying the law -- judges are informed as well by the input and scrutiny that various collateral actors (prosecutors, defense counsel, appellate courts, to name a few) are obligated to provide as part of their own functions. A judge's decision is the culmination of formal proceedings, arguments, and (one hopes) deliberate reflection upon the proof and the law, with a careful eye set towards what a superior tribunal might decide upon review of a potential appeal or writ.

Police work contrasts radically with this model. Much of the officer's work is ”pre-process", and comprises actions that require above all, subtle and rapid reactions that are virtually impossible to define or prescribe a priori. How much force, for example, is justified in effectuating the arrest of a suspect? In practice, an officer must calculate in real time, factors such as the other's physical stature, whether and how he is armed, the consequences of his evading capture, and a host of other data that can only be determined (if at all) in the midst of the encounter. The law offers guides -- principles of proportionality and necessity -- but ultimately leaves the officer, in the heat of the moment, with little more than his reservoir of experience and good judgment. In other words, his discretion.

Second, neither the police hierarchy nor the judicial apparatus, which have defined responsibilities for supervising the actions of police personnel, are present at the moment of decision and action. (Few things highlight the different roles and modalities of judges and police more than the judicial oversight of police. Society mandates judges to second guess the police, and grants them the luxury of time and distance so that they might carefully contemplate what the officer had seconds to decide, perhaps with his own or others' safety at risk. ) Officers are supposed to record some of their actions (how and how often varies from place to place) but there is nothing to compare with the reasoned memorials of judicial decisions. This is particularly true when officers decline to act -- for instance normally, nothing is written up when the cop decides not to stop a car or write a ticket. In other cases, officers can make certain actions, even encounters with civilians, virtually invisible by simply failing to note or report them, even when required to do so. (New York City police officers have apparently used this tactic frequently: although they are required to fill out a form to record each "stop and frisk" encounter, outside investigators determined that frequently no record was made or kept.) In fact, for batter or worse, officers pass the majority of their professional lives acting with little visibility and little scrutiny, but with very high stakes. For these reasons the principal challenges in police reform are to examine, oversee, and occasionally bureaucratize discretion in order to ensure that day in and day out, officers are doing what we hope that they will, and not what we fear.

Accountability is the mechanism, or rather the operative principle behind a series of mechanisms, for regulating the exercise of broad discretion and limiting bad, illegal, or non- performance. In other words, the challenge is to bring the exercise of discretion within a system of institutional and societal controls. Achieving this is the nucleus of sustainable reform, since it tends to insure the return on other investments, such as better salaries, training and equipment. Accountability is also particularly important for Mexico, which has lacked systemic approaches to public administration of police forces, precisely because; it incorporates a focus on the institutional response as well as the individual conduct.

Accountability is a term not easily translated into Spanish. "Rendicion de cuentas", perhaps the term most often used, approximates but does not completely capture the dimensions of the concept as it is intended here. In Mexico in particular, rendicion de cuentas conjures up the notion of making a report, often, if not always, about the use of public resources. It therefore carries a connotation of fiscal responsibility and oversight, and does not, in common usage, necessarily include the idea of a response -- a process or processes which determine and impose consequences if the action is found to be inappropriate or illegal. As used here, accountability connotes review of a broader range of conduct, investigation if necessary, a determination of appropriateness, and the application of sanction or reward as indicated by the circumstances. It also implies an atmosphere of transparency and the provision of information needed to allow the appropriate actors to make informed decisions.

Police should not interpret the focus on their discretion, or the call for greater accountability solely as "stick", or a crippling restriction on their judgment and action. Demonstrated accountability is also a key to credibility, or as one expert, Lawrence Sherman put it, “evidenced-based policing”. As Sherman uses it, the term implies the presence of evaluation and accountability mechanisms, as elements of modern management systems, that permit the police hierarchy to know what officers are doing, with what results, and where to make changes — without which it would be hard to demonstrate impact, and thereby make a convincing case for public support, true civil service status, enhanced budgets, etc. Irrespective of the particular policing strategies employed — be they the so-called zero tolerance, community policing, or problem solving policing — effective accountability mechanisms are needed to assure both the police and the public that the inevitable officer discretion is not abused.

