Us Jfa Sexual Violence Texas Prison 2006

Us Jfa Sexual Violence Texas Prison 2006

• Topics: Sexual Assault

Download original document:
Brief thumbnail

This text is machine-read, and may contain errors. Check the original document to verify accuracy.

The author(s) shown below used Federal funds provided by the U.S.
Department of Justice and prepared the following final report:

Document Title:

Sexual Violence in the Texas Prison System

Author(s):

James Austin ; Tony Fabelo ; Angela Gunter ;
Ken McGinnis

Document No.:

215774

Date Received:

September 2006

Award Number:

2004-RP-BX-0003

This report has not been published by the U.S. Department of Justice.
To provide better customer service, NCJRS has made this Federallyfunded grant final report available electronically in addition to
traditional paper copies.

Opinions or points of view expressed are those
of the author(s) and do not necessarily reflect
the official position or policies of the U.S.
Department of Justice.

This document is a research report submitted to the U.S. Department of Justice. This report has not
been published by the Department. Opinions or points of view expressed are those of the author(s)
and do not necessarily reflect the official position or policies of the U.S. Department of Justice.

The JFA Institute
Washington, D.C./Austin, Texas
Conducting Justice and Corrections Research for Effective Policy Making

Sexual Violence
In The
Texas Prison System

James Austin
Tony Fabelo
Angela Gunter
Ken McGinnis

March 2006
This project was supported by Grant No. 2004-RP-BX-0003 awarded by the National Institute of
Justice, Office of Justice Programs, US Department of Justice. Points of view in this document
are those of the author and do not necessarily represent the official position or policies of the US
Department of Justice.

5 Walter Houp Court, NE Washington, D.C. 20002

Ph. 202-544-4434

www.JFA-Associates.com

The JFA Institute

Acknowledgements
We want to thank a number of persons at the Texas Department of Criminal
Justice who help make this study possible.
First we would like to thank Garry Johnson the former Executive Director of the
Texas Department of Criminal Justice who supported the original concept and
project proposal. His successor Brad Livingston has continued to provide this
same level of support. Ed Owens, Deputy Executive Director and Doug Dretke,
Director, Correctional Institutions Division also provided unlimited access to the
TDCJ facilities, staff and data.
Dimitria D. Pope, Director, RED Group was especially helpful in providing data
and interpretation of the results contained in this report.
We also want to acknowledge the assistance of Monty Hudspeth, Safe Prisons
Program Manager, Kennan Zhiska, Data Management Director, RED Group,
and Karen Hall, Program Specialist, Executive Services who provide important
data over the course of the study.

The JFA Institute

Executive Summary
The Texas Department of Criminal Justice (TDCJ) operates the nation’s
third largest prison system (along with California and the Federal Bureau of
Prisons) with over 150,000 prisoners. Each year approximately 500-600 prisoner
on prisoner sexual assaults are reported by prisoners and staff to the TDCJ. For
each reported assault, detailed information is collected and stored on a specially
created database that was developed as part of the agency’s effort to report,
evaluate and reduce prison sexual assaults.
The study was designed to provide comprehensive assessment of nearly
2,000 officially reported sexual assaults that occurred in the Texas prison
systems between 2002 and 2005. The research design attempts to describe the
attributes of these incidents in terms of victim and perpetrator attributes and the
time and location of such events. We also examined the influence of the “prison
environment” on the prevalence and nature of prison violence, including sexual
assault.
This study had several objectives. First, we wanted to better understand
the number and nature of sexual assault allegations being made in one of the
nation’s largest prison systems. Second, we wanted to understand how the
Texas Safe Prison program operates and its impact on sexual assault within its
numerous prisons. Finally, we wanted to see what lessons could be learned that
would have a positive impact on reducing the number of sexual assaults for both
Texas and other correctional systems.
Overall Rates of Sexual Assaults
1. Texas has the highest reported number of alleged incidents at 550 for a
rate per 1,000 prisoner population of 3.95, almost four times the national
average for the states of 1.05. It also has one of the lowest substantiation
rates (less than three percent).
2. The official alleged sexual assault rate since 1993 hovered between 1.2
and 0.6 per 1,000 inmate population until 1999 when the rate doubled.
Shortly after the passage of the Texas Safe Prison Program, the rate
increased again which is not surprising given that a major objective of the
act was to increase the detection and reporting of such incidents.
3. Specifically, there has an increase in the reporting of “abusive sexual
contacts” for each calendar year since 2001. This increase suggests that
the prisoner awareness effort may be having an impact on a prisoner’s
willingness to report an act of sexual abuse/assault in prison.
4. A significant number of the allegations made by prisoners are not reported
in a timely manner (30 % within the same day) and have no independent

iii

The JFA Institute

witnesses to the allegation. These delay and witness problems are major
problems in attempting to sustain an allegation.
5. Approximately two-thirds of all reported cases of sexual assault allegedly
occurred in an inmate cell, followed by assaults in the shower or bathroom
area, assaults in dorms, and finally assaults occurring in common areas
such as dayrooms, work places, cafeterias, etc.
6. The locations for sustained assaults is more pronounced, with threequarters of confirmed cases occurring in cells, and nearly 12% of
confirmed cases occurring in the shower or bathroom facilities.
7. Just over 50% of the sustained cases included forensic evidence from a
rape kit or a forensic exam. On the other hand, rape kits and forensic
exams were performed in only 20% of all alleged sexual assaults. The
primary reason for not completing the exams was “time lapse”.
Attributes of Victims and Assailants
1. An extremely small percentage of the daily population is classified as a
“victim” or “assailant” (about 2 percent). Such a small “base rate” means
that it is virtually impossible to develop a statistical profile of potential
victims and assailants.
2. White inmates are attacked more frequently than any other race (60%)
while two thirds (68%) of the sustained incidents involved black assailants
followed by 19% being Hispanic assailants and 12% white assailants.
3. In general the victims are younger than the assailants. The average age of
victims in sustained cases is 3 years younger than the assailants, while
the average age of victims in probable cause cases is 4 years younger
than assailants.
4. Although only 12% of the allegations involved a mentally ill or intellectually
impaired prisoner, this is eight times the proportion in the general prisoner
population (1.6%).
5. A disproportionate number of alleged sexual assaults occurring within the
safe-keeping and mentally ill/impaired prisoners (ten times the rate in the
general population). It would appear that a number of these prisoners may
become involved in aggressive sexual behavior after being placed in safekeeping.
6. For males in particular, assailants are more likely to have violent offenses
while victims are more likely to have a conviction for a sex assault offense
or a non-assaultive sex offense.

The JFA Institute

7. For females, victims and assailants in sex assault allegations are more
likely to have violent criminal histories.
8. While there may be some violence among gangs in sex assault
allegations, the vast majority of victims (nearly 97%) are not associated
with any gang. Gang affiliation is highest for assailants whose alleged
sexual misconduct was sustained by TDCJ’s internal review, with 20.5%
of all sustained assailants affiliated with a gang.
9. On average, both victims and assailants tend to have served longer in
TDCJ than all other TDCJ inmates. This may be related to the fact that
TDCJ also houses state jail prisoners, operates SAFPF facilities, and in
general deals with a much wider group of prisoners than those implicated
in sex assault allegations.
Unit Attributes
1. While there is variation in the type of institutions where incidents were
sustained, the majority of cases (86%) were sustained at large, male
prisons operated by TDCJ.
2. Units housing the longer-term prisoners with higher custody levels have
higher numbers and rates of allegations than other facility types.
3. Units housing special needs populations (psychiatric and mentally
retarded) face particular challenges in managing their inmates, with higher
rates of general inmate-on-inmate violence and higher rates of use of
force, in relation to the population size of these facilities.
Qualitative Analysis of Selected Units
1. Staff at all levels was found to have a clear understanding of the
organization’s expectations and goals as it related to Safe Prisons
Program requirements.
2. There did not appear to be any indication of “organizational indifference”
at any of the facilities visited. There may be individuals within the
organization that may ‘look the other way’ when faced with an allegation of
sexual assault, but that was not a prevalent attitude observed within the
facilities visited.
3. The number of allegations at the low rate facilities (Holiday and Murray)
can be attributed to not only to the operational mission of the facilities but
also to the fact that the majority of beds in both of the facilities are located
in dormitory settings which make sexual assaults more difficult to occur

The JFA Institute

undetected by staff. Ironically, cell blocks seem to contribute to sexual
assaults.
4. Solid cell fronts, while permitting privacy for the inmates and reducing
noise within the unit, also provides the degree of privacy that permits
sexual assaults to occur. Unlike older prison designs where the cell fronts
consisted of bars, the solid doors limit visual observation by staff and to a
degree sound proof the cells to the point where staff have difficulty hearing
what is going on in individual cells.
5. The fact that the majority of the sexual assaults occur between 6 a.m. and
6 p.m. can also be understood when observing the level of activity in the
housing units in TDCJ facilities. Even though during these hours the
number of staff assigned to the units is increased, the level of activity that
they are responsible for also increases dramatically. The responsibility of
housing unit staff draws them away from the routine of checking on cell
activity on a routine basis. This provides for periods of time when only the
housing unit picket officer (Control Room Officer) is monitoring the day
rooms and cell fronts. This issue should be reviewed by the TDCJ from a
staff deployment and training standpoint.
6. The low level of allegations occurring at the mental health facilities could
be attributed to staff deployment levels. As would be expected of special
needs treatment facilities the staff to inmate ratios are very high in
comparison to other TDCJ facilities. This results in almost constant
supervision of all the prisoners of the two facilities both during in-cell and
out-of-cell activity periods.
7. The openness of the cell clock design at facilities similar to Darrington
may in fact result in discouraging the reporting of assaults. The lack of
privacy may discourage incidents of sexual assault, but may also deter the
reporting of legitimate allegations of assault.
Discussion and Recommendations
Much of what is reported above matches prior studies. The number of
official allegations of prisoner on prisoner sexual assault in Texas is relatively low
although it is noted that its allegation rate is higher than most states. However,
the higher allegation rate is due, in part, to the recently implemented Safe
Prisons program which is designed to encourage by staff and prisoners the
reporting of alleged sexual assaults.
Like most states the substantiation rate of the allegations is quite low. In
Texas one of the major reasons why the substantiation rate is so low is the delay
in having an allegation reported by the prisoner. As with any criminal
investigation, any undue delay in reporting a criminal act reduces the ability of

The JFA Institute

the investigators to gather physical evidence and interview persons who would
serve as potential witnesses.
The low rates of allegations and substantiated allegations makes it very
difficult if not impossible to develop a statistical profile or risk instrument that
would have strong predictive capabilities. However, there a number of attributes
that distinguish the victims and perpetrators from other prisoners.
The alleged victims are generally younger than their assailants. There is a
strong racial relationship in such incidents where victims are more likely white
while their assailants are more likely black or Hispanic and be gang affiliated.
The assailants are also older, more likely to have lived in urban areas, have been
convicted of a violent crime, are in a higher custody level and have served more
time in prison than their victims.
Not surprisingly the higher security facilities as well as units housing
special needs populations (psychiatric and mentally retarded) face particular
challenges in managing their inmates, with higher rates of general inmate-oninmate violence and higher rates of use of force, in relation to the population size
of these facilities. Incidents are more likely to occur in the day time in housing
units where officers can be easily distracted or have poor line of sight to the
celled housing units.
These profile data on the victims, assailants and facilities where such
allegations are made suggest that one needs to see the crime of prison rape in a
larger context. Specifically, it would appear that prison rape (or the threat of rape)
is just one weapon predatory prisoners will use to impose their influence and
control over other prisoners. The prisoners and prisoners that have the highest
rates of sexual assault also have the higher rates of other serious misconduct
and criminal activity. So it would suggest that in order to reduce prison rape
other forms of serious misconduct must also be addressed.
Finally, the research suggests that the Texas Safe Prison program has
produced a much greater awareness on the part of staff and prisoners on the
issue of prison rape. One of the immediate effects has been a much higher rate
of reporting and a more structured response and investigation to such
allegations. And the data base that has been created allows the TDCJ to monitor
the number, attributes and final dispositions of these incidents. Many states
could benefit from adopting many of the key components of the Texas program.
There are some suggestions that we would suggest to Texas and other
correctional systems that if implemented might serve to reduce sexual assaults.
First, further efforts are needed to provide structured opportunities for prisoners
to report any allegation of sexual assault –either staff on prisoner or prisoner on
prisoner. One specific suggestion is to require case managers to always inquire

vii

The JFA Institute

of the prisoner during a formal session or contact whether any assaults have
occurred or if they are being threatened by other prisoners.
Second, there is a concern on what happens to prisoners who make an
allegation but the case cannot be substantiated. There are a large number of
these cases in Texas. Does staff make any effort to more closely monitor, reassign or supervise the prisoners who have been implicated in the incident? It is
not clear that any such efforts are being made by prison officials in such cases.
Third, the TDCJ should examine the reasons for the large number of
cases in which either the victim or the alleged assaulter was transferred without
any disciplinary or legal action. These are cases which have also been
determined to be unsubstantiated. The questions centers on why do such a
large number of prisoners involved in these alleged incidents end up getting
transferred without any verification of the incident. Are these transfers based on
legitimate concerns or a means of foregoing either discipline or prosecution?
Fourth the categories for designating prisoners in the TDCJ data base
should be expanded so that it has the following four designations:
a.
b.
c.
d.

Potential Victim
Known Victim
Potential Assailant
Known Assailant

The low rate of officially reported sexual assault on prisoners means that it
is not practical nor recommended that a traditional risk scoring system be
attempted. To do so would produce an unacceptable level of “false positives” in
terms of identifying both potential victims and assailants. This not to say that the
factors and attributes that are associated with assailants and victims as
presented in this report cannot be applied in some manner.
Rather, it is recommended that a criteria for applying the “potential” victim
or assailant label should take the form of a check list that takes into account the
factors found to be associated with such allegations in this study. An example of
such a check list is provided in the main report.
Finally, this research did not adequately address the issue of sexual
assault among female prisoners. We found at the women prisons that the
attitudes expressed by the staff suggested that they believe sexual activity was
more common then at male facilities but that such behavior was largely
consensual. We are not persuaded that this is indeed the case. Clearly a
separate and more detailed assessment of sexual assault among female
prisoners is needed.

viii

The JFA Institute

I. Introduction
A.

Background
“The number one fear of those going to prison: being raped…
Prisons are single-sex institutions, but sex roles in prison are a bit
more complicated. In any given facility, there are heterosexuals,
homosexuals, transsexuals, and other sexual variations. Without
doubt there is sex in prison – both consensual and coerced” 1

In September 2003, President George Bush signed the Prison Rape
Elimination Act of 2003 (from here on referred as PREA or Act). This bill was
passed by Congress in response to a growing concern that a significant number
of the over 2 million persons incarcerated in the nation’s state and federal
prisons, local jails and juvenile facilities are being sexually assaulted each year.
Moreover, there is concern that many of these assaults are not being reported to
correctional officials and/or those correctional agencies are not taking sufficient
steps to prevent such crimes from occurring in the future including, using
vigorous prosecution as deterrence.
The Act establishes three programs in the Department of Justice – one
dedicated at collecting national prison rape statistics and conducting research;
one dedicated to the dissemination of information and procedures for combating
prison rape; and a grant program to assist in funding State programs. The
Bureau of Justice Statistics (BJS), the National Science Foundation (NSF) and
the National Institute of Justice (NIJ) are charged with conducting the research.
BJS is now developing a methodology for measuring the number and type of
rapes occurring in the nation’s prison system using both officially reported and
self-reported survey data. The BJS data will be included in an annual report of
the prevalence of prison rape by each state that participates in the data collection
effort. An annual public review of the performance of those prison systems
where the incidence of prison rape greatly exceeds the national average will be
conducted following procedures set by the new law. If a State’s prison system
does not fall into this category, or if it does and that State’s prison officials
participate in the public hearings on their performance, the State will receive
increased funds from certain federal grant programs.
A Prevention and Prosecution Program was created to serve as a
clearinghouse for the provision of information and assistance to those authorities
responsible for the prevention, investigation, and punishment of prison rape. This
program will also provide training and assistance to federal, state and local
prison officials on the prevention, investigation and punishment of prison rape.
The Grant Program will make annual grants (up to $40 million each year) to state
and local programs that enhance the prevention and punishment of prison rape.
1

Ross, Jeffrey Ian and Stephen C. Richards. 2002. Behind Bars: Surviving Prison. Alpha Books:
Indianapolis, IN.

-1-

The JFA Institute

Finally, the new law creates a National Prison Rape Reduction Commission
which will conduct comprehensive hearings and examine all penal, economic,
physical, mental, medical and social issues relating to prison rape in America. At
the conclusion of its review the Commission will issue a comprehensive report on
the subject, including a recommended set of national standards to reduce and
eliminate prison rape. The commission’s recommended standards will address
practices for the investigation and elimination of prison rape including the training
of correctional officers; sexually transmitted disease prevention; identifying,
protecting, screening, isolating, and punishing vulnerable and potentially
offending inmates; and other related issues.
The Commission’s recommended national standards will be independently
reviewed by the Attorney General, who may modify them, and will, thereafter, be
published for notice and comment under the Administrative Procedure Act. Once
the standards proposed by the Attorney General become a final rule, they will
become immediately applicable to the Federal Bureau of Prisons. States may
adopt the standards by statute, and those that do will receive increased funds for
two years from certain federal grant programs. Clearly, the Act could have a
major impact on state and local correctional practices. As noted above, the NIJ is
expected to assist the development of new correctional standards by funding a
number of studies that address the following themes or topics:
1. Identification and evaluation of sexual assault prevention programs;
2. Research on the impact of victimization and its treatment;
3. Research on risk assessment and management of vulnerable
populations; and,
4. Research on HIV/AIDS or other sexually transmitted diseases and
sexual violence.
Many of the public policy and research issues are reflected in the wellpublicized case of Roderick Johnson, a former prisoner in the Texas prison
system, whose allegations have resulted in major litigation in Texas, the passage
of Safe Prison Act in Texas, and have drawn national attention to the problem of
prison rape. 2 Mr. Johnson has claimed that while incarcerated, he became a sex
slave where he was repeatedly raped by a number of prisoners at a particular
Texas prison. Despite what he states were repeated requests to prison officials,
no actions were taken either to transfer him to another prison or to house him in
a location that would protect him from such assaults.
Mr. Johnson’s case highlights the major policy issues surrounding the
phenomenon of sexual assault within correctional facilities that require further
research. Specifically, what factors (individual and environmental) contribute to
the incidence of sexual assaults, were there actions that the agency could have
taken to reduce the level of risk to sexual assault, and what are the costs to
2

“Back from the Brink: Former prison sex salve is striving for new life in Austin”, Austin AmericanStatesman, Sunday, January 11, 2004, pp A1 and A11.

