U.S. Department of Justice Office of Justice Programs Bureau of Justice Statistics May 2012, NCJ 237363 Sexual Victimization Reported by Former State Prisoners, 2008 Allen J. Beck, Ph.D. BJS Statistician Candace Johnson, Ph.D. Principal Research Scientist, NORC Bureau of Justice Statistics James P. Lynch Director BJS Website: www.bjs.gov askbjs@usdoj.gov --------------------------------------------------- This file is text without graphics and many of the tables. A Zip archive of the tables in this report in spreadsheet format (.cvs) and the full report including tables and graphs in .pdf format are available at: http://bjs.gov/index.cfm?ty=pbdetail&iid=2234 --------------------------------------------------- ************************************************************ The Bureau of Justice Statistics is the statistics agency of the U.S. Department of Justice. James P. Lynch is director. ************************************************************ ************************************************************ This compendium was written by Allen J. Beck, BJS Statistician, and Candace Johnson, NORC Research Scientist. Jessica Rexroat, BJS intern, verified the report. ************************************************************ ************************************************************ BJS statisticians Christopher J. Mumola and Paige M. Harrison, under the supervision of Allen J. Beck, were project managers for the National Former Prisoner Survey. NORC, under a cooperative agreement and in collaboration with BJS staff, developed the survey and collected, processed, and analyzed the data. Candace Johnson with Pam Loose directed the project; Kirk Wolter and Ken Copeland guided sampling, weighting, and other statistical support services; Kris Talley and Patt Maugherman led field operations; Angela Herrmann directed telephone center activities; Mike Cooke managed TACASI development and IT support; Marissa Kiss, Stephanie Poland, Mehera Baugher, and Emily Frizzell assisted throughout the project; and Lynda Okeke and Fang Wang prepared statistical tables. ************************************************************** ************************************************************** Vanessa Curto and Jill Thomas edited the report, and Barbara Quinn and Tina Dorsey designed and produced the report, under the supervision of Doris J. James. ************************************************************* May 2012, NCJ 237363 National Former Prisoner Survey, 2008 Sexual Victimization Reported by Former State Prisoners, 2008 Allen J. Beck, Ph.D. BJS Statistician Candace Johnson, Ph.D. Principal Research Scientist, NORC May 2012, NCJ 237363 BJS ************************************************************ ********* Contents ********* Highlights 5 Incidents of sexual victimization 8 Type of coercion and physical injury 12 Circumstances surrounding victimization 14 Variations by sex of former inmates 15 Variations by other individual-level characteristics 16 Variations by selected facility-level characteristics 20 Prison placements and individual-level risk factors 25 Reporting of sexual victimization 30 Post-release responses to victimization 32 HIV testing and results 33 Methodology 35 Appendix A 40 Appendix B 41 Appendix C 43 Appendix D 44 Appendix E 45 ************************************************************* ************** List of tables *************** Table 1. Former state prisoners reporting sexual victimization during most recent period of incarceration, by type of victimization and facility 8 Table 2. Former state prisoners reporting sexual victimization in a community-based correctional facility 10 Table 3. Criminal history and supervision status of persons under active parole supervision, by sex of former inmate 11 Table 4. Type of coercion and physical injury of former state prisoners who reported sexual victimization, by type of incident 12 Table 5. Circumstances surrounding sexual victimization of former state prisoners, by type of incident 14 Table 6. Former state prisoners reporting sexual victimization, by sex of inmate 15 Table 7. Staff sexual misconduct and type of activity, by sex of victim and sex of staff 15 Table 8. Prevalence of sexual victimization, by type of incident and former prisoner demographic characteristics 16 Table 9. Prevalence of sexual victimization, by type of incident and former prisoner criminal justice status and history 17 Table 10. Multivariate logistic regression models of sexual victimization, by type of incident, former prisoner demographic characteristics, and former prisoner criminal justice status and history 19 Table 11. Prison facilities entered and prison placements, by former state prisoners 20 Table 12. Sequence of reported sexual victimization in prison, by sex of victim and type of incident 21 Table 13. Prevalence of sexual victimization during prison placement, by sex of inmate, type of incident, and facility-level characteristics 23 Table 14. Prevalence of sexual victimization during each prison placement, comparing the National Former Prisoner Survey, 2008, and National Inmate Survey, 2008 09 25 Table 15. Prevalence of sexual victimization among male and female placements, by type of incident and former prisoner characteristics 26 Table 16. Final multivariate logistic regression models of sexual victimization among male and female placements, by type of incident and former prisoner and facility characteristics 28 Table 17. Reporting of sexual victimization, by type of incident and persons to whom the incident was reported 30 Table 18. Reasons for not reporting sexual victimization to facility staff, by type of victimization 31 Table 19. Facility responses to the reporting of sexual victimization to staff, by type of incident 31 Table 20. Post-release responses of victims to sexual victimization, by type of incident 32 Table 21. HIV testing and results for former state prisoners, by type of victim 33 Table 22. Current employment, housing, and living arrangements of former inmates under active parole supervision, by victimization status 34 *********************** List of appendix tables *********************** Appendix table 1. Standard errors for table 4: Type of coercion and physical injury of former state prisoners who reported sexual victimization, by type of incident 46 Appendix table 2. Standard errors for table 5: Circumstances surrounding sexual victimization of former state prisoners, by type of incident 47 Appendix table 3. Wald-F statistics for former prisoner characteristics in the final multivariate logistic regression models of sexual victimization, by type of incident 48 Appendix table 4. Wald-F statistics for former state prisoner and facility characteristics in the final multivariate logistic regression models of sexual victimization among male and female placements, by type of incident 49 Appendix table 5. Standard errors for table 17: Reporting of sexual victimization, by type of incident and persons to whom the incident was reported 50 Appendix table 6. Standard errors for table 19: Facility responses to the reporting of sexual victimization to staff, by type of incident 50 Appendix table 7. Standard errors for table 20: Post-release responses of victims to sexual victimization, by type of incident 51 ************* Highlights ************* Prevalence of sexual victimization * An estimated 9.6% of former state prisoners reported one or more incidents of sexual victimization during the most recent period of incarceration in jail, prison, and post-release community-treatment facility. * Among all former state prisoners, 1.8% reported experiencing one or more incidents while in a local jail, 7.5% while in a state prison, and 0.1% while in a post-release community-treatment facility. * About 5.4% of former state prisoners reported an incident that involved another inmate. An estimated 3.7% of former prisoners said they were forced or pressured to have nonconsensual sex with another inmate, including manual stimulation and oral, anal, or vaginal penetration. * About 5.3% of former state prisoners reported an incident that involved facility staff. An estimated 1.2% of former prisoners reported that they unwillingly had sex or sexual contact with facility staff, and 4.6% reported that they "willingly" had sex or sexual contact with staff. * Although the rate of sexual victimization in state prison reported by former inmates (7.5%) was higher than the rate reported by inmates in previous BJS surveys (4.8% in 2008-09), the difference may reflect longer exposure periods (39.4 months and 7.9 months, respectively). Coercion and physical injury * Among victims of inmate-on-inmate sexual violence, a quarter had been physically held down or restrained, and a quarter had been physically harmed or injured. * Half of all victims of staff sexual misconduct said they had been offered favors or special privileges; a third said they had been persuaded or talked into it. * Approximately 16% of victims of unwilling sexual activity with staff, compared to 2% of victims of "willing" sexual activity, reported being physically injured by staff. Individual risk factors * The rate of inmate-on-inmate sexual victimization was at least 3 times higher for females (13.7%) than males (4.2%). * The rate of "willing" sexual activity with staff was higher among males (4.8%) than females (2.6%), and the rate of unwilling sexual activity was higher among females (2.5%) than males (1.1%). * Among heterosexual males, an estimated 3.5% reported being sexually victimized by another inmate. In comparison, among males who were bisexual, 34% reported being sexually victimized by another inmate. Among males who were homosexual or gay, 39% reported being victimized by another inmate. * Female heterosexual inmates reported lower rates of inmate-on-inmate victimization (13%) and staff sexual misconduct (4%) than female bisexual inmates (18% and 8%, respectively). * Among female homosexual or lesbian inmates, the rate of inmate-on-inmate sexual victimization was similar to that for female heterosexual inmates (13%), while the rate of staff sexual victimization was at least double (8%) that for female heterosexual inmates (4%). * The rate of imate-on-inmate sexual victimization for males was higher among non-Hispanic white inmates (5.9%) and inmates of two or more races (9.5%) than non-Hispanic black inmates (2.9%). Among male former state prisoners, the rates of staff sexual misconduct were higher for those of two or more races (11.3%) and black non-Hispanics (6.5%) than for white non-Hispanics (4.5%) and Hispanics (4.0%). * The rate of staff sexual misconduct was higher for male inmates ages 20 to 24 (7.9%) than for male inmates ages 25 to 34 (5.2%), ages 35 to 44 (3.5%), and age 45 or older (2.0%). * Among female former state prisoners, rates of staff sexual misconduct were lower for those ages 35 to 44 (3.1%) and age 45 or older (1.6%), compared to those ages 20 to 24 (6.7%). * Most victims of staff sexual misconduct (87%) reported only perpetrators of the opposite sex. * Among victims of staff sexual misconduct, 79% were males reporting sexual activity with female staff. An additional 5% were males reporting sexual activity with both female and male staff. Facility characteristics * Rates of sexual victimization did not vary among commonly cited characteristics of facilities, including size of facility, facility age, crowding, inmate-to-staff ratios, or gender composition of staff. * Among male former inmates, inmate-on-inmate and staff-on-inmate victimization rates were higher in facilities under a court order or consent decree, higher in facilities reporting a major disturbance in the 12 months prior to the most recent facility census, higher in facilities with medium or greater security levels, and higher in facilities with a primary function of housing general population than in facilities without these characteristics. * Among female former inmates, rates of inmate-on-inmate victimization were lower in community corrections centers, in facilities that permitted 50% or more of their inmates to leave unaccompanied during the day, in minimum or low security facilities, and in privately operated facilities than in facilities without these characteristics. * After controlling for multiple individual-level characteristics **Inmate-on-inmate and staff-on-inmate victimization rates among males were lower in reception and diagnostic centers than in facilities that housed general population, and lower in minimum security facilities than in facilities with higher security levels. **Inmate-on-inmate victimization rates among females were lower in facilities that permitted 50% or more of their inmates to leave unaccompanied during the day than in other confinement facilities, lower in medium security facilities than in maximum or high security facilities, and higher in facilities that had a major disturbance in the census year than in facilities that did not. Sexual victimization and its consequences * Following their release from prison, 72% of victims of inmate-on-inmate sexual victimization indicated they felt shame or humiliation, and 56% said they felt guilt. * The majority of victims of staff sexual misconduct involving unwilling activity said they felt shame or humiliation (79%) and guilt (72%) following their release from prison. More than half (54%) reported having difficulty feeling close to friends or family members as a result of the sexual victimization. * Although the vast majority of victims of staff sexual misconduct (86%) reported at least one incident that they considered "willing," approximately a quarter said they felt guilt (27%) or shame and humiliation (23%) after their release from prison. * Among former inmates who had been tested for HIV (90%), those who had been sexually victimized by other inmates or by staff had significantly higher percentages for HIV positive (6.5% and 4.6%, respectively) than those who had not been victimized (2.6%). * Among former inmates under parole supervision, victims and nonvictims did not differ in their current employment (64% employed) or housing arrangements (92% in house, apartment, trailer, or mobile