click click Table of Contents click


Behind Bars in Brazil




In the decade since Human Rights Watch first examined prison conditions in Brazil, the inmate population has increased at a rapid pace. Beside exacerbating prison overcrowding-a problem that Human Rights Watch originally documented during its 1988 mission to the country-the fast growth of the inmate population has coincided with years of flagrant prison abuses. Thus, our first Brazil prisons report was succeeded by a 1989 newsletter that focused on conditions in a notorious São Paulo jail, followed by a 1992 report on a massive prison massacre in São Paulo. Our annual summary of global human rights conditions, moreover, has consistently condemned Brazil for severe prison overcrowding, horrendous conditions of detention, and summary executions of inmates.

Human Rights Watch's sustained attention to conditions of confinement in Brazil reflects our sense that the mistreatment of prisoners is one of the country's most serious and chronic human rights violations. Despite a few encouraging changes over the years that we have monitored prison conditions in Brazil, the overall picture has been grim. In 1997, in particular, a dramatic series of riots, hostage-taking episodes, and killings in penal facilities across the country forcefully confirmed to us the need for continued international monitoring of Brazil's treatment of prisoners.

The present report is the most comprehensive and detailed review of conditions of detention in Brazil that Human Rights Watch has undertaken to date. Indeed, it is the fruit of the most exhaustive prisons research that the organization has conducted in any country. The conclusions in this report are based on an examination of penal facilities in Brasília and in the states of Amazonas, Ceará, Minas Gerais, Paraíba, Rio Grande do Norte, Rio Grande do Sul, and São Paulo. Although our main focus was on prisons, we also visited a number of jails and police lockups, particularly since, given prison overcrowding, prisoners often spend years in such facilities.

In all, from September 1997 through March 1998, Human Rights Watch researchers visited some forty prisons, jails, and police lockups, interviewing hundreds of prisoners while also meeting with prison authorities, prison staff, members of the Prison Ministry (Pastoral Carcerária, a branch of the Catholic Church that provides assistance to prisoners), judges, lawyers, prosecutors, legislators, academics, and representatives of nongovernmental organizations. Among others, we met with the heads of several state prison systems (in most states, the secretary of justice), the president of the national human rights commission, the author of the final report of the São Paulo Legislative Assembly's Parliamentary Commission of Inquiry into the São Paulo prison system, the national coordinator of the Catholic Church's Prison Ministry, the policeombudsman for the state of São Paulo, and the president of the São Paulo prison guards union.

Although where directly relevant this report touches on general deficiencies in the criminal justice system, it does not purport to be a comprehensive evaluation of the administration of justice in Brazil. Human Rights Watch has, in several previous reports, documented and analyzed the other serious human rights abuses that plague the Brazilian justice system, focusing principally on the issue of police violence.1 The situation of inmates described in this report is, it should be emphasized, an integral part of this larger system. The lives of Brazil's prisoners are directly affected by the decisions of, among others, the country's police, judges, public defenders, and prosecutors. Therefore, although prison reform is clearly needed, a comprehensive approach to reform-one that integrates human rights concerns into all levels of the justice system-is likely to be more effective than a piecemeal one. Given the advantages of such an approach, Human Rights Watch was encouraged by the Brazilian government's 1996 promulgation of a national human rights plan. It is our hope that this report will support the effort to implement the human rights goals articulated in that plan.


Brazil's prison woes represent the logical consequence of two decades of soaring crime rates, mounting public pressure to "get tough" on crime, and continuing neglect by policy makers. The 1997 wave of inmate riots, though sparked by recent events, had its roots in a confluence of historical factors.

The 1964 coup d'etat that ended the rule of President João Goulart was the first in a series of coups in South America that ushered in the region's two decades of authoritarian rule. In Brazil, as elsewhere in the region, the military regime was characterized by grave and systematic human rights violations, including forced disappearances, political killings, and the routine practice of torture on political detainees. In the late 1970s, the military began a top-down process of abertura, or opening; a turning point in this process was the passage of a 1979 amnesty law that exonerated those who committed human rights abuses, provided for the release of many political prisoners, and allowed numerous exiled dissidents to return. In the years that followed, the military government refused to accede to increasing popular demands for an immediate return to democracy, instead controlling a gradual transition in which Brazilians were permitted to choose a president through indirect elections in 1985 and to elect constituents to draft a new constitution in1988. In 1989, in the first direct presidential balloting in nearly thirty years, Brazilians elected Fernando Collor de Mello in a close runoff with leftist candidate Luís Inácio "Lula" da Silva. Three years later, demonstrating the strength of their newly created democratic institutions, Brazilians impeached Collor because of his administration's rampant corruption.

