In the United States, being locked away in long-term solitary confinement is arguably the worst thing that can legally be done to a human being, short of the death penalty. Solitary’s deeply harmful effects on emotional, psychological, and physical well-being are now so well-documented as to be irrefutable, and a growing international consensus equates the practice with torture. Yet there is so little government oversight of U.S. prisons that official data and documentation on the use of solitary confinement in various settings is thin and incomplete, when it exists at all. Despite a decade of work by journalists, advocacy groups, and incarcerated people themselves to expose the realities of solitary confinement, key information is still missing—including something as essential as a comprehensive and accurate count of how many people are in solitary.
Counting the number of people in solitary confinement is notoriously difficult. Americans are incarcerated in one federal and more than fifty state and territorial prison systems, in over 3,000 local jails, and in immigration, military, tribal, civil commitment, psychiatric, and juvenile detention facilities. There are no federal requirements for tracking or reporting this information, and when it is collected at all, each system uses its own often opaque data-gathering techniques and definitions of what constitutes solitary. Until recently, the most authoritative source of information was a Bureau of Justice Statistics (BJS) census, which showed 81,622 people in “restricted housing” in state and federal prisons in 2005. For years, this outdated and incomplete figure was the most widely cited and published solitary count.
However, in recent years, the Association of State Correctional Administrators (ASCA) and the Liman Center at Yale Law School have jointly worked with federal and state departments of corrections to gather more accurate, up-to-date information. This major undertaking has produced three nationwide surveys of solitary confinement, the most recent of which was published this past October. These reports are accepted as the most accurate counts of individuals in solitary confinement today. Yet, as the authors themselves carefully document, these reports are based on imperfect information: the collected data are incomplete; are impossible to compare year-to-year; and exclude major categories of people, including those in local jails and various other detention institutions.
The ASCA/ Liman Center’s October report, Reforming Restrictive Housing: The 2018 ASCA-Liman Nationwide Survey of Time-in-Cell, counted 61,000 people in solitary, a decrease from the organization’s 2016 report, which counted 68,000. Both results are markedly lower than the earliest 2015 report, which provided only a partial count and estimated the total number of people in solitary at 80,000–100,000.
At first glance, this appears to suggest a significant reduction in solitary confinement over the past several years. In fact, upon publication of the report, several prominent media outlets celebrated this dramatic decrease, reporting that “the total segregated population has fallen from up to 100,000 in 2014 to about 61,000 last year,” and that there has been a “healthy drop” that was “badly overdue.” However, the authors’ careful methodology reveals that each of the three reports used slightly different parameters, and as a result, year-to-year figures cannot be directly compared.
For example, in the first report, which found 80,000 to 100,000 people in solitary confinement in the fall of 2014, the authors used data from 34 jurisdictions to estimate the total number of people in segregation in prisons nationwide. That report also counted everyone in segregation, no matter the duration of their stay.
The second report, using data from one year later, identified a much lower figure—67,442 people in segregation. However, as the authors made clear, this report’s data was not an extrapolation, but rather the sum of a limited data set. Since only 96 percent of state and federal prisons provided data, the incarcerated populations of 4 percent of prisons were not included in this figure.
The second report’s final estimate was further deflated by the adoption of a more limited definition of solitary confinement. For this report, the authors counted only those who had been in segregation for “15 continuous days or more”—only a subset of the total segregated population. This change in methodology was based on the United Nations’ Standard Minimum Rules for the Treatment of Prisoners (the Mandela Rules), an agreement banning long-term solitary (beyond 15 days), which was adopted by the UN General Assembly between the publication of the two reports.
Clearly, the rate of solitary confinement is almost certainly not falling as quickly as a side-by-side comparison of the Liman reports might suggest without a careful review of the various reports’ methodologies. Some of the drop does reflect changes in jurisdictions like Colorado (which pioneered reforms that have decreased its segregated population from 1,500 in 2011 to almost 0 today) and California (which decreased its solitary population following a series of hunger strikes and a lawsuit settlement, from a high in 2012 of almost 10,000 to approximately 4,000 five years later). Other states, including Connecticut, Maine, and North Dakota, as well as cities like New York, have also reduced their solitary populations by well over half. Yet so far, these dramatic decreases remain the exception, not the rule. In fact, the report’s authors identify multi-directional trends, noting that eleven states actually saw an increase in the number of people in isolation since 2015, even as in nine of those states, overall prison populations declined.
