How an innocent man dies in Rio de Janeiro

Country/area: Brazil

Organisation: Época

Organisation size: Big

Publication date: 27 Jan 2020

Credit: Rafael Soares

Project description:

The report “How an innocent man dies in Rio de Janeiro”, published in eight pages of Revista Época in January 2020, is the result of a four-month journalistic investigation. Reporter Rafael Soares analyzed, one by one, the investigations of all 195 homicides committed by the Rio de Janeiro police in the month with the highest number of occurrences in the state’s history – July 2019 – and discovered cases with serious indications of police errors and executions summary.

Impact reached:

The result of the journalistic investigation contrasted the speeches of the federal and Rio governments. Both President Jair Bolsonaro and Governor Wilson Witzel were elected by replicating the speech that “greater legal protection” was needed for the police officer who killed on duty. The speech was accompanied by an increase in the number of deaths in operations in Rio, which led to a peak of 195 cases in July 2019 – to date the highest number in the history of the state. However, the serious failures in homicide investigations and evidences of executions revealed by the report opened wide that, unlike the official version, the officers that kill on duty are protected by the criminal justice system.

The report served as the basis for a series of requests made to the Supreme Federal Court (STF) by political parties and civil society organizations in the Fundamental Precept Failure Statement (ADPF) 635, which questions Rio’s security policy and operations in favelas during the pandemic of coronavirus. One of the requests was inspired by the result of the report: the one that claimed that investigations into homicides committed by state agents should be carried out, independently, by the Prosecutor’s Office. The request was unanimously accepted by the STF plenary on August 18th.

Techniques/technologies used:

The journalistic investigation was divided into three stages. The first was the data collection. The reporter obtained access to data of all 151 police investigations (a police inquiry may investigate more than one murder)on homicides committed by officers on duty on July 2019 through the Freedom of Information Act (FOIA). Information such as the investigation’s serial numbers, locations of all investigations, investigated officers, units in charge of the investigation and actual stage of the proceedings were provided by the police.

Then, began the second stage: the documental research. In order to gain access to other important documents of the investigations – such as forensic reports, for example – regular visits to the Homicide Unit and the Prosector’s Office were necessary. By the end of this stage, the reporter was able to collect detailed information on the majority of the investigations and could produce an Excel table with the data. The objective was to understand both if there were indications of irregularities in the police actions as well as to point out flaws in the investigations (lack of expertise at the crime scene or testimonies from victims’ relatives, for example).

The thirdd stage was the field work. The reporter selected eight cases with serious  indications of police misconduct and collected over a month testimony from relatives of victims, witnesses to the crimes and members of the criminal justice system involved in the cases – investigators, prosecutors, public defenders, lawyers and experts. This new stage was intended to advance in relation to the official investigation, hearing new witnesses, and to outline a more detailed profile of the victims.

What was the hardest part of this project?

The hardest part of the project was to overcome the lack of transparency of law enforcement agencies in Brazil. Till 2018, Rio de Janeiro police didn’t respond to FOIA requests. Since then, the answers are rare and often incomplete. Because of that, this project of analizing all investigations on police killings in a month had to be postponed for over two years. The reporter came up with this project’s idea in January 2018, but by that time police didn’t provide the requested data.

For two years, I interviewed lawyers, police officers and prosecutors to understand how worked the digital system of Rio’s police and Prosecutor’s Office that controls the progress of investigations. My goal was to write better requests to get the data. I learned that this sistem is able to filter several informations of each inquiry and it also generates csv editable files. And above all I learned that this files were not protected by any kind of confidenciality. In September 2019, I received for the first time the archives. I started to make the same request every month not to miss any update.

The second stage of the investigation was also a challenge. For two months I had to convince more than a dozen of sources to gave me access to the documents. As the majority of the investigations were ongoing, prosecutors and investigators were not able to provide copies of the investigations. In a lot of cases, I had to copy the content of the documents by hand. In the end of that process, I had acces to inquiries of 135 of the 195 murders committed by Rio police in July 2019.


What can others learn from this project?

This project is a good example of how to combine techniques of tradicional reporting — such as field work, interviews with sources and documental research — with data journalism’s methodology and analysis to overcome an opaque government. In countries like Brazil, where transparency is not a rule followed by governments and access to public data can be hard and scarce, it’s a challenge to get big amounts of data, specially from law enforcement agencies.

In this case, the main goal of the reporter was to analyze how the police investigates itself in the brazilian state with the biggest number of police killings. Therefore, it was fundamental to get information about all inquiries.

If Rio’s police provided all the data (the complete investigations of police killings or more detailled data about each inquiry), it would be a lot easier to analize the material and come up with findings. But as it didn’t happened it was necessary to create a methodology to get the data and then to organize it.


Through the FOIA requests, I was able to get basic information from each inquiry. Then I started to get other necessary data in multiple others sources (documents, interviews, field work). In the end, every information of each inquiry was put in a database, where I could segment and rank everything.

In the end, this project taught me the importance to rely not only on one but on multiple sources to build an journalistic investigation.   

Project links: