Vera identified common themes in agencies’ use of restrictive housing. This section presents key findings related to some or all of the partner sites—though not all of the findings discussed here are relevant to all jurisdictions. Various examples from the jurisdictions illustrate each point.A more detailed picture of each department can be found in the technical reports provided to each site. In particular, the use of protective custody varied widely among Vera’s partners and is addressed in each site’s technical report.

It is important to note that the data presented below refer to the period before the SAS Initiative started and therefore do not reflect changes agencies have made to their policies and practices since then. 

Conditions in Restrictive Housing

  1. Although cell size, recreational areas, and other characteristics of restrictive housing units varied, incarcerated people in these units were typically held in stark, isolated environments with little sensory stimulation or social interaction. Many cells were small and sparsely furnished, and some had no windows or natural light. Opportunities for therapeutic programming or any form of productive activity were scarce.

    In the most-restrictive housing, people were held in their cells for around 23 hours a day. They received up to one hour of out-of-cell recreation, often held in a small caged area or a bare concrete space, sometimes with limited access to fresh air and direct sunlight. In some systems, barred indoor enclosures were used for recreation at times.

Disciplinary Segregation

  1. All sites frequently used restrictive housing to respond to nonviolent infractions (see Table 1, below): 

    • In Nebraska, North Carolina, and Oregon, the most common infractions resulting in disciplinary segregation were low-level nonviolent offenses, and “disobeying an order” was the most common infraction (see Table 1 below). 
    • In Nebraska, “disobeying an order” accounted for 28 percent of such sentences. 
    • In North Carolina, none of the top 10 infractions resulting in segregation sanctions were among the most serious charges (as determined by the Department of Public Safety).
    • In Oregon, 58 percent of disciplinary segregation sentences were for nonviolent infractions.
    • In New York City, a higher proportion of segregation sentences were given in response to violent infractions; even there, however, disobeying an order was still the fifth-most-common charge to receive a restrictive housing sentence.

    Segregation sentences ranged from a couple of days to several months—though in some jurisdictions, a person charged with multiple rule violations may have been required to serve multiple sentences back-to-back. Some people were released to the general population at the end of their assigned sentence and some were released earlier, but others were transferred to other forms of restrictive housing upon completing the sanction, thus extending their stay in restrictive conditions.

    It should be noted that since Vera conducted these analyses, the North Carolina Department of Public Safety has significantly altered its policies governing disciplinary practices and the Nebraska Department of Correctional Services has ended its use of disciplinary segregation altogether.


  1. restrictive-housing-figures-02-v2.jpg#asset:19615

    Punishment of misbehavior was a substantial driver of the restrictive housing populations at Vera’s partner sites. The charts below show the percentages of people in restrictive housing for disciplinary reasons in three jurisdictions, either serving time in these units in response to an infraction or awaiting a disciplinary hearing.

    Many people who end up in restive housing enter through disciplinary segregation before being transferred to administrative segregation or another form of restrictive housing. For example, approximately 90 percent of incarcerated people in Oregon who spent time in any type of restrictive housing first entered through the disciplinary unit. 

    Overall, the number of people who serve time in disciplinary segregation can be high: in Nebraska, for example, 44 percent of all incarcerated people had, at some point during their time in prison, been placed in restrictive housing for disciplinary reasons (as punishment for an infraction or pending an investigation).

  1. This finding is difficult to quantify using administrative data. Still, at many of the facilities Vera visited, staff and incarcerated people reported that they believed people sometimes violated rules with the express purpose of being placed in restrictive housing because they feared for their safety in the general population. The reasons cited for these concerns included belonging to a vulnerable group (such as young people), being targeted by gang members, and a general fear of violence. Perceptions of threat and insecurity therefore appear to increase the number of infractions committed and the number of people in restrictive housing.

Administrative Segregation

The type of restrictive housing commonly referred to as administrative segregation can be used for multiple purposes, including managing someone who is considered a threat to safety and security or holding someone temporarily while certain administrative processes are completed. This report focuses on one use of this housing that Vera found contributed substantially to the population in restrictive housing across partner sites: indefinite administrative segregation to manage people staff considered dangerous or disruptive.

  1. In North Carolina, for example, more than 1,200 people were being held in indefinite administrative segregation at the time of Vera’s initial assessment. For these individuals, as in other jurisdictions, release from restrictive housing was granted only when their cases were reviewed by a staff member or committee and they were judged ready to return to the general population.

    Vera found that such reviews were conducted infrequently. In Oregon, for example, people who were incarcerated typically spent between 60 and 150 days (approximately five months) in the Intensive Management Unit—a form of administrative segregation—before their first review.

    In many jurisdictions, the criteria used to make release decisions were unclear. People were often required to demonstrate that they did not pose a threat to the safety of others in order to be granted release from segregation—for example, by participating in programming or interacting with other incarcerated people and staff. But such opportunities were rare; most agencies offered little in the way of meaningful programming or congregate activity in which the incarcerated person could demonstrate positive behavior. Further, the harmful effects that isolation can have on a person’s mental well-being and behavior may have made it increasingly difficult for people to “earn” their release from administrative segregation through good behavior. Grassian, 2006, 325.  Opportunities to engage in therapeutic programming were often limited to in-cell workbooks or absent entirely. 

    These factors contributed to long stays in administrative segregation. In addition, people were sometimes transferred directly to administrative segregation from disciplinary segregation or other types of restrictive housing, leading to even longer continuous periods in such settings. 


