Header image for news printout


The Special Rapporteur on extrajudicial, summary of arbitrary executions, Philip Alston, issued the following statement in Brasilia at the end of his mission to Brazil on 14 November 2007.

I arrived in Brazil eleven days ago at the invitation of the Government in order to inquire into the phenomenon of extrajudicial executions. I am very grateful to the Federal Government and to the State Governments of São Paulo, Rio de Janeiro and Pernambuco for the unqualified cooperation extended to me.

I met with a wide range of officials including the Minister of Foreign Affairs, the Federal Secretary for Security, the chief judge at the Supreme Federal Court, other judges at various levels, the Governor of Pernambuco, the Prosecutor-General of the Republic, and a great many members of the Ministério Público and Public Defender at the State level. I met with State Secretaries of Public Security, Justice and Prisons, as well as with the respective heads of the Military Police and the Civil Police, Technical-Scientific Police, Federal Police, representatives of the various Internal Oversight offices, and the Ombudsman in each State. I visited a prison in São Paulo, a civil police station and a military police battalion in Rio de Janeiro, a favela in Rio de Janeiro, and a settlement in Pernambuco.

I was privileged to meet with many of the key human rights actors including the UN Country Team, members of the Council for the Defence of the Human Person, the Human Rights Department of the Ministry of Foreign Affairs, the Commission of Human Rights and Minorities of the Federal House of Representatives, the Special Secretary of the National Human Rights Secretariat, and the Human Rights Commission of the Rio de Janeiro Legislative Assembly.

I also received detailed briefings from a wide range of civil society representatives, victims, witnesses, and families of executed persons. Their inputs were invaluable.

In short, I sought and received detailed information from a very wide range of the key actors within Brazilian society.

Before looking at specific problems I would like to make several preliminary points. The first is that the Government’s invitation for me to visit reflects a clear commitment to transparency and to living up to its human rights obligations.

Second, while my task is to report to the UN Human Rights Council on the situation I have found, real solutions to the problems will only come from within Brazil. My most important role then is to assist Brazilian society to acknowledge the scale of the killings that are occurring, and to act as a catalyst to discussions designed to identify effective solutions.

Third, I recognize that my visit takes place within the context of widespread community concern about criminal violence. The cities of Brazil face enormous challenges in keeping their residents safe from the violence of gangs involved in drug trafficking, arms trafficking, and other organized crime. In Rio de Janeiro, such gangs dominate entire communities, subjecting residents to senseless violence and constant repression. In São Paulo, the events of May 2006, in which a gang brought the city to a halt with systematic attacks on public institutions provided a shocking demonstration of the need for more effective policing. I should emphasize that human rights law not only prohibits governments from committing extrajudicial executions, but that it also requires governments to protect their people from murderers. Indeed, one of the central pillars of the idea of human rights has always been the right to life and the freedom from fear. Thus, ensuring security for all citizens is a key role that governments must play. Human security is a part of, and not in competition with, human rights. In the Brazilian context in particular, my findings show that the issues of ending human rights abuses by the police and ensuring effective crime prevention by the police are tightly linked. A key reason for the ineffectiveness of the police in protecting citizens from these gangs is that they too often engage in excessive and counter-productive violence while on-duty and participate in what amounts to organized crime while off-duty.

Fourth, my visit has convinced me that within Brazil there are many powerful forces for good. While my focus is on the ‘disaster areas’, this should not be seen to imply non-recognition of the many positive developments within this society.

Finally, I should note that my present statement is only designed to highlight some, but by no means all, of the issues and recommendations which will be part of my final report, which I expect to be available to the public by March of next year.

Some key concerns

The following statistics, coming from diverse sources, provide a snapshot of some of the key problems:

· Homicide is now the leading cause of death for persons aged 15-44 years. For some time now there have been 45,000-50,000 homicides committed every year in Brazil. Victims are overwhelmingly young, male, black, and poor.

· In Rio de Janeiro and São Paulo, only about 10% of homicides are tried in the courts; in Pernambuco it is about 3%. Of that 10% tried in São Paulo, about 50% are actually convicted.