With accountability as the rubric for looking at the challenge of reform, the question shifts to where to begin, and what other experiences might offer clues.

What lessons do other experiences offer?

The most dramatic reforms tend to arise in post conflict situations (such as Central America), upon independence (as in East Timor), after the fall of security structures of an old regime (for example, Haiti) or with the change of governmental paradigm (Russia after the break-up of the Soviet Union). In such cases, there is the political momentum to make fundamental, as opposed to incremental changes. Often the government can rely upon external actors, such as the United Nations, NATO, or the Organization of American States to deploy forces to shore up security during the transition to a new police. At times donors are prepared to finance much of the process, as well as underwrite significant portions of the new entity's budget. Where a thoroughly discredited old force is to be radically transformed or replaced, the focus tends to be on creating a new police with some measure of professional selection and training, formal independence from the military, and reasonable working conditions — pillars, if you will, of democratic policing. These efforts usually include some initiative around accountability in order to offer some response to a prevailing pattern of rights abuses committed by police officers.

The "rupture" or post-trauma situations just described are similar in some ways and quite different in others from the broad outlines of the Mexican scenario (to the extent that the situation in Mexico, a federal republic of some 32 state-level entities, can be generalized). Mexico cannot contemplate decommissioning entire police forces at once, lacking 'blue helmets' or UN CIVPOL to maintain security while it reinvents the police. And although foreign experts may be welcomed to offer their perspectives, Mexico's reform process will have to be its own, not directed by other nations or donor institutions, as in the other cases. Nonetheless, some of these experiences are relevant.

In stable countries, including those which enjoy a tradition of "democratic" policing, reform tends to be more evolutionary, even if episodic. Important changes have often tended to deal with more detailed issues, the creation or strengthening of internal control or oversight mechanisms, the integration of different mechanisms into systems. In the most successful cases these changes reflect the acceptance of accountability as a central thread in management and public administration, and not just a necessary evil in order to placate public pressure over human rights or corruption scandals. These may prove to be more apt examples, particularly at the federal and Federal District levels.

There are diverse lessons to be drawn from the wide range of experiences. They speak to the steps needed to create the impetus for reform, and the steps needed to diagnose and address the challenge. Two initial recommendations are directed at the process for facilitating learning and agreement.

1. Acknowledge the technical aspects of this field of inquiry and practice, and seek formulas for avoiding political gridlock.

In Northern Ireland, after thirty years of conflict, the parties found themselves in extremely difficult negotiations that had as one of their central issues the reform of a police force that was tainted by violence and impunity according to Catholics who had borne the brunt of their tactics, but which for loyalist Protestants and the British government, offered the main line of defense against IRA terrorism. It became clear that the parties would not likely resolve the question of what to do with the police, and that some alternative process would be required if progress were to be achieved. Faced with a strong possibility of paralysis and failure, the parties found a mechanism which promised a sound technical basis for a reform plan, and a way around political impasse.

The parties agreed upon terms of reference for a commission charged with developing recommendations for reform. The terms of reference articulated not only the inclusive and democratic values that the Commission, which would come to be known by the name of its chief, Christopher Patten, was to emphasize, but also various enhancements of a professional nature that appealed directly to the police themselves. The terms include various popular participation and transparency requirements, ensuring that all interested sectors would have the opportunity to present their concerns and suggestions.

Among the distinguished members of the commission, almost all of whom were well-versed in the field, were foreign individuals. They were included partly to ensure that international experiences would enrich the Commission's consideration, but also to establish presences within the panel that were not partisans of any local cause or group, and who could help ensure the technical focus. Those Commissioners drawn from Northern Ireland, including some from the private sector, ensured that political realities were not entirely ignored.