The transition to democratic rule ensured that opponents of the military regime, many of them the sons and daughters of middle-class Brazilians, were no longer subject to massive human rights violations. Instead, much as in the period prior to the 1964 coup, criminal suspects, the landless, darker-skinned and poorer Brazilians, and others on the margins of mainstream society once again became the principal victims of official violence. As the targets of state abuse came less and less from affluent and established segments of Brazilian society, public support for human rights declined. This shift coincided with the continuing rise in violent crime, explained below, that has plagued Brazil over the past two decades. For many today, the defense of human rights has become synonymous with the defense of bandidagem, or criminality.

The serious economic troubles that buffeted Latin America beginning with the petroleum shocks of the 1970s and debt crisis of the 1980s did not pass lightly over Brazil. In the wake of the Brazilian economic "miracle"-the years of continuous double-digit growth that marked the late 1960s-Brazil's economy slowed in the 1970s and came to a virtual halt in the 1980s. From 1980 to 1993, according to the World Bank, Brazil's annual growth rate averaged just 1.5 percent. At the same time, the gap between rich and poor, already wide, grew steadily. An increasingly large number of those mired in poverty turned to crime, often violent crime. In the state of Rio de Janeiro, where Brazil's crime epidemic is most visible, the number of homicides tripled during the years of recession and economic stagnation, soaring from 2,826 homicides in 1980 to 8,408 in 1994.2 In parallel fashion, the homicide rate in the state of São Paulo soared from 14.62 per 100,000 in 1981 to 44.89 per 100,000 in 1995.3

Another cause of criminality in Brazil has been the growth of the international market for cocaine, which has produced corresponding surges inincarceration for trafficking and related offenses.4 Although there are no reliable measures, by all accounts the drug trade in Brazil is a multimillion-dollar business, and a growing one. In 1994, the Drug Division of Brazil's Federal Police seized 11.8 metric tons of cocaine, a seven-fold increase over the 1.7 tons captured in 1989.5

The crime epidemic has fueled public support for the enactment of harsher criminal laws. The most prominent example of recent legislative initiatives spurred by public concern over crime is the Law of Heinous Crimes (Lei de Crimes Hediondos), passed on July 25, 1990, in response to a rash of kidnappings. The law augmented the sentences imposed for a series of crimes, including kidnapping and drug trafficking, rendering those arrested for such crimes ineligible for pretrial release. Experts differ as to the deterrent effect of this law, but there is little doubt that it has done much to increase the size of the inmate population, particularly of those detained pretrial in police stations and jails.

As this example suggests, crime control policies in Brazil are often influenced more by the immediate political stakes than by any criminological analysis of their efficacy. Indeed, many state authorities (particularly those responsible for crime control) have gained election and continued popularity by championing policies that are aggressively anti-crime, and often anti-criminal suspect. Besides legislative efforts, recent policing trends have also spurred increases in the inmate population. In the months preceeding Rio de Janeiro's 1994 gubernatorial elections, for example, the government launched Operation Rio, a series of joint military-police actions to battle drug traffickers in the city's favelas, or shantytowns. The decision to bring in the military was widely celebrated in the media and by the leading candidates, including Rio's current governor, Marcello Alencar. One consequence of the massive arrests that characterized Operation Rio was a significant short-term increase in the chronic overcrowding of Rio's policelockups and prisons. Many of those arrested were detained for the thirty days allowed under the terms of the Law of Heinous Crimes.6

In 1994, and again in October 1998, Brazilians elected Fernando Henrique Cardoso to the presidency. The election of Cardoso, a sociologist by training who went into exile during the military dictatorship, raised hopes for improvement in Brazil's human rights record. Indeed, his administration has openly acknowledged many of Brazil's serious human rights problems, including its prisons crisis. As this report describes, however, much of Brazil's penal policy is decided at the state, not the federal, level. And regrettably, in those areas in which the federal government is active, Cardoso's administration has not lived up to the expectations that accompanied his election.

Methodology and the Question of Access

This report is one of a series of reports published by Human Rights Watch that describe conditions in prisons worldwide. Besides Brazil, we have investigated and reported on prison conditions in Czechoslovakia (prior to its division into two states), Egypt, Hong Kong, India, Indonesia, Israel and the Occupied Territories, Jamaica, Japan, Mexico, Poland, Romania, South Africa, the former Soviet Union, Spain, Turkey, the United Kingdom, the United States (including a separate short report published on Puerto Rico), Venezuela, and Zaire. We have also published a comprehensive report, the Human Rights Watch Global Report on Prisons, that covers prison conditions in such additional countries as China, Cuba, and Peru. In this report, as in all of our previous prison reports, we assess the government's practices according to the guidelines set forth in the United Nations Standard Minimum Rules for the Treatment of Prisoners, as well as the relevant provisions of international human rights treaties to which the country is a party.

