How Reuters researched and reported Slavery's Descendants

Journalists also limited their research to direct lineal ancestors of the present-day elite rather than building sprawling family trees that included distant cousins.) The Lab tentatively identified slave-owning ancestors for 142 members of the 117th Congress, 16 governors, two Supreme Court justices, President Joe Biden and all living former presidents except for Donald Trump, whose ancestors came to America after slavery was abolished. That information was only a starting point.

How many members of America's political elite are direct descendants of people who were slaveholders? To answer that question, Reuters examined the family ancestries of more than 600 of the country's leading officeholders: members of the 117th Congress, which ended this January; the 50 U.S. governors in office before last November's elections; the nine current justices of the Supreme Court; and President Joe Biden and the five living former presidents. (The analysis of the 117th Congress was essentially a snapshot of its members as of April 2021, when it had 536 members.)

Such an inquiry has never been done. At the outset, Reuters sought assistance from the Record Linking Lab at Brigham Young University in Utah. Led by an economics professor and staffed by student researchers, the lab harnesses the underlying data of FamilySearch.org, a top genealogy website, which says it has digitized billions of records from across the world. (The website is funded by the Church of Jesus Christ of Latter-day Saints, an early pioneer in genealogical research in the United States. The church also founded Brigham Young University.)

The Record Linking Lab uses machine learning to perform automated linking of U.S. census documents, typically for academic papers and large studies, and then staffers manually check smaller sets of records to match them with the names of specific ancestors. For Reuters, the Lab performed an initial search of the ancestries of the 600-plus political notables. It returned a large set of probable cases in which top politicians appeared to be descended from at least one direct lineal ancestor who enslaved people in America.

(Reuters only considered evidence of slaveholding that occurred after the founding of the United States. Journalists also limited their research to direct lineal ancestors of the present-day elite rather than building sprawling family trees that included distant cousins.) The Lab tentatively identified slave-owning ancestors for 142 members of the 117th Congress, 16 governors, two Supreme Court justices, President Joe Biden and all living former presidents except for Donald Trump, whose ancestors came to America after slavery was abolished.

That information was only a starting point. It provided initial roadmaps for a team of about a dozen journalists who spent more than a year developing and then executing a multi-step reporting process that either corroborated the initial links, disproved them, or identified entirely different familial lines tied to slavery. VERIFICATION

The team used a variety of sources to verify each generational connection in a family tree. In cases when the FamilySearch.org model appeared to be incorrect, we tested whether other ancestors in the family tree might have enslaved people. As a central guide, reporters began with U.S. census records, which are searchable from 1790 to 1950 through the FamilySearch website and Ancestry.com, another major online genealogy website in America. (Most of the records from the 1890 census were destroyed by fire and are not available.)

Using other publicly available information such as obituaries, wedding announcements and biographies, the team's first step was to establish an ancestor of a political notable – a parent or grandparent – who appeared in the 1940 or 1950 census. From there, reporters worked back in time through the censuses into the slavery era.

For each ancestor listed in the census documents, we recorded in our database a host of discrete pieces of information, such as first, middle and last names; ages; names of other household members; and place of residence. Following a person from one decade's census to the previous one was straightforward when names, ages and other details matched.

When the family lineage moved from children to parents, however, establishing that the correct person was followed from one census to the next required heavy use of other public records. Those records included wills; birth, marriage and death certificates; grave and cemetery records; contemporary news articles, such as wedding announcements and obituaries; and books, such as regional histories, family biographies and family Bibles. Our journalists accessed the information through Ancestry.com and FamilySearch.org; digitized state archives; books at repositories including the Library of Congress, the New York Public Library and the Daughters of the American Revolution Library; and through public records requests filed directly to counties and states. To deal with discrepancies in the records, such as differences in name spellings and ages, we generally used multiple documents to confirm a given generational linkage. For example, an 1860 census entry might show an 8-year-old boy named John Smith living in the same household as a 45-year-old man, James Smith. The document suggests that James is John's father, but it does not directly state it. To confirm that John was indeed James' son – versus, say, a cousin living with the family – we obtained at least one other document, such as a death certificate for John Smith, in order to substantiate the relationship.