Northern Ireland's own experience taught that however good a commission or its recommendations, it would achieve little if there were no political will to implement its vision. The parties took various measures to prevent repetition of past failures. First, the Good Friday Agreement, which set forth the overall political settlement and created the Patten Commission, was submitted to a public referendum, and obtained 71% of the vote. In one of its provisions, the Agreement underscored the parties' determination that implementation not be frustrated, stipulating that the signatories would "introduce and support legislation" to give effect to specific portions of the Agreement. Although the portions dealing with police matters did not include the same language, a combination of factors — the referendum, the prestige of the Commission and its members, the transparency of its work (the commission conducted more than 40 public meetings) and the detailed terms of reference -- created an environment in which it was virtually impossible to renounce the Commission's report, and still difficult to frustrate it by indirect means. The Commission itself added another device: it suggested an Oversight Commissioner to review implementation of the reform plan, which was created after the Commission delivered its final report.

Mexico might benefit from adapting certain elements of the Northern Ireland approach. The Fox administration has found itself stymied on most legislative fronts, and has not yet demonstrated a convincing habit of consultation or the capacity to push reforms past the objections of opposition parties determined to deny his government easy victories. The Los Pinos approach to legal and police reform holds out little promise so far of broader participation, and may well run aground when the time comes to develop reform legislation. If President Fox used his political capital to promote an agreed process, involving an appropriate commission, whose recommendations would become the shared responsibility to all parties to implement, Mexico might find itself with a recipe for moving forward in a difficult area. A serious group combining the right mix of insider and outsider perspectives that took up the challenge of studying, broadly listening, and proposing, would provide Mexicans with nothing worse than a proposal that was fodder for an enriched debate —and if the parties honored their commitments, perhaps a great deal more. Bequeathing the future of reform to the traditional system of overly personalized politics, quasi-clandestine negotiation and the exclusion of affected sectors of society cannot possibly be the best alternative.

2. Do not limit the learning to the government.

Local officials must take seriously the need to immerse themselves in the mechanics of modern police management and accountability. The challenge of making the police work is real, and complex, as is the question of ensuring public security. Mexico does not enjoy the luxury of great local expertise on the subject, so it falls to the authorities to catalyze a learning process within and outside of the institutions themselves.

Reforming the police is a struggle that will consume years. Every relevant actor in this area is fully aware of this. However there is reason to doubt that official diagnoses will be sufficiently self-critical, or reform plans adequate. Without outside ideas, critiques, and support Mexico would also lack a valuable resource for generating the political agreement to take needed action, and would be far less likely to have needed scrutiny of the implementation.

Mexican human rights groups NGO's have remained on the margin of debate about police reform, leaving a noticeable void in public discourse. Reform-minded officials from various agencies have noted this absence, lamenting what they perceive as a lack of support in legislative and policy battles. Mexico's transition toward pluralistic politics carries with it a certain irony: now with more official recognition of the institutional problems and need for profound changes, the NGO's who for years clamored for such changes have encountered difficulties in taking advantage of relative openness. They suffer from a lack of human and financial resources, as well as the same lack of familiarity with the subject matter that (unfortunately) characterizes many officials.

South Africa offers an illustrative, if unfortunate, example in this regard. During the apartheid era, there evolved an impressive network of local NGO "advice offices" whose presence in the so-called homelands as well as urban centers was a veritable lifeline for people taken away unjustly or otherwise victimized by the security forces. With the transition to multiracial democracy under the presidency of Nelson Mandela, the manifestation of police abuse changed somewhat, but the problem did not vanish. (Deaths in detention at the hands of South African police are as much, if not more, of a problem today as they were under apartheid.) However, five years after the initial transition, the number of advice offices had fallen from over 400 to about 250, casualties of the flow of attention and funds away from the NGO sector once the formal specter of apartheid had ended.

Government wastes the potential of civil society at the risk of bad policies and poor governance. Certainly this is one of the lessons to be drawn from Mexico's last century. In fact, a book on comparative democratic policing, still one of the most recommended texts of its kind, is the result of an NGO effort to bolster Northern Ireland's official reform effort. Human Rights on Duty, published by the Belfast-based Committee for the Administration of Justice (CAJ), compiles information on policing transitions in seven nations, thanks to the collaboration of more than a hundred officials, academics and advocates who completed questionnaires and provided analysis and materials. The aim of this effort was to enrich the Patten Commission process, and educate officials and journalists, as well as a public whose only experience with the police had been with a force created and operated pursuant to a national security doctrine.