In the interests of accuracy and objectivity, Human Rights Watch strongly prefers to base its reporting on first-hand observation of prison conditions and direct interviews with prisoners and corrections officials, although we have devised an alternative research methodology for use in countries whose authorities bar access to outside monitoring. Normally, in conducting prison investigations, Human Rights Watch follows a self-imposed set of rules: investigators undertake visits only when they, not the authorities, can choose the institutions to be visited; when they can be confident that they will be allowed to talk privately with inmates of their choice; and when they can gain access to the entire facility to be examined. These rules ensure that investigators are not shown "model" prisons, "model" prisoners, or the most presentable parts of the facilities under investigation. If entry on these terms is denied, Human Rights Watch reports abuses based on interviews with former prisoners, prisoners on furlough, relatives of inmates, lawyers, prison experts, and prison staff, as well as on documentary evidence.

In Latin America, fortunately, our researchers are normally granted generous access to the penal facilities of their choice. Most Latin American countries have commendably open policies with regard to outside monitoring of prison conditions, with only a few states such as Cuba and Peru remaining hostile to such scrutiny. Indeed, it is perhaps the relative openness of the region's prison systems-not only to human rights organizations but also to humanitarian, religious, and community groups-that stands out as their most notable virtue.

Brazil, with its democratic political structure and official government support for human rights, would appear to present a favorable environment for human rights monitoring. We found, nonetheless, that gaining access to the country's prisons and police lockups was surprisingly difficult. Our researchers faced obstructions ranging from overt denials of access to, more frequently, unnecessary and time-consuming procedural hurdles. Negotiating the terms and conditions of access required countless faxes, dozens of phone calls, and long hours in meetings with state authorities-time that might have been more profitably spent observing conditions in the prisons. Morever, on several occasions our researchers were left with the distinct impression that the sole point of some of these procedural hurdles was to slow our research. Hindering our access in this way rather than simply denying us entry to the prisons does, of course, allow officials to continue to maintain a facade of sympathy for human rights concerns.

To some extent, we acknowledge, our difficulties in obtaining access were a consequence of Brazil's federal system of governance and, in particular, its system of state control of penal matters. Unlike certain other countries whose prison conditions Human Rights Watch monitors, Brazil does not have a centralized prison authority with the power to grant requests for access andfacilitate visits. Rather than a national prison system, each state manages its own set of prisons, jails, and police lockups. Moreover, as this report will describe, responsibility for and oversight of penal matters is divided among a variety of state actors that, individually or collectively, enjoy discretion with regard to granting access. Conducting a national survey of conditions therefore required us to contact a whole host of government authorities from whom our requests for cooperation elicited a considerable diversity of responses. Some authorities took pains to ensure the success of our prison visits; others failed even to respond to our phone calls, or were openly antagonistic. Possibilities for any given visit, moreover, often rested on the cooperation of the authorities directly in charge of the institution: that is, the prison director and guards. In one São Paulo police lockup, for example, the commander refused to allow us to speak to prisoners, responding to the judicial order we showed him by insisting, "I don't care if the President of the Republic personally authorized your visit, I'm the one in charge here!"7

In the end, of all of the states whose penal systems we attempted to investigate, only three-Amazonas, Brasília and Rio Grande do Norte-were entirely open to monitoring. The minister of justice of the state of Amazonas, in particular, was extremely receptive to our interest in the prisons under his jurisdiction and unfailingly helpful in ensuring our free and unconstrained access to examine conditions and meet privately with prisoners. On the other side of the spectrum, the state of Rio de Janeiro wholly barred our entry; indeed, the prison authorities there never even acknowledged our phone calls and faxes.

Perhaps the most dismaying denial of access occurred at Roger prison in Paraíba state. In the months prior to our visit there, a total of eleven prisoners died in three separate incidents, the most bloody of which involved the brutal torture of several prisoners before they were killed. Having met with the coroner who examined the bodies of the dead inmates, we were anxious to interview the surviving eyewitnesses to these killings. When we arrived at Roger prison, however, we encountered three local public defenders-lawyers charged with providing free legal assistance to prisoners-who wanted to block our entry.8 Inparticular, they did not want us to be allowed to speak privately with prisoners, regardless of whether the prisoners themselves wanted to speak with us.

