NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Forum on Global Violence Prevention; Board on Global Health; Institute of Medicine; National Research Council. Contagion of Violence: Workshop Summary. Washington (DC): National Academies Press (US); 2013 Feb 6.

Cover of Contagion of Violence

Contagion of Violence: Workshop Summary.

Show details


Eleanor Taylor-Nicholson, L.L.M.


Barry Krisberg, Ph.D.

Earl Warren Institute on Law and Social Policy at the University of California, Berkeley, School of Law

Among justice system officials and the citizenry at large, one of the most accepted methods for dealing with individual and community violence is punishment of offenders through incarceration. The United States has the largest imprisonment rate of any nation. As of the end of 2011, 2.3 million people (one in 33 adults) were in correctional facilities, either at the state, federal, or county level. The majority of violent offenders are sentenced to prison (BJS, 2011). Besides these adults, more than 75,000 juveniles were held in juvenile incarceration facilities or adult institutions.

The role of prisons in responding to violent crime has grown considerably. In 2008, the number of offenders sentenced to state prison for violent offenses reached 715,400, up from 95,400 violent offenders in 2000. This increase accounted for 60 percent of prison growth during this period (BJS, 2011).

Little is understood, however, about whether prisons really work to reduce or prevent the spread of violence in society. Many people assume that prison prevents violence, at least temporarily, by keeping violent individuals off the street. But what if imprisonment makes matters worse and increases transmission? Few longitudinal studies have examined the effects of incarceration as a factor in the mental and physical health of former prisoners, let alone more nuanced analyses such as the effects of particular types of incarceration facilities or length of imprisonment.

We do know that recidivism is high among former inmates, and we also have some limited indication from studies of prison life about the levels and types of violence among prisoners and former inmates. It is striking that this literature may suggest that incarceration could in fact exacerbate violence in some cases, both within the prison walls and in the broader community. This raises significant questions about the dominant ideology that determines how governments invest in strategies to reduce violence.

Violence in Prisons

“[P]rison is no fairy-tale world. He never said who did it, but we all knew. Things went on like that for awhile…. Every so often; Andy would show up with fresh bruises. The Sisters kept at him—sometimes he was able to fight 'em off, sometimes not.” —The Shawshank Redemption

One big concern about addressing violence through incarceration is that prisons themselves are extremely violent places. While this has long been recognized, quantitative studies only began in the 1970s (Ellis et al., 1974) and epidemiological research more recently has allowed us to better understand the frequency and characteristics of violence in facilities.

Nancy Wolff and Jing Shi, for example, have conducted research in prisons across a northeastern state of the United States to determine frequency of victimization of physical and sexual violence. They defined physical violence in line with the National Violence Against Women and Men surveys to include being hit, slapped, kicked, bit, choked, beat up, or hit with or threatened with a weapon. Of the 20,447 inmates at 14 facilities (13 for males and one for females) surveyed, Wolff and Shi (2009) found that approximately 20 percent of female inmates and 25 percent of male inmates reported being physically assaulted during their current sentence by either another inmate or a guard. In the previous 6 months before the survey, men reported much higher incidence of assault with a weapon than women, and also reported much higher victimization by a staff member; nearly one in four men was assaulted by a staff member in the 6-month period.

Not all prisoners are violent, and not all inmates are victimized. On average, victims in the Wolff and Shi study were in their early 30s, African American, had spent at least 2 years in the facility, had 4 to 5 years remaining to serve, and had spent 8 years in prison since turning 18 (Wolff and Shi, 2009). Victimization was noted to depend on age and vulnerability. Younger inmates were more likely to be targeted either by other inmates (one in three younger inmates was assaulted by another inmate during their sentence compared to one in four among older inmates) or by a corrections officer (36 versus 25 percent). Furthermore, sexual orientation and mental illness/disability were identified as contributing to prison sexual assault in one quarter of all assaults by other inmates.

Other research has demonstrated the powerful “situational” impact of prisons on prisoner and guard behavior. Of these, the most famous is the Stanford Prison Simulation study by Philip Zimbardo and colleagues. In 1971, Professor Zimbardo engaged a group of “normal, average, healthy American college males” in a planned 2-week simulation of a prison. The researchers assigned half of the students to role-play “guards” and rotated on 8-hour shifts, and the other half to play “prisoners” continuously. After just 1 week the project had to be terminated because “it became apparent that many of the ‘prisoners’ were in serious distress and many of the guards were behaving in ways that brutalized and degraded their fellow subjects.” The prisoners demonstrated “learned helplessness” behavior and the guards displayed physical and verbal aggressiveness that was not indicated in their preexperiment personality tests.

