Originally published by The New Yorker
On the morning of December 12, 2008, two inmates at the Reeves County Detention Center, in Pecos, Texas, set a mattress on fire and started a riot. For months, prisoners at the facility, which housed twenty-four hundred male inmates, had been complaining about poor food, abuse in solitary confinement, and inadequate health care. That morning, Jesus Manuel Galindo, a thirty-two-year-old epileptic inmate, was found dead from a seizure brought on by a lack of medication. Galindo was from Mexico, though he had lived in the United States since he was thirteen, and was serving a thirty-month sentence for “illegal reëntry” into the country, after a previous deportation, which is a federal felony offense. Guards had left Galindo in solitary confinement, where his pleas for medical attention were ignored. The insurrection spread throughout the facility but ended the next day. Six weeks later, though, when guards reportedly moved an inmate named Ramon García into solitary confinement after he said that he was ill, another riot broke out; it lasted five days and caused twenty million dollars’ worth of damage. “We spoke with the warden and we told him to take our countryman out of the punishment cell and take him to the hospital,” an inmate later said, according to an account by the Texas Observer. “We told them that if they were not going to do it, then we would do it.”
The Reeves County Detention Center was then the largest private prison in the world, and the “insurrection at Pecos,” as the riots became known, brought national attention to the people who were incarcerated there. It was a population that, until then, had been mostly overlooked by politicians, journalists, and the general public. The inmates were not U.S. citizens, and they would eventually be deported, but first they were serving criminal sentences. Some had committed immigration offenses, and others had been convicted of drug possession, burglary, or assault; a few had been found guilty of violent crimes, including homicide. In Galindo’s case, it was his illegal-reëntry sentence that landed him in federal custody, but a judge “enhanced” his prison term, based on additional charges that Galindo had written a fake check and contacted his ex-wife in violation of a restraining order.
No one understood exactly how the process worked, but, at some point after the men were sentenced, federal prison authorities had segregated them from the general prison population. The logic appeared to be that, because these inmates would eventually be deported, the government need not expend additional resources on them or provide services such as access to drug-treatment programs and literacy courses that were typically given to citizen inmates. Normally, federal rules mandate that inmates be housed within five hundred miles of their families, but, even though many of these men had family members in the United States, they were thousands of miles from them. Although the facility was under the charge of the U.S. Bureau of Prisons, it was administered by the GEO Group, a private-prison company that, in 2008, reported a billion dollars in profits that came largely from contracts with the federal government. The inmates’ medical care had been subcontracted to a private health-care provider whose practices had previously been investigated by the Department of Justice.
There are currently close to nineteen thousand noncitizen inmates being held in ten privately run prisons in seven states. Immigration detention is the responsibility of the Department of Homeland Security (D.H.S.), which detains about forty thousand undocumented immigrants on any given day. But many immigrants who’ve been convicted of crimes, like Galindo, are under the supervision of a different branch of the federal bureaucracy, the Bureau of Prisons (B.O.P.). As the legal scholar Emma Kaufman notes, in a new article in the Harvard Law Review, inmates in foreign-only, or Criminal Alien Requirement (C.A.R.), facilities make up ten per cent of the over-all population in federal prisons, and they have far fewer protections. Kaufman writes, “All foreign prisons are not only places where foreigners are separated from the rest of the penal population. They are also stripped-down institutions with fewer services than other federal prisons.” Because the inmates will be deported at the end of their terms, they have become a “distinct class of prisoners” whose status has led “prison officials to funnel foreign nationals into remote prisons with fewer resources.”
At the Reeves County Detention Center, according to a 2014 report by the American Civil Liberties Union, inmates who were mentally or physically ill were placed in solitary confinement because an infirmary was still under construction. There were complaints, both at Reeves and at other C.A.R. facilities, that inmates were routinely put into solitary for no reason. At Reeves, a quota written into the contract with the GEO Group stipulated that at least ten per cent of the bed space in the facility needed to be reserved for isolation cells. That was double the space allocated at federal facilities run by the B.O.P., and the result, the A.C.L.U. noted, was a “contractual” incentive for putting prisoners “in extreme isolation whenever the prison is filled to capacity.” In 2017, after further complaints, the federal government closed the prison and transferred the inmates to other facilities across the country. There they found vermin infestations, spoiled food, a shortage of medical staff, and cells that were routinely flooded with raw sewage.
Kaufman’s research is the result of years of public-record requests, interviews, and archival work, and her article, titled “Segregation by Citizenship,” is one of the most comprehensive accounts to date of this largely invisible population. According to government figures Kaufman obtained, fifty-three per cent of inmates in these facilities are convicted of drug crimes, thirty-two per cent are convicted of immigration-related infractions, and eight per cent are convicted of violent offenses. The average length of their sentences is six years; a quarter are serving terms of ten years or longer. They come from a hundred and thirty-one countries, though nearly ninety per cent are from Mexico, Central America, Ecuador, Colombia, Cuba, or the Dominican Republic. (The largest group, about seventy per cent of the total, is from Mexico.)