home); however, victims (18%) were somewhat more likely than nonvictims (14%) to be living alone. Sexual Victimization Reported by Former State Prisoners, 2008 The Bureau of Justice Statistics (BJS) conducted the first-ever National Former Prisoner Survey (NFPS) between January 2008 and October 2008. NORC at the University of Chicago, under a cooperative agreement with BJS, collected the data. A total of 317 parole offices in 40 states were randomly included in the survey sample. A total of 17,738 former state prisoners who were under active supervision (i.e., required to contact a supervisory parole authority regularly in person, by mail, or by telephone) participated in the national survey. Interviews from an additional 788 former prisoners were included from the survey test sites. These former inmates had been randomly selected from 16 offices sampled. Based on 18,526 completed interviews, the survey achieved a 61% response rate. (See Methodology for further details.) The NFPS is part of the National Prison Rape Statistics Program, which collects both administrative records of reported sexual violence and allegations of sexual victimization directly from victims. BJS has collected administrative records annually since 2004. We collected victim reports through surveys of adult inmates in prisons and jails in the National Inmate Survey in 2007 (NIS-1) and in 2008-09 (NIS-2) and through surveys of youth held in juvenile correctional facilities in the National Survey of Youth in Custody in 2008-09 (NSYC-1). The NFPS collects data on the totality of the prior term of incarceration, including any time in a local jail, state prison, or community correctional facility prior to final discharge. Because the survey is based on a sample of parole offices and not a sample of prisons, the NFPS is not conducive to providing facility estimates or rankings. The NFPS is designed to encourage a fuller reporting of victimization, by surveying only former inmates, who are not subject to the immediate risk of retaliation from perpetrators or a code of silence while in prison. The NFPS may elicit reports of incidents that were unreported in the previous NIS-1 and NIS-2 surveys of prisoners; however, some reports may be untrue. At the same time, some former inmates may remain silent about sexual victimization experienced while incarcerated, despite efforts to assure victims that their responses will be kept confidential. ************************************************** Other PREA data collections The Prison Rape Elimination Act of 2003 (PREA; P.L. 108-79) requires the Bureau of Justice Statistics (BJS) to carry out, for each calendar year, a comprehensive statistical review and analysis of the incidence and effects of prison rape. The act further specifies that the review and analysis shall be based on a random sample or other scientifically appropriate sample of not less than 10% of all federal, state, and county prisons and a representative sample of municipal prisons. It requires BJS to use surveys and other statistical studies of current and former inmates. To fully meet these requirements, BJS has developed a multiple measure, multi-mode data collection strategy, in which the NFPS is one component. National Survey of Youth in Custody (NSYC) provides facility-level estimates of youth reporting sexual victimization in juvenile facilities, as required under PREA. The first NSYC (NSYC-1) was conducted in 2008-09; the second (NSYC-2) is underway and will be completed by September 2012. NSYC-2 will provide estimates for large facilities that house adjudicated youth and for each of the 50 state systems and the District of Columbia. National Inmate Survey (NIS) gathers data on the incidence of sexual assault in adult prisons and local jail facilities, as reported by inmates. The first NIS (NIS-1) was conducted in 2007, the second (NIS-2) was conducted in 2008-09, and the third (NIS-3) is underway and will be completed by May 2012. Based on inmate allegations, the NIS provides facility-level estimates used to rank facilities as required under PREA. Survey of Sexual Violence (SSV) collects data annually on the incidence of sexual violence in adult and juvenile correctional facilities. Based on administrative data only, the SSV is limited to incidents reported to correctional officials. Begun in 2004, the SSV provides detailed information on incidents that have been substantiated upon investigation. Clinical Indicators of Sexual Violence in Custody (CISVC) In 2010 11, in collaboration with the National Institute of Justice (NIJ) and the Centers for Disease Control and Prevention (CDC), BJS conducted a feasibility study using medical indicators and medical surveillance methodologies. As part of routine medical practice, medical staff in 19 prisons and 11 jails completed a surveillance form for adult males who either made an allegation of sexual violence or displayed clinical conditions consistent with sexual victimization. Results of the 12-month pilot study are expected in 2012 ************************************************************* ********************************** Incidents of sexual victimization ********************************** 9.6% of former state prisoners reported one or more incidents of sexual victimization during the most recent period of incarceration in jail, prison, or a post-release community-treatment facility Among the 18,526 former state inmates participating in the NFPS survey, 2,096 reported experiencing one or more incidents of sexual victimization during their most recent period of incarceration, including the combined time in local jails, state prisons, or post-release community-treatment facilities. Because the NFPS is a sample survey, weights were applied to the sampled offices and offenders under their supervision to produce national-level estimates. The estimated number of former state prisoners experiencing sexual victimization totaled 49,000, or 9.6% of all former state prisoners under active supervision at midyear 2008 (table 1). Among all former state prisoners, 1.8% reported experiencing one or more incidents while in a local jail, 7.5% while in a state prison, and 0.1% while in a post-release community-treatment facility. An estimated 1.4% reported an incident in a facility for which the type could not be determined. 5.4% of former inmates reported an incident with another inmate; 5.3% reported an incident with staff Among former state prisoners, 5.4% (or an estimated 27,300 prisoners nationwide at midyear 2008) reported an incident that involved another inmate, and 5.3% (27,100) reported an incident that involved facility staff. Some inmates (1.1%) reported sexual victimization by both another inmate and facility staff. An estimated 3.7% of former prisoners said they had nonconsensual sex with another inmate, including manual stimulation and oral, anal, or vaginal penetration. An additional 1.6% of former prisoners said they had experienced one or more abusive sexual contacts only with another inmate, including unwanted touching of the inmate's buttocks, thigh, penis, breast, or vagina in a sexual way. An estimated 1.2% of former prisoners reported that they unwillingly had sex or sexual contact with facility staff. An estimated 4.6% said they "willingly" had sex or sexual contact with staff. ************************************************************** National Former Prisoner Survey Protocol Active post-custody supervision Only people who had served time in a state prison, were age 18 or older at the time of the survey, and under active post-release supervision were eligible to be part of the NFPS. The survey includes some people who were under age 18 at the time of their incarceration. These people were sentenced as adults and served time in adult facilities. In 2008, approximately 85% of all persons under some form of post-custody supervision were considered to be under active supervision. Individuals who were ineligible for the survey included those who had absconded, were re-incarcerated, were in a halfway house or community-based treatment center, had a warrant issued for their arrest, were in violator status, or had been transferred to another parole office. Persons under supervision in small offices (under 40 parolees) in remote areas or in specialty offices (e.g., sex offender supervision or treatment facilities) were also excluded from the survey. Overall, the NFPS is a representative sample of approximately 510,800 former state prisoners under supervision at midyear 2008. (See Methodology for sampling information.) Audio computer-assisted self-interview The NFPS interviews, which averaged 23 minutes in length, were conducted using computer-assisted personal interviewing (CAPI) and audio computer-assisted self-interviewing (ACASI) data collection methods. Survey interviewers initiated the personal interview using CAPI to obtain demographic and criminal history information. For the remainder of the interview, respondents interacted with a computer-administered questionnaire using a touch-screen and synchronized audio instructions delivered via headphones. Respondents completed the survey in private at the parole office (or satellite office), with the interviewer in the room but unable to see the computer screen. (See Methodology for further description of the survey protocol.) Voluntary participation with incentives Before the interview, respondents were informed verbally and in writing that participation was voluntary and that all information provided would be held in confidence. They were also informed that if they agreed to participate, they would receive $50 as a token of appreciation for participating in the survey. A second automated consent protocol was administered at the beginning of the ACASI portion of the interview to confirm that the respondent had been properly informed that participating in the survey was voluntary and that they were "ready to continue with the interview." Measuring sexual victimization The NFPS first screened for inmate-on-inmate sexual touching "when they didn't want this to happen" and sexual activity "in which they did not want to participate." Respondents were then asked about the specific sexual activity, including oral, anal, or vaginal penetration, and about the specific contact, including touching of buttocks, thighs, breasts, penis, or vagina in an attempt to hurt or arouse the victim or the perpetrator. (See appendix A and B for specific survey questions.) Reports of inmate-on-inmate sexual victimization were classified as either nonconsensual sexual acts or abusive sexual contacts. The NFPS also asked about staff sexual activity, including staff sexually harrassing the inmate, staring inappropriately at the inmate, and forcing the inmate to undress or brushing private parts when "it was not required by their job." Former inmates were then screened for sexual contact with staff, "whether it was willing or not," or "whether you wanted to have it or not." These contacts may have included oral, anal, or vaginal penetration and other forms of sexual stimulation through rubbing the penis and touching other private parts. (See appendix C and D for specific survey questions.) Reports of staff sexual misconduct involving physical force, threat of force, fear of bodily injury, and being pressured or made to feel they had no choice were classified as "unwilling." Other reports of staff sexual misconduct in which former inmates reported willingly having sex or sexual contact with staff were classified as "willing" even though any sex or sexual contact between inmates and staff is illegal. (See Definition of terms on page 13.) ************************************************************** Former prisoners reported a wide variety of other sexual experiences with staff that were inappropriate: * An estimated 8.9% reported that staff had hassled or harassed them in a sexual way. * 27.9% said that staff had stared or watched them at inappropriate times (e.g., while the inmate was dressing or taking a shower). * 13.5% said that staff had forced them to undress in their presence or had brushed against their private parts when "they did not think it was an accident or it was not required by their job." Nearly a third (32.4%) of all former inmates reported one or more of these types of experiences. While inappropriate, these lesser forms of staff sexual misconduct were not included in the analysis unless combined with reports of "willingly" or unwillingly having sex or sexual contact with staff. (See appendix D for survey items related to staff-on-inmate sexual victimization.) Few former inmates reported experiencing sexual victimization while in a community-based correctional facility Correctional facilities are typically classified as community-based if 50% or more of the residents are regularly permitted to leave unaccompanied by facility staff to work or study in the community. Community-based facilities include entities such as halfway houses, residential treatment centers, restitution centers, and pre-release centers. Although community-based correctional facilities are covered by the Prison Rape Elimination Act (PREA), little data have been available on the rate of sexual victimization in these facilities. Despite the large number of such facilities (529 of the 1,719 state correctional facilities in 2005), they hold few inmates on any single day (54,233 inmates at yearend 2005, or approximately 4% of all state inmates held nationwide) (See Census of State and Federal Correctional Facilities, 2005, October 2008, BJS Web, NCJ 222182, appendix tables 2 and 10.) As a result of their small size (an average of 102 inmates per facility) and the relatively short length of stay for most inmates while in community-based facilities, inmates held in these facilities have been excluded from previous PREA-related inmate surveys. The NFPS provides the first systematic data available on sexual victimization in community-based facilities. An estimated 66,400 former state prisoners (13%) said they had served some time in a community-based treatment facility or halfway house after release from prison. Among those inmates who had served time