Skip navigation

Selected Rehabilitation Programs at the DCJ - Audit Report, TX SAO, 2007

Download original document:
Brief thumbnail
This text is machine-read, and may contain errors. Check the original document to verify accuracy.
John Keel, CPA
State Auditor

An Audit Report on

Selected Rehabilitation Programs
at the Department of Criminal
Justice
March 2007
Report No. 07-026

An Audit Report on

Selected Rehabilitation Programs at the
Department of Criminal Justice
SAO Report No. 07-026
March 2007

Overall Conclusion
Three of five programs auditors evaluated at the
Department of Criminal Justice (Department)
reduced recidivism (re-arrest and reincarceration) rates. A fourth program reduced
re-arrest rates but had higher re-incarceration
rates; while the fifth program did not reduce
either re-arrest or re-incarceration rates.
Recidivism rates of offenders completing the Sex
Offender Treatment Program, the Sex Offender
Education Program, and the Pre-Release
Therapeutic Community program were lower
than the recidivism rates of offenders who were
eligible for these programs but did not
participate. For example, offenders who
completed the Sex Offender Treatment Program
had re-incarceration rates that were 61.6
percent lower than sex offenders who were
eligible for but did not participate in this
program. Also, 80 percent of offenders who
completed this program and were later reincarcerated were re-incarcerated for technical
violations of their parole, not for arrests on new
charges.

Programs Evaluated
The following programs were selected
for review:

ƒ Sex Offender Treatment Program.
ƒ Sex Offender Education Program.
ƒ Pre-Release Substance Abuse
Program.

ƒ Pre-Release Therapeutic Community
program.

ƒ InnerChange Freedom Initiative
program.

These programs are part of the
Department of Criminal Justice’s
(Department) rehabilitation tier, a
group of programs designated by the
Department as having a positive impact
on recidivism rates.

Recidivism
Recidivism rates were calculated using
two different measures: re-arrests and
re-incarcerations.
A re-arrest is when an offender is
arrested for a new criminal offense
within two years of his or her release
from prison. A re-incarceration is when
an offender is sent back to prison within
two years of his or her release as a
result of a new conviction or a technical
violation of parole rules.

The Pre-Release Substance Abuse Program did not
reduce re-arrest or re-incarceration rates for
participants. While the faith-based InnerChange
Freedom Initiative program reduced re-arrest rates, it showed higher reincarceration rates among participants. There are possible reasons for these
results, including the serious nature of the offenses of the participants in the PreRelease Substance Abuse Program and the small number of offenders who
complete the InnerChange Freedom Initiative program.
The Department maintains the data necessary to calculate recidivism rates for
individual programs, but this data resides in multiple places (such as the
Department’s mainframe and standalone computers at the program level) and is
difficult to obtain. In addition, the data is not always complete and may not be
entirely accurate (see Chapter 1-A for more information on recidivism rates).
This audit was conducted in accordance with Texas Government Code, Section 321.0132.

For more information regarding this report, please contact Lisa Collier, Audit Manager, or John Keel, State Auditor, at (512) 936-9500.

An Audit Report on
Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026

The process used by the Department and the Texas Board of Pardons and Paroles
(Parole Board) for assessing and placing offenders in programs is not efficient and
may result in some offenders remaining in prison longer than they should, which
could contribute to prison overcrowding. The Parole Board, as part of its voting
process, recommends which program an offender should complete before release.
However, its recommendations were not implemented in 41.3 percent of the cases
auditors reviewed. This is because the Department’s clinicians assess offenders for
placement in a program after the Parole Board makes its recommendation. The
Department makes the final decision regarding which program an offender is
required to complete before release. Often, the clinicians’ and the Parole Board’s
recommendations differ. As a result, offenders who complete programs early
sometimes remain in prison until their Parole Board-approved earliest possible
release date; other offenders must remain incarcerated past their earliest possible
release date while they complete a program. Both of these situations could
contribute to overcrowding at prisons. Between September 2005 and February
2007, there were 2,176 offenders with program completion dates available in the
Department’s automated system. Of these, 549 (25.2 percent) had not been
released on parole as of February 2007.

Key Points
The Department needs to increase the capacity of some rehabilitation programs.
As of January 2007, there were 10,713 offenders with Parole Board-approved
release dates (or FI-R votes, which stands for Further Investigation-Rehabilitation
votes) who were waiting to complete a program before being released from prison.
Of these, 435 offenders (4.1 percent) were not enrolled yet in a rehabilitation
program. These offenders will not have enough time (based on the length of the
programs they are waiting for) to complete their programs before their tentative
release dates set by the Parole Board. These offenders are not currently enrolled
in a program because of a lack of available space in some programs.
Two programs documented the treatment provided to offenders, but three other
programs lacked sufficient documentation.
The Pre-Release Substance Abuse Program and the Pre-Release Therapeutic
Community program generally documented the treatment provided to offenders as
required by the Department’s policies and procedures. However, the Sex Offender
Treatment Program and the Sex Offender Education Program should improve their
documentation in offenders’ files of the treatment and education provided to
these offenders.
In addition, auditors did not find sufficient documentation at the InnerChange
Freedom Initiative program and, therefore, were unable to assess that program’s
performance or evaluate the support for the treatment provided to program
participants. The InnerChange Freedom Initiative is a faith-based program that is
staffed mostly by volunteers and does not receive state funding.

ii

An Audit Report on
Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026

The Department has implemented the same types of rehabilitation programs shown
to reduce recidivism in other states.
The Department provides rehabilitation programs that are comparable to the types
of rehabilitation programs shown by research studies to reduce recidivism. While
Texas has the programs found by research to be effective, program capacity is not
always sufficient to meet the needs of Texas’s prison population. In addition, drug
courts have seen limited use in Texas. Several studies indicate drug courts in other
states have been effective in reducing recidivism. Vocational education programs
in prison, treatment-oriented intensive supervision programs, and drug treatment
programs in the community have also decreased recidivism.

Summary of Management’s Response
Management agrees with the recommendations in this report and is taking steps to
implement them.

Summary of Information Technology Review
Auditors relied on data from various automated systems at the Department to
calculate recidivism. The majority of the data came from the Department’s
mainframe computer. Auditors did not perform an information technology review
of this computer but relied on work done during prior State Auditor’s Office audits
that addressed the completeness, accuracy, and reliability of data stored on this
computer. These audit reports included An Audit of the Criminal Justice
Information System (State Auditor’s Office Report No. 06-022, February 2006) and
An Audit Report on Performance Measures at Five State Agencies (State Auditor’s
Office Report No. 07-005, December 2006). Auditors concluded from these audits
that the data stored on the Department’s mainframe is generally reliable.
Other data used to calculate recidivism and to select files for testing was obtained
from stand-alone personal computers at the program level. The data on these
computers is stored in either Microsoft Access or Excel databases. Auditors tested
some basic data for completeness and accuracy by selecting samples from these
databases for testing and verifying the data during testing. Auditors also matched
prison unit count sheets to information in the databases to ensure there were not
offenders in the programs who were not represented in the databases. Auditors
found these databases were generally accurate, but they were not always
complete. Not all of the data in these databases was tested.
Some of the data stored on both the Department’s mainframe and stand-alone
computers was not as easy to obtain or use as it should be (see Chapter 1-B for
further information on the data limitations).

iii

An Audit Report on
Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026

Summary of Objectives, Scope, and Methodology
The objectives of this audit were to:
¾

Determine whether the Department collects and maintains sufficient data for
measuring the effectiveness of programs intended to reduce recidivism.

¾

Determine the outcomes for participants in selected programs.

¾

Determine whether there is a documented selection process for program
participation and that the selection of participants is consistent with this
process.

¾

Identify rehabilitation programs in other states that have demonstrated a high
level of success.

The scope for recidivism calculations was the population of offenders who were
released from prison in fiscal year 2004 and who were eligible for or participated
in one of the five programs reviewed. Auditors calculated the re-arrest and reincarceration rates for a two-year period following an offender’s release from
prison. For program testing work and the Parole Board’s FI-R decisions, auditors
reviewed data from fiscal year 2006 and the first quarter of fiscal year 2007.
The audit methodology consisted of visiting all locations of the five selected
programs, conducting staff and offender interviews, analyzing data, reviewing
policies and procedures, and testing offender files at the programs and at four
parole offices.

iv

Contents
Detailed Results
Chapter 1

Three Selected Programs Reduce Recidivism, But
Problems with the Data Make It Difficult to Calculate
Recidivism Rates....................................................... 1
Chapter 2

The Department Needs to Work with the Parole Board to
Improve the Process of Assessing and Placing Offenders
in Programs Prior to Their Release from Prison .................. 5
Chapter 3

The Sex Offender Treatment Program and the Sex
Offender Education Program Need to Improve
Documentation of the Treatment and Education
Provided to Offenders ............................................... 11
Chapter 4

Two Pre-Release Programs Document the Treatment
Provided to Offenders as Required by the Department’s
Policies ................................................................. 14
Chapter 5

The Department Has Implemented the Same Types of
Rehabilitation Programs Shown to Reduce Recidivism in
Other States, But Additional Capacity Is Needed ............... 19

Appendices
Appendix 1

Objectives, Scope, and Methodology.............................. 26
Appendix 2

Comparison of Recidivism Results ................................. 31
Appendix 3

Demographics of Five Selected Programs and
Methodology for Calculating Recidivism Rates .................. 33

Appendix 4

Other Recidivism Studies ............................................ 42
Appendix 5

Recent State Auditor’s Office Work ............................... 48

Detailed Results
Chapter 1

Three Selected Programs Reduce Recidivism, But Problems with the
Data Make It Difficult to Calculate Recidivism Rates
Offenders who completed the Sex Offender Treatment Program, the Sex
Offender Education Program, and the Pre-Release Therapeutic Community
program in fiscal year 2004 were less likely to be re-arrested or reincarcerated in the two years following their release than offenders who were
eligible for these programs but did not participate. Offenders who completed
the Pre-Release Substance Abuse Program and the InnerChange Freedom
Initiative program had less favorable results; however, there are possible
reasons for this, including the serious nature of the offenses of the Pre-Release
Substance Abuse Program participants and the small number of offenders who
complete the InnerChange Freedom Initiative program.
Auditors found it was difficult to obtain the data needed to calculate
recidivism rates for individual programs, and the Department of Criminal
Justice (Department) could not always guarantee the reliability of this data.
This also may have contributed to the varying recidivism rates. The data
resides in multiple places, is difficult to obtain, and may not be complete or
accurate.
Chapter 1-A

Three Programs Appear to Be Effective at Reducing Recidivism
Three programs appear to reduce recidivism for the offenders who complete
them: the Sex Offender Treatment Program, the Sex Offender Education
Program, and the Pre-Release Therapeutic Community program. Auditors
calculated re-arrest and re-incarceration rates for offenders who participated in
these programs and were released from prison in fiscal year 2004. In addition,
auditors calculated re-arrest and re-incarceration rates for offenders who were
eligible for these programs but did not participate and were released from
prison in fiscal year 2004 (these offenders’ information was used for
comparison purposes).
Auditors found that offenders who completed the Sex Offender Treatment
Program, the Sex Offender Education Program, and the Pre-Release
Therapeutic Community program were less likely to be re-arrested or reincarcerated in the two years following their release than offenders who were
eligible for these programs but did not participate (see Table 1 on the next
page).

