Brazil

• Introduction and background
• Part 1: Legal frameworks
• Part 2: Policies and procedures
• Part 3: Comparative indicators
• Summary and recommendations
• References and data sources

Introduction

Abstract

Brazil is a recent democracy, still influenced by its dictatorial period of authoritarianism. The legacies of that period are exceptionally noticeable in the public safety sector, as the abusive use of lethal force by the police is rampant in the country. In 2020, 2.77 civilians were killed by on-duty Law Enforcement Agents (LEAs) using a firearm for every 100 thousand inhabitants. Deaths caused by the police using a firearm were 11.8% of the total number of violent intentional deaths in the country, and 11.28 civilians were killed by on-duty LEAs using a firearm for every single agent killed by gunshot while on duty, demonstrating there is little room for confrontations and strong evidence of extrajudicial use of lethal force.

The Public Prosecutors Office is the organ responsible for police oversight, but it is not very active. The Public Prosecutors Office can decide to either report or archive a case of death caused by a police officer. When a case is reported by the Public Prosecutors Office, it goes to trial — but when it does not, the process reaches its end. Research shows the Prosecutors Office stops over 90% of the police lethality cases from ever undergoing trial in the biggest states of the country (namely, Rio de Janeiro and São Paulo). Meanwhile, although official data production is abundant, most of it remains unpublicized or difficult to obtain, with no evidence that standardization between the 96 police forces is likely to be established in the foreseeable future. Moreover, statistics are perceived by the Brazilian Executive as means to the end of legitimating the dysfunctional anti-democratic weaponizing of the State apparatuses, as lethal force serves as a tool for controlling the poor. As such, statistical fluctuations have rarely ever been perceived as policy-guiding indicators for the need to adjust or change public policies towards more democratic, transparent, and controlled police forces.

Background

Brazil is a federative republic. As such, policy, polity and politics are arranged in a decentralized manner, with different competences and attributions allocated to each level of power. When it comes to the public safety sector, most of the attributions and competences are allocated on the state level.

As a result, data production, data systematization and data publicization are all done by each of the 27 Federation Units on the state level autonomously. Since every state level Federation Unit has its own process and political arenas and agendas, the transparency/opacity of each one varies with time, according to conjunctural and structural circumstance, such as staffing, administrative and technological capacity, as well as political interests of their current governments. The data regarding Public Safety Personnel references 2021.

Government Level	Police Agencies	Legal Competences	Personnel (as of 2021)
National	Federal Police (PF)	Investigating criminal offences against the political and social order or to the detriment of property, goods, services and interests of the Federal or its autonomous entities and public companies, as well as other offences the commission of which has interstate or international repercussions and requires uniform repression, as provided by law. Suppressing illicit trafficking of narcotics and related drugs, as well as smuggling and embezzlement, without detriment to the action of the Treasury and other public bodies in their respective areas of competence. Policing maritime, airspace and land borders. Enacting judicial policing at the Union level.	13,795
	Highway Federal Police (PRF)	Ostensive patrolling of federal highways.	11,575
	Federal Penitentiary Police (PPF)	Enforcing federal penitentiary establishments’ security.	1,000
	Railroad Federal Police (PFF)	Ostensive patrolling of federal railroads.	189
	Legislative Police Departments (DEPOL)	Preserving the public order and public property, as well as preventing and ascertaining criminal offenses committed in the National Congress’ buildings and external areas’ jurisdiction.	459
Statal (26 States plus the Federal District)	Polícia Militar (PM)	Ostensive policing and preserving the public order.	411,241
	Polícia Civil (PC)	Enacting judicial policing and investigating non-military criminal offenses.	109,440
	Polícia Penal (PP)	Enforcing State and district-level penitentiary establishment’s security.	98,248
Sum of all 86 police forces (5x national and 27×3 federative unit):			645,947