in such a facility, 0.9% reported experiencing one or more incidents of sexual victimization while in the facility (table 2). Among all former state prisoners, 0.1% said they had been sexually victimized while in a post-release community-treatment facility. Approximately 65,500 former inmates (13%) reported spending time in a community-based facility before their release from prison. During the time they had been placed in such a facility, 2.0% reported experiencing one or more incidents of sexual victimization: 0.7% reported inmate-on-inmate sexual victimization, while 1.4% reported staff-on-inmate victimization. Sexual victimization rates differed from those previously reported in the NIS-1 and NIS-2 The rate of sexual victimization reported by former state prisoners (9.6%) was substantially higher than the rates reported in the previous BJS National Inmate Surveys which were based on confined state and federal inmates (e.g., 4.5% in the NIS-1 conducted in 2007 and 4.4% in the NIS-2 conducted in 2008-09). The differences may largely reflect longer average exposure time among former inmates. Unlike NFPS, the NIS-1 and NIS-2 provide facility-specific estimates of the prevalence of sexual victimization. As required under the PREA of 2003, these surveys are designed to provide a list of prisons and jails ranked by the prevalence of sexual victimization. To eliminate experiences that may have occurred in other facilities or in the distant past and to control for the varying length of stay, the NIS-1 and NIS-2 asked inmates to provide the most recent date of admission to the current facility. To provide comparative rates, if the date of admission was at least 12 months prior to the date of the interview, inmates were asked questions related to their experiences during the past 12 months. If the admission date was less than 12 months prior to the interview, inmates were asked about their experiences since they arrived at the facility. As a consequence, the average exposure period among state prisoners participating in the NIS-2 was 7.9 months, compared to the average of 39.4 months that the former inmates in the NFPS had served in state prisons prior to their release (excluding time served in a local jail or post-release community-treatment facility). (See Methodology for differences in coverage between NFPS and NIS.) ********************************************************** Criminal history and supervision profile of former inmates In addition to collecting data on the sexual experiences of former state inmates during their most recent period of confinement, the NFPS gathered information on their criminal histories and their current supervision (table 3). Results from the survey include the following: * Among former state inmates under active supervision, 47% had been on parole or community supervision for less than 12 months, 24% for 12 to 23 months, 12% for 24 to 35 months, and 17% for 36 months or more. * About a third of the former inmates under active parole supervision had served time for a violent offense (33%) or a drug offense (30%), a quarter for a property offense (25%), and an eighth for a public-order offense (12%). * Nearly 21% of the former inmates under active supervision had served time for a probation violation; 21% for a parole violation. * Nineteen percent of the former state prison inmates had served less than 12 months in prison or jail before their release from prison; 25% had served 12 to 23 months; 16%, 24 to 35 months; 17%, 36 to 59 months; and 23%, 60 months or more. * More than half of those under supervision (56%) had been released as a result of a parole board decision. * As a condition of their release, more than 90% of the parolees were required to submit to drug testing; 78% were required to be employed. * Participation in a treatment or counseling program was also required of 53% of the parolees for drugs; 39% for alcohol; and 26% for issues other than drugs or alcohol. * Nearly 80% of parolees said they had a face-to-face meeting in the last month with their parole officer in the parole office; 57% said they had a face-to-face meeting outside of the office (at home, at work, or elsewhere); and 48% said they had been contacted by mail, e-mail, or telephone. **************************************************************** ************************************** Type of coercion and physical injury ************************************** Among victims of inmate-on-inmate sexual violence, a quarter had been physically held down or restrained and a quarter had been physically harmed or injured Incidents of inmate-on-inmate sexual victimization involved a wide array of acts with differing levels of coercion and force to get the victim to participate. Although many of the victims were not raped, all of the victims had been sexually abused. Force or threat of force or harm was reported by 59% of victims of inmate-on-inmate sexual victimization about 3.2% of former state prisoners overall (not shown in table). An estimated 53% said they had been threatened with harm or a weapon; 28% had been physically held down or restrained, and 25% had been physically harmed or injured. (table 4). The most common form of coercion other than force or threat of force was being "persuaded or being talked into it" (34%) against their will. A quarter of victims (26%) said the inmate perpetrator had given them a bribe or blackmailed them, and a quarter (24%) said they had been offered or given protection from other inmates. In addition, some victims had been given drugs or alcohol to get them drunk or high (9%), and some had sexual contact to pay off or settle a debt that they owed (8%). Overall, 52% of victims reported one or more of these types of coercion. Among victims of inmate-on-inmate sexual victimization, the most common injuries were anal or vaginal tearing, severe pain or bleeding (12%) and chipped or lost teeth (12%). Although 29% of all victims reported bruises, black eyes, sprains, cuts, scratches, swelling, or welts from one or more incidents involving another inmate, nearly a quarter (23%) also reported more serious injuries Half of victims of staff sexual misconduct said they had been offered favors or special privileges; a third had been persuaded or talked into it; a quarter had been bribed or blackmailed Most victims of staff sexual misconduct said they "willingly" had sex or sexual contact with staff; however, many of these victims also reported sexual contact or activity with staff that involved some form of coercion. Among the estimated 23,300 former prisoners who reported some type of "willing" activity with staff, 11% (2,500) also reported unwilling sexual activity with staff. The most common form of coercion reported by the 23,300 "willing" victims was offers of favors or special privileges by staff (49%), followed by being persuaded or talked into it by staff (34%) and being given a bribe or blackmailed (24%). Overall, 62% of inmates who had engaged in "willing" sex or sexual activity with staff reported some type of coercion or offer of favors, special privileges, or protection by staff. An estimated 6,300 former state prisoners (1.2%) reported unwilling sexual activity with staff during their most recent period of incarceration. These unwilling victims (48%) were more likely than "willing" victims of staff sexual misconduct (5%) to report having been physically forced or threatened by staff. An estimated 44% of unwilling victims said they had been threatened with harm or a weapon, 19% said they had been physically held down, and 15% said they had been physically harmed or injured. Nearly 7 of every 8 inmates who reported unwilling sexual activity with staff said that staff had employed other forms of coercion to engage them in sexual activities or have sexual contact. In specific, 72% of the victims were offered favors or special privileges, 72% were persuaded or talked into it, 60% were given bribes or blackmailed, 34% were offered protection from other staff, and 21% were given drugs or alcohol. These reports suggest that these former inmates experienced a wide array of pressures from staff to have sexual contact against their will. More than 80% of victims who reported force and threat of force also reported one or more of these other forms of coercion (not shown in table). Four percent of victims of staff sexual misconduct reported that they were physically injured Approximately 16% of victims of unwilling sexual activity with staff, compared to 2% of victims of "willing" sexual activity, reported being physically injured by staff. Among victims of unwilling activity, the most commonly reported injuries were bruises, black eyes, sprains, cuts, scratches, swelling or welts (11.8%). Most of these victims of unwilling sexual activity also reported more serious injuries, including anal or vaginal tearing (8.8%), loss or chipping of teeth (6.3%), and receiving internal injuries (4.9%). Though few victims of "willing" sexual activity with staff reported being injured, some (1.6%) reported a serious injury. *********************************************************** Definition of terms Sexual victimization all types of unwanted sexual activity with other inmates (e.g., oral, anal, or vaginal penetration, hand jobs, or touching of the inmate's buttocks, thighs, penis, breasts, or vagina in a sexual way), abusive sexual contacts with other inmates, and both willing and unwilling sexual activity with staff. Nonconsensual sexual acts unwanted contact with another inmate or any contact with staff that involved oral, anal, vaginal penetration, hand jobs, and other sexual acts. Abusive sexual contacts only unwanted contact with another inmate or any contact with staff that involved touching of the inmate's buttocks, thighs, penis, breasts, or vagina in a sexual way. Unwilling activity incidents of unwanted sexual contact with another inmate or staff. Willing activity incidents of willing sexual contacts with staff. These contacts are characterized by the reporting inmates as willing; however, all sexual contacts between inmates and staff are legally nonconsensual. Staff sexual misconduct all incidents of willing and unwilling sexual contact with facility staff, and all incidents of sexual activity that involved oral, anal or vaginal penetration, hand jobs, and other sexual acts with facility staff. ********************************************************* *************************************** Circumstances surrounding victimization *************************************** Inmate-on-inmate victimization occurred most often in the victim's cell; staff-on-inmate victimization occurred most often in a closet, office, or other locked room In the NFPS victims were asked to provide additional information about the circumstances surrounding their victimization, including the number of times it had happened, the number of different perpetrators, when each incident occurred, and where it occurred in each facility. Data provided by inmates who reported sexual victimization by another inmate revealed that * Victimization was more common in the evening (between 6 p.m. and midnight) (52%) than at any other time (table 5). * Almost two-thirds (62%) of the victims said at least one incident took place in their cell or room, and a quarter said an incident took place in another inmate's cell or room (24%). * An estimated 46% of inmate-on-inmate victims had been victimized by more than one perpetrator. About 42% had been victimized once; 31% had been victimized 3 or more times. Most victims (86%) of staff sexual misconduct reported more than one incident; 47% reported more than one perpetrator Data provided by inmates who had been sexually victimized by facility staff also revealed that * Reports of staff sexual misconduct were more common between midnight and 6 a.m. (54%) than at any other time. * More than two-thirds (70%) of the victims said at least one incident had occurred in a closet, office, or other locked room; half (54%) of the victims said an incident had occurred in a work area; and a third (35%) in a shower or bathroom. * Reports of when and where incidents had occurred were similar among willing and unwilling victims. *************************************** Variations by sex of former inmates *************************************** Past BJS surveys of confined prison inmates have consistently found higher rates of inmate-on-inmate victimization among females than males. In the National Inmate Survey, 2008-09, 4.7% of the surveyed female inmates and 1.9% of the male inmates reported being sexually victimized by another inmate. This difference was found to be statistically independent and largely unexplained by covariation with other demographic characteristics (e.g., an inmate's race or Hispanic origin, age, education, marital status, and weight). (See Sexual Victimization in Prisons and Jails Reported by Inmates, 2008-09, tables 6 and 7.) The reports of former prisoners confirm the large and statistically significant difference between male and female rates of inmate-on-inmate sexual victimization (table 6). The rate of inmate-on-inmate sexual victimization among former state prisoners was 3 times higher among females (13.7%) than males (4.2%) When the rate of sexual victimization was limited to nonconsensual sexual acts including only incidents of manual stimulation and oral, anal, or vaginal penetration, the difference between females and males was large. An estimated 10.5% of females reported such incidents with other inmates, compared to 2.7% of males. The rate of "willing" sexual activity with staff was higher among males (4.8%) than females (2.6%), and the rate of unwilling sexual activity was higher among females (2.5%) than males (1.1%) Most victims of staff sexual misconduct (87%) reported only perpetrators of the opposite sex (table 7). Same-sex victimization was more likely to be reported by victims of unwilling staff sexual misconduct than by victims of "willing." Approximately a quarter of the reports of unwilling sexual activity with staff involved male inmates with male staff only (23%) and female inmates with female staff only (3%). **************************************************** Variations by other individual-level characteristics **************************************************** Large differences in sexual victimization were found among former inmates based on their sexual orientation An estimated 3.5% of male heterosexual former inmates reported being sexually victimized by another inmate, and 5.2% reported being victimized by staff (table 8). In comparison, 34% of male inmates who were bisexual reported being sexually victimized by another inmate, and 18% reported being sexually victimized by staff. Thirty-nine percent of male inmates who were homosexual or gay indicated they had been victimized by another inmate and 12% by staff. * Bisexual female former inmates reported a higher rate of inmate-on-inmate victimization (18%) than heterosexual inmates (13%) and lesbian inmates (13%). * Lesbian inmates and bisexual female former inmates had rates of staff sexual misconduct that were at least double (8%) the rate among heterosexual female former inmates. Rates of inmate-on-inmate sexual victimization among males varied by race and Hispanic origin. Rates of inmate-on-inmate victimization for males were higher among white non-Hispanic inmates (5.9%) and multi-racial inmates (9.5%) than among black non-Hispanic inmates (2.9%). Among females, rates of inmate-on-inmate victimization did not vary significantly across other demographic characteristics. Patterns of staff sexual misconduct among male former prisoners varied among multiple demographic characteristics. Rates of staff sexual misconduct were higher for those of two or more races (11.3%) and lower for white non-Hispanic inmates (4.5%) and Hispanic inmates (4.0%), when compared to black non-Hispanic inmates (6.5%). Rates of staff sexual misconduct were also higher among younger male inmates (under age 25) than among older male inmates; higher among those with less than a high school education than those who completed high school; and higher for those who were married than those who were widowed, divorced, separated, or never married. Among female former prisoners, rates of staff sexual misconduct were significantly lower among those ages 35 to 44 (3.1%) and age 45 or older (1.6%), compared to those ages 20 to 24 (6.7%). The only other statistically significant difference among female inmates was that 11.9% of multi-racial inmates reported staff sexual misconduct, which was approximately 2 to 3 times the rate of victimization reported by white non-Hispanic inmates (4.5%), black non-Hispanic inmates (4.2%), and Hispanic inmates (3.8%). Violent, male sex offenders reported high rates of inmate-on-inmate sexual victimization (13.7%) Consistent with findings in previous BJS data collections based on inmates held in prisons and local jails, male former state prisoners whose most serious offense was a violent sex offense reported significantly higher rates of inmate-on-inmate sexual victimization than other inmates (table 9). The rate among male sex offenders (13.7%) was more than twice the rate among other violent male offenders (4.9%), 3 times the rate of property offenders (4.4%), 5 times the rate of public-order offenders (2.6%), and 6 times the rate of drug offenders (2.0%). Among female former inmates, no differences were found in rates of inmate-on-inmate sexual victimization by type of offense for which they had been sentenced. Rates of sexual victimization increased with the length of time that former inmates had served Rates of sexual victimization by other inmates and staff increased with the length of time the former inmate had served during the most recent incarceration in jail, prison, or post-release community-treatment facility. Among males, rates for former inmates who had served 5 years or more were at least 3 times higher than for those who had served less than a year. Among males who had served 10 years or longer, 12% reported having been sexually victimized by other inmates and 17% by staff. Among females who had served 5 to 10 years before their release, 21% reported having been sexually victimized by other inmates and 11% by staff. A quarter of the females who had served 10 years or more reported experiencing staff sexual misconduct. Male inmates with no prior incarceration experience had a higher rate of inmate-on-inmate sexual victimization (6.6%) compared to those with one or more prior incarcerations (3.7%). Rates of staff sexual victimization for males and rates for both inmate and staff sexual victimization for females did not vary by prior incarceration history. Former state prisoners who had served time in 5 or more facilities during their most recent confinement reported the highest rates of staff sexual misconduct Staff sexual misconduct rates varied by the number of facilities in which male and female inmates had served time during their most recent period of incarceration. Approximately 25% of the male former state prisoners and 10% of the female state prisoners had served time in five or more facilities before their release. These inmates reported significantly higher rates of staff sexual misconduct (6.7% or more among males and 7.7% or more among females) than those who had served in three or fewer facilities (4.0% or less among males and females). These differences are consistent with the practice of transferring victims to other facilities in response to substantiated incidents of staff sexual misconduct. In 2006, based on reports by correctional authorities of all substantiated incidents of sexual victimization, nearly a third of the victims of staff sexual misconduct were transferred to another facility 19% of victims in prisons and 48% of victims in jails. (For further details, see Sexual Violence Reported by Correctional Authorities, 2006, April 2007, BJS Web, NCJ 218914, table 12.) The increase in sexual victimization rates by number of facilities may also reflect covariation with total time served, with inmates who serve longer sentences being placed in more facilities. Difference in findings after BJS performed multivariate logistic regressions Among male former prisoners, after controlling for the effects of other individual-level factors * Inmate-on-inmate victimization was higher for white non-Hispanic inmates, persons of two or more races, those age 24 or younger, non-heterosexuals, sex offenders, and those who served 5 years or more than for inmates with other characteristics (table 10) Staff sexual misconduct was higher for black non-Hispanic inmates, persons of two or more races, those age 24 or younger, non-heterosexuals, those with one or more prior incarcerations, and those who served 12 months or more than for inmates with other characteristics. Among female former prisoners, after controlling for the effects of other individual-level factors * Inmate-on-inmate victimization was higher for inmates of two or more races, those age 24 or younger, those with some college education, those with one or more prior incarcerations, and those who had served a year or more but less than 10 years than for inmates with other characteristics. * Staff sexual misconduct was higher for white non-Hispanic inmates, inmates of two or more races, those age 24 or younger, those with one or more prior incarcerations, and those who served 3 years or more than for inmates with other characteristics. * Sexual victimization either by other inmates or by staff was unrelated to sexual orientation. For both male and female former prisoners, the number of facilities in which the inmate had served time was no longer related to reports of staff sexual misconduct. For male former inmates, education levels and marital status were no longer correlated with reports of staff sexual victimization. ************************************************************** BJS used multivariate logistic regression models to determine which inmate characteristics were significant predictors of whether the inmate would be sexually victimized in prison Multivariate logistic regression estimation is a modeling technique used to determine what characteristics are statistically significant for predicting a dichotomous outcome (e.g., an inmate is victimized or not victimized) while controlling for all the other characteristics in the model. BJS used this technique to determine which inmate characteristics were statistically significant for predicting sexual victimization. Based on four separate models, each representing the type of sexual victimization and sex of the former inmate, the variations in rates of sexual victimization by demographic characteristic, sexual orientation, and criminal justice status were found to be largely statistically independent of one another. Estimates are displayed in terms of the conditional marginal probabilities, which represent the probability that a former male or female inmate with a particular characteristic has experienced a specific sexual victimization outcome conditional on the former inmate having the mean value for all other predictors in the model. For example, based on models with demographic and criminal justice status characteristics only, a white male former inmate had a 4.0% chance of having been sexually victimized by another inmate, and a female former inmate had a 12.2% chance (conditional on the inmate having the mean value on all of the other characteristics in the model). For characteristics that are categorical, which is the case for every variable in these logistic regression models, the mean value is a weighted value of the joint distribution of all other characteristics in the respective model. (See Methodology for a discussion of logistic regression ************************************************************** ****************************************************** Variations by selected facility-level characteristics ****************************************************** To understand more fully the correlates of sexual victimization and circumstances surrounding that victimization, BJS included survey questions to identify each of the state facilities in which the inmates had served time during their most recent period of incarceration and asked if the inmates had been sexually victimized (by type of victimization) within each facility. (See appendix E for specific survey questions.) Data collected in prior BJS censuses of state prison facilities (conducted in 1990, 1995, 2000 and 2005) were linked to these reports to determine whether any facility characteristics were associated with variations in rates of sexual victimization. Although data on local jails and post-release community-treatment facilities were also collected, they have been excluded from this analysis. (See Methodology for details on linkage of prison census data.) Three-quarters of former inmates had served time in more than one prison facility; nearly 1 in 8 had served time in 5 or more prison facilities before their release During their period of incarceration, inmates typically served time in more than one facility. Covering only the last continuous period of incarceration, the 18,526 former prisoners interviewed in NFPS reported 63,813 separate placements. Of these, 16,073 involved jail placements (either prior to prison, as an interim stop or as the only placement for some). In addition, 2,422 placements were to post-release local facilities, and 1,525 were to out-of-state facilities. When weighted by the inverse of the probability of selection, the eligible prison-only placements represented more than 1.3 million separate placements (table 11). Overall, an estimated 500,400 former prisoners (98%) had served time in a prison facility in the state in which they were currently under parole or post-custody supervision, excluding 10,400 prisoners who had served time only in a local jail. Approximately three-quarters of these inmates had served time in more than one prison facility; nearly an eighth had served time in five or more prison facilities. Nearly 44% of male victims and 74% of female victims reported that sexual victimization had occurred in the first prison facility they had entered Based on the self reports of victims in the NFPS, sexual victimization frequently occurred in the first facility to which the inmates had been admitted. Among victims, nearly half (49%) said they were victimized in the first prison they had entered (table 12). Female victims (74%) were much more likely than male victims (44%) to report having been victimized in the first prison entered. This difference in victimization may be partially explained by differences among males and females in the likelihood of serving time in more than one facility. An estimated 78% of males, compared to 57% of females, had served time in two or more prison facilities before their release; nearly 25% of males, compared to 7% of females, had served time in four or more prison facilities (not shown in table). Among victims, sexual victimization by other inmates was more likely than sexual victimization by staff to have occurred in the first facility the inmates had entered. More than half (57%) of the estimated 18,800 prisoners who reported sexual victimization by another inmate indicated that it occurred in the first prison facility they entered. Fewer than half (42%) of the 23,800 victims of staff sexual misconduct indicated it occurred in the first facility. Reports of sexual violence did not vary by size of facility, facility age, crowding, inmate-to-staff ratios, or sex composition of staff Further analysis of prison placements provided a detailed risk profile of facility-level characteristics in conjunction with reported sexual victimization. Overall, few measurable differences at the 95% level of statistical confidence emerged among commonly cited facility risk factors (table 13). Sexual victimization rates (defined by type of victimization and sex of inmate) did not increase with * increased size of the facility (based on the average daily population during the 12 months prior to the census closest in time to the inmate's placement) * increased age of the facility (based on year of original construction) * facility overcrowding (based on the ratio of the