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 1

Offenders who completed the Pre-Release Substance Abuse Program were
slightly more likely to be re-arrested or re-incarcerated in the two years
following their release than offenders who did not participate in the program.
The Department has a shortage of substance abuse treatment beds, however,
so only the offenders with the worst substance abuse problems are likely to be
placed in this program. Consequently, these offenders may be more likely to
have severe drug or alcohol addiction problems that cause them to commit
offenses that result in a re-arrest or re-incarceration than offenders not
selected to participate in a substance abuse treatment program.
Offenders who completed the InnerChange Freedom Initiative program were
much less likely to be re-arrested in the two years following their release from
prison than offenders who were eligible for this program but did not
participate. However, offenders who completed this program were slightly
more likely to be re-incarcerated in the two years following their release than
eligible offenders who did not participate in this program. Because there are a
small number of offenders who completed this faith-based program, a high
recidivism rate can result from only a few re-arrests or re-incarcerations. For
example, of the 369 offenders who participated in this program in fiscal year
2004, 78 completed the in-prison portion of the program. Of these 78, 57
completed the aftercare program. Program completion is defined as offenders
who complete at least six months of aftercare in the community. Because of a
lack of records, auditors relied on the program’s staff to identify the offenders
who completed aftercare (see Chapter 4-C for more information on the
InnerChange Freedom Initiative program). Of the 57 offenders completing the
aftercare program, 11 were re-arrested and 8 were re-incarcerated.
Table 1

Recidivism Rates for Offenders within Two Years of Release from Prison
Program

Recidivism
Measure

Completers of the
Program

Non-Completers of
the Program

All Program
Participants

Comparison Group

Sex Offender
Treatment
Program

Re-arrest

16.5%

26.7%

22.4%

17.7%

Re-incarceration

5.9%

6.0%

6.0%

15.3%

Sex Offender
Education
Program

Re-arrest

19.4%

26.7%

20.0%

22.4%

Re-incarceration

12.3%

6.7%

11.8%

15.4%

Pre-Release
Therapeutic
Community

Re-arrest

18.2%

50.0%

18.9%

42.6%

Re-incarceration

18.7%

15.0%

18.6%

26.7%

InnerChange
Freedom
Initiative

Re-arrest

19.3%

39.5%

36.2%

30.6%

Re-incarceration

14.0%

14.8%

14.7%

12.3%

Pre-Release
Substance Abuse
Program

Re-arrest

38.1%

51.7%

38.3%

31.4%

Re-incarceration

19.7%

24.1%

19.8%

15.1%

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 2

See Appendix 2 for a comparison of the State Auditor’s Office’s 2007 results
to those of the Criminal Justice Policy Council in 2003 and Appendix 3 for
demographics of test and control groups and more information on the
methodology used to calculate recidivism rates.
Chapter 1-B

The Data Needed to Calculate Recidivism Is Available, But It Needs
Improvement
The data to calculate recidivism rates is available at the Department, but it
resides in multiple places, such as on the Department’s mainframe computer
and stand-alone computers at the program level, and is difficult to obtain. In
addition, the data is not always complete and portions of the data may not be
entirely accurate.
Much of the electronic data for the Sex Offender Treatment Program, the Sex
Offender Education Program, and the InnerChange Freedom Initiative
program is maintained by program staff at both the Department’s
Rehabilitation and Re-entry Division in Huntsville and at various prison units
where the programs are housed. The data at these locations is maintained on
stand-alone personal computers and does not always match the data
maintained in other locations or on the Department’s mainframe. There are
other problems that make the accuracy and completeness of the program data
questionable. These include:
ƒ

Some mainframe data fields are not completely populated because of
limitations of the Department’s mainframe. Available program completion
dates are not always accurate because the mainframe does not allow
backdated entries. Therefore, the accuracy of the program completion
dates is dependent on staff promptly entering the data into the mainframe.

ƒ

Some mainframe data is overwritten when an offender’s status changes.
For example, an offender’s program enrollment date is overwritten as soon
as a program completion code is entered into the mainframe.

ƒ

Some data in the files on the stand-alone computers is not complete
because it is entered by hand and there are data-entry backlogs.

ƒ

Data in the files on the stand-alone computers is not collected and
maintained consistently across programs.

Recommendation

The Department should work with the Legislative Budget Board, the
Governor’s Office, and other agencies charged with calculating recidivism
rates to identify, process, and maintain the data in a manner that is useful, easy
to access, and ensures the completeness and reliability of the data.
An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 3

Management’s Response

TDCJ agrees to work with the Governor’s Office, the LBB, and other agencies
to determine the appropriate manner in which data should be identified,
processed and maintained. We will also continue our efforts to ensure the
data is useful, easily accessible, complete, and reliable.

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 4

Chapter 2

The Department Needs to Work with the Parole Board to Improve the
Process of Assessing and Placing Offenders in Programs Prior to Their
Release from Prison
The process used by the Department and the Texas Board of Pardons and
Paroles (Parole Board) for assessing and placing offenders in programs is not
efficient and may result in some offenders remaining in prison longer than
they should, which could contribute to prison overcrowding.
The Parole Board, as part of its voting process, recommends which program
an offender should complete before release. However, its recommendations
were not followed in 41.3 percent of the fiscal year 2006 cases auditors
reviewed. This is because the Department’s clinicians assess offenders for
placement in a program after the Parole Board makes its recommendation, and
the clinicians’ recommendation often differs from the Parole Board’s
recommendation.
The Department makes the final decision regarding which program an
offender is required to complete before release. This process may result in
some offenders spending more time in prison than they should. For instance,
some offenders who complete programs early must remain in prison until their
Parole Board-approved earliest possible release date; other offenders must
remain in a program past their earliest possible release date until they
complete the program. Both of these situations could increase the prison
population because they result in some offenders staying in prison past the
time they complete a program or past the point of their projected release date.
Chapter 2-A

The Process Used to Assess and Place Offenders in Programs Needs
Improvement
Offenders are placed in programs either by a recommendation from the Parole
Board or by an assignment from the Department’s Rehabilitation and Re-entry
Division. When the Parole Board votes to recommend that an offender be
placed in a program, this is referred to as an “FI-R vote,” which stands for
Further Investigation-Rehabilitation. The Parole Board then votes on a
tentative release date that allows the offender enough time to complete the
recommended program. In making its recommendation about which program
an offender should complete, the Parole Board relies on summaries from the
institutional parole officers.
After the Parole Board makes its recommendation, the Department’s
clinicians assess the offender for program placement, sometimes selecting a
different program than the one recommended by the Parole Board. The
Department’s recommendation is the one that is implemented. As a result,
3,386 offenders out of 8,202, or 41.3 percent, were placed in or are scheduled
An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 5

to be placed in programs that differ from those recommended by the Parole
Board (see Table 2).
Some offenders who complete a shorter program must wait in prison until
their release date, while others who complete a longer program must remain in
the program until after their release date. Both of these conditions could result
in increases to the prison population because they prolong offenders’ lengths
of stay.
Table 2

Offenders Placed in Programs
Fiscal Year 2006 and the First Quarter of Fiscal Year 2007

Program

Program
Length
(Months)

Total Number of
Offenders in
Program

Number of
Offenders with
Different FI-R Type
and Program Type

Percent of
Offenders with
Different FI-R Type
and Program Type

Changes

3

3,564

1,853

52.0%

Driving While
Intoxicated

4

13

0

0.0%

18

138

17

12.3%

In Prison
Therapeutic
Community

6

184

183

99.5%

Life Skills

3

154

67

43.5%

Pre-Release
Substance Abuse
Program

6

2,039

654

32.1%

Pre-Release
Therapeutic
Community

6

1,557

413

26.6%

Segovia

3

1

0

0.0%

Sex Offender
Education Program

4

161

104

64.6%

Sex Offender
Treatment Program

18

289

92

31.8%

6

14

3

21.4%

Voyager

3 to 12

87

0

0.0%

Youthful Offender
Program

9 to 12

1

0

0.0%

Totals

8,202

3,386

41.3%

InnerChange
Freedom Initiative

Serious Violent
Offender Re-entry
Initiative

Source: Data provided by the Department and analyzed by auditors.

The Department’s available release data for offenders released from programs
between September 1, 2005, and November 30, 2006, contained information
for 2,176 offenders who completed a program. Of these, 549 (25.2 percent)
An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 6

had not been released on parole as of February 2007 (see Table 3 on the next
page). Of those who were released, an offender spent an average of 56 days in
prison after his or her program completion date.
The 549 offenders who were still in prison after completing a program had not
been released on parole for the following reasons:
ƒ

They had not reached their “release no earlier than” date set by the Parole
Board.

ƒ

They had passed their “release no earlier than” date, but their release plans
had not been approved. For example, they were waiting for a halfway
house placement, or they did not have an approved home to which they
could be paroled.

According to Parole Board policy, offenders should be released on parole on
their “release no earlier than” date or upon their program completion date,
whichever is later.
The Department’s State Classification Committee checks to ensure that an
offender meets the Department’s criteria for transfer to a program after the
Parole Board vote and the Rehabilitation and Re-entry Division’s assessment.
This process adds to the time it takes to assess and place an offender in a
program.

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 7

Table 3

Offenders Waiting to Be Released on Parole after Completing a Program
(as of February 2007)

Program

Total Number of
Offenders with Program
Completion Dates

No Program Type

Number of Offenders
with Program
Completion Dates Who
Had Not Been Released
on Parole

Percent of Offenders
with Program
Completion Dates Who
Had Not Been
Released on Parole

147

24

16.3%

1,160

278

24.0%

6

0

0.0%

InnerChange
Freedom Initiative

12

11

91.7%

In Prison Therapeutic
Community

62

43

69.4%

Life Skills

Changes
Driving While
Intoxicated

34

8

23.5%

Pre-Release
Substance Abuse
Treatment Program

402

95

23.6%

Pre-Release
Therapeutic
Community

293

62

21.2%

0

0

0.0%

Sex Offender
Education Program

19

14

73.7%

Sex Offender
Treatment Program

8

6

75.0%

Serious Violent
Offender Re-entry
Initiative

2

1

50.0%

31

7

22.6%

0

0

0.0%

2,176

549

25.2%

Segovia

Voyager
Youthful Offender
Program
Totals

Recommendations

The Department’s Rehabilitation and Re-entry Division and the Parole Board
should work together to assess offenders who are nearing their release date
and recommend a program based on the Department’s clinical expertise.
The placement of offenders into a program will be easier and timelier if the
screening process performed by the State Classification Committee is
performed earlier in the assessment and placement process.
The Department should improve its parole release process to ensure that
offenders who complete programs are released on parole in a timely manner.
An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 8

This would help to ensure that bed space is available for new program
participants. In addition, prompt release on parole would minimize the
chances of offenders being negatively influenced by members of the general
prison population who have not participated in rehabilitation programs.
Management’s Response

TDCJ agrees to work with the Board of Pardons and Paroles to re-evaluate
the established processing procedures. To support a parole release vote,
which includes a recommendation for a specific program, the Department
would complete the final clinical assessment on all eligible offenders prior to
submission to the BPP. We would also notify the BPP in the event their
recommendation for program placement appears to be inconsistent with
TDCJ’s final clinical assessment to allow them the opportunity to determine
whether their program placement recommendation should be revised.
In addition, we would complete an initial screening by the State Classification
Committee prior to submission to the BPP. However, a re-screening will need
to occur after the BPP vote to ensure the offender is still eligible for program
placement at the time of program enrollment. In the event the offender is
determined to be ineligible for program placement after the BPP vote, the
BPP will be notified to allow them to consider the need for vote modification.
The actual parole release of the offender will still be dependent on the
completion of the program and on the offender having an approved parole
release plan. Approval of a parole release plan will sometimes still involve
the limited availability of Half-Way House beds that may delay release after
completion of the assigned program.
Chapter 2-B

The Department Needs to Increase the Capacity of Some
Rehabilitation Programs
As of January 2007, there were 10,713 offenders with Parole Board-approved
release dates (FI-R votes) who were waiting to complete a program before
being released from prison. Of these, 435 offenders (4.1 percent) were not
enrolled yet in a rehabilitation program (see Table 4). These offenders will not
have enough time (based on the length of the programs they are waiting for) to
complete their programs before their tentative release date set by the Parole
Board. These offenders were not enrolled in a program because of a lack of
available space in some of these programs.
According to the Parole Board, offenders with Parole Board-approved release
dates should be enrolled in a rehabilitation program no later than their
specified release date. Programs with the largest number of offenders waiting
to be placed include the Pre-Release Substance Abuse Program, the Pre-

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 9

Release Therapeutic Community program, the Sex Offender Treatment
Program, and the Sex Offender Education Program.
Table 4

Offenders Waiting to Be Enrolled in a Program
Who Will Not Be Able to Complete the Program Before Their Target Release Date
Number of
Offenders

Program
Changes (curriculum available at most prisons)

Percentage of
Offenders
40

9.2%

Driving While Intoxicated

0

0.0%

InnerChange Freedom Initiative

1

0.2%

In Prison Therapeutic Community

0

0.0%

Life Skills

1

0.2%

167

38.5%

68

15.6%

0

0.0%

Sex Offender Education Program

44

10.1%

Sex Offender Treatment Program

49

11.3%

Serious Violent Offender Re-entry Initiative

1

0.2%

Voyager

0

0.0%

Pre-Release Substance Abuse Treatment
Program
Pre-Release Therapeutic Community
Segovia

Youthful Offender Program
No Program
Totals

0

0.0%

64

14.7%

435

100.0%

Source: Unaudited data provided by the Department.

Recommendation

The Department should ensure that offenders with Parole Board-approved
release dates are placed in the appropriate programs on a timely basis so the
offenders can complete the program before or on their “release no earlier
than” dates. To do this, the Department may need to increase the available
capacity for some programs.
Management’s Response

TDCJ agrees to continue to explore ways to increase available program
capacity to enhance the timely placement of offenders.

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 10

Chapter 3

The Sex Offender Treatment Program and the Sex Offender Education
Program Need to Improve Documentation of the Treatment and
Education Provided to Offenders
Although the Sex Offender Treatment Program and the Sex Offender
Education Program are providing treatment and education to eligible
offenders, these programs do not consistently document the offenders’
progress as required by the Department’s policies. Auditors tested 115
randomly selected participants’ files at the prison units providing one or both
of these programs and found that all the offenders were eligible for these
programs. However, many of the files did not contain sufficient
documentation to support the treatment and education provided. For example,
79 percent of the 85 files tested in the Sex Offender Treatment Program did
not contain evidence of the required individual therapy.
In addition, auditors visited 4 parole offices and reviewed 15 files of paroled
sex offenders at each location. Auditors found that there was sufficient
evidence that sex offenders are receiving adequate supervision and treatment
while on parole.
Chapter 3-A

The Department Needs to Ensure That Offenders’ Treatment or
Educational Progress Is Consistently and Adequately Documented
The Sex Offender Treatment Program and the Sex Offender Education
Program do not consistently document the provision of treatment and
education to eligible participants. Auditors tested 25 to 30 program files at
three prison units providing the Sex Offender Treatment Program and found
that:
ƒ

67 (79 percent) of 85 program files did not contain evidence of the
required individual therapy.

ƒ

59 (69 percent) of 85 program files did not contain pre- and post-test
results.

ƒ

19 (22 percent) of 87 program files did not contain closing summaries,
which are treatment summaries provided to the Department’s Parole
Division.

Additionally, auditors tested 30 offender files from the Sex Offender
Education Program and found that 12 (40 percent) of 30 files did not contain
pre- and post-test results. Auditors also found that both the Sex Offender
Treatment Program and the Sex Offender Education Program’s policies and
procedures have not been updated since 2002, so they may not reflect the
current practices within those programs.
An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 11

Recommendations

The Department should:
ƒ

Revise its policies and procedures for the sex offender programs to reflect
current practices.

ƒ

Ensure that program staff consistently document offenders’ progress as
required by the Department’s policies and procedures.

Management’s Response

TDCJ agrees to review and, as necessary, revise the policies and procedures
to ensure they reflect current practice. In addition, documentation
requirements will be reviewed to ensure they are appropriate for the program
and if required they will be revised. Once revisions, if any, are completed
remedial training will be provided to staff and monitoring will be
implemented to ensure documentation is maintained in accordance with
established standards.
Chapter 3-B

The Two Sex Offender Programs Provide Rehabilitation for Sex
Offenders Who Are Released from Prison
The Sex Offender Treatment Program and the Sex Offender Education
Program are intended to reduce a sex offender’s risk of re-offending by
providing need-specific rehabilitative interventions. The Sex
Sex Offender Treatment Program
Offender Treatment Program is an 18-month program that
and Sex Offender Education Program
provides both education and therapy. The Sex Offender Education
The Sex Offender Treatment Program has
Program is a four-month program that provides education but not
a capacity of 484 offenders and is
operated at three prison units: the Goree,
therapy.
Hightower, and Hilltop units. The
programs at the Goree and Hightower units
serve male sex offenders and have a
capacity of 204 and 252 beds,
respectively. The program at the Hilltop
Unit serves female sex offenders and has a
capacity of 28 beds.
The Sex Offender Education Program was
implemented in 2002 at the Hightower
Unit. This program serves male offenders
and has a capacity of 111 beds.
Expenditures for fiscal year 2006 for both
programs were $2,110,187. The
Department’s budget for the two programs
during fiscal year 2007 is $2,189,636.
There are 62 staff positions budgeted for
these programs.
Source: Capacity and staffing data are
from the Department and the
Department’s fiscal year 2008-2009
Legislative Appropriations Request.

The Sex Offender Treatment Program provides treatment to moderate- to
high-risk sex offenders. The program began as a pilot program in

1989 to serve male sex offenders and participation was voluntary.
In 1998, participation in the program became mandatory for
selected sex offenders. In December 2000, the program began
serving female sex offenders.
Most of the offenders entering the program are moderate- to highrisk sex offenders. As of January 12, 2007, there were 4,461
offenders eligible for the program, according to the Department.
Of those, 113 offenders had been screened and assigned to a unit,
but they were waiting for a bed to become available.
The program is based on a cognitive-behavioral model and
consists of three phases. Phase I is the evaluation and treatment
orientation phase and lasts about three months. It includes a

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 12

psychological evaluation of each offender, which is used to develop an
individualized treatment plan. This phase also includes psycho-educational
classes to assist offenders in admitting guilt, accepting
Cognitive Behavioral Model and
responsibility for their offenses, and developing appropriate coping
Psycho-Educational Classes
skills.

The cognitive-behavioral model of
treatment focuses on changing offenders'
deviant thought patterns to change their
patterns of behavior.

Psycho-educational classes, which are a
component of the cognitive-behavioral
model, are designed to educate offenders
about their deviant thinking patterns and
provide information to help offenders
change their unhealthy patterns of
thinking.

Phase II is the intensive therapy phase, which lasts about 12
months. It provides a structured living and treatment environment.
Offenders attend individual and group therapy sessions, which
focus on changing the offenders’ deviant behavior and thought
patterns.

Phase III is the transition and release preparation phase, which
focuses on preparing the offenders for re-integration into society.
This phase lasts about three months.
The Sex Offender Education Program provides education but not therapy.