Part 1: Legal Frameworks

Global treaties

1966 International Covenant on Civil and Political Rights (ICCPR)		State party
ICCPR Optional Protocol 1		State party
1984 Convention Against Torture (CAT)		State party
CAT committee competent to receive individual complaints?		Yes
CAT Optional Protocol 1		State party

Regional treaties

1948 Charter of the Organization of American States		State party
1969 Inter-American Convention on Human Rights		State party
Competence of Inter-American Court on Human Rights		Yes

National legal provisions

Relevant constitutional provisions or general laws

Brazil’s 1988 Constitution enshrines the right to life and to physical integrity. No one shall be subjected to torture or to cruel, inhuman, or degrading treatment or punishment. Furthermore, will be held in custody unless there is a warrant for arrest issued by a judge with the requisite jurisdiction or unless the person was arrested in the process of committing a crime.

The Brazilian Constitution establishes that the Military Police are responsible for the preservation of public order and for policing.

Relevant specific national legislation

Article 25 of the Brazilian Penal Code states that anybody is justified in using, within reason, the means necessary to repel an unfair imminent or ongoing aggression against themself or another person.

The 1941 Criminal Procedure Code establishes the principle of use of force in self-defence, only to be used in case of ‘resistance’ where law enforcement officials can “use whatever means necessary to defend themselves or to overcome resistance”, or to prevent the escape of a prisoner.

A 2014 federal statute calls for security agencies to prioritise the use of less-lethal weapons, provided that their use does not endanger the physical or mental integrity of police officers. Use of lethal force must comply with the principles of legality, necessity, and reasonableness and proportionality. It is not lawful to use firearms against an unarmed fleeing person, or against a vehicle that disrespects a police blockade on public roads, except when there is a risk of death or injury to law enforcement agents or third parties.

Relevant national regulations

The Interministerial Ordinance n. 4226, of 31 December 2010, signed by the Justice Department and Human Rights Department, provides specific regulations regarding the use of force by all the police forces in Brazil. Made up of 25 guidelines, the document states the use of force must be compliant with the principles of legality, necessity, proportionality, moderation and convenience. It states police use of firearms is illegitimate except in cases of legitimate defense of their own or a third person’s life, and demands that they otherwise refrain from resorting to it. The ordinance also requires that a law enforcement officer carry at least two less-lethal weapons. It also defines the types of weapons and techniques that are authorised, depending on technical circumstances.

Legal interpretation and application

UN or other international body decisions or advisory opinions

The UN Committee against Torture expressed deep concern “about the persistent use of excessive force, especially the use of lethal force, by law enforcement and military officials in the context of security operations to combat organized crime” in 2023, and called for “urgent measures to end the use of excessive force, especially the use of lethal force, by law enforcement and military officials”¹.

Also in 2023, the UN Human Rights Committee expressed concerns about excessively lethal policing, disproportionately affecting LGBT and nonwhite people.

Furthermore in 2023, a UN Special Rapporteur affirmed Brazil’s long history of violence against peaceful demonstrations and of criminalizing social movements.

Regional court judgments

The cases in which Brazil has been condemned by the Interamerican Court of Human Rights regarding Law Enforcement Operations include:

1. Nogueira Carvalho and other v Brazil
2. Escher and others v Brazil
3. Gomes Lund and others (Araguaia guerilla) v Brazil
4. Cosme Rosa Genoveva, Evandro de Oliveira and others (“Favela Nova Brasília”) v Brazil
5. Vladmir Herzog v Brazil

National court judgments

The Federal Supreme Court (STF) ordered a suspension of police operations in Rio de Janeiro communities during COVID, except in “absolutely exceptional cases”. Deaths fell significantly until October 2020, as police operations intensified despite the Supreme Court’s decision. In the same case, the plenary of the STF ordered the submission of a plan to reduce police lethality and to control human rights violations by the security forces, to create a judicial observatory on citizen policing within the National Council of Justice (CNJ), and to install GPS equipment and audio and video recording systems in police vehicles and uniforms².