-2-

The JFA Institute

victims in terms of pain, suffering, trauma and public health? Finally, there is a
need to determine what pre-emptive actions can be taken by correctional
agencies to better prevent such crimes from occurring in the future through better
risk assessments, improved reporting standards, staff training, and more
effective classification, housing and staff supervision techniques.
The following report summarizes a study of officially reported sexual
assaults and rapes reported in the Texas prison system conducted by The JFA
Institute, in cooperation with the Texas Department of Criminal Justice (TDCJ). In
addition to a comprehensive assessment of the attributes of the victims and
perpetrators of such crimes an attempt was also made to evaluate the role of the
prison environment in such events.
B.

Study Goals and Objectives

The Texas Department of Criminal Justice (TDCJ) operates the nation’s
third largest prison system (trailing only the California and the Federal Bureau of
Prisons) with over 150,000 prisoners. 3 Each year approximately 500-600
prisoner on prisoner sexual assaults are reported by prisoners and staff to the
TDCJ. For each reported assault, detailed information is collected and stored on
a specially created database that was developed as part of the agency’s effort to
report, evaluate and reduce prison sexual assaults. Among the items captured in
this database are victim and assailant characteristics (including race, sex, age,
and gang affiliation) and incident information (including date and time of incident,
location of the incident and custody level at time of the incident).
The study was designed to provide comprehensive assessment of the
estimated 1,938 officially reported sexual assaults that occurred in the Texas
prison systems from January 1, 2002 through August 31, 2005. With nearly four
years worth of data collected, the research team has been able to determine
what prisoner attributes indicate who is more prone to either becoming a victim or
victimizer of such crimes. TDCJ is hopeful that this analysis can provide a
starting point to create risk screening instruments that may be used by the
department as part of their overall classification and risk assessment systems.
One instrument could be designed to identify prisoners who pose a high risk of
becoming a victim of such crimes while another could be used to identify
potential predators. As will be noted later on, there are many methodological,
legal and ethical issues associated with the labeling of prisoners in such a
manner. However, it may assist prison officials in preventing sexual assaults in
the future to explore if such a screening process is possible and how best to
implement it.
The research design also attempts to examine the influence of the “prison
environment” on the prevalence and nature of prison violence, including sexual
3

This number is based on the TDCJ’s count of active prisoners and is somewhat lower than the
data published by BJS.

-3-

The JFA Institute

assault. In consultation with the TDCJ, we identified several facilities that
represent locations with the highest and lowest rates of sexual assault, different
populations and custody levels. Seven facilities were selected for this qualitative
analysis. In these facilities we conducted extensive sites visits to try to determine
what environmental and management factors may explain variance in the
incidence of sexual assault violence.
This analysis was directed at
understanding management techniques that may impact the actual or reported
number of sexual assaults.
This information may assist TDCJ in making
modifications in terms of how it manages prisoners who either pose a high risk of
being victimized or are more likely to commit such crimes, in the on-going
training of staff in recognizing and managing such risk, and legal actions that can
be taken to respond to such crimes when they do occur.
Thus the overall purpose of the study is to: (a) produce and disseminate
sound empirical research to be used by practitioners and policy-makers on the
attributes of officially reported incidents of sexual assault; (b) explore screening
methods to better identify prisoners most likely to become involved in the such
crimes; and, (c) identify administrative and management methods that may serve
to prevent such assaults from occurring in the future.

-4-

The JFA Institute

II. Review of the Relevant Literature
A.

Overview

Over the past four decades, much has been written about the social
organization, structure and culture of the American prison system. Beginning with
Gresham Sykes classic 1958 book The Society of Captives, Erving Goffman’s
Asylums, and the more contemporary works of John Irwin (The Felon and
Prisons in Turmoil), Hans Toch (Living in Prison: The Ecology of Survival) and J.
Jacobs (Stateville: The Penitentiary in Mass Society), there has been
considerable analysis of how “total” institutions function and how the inmates
who are committed to them adapt to the prison environment. 4
Although these studies typically address the topic of prison violence, little
has been written on the number, type and form of sexual assault occurring within
adult and juvenile correctional facilities. When we conducted a NCJRS literature
search on the topic of prison rape and sexual assault, very few publications were
located. Text books on corrections also give short shrift to the topic of prison
rape. For example, Silverman’s comprehensive 550 page text on corrections
devotes only four pages to the topic of “sex in male prisons.” 5
However, these few studies do offer important information on the
difficulties in measuring the prevalence of prison sexual assault, the types of
prisoners who tend to become involved in such crimes as either a victim or
offender, the influence of the prison environment (design, culture and
management) on prison rape, and what actions correctional officials can take to
reduce the incidence of prison rape. Taken as a whole, any assessment of the
phenomenon of prison rape needs to incorporate a comprehensive research
design that sees the event as the result of an interaction of people and place. – in
other words, certain persons who are more likely to become either victims of rape
or the predators of such prisoners, and certain environmental conditions that
serve to increase or suppress the rate of sexual assault. What follows is a
summary of this literature which provided a context for developing our research
design.
B.

The Prevalence of Prison Rape

One of the major reasons for the passage of the Prison Rape Elimination
Act of 2003 was the recognition that there are no accurate estimates regarding
the extent of sex in adult and juvenile prisons, let alone the extent of sexual
4

See Sykes,G.M. 1958. The Society of Captives. Princeton, NJ: Princeton University Press;
Goffman, E. 1961. Asylums. Garden City, NY: Anchor Books; Irwin, J. 1970. The Felon.
Englewood, NJ: Prentice-Hall; Jacobs, J. 1977. Stateville: The Penitentiary in Mass Society.
Chicago: University of Chicago Press; and Toch, 1977. Living in Prison: The Ecology of Survival.
New York: Free Press.
5
Silverman, Ira J.2002. Corrections : A Comprehensive View. Belmont, CA: Wadsworth
Publishing.

-5-

The JFA Institute

assault. The enabling legislation states that an estimated 13% of the prisoners
now incarcerated in state and federal prisons have been sexually assaulted with
many being exposed to repeated assaults. This would mean that approximately
200,000 of the 1.6 million prison population have been raped.
Our review of those studies that have attempted to measure both reported
and unreported rapes shows that the rates vary dramatically by each study but,
like national victimization studies, the prevalence of rape as reported via self
report and confidential surveys far exceeds the number reported by prison
officials. Those studies that have been done are limited to adult prisoners (mostly
male) and report that while a significant number of prisoners are involved in
sexual behavior, a smaller number report being sexually assaulted, and an even
smaller number of cases are officially recorded by prison officials. 6 For example,
a study done of federal prisoners in the 1980s found that 30% had experienced
at least one homosexual experience but only 1% said that they had been forced
to have sex. 7
In a 1996 study of two maximum security prisons, researchers
found via self-report surveys that 22% of the prisoners had been sexually
assaulted since being imprisoned. 8
Much earlier, Torres found that 30% of the prisoners interviewed in Sing
Sing prison reported being involved in at least one homosexual event but a much
lower number reported being sexually assaulted. Here again the prison in
question was a high security facility that tends to house prisoners who are more
aggressive or require protection from such prisoners. 9
Most recently, NIJ commissioned a review of the literature to sort the
issues regarding how different methodologies and definitions of prison rape result
in different prevalence and incident rates. This report was written by Gerald G.
Gaes and Andrew L. Goldberg and was published in March 2004. 10
Gaes and Goldberg conducted a Meta analysis of all the studies that met
their criteria for inclusion. They found that the average prison lifetime sexual
assault prevalence was 1.91 percent, meaning that this percentage of inmates
has experienced a sexual victimization over a lifetime of incarceration. This
estimate was based primarily on studies which reported completed victimizations.
6

Fagan, T.J., Wennerstrom, D., and Miller, J. (1996). “Sexual assault of male inmates:
Prevention, identification, and intervention. Journal of Correctional Health Care 3(1):49-63.
7
Nacci, P.L. and T.R.Kane. (1983, December). “The Incidence of sex and sexual aggression in
federal prisons.” Federal Probation, 48(1)31-36. Nacci, P.L. and T.R. Kane. (March 1984). “Sex
and sexual aggression in federal prisons: Inmate involvement and employee impact.” Federal
Probation(48(1), 46-53.
8
Struckman-Johnson, C.,Struckman –Johnson, D., Rucker, L., Bumby, K., and Donaldson, S.
(1996). “Sexual coercion reported by men and women in prison.” The Journal of Sex Research.
33(1), 67-76.
9
Rettig, R.P., Torres, M.J., and Garrett, G.R. (1977). Manny: A criminal-addict’s story. Boston:
Houghton Mufflin.
10
Gaes, Gerald G. and Goldberg, Andrew L. (2004). Prison Rape: A Critical Review of the
Literature National Institute of Justice, U.S. Department of Justice, Washington, DC, March.

-6-

The JFA Institute

The authors discussed why prevalence measures fluctuate significantly and the
challenge of pinpointing a general prevalence rate for prison sexual assault:
“These studies use different methods to establish the level of
victimization (questionnaires, interviews, informants, administrative
records); they use different questions, and they use different time
frames. Definitions vary widely from rape to sexual pressure. Some
of these estimates rely on self-reported victimizations, while others
are based on the perceptions of inmates and staff on the overall level
of victimization in the prison. These latter estimates always appear
higher than self reports, and it is unclear what these latter estimates
mean since there is no presumption that inmates or staff actually
witness all of the sexual assaults they claim are occurring. Most
studies fail to report how long the sexual assault victim has been in
prison making it difficult to compare prisons across jurisdictions, due
to the likelihood of different exposure periods” 11
BJS in its most recent study of sexual violence published in July 2005
reported that at present “there are no reliable estimates of the extent of
unreported sexual victimization among prison and jail inmates and youth held in
residential facilities.” 12 Therefore, BJS was required to conduct a national
prevalence study under the Act. This study is now in the design stage as it
involves sophisticated data collection methods using a large representative
sample of correctional facilities to yield reliable victimization estimates for the
prison, jail and juvenile detention systems of each state.
In the meantime, the BJS July 2005 study is the first ever national survey
of administrative records on sexual violence in adult and juvenile correctional
facilities. The statistics compiled for this study are based on incidents reported to
correctional authorities during 2004. In other words, “what officials know” and
“how many allegations were reported.” The survey selected more than 2,700
correctional facilities holding 79% of all adults and juvenile in custody in the
nation.
Figure 1 shows the definitions of sexual conduct used by BJS in collecting
data for their survey. BJS used a comprehensive set of definitions distinguishing
nonconsensual sexual acts from abusive sexual contact, staff sexual misconduct
and staff sexual harassment. It is important to note that over half of the federal
and state prison systems surveyed were able to report along the definitions
below. For nonconsensual sexual acts, 66% of the prison agencies were able to
fully report; 51% for abusive sexual contacts; 67% for staff sexual misconduct;
and 53% for staff sexual harassment. Other jurisdictions were able to report

Ibid., Gaes, Gerald G. and Goldberg, Andrew L. (2004), page 2.
Beck, Allen J. and Hughes, Timothy A. (2005). Sexual Violence Reported by Correctional
Authorities, 2004 Bureau of Justice Statistics, U.S. Department of Justice, Washington, DC, July.
12

-7-

The JFA Institute

partially along the categories or using a combination of the definitions (as for
example, staff sexual harassment combined with staff sexual misconduct).
The BJS survey found that:
•

There were 5,528 allegations of sexual violence reported in the systems
sampled representing 8,210 allegations Nationwide in 2004 13

•

42% of the allegations involving staff sexual misconduct; 37% inmate-oninmate nonconsensual sexual acts; 11% staff sexual harassment; and
10% abusive sexual contact 14

•

The allegation rate of sexual violence per 1,000 inmates in 2004 was 3.15
(national estimate for all allegations) 15

•

The number of nonconsensual sexual acts in state operated prison
facilities reported in the survey was 1,229 for a 1.05 rate per 1,000
inmates and of this number 17.6% were substantiated at the time of the
survey, 45.3% were unsubstantiated and 37.2% were unfounded 16

Table 1 below shows the number of allegations of inmate-on-inmate
sexual violence and the number substantiated as reported by state prison
authorities. The rate of allegations per 1,000 and the percent substantiated was
computed by the authors using the BJS data. The BJA report also has similar
tables for inmate-on-inmate abusive sexual acts and staff sexual misconduct but
these are not presented here. There were 1,246 allegations of inmate-on-inmate
sexual violence reported by state and federal prison authorities as seen in the
table below. In addition, and not shown below, there were 287 allegations of
inmate-on-inmate abusive sexual contacts; 1,506 allegations of staff sexual
misconduct with inmates; and 417 allegations of staff sexual harassment of
inmates. 17
It is important to note that BJS warns that these results cannot be used to
rank states along sexual assaults rates. As the report states:
“The absence of uniform reporting and tracking procedures necessitates
caution when interpreting the 2004 survey results. The data should not be
used to rank systems or facilities. Higher or lower counts may reflect
variations in definitions, reporting capacities, and procedures for recording
allegations and not differences in the underlying incidence of sexual
violence.” 18
13

Ibid., Beck, Allen J. and Hughes, Timothy A. (2005), page 5.
Ibid., Beck, Allen J. and Hughes, Timothy A. (2005), page 1.
15
Ibid., Beck, Allen J. and Hughes, Timothy A. (2005), page 1.
16
Ibid., Beck, Allen J. and Hughes, Timothy A. (2005), Tables 3 and 5.
17
Ibid., Beck, Allen J. and Hughes, Timothy A. (2005), Appendix table 1a and 2 a.
18
Ibid., Beck, Allen J. and Hughes, Timothy A. (2005), Page 4.
14

-8-

The JFA Institute

Figure 1: How Sexual Violent Was Measured by the BJS Sexual Violence
Study of July 2004
How sexual violence was measured
The definition of “rape” as required under the Prison Rape Elimination Act of
2003 was operationalized by disaggregating sexual violence into two categories of
inmate-on-inmate sexual acts and two categories of staff sexual misconduct. The
inmate-on-inmate categories reflected uniform definitions formulated by the National
Center for Injury Prevention and Control, in “Sexual Violence Surveillance: Uniform
Definitions and Recommended Data Elements,” Center for Disease Control and
Prevention. The categories were –
Nonconsensual sexual acts
Contact of any person without his or her consent, or of a person who is unable to
consent or refuse; and
• Contact between the penis and the vagina or the penis and the anus including
penetration, however slight; or
• Contact between the mouth and the penis, vagina, or anus; or
• Penetration of the anal or genital opening of another person by a hand, finger, or
other object.
Abusive sexual contacts
Contact of any person without his or her consent, or of a person who is unable to
consent or refuse; and
• Intentional touching, either directly or through the clothing, of the genitalia, anus,
groin, breast, inner thigh, or buttocks of any person.
Definitions of staff sexual misconduct and staff sexual harassment were based
on “Training for Investigators of Staff Sexual Misconduct,” prepared by the National
Institute of Corrections.
Staff sexual misconduct
Any behavior or act of a sexual nature directed toward an inmate by an employee,
volunteer, official visitor, or agency representative. Romantic relationships between
staff and inmates are included. Consensual or nonconsensual sexual acts include:
• Intentional touching of the genitalia, anus, groin, breast, inner thigh, or buttocks
with the intent to abuse, arouse, or gratify sexual desire; or
• Completed, attempted, threatened, or requested sexual acts; or
• Occurrences of indecent exposure, invasion of privacy, or staff voyeurism for
sexual gratification.
Staff sexual harassment
Repeated verbal statements or comments of a sexual nature to an inmate by
employee, volunteer, official visitor, or agency representative, including:
• Demeaning references to gender or derogatory comments about body or clothing;
or
• Profane or obscene language or gestures.
Source: Beck, Allen J. and Hughes, Timothy A. (2005). Sexual Violence Reported by
Correctional Authorities, 2004 Bureau of Justice Statistics, U.S. Department of Justice,
Washington, DC, July. Page 3.

-9-

The JFA Institute

There are a number of key findings in the BJS report. First as can be seen
in Table 1 below, the inmate-on-inmate sexual violence rate per 1,000 inmates is
extremely low (less than one per 1,000 prisoners and the substantiation rate is
even lower (about 12% of all allegations are substantiated).
Second, although all of the state rates are extremely low, there is
considerable variation among the states even when looking at large and small
prison systems. For example, among the four most populous states of Texas,
California, New York, and Florida there is significant variation in their reporting
and substantiation rates. Texas has the highest reported number of incidents at
550 for a rate per 1,000 of 3.95, almost four times the national average for the
states of 1.05. It also has one of the lowest substantiation rates (less than three
percent). Yet, California, with a larger prison population reported only 23 inmateon-inmate sexual violence incidents for a rate of .14, ten times below the national
average. But California also claims it has substantiated all of its allegations.
New York, with 15 incidents and a rate of .23 and Florida with 75 incidents and a
rate of .97 were also below the national average but have higher substantiation
rates as compared to Texas.
Do these variations in reported incidents and rates reflect differing levels
of sexual assault? Obviously, the severe degree of variation among the states
show that what is being measured here is variation in the reporting systems of
each state. For example, the Texas prison rape prevention program includes
monitors in each prison facility and extensive training and reporting requirements.
It may be that the higher numbers reported in Texas are a result of this
aggressive policy and not necessarily the results of the state prison system
having a higher number of sexual assaults in comparison to the other states.
However, the BJS national incidence study, when completed, will provide a better
answer to this question and a more reliable benchmark to measure prevalence of
sexual assaults among the different state prison systems.
The challenge of defining sexual assaults and encouraging their reporting
in a prison setting is significant. NIJ funded Dr. Mark Fleisher to conduct an
anthropological study of inmate culture to better understand how inmates
perceive and deal with sexual assaults. The preliminary findings of the study
were reported by the National Institute of Corrections (NIC) in their first annual
report required by the Act. 19 This study was designed “to identify and clarify
definitions of sexual activity in prisons, and to help policymakers and practitioners
better understand the differences between consensual, coercive, and predatory
sex in corrections institutions.” 20 The research reports that most often inmates
are unfamiliar with the terms “coercive sex and consensual sex” and the inmate
19

National Institute of Corrections, Annual Report to Congress, Prison Rape Elimination Act
(PREA) Public Law 108-79, September 2004. Appendix A: Rape and Coercive Sex in American
Prisons: Interim Findings and Interpretation on Preliminary Research, page 26-34.
20
National Institute of Corrections, Annual Report to Congress, Prison Rape Elimination Act
(PREA) Public Law 108-79, September 2004, page 5.

- 10 -

The JFA Institute

Table 1: Allegations of Inmate-on-Inmate Sexual Violence Reported by
State or Federal Prison Authorities, by Type, 2004

Total
Federal/b
State
Alabama
Alaska
Arizona
Arkansas
California
Colorado/e
Connecticut
Delaware
Florid
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming

Prisoners

Allegations

1,318,616
151,650
1,166,966
24,768
3,158
26,833
12,655
160,703
16,609
18,814
6,778
77,647
44,026
3,877
4,621
44,379
21,236
8,611
9,181
10,814
16,672
1,986
23,622
10,043
48,111
7,827
11,456
30,139
2,074
4,053
10,152
2,426
23,752
3,703
64,778
35,219
1,176
42,231
17,727
12,678
39,823
3,494
23,321
3,157
14,306
139,148
4,550
1,632
29,514
16,765
3,987
21,560
1,174

1,246
17
1,229
6
0
18
4
23
5
6
3
75
51
6
3
17
18
4
21
7
1
0
3
12
39
13
3
17
2
12
15
0
1
4
15
15
0
86
29
16
9
9
14
2
8
550
18
6
5
12
12
31
3

- 11 -

Rate Per
1,000
0.94
0.11
1.05
0.24
0.00
0.67
0.32
0.14
0.30
0.32
0.44
0.97
1.16
1.55
0.65
0.38
0.85
0.46
2.29
0.65
0.06
0.00
0.13
1.19
0.81
1.66
0.26
0.56
0.96
2.96
1.48
0.00
0.04
1.08
0.23
0.43
0.00
2.04
1.64
1.26
0.23
2.58
0.60
0.63
0.56
3.95
3.96
3.68
0.17
0.72
3.01
1.44
2.56

Substantiated
152
0
152
2
0
4
1
23
3
0
0
2
0
0
3
1
0
4
2
2
0
0
1
2
17
4
0
3
1
0
4
0
0
1
2
0
0
14
2
5
3
3
1
0
2
13
2
1
1
4
11
7
1

Percent
Substantiated
12.2%
0.0%
12.4%
33.3%
0.0%
22.2%
25.0%
100.0%
60.0%
0.0%
0.0%
2.7%
0.0%
0.0%
100.0%
5.9%
0.0%
100.0%
9.5%
28.6%
0.0%
0.0%
33.3%
16.7%
43.6%
30.8%
0.0%
17.7%
50.0%
0.0%
26.7%
0.0%
0.0%
25.00%
13.3%
0.0%
0.0%
16.3%
6.9%
31.3%
33.3%
33.3%
7.1%
0.0%
25.0%
2.4%
11.1%
16.7%
20.0%
33.3%
91.7%
22.6%
33.3%

The JFA Institute

“socio-sexual system is relativistic” and rape “have a slippery and fuzzy
definitions”. 21 Prison sex is akin to an “amoral exchange of service for service
(sex for protection), service for property (sex for cigarettes or commissary items),
or a chain of such exchanges”. 22 Prison rape is a more common topic of
conversation than an actual act. 23 According to Fleisher, the Act “has objectified
definitions of rape and coercive sex; however, this research shows that PREA
rape and coercive sex criteria are dissimilar from inmates’ perceptions and
interpretations of coercive sex and rape. Such dissimilarity points to fundamental
differences between PREA and inmate socio-sexuality.” 24
The most common inmate terms that matches the PREA definition of rape
is that of “predator”. 25 A predator is an “inmate who has the intention of
engaging in violent sex.” 26 Predators punch their victims and rape them when
they are unconscious. “The unconscious rape of a victim is cited as a common
context, however, the frequency of such violent rapes to be relatively rare”. 27
C.

Prisoner Attributes and Risk Assessments

There is widespread belief that correctional officials can do a better job of
identifying prisoners who pose a higher risk to either being raped or perpetrating
such crimes. Prison wardens believe that institutional policies and procedures,
staff training and increased staff supervision are effective means of preventing
sexual assaults between inmates. Wardens that are more aware of the sexual
assault incidents that occur in their institutions are also more likely to believe that
additional staff and training are effective deterrents. 28
During the past decade, prison systems have experienced increased
pressure to improve their approaches to classify prisoners according to custody,
work, and programming needs. Fueled by litigation on rates of violence and
over-crowding, classification systems are viewed as a principal management tool
for allocating scarce prison resources efficiently and minimizing the potential for
violence and/or escape. These systems are also expected to provide a greater
level of safety to staff and prisoners by identifying and housing prisoners
according to the risk they pose.

Ibid., National Institute of Corrections, September 2004, Appendix A, page 20.
Ibid., National Institute of Corrections, September 2004, Appendix A, page 28.
23
Ibid., National Institute of Corrections, September 2004, Appendix A, page 29.
24
Ibid., National Institute of Corrections, September 2004, Appendix A, page 33.
25
Ibid., National Institute of Corrections, September 2004, Appendix A, page 32.
26
Ibid., National Institute of Corrections, September 2004, Appendix A, page 32.
27
Ibid., National Institute of Corrections, September 2004, Appendix A, page 32.
28
Hensley, Christopher; Dumond, Robert W.; Tewksbury, Richard; and Dumond, Doris (2002).
“Possible Solutions for Preventing Inmate Sexual Assault: Examining Wardens’ Beliefs” American
Journal of Criminal Justice, Vol. 27, No1.
22

- 12 -

The JFA Institute

Objective prison classification systems were originally adopted in the
1980s, but by the late 1990s significant modifications improved classification
practices, including new risk assessment measures. The number of prisoners
over-classified in higher custody levels has been reduced; custody decisions are
made more consistently; criteria for custody decisions have been validated;
prisoner program needs are assessed systematically; and institutional safety for
both staff and prisoners has been increased.
While most prison systems have implemented successful objective
classification systems for the general population, less attention has been devoted
to identifying prisoners who require special placement or need to be managed
differently from the so called “typical prisoner”. Such remedies have consisted of
placing prisoners in administrative segregation or protective custody housing
units. Unfortunately, these designations and placements often occur after the
fact, when, because of disruptive and aggressive behavior, these prisoners need
to be removed from the general population for either his own protection or to
protect other prisoners in the general population. This is especially true for
persons who have been sexually assaulted or fear such assaults. Often such
prisoners are removed from the general prison population and assigned to a
segregated protective custody unit.
As suggested earlier, sexual behaviors within a prison are rarely reported
by inmates or are discovered by prison officials. And, because these incidents
are rarely recorded, it creates a “low base-rate” or variance problem making it
difficult to predict which prisoners and under what circumstances such acts will
occur. Thus developing a “risk instrument” for identifying potential sexual assault
victims or assailants, while desirable, may be more difficult to achieve.
There are a limited number of instruments now being used to assess the
risk of either violence or sexual assault (Sexual Violence Risk-20 29 ; the Hare
Psychopathy Checklist, Revised; and Violence Risk Assessment Guide 30 ). But
these instruments have been designed based on recidivism studies of persons
released from prison that subsequently are re-arrested and/or convicted for a
violent crime including sexual assault or rape. They have not been used to
identify prisoners who are prone to become victims or perpetrators of sexual
assault.
Those studies that have been completed have found that certain
demographic factors tend to be associated with these crimes including age, race,
stature, gang affiliation, sexual orientation, prior criminal record and prior
incarcerations, prior prison sexual assaults, mental health status, amount of time
served to date, and gender. More precisely, victims tend to be young, white male
29

Dunne, Felicity. (2000) A Framework for Reducing Reoffending: Differentiated Case
Management In Victorian Corrections.
30
Quinsey, V., Harris, G., Rice, M., Cormier, C. (1998) Violent Offenders. Appraising and
Managing Risk. American Psychological Association, Washington DC

- 13 -

The JFA Institute

prisoners who are of small stature, are not affiliated with a prison gang or
organization, who are just starting their sentences and have a history of mental
health problems.
On the matter of race, it is also generally accepted that while most acts of
violence in the “free community” are intra-racial, rapes in prisons tend to be interracial with a higher proportion of whites being victimized by black prisoners. This
phenomenon has been used by some to argue that rape is not so much an
expression of sexual deprivation but more a reflection of the long-standing race
relations in the United States. For example, Carrol found that among officially
reported rapes, 75% were black assailants and white victims. 31 The same
statistical trend is being reported by the TDCJ with many of the prisoners being
associated with prison gangs.
Fagan et al. found attributes associated with victims, also known as
punks, kids or catchers. These victims tend to be smaller in stature, younger than
their aggressors, seen as weak, passive and easily intimidated, and tend to be
first time prisoners who are not “street wise”, not affiliated with a prison gang or
organization, and often convicted of child molestation. 32 Dumond found certain
categories of male prisoners who were especially vulnerable. These included:
“(1) the young and inexperienced; (2) the physically weak
and small; (3) those suffering from mental illness or
developmental disabilities; (4) those who are not “tough” or
“streetwise”; (5) those who are not gang affiliated; (6) the
homosexual, transgendered, or overtly effeminate; (7) those
who have violated the “code of silence”; (8) those who are
disliked by staff or other inmates; and (9) those who have been
sexually assaulted.” 33
Criminologists have argued that like rape on the “outside”, prison rape has
less to do with the deprivation of normal sexual outlets and more to do with
conquest and control, revenge or retaliation, sadism and degradation, status and
affiliation, and maintaining social hierarchy in prison. With this background
information, the following categories of high risk or special management
populations are instructive. These are shown in Table 2 below. These “special
management” designations relate to the identification and housing/separation,
and supervision of prisoners likely to become either a victim or perpetrator of a
sexual assault. For prison administrators, proper screening for risk and the
31

Carroll, Leo. (1974). Hacks, blacks, and cons: Race relations in a maximum security prison.
Lexington, MA: Heath.
32
Fagan, T.J., Wennerstrom, D., and Miller, J. (1996). “Sexual assault of male inmates:
Prevention, identification, and intervention. Journal of Correctional Health Care 3(1):49-63.
33
Dumond, Robert W. (2003) “Confronting America’s Most Ignored Crime Problem: The Prison
Rape Elimination Act of 2003” in The Journal of the American Academy of Psychiatry and the
Law, Volume 31, Number 3, page 355.

- 14 -

The JFA Institute

management of the “known risk” can serve to prevent such incidents from
occurring.
Finally, it is also well known that similarly designed facilities with similarly
situated prison populations can produce very different rates of prisoner
misconduct both within and across state prison systems. Relative to sexual
assaults, they tend not to be randomly distributed across similarly designed
prisons.
As we will discuss later, the number of sexual assaults in Texas
prison vary by facility, with some producing a disproportionate number of
reported rapes even though the “base rate” is still low for all facilities. The
research presented here tries to examine if such variations in misconduct rates
for prisons that are equivalent in design and prison population attributes relate to
differences in management style adopted by each prison administrator. Issues of
management “attitude” have influenced the approach to the measurement,
prevention and prosecution of prison sexual assaults as discussed below.
D.

Prison Management and “Deliberate Indifference”

Advocates of prison rights have raised the concern that correctional
systems have traditionally ignored the prevalence of sexual assault and actually
allowed it to occur by not taking preventive actions that would reduce the extent
of the problem. In this sense, there is a culture of indifference which not only
results in a high degree of under-reporting but also can also encourage the
phenomena of prison rape. These are situations where prison officials were
aware of a sexual assault or misconduct but refused to intervene. Sexual
assaults were seen by prison officials as part of the prison experience and it was
up to the prisoner to fend for himself. There is also the implication that
homosexual prisoners, who are more likely to become the targets of such rapes,
are merely reporting consensual sexual acts that have became more aggressive
in nature.
Man and Cronan argued that a possible legal remedy for the victims of
prison rape was to claim that correctional administrators, by their failure to
recognize and take pre-emptive actions to prevent such crimes from occurring,
were in violation of the Eighth Amendment of the U.S. Constitution prohibiting
cruel and unusual punishments. 34 This perspective is grounded in the Farmer v.
Brennan 1994 decision. The case involved another situation where prison
officials failed to take action to prevent sexual assaults from occurring. The
prisoner in question had feminine characteristics including breast implants but
was placed in the general population where he was beaten and raped. What has
been now referred to as the “Farmer Deliberate Indifference” standard as defined
by the Supreme Court consists of a two-part test.

Man, Christopher and John P. Cronan. “Forecasting Sexual Abuse in Prison: The Prison
Subculture of Masculinity as a Backdrop for "Deliberate Indifference". Journal of Criminal Law
and Criminology. 92:1. Fall 2001/Winter: 127-185.

- 15 -

The JFA Institute
Table 2: Typology of High-Risk and Special Management Inmates
Category
Security Threat Group

Assessment Method
Subjective assessment based on at
least three sources of independent
objective data as applied to well-defined
agency criteria.
Subjective assessment based on at
least three sources of independent
objective data as applied to well-defined
agency criteria.
Standardized psychometric tests and
clinical judgment by MH staff.

Placement
Administrative Segregation or
General Population -- High
Custody

Chronic Misbehavior
Assaultive

Objective External Class

Non-Assaultive

Objective External Class

General Population – High
Custody, Administrative
Segregation, or Mental Health
Unit
General Population – High
Custody, Administrative
Segregation, or Mental Health
Unit

Likely Victim

Mentally Ill

Predator
Non-Sexual Predator

Sexual Predator

Developmentally Disabled

Subjective assessment based on at
least three sources of independent
objective data as applied to well-defined
agency criteria.
Subjective assessment based on at
least three sources of independent
objective data as applied to well-defined
agency criteria.
Standardized psychometric tests and
clinical judgment by MH staff.

Protective Custody or
Restricted General Population
Facilities
Mental Health Unit and/or
Administrative Segregation

General Population – High
Custody, Administrative
Segregation, or Mental Health
Unit.
General Population – High
Custody, Administrative
Segregation, or Mental Health
Unit.
General Population (all custody
levels) or Mental Health Unit.

First, the victim (prisoner) must show that the circumstances of his
incarceration present a substantial risk of serious harm. Then one must establish
that prison officials were aware of the situation but failed to provide a remedy.
Thus they acted with “deliberate indifference” in failing to keep the inmate/victim
safe from harm. Since certain conditions are known to contribute to violence and
sexual assault, prison officials have a constitutional obligation to intervene in
order to keep prisoners safe from harm. If no action is taken, one can claim that
such inaction constitutes “deliberate indifference” and is therefore actionable in a
court of law.
In a more general sense, and related to correctional policies, advocates
have also argued that (a) overcrowding which limits the ability of prison officials
to properly classify inmates in appropriate housing, (b) inadequate staffing and
supervision of inmates, (c) lenient disciplinary actions against perpetrators, (d)
weak investigation techniques of alleged sexual assaults and (e) indifference

- 16 -

The JFA Institute

from prison authorities may aggravate conditions that facilitate sexual assaults or
intimidation that turns into apparently consensual sex. 35
E.

The Risk to Public Health: Transmission of AIDS and Other
Communicable Diseases

Finally, there is the concern of the extent to which sexual assault further
increases the transmission of HIV or AIDS among the prisoner population and
the general public once the prisoner who has been assaulted and infected is
released. Especially in the case of un-reported sexual assault, the victim may be
reluctant to report any un-protected sexual activity. Similarly, infected predators,
unless identified and segregated from the general population, may also be
infecting other prisoners. One example is Michael Blocker who was a prisoner in
the Illinois prison system who claimed that he was made a sex slave while
incarcerated at the Menard Correctional facility and that prison officials failed to
protect him from repeated assaults. He was subsequently paroled but was later
found to be HIV infected. He successfully sued the prison administration and
was awarded $1.5 million in damages. 36 In part due to this nationally publicized
case, correctional agencies began to routinely test their prisoner population at
admission..
The BJS reports that a relatively small percent of the nation’s prison
In 2002, there were about 22,317 state
population is infected with HIV. 37
prisoners with HIV or AIDS of which the Texas prison system reported 2,528.
This represents about 2% of the entire prison system (both national and in
Texas). Further, the rate of infections has remained relatively stable since 1995
when there were 24,256 such cases. Nonetheless, after “natural causes”, AIDS
is the leading cause of death among prisoners. The report is not able to
determine the extent to which these prisoners were infected prior to their
incarceration or were infected while incarcerated through either consensual or
forced sex. According to monthly health reports filed by the Texas Department of
Criminal Justice, about 48 percent of all prisoners diagnosed with HIV or AIDS
caught the virus through intravenous drug use. Here again, however, prison
officials do not know how widespread drug use is behind bars, or how many
users are exposed to HIV in prison, and whether such infections result from
sexual assaults. 38

Human Rights Watch (2003). No Escape
From thief to cellblock sex slave: A convict’s testimony. (1997, October 19), New York Times,
Section4, p.7.
37
Maruschak, Laura M.(2004) HIV in Prisons and Jails, 2002. Bureau of Justice Statistics, U.S.
Department of Justice, Washington, DC. December.
38
Rodriguez, Brenda “HIV, AIDS, and Rape in Texas Prisons” from States of Confinement:
Policing, Detention, and Prisons, pp. 159-171, 2000, Joy James, ed.
36

- 17 -

The JFA Institute

III.

Texas Prison Safe Program

Overview Texas Prisons

Texas has the second largest prison system in the country after the
federal system. The system is operated by the Texas Department of Criminal
Justice (TDCJ). This state agency administers the correctional institutions
(through its Correctional Institutions Division), the parole system, and the division
that sets probation standards and distributes state probation funding.
At mid-year 2004, Texas had 169,110 prisoners in different types of
correctional institutions, second only to the 179,210 prisoners in the federal
system. Texas had an incarceration rate at 704 per 100,000 population; second
only to Louisiana’s 814. 39 Texas prison officials operate 106 facilities of which
51 are prisons, seven are prisons run by private operators, 22 are a special type
of prison, like a transfer facility, medical facility or pre-release center, and the rest
are facilities housing the equivalent of fourth degree felons (16 state operated
and 5 privately operated “State Jails”) or substance abusing prisoners in
treatment (five facilities called Substance Abuse Felony Punishment Facilities).
TDCJ’s annual budget now exceeds $2.1 billion. 40
In terms of the classification, the agency has established a Classification
and Records Central Office which oversees the custody level assigned to each of
the approximately over 70,000 persons admitted to TDCJ prisons and states
jails. Each prisoner admitted undergoes an extensive assessment at one of the
state’s reception facilities. That process results in a prisoner typically being
assigned to one of five security levels with G1 (General Population 1) being the
lowest security level and G5 the highest. Numerous other custody codes exist to
protect inmates and staff in the management of the prisoner population.
In addition to the security level designation, a number of special
management flags have traditionally been available to be applied to prisoners
who pose a possible risk to the security of the prison system. These are referred
to as “security precaution designator” and include Escape (ES), Staff Assault
(SA), or Hostage (HS). Notably, prior to the adoption of the Safe Prisons
Program as described in the next section, there were no flags to designate a
person as either a possible sexual predator or a possible victim to such assaults.
All prisoners are reviewed thereafter on an annual basis for purposes of updating
and revising the designated security.
TDCJ maintains the information above in a comprehensive database that
also captures the relevant demographic and criminal record data, retains each
prisoner’s history of transfers within the prison system, participation in a variety of
39

Harrison, Paige M. and Beck, Allen J. (2005). Prison and Jail Inmates at Midyear, 2004 .
Bureau of Justice Statistics, U.S. Department of Justice, Washington, DC. April.
40
Texas Department of Criminal Justice (2004). Statistical Report, Fiscal Year 2004. December.

