Beck Presents BJS PREA Report at Grapevine Conference

By Lia Gormsen

Between April and August 2007, the Bureau of Justice Statistics conducted the first National Inmate Survey (NIS) to fulfill requirements mandated by the government in the Prison Rape Elimination Act (PREA) of 2003. Allen J. Beck, Ph.D., chief statistician at BJS, presented the results from the survey at ACA’s Winter Conference in Grapevine, Texas. Among the study findings were that 4.5 percent, an estimated 60,500 state and federal prison inmates, experienced one or more incidents of sexual victimization involving other inmates or staff.

Each year, BJS is required to carry out a comprehensive statistical review and analysis of the incidence of rape in U.S. prisons. With NIS, the bureau sought to provide a ranked listing of state and federal prisons according to the incidence of prison rape. The study was conducted in 146 randomly selected adult confinement facilities that included at least one facility from each state. A sample of 23,398 inmates completed self-interviews on touch-screen laptops by following audio instructions delivered via headphones. The survey was unique in this way, Beck explained, because the data was produced independently. Past surveys on prison rape have relied on data tabulated from facility administrative records that document incidents reported by inmates to correctional authorities. Because some victims may be afraid to report incidents and because facility records may vary based on the way rape is defined, reported and recorded, previous facility-to-facility comparisons were not considered reliable, Beck explained. Even with NIS, it is still “statistically not possible to rank facilities” without talking directly to each inmate, Beck said. However, in providing anonymity to respondents, NIS encouraged fuller reporting by victims. That, in combination with the uniform nature of the survey in terms of defining and recording incidence, allowed BJS to group the facilities with the highest and lowest rates of sexual victimization. The 10 facilities with the highest prevalence of sexual victimization had rates of 9.3 percent or higher, and three of these facilities had incidents of staff sexual misconduct that exceeded 10 percent. The six facilities with the lowest prevalence had no reports of sexual victimization.

Overall, more inmates reported sexual victimization involving staff than involving other inmates — 2.9 percent of incidents involved staff, whereas 2.1 percent involved another inmate. Beck found that “strong gender differences emerged” when comparing facilities with high staff-on-inmate victimization to those with high inmate-on-inmate victimization. Female facilities had much higher rates of inmate victimization, whereas male facilities had higher rates of staff victimization. The “good news,” Beck said, was that across the board, the survey uncovered very low levels of physical injury accompanying victimization. Half of 1 percent reported being injured by an inmate during an incident, and three-tenths of 1 percent reported being injured by a staff member.

Prevalence vs. Incidence

Although PREA required BJS to rank facilities based on the incidence of prison rape, Beck explained that it is much easier to measure prevalence rates than incidence rates. Here prevalence refers to the total number of inmates victimized, and incidence denotes the total number of times that each inmate was victimized. According to Beck, requiring “serial” victims to recall each individual time they were involved in an incident brings ethical challenges because it “increases the likelihood of retraumatizing those individuals.” BJS believed a more reasonable burden would be asking inmates to categorize the number of incidents as: one time; two times; three to 10 times; or more than 11 times.

When considering the prevalence data, Beck asked workshop attendees to pay careful attention to the confidence intervals accompanying each percentage. He also said that the standard errors are large because of the relatively small number of victims reporting on incidences. In the future, Beck said BJS will work on increasing the amount of inmates surveyed to garner more precise results.

Study Limitations

In his concluding remarks, Beck emphasized that the study “is simply a methodology applied in a uniform way across various settings.” All reported incidences were only allegations — BJS conducted no investigations and, therefore, cannot prove whether the incidents actually occurred. Beck noted that investigations often prove fruitless anyway; half of all investigated allegations of sexual misconduct in prisons cannot be substantiated, he said. “What we have is a snapshot — a look at sexual victimization that gets beyond all that variation from facility to facility and policies on reporting and willingness to report, all those variations from facility to facility that undermine interpretation.” Beyond that, Beck said, the study simply reveals what certain inmates report experiencing during a certain time.

Beck suggested that the study could be interpreted in much broader terms. In as much as it is an avenue for inmates to voice their displeasures, he said, the survey may be a good reflection of institutional climate. For example, Beck believes it could be a measure of how inmates feel about staff and their interactions with staff, particularly during strip searches when inmates often feel vulnerable.

Up and Coming at BJS

In addition to the facility rankings, NIS provided the statisticians at BJS with a “wealth of data at the individual level” that will enable the corrections community to better understand sexual violence in prisons, Beck said. BJS is in the process of compiling individual reports for those facilities with the highest victimization rates so that administrators can better understand the study’s findings. Beck and his team are also writing a report concerning larger questions about the characteristics of victims, perpetrators and the circumstances surrounding victimization.

A summary of the report, as well as a link to the full report, is available on the BJS Web site at www.ojp.usdoj.gov/bjs/abstract/svsfpri07.htm.

Lia Gormsen is assistant editor of On the Line.

Click for Executive Director's Corner