In July 2013, the construction worker Amarildo Dias de Souza, a habitant of the Rocinha favela, was taken by the Rio de Janeiro Military Police to a UPP unit (Unit of Pacifying Police) for questioning. He was never seen again. His family pressed for action and after three months of investigation the Rio de Janeiro´s Public Prosecution Service initiated a prosecution against 25 military policemen for the homicide of Amarildo. Twelve of the twenty-five police officers charged were convicted of torture followed by death and of hiding his body.

When the São Paulo Military Police intervened in a 1992 dispute between detainees, they killed 111 inmates in what is now known as the “Carandiru Massacre”. Some of the police officers involved were convicted at trial, but on appeal all convictions have been annulled the convictions. This is not the final outcome, as the Supreme Court may reinstate the convictions, or new trials may be held.

Oversight bodies

Military courts have jurisdiction over actions by the Military Police, which is in charge of public order.

In São Paulo state, a Special Action Group for Public Security and External Control of Police Activity (GAESP) was created in 2022 within the Public Prosecutor’s Office as an external control of police activity (including municipal guards), with a broad remit and aiming at the collective protection of public security through administrative and judicial measures.

In Rio de Janeiro state, the Public Prosecutor’s Office created the General Coordination of Public Security of the Public Prosecutor’s Office of the State of Rio (CGSP) in 2021, which has amongst it tasks to control the regularity, suitability, and efficiency of police activity.

The first Brazilian Police Ombudsman was created in São Paulo 1995 through an initiative of the São Paulo State Council of Human Rights. As of 2005, 14 of the 27 Federative Units (or states) had implemented Police Ombudsmen via State Laws and Decrees. The list of states with Police Ombudsmen and their creation year is as follows:

1. São Paulo, created by the São Paulo State Decree number 39,900, Jan. 1^st, 1995;
2. Pará, created by the Pará State Law number 5,944, Feb. 6th, 1996;
3. Minas Gerais, created by the Minas Gerais State Law number 12,622, Sep. 25th, 1997;
4. Espírito Santo, created by the Espírito Santo Complimentary Law number 297, Jul., 27th, 2007;
5. Rio de Janeiro, created by the Rio de Janeiro State Law number 3,168, Jan. 12th, 1999;
6. Rio Grande do Sul, created by the Rio Grande do Sul State Decree number 39,668 Aug. 17th, 1999;
7. Mato Grosso, created by the Mato Grosso State Law number 7,286, May 23rd, 2000;
8. Paraná, created by the Paraná State Decree number 2,026, May 9th, 2000;
9. Pernambuco, created by the Pernambuco State Decree number 22,149, Mar. 2000;
10. Rio Grande do Norte, created by the Rio Grande do Norte State Law number 7,851, Jun. 28th, 2000;
11. Goiás, created by the Goiás State Law number 14,383, Dec. 31st, 2002 ;
12. Santa Catarina, created by the Santa Catarina State Complimentary Law number 243, Jan. 30th, 2003;
13. Bahia, created by the Bahia State Decree number 7,623, Jun., 25th 1999;
14. Ceará, created by the Ceará State Law number 13,093 Jan. 8th, 2001.

Part 2: Policies and Procedures

Data collection and publication by official agencies

1. Are the number of deaths following any police use of force
	Collected?	Good, Robust
	Accessible through existing publicly available information?	Good, Robust
	Is this a legal requirement?	Good, Robust
	Can such information be requested from the authorities when not publicly available?	Good, Robust
	If one can request it, what is the likelihood this information would be released?	Good, Robust
2. If published, to what extent can the number of deaths be readily determined from official statistics?		Partial, Medium
3. Is it possible to identify specific individuals killed in official records?		Partial, Medium
4. Is demographic and other information for the deceased
	Collected?	Good, Robust
	Accessible through existing publicly available information?	Limited, Poor
	Is this a legal requirement?	No Provisions
	Can such information be requested from the authorities when not publicly available?	Good, Robust
	If one can request it, what is the likelihood this information would be released?	Good, Robust
5. Is demographic and other information on officers in use of force incidents
	Collected?	Good, Robust
	Accessible through existing publicly available information?	No Provisions
	Is this a legal requirement?	Good, Robust
	Can such information be requested from the authorities when not publicly available?	Good, Robust
	If one can request it, what is the likelihood this information would be released?	Good, Robust
6. Is information on the circumstances
	Collected?	Partial, Medium
	Publicly available?	Partial, Medium
	Is this a legal requirement?	Partial, Medium
	Can such information be requested from the authorities when not publicly available?	Good, Robust
	If one can request it, what is the likelihood this information would be released?	Partial, Medium
7. Is information about the type(s) of force used
	Collected?	No Provisions
	Accessible through existing publicly available information?	No Provisions
	Is this a legal requirement?	No Provisions
	Can such information be requested from the authorities when not publicly available?	Good, Robust
	If one can request it, what is the likelihood this information would be released?	No Provisions

Data quality of official sources

8. How reliable are the sources used to produce official statistics about deaths?		Good, Robust
9. Are there mechanisms for internal quality assurance / verification conducted		Good, Robust
10. Is the methodology for data collection publicised?		Good, Robust
11. How reliable are the overall figures produced?		Good, Robust

Data analysis and lessons learnt

12. Do state or police agencies analyse data on the use of lethal force, to prevent future deaths?		No Provisions
13. Is there evidence that state/ police agencies act on the results of their analysis, including applying lessons learnt?		No Provisions
14. Are external bodies are able to reuse data for their own analyses?		Good, Robust
15. Do external, non-governmental agencies collect and publish their own statistics on deaths following police use of force?		Good, Robust

Investigations by official agencies

16. Is there a legal requirement for deaths to be independently investigated?		Good, Robust
17. If so, which organisation(s) conduct these investigations?		The Civil Police, and sometimes the Office of the Public Prosecutor
18. In the year in question, how many deaths following police use of force have been investigated by the organisation(s) specified in question 17?		Data unavailable
19. Are close relatives of the victims involved in the investigations?		Limited, Poor
20. Investigation reports into deaths
	Are publicly available?	Partial, Medium
	Give reasons for the conclusions they have reached?	Partial, Medium
	Is this a legal requirement?	No Provisions
	Can such information be requested from the authorities when not publicly available?	Good, Robust
	If one can request it, what is the likelihood this information would be released?	Partial, Medium
21. Is there information available on legal proceedings against agents / officials, pursuant to deaths?		Limited, Poor
22. Is there information available on legal proceedings against state agencies, pursuant to deaths?		Limited, Poor
23. Is there information available on disciplinary proceedings against agents/ officials, pursuant to deaths?		Partial, Medium
24. Number of prosecutions against agents / officials involved in the last ten years?		Data unavailable
25. Number of convictions against agents / officials involved in the last ten years?		Data unavailable
26. Number of prosecutions against agencies involved in the last ten years?		Data unavailable
27. Number of convictions against agencies involved in the last ten years?		Data unavailable
28. Number of cases in which states have been found to have breached human rights law on the use of lethal force?		Data unavailable

Detailed elaboration

Data collection and publication by official agencies

The number of deaths following any police use of force can be requested by anyone, any time via Access to Information Law. Requests regarding deaths following any police use of force are regularly answered, even though such official data might not be actively publicized by its holders. The Public Safety Departments of Amapá, Acre’, Rondonia, and Roraima all have their own websites, none of which contain any data regarding Public Safety statistics.