- 18 -

The JFA Institute

prison treatment programs, and the prisoner’s disciplinary conduct history.
These data are used later for the analysis in this study.
B.

Texas Safe Prison Program

In 2001, the Texas Legislature enacted a provision as part of TDCJ’s
funding that required the establishment of a Safe Prisons Program. The primary
program goal is to prevent and limit the number of sexual assaults by inmates on
inmates. According to the legislative directive:
“Strategies to prevent sexual assaults that may be used in
the Safe Prisons Program include, but are not limited to, use of
protective custody; use of an inmate’s assault history in making
cell assignments; use of an inmate’s likelihood of victimization in
cell assignments; education of correctional officers on the
importance of preventing sexual assault, including prosecution;
and use of surveillance cameras.” 41
Sexual assault was defined as:
Forcing another person, by violence or threats of violence, to
perform a sexual act (a sexual act is any intentional contact
between the genitals, anus, groin, breast, inner thigh, or
buttocks of one person and the genitals, anus, groin, breast,
inner thigh, buttocks, mouth, or hands of another person),
sexual fondling, or sexual assault with an object. The term
Sexual Abuse includes the acts relating to sexual assault as
described in the Texas Penal Code. 42
TDCJ’s definition of sexual assault is quite broad, and includes both
nonconsensual sexual acts and abusive sexual contacts, as defined by BJS, in
its scope. It is worth noting that some acts that fall under this definition of “sexual
assault” do not meet the Texas Penal Code criteria for prosecution under state
law. TDCJ uses this terminology for classification and disciplinary purposes,
while the Office of the Inspector General (OIG) also investigates cases that meet
state criteria for possible criminal prosecution outside of the agency.
As a method to monitor program implementation and effective operations,
the legislature also included a provision that requires the agency to report to the
Legislative Budget Board and the Governor the number of sexual assaults by
inmates on inmates and the actions taken on each assault. Lastly, the rider
created a Safe Prisons Coordinator who reports directly to the TDCJ Executive
Director.

41
42

Safe Prisons Rider, General Appropriations Act, 78th Texas Legislature, Rider 59
Texas Department of Criminal Justice (2004). Safe Prisons Program Update

- 19 -

The JFA Institute

Figure 2 below depicts the operational structure of the Safe Prisons
Program. The agency created a Safe Prisons Program Council of correctional
experts to provide guidance to the program and designated in each prison unit a
Unit Safe Prison Program Coordinator who is responsible for the overall
monitoring of the program at the facility level. The program also extended
victims rights to victimized prisoners, which includes offering a victim
representative if the investigation determines that a sexual assault exam was
required.
Since the passage of the act the agency has further enhanced and
defined program components, created new policy directives and enhanced
others that are geared towards maintaining a safe and secure prison
environment for prisoners, established a data tracking system, implemented selfmonitoring, created an additional administrative position to manage the day-today program operations and instituted an advisory committee to address specific
policy issues related to prison sexual assaults. A brief description of each
strategy is included below.
Education – The TDCJ Training Department has modified its Pre-Service
and In-Service curricula to include a module that addresses the Safe Prisons
Program for staff. The four-hour training block addresses prisoner sexual assault
(prevention and care after an assault), prisoner extortion and prisoner life
endangerment, as well as specific training on attendant policy directives. An
“Institutional Character Profile” (ICP) has been instituted to identify the attitudes
of staff and prisoners as they related to sexual assaults in the prison setting and
then introduce education and training to address these attitudes. Additional
information is also provided to staff agency-wide through agency publications
and website. Prisoners are also educated on the risks of sexual assaults and the
prosecution process at the reception centers and upon arrival at their unit.
Housing Assignments – Several policies have been created to
appropriately classify prisoners for housing and job assignments. These
assignments are based on information reported by the prisoner, previous
assault/criminal history or information obtained from official documents. Although
there is no objective based system for identifying at-risk prisoners, prisoners who
meet specific criteria may be placed in protective custody and/or safekeeping
status.
Facility Enhancements – The TDCJ is constantly reviewing information
on how it can best address sexual violence in its institutions. One strategy used
is the installation of surveillance cameras at specific locations on units. Using
available data, the administration has invested in placement of various types of
cameras for housing areas, recreation yards or isolated areas. Additional
cameras are planned for implementation as data dictates.

- 20 -

The JFA Institute

Figure 2: Operational Structure of the Texas Safe Prisons Program

Safe Prisons Program
Coordinated by Director of
Correctional Institutional Division (CID)

October 2003
Safe Prisons Management Office

Monitor alleged incidents to assure
agency compliance with policies

November 2003
Safe Prisons Program Council

Facilitates training and awareness
programs with staff and prisoners

Group of criminal justice professionals
providing guidance

Identifies issues for further policy
development and collects statistical
information on alleged sexual assaults

Unit Safe Prison Program Coordinator
Perform facility based initiatives
Responsible for overall monitoring of the Safe
Prison Program at the facility level

Source: Memorandum, Brad Livingston, Interim Executive Director, Texas Department of Criminal
Justice, “Summary of the State Prisons Program”.

- 21 -

The JFA Institute

Statistics – The agency has several databases that contain varying
degrees of information on sexual assaults. The Emergency Action Center (EAC)
collects limited data on all serious incidences, including alleged sexual assaults.
In addition to the EAC statistics, data is maintained at the unit level on all
prisoners identified or requesting protective custody/transfers as a result of a
sexual assault. The Safe Prison Program Manager has established a database
on all sexual assaults that are reported and/or investigated by their offices. In
addition, this database contains specific data regarding the disposition of the
case at the institutional level. Additional data can also be found in the Office of
the Inspector General (OIG), which independently investigates sexual assault
cases and decides if there is enough evidence to substantiate prosecution of
perpetrators. Such cases are then referred to the local District or County Attorney
for possible prosecution.
As a summary of the program states: “All allegations of sexual assault are
referred to the OIG for investigation, those prisoners identified as potential
predators by the OIG are then identified by the Safe Prison Program Manager on
the TDCJ Mainframe System. The identification aids the facilities when
considering housing or programs for the prisoner and parole when determine
post release programs when the prisoner leaves custody.” 43
The Safe Prisons Program Manager independently reviews each allegation of
sexual assault reported on the EAC report, and monitors the actions of the OIG
and the Classification Office to ensure that the incident has been appropriately
reviewed and that appropriate actions have been taken with respect to referral to
the prosecutor, and any changes to the prisoners’ classification status and
housing location.
Figure 3 below shows the computer code definitions used to identify potential
victims and predators. These codes can be used to monitor and house prisoners
with a history of sexual assault or history of allegations of sexual assault. 44
Table 3 below shows the sexual assault investigation checklist delineating all the
processes that need to be completed when an inmate alleged a sexual assault.
Table 4 shows the prisoner protection investigation checklist.
Note that the information collected as part of the checklist is not the same
as the data collected by the Safe Prison Management Office for its database.
Consequently, the interview questions concerning extortion and coercion were
not available for this study.

The Texas Safe Prison Program seems to integrate the key elements that
can be defined as a “best practices” approach to dealing with the prevention,
43

Source: Memorandum, Brad Livingston, Interim Executive Director, Texas Department of
Criminal Justice, “Summary of the State Prisons Program”.
44
Texas Department of Criminal Justice (2004). Safe Prisons Program Update

- 22 -

The JFA Institute

victim services and investigation of prison sexual assaults (Figure 4). The extent
to which these practices have been appropriately implemented, with adequate
funding and accountable management practices, was not an issue examined
here although the present research indirectly addresses some implementation
issues.

Figure 3: Computer Codes Used in Texas Prison System Main Database to
Identify Potential Perpetrators and Victims of Sexual Assault
PD – Sexual Predator
• Prisoner found guilty of Sexual Assault or serving a sentence for
Sexual Assault while incarcerated in an adult penal institution
PP – Potential Sexual Predator
• Prisoner has a history of alleged sexual assaults as the assailant.
This matter has been investigated by OIG and found to meet the
penal code elements. A history is defined as more than two
allegations.
SV – Sex Victim / Potential Sex Victim
• Prisoner is the victim of a sexual assault which has resulted in the
finding of guilt of the assailant. Prisoner is a potential sex victim as
evidence by a history of sexual assault allegations. This matter
has been investigated by OIG and found to meet the penal code
elements. A history is defined as more than two allegations.

- 23 -

The JFA Institute

Table 3: Texas Safe Prison Program Sexual Assault Investigation Checklist
Date
Completed

Time Completed
(circle am/pm)
(a.m./p.m.)

Procedures
1.

Upon the allegation being made, immediately escort the alleged
victim to the medical department. Notify medical staff of the
prisoner’s allegation. OIG will make the determination as to the
necessity of completing a rape kit.
(a.m./p.m.) 2.
If there are no visible injuries, take one photograph of the front and
one of the back of the prisoner while he is wearing State-issued
underwear. If there are injuries, take photographs of the injuries as
well as the front and back of the prisoner. All photographs are to be
immediately labeled and signed by the security staff member taking
the photographs.
(a.m./p.m.) 3.
Notify the Office of Inspector General
(a.m./p.m.) 4.
Notify Duty Warden
(a.m./p.m.) 5.
Determine the location of the alleged assault and protect and isolate
it as a crime scene. Take photographs and videotape of the area.
(a.m./p.m.) 6.
If possible, collect the clothing the alleged victim was wearing at the
time of the assault or immediately after and place the clothing in
separate paper sacks. Properly label the items utilizing the evidence
tags in Central Control. Limit the number of people who handle
these items to the absolute minimum and complete the chain of
custody.
(a.m./p.m.) 7.
Identify the alleged assailants and immediately place them in PreHearing Detention.
(a.m./p.m.) 8.
Take photographs of the assailants and immediately label them.
(a.m./p.m.) 9.
Collect the clothing of the assailants and place the items in separate
paper sacks. Label the sacks utilizing the evidence tags in Central
Control. Limit the number of people who handle this evidence to the
absolute minimum.
(a.m./p.m.) 10. Report the incident to the Emergency Action Center (EAC). This
must be done within three (3) hours of when the alleged incident
was reported by the prisoner.
(a.m./p.m.) 11. Identify all staff who would have been in the vicinity of the alleged
assault and interview them. Obtain written statements from all staff
members.
(a.m./p.m.) 12. Draft a written report documenting the entire incident: before,
during, and after. Include all of the information obtained in the
previous steps. The names of the staff who collected evidence,
including photographs and videotapes, should be documented with
what they specifically did and outlined in the text.
(a.m./p.m.) 13. Complete the TNG-93 and send e-mail to EAC and all other
pertinent staff.
(a.m./p.m.) 14. Complete a Prisoner Protection Investigation in accordance with AD04.69 and turn into the Chief of Unit Classification.
(a.m./p.m.) 15. Inform medical personnel to refer the victim to an Agency
psychologist..
Source: Texas Safe Prisons Management Office, January 2005

- 24 -

The JFA Institute

Table 4: Texas Safe Prison Program Prisoner Protection Investigation
Checklist
Check if
Completed

Responses/Comments

Issues
1.
2.
3.
4.
5.

8.
9.
10.

11.

12.

13.
14.
15.
16.

- 25 -

Name of offenders threatening you, AKA or
housing area
If name is not known, can you identify
offenders from a video or a photo lineup?
Are the offenders threatening you gangrelated?
What STG if any?
Are you gang-related?
If yes:
a. What gang?
b. Are you active?
c. Are you an ex-member?
(complete extortion checklist)
Are you paying protection?
If yes:
a. Offender’s names/AKA
b. Is he STG-related?
c. What are you paying with?
Have you been involved in sexual activities
either voluntary/non-voluntary?
(If nonvoluntary,
complete
sexual
assault
checklist.)
Were you housed with a particular group in
County Jail?
Were you housed in protective custody in
county jail?
Were you involved in any incidents in
county jail (fights, assaults, extortion, or
gang-related activities)?
Contact the County Jail to verify any
information to substantiate the offender’s
allegations.
Were you assaulted?
If yes:
a. Check with medical records to verify
any injuries.
b. If there are visible injuries, have the
offender checked by the medical staff and
note all injuries on the investigation.
Check the classification folder for any
previous offender protection investigations.
Get any witness statements either from
offender or staff.
Verify as much information from the
offender as possible.
Is there any other pertinent information that
you could provide for this investigation?

The JFA Institute

Figure 4: Best Practices Elements That Are Integrated into the Texas Safe
Prison Program
Investigation
•
•
•

Protocol to guide investigations
Internal policies to promote accountability among inmates who commit
sexual violence
Specialized training for investigators and prosecutors

Victim services
•
•

Provision for crisis intervention and safe-keeping for victims
Mental health care for victims

Prevention
•
•
•
•
•

Technologies and equipment for reducing opportunities to victimize
Classifications systems that identify potential aggressors and victims
Expanded use of specialized housing and placement options
Focused inmate orientation and education
Specialized pre-service training and ongoing training for correctional staff

Record Keeping
•
•

Computerized database to track allegations of sexual assaults that can
serve as a reporting system
Computerized classification codes that can alert staff during classification
decisions as to potential perpetrators and victims

- 26 -

The JFA Institute

IV.

Research Design and Method

Overall Approach

The study is designed to provide a comprehensive analysis of sexual
assaults that have been reported officially within the Texas prison system
between January 1, 2002 and August 31, 2005. Our interest is to provide a rich
description of these events and then make comparisons to other prisoner
populations to better understand what types of prisoners and under what
conditions sexual assaults are most likely to occur. From this analysis, we are
hopeful that better risk assessment tools can be developed and better
management techniques deployed that will serve to reduce these crimes.
However, as we will discuss later, the incidence of alleged sexual assaults is
relatively low making it difficult to have statistically significant variance to identify
risk factors.
The research design has two major components: (a) the quantitative
analysis of approximately 1,938 officially reported incidents of sexual assault or
rape from January 1, 2002 through August 31, 2005 and, (b) observations of the
prison environment at seven selected prisons. The sections below explain this
methodology in more detail.
B.

Quantitative Analysis of Officially Reported Sexual Assaults

The quantitative analysis consists of analyzing 1,938 cases of alleged
sexual assault reported to the Safe Prison Management Office (SPMO) between
January 1, 2002 and August 31 2005. As noted earlier, this office established a
database on all reported sexual assaults that formed the basis of the statistical
analysis. The office developed its own stand-alone Access database with details
pertaining to the alleged incidents. In addition, the database contains race, sex,
age, custody level, gang affiliation, and prison identification number for victims
and alleged assailants. It also contains investigative information showing
whether the department treated the allegation as a disciplinary case or otherwise.
TDCJ also provided a file of all prisoners on hand in a TDCJ facility as of
June 30, 2005. The research team matched the information in the SPMO
database with this file in order to make prisoner comparisons. In addition to
demographic data, the on hand file contained offense of record, sentence length,
and prison record information for the entire TDCJ population. Such data will help
assess if there are any prisoner attributes that are associated with being either a
victim or perpetrator of such incidents. Some preliminary analysis provided by
the TDCJ suggests that among the victims and victimizers, the former are
disproportionately younger, white, not affiliated with a prison gang, slight build by
virtue of their height and weight, and serving their first prison term. Conversely,
the alleged predators are disproportionately older, Black or Hispanic, affiliated
with a prison gang, and serving their second or third prison term. Such analysis

- 27 -

The JFA Institute

will be used for purposes of determining if a risk assessment tool can be devised
to identify those prisoners who are more likely to become either a victim or
predator. 45
In addition, the research team also reviewed unit specific aggregate data
related to incidents reported to the Emergency Action Center (EAC). Each prison
unit is required to report to the EAC major inmate misbehaviors like assaults,
escapes, and disturbances. It is also required to report major use of force
incidents and alleged staff misbehaviors. This information is used here to
determine if facilities with high numbers of alleged sexual assaults also have high
levels of other violent inmate incidents.
C.

Supplemental Analysis of Prison Setting and Environmental Factors

The second component of the research is oriented at attempting to
understand the operational practices that can impact, both positively and
negatively, the prevention of sexual assaults. As noted earlier, efforts to reduce,
prevent and eventually eliminate the incidence of sexual assault in a correctional
environment must involve the identification and understanding of the
management and operational practices surrounding such events. This type of
assessment may enhance the department’s understanding of how its operational
practices can be modified in order to improve its ability to deter these incidents in
the future. It may also contribute to the national understanding and knowledge of
the type of operational practices that need to be examined, modified, and/or
replicated in order to reduce incidents of sexual assaults within correctional
environments.
To accomplish the above, seven facilities were selected for site visits. The
site visit included:
•

Review of location of alleged incidents and issues related to location and
inspection of the physical plant

•

Review of operational practices in place at the time of the reported
incident

•

Interviews with supervisory staff of the institution

The questions framing the qualitative investigation were the following:
•

Is the physical plant configuration of the specific unit a significant factor in
terms of the ability of staff to prevent assaultive behavior, observe
prisoners at all times, provide proper supervision, etc.?

Texas Department of Criminal Justice (2004). Safe Prisons Program Update

- 28 -

The JFA Institute

•

Were there any “breakdowns” in the procedures controlling the movement
of prisoners?

•

What was the nature of the staffing and personnel practices in place at the
time of the incidents, and were these practices a contributing factor to the
alleged incident?

•

Did the classification and placement decisions of prisoners involved in the
incidents allow for the proper separation of prisoners?

•

Were prisoner screening procedures in place to identify both sexual
predators and their potential victims?

