How Agência Pública Found the Descendants of Alleged Slave Owners in Brazil

This report was originally published in Portuguese in Agência Pública.

Este relatório foi originalmente publicado em português na Agência Pública.

Investigative project included academic researchers and journalists in a quest for historical documentation in the country that received the most enslaved Africans 

In July 2023, Agência Pública, Brazil’s largest nonprofit newsroom and most awarded investigative news agency, launched an ambitious project to map the genealogical roots of Brazil's main political leaders. Inspired by the Slavery’s Descendants special published by news agency Reuters in June, our investigation aimed to identify which Brazilian authorities —Senators, Governors, and Presidents—have family ties to former slave owners. The goal was not just to stir up the past but to document direct connections to slavery through genealogical sources and historical records, verifying the impact that the slave economy has still today in the country that received more African slaves than any other.

Slavery in Brazil unfolded differently than in other countries. It lasted over 300 years, and enslaved labor was extensively used throughout the territory. An estimated 5.8 million slaves were trafficked between the 1530s and the mid-1800s. 

---

As a nonprofit journalism organization, we depend on your support to fund more than 170 reporting projects every year on critical global and local issues. Donate any amount today to become a Pulitzer Center Champion and receive exclusive benefits!

---

Access to documents from this period follows its own logic. In the United States, there is a structured national census form with information about the enslaved, which the Reuters team used to verify slave owners. In Brazil, such information is dispersed, making it impossible to rely on a single source for this type of verification.

Due to these and other particularities, our team realized that following the steps of foreign journalistic investigations would not suffice to map the genealogy of high-ranking members of power in Brazil. We needed to develop a methodology tailored to our reality.

Our first step was to seek out those with theoretical expertise on methodologies and sources for this type of research, namely, specialists in genealogy. We assembled a multidisciplinary team consisting of two journalists and four researchers from the Federal University of Paraná (UFPR), who dedicated a year to genealogical research. As Agência Pública's data analyst, I was tasked with coordinating this team. We also received support from researchers in genealogy, archival science, and history from various institutions, including the University of Campinas (Unicamp), the Federal University of Alagoas (UFAL), and other academic and research institutions.

Before even outlining the steps we needed to follow in the genealogical research, we had to answer some fundamental questions to assess the project's feasibility. One of the most crucial was: Do we have public databases on enslaved people and enslavers in Brazil?

The answer is yes, they exist. Between July and October 2023, we conducted extensive mapping of these sources and compiled a documentary database for the research. This database includes documents such as nominative lists of inhabitants or population bundles, which are 19th-century population censuses, and post-mortem inventories, which detailed the assets left by deceased individuals, including enslaved people, on the properties of the elite of the time, akin to modern wills.

We also included baptism, birth, and marriage records stored in notary offices, digital archives, and the FamilySearch website, which ultimately became our main research tool, as I will detail later. Additionally, we consulted public archives and registries from various states during the investigation. Our source database also incorporates digital tools such as DivulgaCand (a database of candidates from the Superior Electoral Court that gathers information on Brazilian candidacies) and Google Scholar. We also utilized the collaborative database SlaveVoyages, which documents ships and routes of the transatlantic slave trade.

Among the 116 politicians whose pasts were traced, we found 33 officials with links to enslavers — more than 28%.

For this investigation, Agência Pública defined a sample of Brazilian politicians and authorities, including senators from the 57th Legislature (2023–2027), sitting governors, and presidents since Brazil's re-democratization. The list includes a total of 116 names.

We documented over 200 family relationships and examined around 500 documents. Our focus was on tracing direct links to slaveholders in the genealogical trees—only direct ancestors such as parents, grandparents, great-grandparents, and so on were considered to establish ties to slave owners.

However, in many cases, data about relatives like uncles or cousins proved helpful in corroborating genealogical connections. 

This type of cross-referencing, widely used in genealogical research, is facilitated by online tools such as the FamilySearch website. This platform is managed by a nonprofit organization with a fascinating background—it is run by The Church of Jesus Christ of Latter-day Saints. The platform provides access to millions of historical records, including immigration documents, church and civil certificates, censuses, and other vital records from various countries. Users can build and share family trees, collaborate with others, and access an extensive collection of digitized historical documents.

The church's interest in genealogy stems from the Mormon belief that family relationships are eternal and continue after death. In Brazil, FamilySearch partners with the Immigration Museum in São Paulo. Henrique Trindade, the museum's Training and Educational Coordinator, explains that there are different ways to access FamilySearch documents. Many digitized documents are public and can be accessed from any device by creating a free account. Exclusive access is also available in affiliated libraries or Family History Centers.

The Immigration Museum is the first affiliated library of FamilySearch in Brazil. “You can view some records in these affiliated libraries. That’s because FamilySearch’s contracts with institutions [like notary offices, parishes, or archives] providing the documents establish restricted access,” explains Trindade. Agência Pública visited the Center for Preservation, Research, and Reference at the museum several times during the research and verification process.