The Sex
Offender Education Program is designed for low-risk sex offenders. Its
structure is similar to Phase I of the Sex Offender Treatment Program.
Offenders complete psycho-educational classes, but they do not attend group
or individual therapy sessions. To successfully complete the program,
offenders are required to actively participate in the classes and demonstrate an
understanding of the required educational components. As of January 12,
2007, the program had screened 485 eligible offenders who were waiting for
an available bed, according to the Department.

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 13

Chapter 4

Two Pre-Release Programs Document the Treatment Provided to
Offenders as Required by the Department’s Policies
The Pre-Release Substance Abuse Program and the Pre-Release Therapeutic
Community program generally are providing treatment, education, or work
experience as intended. Both programs have documented selection criteria for
program participation. Auditors evaluated the files of 30 participants in the
Pre-Release Substance Abuse Program and 30 participants in the Pre-Release
Therapeutic Community program and found all participants met the programs’
selection criteria.
However, auditors found the InnerChange Freedom Initiative program lacked
sufficient documentation to assess program performance or the level of
treatment provided to program participants. Auditors evaluated the files of 30
participants and verified that each participant signed a release stating he
volunteered to participate in the program. The InnerChange Freedom Initiative
program is a faith-based program staffed primarily by volunteers; it does not
receive state funding.
Chapter 4-A

The Pre-Release Substance Abuse Program Provides Treatment to
Chemically Dependent Offenders
Auditors found sufficient documentation that treatment and education or work
experience are being provided by the Pre-Release Substance Abuse Program
and that participants are required to comply with the requirements
The Pre-Release Substance
for completing the program. Auditors tested files for 30 current
Abuse Program
program participants and found that all of these offenders met the
The Pre-Release Substance Abuse
eligibility criteria for program participation.
Program was started in 1996 and is at
the LeBlanc Unit in Beaumont, Texas. It
operates 1,008 beds for male offenders.
Expenditures for fiscal year 2006 were
$1,112,082; the budget for expenditures
in fiscal year 2007 is $1,535,059.
Between September 1, 2005, and
November 30, 2006, there were 2,193
offenders enrolled in the program.
During the same time period, there
were 2,073 graduates of the program.

The Pre-Release Substance Abuse Program has a staffing shortage.
Unit staffing reports indicate that the program is intended to operate
with 65 staff members, but it is currently operating with 32 staff
members. Between September 1, 2005, and November 30, 2006, the
program had a high of 46 staff members (in September 2005) and a
low of 31 staff members (in October 2006).

Sources: The Department’s program
capacity and graduate data and the
Department’s Legislative Appropriations
Request for the 2008-2009 biennium.

The Pre-Release Substance Abuse Program is an intensive, sixmonth program for offenders with serious substance abuse or
chemical dependency problems, as well as anti-social characteristics.
To participate in the program, offenders must have a minimum of seven
months left in prison before release. In addition, to be eligible for this
program, a substance abuse screening instrument must show that the
participants are chemically dependent.
The program is divided into three phases, all of which use standard work
books, group meetings, and therapy. The first phase, Orientation, lasts about
An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 14

30 days. During this phase, participants are introduced to the program’s
structure, rules, language, and thinking patterns.
The second phase, Cognitive Intervention, lasts about 90 days. During this
phase, participants identify their basic needs and the negative processes they
previously used to fulfill these needs. They also learn about the
addiction/offender cycle and how to stop it, and how to identify various
thinking errors.
The third phase, Relapse Prevention, lasts about 60 days. During this phase,
participants address relapse, stress, and anger issues and prepare to return to
work upon their release from prison.
Chapter 4-B

The Pre-Release Therapeutic Community Program Provides
Rehabilitation to Multiple Needs Offenders
Auditors found sufficient documentation that treatment and education or work
experience are provided by the Pre-Release Therapeutic Community program
and that participants are required to comply with the requirements for
completing the program. Auditors tested files for 30 current
program participants and found that all of these offenders met
The Pre-Release Therapeutic
Community Program
the eligibility criteria for program participation.
Started in 1997, the Pre-Release Therapeutic
Community program is at the Hamilton Unit in
Bryan, Texas, and has 600 treatment beds and
566 pre-treatment beds for male offenders.
Expenditures during fiscal year 2006 were
$574,445; the budget for expenditures for fiscal
year 2007 is $996,341.
Between September 1, 2005, and November 30,
2006, there were 1,452 offenders enrolled in
the program. During the same time period,
1,421 offenders graduated from the program.

The Pre-Release Therapeutic Community program has a staffing
shortage. Allotted to have 32 staff members, the program
currently has 26 treatment and program staff members,
according to the Department. During April 2006, the program
operated with 19 staff members.

The Pre-Release Therapeutic Community program is an
intensive, six-month treatment program for offenders. The
Sources: The Department’s program capacity
therapeutic community is “a positive, self-contained
and graduate data and the Department’s
Legislative Appropriations Request for the 2008environment where offenders who have similar treatment needs
2009 biennium.
live and work together toward a common goal of addiction
recovery, positive behavior, and life change through peer, group, and
hierarchical structure,” according to program literature. The program provides
pre-release treatment to offenders who are within seven months of release.
Individual treatment plans are developed and monitored to ensure appropriate
programming and support services, and a continuum of care plan for release is
developed through coordination with the Department’s Parole Division.
Program participants are housed in a dormitory setting to promote group
counseling and social interaction. They hold each other accountable for any
rule infractions. Some participants accept various responsibilities within their
housing unit, including acting as class facilitators.

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 15

The program has three phases. The first phase, Orientation, takes about 30
days. During this phase participants receive an overview of the program,
including program goals and principles. Participants also learn the tools, rules,
and structure used by the program.
The second phase of the program takes about 90 days and focuses on
identifying an offender’s basic needs, recognizing and breaking the
addiction/offender cycle, acknowledging an offender’s thinking errors, and
taking corrective action.
The third phase of the program takes about 60 days and focuses on relapse
prevention. Participants learn about relapse and stress issues, learn how to
deal with anger, and make plans to return to work upon being released from
prison.
Participants spend the day in a structured treatment schedule. A typical day
lasts from 6 a.m. to 8 p.m. and includes group therapy sessions; phase-specific
classes; support groups, such as Alcoholics Anonymous and Narcotics
Anonymous; and self-help programs. Chemically dependent participants use a
workbook to complete their lesson plans. Participants who are determined not
to be chemically dependant use the same workbook, but with altered lesson
plans to address criminal behavior and the offender cycle.
Chapter 4-C

The InnerChange Freedom Initiative Program Lacks Sufficient
Documentation to Assess the Program’s Performance
The InnerChange Freedom Initiative program is a faith-based program staffed
primarily by volunteers. It does not receive state funding. Offenders
voluntarily participate in the program. Auditors did not find sufficient
documentation to be able to assess this program’s
performance or evaluate support for the level of treatment
The InnerChange
Freedom Initiative Program
provided to program participants.
The InnerChange Freedom Initiative
program started in 1997 and is at the
Vance Unit in Richmond, Texas. As of
December 2006, there were 220 offenders
enrolled in the program.
The program is funded by Prison
Fellowship Ministries, a non-profit
organization. It does not receive state
funding.
The program’s director stated that the
program has a budget of $800,000:
$600,000 to operate the program and
$200,000 for the aftercare component.
There are seven paid employees at the
Vance Unit; the remaining staff members
are volunteers.

Because documentation in offenders’ files was limited, the
only criteria for enrollment that auditors could test was
whether the files contained a signed consent form
acknowledging that the offenders volunteered for the
program. Auditors selected a random sample of 30 offenders
enrolled in the program and reviewed their files; all 30 files
tested contained a signed consent form.

Files did not include sufficient documentation regarding
offenders’ daily activities to determine whether offenders
were receiving treatment as prescribed by the program.
Source: Unaudited information provided by
Because files maintained by the program’s staff at the Vance
program staff.
Unit and at the aftercare centers contained minimal
documentation, auditors also were unable to test whether offenders shown to
An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 16

have completed this program actually met the program’s completion
requirements, including aftercare participation.
To participate in the InnerChange Freedom Initiative program, offenders
must:
ƒ

Volunteer for the program.

ƒ

Be within 18 to 24 months from release.

ƒ

Sign consent forms for release of information and media release.

ƒ

Have the appropriate minimum security clearance.

ƒ

Satisfactorily complete the introduction and orientation phases.

ƒ

Have the potential to read at the sixth-grade level.

ƒ

Have the mental ability to function in the program.

ƒ

Plan to be paroled in the Houston or Dallas-Fort Worth area or
surrounding counties.

Offenders’ files included a signed participation form, notes from the
offenders’ volunteer counselors, and various quizzes and tests related to the
program’s curriculum. Auditors observed participants completing one of the
program’s tests, which covered the offender’s knowledge of the books of the
Bible.
According to program literature, the InnerChange Freedom Initiative program
is “a Christ-centered, Biblically rooted, faith-based program” that “seeks to
equip members with the understanding through the Word of God that sin is
the root of their problem.”
While the InnerChange Freedom Initiative program is a Christian program
(and all volunteers are Christian), members of other denominations can enroll
in it. Christians make up the majority of participants, but currently the
program also has participants of the Islamic, Jewish, Buddhist, and other
faiths.
The program consists of four main phases. Offenders start the program with a
group of their peers, called a “class,” and continue to move through the
program with that group. A new class begins every three months. The
program does not have a tailored curriculum. The program’s curriculum
includes a series of tests offenders must complete. But if the offenders fail the
tests, they can still advance through the program.
Because a class advances through the program and completes the program
together, auditors were unable to determine whether individual offenders
An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 17

received treatment as prescribed by the program, had properly mastered the
curriculum, and met the program completion requirements. Also, because
offenders are not required to pass tests on program information, it is not
possible to state whether offenders properly understand and are mastering the
program’s curriculum.
The first phase of the InnerChange Freedom Initiative program lasts 12
months and attempts to help the offender facilitate a life transformation by
eliminating the thinking processes that resulted in his incarceration and
rebuilding the offender’s value system. During this phase, assigned peer
groups meet weekly to assist offenders in understanding accountability and
affirmation. Based on each individual’s needs, offenders are enrolled in
various educational programs. Mentoring relationships are established with
men from the community where the offender will be paroled. Each offender is
expected to meet with his mentor for at least two hours per week.
Phase two lasts six months. Program participants complete public service in
the community by working off-site for most of the day. Participants continue
to attend peer group meetings but are coached in leadership roles. They also
are encouraged to complete General Educational Development (GED) classes
and spend classroom time developing skills needed for re-entry into society.
Offenders continue to meet with their mentors on an individual basis. They
also create a written life plan.
Phase three is supervised work release and lasts from the time the offender
completes the program until he is released from prison. Consequently, the
length of this phase time varies from offender to offender. One offender
currently residing at the Vance Unit completed the program over two years
ago, but he still remains in the unit pending parole.
Phase four is aftercare, which usually lasts six to nine months after the
offender is released into the community on parole. The purpose of this final
phase is to successfully re-integrate the offender into the community. If the
offender fulfills the aftercare requirements—which include remaining in good
standing with his parole requirements, making restitution payments,
participating in productive work, and becoming a member of a church—the
offender will graduate from the program. Mentoring relationships continue
throughout the aftercare phase.