•

Were prisoners screened for potential vulnerabilities or predator
tendencies, and was this information documented properly for use in
placement and management decisions relative to the prisoners?

•

Did the training curriculum of the institution provide sufficient information
relative to the prevention and investigation of sexual assaults?

•

Were the investigative policies, procedures and practices adequate to
ensure the proper investigation of any reported allegation of sexual
assault?

•

Were practices in place to record allegations, both substantiated and
unsubstantiated so that prisoners with histories or tendencies of predatory
behavior or vulnerabilities can be identified and closely monitored when
appropriate?

•

Were prisoners who have been identified as at risk for sexual victimization
monitored and counseled when appropriate?

•

What were the policies and practices of the institution relative to reporting
the occurrence or allegation of sexual assault to designated law
enforcement personnel?

Table 5 below shows the characteristics and location of the prison units
selected for the site visits and the number of incidents of alleged sexual assaults
in each unit in calendar year 2004, the last year for which 12 months of data was
available. The prison units selected for the site visit were selected in consultation
with the director of the TDCJ Correctional Institutions Division, the Deputy
Director of TDCJ and the manager of the SPMO. As will be shown later, there is
variance in the number and rates of allegations being made by prisoners in
Texas by facilities or units. Some of this variance may be related to the fact that
some facilities are intended to house prisoners either for protective custody or

- 29 -

The JFA Institute

disciplinary administrative reasons. But even within the same unit types, there
are differing numbers and rates of reported sexual assaults.
The units were selected to represent facilities of different configurations,
capacity, functions and different rates of incidence of alleged sexual assaults.
The average alleged sexual assault incident rate reported to BJS in 2004 for
Texas was 0.39 per 100 prisoners. 46 Five of the units selected for the site visits
have higher than the system wide average incidence rate while two have lower
than the average. It is important to note that with over 100 prison units in Texas,
the selection of the units was not intended to provide a random sample of facility
types representing the distribution of facilities, population, or incidents. The
scope of the study did not allow for this methodology.

Table 3: Characteristics and Location of Units Selected for Site Visits and Number
and Rate of Incidents of Alleged Sexual Assaults in Each Unit in Calendar 2004

Date Unit
Established

Population
as of
6/30/05

2004 Incidents
of Alleged
Sexual
Assaults
(Rate Per 100
Inmates)

1917

1,805
Males

9
(0.50)

1990

2,882
Males

36
(1.25)

1995

1,197
Females

16
(1.34)

G1 – G5, Administrative
Segregation

1988

483
Co-gender

4
(0.83)

Mental Health, co-located
with Hodge Unit

Huntsville,
Texas

1994

2,060
Male

2
(.097)

G1, G2, G4,
Administrative
Segregation, Intake

Jester IV
(Psychiatric
Facility)

Richmond,
Texas
(Southwest
of Houston)

1993

487
Male

9
(1.85)

Mental Health, co-located
with Jester I, Jester III
and Vance Unit)

Polunsky
(Prison)

Livingston,
East Texas

1993

2,845
Male

9
(0.32)

G1 – G5, Death Row,
Administrative
Segregation

Unit

Location

Darrington
(Prison)

Rosharon,
East Texas

Hughes
(Prison)
Murray
(Prison)
Skyview
(Psychiatric
Facility)
Holliday
(Transfer
Facility)

Gatesville,
Central
Texas
Gatesville,
Central
Texas
Rusk,
North East
Texas

Custody/ Other

G1-G5, Administrative
Segregation, Outside
Trusty Camp with 321
bed capacity
G1 – G5, Safekeeping,
Administrative
Segregation

Source: Texas Department of Criminal Justice web site,
www.tdcj.state.tx.us/stat/unitdirectory/da.htm

Ibid., Beck, Allen J. and Hughes, Timothy A. (2005), Appendix table 1a and 2 a.

- 30 -

The JFA Institute

V. Analysis of Alleged Prisoner on Prisoner Sexual Assaults
The statistical analysis has been organized into three main sections: (a)
analysis of the alleged prisoner on prisoner incidents; (b) analysis of the
prisoners implicated in the allegations; and, (c) analysis of the units where sexual
assaults were alleged, and a more general analysis of inmate-on-inmate violence
across units.
It is important to note that much of the analysis from the SPMO database
relied on a description field (“motive”) as opposed to other available columns.
Many inconsistencies in the reporting of incidents by each facility led the
research team to examine each case by hand and code the information for
analytical purposes. The unit where the incident was alleged to have occurred
was confirmed in the “motive” field, as was the date of the incident, and the
location within the facility. In many cases, the columns that were to contain these
details instead contained where and when the allegation was made. The
research team also used the “motive” field to identify whether injuries were
sustained, whether a forensic exam or rape kit was completed, and whether the
incident met the BJA criteria for nonconsensual sexual activity or abusive sexual
contact. As will be discussed later, the definitions and data collection procedures
need to be improved for more accurate reporting of sexual assaults.
A. Overview of Prisoner on Prisoner Incidents
One of the issues we have tried to explore is the extent to which
implementation of new programs and policies may impact the number of
incidents being reported to prison officials. Texas offers such an example. Table
6 listed below summarizes the number of sexual assaults since 1993. As shown
in the table, the official sexual assault rate hovered between 1.2 and 0.6 per
1,000 inmate population until 1999 when the rate doubled. Shortly after the
passage of the Texas Safe Prison Program, the rate doubled again.
The TDCJ reported that that the reporting standards for sexual assault
were more conservative prior to 2002. Historically, there had to greater certainty
that a sexual assault had actually occurred before it would be reported by staff or
prisoners. One of the objectives of the TSPP was to accept any report of a
sexual assault or sexual abuse regardless of the credibility or preliminary
evidence surrounding the allegation.
Another thing to consider when exploring why the rates went up so dramatically
is the nature of the allegation. Several of the alleged cases of nonconsensual
sexual acts from 2002 through 2005 involved oral sex only, which often occurs
without any injuries noted. Inmates comply because of the threat of violence,
which makes it difficult to differentiate between consensual and nonconsensual
acts. Since allegations prior to 1999 required physical or visible evidence, it
would not be surprising if few allegations of oral sex were reported. Specifically,

- 31 -

The JFA Institute

in 2002, 44 of the nonconsensual sexual acts involved oral sex only, follows by
46 cases in 2003, 90 cases in 200 and 75 in 2005. But even with these
considerations of sexual behavior that were not being reported prior to 2002, it
would appear that the Texas program has increased the number of such
allegations being made by prisoners.
Table 6: Number and Rates of Alleged Prisoner on Prisoners Sexual
Assault 1993-2005
Year
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005

N
75
84
131
84
87
89
237
234
292
460
425
609
473

Rate/1,000
Inmate
Population
1.2
62,855
1.0
84,355
1.2
107,587
0.6
130,413
0.6
135,895
0.6
143,085
1.6
146,574
1.6
150,309
2.0
146,244
3.2
146,244
2.8
152,602
3.9
154,978
NA
NA

Since 2002, the TDCJ has been maintaining a detailed data base that
allows one to track the final disposition of each allegation. Specifically, the data
base provides a “case status” for each prisoner implicated in the sex assault
allegation. This field indicates whether the internal investigation into the
prisoner’s conduct is still Active, or whether TDCJ’s investigation resulted in a
Closed Case (no finding), a Disciplinary for Consensual Activity, a Unit Transfer
or placement into Safekeeping status, or whether the allegation was Sustained
for the victim and/or prisoner. This “sustained” resolution essentially indicates
whether a victim was classified as a “Sexual Victim” or whether an assailant was
classified as a “Sexual Predator” or “Potential Sexual Predator”. None of these
resolutions reflects a legal finding, but rather a disciplinary and classification
finding for purposes of prison management. Although the data in the database is
stored for each prisoner, the results have been summarized at the incident level
for presentation here. Table 7 highlights these case resolutions.

- 32 -

The JFA Institute

Table 7: TDCJ Incidents by Case Resolution Status,
January 1, 2002 through August 31, 2005
Active or
Unknown

Closed
(No finding)

Disciplinary
Received

Inmate
Transfer/
Safekeeping

Sex Assault
Sustained

Total

2002

3 (0.7%)

280 (60.9%)

8 (1.7%)

158 (34.3%)

11 (2.4%)

460

2003

5 (1.2%)

267 (62.8%)

8 (1.9%)

137 (32.2%)

8 (1.9%)

425

2004

3 (0.5%)

398 (65.4%)

24 (3.9%)

170 (27.9%)

14 (2.3%)

609

2005

0 (0.0%)

309 (69.6%)

26 (5.8%)

99 (22.3%)

10 (2.3%)

444

Total

11 (0.6%)

1254 (64.7%)

66 (3.4%)

564 (29.1%)

43 (2.2%)

1938

Report
Year

The number of sustained incidents has increased slightly since TDCJ
began tracking the allegations in 2002, but the rates are consistent across years,
with the exception of a lower rate in 2003. For the analysis in this section, the
research team summarizes findings for all alleged assaults, as well as sustained
allegations (2.2% overall) and allegations where the alleged victim was
transferred to another unit or placed in Safekeeping (29.1% overall). For these
latter cases “there was not enough evidence to support an occurrence of sexual
assault” according to TDCJ’s definition, but the victim presented certain
vulnerabilities or the alleged victim/assailant combination was found to be
problematic. In subsequent tables, this subset of cases will be referred to as
“Probable Cause” incidents.
As mentioned previously, the research team used the “motive” field to
classify allegations according to the BJA definitions of nonconsensual sexual
acts and abusive sexual contacts. Table 8 below shows this differentiation, and
illustrates an increase in the reporting of abusive sexual contacts for each
calendar year. This increase suggests that the prisoner awareness effort may be
having an impact on a prisoner’s willingness to report an act of sexual
abuse/assault in prison. Because the researchers relied on the “motive” field, the
BJA breakdown should be viewed as an approximation; not all of the descriptions
were detailed as to the nature of the sexual assault. Where there was a lack of
detail, the research team classified the event as a nonconsensual sexual act.
Given the number of allegations, and given that less than 3% are
sustained, we examined possible factors that may have affected the
department’s ability to sustain an allegation. The next set of tables illustrate
some of the difficulties faced by institutions in attempting to prove an allegation of
sex assault, which may assist in establishing management practices or
developing prisoner awareness programs in the future.
As shown in Table 9 a major reason for not being able to sustain an
allegation is the delay in a prisoner reporting the incident to prison officials. This
table shows the lapse in time from when an incident allegedly occurred to when
that incident was reported to TDCJ officials, as confirmed by the narrative in the

- 33 -

The JFA Institute
Table 8: TDCJ Sex Assault Allegations Classified by BJA Definition,
January 1, 2002 through August 31, 2005
All Incidents
Report
Year

Sustained Incidents

Probable Cause Incidents

Nonconsensual
Sexual
Act

Abusive
Sexual
Contact

Nonconsensual
Sexual
Act

Abusive
Sexual
Contact

Nonconsensual
Sexual
Act

Abusive
Sexual
Contact

2002

427

149

2003

412

134

2004

543

151

2005

349

Total

1,731

207

517

Table 9: Time Lapse from Incident Occur Date to Incident Report Date,
January 1, 2002 through August 31, 2005
2002
All Incidents

Sustained Incidents

Probable Cause Incidents

Same Day

161

35.0%

72.7%

32.3%

Within 2 days

18.3%

9.1%

18.4%

3 to 7 days

16.5%

0.0%

17.1%

8 to 30 days

15.4%

9.1%

21.5%

1 month +

14.8%

9.1%

10.8%

Total

460

158

2003
All Incidents

Sustained Incidents

Probable Cause Incidents

Same Day

103

24.2%

25.0%

22.6%

Within 2 days

101

23.8%

25.0%

23.4%

3 to 7 days

17.4%

25.0%

17.5%

8 to 30 days

16.9%

12.5%

16.8%

1 month +

17.6%

12.5%

19.7%

Total

425

137

2004
All Incidents

Sustained Incidents

Probable Cause Incidents

Same Day

155

25.5%

50.0%

24.1%

Within 2 days

137

22.5%

21.4%

24.1%

3 to 7 days

14.9%

7.1%

14.1%

8 to 30 days

14.9%

14.3%

14.7%

1 month +

135

22.2%

7.1%

22.9%

Total

609

- 34 -

170

The JFA Institute

2005
All Incidents

Sustained Incidents

Probable Cause Incidents

Same Day

162

36.5%

50.0%

33.3%

Within 2 days

20.7%

10.0%

21.2%

3 to 7 days

10.8%

10.0%

13.1%

8 to 30 days

10.6%

20.0%

9.1%

1 month +

21.4%

10.0%

23.2%

Total

444

Total
All Incidents

Sustained Incidents

Probable Cause Incidents

Same Day

581

30.0%

51.2%

156

27.7%

Within 2 days

414

21.4%

16.3%

123

21.8%

3 to 7 days

289

14.9%

9.3%

15.6%

8 to 30 days

281

14.5%

14.0%

16.1%

1 month +

373

19.2%

9.3%

106

18.8%

Total

1938

564

“motive” field. A majority of the sustained cases were reported on the
same day as the assault, or within 2 days of the assault. However, the time
lapse for probable cause cases resembles the time lapse for all reported
incidents.
Other factors affecting whether a case can be sustained include the
location where the assault was alleged to have occurred, as shown in Table 10
below. TDCJ’s Safe Prisons Program policy includes interviews with staff and
inmates who were possible witnesses to the incident. However, it appears that
there are few witnesses to an incident in the majority of reported cases of sex
assault. Approximately two-thirds of all reported cases of sexual assault
allegedly occurred in an inmate cell, followed by assaults in the shower or
bathroom area, assaults in dorms, and finally assaults occurring in common
areas such as dayrooms, work places, cafeterias, etc. The locations for
sustained assaults is more pronounced, with three-quarters of confirmed cases
occurring in cells, and nearly 12% of confirmed cases occurring in the shower or
bathroom facilities. Section VI of this report discusses the configurations of
various institution types, and why some facilities with certain cell designs might
contribute to the incidence of sex assaults while others might reduce the
incidence.
The issue of evidence collection is addressed in Tables 11 and 12. TDCJ
determines whether to administer a rape kit or perform a forensic exam based on
the nature of the allegation and on whether evidence is likely to be available. For
cases of fondling or abusive sexual contact, such evidence will not be available

- 35 -

The JFA Institute

to collect. For cases reported several days or weeks after the incident, this
evidence will not be available either.
Note that just over 50% of the sustained cases included forensic evidence
from a rape kit or a forensic exam. On the other hand, rape kits and forensic
exams were performed in only 20% of all alleged sexual assaults. For the
remaining 80% of all alleged assaults, the reasons why these tests were not
performed are listed in Table 11 below.
Time lapse from the incident occurring to the report date is the reason
cited most frequently for not completing a rape kit or conducting a forensic exam,
followed by nature of the allegation.
As part of the Safe Prisons Program protocol, the unit medical department is
charged with examining a prisoner immediately after an assault is reported.
Table 13 shows whether the medical exams revealed injuries to the victims or
assailants. (Injuries Unrelated to Allegation refer to cases where injuries were
found, but they were determined to have resulted from some other altercation
with a prisoner, or from self-inflicted wounds.) The data show that injuries are
noted in less than a quarter of all sustained allegations, and in only 10% of all
alleged assaults. Such a finding seems to confirm the difficulty in distinguishing
between consensual and nonconsensual activity.

Table 10: Location within the Facility of the Alleged Sexual Assault,
January 1, 2002 through August 31, 2005
Location

All Incidents

Sustained Incidents

Probable Cause Incidents

Cell

1298

67.0%

74.4%

382

67.7%

Dorm

152

7.8%

4.7%

6.0%

Shower/Bathroom

170

8.8%

11.6%

9.9%

Common Area

127

6.6%

4.7%

5.7%

Unspecified

191

9.9%

4.7%

10.6%

Total

1938

- 36 -

564

The JFA Institute

Table 11: Whether a Rape Kit or Forensic Exam Was Performed,
January 1, 2002 through August 31, 2005
All Incidents

Sustained Incidents

Probable Cause Incidents

Rape Kit

220

11.4%

25.6%

12.6%

Forensic Exam

175

9.0%

25.6%

10.3%

None

1314

67.8%

46.5%

371

65.8%

Unspecified

229

11.8%

2.3%

11.3%

Total

1938

564

Table 12: Reasons Why a Rape Kit or Forensic Exam Was Not Performed,
January 1, 2002 through August 31, 2005
Reason

All Incidents

Sustained Incidents

Probable Cause Incidents

Time Lapse

903

58.5%

71.4%

274

63.0%

Nature of Incident

226

14.6%

9.5%

14.3%

Other Reason

3.6%

4.8%

2.1%

Victim Refused

3.5%

0.0%

2.3%

Unspecified

304

19.7%

14.3%

18.4%

Total

1543

435

Table 13: Whether the Medical Exam Revealed Injuries to Victim or
Assailant, January 1, 2002 through August 31, 2005
Injury Level

All Incidents

Sustained Incidents

Probable Cause Incidents

Injuries Found

200

10.3%

23.3%

13.1%

Injuries Unrelated

2.6%

2.3%

3.0%

No Injuries Found

935

48.2%

23.3%

252

44.7%

Exam Not Done

4.3%

2.3%

2.7%

Unspecified

669

34.5%

48.8%

206

36.5%

Total

1938

564

Alleged Victims, Alleged Assailants, and all Other TDCJ Inmates

This next section presents findings from a demographic analysis and from
other differentiations among the inmate populations. The first four tables
continue to rely on the SPMO database for a comparison of prisoners implicated
in sex assault allegations.
After that, the analysis expands to include
comparisons with the TDCJ prisoner population on hand as of June 30, 2005.

- 37 -

The JFA Institute
Table 14 reports on the racial composition of sexual assaults for each incident,
as reported in the SPMO database. “Mixed” refers to a situation in which more than one
inmate is implicated as an assailant in the allegation, and the racial composition of the
assailants spans more than one racial group. “Other” typically refers to Asian or Native
American inmates.

