In all likelihood, most Durhamites will never see Butner Federal Correctional Complex.
North Carolina’s only federal prison is just 11 miles from Brightleaf Square, straddling the line between Durham County and neighboring Granville County. But the roads that lead to it are almost deserted. Acres of forest isolate the complex, and driving northeast on Old Oxford Road, signs for a pharmaceutical business park and the entrance to Historic Stagville plantation are among the only hints of civilization.
Perhaps this remoteness is part of why, between 1990 and 2010, the U.S. Census incorrectly counted all Butner prisoners as residing in Granville County—despite a substantial portion of them living on the Durham side of the county line.
Census data guides over $1 trillion in federal spending each year and determines the shape and size of counties’ voting districts.
The miscount, which involved thousands of federal detainees, skewed such decisions. It warped Granville County voting maps and caused at least one federal department to misallocate funds. And it throws Durham and Granville into a nationwide debate over whether the census should count prisoners as residents of the counties where they’re incarcerated—a practice that activist Gino Nuzzolillo calls “a remnant of white supremacy.”
Five facilities—a medical center, a prison camp, a low-security prison and two medium-security prisons—comprise Butner Federal Correctional Complex (FCC). The buildings cluster together at the south end of the institution’s 1,600 acres of mostly undeveloped land.
The Durham/Granville county line has bisected the complex since it opened in 1977. A Durham Geographic Information Systems map shows the line cutting through the oldest part of Butner FCC, Federal Correctional Institution I (FCI-I), with most of FCI-I’s buildings falling on the Durham side.
A Bureau of Prisons (BOP) population report in February 2000 counted 825 prisoners as residing in this medium-security facility. The report also counted 302 prisoners at Butner Federal Prison Camp, which is entirely in Durham County.
All Durham federal prisoners resided in Mangum Township, the county’s northernmost subsection. However, the 2000 decennial census—taken just two months after this population report—counted zero detainees in Mangum Township. So did the 1990 census.
In other words, two decades of censuses failed to count any of Durham’s hundreds of federal prisoners as residing in the county.
Initially, the Census Bureau repeated its miscount in 2010. But administrators noticed something amiss while reviewing that year’s data.
The Bureau quietly issued a correction. Mangum Township didn’t have zero prisoners—it had 2,387 prisoners. Conversely, Granville County’s recorded incarcerated population dropped from 5,631 to 3,244.
The Census Bureau did not update its publicly available datasets to reflect this information, and no major media outlets appear to have reported the error.
Durham Planning Manager Scott Whiteman, who assumed his current position in 2015, said that he “can’t confirm” whether the Census Bureau notified Durham County when it discovered the mistake. However, he believes he might have once heard his predecessor mention the miscount.
“I’m glad it’s correct now,” he said.
The miscount had a measurable impact on Granville County’s voting districts.
The 2010 redistricting cycle found North Carolina legislators embroiled in controversy over gerrymandered maps. While state officials tussled over where and how to draw congressional district lines, Granville County drafted its voting maps using the original 2010 census data—data that incorrectly counted over 2,000 Durham County residents.
These residents made up about 4% of Granville County’s recorded population.
Most federal prisoners can’t vote, so the miscount had little impact on how many voters cast ballots in the county. However, Granville County uses census data to draw the boundaries of its seven voting districts. Each of these districts elects its own county commissioner and school board member.
Since Butner FCC’s district had fewer residents than census data indicated, this meant that the area was overrepresented in the county government. Relative to its population, it was bigger than it should have been, and its residents’ votes carried more weight than votes elsewhere in the region.
County officials drafted a new district map in March 2013 after the Census Bureau notified them of the mistake, according to a county press release. Reporting on a tangentially related story this October, Charlotte news station WFAE obtained documents showing where the county adjusted district lines.
Officials carved off a sizeable chunk of District 7, which contains the town of Butner, and gave it to District 3, which contains the federal prison. They then shifted the borders of two other voting districts to reestablish a balance.
Prior to the 2013 correction, Granville County’s districts had presumably reflected flawed census data since 1990.
The error’s economic consequences are more difficult to measure.
On the Granville side of the county line, a five-minute drive from the federal prison, the town of Butner is an unusual place. In contrast to the deteriorating infrastructure that mars many of North Carolina’s small towns, Butner’s town hall and post office are modernist gems.
Auburn accent colors, large windows and geometric metal features distinguish the buildings. A plaque outside the town hall dates its construction to 2011, while a flag across the street declares that Butner incorporated in 2007.
Butner owes most of its growth to the numerous federal and state institutions that cluster around it. In addition to the federal prison, the town of some 8,000 people also contains a state psychiatric hospital, a substance abuse treatment center, a regional hospital, a center for people with disabilities and a home for at-risk veterans.
But flawed census data also played a role in the town’s economic success.
In 2017, an analysis by The George Washington University found that the federal government relied on census counts to distribute $1.5 trillion to over 300 programs nationwide. These include heavy hitters like highway planning and construction, the Supplemental Nutrition Assistance Program and low-income housing loans.
Not all of these programs distribute funds on a county-by-county basis. For instance, a spokesperson at the Department of Health and Human Services said that none of that agency’s numerous programs would have been affected by the miscount.
At the North Carolina Department of Transportation, however, a spokesperson said the oversight probably impacted two kinds of budget allocations: 5311 funds and the Rural Operating Assistance Program. Both of these provide public transportation in rural areas.