number of inmates the facility actually held to the official rated capacity) * increased inmate-to-staff ratios in the facility (based on the number of inmates held to the total number of payroll, non-payroll, and contract staff) * decreased proportion of female to male staff. Sexual victimization rates varied by type and primary function of the facility and by indicators of facility disorder Among male former state prisoners, inmate-on-inmate and staff-on-inmate victimization rates were higher in facilities * under a court order or consent decree (to limit the number of inmates, to address crowding, or in response to the totality of conditions at the time of the census) than in other facilities * reporting a major disturbance (involving five or more inmates and which resulted in serious injury or significant property damage) in the 12 months prior to the census than in other facilities * with medium or greater security levels than in facilities with minimum or low security levels * with a primary function of housing general population than in facilities with specialized functions (e.g., reception or diagnostic centers, community corrections centers, medical treatment, and alcohol or drug treatment). Among female former state prisoners, reports of inmate-on-inmate victimization were lower in community corrections centers (1.3%) and facilities that permitted 50% or more of their inmates to leave unaccompanied during the day (1.3%), than in facilities holding general population (5.3%) or in general confinement facilities (5.0%). Inmate-on-inmate sexual victimization rates were also lower in minimum or low security facilities (3.0%) than in medium security facilities (5.6%) and lower in privately operated facilities (2.0%) than in state facilities (5.0%). Reports of staff sexual misconduct among female former state prisoners did not vary significantly across facility characteristics, except for being lower in minimum or low security facilities (1.0%) than in medium security facilities (3.0%) and higher in facilities under a court order (6.3%) than in other facilities (2.0%). Whether a facility was publicly operated or privately operated was an inconsistent predictor of victimization. While reports by male former state prisoners of staff sexual misconduct were higher in privately operated facilities (4.6%) than in state facilities (2.5%), reports by female former prisoners did not differ statistically between state and privately operated facilities. Rates of inmate-on-inmate victimization among women, but not among men, were lower in privately operated facilities than in state facilities. When an inmate had been placed in the facility was a strong predictor of sexual victimization among male former prisoners. As measured by the census year closest to the time of placement, male former state inmates were nearly 4 times more likely to report victimization by another inmate and 3 times more likely to report victimization by staff if they had entered the facility around the 1990 census than if they had entered around the 2005 census. Female inmates who entered the prison facility around 1990 were nearly 8 times more likely to report staff sexual misconduct (12.5%) than if they had entered around 2005 (1.6%). However, rates of inmate-on-inmate sexual victimization among females did not differ by year of placement. **************************************************** Prison placements and individual-level risk factors **************************************************** While individual-level factors related to sexual victimization have been examined by type of victimization and sex of inmate for the entire period of incarceration, they may also be examined for each prison placement. Such a placement-based examination takes into account the risk of victimization related to individual-level factors, regardless of the duration of time served during a specific prison placement, and permits analysis of the independent contribution of individual-level and facility-level factors. Unlike the earlier analysis based on individuals (510,800 former inmates nationwide), the placement analysis is based on prison placements (more than 1.25 million prison placements). Individuals are represented multiple times, based on their number of separate prison placements (up to a maximum of 15 times per inmate). An incident of sexual victimization was reported in 4.0% of prison placements Among the estimated 1.15 million placements of male inmates, 1.6% involved inmate-on-inmate sexual victimization and 2.6% involved staff-on-inmate victimization (table 14). Among the estimated 102,500 placements of female inmates, 4.8% involved sexual victimization by another inmate and 2.4% involved sexual victimization by staff. Overall, in 4.0% of the more than 1.25 million prison placements, former prisoners reported at least one experience of sexual victimization by another inmate or facility staff. These rates of victimization among all prison placements are similar to those reported by inmates in NIS-2 based on the current facility (during the last 12 months or since admission, if shorter). Based on the NIS-2, 2.0% of currently confined male inmates reported sexual victimization by another inmate and 3.2% by staff; 5.1% of currently confined female inmates reported sexual victimization by another inmate and 2.3% by staff. Patterns of sexual victimization based on prison placements also resembled those observed for individuals based on their entire period of incarceration in NFPS, except that the rates were somewhat lower across categories defined by demographic and criminal justice characteristics: * Large differences in sexual victimization during prison placement were found among inmates based on their sexual orientation (table 15). * Among male former inmates, reports of inmate-on-inmate sexual victimization during prison placement varied by race and Hispanic origin. Reports were higher among white non-Hispanic inmates (2.4%) and multi-racial inmates (4.0%) than among black non-Hispanic inmates (0.9%). * Among male inmates, violent sex offenders reported higher rates of inmate-on-inmate sexual victimization (5.9%) than other violent offenders (1.4%), property offenders (1.7%), public-order offenders (1.0%), and drug offenders (0.8%). * Among male former inmates, those with a prior incarceration (1.4%) were less likely to be victims of other inmates than those with no prior incarceration (2.5%). * Among female former inmates, rates of inmate-on-inmate victimization during prison placement did not significantly vary across demographic characteristics, except among inmates of two or more races. * Among placements of male former prisoners, reports of staff sexual misconduct were higher among black non-Hispanic inmates (3.3%) than among white non-Hispanic inmates (2.0%) and Hispanic inmates (1.7%). The rates were also higher among younger male inmates (under age 25) than among older male inmates (age 35 or older). * Among placements of female former prisoners, rates of staff sexual misconduct were significantly lower among those ages 35 to 44 (1.5%) and age 45 or older (1.1%), compared to those ages 20 to 24 (4.0%). * The only other statistically significant difference among placements of female former inmates was that 8.1% of multi-racial inmates reported staff sexual misconduct, compared to 2.4% or lower among inmates who were white, black, or Hispanic. ************************************************** Independent contributions of individual-level and facility-level factors to victimization ************************************************** Variations in rates of sexual victimization among groups of inmates based on the characteristics of facilities in which they were placed overlap somewhat with variations based on individual risk factors. Multivariate regression models were developed to incorporate selected individual-level and facility-level factors. Models for each type of sexual victimization were developed separately for male and female former prisoners based on prison placements (table 16). In each of the logistic regression models, the conditional probabilities represent the probability that a former inmate during a prison placement with a particular individual or facility characteristic experienced sexual victimization, conditional on the inmate having the mean value for all other predictors in the model. For example, based on the final multivariate model, a male former state prisoner who served time for a violent sex offense had a 2.2% chance of having been sexually victimized by another inmate during a prison placement, given that he was at the mean of the joint distribution of other individual characteristics (i.e., race or Hispanic origin, marital status, sexual orientation, and time served) and facility characteristics (i.e., type, security level, census year closest to placement, and placement order). Variations in sexual victimization rates were strongly related to sexual orientation after controlling for other factors An inmate's sexual orientation remained an important correlate of victimization, except for staff sexual misconduct reported by females. Male inmates who identified themselves as bisexual were at least 13 times more likely to report being sexually victimized by another inmate and 4 times more likely to report being sexually victimized by staff than male inmates who identified themselves as straight or heterosexual. After controlling for other factors, an inmate's race or Hispanic origin remained an important predicator of sexual victimization. Based on the multivariate placement models for male former inmates * White non-Hispanic inmates (1.4%) and inmates of two or more races (2.0%) had higher rates of inmate-on-inmate victimization than black non-Hispanic and Hispanic inmates (each 0.6%). * Black non-Hispanic inmates (2.0%) had higher rates of staff-on-inmate victimization than white non-Hispanic (1.4%) and Hispanic inmates (1.2%). Based on the multivariate placement models for female former inmates * Inmate-on-inmate victimization rates were lower for black non-Hispanic inmates (2.8%) than for white non-Hispanic inmates (3.9%), Hispanic inmates (4.6%) and inmates of two or more races (8.7%) * Staff-on-inmate victimization rates were higher for inmates of two or more races (6.1%) than for white non-Hispanic (1.7%), black non-Hispanic (1.1%) and Hispanic inmates (1.1%). Facitily-level factors remained significant after controlling for inmate characteristics * Among male former inmates, inmate-on-inmate and staff-on-inmate victimization rates were lower in reception or diagnostic centers than in facilities that housed general population. For inmate-on-inmate victimizations, rates were lower in minimum security facilities than in facilities with higher security levels. * Among female former inmates, the inmate-on-inmate victimization rate was lower in facilities that permitted 50% or more of their inmates to leave unaccompanied during the day (than in other confinement facilities), lower in medium security facilities (than in facilities with maximum, close, or high security levels), higher in facilities that had been built before 1951 or after 1985 (than in facilities built between 1951 and 1985), and higher in facilities that had a major disturbance in the census year (than in those that did not). * Among female former inmates, few facility-level factors were linked to staff sexual misconduct. Rates were lower in minimum or low security facilities than in medium security facilities, and lower in facilities built between 1951 and 1985 than in facilities built before 1951. The final regression models confirm that the best predictor of staff sexual misconduct among female former inmates was length of time served, with the rate of staff sexual misconduct during prison placements generally increasing with total time served. *********************************** Reporting of sexual victimization *********************************** In addition to identifying key individual-level and facility-level risk factors associated with variations in sexual victimization rates, the former inmate survey provided data on the responses of victims and the impact on victims while they were still in prison and after their release from prison. The survey included questions about reporting victimization to facility staff and persons other than facility staff, reasons for not reporting and what happened if they did report to facility staff, and the continued impact on the victims following their release from prison. Two-thirds of victims of inmate-on-inmate sexual victimization said they reported at least one incident to facility staff or someone else Detailed data suggest that victims of inmate-on-inmate victimization were most likely to have reported at least one incident to another inmate (49%), followed by having reported to a family member or friend outside the facility (28%) and to correctional officers (24%). Few victims (1.4%) had reported their experience on a telephone hotline (table 17). An estimated 37% of victims of inmate-on-inmate sexual victimization reported at least one incident to facility staff 58% of those injured, compared to 29% of those not injured (not shown in table). Reporting to medical and healthcare staff was even less common (14% of the victims said they reported an incident to medical personnel). Among those physically injured by another inmate, a third (33%) said they reported to medical staff (not shown in table). 22% of unwilling victims of sexual activity with staff, compared to 3% of "willing" victims, said they had reported an incident to facility staff or someone else "Willing" sexual activity with staff was rarely reported to any facility staff (2%) or to other persons (2%). When sexual activity with staff was reported, it was typically unwilling and divided nearly equally between reporting to staff (21%) and someone other than staff (20%). Most common reasons for not reporting sexual victimization by other inmates linked to embarrassment, shame, and not wanting others to know Victims who never reported their experience of sexual victimization to anyone at the facility were asked their reasons for not reporting. The most common reason for victims not reporting sexual victimization by another inmate was that they "didn't want anyone else to know about it" (70%), followed by feeling "embarrassed or ashamed that it happened" (68%) (table 18). The most common reasons for not reporting staff sexual misconduct were that they "had the sex or sexual contact willingly" (80%) and they "didn't want the staff person to get in trouble" (71%). Almost 40% of victims of inmate-on-inmate sexual victimization and staff sexual misconduct said they were afraid of being punished by staff. Nearly a quarter of inmates victimized by another inmate or by staff said they were afraid of being charged with making a false report. 37% of victims who reported being victimized by other inmates said facility staff did not respond Among victims who reported inmate-on-inmate sexual victimization to authorities * 34% were moved to administrative segregation or protective housing, 24% were confined to their cell, 14% were assigned a higher level of custody, and 11% were placed in a medical unit or hospital. *18% were offered a transfer to another facility. * More than a quarter (29%) said they were written up; and 28% said they spoke to an investigator (table 19). Though few in number, 86% of the victims who reported staff sexual misconduct to authorities said the facility responded: more than half (54%) of the victims spoke to an investigator, and nearly half (46%) said that they were "written up" for the incident. More than 40% said they were moved to administrative segregation or protective housing. **************************************** Post-release responses to victimization **************************************** Following their release from prison, victims of inmate-on-inmate sexual victimization * had feelings of shame or humiliation (72%) and guilt (56%) (table 20) * discussed their victimization with intimate partners (27%) or friends (26%) * saw a therapist or counselor (15%), participated in a self-help group for emotional or mental health problems (14%), or enrolled in a treatment program other than counseling (8%). Victims of staff sexual misconduct that involved unwilling activity reported many similar feelings or thoughts about their victimization following their release from prison: * 79% said they felt shame or humiliation. * 72% said they felt guilt. * 54% reported having difficulty feeling close to friends or family members. Although the vast majority of victims (86%) of staff sexual misconduct said they had at least one incident that they considered "willing," more than a quarter (27%) said they felt guilt and 23% said they felt shame or humiliation after their release from prison. ************************ HIV testing and results ************************ The former inmate survey also provided self-report data on HIV testing and outcomes for victims of inmate-on-inmate and staff-on-inmate sexual victimization. * Among former inmates tested for HIV, a significantly higher percentage of those who had been sexually victimized by other inmates (6.5%) or by staff (4.6%) were HIV-positive, compared to those who had not been victimized (2.6%) (table 21). * Among all victims tested for HIV, 5.1% of males and 3.9% of females were HIV-positive (not shown in table). * Among all victims tested for HIV, 15.5% of non-heterosexual inmates were HIV-positive, compared to 3.8% of heterosexual victims (not shown in table). Although the rates of HIV infection are high among former prisoners, when these inmates contracted HIV is unknown. Inmates were not asked if they became HIV-positive while incarcerated. Through the National Prisoner Statistics program, BJS has collected data on the number of inmates HIV-positive and the number with confirmed acquired immune deficiency syndrome (AIDS) since 1991. At yearend 2008, 20,499 inmates in state prisons (or 1.6% of all inmates in custody) were known to be HIV-positive or had confirmed AIDS. (See HIV in Prisons, 2007-08, NCJ 228307, BJS Web, December 2009.) Data are not available on the percentage of inmates who were HIV-positive when they were admitted to prison. ***************************************** Current employment, housing, and living arrangements of former inmates ***************************************** Following their release from prison, victims and nonvictims did not differ in their employment and housing arrangements, but victims were more likely than nonvictims to be living alone The NFPS asked former prisoners about their current employment status, housing, and living arrangements while on post-release supervision. Except for a slightly higher percent of victims living alone (18%) than nonvictims (14%), few differences were reported in their current employment status or housing arrangements: * Two-thirds of both victims and nonvictims were employed; at least half of each group were employed full-time. * A quarter of victims and nonvictims were looking for work. * More than 90% of victims and nonvictims were currently living in a house, apartment, trailer or mobile home. * An estimated 2% of victims and nonvictims said they were in a shelter, homeless, or on the street (table 22). ************ Methodology ************ The National Former Prisoner Survey (NFPS) was conducted between January 2008 and October 2008. The data were collected by the NORC at the University of Chicago under a cooperative agreement with the Bureau of Justice Statistics. A total of 317 total parole offices in 40 states were included in the survey sample. Within each of the sampled offices, respondents were selected at random from rosters of eligible former prisoners. A total of 17,738 former state prisoners under active supervision participated in the survey. An additional 788 former prisoners were included from survey test offices. These former inmates had been randomly selected from 16 sampled offices. Based on 18,526 completed interviews, the survey achieved a 61% response rate. The interviews, which averaged 23 minutes in length, were conducted using computer-assisted personal interviewing (CAPI) and audio computer-assisted self-interviewing (ACASI) data collection methods. Survey interviewers initiated the personal interview using CAPI to obtain demographic and criminal history information. For the remainder of the interview, respondents interacted with a computer-administered questionnaire using a touch-screen and synchronized audio instructions delivered via headphones. Respondents completed the survey in private at the parole office (or satellite office), with the interviewer in the room but unable to see the computer screen. The interviews were completed in 13 rounds of data collection. The survey was typically scheduled for 2 weeks at each parole office. Approximately 2 to 3 weeks before each round of interviews, sampled respondents were contacted by mail. The respondents were introduced to the survey, provided an approximate data collection date, informed of the nature of the survey, and given a toll-free number to set up an appointment. Two weeks prior to the interview period, NORC appointment setters called respondents to set up interview appointments with those respondents who had not used the toll- free number. Before the interview, respondents were informed verbally and in writing that participation was voluntary and that all information provided would be held in confidence. They were also informed that if they agreed to participate, they would receive $50 as a token of appreciation for participation in the survey. A second automated consent protocol was administered at the beginning of the ACASI portion of the interview to confirm that the respondent had been properly informed that participation in the survey was voluntary and that they were "ready to continue with the interview." Interviews were conducted in either English (99.4%) or Spanish (0.6%). Twenty interviews (0.1%) were lost because the sample parolee could not speak English or Spanish. Sample design The NFPS used a multistage stratified sample design with probabilities of selection proportionate-to-size (PPS). The sampling frame was developed based on the 2006 Census of State Parole Supervising Agencies, conducted by the U.S. Census Bureau on behalf of BJS. The original sampling frame consisted of two subframes: a field office group (FOG) level frame of 800 field office groups and a district level frame of 101 districts. Each of these subframes was sampled separately to produce the sampled offices. A two-stage stratified sample design was used for selection from the office subframe, while a three-stage design was used for the district subframe. Sampling from the field office subframe Prior to selection of offices, the field office sampling frame was modified: * 40 offices were considered ineligible and deleted, including administrative offices not actively supervising parolees; specialty offices, such as those supervising sex offenders only; and offices supervising only individuals committed to treatment or other confinement facilities. * 8 offices in the District of Columbia were deleted, as most parolees had served time in federal facilities; 10 offices in Alaska and 4 offices in Hawaii, except for the largest in each state, were deleted. * 34 offices, with a parolee population under 40 and that had no other in-state office within 75 miles, were deleted. * 27 offices included in the sample frame for the NFPS pretest were excluded. (See page 36 for details on the inclusion of pretest sample data in the final survey.) * Offices with a parolee population under 40 were combined with another office within 75 miles for sampling purposes to form FOGs. Field offices and FOGs were selected from the remaining frame of 800 offices using a PPS sampling procedure: * 34 FOGs were selected with certainty, due to their large size, and allocated a sample size of 260 parolees. * 117 FOGs were selected with non-certainty and allocated a sample size of 130 for offices with populations of 130 or greater, or the actual population sizes for offices with populations of less than 130. * 37 additional FOGs were selected as reserve offices and activated in order of selection as needed to offset office refusals and potential ineligible offices on the frame. A total of 151 FOGs (including 7 involving combined offices) were selected. Data collection was completed in 157 FOGs (after exclusion of 4 ineligible administrative offices and inclusion of 10 reserve sample offices.) All eligible sampled offices agreed to participate. Sampling from the district office sub-frame The 2006 census provided data only at the district or regional office in 19 states. A total of 50 districts were selected from the list of 101 offices using PPS sampling procedures (including 35 certainty and 15 non-certainty districts). An additional 20 reserve districts were selected to offset potential refusals and ineligible listings. Eleven of the districts on the reserve list were activated. All sampled state offices granted approval to participate in the survey; however, one sampled district was excluded due to lateness of the request. Following selection of district offices, field offices and population sizes within each selected district were enumerated. The same grouping and elimination criteria as listed above for parole offices were used for FOGs within each district, except that the distance cutoff was increased to 120 miles. Offices with fewer than 40 parolees were only eliminated if there was no other office with 120 miles for grouping. Three offices were deleted due to the distance and size criteria. Access to one office was denied by the district and was replaced by another randomly selected office within that district. The remaining eligible FOGs within each sampled district were selected using PPS sampling procedures. A total of 81 FOGs were selected (including 21 with combined field offices). Parolee selection After approvals from states and offices were received, a roster was requested of eligible former prisoners under active supervision by the parole office. Only persons who had served time in a state prison and were age 18 or older were eligible. Individuals who had absconded, were reincarcerated or committed, were in a halfway house or community treatment center, had a warrant issued for their arrest, were in violator status, or had been transferred to another parole office were ineligible. Prior to sample selection, each roster was sorted by sex. Males and females were then randomly selected for inclusion in the survey. The target sample size was 260 (196 males and 64 females) in large field offices and 130 (98 males and 32 females) in smaller offices. Female former prisoners were sampled at a higher rate than males to ensure that sufficient numbers were included to provide reliable national estimates. For offices with 130 or fewer eligible parolees, all former prisoners were selected. In total, 34,782 former prisoners were selected. Response rates Of the total number of former prisoners selected for the national survey, 17,738 eligible and available respondents completed the CAPI and ACASI portions of the interview: * 4,318 (12.4%) refused to participate in the survey for reasons such as work, no transportation to the office, unwillingness to discuss prison experiences, wanting to avoid the parole office, dissatisfaction with the system, in trouble, health issues, and other unknown reasons. * 5,664 (16.3%) could not be reached due to absence of working phone numbers, wrong phone numbers, failure to return phone messages, or lack of interest communicated by the parole officer or another person. * 3,338 (9.6%) were found to be ineligible for reasons such as death; deportation; under age 18; not on parole; under warrant or absconder status; transferred to another office; incarcerated at time of sample selection; or under warrant, violator or absconder status after sample selection. * 2,151 (6.2%) were unavailable due to change in status after sample selection, primarily due to incarceration, transfer, or completion of parole supervision. * 1,571 (4.5%) did not participate for other reasons, primarily due to missing scheduled appointments. Overall, 60.6% of eligible and available sampled former prisoners participated in the survey. Inclusion of pretest sample sites and respondents Due to the extensive and burdensome nature of the pretest, and the similarities in the pretest and national survey design and content, data from the pretest were included in the final survey results. PPS sampling methods were employed: * 16 offices from a list of 29 offices responsible for supervision of post-release offenders were selected using PPS sampling methods. * The initial protocol called for randomly sampling 75 cases from each office, with a goal of obtaining 50 completed interviews per office. * A reserve sample of 50 former prisoners per office was drawn for activation in case of high numbers of ineligible cases and non-response. * After interviewing in 8 offices, the initial sample sizes were increased to 100 to address the large numbers of ineligible cases on the office rosters. * Sample sizes varied from 100 to 135 across 15 of the sampled offices. One office, with a population of 67, had a sample size equal to its population. In total, 1,745 former prisoners were selected in the pretest. Interviews were completed by 788 parolees (representing 64.5% of the 1,222 eligible sampled cases). Weighting and non-response adjustments Responses from interviewed former prisoners were weighted to provide national-level estimates. The construction of parolee base weights took into account the probability of selection at each stage, reflecting two stages in the field office subframe and three stages in the district subframe: * Within the district sample, selection probabilities were adjusted at the district and office levels for non-response. * Within the field office sample, all selected FOGs participated. * Probabilities of selection from each subframe took into account the impact of certainty and non-certainty districts and offices. * Within selected FOGs in both subframes, males and females were selected at different rates; consequently, the probabilities of selection were computed separately. * The final parolee base weights in the district sample represent the inverse of the product of the non-response adjusted probabilities from the district and FOG stages times the probability of selection of male and female parolees within field offices. The base weights in the field office subframe represent the inverse of the product of the probability of selection at the FOG and parolee stages. A series of adjustments were applied to the initial base weights to compensate for non-response among sampled parolees. To produce final weights, the initial weights given to sampled parolees who did not respond to the survey were distributed to responding cases. A response propensity model was constructed and used to form adjustment cells within which these weights were reallocated: * Variables used in the non-response adjustments were race, age, years in prison, years since release, and required number of office visits per month. * Predicted probabilities obtained from the model were used to construct five non-response groupings, based on the quintiles of the probability distribution. * Subsequent weight adjustment was done within each group, so that the weights of nonresponding individuals within a range of propensities were allocated to like individuals. Following procedures to address undesirably large weights, final post-stratification was introduced to adjust the weights to known control totals. All states in the sample frame were divided into seven strata, and control totals for each stratum were produced based on the midyear 2008 parolee counts. The final weights in the national survey (including pretest offices) averaged 27.57, resulting in a national estimate of approximately 510,800. Standard errors and confidence intervals As with any survey estimates, the NFPS estimates are subject to error arising from the fact that they are based on a sample rather than a complete enumeration. A common way to express this sampling error is to construct a 95%-confidence interval around each survey estimate. Typically, multiplying the standard error by 1.96 and then adding or subtracting the result from the estimate produces the confidence interval. This interval expresses the range of values that could result among 95% of the different samples that could be drawn. To facilitate the analysis, rather than provide the detailed estimates for every standard error, differences in the estimates of sexual victimization for subgroups in these tables have been tested and notated for significance at the 95%-level confidence. For example, the difference in the rate of inmate-on-inmate sexual victimization among female prisoners (13.7%), compared to male prisoners (4.2%), is statistically significant at the 95% level of confidence. (See table 6 on page 15.) In all tables providing detailed comparisons, statistically significant differences at the 95% level of confidence or greater have been designated with two asterisks (**). Measurement of sexual victimization The survey of sexual victimization relied on former inmates reporting their direct experience, rather than reporting on the experience of other inmates. Questions related to inmate- on-inmate sexual activity were asked separately from questions related to staff sexual misconduct. The ACASI survey began with a series of questions that screened for general unwanted sexual activities with other inmates. To fully measure all sexual activities, a question related to touching of specific body parts in a sexual way was followed by a general question regarding any sexual activity with another inmate in which the respondent did not want to participate. (For specific screening questions, see appendix A.) Respondents who answered "yes" to either of the sex-specific screener questions were then asked questions related to the touching of body parts in a sexual way and questions related to acts of oral, anal, and vaginal sex. (For survey items related to specific inmate-on-inmate victimization categories, see appendix B.) The nature of coercion (including physical force, pressure, and other forms of coercion) was measured for each type of reported sexual activity. ACASI survey items related to staff sexual misconduct were asked somewhat differently. Respondents were first asked if during their confinement staff had sexually harassed them, watched them at an inappropriate time while dressing or taking a shower, or forced them to undress or touched their private parts in an inappropriate way. (See survey items in appendix C.) Regardless of the responses to the initial questions, all respondents were then asked two screener questions to determine if they ever "willingly" or unwillingly had sex or sexual contact with any facility staff. Respondents who answered "yes" to either of the screener questions were then asked questions related to "willing" and unwilling sex or sexual contacts, the nature of coercion, and the specific types of sexual contacts including oral, anal, and vaginal sex. (For survey items related to specific staff-on-inmate victimization categories, see appendix D.) Based on their responses to the survey items, victims were classified in each of four general variables: (1) inmate-on- inmate nonconsensual acts, (2) inmate-on-inmate abusive sexual contacts only, (3) staff sexual misconduct unwilling; and (4) staff sexual misconduct willing. The entire ACASI questionnaire (listed as the National Former Prisoner Survey) is available on the BJS website at http://www.bjs.gov. Prison placements In order to assess the impact of facility factors on the likelihood of a former prisoner reporting inmate-on-inmate sexual victimization and staff sexual misconduct, a separate record for each prison placement was constructed. Overall, the 18,526 former prisoners in the survey reported entering separate prison facilities 45,318 times during the period of incarceration prior to their release from custody. Depending on the number of state prisons entered, each respondent is represented between 0 and 15 times. Facility-level data from BJS prison censuses conducted in 1990, 1995, 2000, and 2005 were merged to each placement record, based on the census closest to the period of incarceration. Individual-level victimization variables and other individual- level characteristics reported in the survey were included with each placement. Of all 45,318 prison placements, 96% were matched to an appropriate census data file. When weighted to reflect initial probabilities of selection and non-response adjustments, data were available for 1,254,998 of the prison placements (approximately 94% of all placements). Local jail placements (prior to prison, as an interim stop, or as the only placement) were excluded. Placements in hospitals and out-of-state facilities were also excluded. Linkage of placements to prison census data Prison facilities were identified from an initial list provided by the 2000 Census of State and Federal Correctional Facilities. Information on these facilities was enhanced by the 1990 and 1995 censuses and contacts in departments of correction in each state. The final list was programmed to appear in the CAPI portion of the survey prior to the start of ACASI. If the respondent provided a facility name that could not be identified on the list after attempts to locate the facility by name of city, the interviewer added the facility to the list. Initial rules were used to define the most appropriate census file for linking the facility information. Each inmate provided admission and release dates for the most recent period of incarceration, not for individual facilities. The following rules were applied to determine the reference year for linkage, either the year of admission or the average year (i.e., the midpoint between the admission and release date): * If the victim was placed in only one facility, the average year was used. * If the victim was in multiple facilities, the admission year was used for the first facility placement and the average year for all other placements. If the reference year was-- * 1992 or earlier, facility characteristics were drawn from the 1990 census * 1993-1997, facility characteristics were drawn from the 1995 census * 1998-2002, facility characteristics were drawn from the 2000 census * 2003 or later, facility characteristics were drawn from the 2005 census. All unmatched facilities from the interviewer entry list were identified and linked to an appropriate census file. Logistic regression models Multivariate logistic regression estimation is a modeling technique used to determine what characteristics are statistically significant for predicting a dichotomous outcome (e.g., victimized or not victimized) while controlling for all the other characteristics in the model. NFPS used this technique to determine what inmate-level characteristics were significant predictors of inmate-on- inmate sexual victimization and staff sexual misconduct over the entire period of confinement. For each type of victimization, and for males and females separately, selected parolee characteristics were examined, including demographic characteristics, sexual orientation and history, and criminal justice status and history. For each type of victimization, a logistic model was iteratively run under a backwards selection technique until only predictors that were significant at the 95% level of confidence remained. (See table 10 on page 19.) NFPS also used this technique to determine what facility-level and inmate-level characteristics were significant predictors of inmate-on-inmate sexual victimization and staff sexual misconduct in each of the state prisons where the former inmates had served time. Based on a separate record for each prison placement for the 18,526 former prison inmates interviewed, logistic regression models were developed to model the effects of individual and prison-level characteristics. For each type of victimization, a logistic model was developed to isolate specific individual and facility factors that remain significant at the 95% level of confidence. (See table 16 on page 28.) Conditional marginal probabilities In each of the logistic regression models, the conditional marginal probability represents the probability that an inmate with a particular characteristic has experienced sexual victimization (by type) conditional on the inmate having the mean value for all other predictors in the model. For example, based on a combination of individual characteristics only, a white male former prisoner had a 4.0% chance of being victimized by another inmate given that he was at the mean of the joint distribution of age, sexual orientation, offense, and total time served. For purposes of estimating the conditional marginal probabilities, the overall log odds of incurring the event for a given level of race/Hispanic origin was calculated from the estimated linear model by specifying the value of the race/Hispanic origin variable as the level of interest and then specifying all other variables in the model to be the weighted mean in the population. For example, the simple logistic model for the binary variable Y (experienced sexual victimization or not) is where, ài is the parameter of the race/Hispanic origin category for the individual i ( i=1 n, n is the total number of respondents). To calculate the log odds for a specific race/Hispanic origin = r; use fixed àr and x, where, . For other categorical covariates x except race/Hispanic origin, the mean value for each dummy variable is the weighted percentage of the corresponding level. Based on the obtained log odds, the conditional marginal probability was then calculated for the specific level of the race/Hispanic = r as These estimates (Mr) represent the expected risk of victimization for a former inmate, conditional on the inmate belonging to a particular group (defined by each characteristic in the model) and having the mean value on all of the other characteristics in the model. The conditional marginal assumes that x is constant, and the variance is due to the variances of . Significance tests For