Specific Populations

In addition to understanding the reasons people are placed in restrictive housing, it is also important to look at who is placed there. Doing so helps identify groups that are disproportionately affected by a system’s policies and practices, as well as populations that may have unmet needs. More research is needed about groups of people who are likelier to be held in restrictive housing and why.Benjamin Steiner and Calli M. Cain, “The Relationship Between Inmate Misconduct, Institutional Violence, and Administrative Segregation: A Systematic Review of the Evidence,” in Restrictive Housing in the U.S.: Issues, Challenges, and Directions (Washington, DC: U.S. Department of Justice, National Institute of Justice, 2016, NCJ 250315).

Echoing disparities seen throughout the criminal justice system, Vera’s analysis found that people with mental health needs, young men, and people of color were more likely to be held in restrictive housing than other incarcerated people.

  1. Identifying how frequently people with mental illness are sent to restrictive housing can be difficult; some data systems do not record incarcerated people’s mental health status, and systems that do may include only limited information. During the initiative, data on mental health needs of people in Nebraska prisons or the Middlesex County Adult Correction Center were unavailable. This lack of clear, precise data makes assessment of a system difficult. It also presents challenges for the unit officers who are responsible for people who are incarcerated, and for the disciplinary hearing officers who try to appropriately respond to or adjudicate their behavior. Still, Vera’s assessment suggests that people with mental health needs were often placed in restrictive housing.

    In Oregon, for example, the majority (61 percent) of people in disciplinary segregation units were identified as having mental health needs ranging from mild to severe.By administrative rule in Oregon, people with severe mental health needs are not sanctioned to stays in disciplinary segregation units beyond 30 days. Facilities are required to divert people with severe mental health needs to other environments at or before the 30-day mark.


    In North Carolina, incarcerated people are classified using a mental health scale from 1 (no mental health needs) to 5 (most-significant mental health needs). As Figure 4 shows, 41 percent of people with the most serious health needs (designated “M5”) were in segregation in June 2015. 


  1. Overall, people of color at Vera’s partner sites had higher rates of contact with restrictive housing than white people did—especially with the most severe types of this housing—and were underrepresented in more treatment-oriented forms of restrictive housing and other less-stringent alternatives.Pathways to restrictive housing are complex and many factors and decision points contribute to someone being placed there. Vera’s analysis describes the outcomes of the use of restrictive housing; it does not assess the determinants for those outcomes or make inferences about causality.

    Figure 5 shows, for example, that in Nebraska prisons, a combined group of Asian Americans, Pacific Islanders, Latinos, and Native Americans had the highest rates of contact with restrictive housing, with 17 percent in the most-restrictive settings as compared to 9 percent of white people.Vera used the racial and ethnic categories that existed in the administrative data provided by each site.  In addition, only 4 percent of black people were in the less-stringent forms of restrictive housing (the majority of which was considered “protective custody”), as compared to 9 percent of white people. 


    Similarly, in New York City, black people were admitted to punitive segregation at 5.7 times the rate that white people were; however, they were less likely to be admitted to units designed as alternatives to restrictive housing for people with severe mental illnesses who had committed infractions, entering those units at 0.6 times the rate that white people did. 

  1. Youth and younger adults were more likely than older people to be placed in restrictive housing.

    • In North Carolina, 8 percent of adults age 26 or older were held in restrictive housing, while 17 percent of young adults (ages 18 to 25) and nearly a third (32 percent) of youth age 17 or under were in restrictive housing (see Figure 6). 
      • Note: North Carolina has since ended the use of restrictive housing for those under 18. (Similarly, New York City has eliminated the use of disciplinary segregation for people who are 21 or younger.)


    • Nebraska showed a similar trend, with only 6 percent of adults age 25 and over in restrictive housing, compared to 13 percent of people under age 25 (see Figure 7).


    • In Oregon, people ages 18 to 25 represented 11 percent of the total prison population but 30 percent of those held in disciplinary segregation. By contrast, people ages 41 and older made up 44 percent of the prison population but just 15 percent of those in disciplinary segregation.
  1. In all five jurisdictions, women were placed in restrictive housing at a lower rate and for shorter periods, on average, than men were. But in Oregon and North Carolina, where mental health information about women in the prison systems was available, the level of mental health needs among these women was high.  

    • In North Carolina, 5 percent of women were in some form of restrictive housing. These women had higher rates of serious mental health needs, with 38 percent requiring some level of psychiatric treatment, as compared to 18 percent of women in the general population.
    • In Oregon, 84 percent of women in restrictive housing had been diagnosed with a mental illness that required treatment, compared to 53 percent of the total female population. And 27 percent of those in restrictive housing were assigned the department’s highest indicator of mental health need, while this was true for 11 percent of women in the general population.

Release from Restrictive Housing to the Community

  1. This practice moves people from an environment of extreme isolation and sensory deprivation into one requiring autonomy and complex social interactions, which likely increases the difficulty of an already-challenging transition.

    Reentering the community from prison or jail is often a difficult process—psychologically, emotionally, and in practical terms, especially in regards to securing housing, health care, and employment. Nevertheless, this is a practice that Vera regularly observed, often without adequate re-entry programming and preparation for release:

    • In Oregon, for example, Vera identified 348 people who were released directly to the community from restrictive housing during an 18-month period, after spending an average of almost five months in such housing immediately prior to their release. 
    • In North Carolina in 2015, 1,892 people were released from restrictive housing directly to the community—that’s roughly 35 people every week. Among them, 15 percent had spent more than six months in such housing immediately prior to their release.