· In the first 6 months of this year in Rio de Janeiro, the police recorded 694 “acts of resistance followed by death”. This is very often a euphemism for extrajudicial executions by the police killings and it is a category which virtually ensures that impunity will follow.

· In Pernambuco, 61 prison deaths were reported during the first ten months of 2007, with 23 occurring in the Anibal Bruno prison, including more than a dozen killings this year, several of which occurred earlier this week.

· In the same state, a reliable estimate is that 70% of all homicides are committed by death squads, and many of those death squads are made up of policemen and former policemen. The 197 people who have been arrested this year for death squad activity represent only the tip of an iceberg.

· Recently, some 2,000 files which were turned over to the Public Prosecutor by the police in Pernambuco expired because the police had delayed action for so long that the statute of limitations on the crimes had passed.

· Over-crowding in Brazilian prisons is now so bad that the occupancy rate is often 3 or more times as many prisoners as the facility was designed to hold. Is it any wonder that riots take place?

What are the main problems that are represented by these snapshots?

· Very high homicide rates, and high rates of impunity: This includes violent killings by individuals, lethal confrontation between drug traffickers and other gangs, and the killings of police and other officials by criminals. These killings have sown widespread fear and insecurity among the general population, but remarkably little is being done in the vast majority of such cases to investigate, prosecute, and convict the culprits. The low rate of cases going to trial illustrates the system’s failure in this regard.

· Killings by vigilante groups, death squads, extermination groups, and militias: These generally consist of off-duty police, ex-police, firefighters, and private citizens involved in activities such as: (a) contract killing; (b) taking over a geographical area, and extracting ‘protection’ money from residents, often under threat of death; and (c) killing or issuing death threats on behalf of landowners to landless workers or indigenous persons as a result of land disputes. (This last issue is especially complex and will be addressed in detail in my final report.)

· Prison killings: Including: (a) inmates killing other inmates; (b) security guards killing prisoners; and (c) inmates killing security guards. I will return to this issue below.

· Killings of police: the police in Brazil clearly operate at significant risk to their lives in many situations. The number of police killed is totally unacceptable and all appropriate lawful measures need to be adopted to prevent such deaths. But we also need to look carefully at the figures. In Rio in 2006 for example the statistics show that 146 police were killed. But only 29 of those were killed while on-duty. The remaining 117 were killed when off-duty. A very significant proportion of those 117 is likely to have been engaging in illegal activities when killed.

· Police killings: These killings are of major concern because they indicate a degree of lawlessness which undermines other efforts to reduce homicides and other forms of criminal activity. They break down into two categories.

Extrajudicial executions by on-duty police

In most cases killings by on-duty police are not included at all in the homicide statistics. Instead they are registered as “acts of resistance” or as cases of “resistance followed by death”. In theory, these are instances in which the police have used necessary and proportionate force in response to the resistance of criminal suspects to the orders of law enforcement officers. In practice the picture is radically different. The determination as to whether an extrajudicial execution or a lawful killing has occurred is first made by the policeman himself. Only rarely are such self-classifications seriously investigated by the Civil Police. I have received many highly credible allegations that specific “resistance” killings were, in fact, extrajudicial executions. This is reinforced by studies of autopsy reports, and by the fact that the ratio of civilians killed to police killed is astonishingly high.

This and other problems are well illustrated by the events involving some 1,350 police that took place in the Complexo do Alemão community of Rio de Janeiro on 27 June 2007. Nineteen individuals died in the so-called “mega-operation” that took place. I interviewed the relatives of eight of those victims and reviewed a range of other reports on the incident. I also spoke with those responsible for directing the operation and to the officers at the police station that is responsible for investigating the killings. They provided me with no evidence that any sustained investigation has been undertaken. They confidently asserted that nearly all of those killed had criminal records. This “fact” could not have been known positively to the police when they killed the individuals. The assertion was firmly denied in statements made to me by families of victims in several of the cases, including that of the killing of a 14 year old boy. And, even if each of the victims had had a criminal record, the appropriate response is arrest not execution. The investigators have apparently failed to ascertain which police fired even a single one of the shots, and have not reconstructed the circumstances in which each of the 19 died. (Their response to these deaths in Complexo do Alemão appears to be typical: The officers I spoke with could not recall that station ever having ever concluded that a single case in which a policeman alleged that he had lawfully killed in response to a suspect’s violent resistance was actually a homicide.)