By taking an active role in the process, the CAJ and other NGOs gained influence. The participation of credible NGOs also helped sustain a certain level of social support for the reform plan, including some of its inevitable compromises. Some who had traditionally denounced the police as irretrievably evil invested time and resources in the effort to find lasting means of making them better. In a context so polarized, with such visceral rejection of most official initiatives, this sort of participation was significant.

To those Mexican officials who ask what can they do to achieve some of the same results, the answer is simple: initiate a joint learning effort, bringing together officials from different agencies and civil society, including NGO representatives. Mexican officials are already developing a taste for comparative study : a team of officials from Los Pinos has recently traveled to Texas and to Chile to look at judicial systems, and a European tour is now contemplated. There are other examples of the authorities' study abroad. But to date few, if any initiatives include those outside of government circles, and there appears to be little sharing of the lessons that were brought back.

A joint initiative might follow the Los Pinos example of traveling to different places where the experiences are illustrative. This might be complemented by comparative seminars, convened to bring experts from abroad to different fora in Mexico. Working groups might review and debate important texts, in order to air different ideas and identify a few areas for more focused inquiry. NGOs that took part might obtain the building blocks of a new and appropriate role for themselves, an important advance in an area that has noticeably lacked their analysis and proposals. For the government's part, it would be a worthwhile and serious investment of time and money to catalyze a tangible collaboration between officials and the non-governmental community, something that they have sought, at least at a rhetorical level. No party would have to give up its rights to dissent, loudly if necessary. Only the right of abstention from matters of public concern would be compromised,

3. Adopt the language of rights and security

The very shape of public discourse about citizen security is one of the most crippling obstacles to joining forces in this area. Naturally enough, two principal values dominate discussion: the protection of the populace from crime and criminals (which implies, in turn, the prevention, persecution and punishment of crime), and clear respect for the rights of every person, whether victim, witness or the accused. When, as inevitably happens, the two notions are posited as incompatible, and in competition, opposing camps arise, each one emphasizing a different value. This paper doesn’t address the many arguments against such a dichotomy but suggests instead that all actors should acknowledge the importance of both values, in public and at each opportunity.

Mexican human rights groups have achieved something very important in this regard: public comment from police officials rarely omits some mention of the relevance of protecting human rights, or fails to assert that it is official policy and practice to do so. Although one could argue about the consistency with which such a policy is actually applied, it is important to recognize that official language has evolved, and congratulate those authorities who have taken the first step of adopting the rhetoric of rights protection.

At the same time, NGOs rarely reciprocate by underscoring the legitimate and necessary battle against crime. By assimilating the idea that the police must be effective as well as fair, rights advocates could help change the way they see and are seen by the police. Without abstaining from criticism, they could contribute to a basis for more productive interchanges about how to satisfy both values to a greater degree something that should interest both the activist and the officer.

4. Develop information mechanisms and practices.

One of the most often stated concerns of Mexican police officials is how to change negative public perceptions of the police. It is readily apparent to them that this will demand more than changing the police themselves; some frustrated officials complain that little notice is taken of the changes they have wrought. Information policies (relatively underemphasized by reforms to date) can help achieve both aims -- positive change, and having the public begin to perceive it. Internally, there is room for great improvement in the management of data about, for example, officer conduct, supervision, resource use. More important, this information could be tied to consequences (positive and negative) that would represent the incentive for improvement. And ultimately, processed data could highlight trends and impacts, strengths and weaknesses that point to the need to sustain or change certain polices and operational methods -- in other words, give police officials a continual sense of institutional health.

Addressing accountability issues is not the only option officials have open to them as they seek to win public confidence. There are numerous policing strategies available, each with its adherents. Before banking on beating back crime, consider that the crime indices are historically not terribly elastic in the short term, or subject to steep declines solely as a result of police actions. Police can undoubtedly affect crime, but their impact on a decidedly social phenomenon is necessarily partial, and still not entirely understood or necessarily replicable even after dramatic declines like that experienced in cities across the United States. It is even less likely that the results will materialize in time for demandingly short political cycles, and virtually impossible that this would occur with a vastly flawed police that have not come closed to overcoming their multiple defects. If officials can hardly rely on vastly improved crime statistics that will speak for themselves, it is worth considering a different way of demonstrating the will to change: a transparent process of cleansing and strengthening the institution, punishing bad conduct and rewarding good.