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 18

Chapter 5

The Department Has Implemented the Same Types of Rehabilitation
Programs Shown to Reduce Recidivism in Other States, But Additional
Capacity Is Needed
A literature review of recidivism research from other states indicates that
multiple rehabilitation programs have been shown to reduce the recidivism
rates among program participants. Auditors identified these programs and
compared them to the programs provided by the Department. In each case, the
Department provides a comparable rehabilitation program to those shown by
research studies to reduce recidivism. While Texas has these types of
programs, capacity is not always sufficient to meet the needs of Texas’s
prison population. Rehabilitation programs with limited capacity include inprison substance abuse treatment, substance abuse treatment while on
community supervision, and drug courts.
Chapter 5-A

Research Has Identified Effective Rehabilitation Programs
Research conducted in recent years indicates that certain rehabilitation
programs have been successful in reducing recidivism rates among program
participants. While the number of rigorous, high-quality evaluations is limited,
those that do exist often reveal positive results for certain types of
rehabilitation programs.
Studies report successful outcomes for substance abuse treatment in the
community. Similarly, studies have shown that certain types of treatment for
sex offenders are effective in reducing recidivism rates among program
participants. Among the program types found to be effective are:
ƒ

In-prison therapeutic community treatment for drug-involved offenders.

ƒ

In-prison therapeutic community treatment for drug-involved offenders
with follow-up community treatment.

ƒ

Cognitive behavioral therapy for the general offender population.

ƒ

Prison-based cognitive behavioral treatment for sex offenders.

ƒ

Non-prison-based sex offender treatment.

ƒ

Vocational education programs.

ƒ

Correctional industry programs.

ƒ

In-prison therapeutic drug rehabilitation with work release programs.

ƒ

Halfway houses.

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 19

ƒ

Pre-release programs.

ƒ

Drug courts.

ƒ

Programs for offenders dually diagnosed with substance abuse and mental
health disorders.

ƒ

Intensive supervision with a focus on treatment.

ƒ

Drug treatment in local jails.

ƒ

Basic adult education programs in prison.

ƒ

Employment training and job assistance in the community.

Conversely, research also has revealed the ineffectiveness of other types of
rehabilitation programs. For example, adult boot camps and the Scared
Straight Program have been shown to be ineffective in reducing the recidivism
rates among program participants.
In October 2006, the Washington State Institute for Public Policy published
Evidence-Based Public Policy Options to Reduce Future Prison Construction,
Criminal Justice Costs and Crime Rates. That report describes the results of a
comprehensive statistical review of program evaluations that were determined
to be of sufficiently rigorous research. The researchers evaluated studies of
254 programs for adult offenders. Table 5 summarizes the results of this
review.
Table 5

Programs’ Impact on Offender Recidivism Rates

Type of Program

Percent Change in Recidivism
a
Rates

Vocational education in prison
Intensive supervision: treatment-oriented programs

-9.0%
-16.7%

General education in prison (basic education or post-secondary)

-7.0%

Cognitive-behavioral therapy in prison or community

-6.3%

Drug treatment in community

-9.3%

Correctional industries in prison

-5.9%

Drug treatment in prison

-5.7%

Adult drug courts

-8.0%

Employment and job training in the community

-4.3%

Electronic monitoring to offset jail time
Sex offender treatment in prison with aftercare
Intensive supervision: surveillance-oriented programs
Washington state’s Dangerously Mentally Ill Offender program

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 20

0.0%
-7.0%
0.0%
-20.0%

Programs’ Impact on Offender Recidivism Rates

Type of Program

Percent Change in Recidivism
a
Rates

Drug treatment in jail

-4.5%

Adult boot camps

0.0%

Domestic violence education/cognitive-behavioral treatment

0.0%

Jail diversion for mentally ill offenders

0.0%

Life skills education programs for adults

0.0%

a

This is the average change in recidivism rate compared to recidivism rates of offenders receiving no
treatment or treatment as usual achieved by a typical program in each category. A negative value
indicates a statistically significant reduction in crime; 0 percent means the typical program does not
achieve a statistically significant change in recidivism rates.
Source: Steve Aos, Marna Miller, and Elizabeth Drake. Evidence-Based Public Policy Options to Reduce
Future Prison Construction, Criminal Justice Costs, and Crime Rates. Olympia: Washington State
Institute for Public Policy, 2006.

Chapter 5-B

Texas Has Implemented the Rehabilitation Programs Found to Be
Effective by Research Studies
During the State Auditor’s Office’s review of recidivism studies, auditors
identified the types of programs research indicates are successful at reducing
recidivism. In each case, auditors identified a comparable program currently
operating in Texas. For example, drug treatment in the community is provided
in Texas through programs such as the Treatment Alternatives to Incarceration
program, Substance Abuse Felony Punishment facilities, aftercare for the InPrison Therapeutic Community program, and the Substance Abuse
Counseling Program. Drug treatment in prison is provided through the In
Prison Therapeutic Community program, the Pre-release Therapeutic
Community program, and Pre-release Substance Abuse Program. Vocational
education in prison is provided by the Windham School District, the War
Against Recidivism program, and the Pre-release Therapeutic Community
program.
Table 6 on the next page lists programs found to be effective by the
Washington State Institute for Public Policy and compares those programs to
current programs in Texas.

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 21

Table 6

Texas Programs That Are Comparable to Successful Programs in Other States
Successful
Programs in
Other States

Comparable Texas Programs

Vocational education
in prison

ƒ The Windham School District offers vocational education in prison.
ƒ War Against Recidivism (WAR) program.
ƒ Pre-release Therapeutic Community (PRTC) program.

Intensive supervision:
treatment-oriented
programs

ƒ Substance Abuse Felony Punishment (SAFP) program.
ƒ Substance Abuse Counseling Program/Intermediate Sanction Facility (SACP/ISF).
ƒ Diversion program funding is provided to community supervision correction departments for reduced
caseloads, progressive sanctions, residential treatment and sanction beds, and aftercare caseloads.

ƒ Sex Offender Specialized Caseloads.
General Education in
Prison (basic education
or post-secondary)

ƒ The Windham School District provides general education in prison.

Cognitive-behavioral
Therapy in prison or
community

ƒ
ƒ
ƒ
ƒ
ƒ

Sex Offender Treatment Program (SOTP).

ƒ
ƒ
ƒ
ƒ

Treatment Alternatives to Incarceration (TAIP) program.

Drug Treatment in
community

Pre-release Therapeutic Community (PRTC) program.
Cognitive Intervention program at the Windham School District.
Turning Point Program in District Resource Centers (DRC).
Thinking for a Change program.
Substance Abuse Felony Punishment (SAFP) program.
Aftercare for In-Prison Therapeutic Community (IPTC) program.
Substance Abuse Counseling Program (SACP).

Correctional Industries
in Prison

ƒ Texas Correctional Industries.

Drug Treatment
in Prison

ƒ In-Prison Therapeutic Community program.
ƒ Pre-release Substance Abuse Program (PRSAP).
ƒ Pre-release Therapeutic Community (PRTC) program.

Adult Drug Courts

ƒ Texas currently has 49 adult drug courts operating or planned in 39 counties. Some state funding is
provided through the Community Justice Assistance Division.

Employment and
job training in the
community

ƒ Project RIO (Reintegration of Offenders).

Sex offender
treatment in Prison
with aftercare

ƒ Sex Offender Treatment Program (SOTP).

Washington’s
Dangerously Mentally
Ill Offender program

ƒ This program is in the early stages in Texas and has not yet demonstrated cost savings. The

Department’s Texas Correctional Office on Offenders with Medical or Mental Impairments (TCOOMMI)
program has elements of Washington’s Dangerously Mentally Ill Offender program; but the two
programs do not fully align with each other.