Many of the documents reviewed by Reuters for this project were more than 100 years old and handwritten in cursive that ranged from the clear to the hieroglyphic. In instances where the writing was indecipherable, Reuters did not rely on that document for its reporting. In cases where we traced a lineage from a politician today to a direct ancestor in the 1860 household census, we then sought to determine whether that household enslaved people.

In 1850 and 1860, the number of people enslaved by a head of household was noted on a "slave schedule" form – essentially an addendum to the U.S. census for those years. The slave schedule only included the name of the enslaver, and the number of people he or she enslaved, along with the age and gender of the enslaved. In each case, the first step was to establish whether there was a match between the name of the ancestor listed on the household census and the name and location of the slaveholder listed on the slave schedule. For example, that householder John Smith from the 1860 U.S. census lived in the same county as the John Smith listed in that county's slave schedule. But with multiple people having similar names in some areas, we needed additional verification to prove a link.

One measure was to search for wills of slaveholders that listed enslaved people. We also examined tax documents that listed enslaved people held by heads of household. REVIEW

After that first round of verification, a small team of journalists vetted the results. Next, the project's editor checked the work. Finally, Reuters arranged to have each case reviewed by two board-certified genealogists: LaBrenda Garrett-Nelson and Scott Wilds. Garrett-Nelson is a fellow of the American Society of Genealogists, and current trustee and former president of the Board for Certification of Genealogists. She previously worked as a tax attorney before becoming a professional genealogist who specializes in African American genealogy and methodology.

Wilds is also a board-certified genealogist and trustee of the Board for Certification of Genealogists with more than 40 years of research experience. Since 1990, much of his work has focused on the ancestry of enslaved African Americans. He is a descendant of slaveholders in South Carolina and elsewhere. In some cases, Reuters identified politicians with multiple slaveholding ancestors. In such instances, we asked the board-certified genealogists to examine the lineage that traced to the ancestor who enslaved the most people.

Garrett-Nelson and Wilds scrutinized the documentation we used in establishing each step that linked a contemporary politician to a slaveholding ancestor. In testing our work, they sometimes engaged in their own research about the family in question. Based on their expertise, they endorsed our conclusions when they judged that a link from politician to enslaver was "supported and accurate based on the information available" and free of "substantive conflicts that dispute the findings that the notable cited in the subject line is a lineal descendant of the slaveholder cited."

They affirmed that we identified 118 leaders with ancestral ties to slavery: 100 members of the 117th Congress

11 governors 5 presidents: President Biden along with four of the five living former presidents – all except for Donald Trump, whose ancestors came to America after slavery was abolished.

And 2 of the nine sitting Supreme Court justices. In many cases, Reuters identified politicians for whom there was strong evidence of an ancestral slaveholder, but insufficient underlying documentation to be certain. Those notables were not included in the Reuters analysis. And because other records that could demonstrate slaveholding have been lost or destroyed over time, "it's a great possibility that you have an undercount," said Tony Burroughs, a genealogist who specializes in helping Black Americans trace their ancestries.

DISCLOSURE For each political leader we tied to an ancestral slaveholder, Reuters shared with them or their offices a cover letter describing the project, a copy of the family tree we created, and the underlying documentation – such as wills, census records and slave schedules – showing that the ancestor in question enslaved people.

Each politician or their representatives had the material by May 1, giving them more than a month to review, challenge and comment on it. Reuters made multiple attempts to discuss our findings with each notable. None disputed our findings that at least one of their ancestors had enslaved people. (Edited by Blake Morrison)

(This story has not been edited by Devdiscourse staff and is auto-generated from a syndicated feed.)