White inmates are attacked more frequently than any other race, by all the
racial groups listed above. Nearly 60% of sustained incidents involved a white
victim, with 42% coming from black assailants and 9% coming from white
assailants followed by 7% Hispanic assailants. Close to 50% of probable cause
incidents involved a white victim. Moreover, two-thirds (67.5%) of the sustained
incidents involved black assailants, while 18.7% involved Hispanic assailants and
11.6% involved white assailants. These results are consistent with the findings
from other research efforts, notably Carroll and Dumond.
Additional demographics purported to be a factor in sex assaults include
the age of victims and assailants. Table 15 presents the findings from this
research effort with regard to the issue of age, showing whether the alleged
assailant was older, younger, or approximately the same age as the victim. For
this analysis, the terms “older” and “younger” included cases where prisoners’
age differed by more than a year in either direction, while prisoners of
approximately the same age were within a year of each other. “Unknown” refers
to incidents where the victim was unable to identify specific assailants. Twothirds (67.4%) of the sustained incidents involved assailants who were at least a
year older than their victims. This finding also seems to confirm the stereotype of
the prison rape victim.
While there is little difference in average age among all alleged victims
and assailants, the results are more pronounced for sustained incidents and
probable cause incidents (Table 16). The average age of victims in sustained
cases is 3 years younger than the assailants, while the average age of victims in
probable cause cases is 4 years younger than assailants. The average age of
assailants is not particularly old, possibly reflecting the alleged gang affiliation of
these prisoners.
Table 17 addresses whether the victims are disproportionately mentally ill
or intellectually impaired at the time the incident was reported to officials. The
data has been divided according to whether the victim was mentally
ill/intellectually impaired, whether the assailant was mentally ill/intellectually
impaired, or whether both victim and assailant were mentally ill/intellectually
impaired.

- 38 -

The JFA Institute

Table 14: Racial Composition of Victims and Assailants in Sex Assault
Allegations, January 1, 2002 through August 31, 2005
Racial Composition

All Incidents

Sustained Incidents

Probable Cause Incidents

Black on Black

272

14.0%

11.6%

10.8%

Black on Hispanic

167

8.6%

14.0%

8.3%

Black on Other

0.3%

0.0%

0.5%

Black on White

449

23.2%

41.9%

174

30.9%

Hispanic on Black

3.2%

4.7%

3.0%

Hispanic on Hispanic

3.8%

7.0%

3.7%

Hispanic on Other

0.1%

0.0%

0.2%

Hispanic on White

136

7.0%

10.1%

Mixed on Black

0.6%

0.0%

0.5%

Mixed on Hispanic

0.6%

0.0%

0.5%

Mixed on Other

0.1%

0.0%

Mixed on White

1.8%

2.3%

3.0%

Other on White

0.1%

0.0%

White on Black

2.5%

0.0%

1.1%

White on Hispanic

2.4%

2.3%

2.1%

White on Other

0.3%

0.0%

White on White

131

6.8%

9.3%

6.4%

Unspecified Asslnt

480

24.8%

0.0%

106

18.8%

Total

1938

564

Table 15: Age Variance of Victims and Assailants in Sex Assault
Allegations, January 1, 2002 through August 31, 2005
Assailant Age

All Incidents

Sustained Incidents

Probable Cause Incidents

Assailant Older

724

37.4%

67.4%

231

41.0%

Assailant Younger

500

25.8%

20.9%

148

26.2%

Approx Same Age

232

12.0%

9.3%

14.0%

Assailant Unknown

482

24.9%

2.3%

106

18.8%

Total

1938

- 39 -

564

The JFA Institute

Table 16: Average Age of Victims and Assailants in Sex Assault
Allegations, January 1, 2002 through August 31, 2005
All Incidents

Sustained Incidents

Probable Cause Incidents

Victims

32.1 years

29.7 years

30.8 years

Assailants

33.1 years

35.4 years

Table 17: Whether Victim or Assailant Was in a Custody Class of Mentally
Ill or Intellectually Impaired, January 1, 2002 through August 31, 2005
Mental Illness Level

All Incidents

Sustained Incidents

Probable Cause Incidents

Not classified MI or II

1713

88.4%

93.0%

532

94.3%

MI victim

143

7.4%

2.3%

4.3%

MI assailant

0.2%

2.3%

0.2%

MI victim & assailant

2.6%

2.3%

1.1%

II victim

0.4%

0.0%

II assailant

0.1%

0.0%

II victim & assailant

0.9%

0.0%

0.2%

Total

1938

564

Though 12% of the allegations involved a mentally ill or intellectually
impaired prisoner, only 7% of the sustained allegations were found to involve this
special population of prisoners. Nonetheless, the findings show that three of the
43 sustained cases involved mentally ill inmates, a significant result given the low
proportion of prisoners classified as mentally ill among the entire TDCJ
population. None of the sustained cases involved an intellectually impaired
prisoner.
To examine this issue more in more detail, Table 18 provides a complete
comparison of prisoner custody class codes for alleged victims and assailants at
the time the incident was reported with all other TDCJ inmates (the comparison
group) who were on-hand in a TDCJ facility on June 30, 205. The custody class
data has been aggregated to simplify its presentation in this report. Note that the
table above uses the victims and assailants from the SPMO sex assaults
database, where a single prisoner may be implicated in more than one incident.
Because this analysis evaluated custody codes at the time of the reported
incident, it was important to include all incidents in the analysis. Further
comparisons between victims, assailants, and the rest of the TDCJ population
will select unique prisoners from the SPMO sex assaults database.
.

- 40 -

The JFA Institute

Table 18: Custody Class of Victims and Assailants at the Time of the
Alleged Incident, and Custody Class of All Other TDCJ Inmates as of June
30, 2005
Custody Level

Victims

Assailants

All Other Inmates

Prison – Minimum

598

30.7%

611

35.2%

99,951

67.0%

Prison – Medium

365

18.8%

333

19.2%

9,489

6.4%

Prison – Close/Max

279

14.3%

264

15.2%

4,012

2.7%

State Jail – Low Risk

0.2%

1,115

0.7%

State Jail – Med Risk

4.2%

5.3%

12,610

8.5%

State Jail – High Risk

0.2%

277

0.2%

Safekeeping Pop

295

15.2%

287

16.5%

2,505

1.7%

Admin Segregation

2.6%

1.3%

9,389

6.3%

Mentally Ill/Impaired

220

11.3%

4.9%

2,394

1.6%

Other Custody Class

2.5%

2.0%

7,444

5.0%

1,946

1.3%

1,736

1.1%

149,186

97.6%

Total

The significant finding here is the disproportionate number of alleged
sexual assaults occurring among the safe-keeping and mentally ill/impaired
prisoners. One would have expected a lower rate among the safe-keeping
population since that is one major reason why they are so classified and housed.
It would appear that a number of these prisoners may become involved in
aggressive sexual behavior after being placed in a safe-keeping housing unit.
But the other major finding to be gleaned from Table 18 is that an
extremely small percentage of the daily population is classified as a “victim” or
“assailant”. Such a small “base rate” means that it is virtually impossible from a
statistical perspective to develop a purely statistical profile of potential victims
and assailants. We will discuss in greater detail the implications of this finding.
The following tables compare these three categories of prisoners by race,
offense of record, sentence length, county of conviction, gang affiliation, and time
served from most recent admission date. In many cases, special attention has
been paid to the issue of gender in relation to prison sexual violence.
Accordingly, many of the tables will also separate out the results for males and
females in order to make comparisons along those lines.
As noted earlier in this report, white inmates are disproportionately the
victims in sex assaults, while black inmates are disproportionately the assailants.
Table 19 indicates that Hispanic inmates are under-represented in the sex
assault allegations as compared to the overall Hispanic population within TDCJ.
Though not shown in the table above, the breakdown by race in sustained
assaults is even more pronounced. For the sustained cases, 66.7% of the

- 41 -

The JFA Institute

victims were white, followed by 20.0% Hispanic victims, and 13.3% black victims.
For assailants, 64.7% were black, followed by 20.6% Hispanic and 14.7% white.
Table 19: Race of Victims, Assailants, and All Other TDCJ Inmates on Hand
as of June 30, 2005
MALES
Victims

Assailants

All Other Inmates

Black

285

25.8%

668

60.1%

52,342

38.1%

Hispanic

215

19.5%

248

22.3%

41,974

30.6%

White

601

54.4%

194

17.5%

42,322

30.8%

Other

0.3%

0.1%

678

0.5%

Total

1,105

0.8%

1,111

0.8%

137,316

98.4%

FEMALES
Victims

Assailants

All Other Inmates

Black

39.5%

60.0%

4,517

38.1%

Hispanic

18.4%

18.7%

2,175

18.3%

White

42.1%

21.3%

5,121

43.1%

Other

0.0%

0.5%

Total

0.6%

11,870

98.8%

Table 20 suggests that the prisoner’s offense of record is also a factor in
predicting the inmate’s behavior while incarcerated. The table categorizes
prisoners according to Violent, Sex, Property, Drug, DWI, and Other offenses.
For this analysis, sex assaults have been separated out from the Violent
category. For males in particular, assailants are more likely to have violent
offenses of record than the TDCJ population overall, while victims are more likely
than the TDCJ population overall to have a conviction for a sex assault offense or
a non-assaultive sex offense. For females, the data simply suggests that victims
and assailants in sex assault allegations are more likely to have violent criminal
histories than the overall female population in TDCJ. The results below seem to
confirm other studies, suggesting that male victims are more likely to have a
history of sex offenses.

- 42 -

The JFA Institute

Table 20: Offense Category of Victims, Assailants, and All Other TDCJ
Inmates on Hand as of June 30, 2005
MALES
Victims

Assailants

All Other Inmates

Violent

387

35.0%

565

50.9%

50,585

36.8%

Sex Assault

290

26.2%

185

16.7%

17,359

12.6%

Sex (NonAssaultive)

7.0%

3.2%

5,301

3.9%

Property

223

20.2%

207

18.6%

24,944

18.2%

Drug

6.5%

7.6%

26,315

19.2%

DWI

1.2%

0.3%

5,578

4.1%

Other

3.9%

2.8%

7,234

5.3%

Total

1,105

0.8%

1,111

0.8%

137,316

99.4%

FEMALES
Victims

Assailants

All Other Inmates

Violent

48.7%

45.3%

3,322

28.0%

Sex Assault

2.6%

4.0%

211

1.8%

Sex (NonAssaultive)

1.1%

1.3%

0.5%

Property

17.1%

21.3%

2,930

24.7%

Drug

21.1%

24.0%

4,015

33.8%

DWI

0.0%

2.7%

342

2.9%

Other

9.2%

1.3%

988

8.3%

Total

0.6%

11,870

98.8%

As noted in the literature review Dumond also suggested that gang
affiliation played a role in victimization. Table 21 below presents the confirmed or
suspected gang affiliation for prisoners in the SPMO database. Gang affiliation
information was not available for all of the TDCJ inmate population. As the table
illustrates, while there may be some violence among gangs in sex assault
allegations, the vast majority of victims (nearly 97%) are not associated with any
gang. The table below shows this breakdown for sustained and probable cause
incidents. Gang affiliation is highest for assailants whose alleged sexual
misconduct was sustained by TDCJ’s internal review, with 20.5% of all sustained
assailants affiliated with a gang.

- 43 -

The JFA Institute

Table 21: Gang Affiliation of Victims and Assailants from January 1, 2002
through August 31, 2005
Affiliation
No gang affiliation
Gang affiliation
Total

Victims

Assailants

1,879

96.6%

1,493

86.0%

3.4%

243

14.0%

1,946

1,736

Specific Gang Breakdown
Crips

31.3%

37.0%

Bloods

8.9%

16.9%

Mandingo Warriors

3.0%

10.7%

Mexican Mafia

13.4%

4.5%

Other affiliations

43.4%

30.9%

Total

243

Table 22: Gang Affiliation of Victims and Assailants from January 1,
2002 through August 31, 2005
Affiliation
No gang affiliation
Gang affiliation
Total

Sustained Incidents

Probable Cause Incidents

Victims

Assailants

Victims

Assailants

33 (94.3%)

35 (79.5%)

547 (96.5%)

25 (89.3%)

2 (5.7%)

9 (20.5%)

20 (3.5%)

3 (10.7%)

567

Table 23 presents the average time served for victims, assailants, and all
other TDCJ inmates. For this analysis, time served refers to amount of time
incarcerated in TDCJ as of the prisoner’s most recent receive date. The results
might be somewhat impacted for prisoners who have been released on parole
prior to the last receive date. However, the findings here confirm other research
efforts: assailants have served a longer time in prison than victims.
Unfortunately, we were not able to distinguish in the data whether the victims or
assailants are serving their first sentences or subsequent sentences.
The table shows that assailants in sex assault allegations tend to have
been incarcerated for longer periods in TDCJ than victims, even for assailants
whose alleged conduct was not able to be sustained by an internal review. On
average, both victims and assailants tend to have served longer in TDCJ than all
other TDCJ inmates. This may be related to the fact that TDCJ also houses
state jail prisoners, operates SAFPF facilities, and in general deals with a much
wider group of prisoners than those implicated in sex assault allegations.

- 44 -

The JFA Institute

Table 23: Average Time Served for Victims, Assailants, and All Other TDCJ
Inmates on Hand as of June 30, 2005
Average Time Served

Victims

Assailants

All Other Inmates

All Prisoners

5.6 years

8.1 years

4.1 years

Sustained Cases

5.8 years

8.4 years

n/a

Probable Cause Cases

5.5 years

6.4 years

n/a

C. Unit Analysis
A brief examination into the units is provided in this section, with more
descriptive analysis to be provided in Section VI of this report. For all tables in
this section, the results refer to units where the assaults allegedly took place, as
opposed to where the incidents were reported. Table 24 shows the top ten units
with the highest number of sex assault allegations, in descending order. All of
these institutions listed above share the following characteristics: They are all
male prison facilities operated by TDCJ with capacities from 2,800 to 3,700
inmates. It is therefore not surprising that these facilities generated the highest
number of allegations.
Units with the highest rate of allegations per 100 inmates are provided in
Table 25 below. The data has been sorted in descending order based on an
average of the rates for 2002, 2003, and 2004 – the years with complete data
available. Rate results from 2005 are skewed downward because only 8 months
of data was available.
Table 24: Top Ten Units Where Sex Assault Allegations Were Alleged to
Have Occurred from January 1, 2002 through August 31, 2005
Units

2002

2003

2004

2005

Total

Allred

142

Telford

120

Hughes

117

Stiles

Clements

Michael

Robertson

Connally

McConnell

Beto

All Other Units

231

195

326

265

1,017

460

425

609

444

1,938

Total

- 45 -

The JFA Institute

Table 25: Top Ten Units by Incidence Rate Where Sex Assault Allegations
Were Alleged to Have Occurred from
January 1, 2002 through August 31, 2005
Units

Population

2002

2003

2004

2005

Total

Jester IV

487

2.053

1.232

1.848

0.411

1.711

Skyview

483

1.863

1.035

0.828

1.242

Hughes

2,882

0.937

1.180

1.249

0.694

1.122

Telford

2,806

1.354

0.998

0.927

0.998

1.093

Allred

3,633

0.963

1.046

0.853

1.018

983

0.916

1.221

0.712

0.610

0.949

Stiles

2,861

1.014

0.664

1.049

0.559

0.909

McConnell

2,847

0.738

0.702

0.773

0.281

0.738

Robertson

2,841

0.634

0.774

0.598

0.727

Connally

2,836

0.494

0.705

0.952

0.388

0.717

Montford

Population figures as of June 30, 2005.

Four of the institutions from Table 24 are again listed in Table 25.
However, with this rate-based perspective, we find that three of the top ten
institutions in terms of rates of allegations were alleged to have occurred in
psychiatric facilities: Jester IV, Skyview, and Montford. In part, this finding
reflects the lower population than the other seven institutions. However, the
population housed in psychiatric facilities may be more likely to be a victim or
assailant in such allegations, or they may be more likely to allege such an
incident given the medications and nature of mental illnesses. Section VI will
explore this issue in more detail.
While the above tables present findings for all the alleged incidents, Table
26 below lists the 23 units where incidents have been sustained through internal
investigation over the past four years. Only 43 of the 1,938 allegations were
sustained, and only one of these incidents has been sustained at a psychiatric
facility during the study period.

- 46 -

The JFA Institute

Table 26: All Units Where Sex Assault Allegations Were Sustained by
Calendar Year from January 1, 2002 through August 31, 2005
Operated By/
Type of Facility/
Gender

2002

2003

2004

2005

Total

Allred

TDCJ/Prison/Male

Lewis

TDCJ/Prison/Male

Hughes

TDCJ/Prison/Male

McConnell

TDCJ/Prison/Male

Robertson

TDCJ/Prison/Male

Stiles

TDCJ/Prison/Male

Eastham

TDCJ/Prison/Male

Michael

TDCJ/Prison/Male

Wynne

TDCJ/Prison/Male

Connally

TDCJ/Prison/Male

Daniel

TDCJ/Prison/Male

Estes

Private/Prison/Male

Ferguson

TDCJ/Prison/Male

Jester IV

TDCJ/Psychiatric/Male

Lychner

TDCJ/State Jail/Male

Middleton

TDCJ/Transfer/Male

Murray

TDCJ/Prison/Female

Plane

TDCJ/State Jail/Female

Polunsky

TDCJ/Prison/Male

Roach

TDCJ/Prison/Male

Scott

TDCJ/Prison/Male

Smith

TDCJ/Prison/Male

Wallace

TDCJ/Prison/Male

Units

Total

While there is variation in the type of institutions where incidents were
sustained, the majority of cases (86%) were sustained at large, male prisons
operated by TDCJ. Only one incident was sustained at a privately operated
facility. Two incidents were sustained at TDCJ-operated State Jail facilities, one
housing females and the other housing males. A total of two incidents were
sustained among female prisoners. One incident was sustained at a transfer
facility. Finally, as noted earlier, one incident was sustained at a psychiatric
facility.

- 47 -

The JFA Institute

It has been suggested that units where sex assault allegations occur are
also units where other incidences of violence are occurring, and that sex assaults
can be viewed as another example of violence in prisons. TDCJ captures in its
Emergency Action Center database serious prisoner assaults, prisoner assaults
involving weapons, and major uses of force, among other measures of violence.
Table 27 presents an analysis of these incidents for calendar years 2002, 2003,
and 2004. Each year provides a summary of serious prisoner assaults, prisoner
assaults with weapons, major uses of force, and sex assault allegations. An
average of each year’s rates of major uses of force (MUF) per 1,000 inmates was
calculated to mitigate the effects of variation in unit capacity. The 10 units with
the highest MUF rates are presented for discussion.
Once again, the three psychiatric facilities are represented in this table, as
well as a facility housing mentally retarded inmates (MROP). The female facility
listed above houses females on death row, as well as all the standard G1-G5
custody levels and prisoners in administrative segregation. The other five
facilities house male prisoners of all custody levels. All of the facilities listed
above are operated by TDCJ.