Other documents were only available for consultation at Family History Centers, located within Mormon churches. There are 387 such centers in Brazil. 

For each record identified in FamilySearch trees, we meticulously verified details such as full names and surname variations, taking into account changes due to marriage, adaptations, or even different spellings. We also cross-referenced dates and places of birth, marriage, and death, along with records of parentage, to confirm the identity and lineage of each ancestor of the politicians and officials we could trace.

We compared the records found on FamilySearch with the attached files within the platform itself. All data were organized into spreadsheets and categorized using a universal genealogical numbering system known as the Sosa-Stradonitz method.

Throughout our research, we faced challenges that, according to José Victor Maritan, a researcher at the Population Studies Center of the University of Campinas (Unicamp), are common in genealogical research. One major difficulty is the lack of unique records for each Brazilian citizen. These records become scarcer as we delve deeper into earlier generations. For instance, the Brazilian unique individuals database was only launched in.

To overcome this, Maritan adopted a research method similar to ours at Pública. “I went to public archives and converted images of documents from the 18th and 19th centuries into spreadsheets. This makes it easier to organize family trees, reference data, and estimate generations,” he explains.

A significant risk in nominative research is encountering homonyms (people with the same name). This is particularly challenging with very common names, like "Silva"—one in every 40 Brazilians has this surname—making it harder to confirm the identity of the person in question.

To avoid these pitfalls, we verified additional information such as the names of spouses, relatives, and places of birth. When this information wasn’t sufficient to confirm a familial link, we excluded that lineage from the study. This happened with the genealogy of President Luiz Inácio Lula da Silva.

We found indications that Lula's family tree might include an ancestor who was a slaveholder, but the available data were insufficient to confirm this connection. The region of Garanhuns in Pernambuco, where Lula and part of his family lived, was initially settled by Micael de Amorim Souto and Maria Paes Cabral, a couple reported to have owned at least five slaves in 1749. However, we could not verify the relationship between Micael and Maria, potential sixth great-grandparents of Lula, and the rest of the connections displayed on genealogical websites due to the lack of documents.

Search for Slaveholders

For genealogies predating 1888, the year slavery was abolished in Brazil, we cross-referenced genealogical data with information about slave ownership. This research employed Open Source Intelligence techniques across various platforms, including advanced searches combining terms like “cativo” (captive), “africano” (African), “escravo” (slave), and “alforria” (manumission) with the names of ancestors.

For instance, in the genealogy of São Paulo Governor Tarcísio de Freitas, evidence was confirmed through baptismal records (parish documents identifying each baptism performed by a particular church, including information such as parents, godparents, age, and place of origin) of enslaved individuals in the state of Rio de Janeiro. Anna Rosalina Pacheco de Freitas, his great-great-grandmother, is mentioned in the baptism record of Martha, the daughter of her enslaved woman Delfina.

In other cases, involvement with the slave system was confirmed through population censuses or references in studies based on historical documents. Senator Marcos Pontes, whose family comes from the interior of São Paulo, has among his ancestors the militia officer and plantation owner Joaquim da Silveira Leite, recorded in the Itu census as a “chefe do fogo” (head of household). This record, cited in the article “Senhores de Terras da Vila de Itu em 1817” by historian Leandro Antonio de Almeida of the Federal University of Recôncavo da Bahia (UFRB), notes that the alferes lived with his wife, Ana Joaquina, five children, and nine enslaved people.

Verification and Validation Process

At the end of the research, each genealogy underwent a re-check. We established validation criteria, such as the existence of supporting documents or references from academic research where the author of the study could be identified. Genealogy books or projects like Compartilhar and Descendants' websites, which house research information, were considered in some cases if the relationship was mentioned across multiple sources.

Then, we used artificial intelligence to help decipher documents written over a hundred years ago in cursive handwriting and on papers in various states of preservation. Although the transcription was not perfect, it allowed us to identify patterns and essential information from the documents.

Another crucial step was ensuring that each politician and authority mentioned received, at least 15 days before publication, an email detailing their connection to the enslaver ancestor, along with their genealogical tree and supporting documents. Each authority or their representative had until November 11 to provide comments or challenge the genealogy presented. All responses we received are available for reading in this special report.

This is a pioneering journalistic effort, which I also consider courageous for addressing a debate that many Brazilian authorities prefer to avoid: how individuals who benefited from slavery have left descendants in politics. We recognize that our findings are just a small glimpse of a much larger picture, given the limitations of access to genealogical records. There are still many stories to be told about how power in Brazil was built on the backs of enslaved labor and how this historical debt has never been paid—or perhaps can never be. We hope this marks just the beginning of our journey in pursuit of these answers.

Edits by Mariama Correia