Police departments should create tracking and reporting mechanisms that would actively inform, via the Internet and other media, about various matters, including the manner in which the police have resolved citizen complaints or other administrative processes covering misconduct. This can be done without providing the public with any identifying data about the victim or alleged offender that might compromise confidentiality policies. (This is, in fact, the practice of many police forces, although it is generally considered good practice to provide, privately, the complainant with more detailed information.) Officials can create virtuous competitions by ordering the data according to police precincts or administrative units, so that the public becomes aware of which units stand out with fewer complaints, or with the most appropriate handling of those which do arise. Simple questionnaires given to complainants can also yield valuable information about the public's experiences with the police and the accountability process, which can be tracked across both time and geography.

In general, more agile and transparent information systems and practices will help police forces to achieve several objectives central to the main goal of improving public perceptions of the police, for example:

• Evaluate the effort and results achieved by officers and their supervisors.
  
• Follow and interpret patterns of conduct at the individual, unit and institutional level, which will provide the feedback needed to modify recruitment and training practices, or operational rules. In short, it provides an informational foundation for deciding how to make the institution better.
  
• Sensitize and educate the public about police efforts to acknowledge and address the institutional problems, thereby helping to adjust expectations to accord with the profound difficulties of this transition. It is worth noting that a public that is well-informed on such points can be an important source of support for a reform process that will inevitably encounter some resistance from within.
  
• Aid the police in obtaining a reliable understanding of what actions and services the public expects from the police. Here too, more information may yield different demands on the police. If the police offer a compelling menu of services (prevention, assistance, etc.) conceivably the demand might evolve towards a force that defines its mandate less exclusively in terms of arrests, instead including a range of other functions.

5. Create police-specific external audit and advisory mechanisms.

Mexican officials should consider the creation, as an early priority, an autonomous public entity to audit the performance of accountability and control mechanisms that apply to police forces. The official human rights commissions cannot play this role, given their government-wide mandates and individual case orientation toward individual cases. A police-specific organ is needed, with personnel whose training and experience afford an appreciation of the difficulties of police work, as well as the capacity to evaluate institutional performance.

Having such expertise and oversight can bring powerful benefits to police managers, and in turn, to the rank and file. One dramatic example concerns the Special Counsel's office for the Los Angeles County Sheriff's Department (LASD), in California. Los Angeles County contains 88 municipalities, including the city of the same name. The LASD has the responsibility for policing almost 50 of these municipalities, while the separate Los Angeles Police Department (LAPD) covers the city of Los Angeles.

The two forces share significant similarities. Police personnel levels are comparable (LASD 8,900; LAPD 10,000) as are the respective populations they patrol (LASD 2.7 million; LAPD 3.5 million). At the time of the 1991 Rodney King incident (involving an African-American man who was brutally beaten by four LAPD officers in the presence of at least ten of their colleagues) both departments had serious brutality. However, motivated by the high costs of police brutality litigation, which were running into the tens of millions and threatened to bankrupt Los Angeles County, officials decided to name a Special Counsel to examine institutional issues within the LASD. The Special Counsel, which functions as an outside auditor, was given broad access to LASD data, processes and personnel, and publishes serious semi-annual reports whose critical findings are always accompanied by detailed suggestions for improving department management and avoiding the problems that have plagued the force.

The results for the LASD, and the marked difference from post-Rodney King events around the LCPD, which stumbled from scandal to scandal, are impressive. LASD halved the number of civilians killed by its officers between 1991 and 1998. For years after, the LAPD didn't even maintain complete data on its officer-involved shootings. However, six years after the LASD created its Special Counsel, LAPD's total number of officer-involved shootings was 30% higher than the LASD's. (In the decade 1991-2001 the annual number of persons shot by an LASD officer dropped from 56 to 12, while the number of officers killed by gunfire dropped from 10 to 2.) Indeed, LASD's relatively open approach to its officers' misconduct, while LAPD tended to whitewash bad incidents, and the LASD's decreasing numbers of officer-involved shootings while LAPD's rose between 1991 and 1998 prompted the LA Weekly to ask in 1999, "Why can't LAPD be more like the Sheriff's Department?"