Table 27: Top Ten Units by MUF Rate Where Sex Assault Allegations Were
Alleged to Have Occurred from
January 1, 2002 through August 31, 2005
Units

Unit Type

Pop

Gender

2002

2003

2004

Avg MUF

Jester IV

Psychiatric

487

Male

107

106

186.86

Skyview

Psychiatric

483

Co-gender

149.76

Connally

Prison

2,836

Male

292

325

353

92.26

Clements

Prison

3,628

Male

377

301

559

95.37

Montford

Psychiatric

983

Male

131

100

89.52

Lewis

Prison

2,163

Male

207

200

257

88.61

Hodge

MROP

821

Male

77.55

Robertson

Prison

2,841

Male

261

262

237

76.97

Mt. View

Prison

604

Female

76.16

Darrington

Prison

1,805

Male

131

139

185

75.35

Population figures as of June 30, 2005.

- 48 -

The JFA Institute

Summary of Findings

Because nearly four years of sex assault allegations were available for
this study, there is a higher base rate for analyzing reported assaults. In
addition, there were enough sustained cases or cases where probable cause
warranted the removal of the alleged victim from the cell assignment or facility,
that some meaningful conclusions can be drawn.
•
•
•
•
•
•
•
•
•
•
•
•
•

Incidents are much more likely to occur in a cell than in a more public
place such as a shower or dorm environment.
Incidents that are reported immediately or within a day or two have a
higher chance of being sustained.
Incidents where rape kits or forensic exams were performed also have
a higher chance of being proven.
Victims are generally younger than their assailants.
Victims are more likely white while their assailants are more likely
black or Hispanic.
Assailants have a higher likelihood of gang affiliation than victims.
Assailants come from larger cities/counties, perhaps a correlation to
the gang affiliation finding.
Assailants have served more time in prison than victims, anywhere
from two to four years longer on average.
Although the data concerning mentally ill prisoners was not extensive,
there is some evidence supporting the conclusion that mentally ill
prisoners are more at risk of assault.
Prisoners with criminal records involving violent crimes are more likely
to perpetrate assaults against prisoners with criminal records involving
sex crimes (assaultive and non-assaultive).
Prisoners with longer sentence lengths are more likely to perpetrate
assaults against prisoners with less severe sentence lengths.
Units housing the longer-term prisoners with higher custody levels
have higher numbers and rates of allegations than other facility types.
Units housing special needs populations (psychiatric and mentally
retarded) face particular challenges in managing their inmates, with
higher rates of general inmate-on-inmate violence and higher rates of
use of force, in relation to the population size of these facilities.

Process and Data Issues

Since its inception in 2002, TDCJ has made numerous enhancements to
the Access database to improve the data collection effort. For example, the
database now distinguishes between an incident report date and time and an
incident occurrence date and time. Enhancements in 2005 added fields
concerning whether an incident was accepted by OIG, whether an incident
involved fondling only (abusive sexual contacts), whether a rape kit was utilized,
and whether a prisoner was found guilty of a disciplinary violation. This
- 49 -

The JFA Institute

information will be useful for future studies, and also for reporting allegations to
federal agencies.
In addition to the enhancements already in place, TDCJ’s database could benefit
from a few additional ones:
•
•
•

•

If a rape kit was not utilized in the investigation, why not? A “drop down”
list of possible reasons would be best for analytical purposes.
Capture the type of incident using the BJA definitions for easier federal
reporting, and possibly also include a field that summarizes the nature of
the assault, including fondling, oral sex, penetration, and other options.
Capture whether a prisoner alleged coercion or extortion as part of the
assault, in order to begin differentiating between the consensual, coercive,
and predatory sexual conduct in prisons. Currently, there is no meaningful
way to distinguish between these three types of inmate behaviors. TDCJ
is able to identify predators electronically, but does not appear able to
identify cases involving extortion or coercion other than through the
narrative description.
Capture how the prisoner communicated the allegation to officials in the
institution – during a classification hearing (months after the fact), upon
intake at a different facility, in an I-60 form, in a grievance form,
communications with guards or management at the facility, or other
means.

For all of these suggestions, the quality of the data matters. It is best to
provide a list of responses from which users can choose, rather than allowing
users to enter the data in a text field. Many of the problems with the current
system involve such text fields, from gang affiliation to injuries noted, making
meaningful analysis more difficult.
Special care needs to be made in educating users of the system. In
particular, users need to be sure that they enter data related to the incident
occurrence, including date, time, facility, and custody code at the time of the
alleged assault.

- 50 -

The JFA Institute

VI.

Qualitative Analysis of Units

Site Visit Selection and Protocol

As noted earlier, efforts to reduce, prevent and eventually eliminate the
incidence of sexual assault in a correctional environment must involve the
identification and understanding of the management and operational practices
surrounding such events.
This type of assessment may enhance the
department’s understanding of how its operational practices can be modified in
order to improve its ability to deter these incidents in the future. It may also
contribute to the national understanding and knowledge of the type of operational
practices that need to be examined, modified, and/or replicated in order to
reduce incidents of sexual assaults within correctional environments.
To accomplish the above, seven facilities were selected for in-depth site
reviews. These facilities included the following:
•
•
•
•
•
•
•

Darrington – general population, administrative segregation
Jester IV – mental health unit
Holilday – transfer unit
Hughes – general population, administrative segregation, safe
keeping
Murray – female prisoners
Polunsky – general population, administrative segregation, death
row
Skyview – mental health unit

These units were selected in consultation with the director of the TDCJ
Correctional Institutions Division, the Deputy Director of TDCJ and the manager
of the SPMO. There is a variance in the function, design, capacity, custody level,
and the number and rates of allegations made by prisoners within these units.
It is important to note that with over 100 prison units in Texas, the
selection of the units was not intended to provide a random sample of facility
types representing the distribution of facilities, population, or incidents. The
scope of the study did not allow for this methodology.
The site visit to these selected facilities included a comprehensive review
of the following factors and issues:
•
•
•

Operational practices in place at the time of the reported incidents
Physical inspection of the facility in order to fully understand how and
where the incidents had occurred
Interviews with supervisory staff of the institution

- 51 -

The JFA Institute

•
•
•
•
•
•
•

•
•
•

•

The physical plant configurations and/or location of incidents in terms of
the ability of staff to prevent assaults, observe prisoners at all times, and
provide proper supervision
Review the degree that alleged perpetrators had access to the location of
the incident and, if the location was restricted, how did the perpetrator
obtained access
Review of the procedures controlling the movement of prisoners
Determine the staffing deployment and personnel practices of the facility
and if these practices could be a contributing factor to a sexual assault
Review the prisoner classification and placement processes and if these
processes insure proper separation of prisoners
Review the prisoner screening procedures in place that are intended to
identify both sexual predators and their potential victims
Review processes that are intended to insure that prisoners are screened
for potential vulnerabilities or predator tendencies and whether this
information was documented for use in placement and management
decisions relative to the prisoners
Review the training curriculum relative to the prevention and investigation
of sexual assaults and implementation of the Safe Prisons program
Review the investigative policies, procedures and practices that are
designed to ensure the proper investigation of any reported allegation of
sexual assault
Determine if practices were in place to record allegations, both
substantiated and unsubstantiated so that prisoners with histories or
tendencies of predatory behavior or vulnerabilities can be identified and
closely monitored when appropriate
Review the policies and practices of the institution relative to reporting the
occurrence or allegation of sexual assault to designated law enforcement
personnel
Observations and Findings

TDCJ policy mandates that staff has a duty to protect prisoners in their
custody and that operational practices, policies and procedures should be geared
to ensure that protection is provided to all prisoners. Prison environments are
complex organizations that are affected by a wide range of factors including the
physical design of the units, complement and deployment of staff, and training
and experience of staff, among others. The intent of this section is to summarize
the overall assessment of the ability of the TDCJ facilities to achieve maximum
effectiveness of its Safe Prison Program initiative. The following will identify
some of the key strengths and weaknesses of the organization as it relates to the
Safe Prisons Program.

- 52 -

The JFA Institute

Organizational Commitment to the Reduction of Sexual Assault

The existence of policies and procedures that mandate specific action on
the part of the organization and its staff does not always guarantee the outcomes
expected by the organization’s leadership. The fact that a policy has been
developed and issued does not necessarily result in the desired outcomes,
especially in an organization as large and geographically dispersed as the TDCJ.
It is important, therefore, to determine to what degree the entire organization,
including those at the institutional line level, have embraced and implemented the
policy requirements associated with the Safe Prisons program.
During the course of the facility visits, staff at all levels were contacted
both formally and informally to determine their knowledge and understanding of
the policy requirements, and to subjectively determine the level of commitment
by the organization to the Safe Prisons ideals. This qualitative review found:
•

Staff at all levels had a clear understanding of the expectations and goals
of the organization as it related to Safe Prisons

•

Inmates were aware of the elements of the program and the processes to
be utilized to report an allegation

•

Classification staff, Safe Prisons staff and housing unit staff appeared to
communicate well regarding issues relative to housing placement of those
who appeared vulnerable and with those who were identified as potential
predators

•

Signs were prominently posted at numerous locations within the facilities
reminding inmates of the provisions of the Safe Prison Program in relation
to reporting allegations

•

There did not appear to be any indication of “organizational indifference”
at any of the facilities visited. There may be individuals within the
organization that may ‘look the other way’ when faced with an allegation of
sexual assault, but that was not a prevalent attitude observed within the
facilities visited

•

Observations during the monthly and quarterly Safe Prison Program
meetings at the institutional level indicated that there was a high level of
communication, cooperation, and exchange of information between staff
and operational and program units, and a willingness to seek out
improvements to the processes related to the program

•

The impact of the Safe Prison program requirements were found at all
operational levels and functions within the units examined

- 53 -

The JFA Institute

•

From the intake processes at the Holliday Unit through the housing
placement decisions, transfer decisions, and program placement
recommendations at each unit, the need to be alert and aware of the
organization’s goal to eliminate sexual assault incidents was apparent.
Investigative Processes

The Safe Prisons policies detail in very specific terms the investigative
process to be followed when an allegation of sexual assault comes to the
attention of staff within the TDCJ. There is a clear protocol that guides the
investigators through the initial review of the allegation. It was reported that each
of the Safe Prisons staff have completed the required training as specified and
many had prior investigative experience either in public law enforcement or as
the designated Strategic Threat Group officer. As noted previously, there was
not an ability to review investigative files so there was no ability to conduct an
assessment of the quality of the individual investigations. However, the process
utilized to investigate and process case was reviewed with the Safe Prison
Program Coordinators at each unit visited. The process appears efficient and
effective in the management and investigation of allegations of this nature
3.

Data Collection and Analysis

The TDCJ has a great deal of data available related to the allegations of
sexual assault within its facilities. These data are in both electronic and hard
copy formats. The data reported in this document represents an example of
capabilities of the TDCJ database to collate and analyze data. At the institutional
level there are additional logs that document allegations and the dispositions of
these cases.
The Prisoner Protection Investigation Log (SPP-01) tracks investigation
requests. This information is maintained at each unit within the TDCJ but is not
collated into one single database department wide. TDCJ policy requires that the
Unit Safe Prisons Project Coordinator shall maintain a file on victims, predators,
and all prisoners who are being investigated. Some of these files are maintained
manually while others are maintained locally in an electronic format. Upon
transfer to another facility the hard copy safe prisons file is forwarded to the
receiving unit.
What is not available is a single database, accessible to selected staff at
all units, of all allegations filed. At present the units only have electronic access
to allegations filed at their respective units. We suggest that it is important that
intake staff, classification staff, Safe Prisons staff, and administrative staff have
access to the number and type of allegations filed that would include the name of
the victim and the name of the alleged predator, the location and the
investigation outcomes, staff and inmate witnesses, investigative officer, etc.
This information is presently available in a variety of locations and formats in

- 54 -

The JFA Institute

TDCJ but it is not available in a single database that permits access at the unit
level. Staff should be able to determine how many system—wide allegations
have been filed, the nature of these allegations, and the outcomes of the
investigation as they review a prisoner for placement within the TDCJ units.
The following outlines examples of the capabilities of a database to accomplish
the above purposes:
•

Permits the identification all allegations filed against a specific prisoner

•

Identifies all allegations filed by a specific prisoner, either as a predator or
a victim

•

Allow for sorting of allegations by location, date, time of day, and other
relevant variables

•

List the investigative staff and the disposition of each case closed by the
specific investigator

These examples represent only a few of the analytical reports that can be
utilized by the TDCJ in its effort to improve its monitoring capability of the
reporting and handling of allegations of sexual assault. The available data can be
utilized to assist intake staff, classification staff, and others involved in the
decision making of placement of prisoners within units and specifically to housing
units. The TDCJ should examine its existing database and hard copy records
and initiate efforts to expand its capability by making the Prisoner Protection
Investigation Log a department wide database whose information is retrievable at
each unit.
4.

Facility Designs and Staff Deployment

The seven facilities selected for the on-site reviews represented a variety
of physical plants and construction designs. This permitted an examination of
how the physical design of the facility could either contribute to the incidence of
sexual assault or aid in the prevention of the occurrence of these events. Those
facilities that were similar in design had significantly different operational
missions and philosophies so it still represented an opportunity to contrast the
impact at each individual facility. The following summarizes the key findings and
observations at each facility.
Holliday and Murray
The Holliday and the Murray units are identical in design. The exception
involves the 50-bed Administrative Segregation Unit at the Holliday Unit and the
130-bed capacity Administrative Segregation Unit at Murray. The majority of
beds at both facilities are in dormitories. Holliday is a transfer unit (intake and

- 55 -

The JFA Institute

transfer processing center) and as such the movement in and out of the facility is
extremely high and results in a reported average length of stay of less than six
months. The Murray Unit is a facility housing female prisoners, and with the
exception of the Close Custody Unit, is of similar design and configuration to the
Holliday Unit.
The number of allegations at both the Holliday and Murray Units were
extremely low. This may be attributed to not only the operational mission of the
facilities but also the fact that the majority of beds in both of the facilities are
dormitories. As reported earlier in this report, the incidence of allegations housed
in dormitories was very low as compared to cells.
Staff at each of these facilities reported that the dormitory’s lack of privacy
and more direct visual supervision by staff contributed to the low number of
allegations. While dormitory housing is not considered the ideal in terms of
prisoner privacy or for the management of prisoners in a security setting, it
evidently does discourage sexual assault. The thought that open dorms with
open showers, toilets, etc would facilitate sexual assault is countered by the lack
of privacy and the improved visual supervision that the design offers to staff.
This however, only can be a deterrent factor if the units are properly supervised
and staffed. Both Units appeared to have adequate staff deployed in a manner
that would insure proper supervision of the units.
As noted the Murray Unit housed female prisoners. We observed that
there was an underlying attitude among some of the line staff at the Unit that
there was a high incidence of consensual sexual behavior occurring among the
prisoner population. Informal discussions with line staff indicated that this was
expected and a part of the life style of female prisoners. It was reported that they
felt that the incidence of sexual assault was low because of the willingness of the
female prisoners to engage in consensual sex with other prisoners.
This attitude was not apparent among supervisory staff nor was it
expressed by those involved in investigations and Safe Prisons implementation.
It is an issue that should be addressed by the TDCJ either through further
discussion with staff or through initiation of a formal cultural assessment of the
facility. It was unclear if the attitudes expressed by these staff members had
affected the reporting or investigation of sexual assault at the Unit.
Hughes and Polunsky
The Hughes and Polunsky Units, which are identical in design and similar
in mission and operational philosophy, are multi-custody units that house inmates
from G-1 up to G-5 (trustees up to maximum security). In addition, the Polunsky
unit houses death row prisoners. The majority of inmates are housed in cells that
are double-bunked. Each unit has an ‘expansion dorm’ that is a large dormitory
for G-1 and G-2 inmates.

- 56 -

The JFA Institute

Neither facility reported an allegation of sexual assault in the dorm
housing units in 2004 or through August 2005. This is consistent with what is
reported department wide. As with Holiday and Murray, staff reported that the
lack of privacy was a contributing factor, but also indicated that the classification
and selection process of those housed in these units prevented predators or
potential victims from being housed in the dormitory units.
The vast majority of allegations at these facilities occurred in the housing
units that utilized cells and specifically in double-bunked cells. In reviewing this
finding with staff it was suggested that the solid cell fronts, while permitting
privacy for the inmates and reducing noise within the unit, also provides the
degree of privacy that permits sexual assaults to occur. Unlike older prison
designs where the cell fronts consisted of bars, the solid doors limit visual
observation by staff and to a degree sound proof the cells to the point where staff
have difficulty hearing what is going on in individual cells. The solid cell fronts
are considered by many to be advancement in prison cell design and
construction and yet, in the case of preventing assaults in cells, this design may
not provide a deterrent due to the privacy it provides to its occupants. TDCJ is
experimenting and testing alternative cell fronts for these types of units. Among
these are the clear Lexan covers that would permit improved viewing into the
cells.
The fact that the majority of the sexual assaults occur between 6 a.m. and
6 p.m. can be understood when observing the level of activity in the housing units
of these facilities. Even though during these hours the number of staff assigned
to the units is increased, the level of activity that they are responsible for also
increases dramatically. The responsibility of housing unit staff draws them away
from the routine of checking on cell activity. This provides for periods of time
when only the housing unit picket officer (Control Room Officer) is monitoring the
day rooms and cell fronts. Because of the solid doors the picket officers’ ability to
monitor activity within the double-bunked cells is extremely limited. This provides
an opportunity for both consensual sexual activity and sexual assaults to occur
without observation of staff. This issue needs to be reviewed and addressed
both from a staff deployment approach and from a training issue with all housing
unit staff.
Skyview and Jester IV
The Skyview and Jester IV facilities serve as one of the primary
assessment, diagnostic and treatment facilities for mentally ill inmates. Jester IV
has a capacity of 505 inmates while Skyview has a capacity of 528. The physical
designs are significantly different. Jester IV is a facility that houses all of its
population in cells while the majority of the Skyview population is housed in small
dormitory housing units. Skyview does have a large close custody unit for those

- 57 -

The JFA Institute

who are experiencing acute mental illness or those who represent a security risk.
This is a cell unit similar to that found in the more modern Jester IV Unit.
The incidence of sexual assaults reported at these two facilities is high in
comparison to other TDCJ facilities. What is unusual is the fact that the vast
majority of these occur prior to arriving at the two mental health facilities. They
are attributed to these two facilities only because the allegations are reported to
staff of the two facilities. When the allegations attributed to the sending
institutions are removed, as was done for the analysis in Section V, the actual
incidence of sexual assaults occurring on these two units is low.
There are similarities in the operation of these two facilities that can
explain to a certain degree the low incidence of sexual assaults occurring at the
units. As would be expected of special needs treatment facilities, the staff to
inmate ratios is very high in comparison to other TDCJ facilities. This results in
almost constant supervision of all the prisoners of the two facilities. When
inmates are out of the cells for either group activity or treatment programming,
the prisoners are under constant supervision. This is combined with the fact that
very few of those housed in cells are double-bunked. The opportunity to engage
in behavior involving sexual assault is limited because of the single cells and
open dormitory configuration. The staff/inmate ratio permits constant supervision
at all times. The existence of these factors may help explain the low incidence of
sexual assaults occurring among those housed at these two facilities.
Staff of the two facilities provided some theories to explain why incidents
that occurred primarily at facilities in which the prisoner was previously housed
were now reported in these two facilities. The researchers did not interview any
inmates as part of this research but the staff referred to their conversations with
inmates and well as their own views. Some of the theories stated were the
following:
•

Mentally ill prisoners felt much more comfortable reporting these types of
incidents to treatment staff versus the security staff found at the sending
institutions; therefore, when inmates arrive at the mental health facilities
they tend to report incidents that had occurred in the past.