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 22

Chapter 5-C

Research Shows that a Combination of Treatment and Intensive
Supervision Is Effective
Research shows that combining more intensive supervision with drug
treatment is an effective strategy in reducing recidivism. As Table 5 in
Chapter 5-A shows, “intensive supervision: treatment-oriented programs”
have been shown to reduce recidivism by 16.7 percent.
An example of this type of approach is the “Breaking the Cycle” program
implemented at locations in Maryland, Alabama, Florida, and Washington.
Through several actions, including early intervention, graduated sanctions and
incentives, judicial oversight, and collaboration among criminal justice and
drug treatment agencies, these programs encourage offenders on supervision
to abstain from drugs and participate in drug treatment. The outcomes, while
not positive on all measures at all implementation sites, have been
encouraging. Researchers studying the program implemented in Maryland
concluded that the program was an effective strategy for reducing drug arrests
among probationers and parolees with drug-related conditions as part of their
supervision requirements. Additionally, a 2003 study of programs in Alabama,
Florida, and Washington showed that program participants in two of three
sites were less likely than a comparison group to be arrested within a year of
entering the program. The authors of the study speculate that an increase in
enforcement during the years of the study may account for the negative
findings at the third site.
The approach of providing intensive supervision with drug treatment is used
in Texas by both the Department’s Community Justice Assistance Division
and its Parole Division. For probation supervision, the 79th Legislature
appropriated approximately $28 million per year in new diversion program
funds. The funds are allocated to community supervision and corrections
departments that use a progressive sanctions model. The progressive sanctions
model must include reduced and specialized caseloads, increased judicial
involvement in the administration of sanctions and incentives, graduated
sanctions, and various treatment options. The Parole Division’s Substance
Abuse Counseling Program also combines treatment with graduated sanctions.
Chapter 5-D

Capacity Is Limited for Some Rehabilitation Programs in Texas
While Texas has the programs found by research to be effective, program
capacity is not always sufficient. Programs with limited capacity include inprison substance abuse treatment, substance abuse treatment while on
community supervision, and drug courts.
The Sunset Advisory Commission noted in its October 2006 Sunset Advisory
Commission Staff Report that 59 percent of prisoners are chemically
dependent, but only 5 percent of potential program participants are admitted to
An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 23

substance abuse programs each year. Similarly, the Sunset Advisory
Commission reported that 50 percent of probationers on direct supervision
were on probation for drug or alcohol offenses, but only 9 percent received
residential substance abuse treatment, and only 16 percent received out-patient
treatment.
The Department requested an expansion of the Substance Abuse Felony
Punishment (SAFP) program in its Legislative Appropriations Request for the
2008-2009 biennium. In that request, the Department reported that there are
3,250 SAFP treatment beds located in seven facilities across the state. It also
reported that, as of May 2006, there were 1,075 offenders on this program’s
waiting list. The Department is also seeking an expansion of the In-Prison
Therapeutic Community program, treatment for offenders convicted of
driving while intoxicated, and out-patient substance abuse treatment for
probationers.
Drug courts are another example of an effective program that has seen limited
use in Texas. Several studies indicate drug courts in other states have been
effective in reducing recidivism. A 2003 study of New York state adult drug
courts concluded that the six largest and oldest drug courts—in the Bronx,
Brooklyn, Queens, Suffolk, Syracuse, and Rochester—averaged a 29 percent
reduction in recidivism over a three-year, post-arrest period.
A 2005 study by the U.S. Government Accountability Office reported that 10
of 13 drug court programs resulted in statistically significant reductions in
overall re-arrest rates among drug court program participants.1 Participants in
these 10 drug court programs had re-arrest rates ranging from nearly 10
percent to 30 percent lower than those of a comparison, non-participant group.
There were mixed results in one of the three remaining programs; the other
two programs did not show statistically significant reductions in overall rearrest rates.
The Criminal Justice Policy Council (Council) released its evaluation of the
implementation of drug courts in Dallas, Jefferson, and Travis counties in
2003.2 The Council found similar results to those of evaluations of drug
courts in other states: offenders completing the drug court programs in these
three counties had a 28.5 percent re-arrest rate three years after entry,
compared to a 65.1 percent re-arrest rate for those not completing the program
and a 56.8 percent re-arrest rate for those not participating in the program.

1

See Adult Drug Courts: Evidence Indicates Recidivism Reductions and Mixed Results for Other Outcomes, GAO Report
#GAO-05-219, February 2005.

2

See Initial Process and Outcome Evaluation of Drug Courts in Texas, Criminal Justice Policy Council, January 2003.
An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 24

Texas does not use drug courts to the extent that many other states do.
According to a November 2006 American University report3, 47 of Texas’s
254 counties (18.5 percent) had a drug court operation in place or planned,
whereas drug court operations were in place in 1,217 of the nation’s 3,140
counties (38.76 percent). These figures include juvenile, family, and adult
drug courts. According to the Texas Governor’s Office, 49 adult drug courts
currently are operating or planned in 39 Texas counties.
Currently, only counties with a population of more than 550,000 are required
to establish a drug court program. House Bill 530, filed during the 80th
legislative session, would expand the use of drug courts in Texas.

3

See Bureau of Justice Assistance, Drug Court Clearinghouse Project at American University: Summary of Drug Court Activity
by State and County, American University, November 22, 2006.
An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 25

Appendices
Appendix 1

Objectives, Scope, and Methodology
Objectives
The audit objectives were to:
ƒ

Determine whether the Department of Criminal Justice (Department)
collects and maintains sufficient data for measuring the effectiveness of
programs intended to reduce recidivism.

ƒ

Determine the outcomes for participants in selected Department
rehabilitation programs.

ƒ

For the selected programs, determine whether there is a documented
selection process for program participation and that the selection of
participants is consistent with that process.

ƒ

Identify rehabilitation programs in other states that have demonstrated a
high level of success.

Scope
The scope of this audit included the following programs:
ƒ

Sex Offender Treatment Program.

ƒ

Sex Offender Education Program.

ƒ

Pre-Release Substance Abuse Program.

ƒ

Pre-Release Therapeutic Community program.

ƒ

InnerChange Freedom Initiative program.

The scope of recidivism calculations included offenders who:
ƒ

Entered or were eligible but chose not to enter one of the five selected
programs from September 1, 2002, to August 31, 2004;

ƒ

Were expected to complete one of the selected programs during fiscal year
2004; and

ƒ

Were released or paroled during fiscal year 2004.

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 26

Re-arrest calculations included data from the Department of Public Safety
containing all arrests for the selected population from September 1, 2003, to
August 31, 2006.
Re-incarceration calculations included data from the Department containing
all incarcerations from September 1, 2003, to August 31, 2006.
Calculations of Parole Board votes included offenders who received a
“Further Investigation – Rehabilitation” (FI-R) vote between September 1,
2005, and November 30, 2006.
Program testing included all offenders participating in one of the selected
programs or released from one of these programs between September 1, 2005,
and November 30, 2006.
Methodology
The audit methodology consisted of conducting interviews; collecting and
reviewing information; and performing tests, procedures, and analyses against
predetermined criteria for the selected programs.
Information collected and reviewed included the following:
ƒ

Interviews with management and staff of the Department and with current
participants of the selected programs.

ƒ

Documentary evidence such as:

ƒ

Œ

Policies and procedures for programs.

Œ

Prior reports from the Criminal Justice Policy Council, including
Evaluation of the Performance of the Texas Department of Criminal
Justice Rehabilitation Tier Programs, 2001, and The Second Biennial
Report on the Performance of the Texas Department of Criminal
Justice Rehabilitation Tier Programs, February 2003.

Procedures and tests conducted included the following:
Œ

Determined whether offenders receiving FI-R votes were placed in
programs with sufficient time allowed for them to complete the
program before their mandatory release dates.

Œ

Identified which offenders were enrolled in a program other than the
program specified by their FI-R vote type.

Œ

Determined whether any offenders with FI-R votes were released on
parole before completing a program.

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 27

Œ

Determined whether any offenders were still waiting for placement in
a program even though their specified release dates had passed.

Œ

Analyzed database information from the Sex Offender Treatment
Program and the Sex Offender Education Program to determine the
types of offenders participating in these programs.

Œ

Analyzed paroled sex offender database information to determine the
number of paroled sex offenders who participated in the Sex Offender
Treatment Program or the Sex Offender Education Program.

Œ

Analyzed class cancellations to determine what percentage of group
sessions and psychological education classes were canceled in the Sex
Offender Treatment Program.

Œ

Analyzed personnel information for each of the selected programs to
calculate the staff turnover rate for each program.

Œ

Tested a random sample of information for current program
participants from each of the five selected programs to:
y

Determine whether the offenders met the selection criteria for
program enrollment.

y

Determine whether there was sufficient documentation that the
offenders are receiving treatment as designated by the program.

Œ

Tested information for a random sample of offenders who completed
each of the five selected programs to determine whether they met the
completion requirements before being released from the program.

Œ

Tested a random sample of information for offenders who completed
the Sex Offender Treatment Program and who were required to receive
aftercare to determine whether those offenders received the aftercare
as required.

Œ

Analyzed data provided by the Department and the Department of
Public Safety to determine the rates of recidivism, as well as
demographics for:
y

Offenders who participated in and completed one of the five
programs.

y

Offenders who participated in but did not complete one of the five
programs.

y

Offenders who participated in one of the five programs (regardless
of whether they completed a program).

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 28

y

Œ

Offenders who were eligible for but did not participate in one of
the five selected programs.

Visited the following prison units to observe and document program
operations and to determine whether treatment was being provided in
accordance with the daily activity schedules:
y

Thomas Goree Unit – Huntsville, Texas.

y

J. W. Hamilton Unit – Bryan, Texas.

y

L. V. Hightower Unit – Dayton, Texas.

y

Hilltop Unit – Gatesville, Texas.

y

Roger Leblanc Unit – Beaumont, Texas.

y

Carol S. Vance Unit - Richmond, Texas.

Criteria used included the following:
ƒ

The Department’s Substance Abuse Treatment Operations Manual,
revised September 2006.

ƒ

The Department’s Programs and Services Division Sex Offender
Evaluation, Education and Treatment Program Policies and Procedures,
revised April 2002.

ƒ

InnerChange Freedom Initiative Policies and Procedures Manual.

ƒ

The Department’s Parole Division policies and procedures for sex
offender supervision and treatment.

Project Information
Audit fieldwork was conducted from December 2006 through February 2007.
This audit was conducted in accordance with generally accepted government
auditing standards.
The following members of the State Auditor’s staff performed the audit:
ƒ

Sandra Q. Donoho, MPA, CIA, CISA, CFE (Project Manager)

ƒ

Sherry Sewell, CGAP (Assistant Project Manager)

ƒ

Kirby Cossey

ƒ

Olin Davis, MBA

ƒ

Arby Gonzales

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 29

ƒ

Lucien Hughes

ƒ

Amadou N’gaide, MBA

ƒ

Ashley Rutherford

ƒ

Sajil Scaria

ƒ

Michael Stiernberg

ƒ

Lisa Thompson

ƒ

Jennifer Wiederhold, CGAP

ƒ

J. Scott Killingsworth, CIA, CGFM (Quality Control Reviewer)

ƒ

Lisa R. Collier, CPA (Audit Manager)

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 30

Appendix 2

Comparison of Recidivism Results
Tables 7 and 8 show the recidivism rates calculated in 2007 by the State
Auditor’s Office and calculated in 2003 by the Criminal Justice Policy
Council (Council).
Table 7

Re-arrest Rates for Program Participants and Control Groups
2007 State Auditor’s Office Results and 2003 Criminal Justice Policy Council Results
Pre-Release
Substance Abuse
Program

Group

SAO
Results
(2007)

Council
Results
(2003)

Pre-Release
Therapeutic
Community
SAO
Results
(2007)

Council
Results
(2003)

Sex Offender
Education Program

Sex Offender
Treatment Program

InnerChange
Freedom Initiative

SAO
Results
(2007)

SAO
Results
(2007)

SAO
Results
(2007)

Council
Results
(2003)

Council
Results
(2003)

Council
Results
(2003)

Completers

38.1%

35.0%

18.2%

37.1%

19.4%

n/a

a

16.5%

13.1%

19.3%

17.3%

Non-Completers

51.7%

36.8%

50.0%

45.4%

26.7%

n/a

a

26.7%

20.2%

39.5%

50.0%

All Program
Participants

38.3%

35.2%

18.9%

40.3%

20.0%

n/a

a

22.4%

18.5%

36.2%

36.2%

Control Group

31.4%

40.8%

42.6%

38.4%

22.4%

n/a

a

17.7%

21.2%

30.6%

34.9%

a

The Sex Offender Education Program was not evaluated by the Criminal Justice Policy Council in 2003.