The NCDOT determines how much money to give these funds based on the U.S. Census’s county population estimates. The miscount would have resulted in more money going to Granville County and less to Durham County between 1990 and 2010.
Fortunately, N.C. State Demographer Michael Cline said that the federal government waited for the Census Bureau to review its data before allocating funds based on the 2010 numbers. So that year’s error had no bearing on federal funding.
Who was responsible?
On a cold November morning, one of Butner FCC’s medium-security prisons appears to be an unassailable fortress.
A double layer of chain link fence encircles the buildings’ gray concrete slabs. Lengths of razor wire cover the fences and fill the gap between them, glittering in the sunlight like an enormous steel serpent.
A bus with black tinted windows rumbles along a path just inside the fence. A few crows caw from the woods. Otherwise, silence.
Cline said that the Census Bureau counts prisoners “administratively.” Instead of having a census taker go from cell to cell, a prison official simply sends the Bureau a record of their institution’s population.
But who was that official? And what Census Bureau employee was in charge of Butner FCC’s data?
Cline, who has worked with the Census Bureau since 2017, said he didn’t know what took place at his agency 10, 20 and 30 years ago. And the prison complex was even less forthcoming.
Representatives did not respond to multiple emails, and the phone line for the public information officer’s facility appears to be nonoperational. Phone calls rang for several minutes and then dropped.
The BOP website lists all of Butner FCC’s subsections as being in Granville County, making no mention of Durham County. It’s possible this played a role.
But it’s unclear who specifically was responsible for the census miscount.
It is similarly unclear whether the 1980 census, the first census to count the prison complex, also misrecorded Butner FCC residents.
That year’s count tallied 385 federal and state prisoners in Granville County and 103 in Durham County. But it failed to record prisoners by township, removing a key piece of evidence.
With more recent data, it would be possible to compare the available numbers against population records from the BOP. But a spokesperson said over email that the BOP information office has no access to records from before 2001—leaving no obvious way to verify whether the miscount spanned an additional 10 years.
How to challenge a census miscount
During a two-and-a-half year window following each decennial census, municipalities can present challenges to recorded data.
The Count Question Resolution Operation (CQR) opened in December and will run through June 2023. Cline said that county governments can present a CQR challenge if they believe their region has been misrepresented.
The Census Bureau investigates these challenges and releases corrections.
Private citizens can’t file a CQR challenge on their own. But they can contact their county government if they find something suspicious in census data.
For Durham County challenges, Cline recommended reaching out to City Manager Wanda Page. Her email address is email@example.com, and her office’s phone number is 919-560-4222.
Where do prisoners “actually reside”?
Some people disagree with the practice of counting prisoners as residents of the county where they’re imprisoned, rather than the county they originally came from.
One vocal reform proponent is the Durham-based Southern Coalition for Social Justice (SCSJ). This group, which advocates for the rights of communities of color, has filed a lawsuit alleging racial gerrymandering in North Carolina’s current district maps.
SCSJ spokesperson Gino Nuzzolillo said in an interview that prisoners feel little connection to the community surrounding their prison and rarely remain in-county after they’re released.
Counting detainees as residing in their prison’s county “unfairly” boosts that region’s political clout, Nuzzolillo said, while also reducing the political power of prisoners’ counties of origin. Since U.S. prisons disproportionately detain Black and Brown people, while towns that house prisons tend to be mostly white, Nuzzolillo sees this as a racial equity issue as well as a political one.
The system’s apparent unfairness leads Nuzzolillo and other activists to refer to the current counting method as “prison gerrymandering.”
Donald Murphy, a BOP spokesperson, did not answer emailed questions about whether the BOP supports existing policies. He instead directed The 9th Street Journal toward a webpage that listed which states currently allow felons to vote and summarized recent legal developments on the issue.
But Jason Torchinsky, general counsel to the National Republican Redistricting Trust, defended the practice in an August interview with the politics news website The Hill.
He argued that prisons “generally rely on local communities for food, water, power, health care [and] support services,” and that government funding should reflect that. Also, since many prisoners come from urban areas, Torchinsky said that failing to count them where they “actually reside” would give undue power to Democratic politicians.
The future of prison census counts
Nevertheless, many states have recently changed their policies.
In 2011, Maryland and New York became the first states to pass laws requiring the census to count prisoners in their counties of origin, according to the Prison Gerrymandering Project. Eight other states followed suit in time for the 2020 census, and a ninth will implement changes in 2030.
North Carolina policies remained unchanged during the most recent redistricting cycle, which concluded in November, but the SCSJ lobbied counties around the state to disregard prison populations when drawing district lines. This way, at least in local elections, all votes in a county would carry equal weight.
It’s unclear whether Granville County followed the SCSJ’s advice. When asked, local officials directed The 9th Street Journal to Brooks Pierce law firm, the private company that drew the first draft of their district maps. But a representative at Brooks Pierce declined to comment for this story.
On Nov. 16, the day after Granville County finished revising its maps, County Commissioner Jimmy Gooch described the redistricting process as “a headache” involving many moving parts. He expressed relief that it was over.
“None of us liked the end product,” he said in an interview, “but you live with it.”
Nuzzolillo, meanwhile, said the SCJS will continue advocating for reforms, although changes to census policies won’t affect anything until the next count in 2030. He stressed the importance of fair counts and fair districts for the functioning of democracy.
“The ability to freely cast a ballot,” he said, “means very little if you are doing it in a voting district where your vote has been diluted.”