I asked the head of the Civil Police in Rio de Janeiro about the findings of an independent autopsy report which strongly suggested that some of the individuals had been extrajudicially executed by the police. His response was to attack the credentials of the ‘out-of-state’ experts and to question their constitutional right to undertake such an analysis. I requested from him, but have not yet received, a scientifically credible response to the report.

Many of those with whom I spoke from the Government and Police in Rio de Janeiro considered the action in Complexo do Alemão to be a model for future action, and most asserted its success. Indeed, person after person casually used the terminology of “war”. But “war” cannot be fought against selected criminal individuals. It is fought against entire communities. The language of war provides a convenient justification for a military-style invasion, and for a strategy that focuses only on force and confrontation. I sought to discover why exactly the operation had been undertaken. Many reasons were offered, but there was little consistency. At the end of the day, unspecified intelligence reports were usually cited by way of justification. The actual results achieved are noteworthy. The most important major drug dealers and traffickers were not arrested or killed. I was told “many” weapons were seized. Given that the community was said to have been awash in arms I was staggered to hear that a 24 hour occupation by 1,350 men yielded: 2 machine guns, 6 handguns, 3 rifles, 1 submachine gun, 2,000 cartridges, and 300 kilograms of drugs. Not a single policemen was killed and few were hurt, but the “resistance” encountered apparently necessitated the killing of 19 persons.

To the extent that the Complexo do Alemão operation reflects the main strategy of the State Governor, it is politically driven and amounts to policing by opinion poll. But, it is popular among those who want rapid results and displays of force. The irony is that it is counter-productive. Various senior police officers with whom I spoke were highly critical of the “war” approach. The Military Police forces involved have had little relevant training in the use of non-lethal weapons, there have been no attempts to develop community-based policing in this area, and almost no sustained social services are provided by the state to the people of the affected community.

Extrajudicial executions by off-duty police

State police, especially state military police, routinely work second jobs while off-duty. Some form “militias”, “extermination groups”, or “death squads” and other groups that engage in violence, including extrajudicial executions, which occur for several reasons. First, their protection rackets – in which shopkeepers, transport providers, and others are coerced into paying money to the group – are violently enforced. Second, to prevent gangs from undermining their control, persons suspected of providing information to, or otherwise collaborating with, gangs are killed. Third, while such groups do not generally begin as death squads per se, the already illicit relationship that they have developed with the more powerful and affluent elements of the community frequently results in their engaging in murder for hire.

For the residents of these communities, control by a “militia” is nearly indistinguishable from control by gangs and traffickers.

Prison violence

High levels of riots and killings in prisons are the result of a number of factors.

Severe overcrowding in prisons contributes to inmate unrest, and also to the inability of guards to effectively prevent weapons and cell-phones from being brought into prisons. Wardens are insufficiently trained and supervised. Low levels of education and work opportunities also contribute to unrest, as does the failure to ensure that inmates are transferred from closed to open prisons when they are entitled to do so. Delays in processing transfers, together with warden violence and poor general conditions encourage the growth of gangs in prisons, which can justify their existence to the prison population at large by purporting to act on behalf of prisoners to obtain benefits and prevent violence.

There are many bodies with the power to investigate prisons conditions, but none are doing their job adequately. This lack of external oversight permits poor prison conditions and abuses of power to continue. Requirements in some places to identify with one gang faction facilitates the growth of gang-identification, and gang related activity. While some role for factions in the prison system may be unavoidable in the short-term, it contributes to the growth of gangs and elevates crime rates more generally.