Sources: 2007 State Auditor’s Office’s results were calculated using unaudited data provided by the Department of Criminal Justice. The 2003 Criminal
Justice Policy Council’s Results were obtained from The Second Biennial Report on the Performance of the Texas Department of Criminal Justice
Rehabilitation Tier Programs, February 2003.

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 31

Table 8

Re-incarceration Rates for Program Participants and Control Groups
2007 State Auditor’s Office Results and 2003 Criminal Justice Policy Council Results
Pre-Release
Substance Abuse
Program

Group
Completers
Technical Violations
Non-Completers
Technical Violations
All Program
Participants
Technical Violations
Control Group
Technical Violations
a
b

SAO
Results
(2007)

Pre-Release
Therapeutic
Community

Council
Results
(2003)

SAO
Results
(2007)

Council
Results
(2003)

Sex Offender
Education Program

Sex Offender
Treatment Program

InnerChange
Freedom Initiative

SAO
Results
(2007)

SAO
Results
(2007)

SAO
Results
(2007)

Council
Results
(2003)

19.7%

21.0%

18.7%

19.9%

12.3%

n/a

13.7%

n/a

a

18.9%

n/a

a

52.6%

n/a

24.1%

19.5%

15.0%

28.9%

6.7%

n/a

0.0%

n/a

a

0.0%

n/a

a

100.0%

n/a

19.8%

21.0%

18.6%

23.4%

11.8%

n/a

13.5%

n/a

a

18.6%

n/a

a

78.6%

n/a

15.1%

29.0%

26.7%

21.9%

15.4%

n/a

23.0%

n/a

a

31.5%

n/a

a

48.7%

n/a

b

a
b

Council
Results
(2003)

Council
Results
(2003)

5.9%

16.7%

14.0%

8.0%

80.0%

n/a

a

25.1%

n/a

a

6.0%

19.0%

14.8%

36.3%

a

28.6%

n/a

a

18.6%

n/a

b

6.0%

18.5%

14.7%

24.3%

a

50.0%

n/a

a

19.6%

n/a

b

15.3%

27.7%

12.3%

22.3%

a

58.7%

n/a

a

30.9%

n/a

The Criminal Justice Policy Council did not break out technical violations that caused violated to be re-incarcerated.
The Sex Offender Education Program was not evaluated by the Criminal Justice Policy Council in 2003.

Sources: 2007 State Auditor’s Office’s results were calculated using unaudited data provided by the Department of Criminal Justice. The 2003 Criminal
Justice Policy Council’s Results were obtained from The Second Biennial Report on the Performance of the Texas Department of Criminal Justice
Rehabilitation Tier Programs, February 2003.

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 32

a

a

a

Appendix 3

Demographics of Five Selected Programs and Methodology for
Calculating Recidivism Rates
Tables 9 through 13 show the demographics of the test and control groups the
State Auditor’s Office used to calculate program recidivism results using data
provided by the Department of Criminal Justice (Department).
Table 9

Sex Offender Treatment Program
a
Demographics of Test and Control Groups
Demographic
Characteristic

Test Group
(N=201)

Control Group
(N=302)

Gender
Male

97.5%

98.7%

2.5%

1.3%

Hispanic

20.9%

21.0%

Black

33.3%

33.3%

45.3%

45.3%

0.5%

0.3%

Female
Race

White
Other

b

Age Entered Prison
35 and Under

46.3%

34.3%

Over 35

53.7%

65.7%

82.6%

100.0%

Drug

7.0%

0.0%

Property

8.0%

0.0%

Other

2.5%

0.0%

18.9%

5.7%

Offense Type
Violent

Release Code
Parole
Mandatory Supervision

25.4%

31.0%

Discharge

51.8%

44.7%

4.0%

18.7%

Other
a
b

Because of rounding, the columns may not total 100 percent.
Race codes I (Indian), O (Other), A (Asian) and U (Unknown) are counted under “Other.”

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 33

Table 10

Sex Offender Education Program
a
Demographics of Test and Control Groups
Demographic
Characteristic

Test Group
(N=170)

Control Group
(N=1,630)

Gender
Male

100.0%

100.0%

0.0%

0.0%

Hispanic

22.9%

22.8%

Black

30.0%

29.9%

47.1%

47.2%

0.0%

0.0%

Female
Race

White
Other

b

Age Entered Prison
35 and Under

27.7%

36.6%

72.4%

63.4%

82.4%

100.0%

Drug

8.8%

0.0%

Property

5.3%

0.0%

Other

3.5%

0.0%

39.4%

4.3%

Over 35
Offense Type
Violent

Release Code
Parole
Mandatory Supervision

17.1%

26.8%

Discharge

38.8%

48.4%

4.7%

20.5%

Other
a
b

Because of rounding, the columns may not total 100 percent.
Race codes I (Indian), O (Other), A (Asian) and U (Unknown) are counted under “Other.”

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 34

Table 11

InnerChange Freedom Initiative
a
Demographics of Test and Control Groups
Demographic
Characteristic

Test Group
(N=369)

Control Group
(N=554)

Gender
Male

100.0%

100.0%

0.0%

0.0%

Hispanic

15.7%

15.7%

Black

42.8%

42.9%

41.5%

41.4%

0.0%

0.0%

35 and Under

67.2%

55.2%

Over 35

32.5%

44.9%

Violent

33.6%

26.0%

Drug

36.0%

38.7%

Property

18.4%

22.8%

6.8%

12.5%

Parole

36.3%

33.3%

Mandatory Supervision

31.4%

40.9%

Discharge

25.5%

10.9%

6.8%

15.0%

Female
Race

White
Other

b

Age Entered Prison

Offense Type

Other
Release Code

Other

Geographic Locations
Houston Area

44.4%

51.2%

Dallas Area

22.8%

48.8%

Other Area

32.8%

0.0%

a
b

Because of rounding, the columns may not total 100 percent.
Race codes I (Indian), O (Other), A (Asian) and U (Unknown) are counted under “Other.”

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 35

Table 12

Pre-Release Therapeutic Community
a
Demographics of Test and Control Groups
Demographic
Characteristic

Test Group
(N=898)

Control Group
(N=1,347)

Gender
Male

100.0%

100.0%

0.0%

0.0%

9.0%

9.0%

47.4%

47.5%

43.1%

43.1%

0.5%

0.5%

35 and Under

41.0%

45.8%

Over 35

59.0%

54.2%

Violent

21.3%

23.2%

Drug

47.8%

46.9%

Property

17.3%

21.8%

Other

13.7%

8.1%

95.6%

100.0%

Mandatory Supervision

1.5%

0.0%

Discharge

0.1%

0.0%

Other

2.9%

0.0%

Female
Race
Hispanic
Black
White
Other

b

Age Entered Prison

Offense Type

Release Code
Parole

a
b

Because of rounding, the columns may not total 100 percent.
Race codes I (Indian), O (Other), A (Asian) and U (Unknown) are counted under “Other.”

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 36

Table 13

Pre-Release Substance Abuse Program
Demographics of Test and Control Groups
Demographic
Characteristic

Test Group
(N=1,729)

a

Control Group
(N=2,594)

Gender
Male

100.0%

100.0%

0.0%

0.0%

Hispanic

10.5%

10.5%

Black

43.8%

43.8%

45.5%

45.5%

0.2%

0.2%

35 and Under

42.1%

47.5%

Over 35

57.9%

52.5%

Violent

17.8%

24.4%

Drug

49.8%

21.7%

Property

17.6%

46.6%

6.8%

7.3%

96.1%

100.0%

Mandatory Supervision

1.5%

0.0%

Discharge

0.2%

0.0%

Other

2.3%

0.0%

Female
Race

White
Other

b

Age Entered Prison

Offense Type

Other
Release Code
Parole

Chemical Dependency
Inpatient

21.1%

63.3%

Outpatient

74.1%

36.7%

4.9%

0.0%

Other
a
b

Because of rounding, the columns may not total 100 percent.
Race codes I (Indian), O (Other), A (Asian) and U (Unknown) are counted under “Other.”

Recidivism Calculation Methodology
The following methodology was developed with assistance from the
Department and was agreed upon before auditors calculated the recidivism
rates.
Data Collection

Auditors obtained demographic information from the Department for:
An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 37

ƒ

All offenders who participated in at least one of the five selected programs
from September 1, 2002, to August 31, 2004, were expected to complete
the program during fiscal year 2004, and were released from prison during
fiscal year 2004. This included both offenders who completed and did not
complete the programs.

ƒ

All offenders who were eligible for at least one of the five selected
programs from September 1, 2002, to August 31, 2004, (and would have
been expected to complete the program during fiscal year 2004 if they
participated) but did not participate in the program and were released
during fiscal year 2004.

In cases in which an offender had more than one release date in a fiscal year,
auditors used only the record from the first release.
The requested information contained the following details about each
offender:
ƒ

First name.

ƒ

Last name.

ƒ

Middle initial.

ƒ

State Identification (SID) number.

ƒ

The offender’s Department of Criminal Justice identification number.

ƒ

Program type

ƒ

Date of birth.