There is an interesting footnote to this comparison, which sheds some light on the question of the relationship between greater accountability and police aggressiveness against crime. Data covering 1997 to 2001 show that the LASD had a higher index of arrests (expressed as a percentage of crime) than the LAPD. (Since 1998, when it was forced by the worst scandal in its history, to implement more serious investigations of civilian complaints, part of a stricter accountability regimen, the LAPD now suffers fewer cases of excessive force, lower litigation costs, and bit by bit, projects a more positive image before the community. )

Mexico's existing SECODAM, the anti-corruption agency, and the counterpart Controlarias Internas in federal agencies) are not a substitute for an effective accountability auditor. These are internal bureaus whose focus has largely reflected the fiscal connotation of rendición de cuentas, and have had little to do with evaluating progress on other key police-specific fronts, such as the development of systems to review use of force, officer safety practices, etc. Moreover, Mexico's experiments in accountability and anti-corruption have largely reproduced the broader Latin American experience: the creation of innumerable additional requirements, and rigid bureaucratic layers that have diminished both efficiency and fairness without producing a corresponding increase in service. In the words of Esteban Moctezuma Barragán and Andrés Roemer, authors of a recent work on public service in Mexico, as a result the government has to pay more to do the same. A review of various annual reports from different agencies also underscores the continuing and urgent need for mechanism that can offer the public a clear and coherent vision of the specific problems and a meaningful description of the actions taken or needed.

Police auditor or complaints bureau?

An audit mechanism promises serious benefits, which from the perspective of strengthening institutional systems, make it a priority. By its nature, the auditor reviews the performance of control mechanisms; in large part this touches on the effectiveness of internal supervision. Academic studies, officer surveys, and the observations of experience all underline the important role of supervisors and commanders in defining what will be internally understood as good or bad performance. An innovation designed to evaluate management's success in creating a culture of respect and accountability would, from this perspective, be an even more urgent priority than the creation of an external entity devoted to receiving civilian complaints against the police. A complaints bureau will almost certainly be necessary (notwithstanding the current human rights commissions) but its recommendations would have much greater impact if the internal systems were in place to deal with them — something that a well-designed audit mechanism can help to bring about.

Even so, authorities at every level of government should seriously study the possibility of a police ombudsman. Other nations have experimented with diverse forms of this model: Northern Ireland named a police ombudsman as a result of its Good Friday Accord, while South Africa created an independent Complaints Directorate to complement several internal mechanisms designed to regulate and monitor use of force. Queensland, Australia created a Criminal Justice Commission to review police conduct in the wake of egregious abuse and corruption scandals. Following inquiries by an ad hoc public committee and various government reports throughout the 1980s, Israel created an external complaints board in 1992, while Colombia, with one of the highest rates of state violence in the world, named a (short-lived) complaints commissioner with internal and external elements. Even New York City's police, who had long fought an independent Civilian Complaint Review Board, finally accepted an autonomous Board in 1993. In 2001 former mayor Rudolph Giuliani responded to widespread concern over the Police Commissioner's frequent decisions to ignore the Board's findings that punishment was warranted, proposing to augment the Board's authority to prosecute the cases administratively, instead of the police officials who previously had this responsibility. Canada has had a complaints board solely for the fabled Royal Canadian Mounted Police (RCMP) since 1988, which conducts investigations that are separate from but complementary to the RCMP's own internal disciplinary procedures. Peru has created a police ombudsman of a different sort: the office is dedicated to defending the rights of officers, and handles cases ranging from challenging abuses of administrative disciplinary processes to assisting surviving spouses to secure death benefits.