Data quality of official sources

Most states possess data regarding deaths due to police interventions at this point³. On the other hand, data on law enforcement agents killed, either on or off-duty, have little coverage and pervasiveness. The consolidated data tends to be the most trustworthy, as the more specific the data selection on the microdatabases, the more information is lost⁴. This phenomenon is aggravated by the fact that each state has its own database, but these are not compiled according to common standards.

As a result, the information is recorded differently in each state. The states’ databases also have different coverage, and some of them are of higher quality than others.

Data analysis and lessons learned

Unfortunately, the data produced by each of the State’s Public Safety Departments and analyzed yearly by the Brazilian Forum on Public Safety, brought up no novelties regarding the pattern of use of force, nor the victims’ profile. The use of force by the police is highly uneven throughout the country, with less than half the states (Amapá, Bahia, Goiás, Pará, Rio de Janeiro and Sergipe, i.e, 6 states out of 27) concentrating most (66%) of the deaths due to police intervention (regarding data on 2022). The most important and unexpected lesson learnt through the data analysis performed by the Brazilian Forum on Public Safety year after year while producing our multiple editions of the Yearbook on Public Safety, has been noticing how difficult it is for the press to cover a statistically significant number of cases in a country as violent and as continental in size as Brazil.

Investigations by official agencies

In general, the Civil Police tend to neutralize the responsibility of officers involved in cases of use of force resulting in deaths, rendering them into legal jargon, as well as directing the investigations towards the (sometimes alleged) previous criminal past of the victims instead of their killing.

Legal proceedings, prosecutions and convictions

In general, the Civil Police tends to neutralize the responsibility of officers involved in cases of use of force resulting in deaths, rendering them into legal jargon, as well as directing the investigations towards the (sometimes alleged) previous criminal past of the victims instead of their killing.

There is little data production on prosecutions and convictions in Brazil. Research produced by the Brazilian Forum on Public Safety demonstrated that the Prosecutors Offices of the States of São Paulo and Rio de Janeiro took forward as little as 10% of the cases of deaths due to police interventions, archiving the remaining 90%⁵.

Non-official sources

We have conducted multiple instances of compilations of data on police use of force using data from the press for the Monitor del Uso de la Fuerza Letal en America Latina initiative. Each attempt made it clear that data from the press was less pervasive than official data⁶.

Part 3: Comparative indicators

I-1a. CK: Number of civilians killed by law enforcement agents on duty, by gunshot	5619
I-1b. CKt: Number of civilians killed by law enforcement agents, regardless of means and whether or not on duty	6429
I-1c. CW: Number of civilians wounded by law enforcement agents on duty, by gunshot	Data unavailable
I-1d. CWt: Number of civilians wounded by law enforcement agents, whether or not on duty and regardless of means	Data unavailable
I-2. CK per 100000 inhabitants	2.27
I-3. CK per 1000 law enforcement agents	11.28
I-4. CK per 1000 arrests	Data unavailable
I-5. CK per 1000 weapons seized	53.03
I-6. AK: Number of law enforcement officers unlawfully killed on duty by firearm	22
I-6b. AKt: Number of law enforcement officers unlawfully killed, whether or not on duty and regardless of means	161
I-7. AK per 1000 agents	0.05
A1. Percentage of homicides due to state intervention	11.8
A2. Ratio between CK and AK	251.81
A3. Civilian lethality index: Ratio between CK and CW	Data unavailable
A4. Lethality ratio: Ratio between Civilian lethality index and law enforcement agents lethality index	Data unavailable
A5. Average number of civilians killed by intentional gunshot, per incident	Data unavailable

The table provides comparative indicators for deaths following police use of force, benchmarked against various figures for the population as a whole. The data regarding use of force references 2022, and was gathered via the Brazilian equivalent of the Freedom of Information Act, by sending requests to all of the states’ Public Safety Departments. The data regarding Public Safety Personnel references 2021, and was gathered by sending FOIA requests to all of the states’ Public Safety Departments, the Transparency Portal of the Federal Government, the Federal Police Department, the Highway Federal Police Department, National Penitentiary Department, and the Justice and Public Safety Ministry. The collected data was then consolidated by the Forum Brasileiro de Segurança Pública (Brazilian Forum on Public Safety)⁷.