each of the regression models, variances for the estimates took into account the NFPS sample design. These variances were computed with weighted data in SUDAAN using the Taylor Linearization method.Wald F-statistics were calculated to test for statistical significance of the effects of each individual-level and facility-level characteristic (appendix tables 3 and 4). The Wald F-statistics were used to test the null hypothesis that all regression coefficients are equal to zero for each characteristic (i.e., the probability of experiencing victimization is the same across all categories of the selected characteristic), conditional on all other inmate or facility-level characteristics being included in the model. NFPS and NIS sample differences Findings in the NFPS are limited to persons released from state prison to a period of post-custody supervision (including parole, probation following incarceration, community supervised release, or other form of supervision). Approximately 24% of all sentenced state prisoners released in 2008 were released unconditionally. These former inmates, some of whom may have experienced sexual victimization while in prison, were not surveyed. Because these former inmates were not subject to post-custody supervision, and correctional authorities lacked their current addresses or phone numbers, they could not be easily contacted and interviewed after their release from prison. The NFPS findings should not be generalized to all former prisoners. Based on selected demographic characteristics reported in BJS's National Corrections Reporting Program in 2008, inmates released conditionally were similar to those released unconditionally. Nearly the same percentage were male (89.2% compared to 89.4%), under age 30 (34.9% compared to 35.2%), age 50 or older (10.1% compared to 10.2%), and sentenced for a violent offense (25.4% compared to 26.6%). However, inmates released unconditionally had served significantly longer in prison than those released conditionally. An estimated 23.2% of inmates released unconditionally had served more than 3 years, compared to 15.9% of inmates released conditionally. In addition, 12.5% of unconditional releases had served fewer than 6 months, compared to 32.5% of conditional releases. Caution is also recommended when comparing NFPS findings with NIS-1 and NIS-2 findings, due to differences in coverage and differences in composition among former and current inmates. In contrast with the NIS sample, the NFPS sample does not include inmates who may never be released. (Based on data from the 1997 Survey of Inmates in State Prison, approximately 5% of all inmates in state prison were not expected to ever be released.) In addition, the cross-section of former inmates is comprised of less serious offenders with shorter maximum sentences than the cross-section of inmates in the NIS sample. For example, 33% of former inmates in NFPS had served time for a violent offense, compared to 56% of state inmates in NIS-2 who were serving time for a violent offense. ************* Appendix A ************* Survey items that screen for inmate-on-inmate sexual victim ization, National Former Prisoner Survey, 2008 C11. [MALES] During your confinement, did another inmate touch your butt, thighs, penis, or other private parts in a sexual way or did you touch theirs, when you did not want this to happen? C11a. [FEMALES] During your confinement, did another inmate touch your butt, thighs, breasts, vagina, or other private parts in a sexual way or did you touch theirs, when you did not want this to happen? [IF C11 NE Yes or C11a NE YES] We recognize that you may find questions regarding sexual activity sensitive. However, no one will know how you answered these questions. Please remember that your responses are confidential. C11b. During your confinement, were you involved in any sexual activity with another inmate in which you did not want to participate? ************* Appendix B ************* Survey items related to inmate-on-inmate sexual victimization categories, National Former Prisoner Survey, 2008 Males Nonconsensual sexual activity oral sex C12. Did the sexual activity ever involve oral sex or a blow job in which your mouth was on another inmate's penis or other private parts? You indicated that you were involved in unwanted sexual inci dents, but that oral and anal sex did not occur. During the unwanted sexual incident or incidents, which of the following things happened? Please indicate all that happened. C15f. Another inmate put their mouth on your penis or other private parts. C15g. You put your mouth on another inmate's butt or thighs. Females C12d. Did the sexual activity ever involve oral sex in which your mouth was on another inmate's vagina or other private parts? You indicated that you were involved in unwanted sexual inci dents but [if applicable] oral, [if applicable] vaginal, or [if applicable] anal sex did not occur. During the unwanted sexual incident or incidents, which of the following happened? Please indicate all that happened. C16g. Another inmate put their mouth on your vagina or other private parts. C16h. You put your mouth on another inmate's butt or thighs. Nonconsensual sexual activity anal sex Males C14. Did the sexual activity ever involve anal sex in which another inmate inserted their finger, penis, or an object into your butt? You indicated that you were involved in unwanted sexual inci dents, but that oral and anal sex did not occur. During the unwanted sexual incident or incidents, which of the following things happened? Please indicate all that happened. C15h. You inserted your finger, penis, or an object into another inmate's butt. Females C14a. Did the sexual activity involve anal sex in which an other inmate inserted their finger or an object into your butt? You indicated that you were involved in unwanted sexual inci dents but [if applicable] oral, [if applicable] vaginal, or [if applicable] anal sex did not occur. During the unwanted sexual incident or incidents, which of the following happened? Please indicate all that happened. C16j. You inserted your finger or an object into another inmate's butt. Nonconsensual sexual activity vaginal sex Female C13. Did the sexual activity ever involve vaginal sex in which another inmate inserted their finger or an object into your vagina? You indicated that you were involved in unwanted sexual inci dents but [if applicable] oral, [if applicable] vaginal, or [if applicable] anal sex did not occur. During the unwanted sexual incident or incidents, which of the following happened? Please indicate all that happened. C16i. You inserted your finger or an object into another inmate's vagina. Males Nonconsensual sexual activity manual/hand job You indicated that you were involved in unwanted sexual inci dents, but that oral and anal sex did not occur. During the unwanted sexual incident or incidents, which of the following things happened? Please indicate all that happened. C15d. Another inmate rubbed your penis in an attempt to arouse you. C15e. You rubbed another inmate's penis with your hand. C15i. You were forced to masturbate while someone watched. Females You indicated that you were involved in unwanted sexual inci dents but [if applicable] oral, [if applicable] vaginal, or [if applicable] anal sex did not occur. During the unwanted sexual incident or incidents, which of the following happened? Please indicate all that happened. C16e. Another inmate rubbed your breasts, butt, or vagina in an attempt to arouse you. C16f. You rubbed another inmate's breasts or vagina with your hand. Abusive sexual contact only Males You indicated that you were involved in unwanted sexual inci dents, but that oral and anal sex did not occur. During the unwanted sexual incident or incidents, which of the following things happened? Please indicate all that happened. C15a. Another inmate touched your butt or thighs. C15b. Another inmate touched your penis or scrotum. C15c. Another inmate violently grabbed or touched your penis or scrotum in an attempt to hurt you. Females You indicated that you were involved in unwanted sexual inci dents but [if applicable] oral, [if applicable] vaginal, or [if applicable] anal sex did not occur. During the unwanted sexual incident or incidents, which of the following happened? Please indicate all that happened. C16a. Another inmate touched your butt or thighs. C16b. Another inmate touched your breasts. C16c. Another inmate touched your vagina. C16d. Another inmate violently grabbed or touched your breasts or vagina in an attempt to hurt you. ************* Appendix C ************* Survey items related to staff sexual misconduct, National Former Prisoner Survey, 2008 E1. During your confinement, did staff hassle or harass you in a sexual way? E2. Did staff ever stare at you or watch you at inappropriate times (for example, while you were dressing or taking a shower)? E3. [FEMALES] Did a staff member ever force you to undress in their presence or hit or brush against your breast or other private parts when you did not think it was an accident or it was not required by their job? E3b. [MALES] Did a staff member ever force you to undress in their presence or hit or brush against your private parts when you did not think it was an accident or it was not required by their job? E4. Did you ever willingly or unwillingly have sex or sexual contact with any facility staff person? [IF E4 NE Yes] Please remember that your responses are confi dential. Neither you nor the staff person involved will be identified through this survey. No one will know you answered these questions. E4a. During your confinement did you have sex or any sexual contact with a staff person, whether it was willing or not? ************* Appendix D ************* Survey items related to staff-on-inmate sexual victimization categories, National Former Prisoner Survey, 2008 Males Nonconsensual sexual activity oral sex E15. Did the sexual contact involve oral sex or a blow job in which your mouth was on the male staff person's penis or other private parts or his mouth was on your private parts? E15a. Did the sexual contact involve oral sex or a blow job in which the female staff person's mouth was on your penis or other private parts or in which your mouth was on her private parts? Females E15b. Did the sexual contact involve oral sex in which your mouth was on the male staff person's penis or other private parts or his mouth was on your private parts? E15c. Did the sexual contact involve oral sex in which your mouth was on the female staff person's breasts, vagina, or other private parts or her mouth was on your private parts? Nonconsensual sexual activity anal sex Males E16. Did it involve anal sex in which the male staff person inserted his finger, penis, or an object into your butt or you inserted your finger, penis, or object into his? E16a. Did it involve anal sex in which the female staff person inserted her finger or an object into your butt or you in serted your finger, penis, or an object into hers? Females E16b. Did it involve anal sex in which the male staff person inserted his finger, penis, or an object into your butt or you inserted your finger or an object into his? E16c. Did it involve anal sex in which the female staff person inserted her finger or an object into your butt or you in serted your finger or an object into hers? Nonconsensual sexual activity vaginal sex Females E13. Did you have vaginal sex with a female staff person in which your penis, finger, or an object was inserted into her vagina? E13a. Were you involved in vaginal sex in which a male staff person inserted his penis, finger, or an object into your vagina? E13b. Were you involved in vaginal sex with a female staff person in which she inserted her finger or an object into your vagina or you inserted your finger or an object into hers? Nonconsensual sexual activity manual/hand job Males E14. Did you give or receive a hand job in which a male staff person rubbed your penis with his hand or you rubbed his? E14a. Did you receive a hand job in which a female staff person rubbed your penis with her hand? Females E14b. Did you give the male staff person a hand job in which you rubbed his penis with your hand? Abusive sexual contact only Male E12. Did a male staff person touch your penis or other private parts in a sexual way or did you touch his? E12a. Did a female staff person touch your penis or other private parts in a sexual way or did you touch her private parts? Females E12b. Did a male staff person touch your breasts, vagina, or other private parts in a sexual way or did you touch his private parts? E12c. Did a female staff person touch your breasts, vagina, or other private parts or did you touch hers? ************* Appendix E ************* Survey items that identify prison placements and measure sexual victimization in each facility, National Former Pris oner Survey, 2008 B5. What was the name of the facility in which you were placed in [Month/Year]? B6. Were you placed in any other facilities between [date of admission] and your release in [Month/Year]? B5. [REPEAT, up to 25 times] What was the name of another facility in which you were placed? For each type of inmate-on-inmate sexual victimization FBM12a. [MALES] You indicated you had unwanted [oral sex or gave a blow job because you were physically forced, afraid, or someone threatened to hurt you physically]. Where did this unwanted [oral sex or blow job occur]? [READ LIST OF FACILI TIES] FBM12b. [FEMALES] You indicated you had unwanted [oral sex because you were physically forced, afraid, or someone threat ened to hurt you physically]. Where did this unwanted [oral sex] occur? [READ LIST OF FACILITIES] For all forms of staff-on-inmate sexual victimization ------------------------------------------------------ E6. At which facility or facilities did you have sexual contact with staff? [READ LIST OF FACILITIES] *************************************************** Office of Justice Programs Innovation -- Partnerships -- Safer Neighborhoods www.ojp.usdoj.gov **************************************************** ********************* 5/7/2012 JER/ 2:35 ******************