The truth of citizen oversight mechanisms is that they have not proven to be sufficient by themselves to bring about the institutional changes needed. For one thing, it is difficult to obtain political agreement over their powers, particularly the tools needed to remain effective in the face of resistance. The essential lesson from experience seems to be that such mechanisms are part of the solution, and should necessarily share authority with the police. They represent a potentially powerful source of oversight, but are not a substitute for the proper internal application of the rules norms.

Perhaps the most prudent approach would be to opt for creation of an audit mechanism, while examining different options for a police ombudsman. The experiences of Mexico's human rights commissions should be evaluated, and attention given to the relationship between them, the audit mechanism and an eventual ombudsman.

Different models can be used to fulfill the function of receiving and investigating civilian complaints. Perhaps more important than the specific design is achieving a balance of different internal and external forces. Police officers and government officials acknowledge privately that many police reject the human rights commissions and NGOs, considering them implacably hostile. Although understandable, this attitude appears to be based on a flawed perception. To the extent that such organizations appear to the police to be too critical, too closely scrutinizing their acts, it probably reflects the rule that public politics abhor a vacuum. Police forces in Mexico do not offer the public sufficient tangible evidence that they are engaged in self-correction, and therefore other actors feel compelled (often by public pressure) to demonstrate that they are filling the gap. Consider an alternative, vastly preferable scenario: internal accountability systems begin to demonstrate that they are exercising effective control over police actions, and external actors begin to feel freer to change the emphases in their work, devoting resources to a different type of focus on the police, or turning to another set of issues altogether. For example, outside groups would be unlikely to feel pressure to continue as before if an audit mechanism began reporting on better discipline and improved police performance.

6. Design a federal entity to support experimentation and study

Federated republics face challenges relatively unknown to unitary systems. Responsibilities and resources divided among so many diverse jurisdictions, and the variety of normative frameworks and institutional designs can make coordination extremely difficult. On the other hand federations offer the extraordinary opportunity to tae advantage of the different experiences at the state and municipal levels. Federal officials, acting in their unique national capacity, should create an entity that will harvest this potential by promoting study, experimentation, and dissemination of the results of local efforts.

Washington, D.C. offers a useful example. As in Mexico the U.S. federal government and its different police forces (the FBI, INS, Border Patrol, etc.) have limited jurisdiction, touching only federally defined crimes, while the vast majority of street crime is in the hands of local forces. However, since 1968 Washington has carved out an important supporting role which has been essential to the development of state and local policing techniques.

In the 1960's, after decades without figuring on the political radar, crime (and the accompanying multiple social tensions) prompted President Lyndon Johnson to name a technical commission (see recommendation 1, above) on "Law Enforcement and the Administration of Justice" which in a sentence captured a good deal of the problem of then-prevalent antiquated law enforcement practices: "the revolution of scientific discovery has largely bypassed the problems of crime and crime control”. Congress responded by creating the National Institute of Justice (NIJ). In its 35 years of operation, the NIJ has financed innumerable studies and pilot projects, carefully documented and accessible to scholars and police officials and interested readers. It is worth noting that the Institute typically does not vouch for or control the findings, opinions or policies advanced by the projects, which often touch on issues that are sensitive for police officials at all levels of government. The net effect is to provide a rich assortment of academic study and concrete experience distilled from federalisms great asset: an array of state and local laboratories for testing new ideas and —with the support of the NIJ — an ongoing virtual conversation among them.

* * *

These recommendations do not pretend to be solutions. They are ideas, and even if implemented they will not begin to exhaust the decisions that face police officials as they confront the policing challenges before them. Even within the area of accountability, there are difficult choices — what sort of mechanisms are needed (and wanted), what level of public transparency, what mix of internal and external controls. This paper aims to help to define some of the questions and offer a basis for tackling them. It is virtually impossible to predict where the process of reform will lead, but perhaps not so to determine where to begin. Mexico's challenges are its own, but they are not entirely unique, and there is much to draw on from the world around it.

¹ Translated and adapted from remarks given at the "First Conference for a Rights-Respecting Culture in the Pursuit of Justice", convened by the office of the Attorney General of the Federal District to celebrate the 54th anniversary of the Universal Declaration of Human Rights, December 10, 2003.