The criminal justice system’s response to extrajudicial executions

A necessary reform to address the problem of extrajudicial executions committed by the police is to change the strategies and culture of policing. This approach is immensely important, and my final report will make a number of specific recommendations in this regard. However, another equally important approach is to ensure that when extrajudicial executions occur, those policemen responsible are convicted and imprisoned, and to to ensure that the victims get justice and the guilty cannot kill again. It is then disturbing that, as I noted earlier, very few homicides result in convictions.

This might suggest that the criminal justice system is hopelessly broken. This is not the case. A successful conviction for murder is the end result of a process handled by a number of institutions: Typically, that process would be that the Civil Police effectively finds witnesses and preserves the crime scene; those witnesses may then require protection; the Technical-Scientific Police gather specialized forensic evidence; the Ministério Público builds a case against the suspect; a jury court hears the evidence and finds the suspect guilty. Every step depends on those that come before it. If any institution fails to act in an effective manner, the whole process is a failure. The bad news is that one or more institutions generally do fail.

The good news is that all of the institutions include a significant number of competent personnel, and that some of the institutions generally function quite well. For example, I have been especially impressed with the professionalism and dedication of the Ministério Público. Similarly, while the witness protection programs suffer from both funding shortfalls and institutional defects, they do succeed in protecting a large number of witnesses.

In my final report, I will make a number of specific recommendations regarding how the criminal justice system should be reformed so as to effectively prosecute extrajudicial recommendations. However, as a preliminary observation I would note both that the criminal justice system is in desperate need of large-scale reform and that such reform is completely feasible. Brazilian society should feel a sense of great urgency in making these reforms, but it should also feel confident that if it acts with urgency, it will succeed.

Preliminary conclusions and recommendations

My report will include detailed recommendations to the federal and state governments for reforms in approaches to policing, and in how the criminal justice system functions. However, a few preliminary recommendations are in order:

Police pay: Low pay for police leads to a lack of professional pride and encourages police to engage in corruption, to take second jobs, and to form “extermination groups”, “death squads”, “militias” and other vigilante groups to supplement their pay. Reforms must include higher salaries.

Investigations into police killings: The Civil Police and the internal affairs services of the Civil and Military Police must effectively investigate killings committed by the police. The current system in many states of immediately classifying police killings as “acts of resistance” or “resistance followed by death” cases is completely unacceptable. Every killing is a potential murder and must be investigated as such.

Forensics: Forensics police and institutions are under-resourced and under-equipped, and they lack independence. To ensure effective prosecutions, this must be changed.

Witness protection: Witnesses to extrajudicial executions committed by the police and organized crime legitimately fear reprisals for testifying. This fear is increased when the police remain on duty throughout the investigations. There is much that is impressive about the current witness protection programs, but their inadequacies must also be candidly recognized and urgently addressed.
Ombudsmen: In the states that I visited, the police ombudsmen lacked true independence or the ability to gather facts on their own. This must be changed: The police require genuine external as well as internal oversight.
Public prosecutors: The Ministério Público is a dedicated and professional body. It must play a key role from the very outset in the investigation of every single incident involving killing by the police.
Prison monitoring: Prisoners have an immense and justified fear of reporting the violence to which they are subjected. The many institutions required by law to monitor prison conditions — most notably including judges of penal execution — are unable or fail to play this role in any adequate manner. The number of such judges must be increased, and the manner in which they work must be greatly improved.
Prison administration: Prisons must be run by the wardens, not by the inmates. This is an undeniably complex issue, and the safety of prisoners must always be the first priority. However, some practices must be discontinued immediately. The practice in Rio de Janeiro of forcing new prisoners who have never belonged to any gang to choose one upon entry into the system is both cruel and needlessly swells the size of the gangs. The human rights of the prisoner and of society at large are violated.
The people of Brazil did not struggle valiantly against 20 years of dictatorship, nor did they adopt a federal Constitution dedicated to restoring respect for human rights, only in order to make Brazil free for police officers to kill with impunity in the name of security. It is imperative that the Federal and State Governments implement sustained reforms in the directions I have indicated in order to enhance the security of ordinary citizens and promote respect for human rights.