Juvenile facilities are not immune from these challenges. Mendel (2011) conducted an analysis of the court-sanctioned remedies ordered to address violent or abusive conditions in juvenile facilities, as well as of reports written by reputable media outlets. The Annie E. Casey Foundation found there had been documented “systemic violence, abuse and/or excessive use of isolation or restraints,” as opposed to isolated incidents, in a number of states, plus the District of Columbia and Puerto Rico. The findings are as follows: since 1970, 39 states; since 1990, 32 states, and since 2000, 22 states.

Root Causes Left Untreated

The causes of violent behavior are rarely treated effectively in prisons. A number of researchers have argued, for example, that prison violence may often be related to untreated mental illness. Emotionally disturbed inmates or inmates who require mental health services have been found to commit prison infractions disproportionately compared to other inmates. Because the correlation between prison infractions and violence is often high, these inmates are disproportionately involved in violent incidents as well (James and Glaze, 2006).

Preexisting mental illness is not limited to adult inmates. In one early study of juveniles, 85 boys detained in California for mostly violent offenses were given a standard psychiatric screen, a semi-structured interview for posttraumatic stress disorder (PTSD), and self-report questionnaires measuring personality traits and defenses. A sex- and age-matched group was used for comparing psychometrics. The results indicated that 32 percent of the inmates fulfilled criteria for PTSD, and 20 percent partial criteria. Half of the subjects said witnessing of interpersonal violence was the traumatizing event (Steiner et al., 1997), indicating the vulnerability of these incarcerated youth to exposure to violence.

Incarceration may in some cases exacerbate mental illness or emotional frailty of inmates. Craig Haney, also one of the researchers in the Zimbardo experiment, has subsequently written further about the psychological impact of incarceration (Haney, 2002), and identified the following common symptoms among his clients:

  • dependence on institution, loss of capacity/judgment;
  • hypervigilance, distrust, suspicion;
  • emotional over-control, alienation, and psychological distancing;
  • social withdrawal and isolation;
  • incorporation of exploitative norms; and
  • diminished sense of self-worth and personal value.

Importantly, Haney notes that these effects vary from individual to individual, and may not necessarily be permanent. This echoes another considerable body of research that has explored the ways different prisoners adjust to life in prison and noted that some prisoners even improve functioning (Bukstel and Kilman, 1980). A Canadian study found that some prisoners saw being in prison as a chance to turn their lives around, and many inmates, while resenting imprisonment, expected their lives to improve after release (Zamble and Porprino, 1988). There is little research on the attributes of prisons that are helpful or hurtful in terms of postrelease adjustment. An assumption has been that smaller prisons, which have more education and treatment services and less restrictive custody situations, are less criminogenic, but the research on these issues still needs to be conducted.

Release and Reentry

Of all those incarcerated in U.S. prisons, more than 93 percent will return home eventually; more than 700,000 persons are released from prisons each year (see, generally, Travis and Waul, 2003). However, recidivism rates are high. One study found that within 3 years, 67 percent of returning prisoners were rearrested for a serious offense and 52 percent were returned to prison for a new criminal offense (Langan and Levin, 2002). These rates are highest for nonviolent criminals (robbery, burglary, larceny, motor vehicle theft), but violent criminals also recidivate. Overall, 1990s data showed that released prisoners had at least a 53 times higher homicide rate than the general population (Langan and Levin, 2002).5

Some evidence has also emerged of high levels of family violence among current and recently released inmates. White et al. (2002) reported that 1 in 3 men incarcerated in federal prisons for low-risk crimes admitted recent physical violence against intimate female partners and 1 in 10 reported severe violence toward women. Other studies have found that domestic violence perpetrated by recently released inmates was related to frustration at joblessness, changed relationship circumstances, and displaced anger at incarceration (Oliver and Hairston, 2008).


Although there is good reason to assume that sources of violence transmission spread violence among prisoners, family members, the children of prisons, and in the communities where released inmates return, the research on this key topic is underdeveloped. What longitudinal studies are available rarely employ the experience of incarceration as an independent variable. The criminal justice community assumes that incapacitation is the major tool to stop violence in society. Incarceration consumes an enormous amount of government funds in lieu of spending on community-based violence prevention programming.

With the movement away from a pure criminal justice model to one that is informed by public health principles, the primacy of imprisonment will need to be reevaluated. In the past, other public health issues, such as tuberculosis, polio, mental illness, and HIV, were responded to with incapacitation and isolation of affected individuals. This approach was not very effective in terms of curtailing the problems in the community.

We need much rigorous research on the ways in which the prison experience increases exposure to violence both within and outside the walls. There is an urgent need to examine how prisons and reentry programs can be redesigned to stem the contagion of violence.



Of the 272,111 prisoners released in 1994, 719 were rearrested for homicide in 13 states in 1995, 8.4 percent of all the homicides in those states.

Copyright 2013 by the National Academy of Sciences. All rights reserved.
Bookshelf ID: NBK207242


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (2.2M)

Recent Activity

    Your browsing activity is empty.

    Activity recording is turned off.

    Turn recording back on

    See more...