Unbeknownst to us, the previous day the public defenders had filed a five-page motion in state court for a "preventive habeas corpus" to bar our prisoner interviews. Describing these interviews as "interrogations," the motion purported to protect the prisoners' right to counsel. Before we had the opportunity to enter the prison, a judicial order arrived granting the motion. The order stated, however, that prisoners had the right to the assistance of a public defender during our interviews, not that a public defender had to be present. Accordingly, prisoners could, if they preferred, waive their right to counsel and choose to have private interviews with us. Despite the order's clear wording, the prison authorities, following the lead of the public defenders, did not permit this possibility, but instead interpreted the order as requiring the public defenders to be present during all of our prisoner interviews. Because, as described above, confidentiality is a basic rule of our inmate interviews, we were forced to leave the prison without speaking to numerous eyewitnesses to the killings.9 (Fortunately, we had already interviewed prisoner eyewitnesses who had previously been transferred to a different prison.)

Despite obstacles such as these, Human Rights Watch gathered the bulk of the information contained in this report during detailed inspections of the country's prisons, jails, and police lockups-of which we saw a total of some forty facilities-and extensive interviews with prisoners. We spent a day at nearly every prison and jail we visited, and several hours at every police lockup. During our visits, we normally viewed the entire facility, including punishment cells and other segregation areas, the infirmary, the kitchen, the recreation areas, the bathrooms, and, of course, the prisoners' living quarters. Some of our discussions with prisoners were conducted in informal groups, as we walked with prisoners through the cell blocks, but a great many of them were one-on-one interviews held outside of anyone else's hearing.

Finally, numerous documentary sources provided additional information on Brazil's prison policies, rules and conditions. Such sources included the national prison census, relevant newspaper articles, case materials from the prosecution of abusive police and prison guards, coroners' reports, academiccommentaries on Brazil's prison laws, and abundant written materials provided by the National Prison Ministry.10

International Human Rights Standards
Governing the Treatment of Prisoners

The chief international and regional human rights documents binding on Brazil clearly affirm that human rights extend to persons who are incarcerated. The International Covenant on Civil and Political Rights (ICCPR), the Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, and the American Convention on Human Rights, all of which Brazil has ratified, all prohibit torture and cruel, inhuman, or degrading treatment or punishment, without exception or derogation.11 Both the ICCPR and the American Convention require that "the reform and social readaptation of prisoners" be an "essential aim" of imprisonment.12 They also mandate that "[a]ll persons deprived of their liberty shall be treated with respect for the inherent dignity of the human person."13

Several additional international documents flesh out the human rights of persons deprived of liberty, providing guidance as to how governments may comply with their obligations under international law. The most comprehensive such guidelines are the United Nations Standard Minimum Rules for the Treatment of Prisoners, adopted by the Economic and Social Council in 1957. Other relevantdocuments include the Body of Principles for the Protection of All Persons Under Any Form of Detention or Imprisonment, adopted by the General Assembly in 1988, and the Basic Principles for the Treatment of Prisoners, adopted by the General Assembly in 1990. It is worth noting that although these instruments are not treaties, they provide authoritative interpretations as to the practical content of binding treaty standards.14

These documents reaffirm the tenet that prisoners retain fundamental human rights. As the most recent of these documents, the Basic Principles, declares:

Except for those limitations that are demonstrably necessitated by the fact of incarceration, all prisoners shall retain the human rights and fundamental freedoms set out in the Universal Declaration of Human Rights, and, where the State concerned is a party, the International Covenant on Economic, Social and Cultural Rights, and the International Covenant on Civil and Political Rights and the Optional Protocol thereto, as well as such other rights as are set out in other United Nations covenants.15

Endorsing this philosophy in 1992, the United Nations Human Rights Committee explained that states have "a positive obligation toward persons who are particularly vulnerable because of their status as persons deprived of liberty" and stated:

[N]ot only may persons deprived of their liberty not be subjected to [torture or other cruel, inhuman or degrading treatment or punishment], including medical or scientific experimentation, but neither may they be subjected to any hardship or constraint other than that resulting from the deprivation of liberty; respect for the dignity of such persons must be guaranteed under the same conditions as for that of free persons. Persons deprived oftheir liberty enjoy all the rights set forth in the [ICCPR], subject to the restrictions that are unavoidable in a closed environment.16

Significantly, the Human Rights Committee has also stressed that the obligation to treat persons deprived of their liberty with dignity and humanity is a fundamental and universally applicable rule, not contingent on the material resources available to the state party.17

1 See, for example, Human Rights Watch, Police Brutality in Urban Brazil (New York: Human Rights Watch, 1997).

2 Alba Zaluar, "The Drug Trade, Crime and Policies of Repression in Brazil," Dialectical Anthropology, vol. 20, 1995, p. 96.