In compiling this table, the main deviation from the comparative indicators’ wording is that for Brazil the category of violent intentional deaths was used instead of total homicides. Since deaths caused by police officers are pre-emptively protected under the law until proven otherwise, they are technically not homicides. As such they constitute their own category of deaths within official nomenclature in Brazil. As a result, within this table the reference to total homicides aggregates both victims of initially non-criminal (police-caused) and criminal violent intentional deaths. Both are treated as violent intentional lethal crimes.

Number of civilians killed and injured

Note: Since the microdata⁸ didn’t contain information on whether Law Enforcement Agents who killed were on or off-duty, but did contain info on whether the victims had been killed with a firearm or not, we multiplied the percentage of victims killed by police officers using firearms (90%) by the number of civilians killed by on-duty Law Enforcement Agents.

Summary and recommendations

Brazil has an accountable legal framework that allows for any citizen to request any information of the government at any time, without having to state their reason for such request. However, while there is good compliance with Freedom of Information Law, the quality of public safety information varies significantly from state to state.

Since every state has its own information system and there is no standardisation in the way data is collected, it is difficult to aggregate statistics and compare police forces. The Federative Union provides a nationwide universal public safety system (Sistema Único de Segurança Pública – SUSP), however it has never been implemented ever since its inception, in 2018.

The use of lethal force in Brazil could be improved through the following measures:

• Releasing information pertaining to police activity, such as number of people stopped and searched by the police and their profile, number of people hurt in police interventions, etc. Such information could be provided either by the Civil and Military Polices or the Public Safety Department of each state;
• Providing information via active transparency, i.e. by making information on police use of force readily available for citizen consultation. Such information could be provided either by the Civil and Military Police or the Public Safety Department of each state;
• Improving coverage on demographic data regarding victims and authors of police lethal use of force. Such improvement could be provided by the Civil Police by having their officers fill out the forms correctly.
• Releasing data on the Public Prosecutor’s Office’s processing of police lethality cases.

Such information could be provided by the Public Prosecutor’s Office of each state.

• Implementing SUSP in order to standardise and centralise information.

References, data sources and downloads

The above report is slightly updated from the Brazil Lethal Force Monitor report (PDF, 150KB) (2024). There are a few links to www.policinglaw.info not in the PDF, as well as edits for clarity.

1 Source: The “Global caselaw” section of https://www.policinglaw.info/country/brazil

2 Source: The “Regional caselaw” section of https://www.policinglaw.info/country/brazil

3 The Brazilian Forum on Public Safety’s experience demonstrated that constant information access requests, with media coverage on the states’ transparency has made the states progressively more cooperative and transparent.

4 A microdatabase consists of a database formed by data in which each line stands for one observation. Therefore, consolidated data is such bases’ data grouped and filtered. A microdatabase of deaths due to police intervention, for an instance, presents each victim killed as one of its constituent lines. The total number of deaths due to police intervention on the other hand, needs to be extracted and group by performing different operations on the microdata, thus being a form of consolidated data.

5 For more information, please check: https://noticias.uol.com.br/cotidiano/ultimas-noticias/2021/11/12/denuncias-e-arquivamentos-mortes-policiais-rj-sp.htm

6 Cano, Ignacio; Forné, Carlos, Silva; Correa, Catalina Pérez. (2024) Monitor of Use of Lethal Force in Latin America and the Caribbean: A Comparative Study of Brazil, Chile, Colombia, El Salvador, Jamaica, Mexico, Peru, Trinidad and Tobago and Venezuela. Available at: https://monitorfuerzaletal.com/docs/MFL2024_Regional_Report.pdf.

7 Documents available from: https://forumseguranca.org.br/

8 See footnote 4 above.