•

Inmates viewed staff at the two mental health facilities as much more
empathetic to allegations of this nature and more professional in the
handling of any investigation into the allegations.

•

Sexual assaults were reported as a means to justify the behavior that had
occurred prior to placement at the facility. For an example, mental health
staff reported that a suicide gesture could be offered as the result of being
the victim of a sexual assault.

- 58 -

The JFA Institute

•

Inmates alleged sexual assaults as a means to manipulate placement to
preferred housing at the two mental health facilities. Most inmates and
staff acknowledge that the conditions and environment at the two facilities
are preferable to that found at other high security facilities within the TDCJ
and that inmates may be using the allegation of a sexual assault as a
means to retain placement of these two facilities.

The above are anecdotal explanations to the high incidence of reporting at
these two facilities and do not represent any assumptions that can be verified
through research and investigation of each case. The TDCJ should establish a
process within the investigation protocol to determine and record the reasons for
the late reporting of these incidents and determine if any action or policy change
is warranted.
On the surface there does not appear to be any common
elements to where or when these allegations made upon arrival at the mental
health facilities are coming from. A variety of originating facilities are involved so
they are not originating at one or two facilities in which a larger problem may be
occurring. Further examination of this issue may bring some solid explanation to
light.
Darrington
Darrington is a 1,791 bed facility that was originally constructed in 1917.
The majority of housing units were constructed in the 1930’s. The facility
includes a 321 bed trustee camp located outside the perimeter of the main
facility. The facility houses G-1 (trustee) through G-5 male prisoners in a
combination of dormitories and traditional cellblocks.
Darrington was selected for review due to the low incidence of reported
sexual assaults. What differentiates this facility from units that house inmates of
similar security levels and demographics like those found at Hughes and
Polunsky, is the design of the facility and in particular the housing units.
The facility is a telephone pole design with all the housing units situated
off the main corridor of the facility. The cellblocks are the traditional three-tier,
open cell units. There were 20 double bunked cells on each of the three tiers
although one cell on each tier was no longer used for housing. Unlike Polunsky
and Hughes the cell fronts were the traditional bars and not closed steel fronts.
The permitted clear observation into the cells by staff and also created the
potential to hear any disturbance or assault occurring in the cells.
The general population units were staffed with one correctional officer
whose work station was located on the lowest tier near the entrance to the
cellblock. At this location his/her observation of the cells was somewhat limited.
However, a roving officer could observe all locations within the unit. This was
clearly a difficult unit to supervise by one officer. Staff reported that it is always
their intent to staff the units with two officers, but severe staff shortages have

- 59 -

The JFA Institute

made compliance with this a rare occurrence. On the occasion of the site visit to
this unit the temperature outside was extremely high and large fans were being
employed to circulate air within the units. This created a great deal of noise that
further limited the officers’ ability to monitor the unit. Still, staff felt that the open
cell fronts reduce the level of privacy within the cellblocks and thus probably
serve as a deterrent to sexual assault. Not only can a patrolling officer observe a
potential assault, other inmates in adjacent cells would be able to hear and report
any assault within the cells of the units.
The configuration of this unit differentiates it from other facilities with
higher incidence of reported sexual assaults. It should also be noted that the
lack of privacy within the cellblocks may also serve to discourage the reporting of
incidents to staff. The design of the units precludes doing almost anything in
private if the units are properly staffed. This is compounded by the fact that
virtually all the activities for the majority of inmates assigned to the cellblocks
occur in groups or within the cellblock. Very little out of cell time is also permitted
to this particular group of prisoners further reducing the opportunity to report
allegations or concerns in private. It would be difficult to report an allegation of
sexual assault and assure the privacy of the complaint in this specific situation.
Prisoners may fear retaliation from other inmates if it is discovered that they have
reported an allegation of sexual assault. The benefits of the openness of the
design of these units may in fact result in discouraging the reporting of assaults.
The TDCJ should continue to examine methods in which inmates can report
allegations of sexual assault without fear of being discovered while assigned to
the unit in which the alleged incident originated.
5.

Intake/ Classification

The Safe Prisons program relies upon the cooperation and communication
of a variety of staff within the institutions in order to achieve its goal of reducing
or eliminating sexual assaults. At all the facilities visited the communication links
between the classification staff and the Safe Prisons staff appeared to be strong
and effective.
There are no specific concerns or issues with the
Intake/Classification process as observed in this qualitative review.
The following summarizes these some key observations in the area:
a. During the TDCJ intake process, each inmate record is reviewed and the
inmate is interviewed by the Security Threat Group (STG) Office and the
Unit Safe Prisons Project Coordinator (see Incoming Chain Interview –
Form SPP-08) to determine if there is a potential or a history of the
prisoner being either a victim or predator. If indicators are present the
separation of the prisoner begins immediately upon intake.
b. The inmate’s institutional record when applicable contains information and
documents relative to a history of predatory behavior, victimization, and

- 60 -

The JFA Institute

need for safekeeping placement and is available for review during all
classification and placement hearings.
c. Staff assigned to implement and monitor the decision of the classification
staff in terms of housing placement utilize ‘tags’ on the placement board to
track the housing assignment of those with special designations such as
predator, safekeeping prisoner, etc. This permit a visual check of the
location of all prisoners who need to maintain separation within the facility.
d. In addition to the normal classification and risk scales the Classification
staff utilizes a department-wide standard of not housing together those
whose weight and age are significantly different. The standard at present
is 40 lbs and 9 years. Inmates who exceed this range in comparison to
the cellmate cannot be housed together. The result is that an inmate who
weighs 155 cannot be housed with a prisoner who weighs more than 195.
This is intended to reduce the vulnerability of smaller and less mature
inmates.
6.

Staff Education and Training

The TDCJ Correctional Training and Staff Development Department has
issued lesson plans and curriculum for a variety of training programs intended to
educate staff on the elements of the Safe Prison Project and to improve their
ability to identify and respond to all elements of sexual assaults in a correctional
environment. These programs are offered during both pre-service training and
in-service training for existing staff annually. The curriculum is comprehensive
and includes handouts, training manuals, crime victims brochures, and other
related materials.
The scope of this review did not allow for the comprehensive examination
of the effectiveness of this training material. However, informal discussions with
staff throughout the seven units did permit an assessment of the knowledge and
understanding of the elements of both the provisions of the Safe Prison program
and the dynamics surrounding sexual assault in a correctional setting.
Staff at all levels and functions were found to be knowledgeable and
understanding of the purpose, scope and requirements of the Safe Prisons
policies. Staff understood the protocol to be utilized in the event an incident or
allegation of sexual assault was observed or reported to them. Safe Prison staff
was knowledgeable and trained in the provisions of the policies and the
procedures to utilize in handling an allegation. Classification staff, administrative
staff, and other ancillary staff were found to be informed and understanding of
the purpose of the policies and their specific requirements and responsibilities in
the prevention and investigation of sexual assault.

- 61 -

The JFA Institute

Summary of Key Findings for the Unit Analysis
1. Staff at all levels was found to have a clear understanding of the
organization’s expectations and goals as it related to Safe Prisons
Program requirements.
2. There did not appear to be any indication of “organizational indifference”
at any of the facilities visited. There may be individuals within the
organization that may ‘look the other way’ when faced with an allegation of
sexual assault, but that was not a prevalent attitude observed within the
facilities visited.
3. The TDCJ should examine its existing database and hard copy records
and initiate efforts to expand its capability by making the Prisoner
Protection Investigation Log a department wide database whose
information is retrievable at each unit.
4. The number of allegations at both the Holiday and Murray Units were
extremely low and can be attributed to not only to the operational mission
of the facilities but also to the fact that the majority of beds in both of the
facilities are located in dormitory settings. As reported earlier in this
report, the incidence of allegations housed in dormitories was very low as
compared to cells.
5. The vast majority of allegations within the TDCJ occur at housing units
that utilized cells and specifically in double-bunked cells. In reviewing this
finding with staff it was suggested that the solid cell fronts, while permitting
privacy for the inmates and reducing noise within the unit, also provides
the degree of privacy that permits sexual assaults to occur. Unlike older
prison designs where the cell fronts consisted of bars, the solid doors limit
visual observation by staff and to a degree sound proof the cells to the
point where staff have difficulty hearing what is going on in individual cells.
6. The solid cell fronts are considered by many to be advancement in prison
cell design and construction and yet, in the case of preventing assaults in
cells, these solid cell fronts may not be a deterrent to prevent sexual
assaults due to the privacy it provides to its occupants.
7. The fact that the majority of the sexual assaults occur between 6 a.m. and
6 p.m. can also be understood when observing the level of activity in the
housing units in TDCJ facilities. Even though during these hours the
number of staff assigned to the units is increased, the level of activity that
they are responsible for also increases dramatically. The responsibility of
housing unit staff draws them away from the routine of checking on cell
activity on a routine basis. This provides for periods of time when only the
housing unit picket officer (Control Room Officer) is monitoring the day

- 62 -

The JFA Institute

rooms and cell fronts. This issue should be reviewed by the TDCJ from a
staff deployment and training standpoint.
8. The low level of allegations occurring at the mental health facilities could
be attributed to staff deployment levels. As would be expected of special
needs treatment facilities the staff to inmate ratios are very high in
comparison to other TDCJ facilities. This results in almost constant
supervision of all the prisoners of the two facilities both during in-cell and
out-of-cell activity periods.
9. The TDCJ should establish a process within the investigation protocol to
determine and record the reasons for the late reporting of incidents as was
found at the mental health units and determine if any action or policy
change is warranted. As noted in the report, a large percentage of cases
were reported only after a significant period of time had lapsed. The TDCJ
should review the reasons for these time delays to ensure that they are
not a result of breakdowns in the either the reporting or investigative
process. Specific attention should be paid to the reporting processes of
those incidents reported at the mental health facilities but having occurred
at the sending institution.
10. The openness of the cell clock design at facilities similar to Darrington
may in fact result in discouraging the reporting of assaults. The lack of
privacy may discourage incidents of sexual assault, but may also deter the
reporting of legitimate allegations of assault. The TDCJ should continue
to examine methods in which inmates can report allegations of sexual
assault without fear of being discovered while assigned to the unit in which
the alleged incident originated.
11. The TDCJ should examine the reasons for the large number of cases in
which either the victim or the alleged assaulter was transferred without
any disciplinary or legal action. These are cases which have also been
determined to be unsubstantiated. The questions centers on why do such
a large number of prisoners involved in these alleged incidents end up
getting transferred without any verification of the incident. Are these
transfers based on legitimate concerns or a means of foregoing either
discipline or prosecution?

- 63 -

The JFA Institute

VII.

Summary and Discussion

This study had several objectives. First, we wanted to better understand
the number and nature of sexual assault allegations being made in one of the
nation’s largest prison systems. Second, we wanted to understand how the
Texas Safe Prison program operates and its impact on sexual assault within its
numerous prisons. Finally, we wanted to see what lessons could be learned that
would have a positive impact on reducing the number of sexual assaults for both
Texas and other correctional systems.
Much of what has been reported here matches prior studies. The number
of official allegations of prisoner on prisoner sexual assault in Texas is relatively
low although it is noted that its allegation rate is higher than most states.
However, the higher allegation rate is due, in part, to the recently implemented
Safe Prisons program which is designed to encourage by staff and prisoners the
reporting of alleged sexual assaults.
Like most states the substantiation rate of the allegations is quite low. In
Texas one of the major reasons why the substantiation rate is so low is the delay
in having an allegation reported by the prisoner. As with any criminal
investigation, any undue delay in reporting a criminal act reduces the ability of
the investigators to gather physical evidence and interview persons who would
serve as potential witnesses.
The low rates of allegations and substantiated allegations makes it very
difficult if not impossible to develop a statistical profile or risk instrument that
would have strong predictive capabilities. However, there a number of attributes
that distinguish the victims and perpetrators from other prisoners.
The alleged victims are generally younger than their assailants. There is a
strong racial relationship in such incidents where victims are more likely white
while their assailants are more likely black or Hispanic and be gang affiliated.
The assailants are also older, more likely to have lived in urban areas, have been
convicted of a violent crime, are in a higher custody level and have served more
time in prison than their victims.
Not surprisingly the higher security facilities as well as units housing
special needs populations (psychiatric and mentally retarded) face particular
challenges in managing their inmates, with higher rates of general inmate-oninmate violence and higher rates of use of force, in relation to the population size
of these facilities. Incidents are more likely to occur in the day time in housing
units where officers can be easily distracted or have poor line of sight to the
celled housing units.
These profile data on the victims, assailants and facilities where such
allegations are made suggest that one needs to see the crime of prison rape in a

- 64 -

The JFA Institute

larger context. Specifically, it would appear that prison rape (or the threat of rape)
is just one weapon predatory prisoners will use to impose their influence and
control over other prisoners. The prisoners and prisoners that have the highest
rates of sexual assault also have the higher rates of other serious misconduct
and criminal activity. So it would suggest that in order to reduce prison rape
other forms of serious misconduct must also be addressed.
Finally, the research suggests that the Texas Safe Prison program has
produced a much greater awareness on the part of staff and prisoners on the
issue of prison rape. One of the immediate effects has been a much higher rate
of reporting and a more structured response and investigation to such
allegations. And the data base that has been created allows the TDCJ to monitor
the number, attributes and final dispositions of these incidents. Many states
could benefit from adopting many of the key components of the Texas program.
There are some suggestions that we would suggest to Texas and other
correctional systems that if implemented might serve to reduce sexual assaults.
First, further efforts are needed to provide structured opportunities for prisoners
to report any allegation of sexual assault –either staff on prisoner or prisoner on
prisoner. One specific suggestion is to require case managers to always inquire
of the prisoner during a formal session or contact whether any assaults have
occurred or if they are being threatened by other prisoners.
Second, there is a concern on what happens to prisoners who make an
allegation but the case cannot be substantiated. There are a large number of
these cases in Texas. Does staff make any effort to more closely monitor, reassign or supervise the prisoners who have been implicated in the incident? It is
not clear that any such efforts are being made by prison officials in such cases.
Third, the TDCJ should examine the reasons for the large number of
cases in which either the victim or the alleged assaulter was transferred without
any disciplinary or legal action. These are cases which have also been
determined to be unsubstantiated. The questions centers on why do such a
large number of prisoners involved in these alleged incidents end up getting
transferred without any verification of the incident. Are these transfers based on
legitimate concerns or a means of foregoing either discipline or prosecution?
Fourth the categories for designating prisoners in the TDCJ data base
should be expanded so that it has the following four designations:
e.
f.
g.
h.

Potential Victim
Known Victim
Potential Assailant
Known Assailant

- 65 -

The JFA Institute

The low rate of officially reported sexual assault on prisoners means that it
is not practical nor recommended that a traditional risk scoring system be
attempted that would assign prisoners to the potential victim and potential
assailant categories. To do so would produce an unacceptable level of “false
positives” in terms of identifying both potential victims and assailants. This not to
say that the factors and attributes that are associated with assailants and victims
as presented in this report cannot be applied in some manner.
Rather, it is recommended that a criteria for applying the “potential” victim
or assailant label should take the form of a check list that takes into account the
factors found to be associated with such allegations in this study. An example of
such a check list is shown on the following page. This checklist, which needs to
be customized for each state would serve to ensure that all prisoners are being
assessed in a systematic manner on the extent to which they may later become
involved in sexual assault incidents. The concept is that prisoners so identified
would be monitored for a set period (at least six months) of time to determine
whether the designation should be retained or removed.
Finally, this research did not adequately address the issue of sexual
assault among female prisoners. We found at the women prisons that the
attitudes expressed by the staff suggested that they believe sexual activity was
more common then at male facilities but that such behavior was largely
consensual. We are not persuaded that this is indeed the case. Clearly a
separate and more detailed assessment of sexual assault among female
prisoners is needed better understand the dynamics of sexual assault among
females.

- 66 -

The JFA Institute

SEXUAL ASSAULT VICTIM/ASSAILANT PROFILE CHECKLIST
I. Identifying Data
Date Of Referral: ___/_____/_____ Referred By:____________________________
Facility: _______

Inmate Name:_________________ DOC Number: ___________

Reason for Referral: ___________________________________________________
______________________________________________________________________
______________________________________________________________________
II.

Possible Victim Factors
1.
2.
3.
4.
5.
6.
7.
8.
9.

Yes/No
_____
_____
_____
_____
_____
_____
_____
_____
_____

Youthful Age (18 or younger)
Elderly (60 or older)
Physical stature (5’6” or less and/or LT 140 lbs)
Developmental Disability/Mental Illness
First Incarceration Ever
Homosexual/Bi-sexual
History of sexual abuse
History of facility consensual sex
Prior History of Protective Custody (adult/juv)

If three or more “Yes”, enter alert code for Vulnerable Inmate
Comments:____________________________________________________________
______________________________________________________________________
______________________________________________________________________
III.

Possible Predatory Factors
Yes/No
______

1. Current or Sexual Assault Conviction
2. Past history of institutional predatory behavior

______
______
______
______

3. Any History of Sexual Abuse
4. Any History of Physical Abuse
5. Gang affiliation
If two or more “Yes”, enter alert code for Potential Aggressor

Comments:___________________________________________________________________
____________________________________________________________________________
_________________________________________________________________________
Completed By: _______________________________

- 67 -

Facility: ____________

The Habeas Citebook: Prosecutorial Misconduct Side

The Habeas Citebook Ineffective Counsel Side