Further, as the Liman reports authors acknowledge, the current data-gathering efforts on solitary all share an inherent flaw: the reports rely entirely on self-reported data from various departments of corrections, and are methodologically generous in deferring to state prison agencies’ own definition of such critical concepts as “serious mental illness” and degrees of isolation. Any solitary confinement count will only be as reliable as the self-reported data it is based on, and data that can be easily manipulated will likely produce unreliable results. Data collection practices vary widely across states, with no robust statistical infrastructure in place and no oversight.
This leaves the data—already politically charged, contentious, and poorly defined—especially vulnerable to distortion. As Keri Blakinger, who covers criminal justice for the Houston Chronicle, pointed out in a tweet last October, Texas underreported its own solitary numbers by excluding those in punitive solitary, and possibly excluding those in pre-hearing solitary and psychiatric seclusion. (Already, Blakinger said, too much time had passed for the state department of corrections to be sure which groups were counted.)
The numbers of individuals with mental illness who are held in isolation are particularly susceptible to underreporting. For the purpose of the most recent report, the authors allowed each prison system to use its own definition of “serious mental illness,” and found as a result that there are likely “more than 4,000 prisoners identified as seriously mentally ill and in restrictive housing.” The number, however, would almost certainly be far higher if there were an objective, reasonable standard for identifying “serious mental illness.”
Texas, for example, responded to the survey by claiming that they “did not define ‘serious’ mental illness” and instead reported the number of individuals “on an inpatient mental health caseload”—a much higher level of care of than most people with serious mental illness receive, and thus a much higher definitional threshold to meet. As a result, the figures for Texas that made it into the report cite implausibly low percentages of people in prison with serious mental illness (1.1 percent) and people in isolation with serious mental illness (0 percent).
As noted, the most widely cited reports, including Liman’s, do not count those in solitary in non-prison environments. In jails alone, there are likely tens of thousands of additional people in solitary. According to a survey conducted by the Bureau of Justice Statistics in 2011-12, the most recent period for which the data was collected and published, there were roughly 20,100 individuals in solitary confinement in jails across the United States.
The number of children in solitary would similarly have been excluded from Liman’s reports, as the vast majority of children in custody are held in juvenile facilities. In a one-day survey in the fall of 2016, the Office of Juvenile Justice and Delinquency Prevention identified 45,567 children in the custody of “residential placement facilities.” And a 2014 report from the same department found that nearly half of these juvenile detention facilities use solitary confinement as a disciplinary method. Earlier surveys have found that as many as 35 percent of children in residential placements had been placed in solitary confinement.
The demand for detailed and reliable data on solitary confinement is only bound to increase in the future, as incremental reforms of the practice raise new questions regarding acceptable levels of prison isolation. For example, a growing number of states have introduced “step-down” programs to transition people out of solitary, with gradually increasing amounts of out-of-cell time. But in some cases, participants have remained indefinitely in the early “steps,” where they may only be outside their cells for three or four hours. The same is true for individuals in some “mental health” units that are supposed to serve as alternatives to solitary.
Typically, it is the responsibility of government-funded and publicly accountable bodies to track such data and make it publicly available. However, in the case of jails and prisons, correctional departments across the country have proven unwilling to demonstrate transparency and report data to the public. Some states, such as Maryland, Maine, Massachusetts, Nebraska, and Nevada, have enacted laws requiring their departments of corrections to track the use of solitary, and others have introduced similar legislation. The federal First Step Act, which has passed both the Senate and House of Representatives, has a similar, though vague, reporting requirement for the federal Bureau of Prisons.
But until this becomes the norm—and until such requirements are made even more reliable through independent oversight—nonprofit ventures like the ASCA/Liman reports will have to take on the task and do the best they can with elusive and imperfect data. And we will not truly know how many people wake up each morning, in the United States, to the soul-crushing isolation of solitary confinement.