The government began documenting the nationalities of federal inmates in the mid-nineteenth century, but the idea of opening noncitizen prisons did not take root until 1980. That year, two immigration crises collided: thousands of Haitian refugees, most of them headed for Florida, were fleeing the regime of Jean-Claude Duvalier just as, in April, Fidel Castro announced that any Cubans who wanted to leave the country were free to do so. Over the next six months, in what became known as the Mariel boatlift, a hundred and twenty-five thousand Cubans arrived in Florida. Castro also opened Cuba’s mental facilities and prisons, and, by initial U.S. government estimates, more than two thousand patients and convicts were among the Mariel Cubans. Police departments across the country claimed that the “Marielitos” were responsible for a disproportionate share of rising local crime, and special task forces were set up to deal with them. Further contributing to the sense of national alarm, local law-enforcement authorities insisted that previous estimates were wrong and that as many as forty thousand criminals had come to the U.S. in the Mariel boatlift. Federal officials disputed those figures but could never disprove them.
The concern about immigrant criminals coincided with the start of a period of harsher drug-offense and sentencing laws, which led to a rapid growth of the prison population. Local law enforcement began coördinating with federal immigration authorities, Kaufman writes, and, by the middle of the decade, they were routinely entering prisons in search of deportable inmates. Throughout the nineteen-nineties, new laws also vastly expanded the range of crimes that could trigger deportation, even for immigrants with green cards and other forms of legal status. Federal prosecutions for immigration crimes also started to increase, as did the harshness of the penalties. As a result, the foreign population in federal prisons exploded, from about seven thousand, in 1989, to more than twenty thousand, in 1994. In 1996, the B.O.P. and the Immigration and Naturalization Service stationed immigration agents at eleven installations inside federal prisons. These installations were “targeted at the Mexicans,” an assistant director at the B.O.P. said at the time. In 1999, the B.O.P. announced the first all-foreign prison to house “the criminal alien population,” and, in June, it took bids from private companies to construct two facilities, in California and New Mexico.
Three corporations now run ten of the extant noncitizen prisons: Corrections Corporation of America (now called CoreCivic), the GEO Group, and the Management and Training Corporation. In January, 2016, Seth Freed Wessler published an investigation in The Nation on medical neglect in the facilities, where more than a hundred inmates died between the late nineteen-nineties and 2014. Through a series of public-record requests, he obtained from the B.O.P. more than nine thousand pages of inmates’ medical records from eleven prisons. (One has since closed.) At the time, Wessler reported, there were nearly twenty-three thousand noncitizen inmates in the bureau’s custody, the majority of whom were in prisons run by the GEO Group. “These prisons operate without the same systems of accountability as regular Bureau of Prisons facilities, and prisoners suffer,” Carl Takei, a senior staff attorney at the American Civil Liberties Union, told Wessler.
In August of 2016, the Justice Department announced that it would phase out the use of private prisons, and it appeared that the noncitizen facilities would cease to operate. Once Donald Trump entered office, however, he reversed that decision, and the stock prices of the two largest private-prison companies, the Geo Group and CoreCivic, soared. In May of 2017, the Justice Department solicited a round of new bids. The Trump Administration has been a particular boon for the success of these companies, as they operate not only correctional facilities for the B.O.P. but also immigration jails for the D.H.S.
When I spoke to Emma Kaufman recently by phone, she stressed the potential implications of the data she uncovered showing that nearly all of the noncitizen inmates in these segregated prisons are Latinos. This is way out of proportion with the makeup of the general federal-prison population, where Latinos comprise about thirty per cent of all inmates. “There’s a clear overrepresentation of Latinos, and the question is why,” Kaufman said. She is concerned that, if someone looks Latino, prison authorities might just assume that he isn’t a U.S. citizen. In theory, the process would seem relatively straightforward—when authorities book an individual into custody, they should be able to determine from government databases whether that person is a citizen or a legal resident—but, in practice, errors occur and are extremely difficult to correct. A database could contain outdated or inaccurate information. Consider the context of immigration enforcement: according to a study published in 2011 by Jacqueline Stevens, a professor of political science at Northwestern University, more than twenty thousand U.S. citizens were arrested or deported by immigration agents between 2003 and 2010—a figure that, Stevens wrote, “may strike some as so high as to lack credibility.”
Kaufman has spent years trying to get the government to supply more information about the segregated facilities. And yet significant questions persist. “We don’t really know the answers, because the B.O.P. designation process is a black box,” she told me. “We popped the lid. Now we need to look inside.”