3 Túlio Kahn, "Table of Contents de Criminalidade," Revista do Ilanud, No. 2, 1997.

4 Brazil functions primarily as a transit country: cocaine produced in Colombia, Bolivia, and Peru enters Brazil by land, air, and river; it is then shipped from major transportation hubs such as Rio de Janeiro and São Paulo to consumer markets in Europe and the United States. There is also an important internal market for cocaine, largely among the urban middle and upper classes and tourists.

5 United States Department of State, Bureau for International Narcotics and Law Enforcement Affairs, "International Narcotics Control Strategy Report," March 1995, p. 73.

6 As we wrote in "Brazil: Fighting Violence with Violence: Human Rights Abuse and Criminality in Rio de Janeiro, " A Human Rights Watch Short Report, Vol. 8, No. 2(B), January 1996, p. 19:

Under Brazilian law (Law No. 7.960/89 and Law No. 8.072/90), a judge may order a person's preventive detention for up to thirty days if there are"fundadas razões"-the rough equivalent of probable cause in Anglo-American law-concerning the person's participation in one of several enumerated crimes, including drug trafficking. According to the terms of the law, such detention is justified only when necessary to assist the conduct of an official police investigation . . . . In practice, however, during Operation Rio persons were detained for thirty days even though no criminal investigations were opened.

7 Human Rights Watch interview, Darci Sassi, sixteenth police precinct, São Paulo, November 19, 1997.

8 We should note, however, that the head of the local public defender's office, who was present during part of this incident, told us that the most vocal of the three public defenders had in fact been suspended from his duties as of one month previously. In other words, he was not acting in any official capacity. Human Rights Watch interview, Gilberto Sarmento, Roger prison, João Pessoa, Paraíba, December 11, 1997.

9 Four months after we were barred, researchers from Amnesty International who were attempting to visit Roger prison met with a similar denial of access. Given the number of deaths that have occurred there in the past year, the authorities' consistent refusal to permit outside monitoring is extremely troubling.

10 The following sources were of particular note: Ministério da Justiça, Conselho Nacional de Política Criminal e Penitenciária, Censo penitenciário de 1995 (Brasília: Conselho Nacional de Política Criminal e Penitenciária, 1997); Conferência Nacional dos Bispos do Brasil, A fraternidade e os encarcerados: Cristo liberta de todas as prisões (São Paulo: Editora Salesiana Dom Bosco, 1997); Deputado Wagner Lino, Relatório Final da Comissão Parlamentar de Inquérito sobre os Estabelecimentos Prisonais do Estado de São Paulo (São Paulo: Assembléia Legislativa do Estado de São Paulo, 1996); and Julio Fabbrini Mirabete, Execução penal: comentários a Lei _ 7.210, de 11-7-84 (São Paulo: Atlas, 1997).

11 The Brazilian Constitution similarly provides, "No one shall be subject to torture or inhuman or degrading treatment." Brazilian Constitution, art 5, sec. 3.

12 American Convention, art.5(6); accord ICCPR, art. 10(3) ("The penitentiary system shall comprise treatment of prisoners the essential aim of which shall be their reformation and social rehabilitation.").

13 American Convention, art. 5(2); accord ICCPR, art. 10(1) ("All persons deprived of their liberty shall be treated with humanity and with respect for the inherent dignity of the human person.").

14 See, for example, the U.N. Human Rights Committee's decision in Mukong v. Cameroon, in which it cites various violations of the Standard Minimum Rules in ruling that the complainant was subject to cruel, inhuman and degrading treatment. Mukong v. Cameroon (No. 458/1991) (August 10, 1994), U.N. Doc. CCPR/C/51/D/458/1991.

15 Body of Principles, art. 5.

16 U.N. Human Rights Committee, General Comment 21, paragraph 3. The Human Rights Committee, a body of experts established under the ICCPR, provides authoritative interpretations of the ICCPR though the periodic issuance of General Comments.

17 Ibid., paragraph 4; see also Mukong v. Cameroon (No. 458/1991) (August 10, 1994), U.N. Doc. CCPR/C/51/D/458/1991 (stating that minimum requirements regarding floor space, sanitary facilities, provision of food, etc., must be observed, "even if economic or budgetary considerations may make compliance with these obligations difficult").

click Table of Contents click