ƒ

Race.

ƒ

Gender.

ƒ

Release date.

ƒ

Release type.

ƒ

Offense type.

ƒ

Custody level.

ƒ

Program completion date.

ƒ

County of residence (for InnerChange Freedom Initiative program
participants only).

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 38

ƒ

Substance abuse assessment score (for Pre-Release Therapeutic
Community program and Pre-Release Substance Abuse Program
participants only).

Auditors also obtained data from the Department on all admissions to prison
or state jails, not including Substance Abuse Felony Punishment facilities or
Intermediate Sanction Facilities, from September 1, 2003, to August 31, 2006.
This data was used to calculate re-incarceration rates. All cases in which an
offender was received on the same date as his or her release were excluded
because these were transfers to another division.
Auditors compiled a list of the offenders in the population by program,
including both the test group and the control group, using the offender’s SID
number, last name, first name, race, gender, and date of birth as identifiers.
Criminal history records for each offender, provided by the Department of
Public Safety, were used to calculate re-arrest rates.
Data Integrity Testing

Auditors performed selected testing of the data obtained from standalone
computer databases at the Department program level to ensure the
completeness and reliability of the data. These tests included matching a
sample of data to hard-copy files at the program level and/or to data on the
Department’s mainframe. Auditors relied on prior audit testing of the
Department’s mainframe, which found that the mainframe data was generally
reliable.
Data Analysis

The collected data was sorted into the following test and control groups:
ƒ

All offenders released from prison in 2004 who participated in and
completed a selected program.

ƒ

All offenders released from prison in 2004 who participated in but did not
complete a selected program.

ƒ

All offenders released from prison in 2004 who qualified for a selected
program but did not participate in one. This was the control group.

Recidivism Calculation

For each selected program, auditors used the population of offenders released
during fiscal year 2004 to calculate the recidivism rate for each group as
follows:
ƒ

Calculated the re-arrest and re-incarceration rates for offenders who
completed a selected program by calculating the percent of these offenders

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 39

who were released in 2004 and were re-arrested or re-incarcerated during
the two years following their release date.
ƒ

Calculated re-arrest and re-incarceration rates for offenders who started
but did not complete a selected program by calculating the percent of these
offenders who were released in 2004 and were re-arrested or reincarcerated during the two years following their release date.

ƒ

Calculated the re-arrest and re-incarceration rates for offenders who
participated in a selected program (including completers and noncompleters) by calculating the percent of these offenders who were
released in 2004 and were re-arrested or re-incarcerated during the two
years following their release date.

ƒ

Calculated re-arrest and re-incarceration rates for offenders who were
eligible for a selected program but never participated in the program by
calculating the percent of these offenders who were released in 2004 and
were re-arrested or re-incarcerated during the two years following their
release date.

ƒ

Calculated the rate of offenders who were re-incarcerated for technical
violations by calculating the percent of offenders who were released in
2004 and re-incarcerated during the two years following their release date
but were not re-arrested.

ƒ

Calculated the re-arrest and re-incarceration rates by comparing the test
and control groups of offenders from each program to Department
admissions data and the Department of Public Safety criminal history
records. This included:
Œ

Determining the number of months between an offender’s release date
and any re-admission to the Department’s prisons.

Œ

Determining the number of months between an offender’s release date
and any subsequent arrest.

These calculations included only those re-admissions and arrests that occurred
within 24 months of the offender’s release from prison.
For each selected program, auditors created a table to compare the re-arrest
and re-incarceration rates, calculated the variances among different groups,
and drew conclusions on the following groups:
ƒ

Offenders who successfully completed a selected program. (For the
InnerChange Freedom Initiative program, auditors considered six months
of participation in the aftercare component to be successful program
completion.)

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 40

ƒ

Offenders who started but did not complete a selected program.

ƒ

Offenders who participated in a selected program, regardless of whether
they completed the program.

ƒ

Offenders who were eligible for a selected program but did not participate.
(This group was used as a control group for comparison purposes.)

ƒ

Offenders from the above groups (completers, non-completers, and
eligible but did not participate) who were re-incarcerated but not rearrested. (This situation was considered a technical violation for purposes
of this audit.)

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 41

Appendix 4

Other Recidivism Studies
Recent program evaluations address recidivism for various offender programs.
Summaries of three reports are included in this appendix. The three reports
cover the Windham School District, which is the Department of Criminal
Justice’s (Department) educational component for offender education; Project
Rio, which is a job placement program for offenders that is offered through
the Texas Workforce Commission; and the Substance Abuse Felony
Punishment Facilities and In Prison Therapeutic Community programs.
Windham School District Evaluation Report, Legislative Budget Board, January
2007.

Evaluations of training services provided by the Windham School District
were mandated by House Bill 2837 (79th Legislature, Regular Session). The
Legislative Budget Board was required to report the evaluation results to the
Legislature. House Bill 2837 specifically required the Department and the
Windham School District to address the type of training services provided, the
type of employment obtained upon release, whether employment was related
to training received, the difference between earnings on the date employment
is obtained and on the first anniversary of that date, and employment retention
factors.
According to the Department, during the 2006-2007 biennium, $11.2 million
was budgeted for vocational training programs and approximately 3,500
offenders participated in vocational training each day. As of December 6,
2006, 15,264 offenders were pending enrollment in vocational programs.
Between April 1, 2004, and March 31, 2005, 64,364 offenders were released
from the Department’s facilities. Of those released, 31,429 (49 percent) were
excluded from the study for various reasons. The Windham School District
tracked the remaining 32,935 offenders for employment upon release,
occupation, and earnings. About 14 percent of offenders, or 4,747 offenders
in the study, were in the vocational group. Examples of the study’s
conclusions included the following:
ƒ

Employment status after release. Of the offenders in the study, 4,747 were
classified in the vocational group and 28,188 were in the non-vocational
group. About 68 percent of the vocational group (compared with 58
percent of the non-vocational group) were employed within one year of
release.

ƒ

Training related to future occupation.

Of the 3,208 vocationally trained
offenders who were employed within one year of release, 2103 (66
percent) were employed in an occupation related to their vocational
training.

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 42

ƒ

First anniversary of employment.

ƒ

Earnings.

ƒ

Retention factors.

The Windham School District tracked
offenders through one year of employment, from the date of initial
employment through the first anniversary of that date. Of the vocational
group, 53 percent were still employed on their first anniversary (compared
with 46 percent of the non-vocational group).
The Windham School District compared the first- and fourthquarter earnings of the employed vocational and non-vocational groups
and calculated the average salary difference. Approximately 41 percent of
the employed vocational group received an earnings increase during the
year (compared with 34 percent for the non-vocational group).

Œ

Œ

Œ

These were divided into the following categories:

Percent of offenders who retained employment for three consecutive
quarters by age group for the vocational group and the non-vocational
group. This category was further divided into the following groups:
y

Offenders younger than 25 years of age: 55.9 percent for the
vocational group (compared with 44.4 percent for the nonvocational group).

y

Offenders between 25 and 34 years of age: 59.1 percent for the
vocational group (compared with 49 percent for the non-vocational
group).

y

Offenders 35 years of age and older: 54.6 percent for the
vocational group (compared with 40 percent for the non-vocational
group).

Percent of offenders who retained employment for three consecutive
quarters by educational level for the vocational group and the nonvocational group. This category was further divided into the following
groups:
y

No GED/high school diploma: 49.5 percent for the vocational
group (compared with 43.4 percent for the non-vocational group).

y

GED/high school diploma: 57 percent for the vocational group
(compared with 48.4 percent for the non-vocational group).

Percent of employed vocational group with industry certification who
retained employment for three consecutive quarters: 61.7 percent
retained employment and 38.3 percent did not.

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 43

Project Reintegration of Offenders, Long-Term Results for 2000-2001 Exit
Cohorts.

The following is a summary of a program evaluation of Project Rio performed
by the Texas Workforce Commission. This summary was written by the
Texas Workforce Commission and is reprinted with its permission:
Project Reintegration of Offenders (Project RIO) is designed to reduce
recidivism rates by assisting ex-offenders [to] find employment. The program
is jointly administered by the Department [of Criminal Justice], the Texas
Workforce Commission, the Youth Commission, and the Windham School
District. The program equips ex-offenders and adjudicated youth with the
necessary skills, attitudes, and abilities to re-enter the labor market, and guides
them toward post-release job opportunities.
To evaluate the effectiveness of Project RIO, the Texas Workforce
Commission obtained data from the Department [of Criminal Justice] to
produce a longitudinal study covering 60,000 individuals released between
July 1, 2000, and June 30, 2001. The entered employment rate was based on
the Texas Workforce Commission’s unemployment insurance wages for the
fourth quarters of 2001, 2003, and 2005. Recidivism data was obtained from
the Department [of Criminal Justice]. Records were divided into four cohorts
based on the extent of participation in Project RIO:
ƒ

Texas Workforce Commission Post-Release RIO. (Offenders received
Project RIO services only after their release.)

ƒ

Pre-Post RIO. (Offenders received Project RIO services both before and
after their release.)

ƒ

Pre-Release RIO. (Offenders received Project RIO services only before
their release.)

ƒ

Non-RIO. (Offenders did not receive Project RIO services.)

The study’s findings demonstrate that Project RIO participation by offenders
results in higher rates of employment and dramatically lower rates of
recidivism, with the best outcomes occurring when individuals participate in
Project RIO both during incarceration and after release.
Employment.

The entered employment rate for the Pre-Post RIO cohort is
approximately twice that of the Non-RIO cohort for one, three, and five years
post-release. The Texas Workforce Commission Post-Release RIO cohort
produced superior results to the Pre-Release RIO cohort, but even the PreRelease RIO cohort entered employment rates one-fourth higher than the NonRIO cohort rates. [See Table 14.]

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 44

Table 14

Employment Rates for Project RIO
Percent Employed
Categorical Label

2001

2003

2005

TWC Post-Release Rio

49.5%

34.7%

34.0%

Pre-Post RIO

63.1%

46.0%

40.3%

Pre-Release RIO

44.8%

33.7%

30.8%

Non-RIO

32.3%

22.4%

24.0%

42.8%

30.5%

29.9%

Totals

Recidivism.

Ex-offenders in the Pre-Post RIO cohort, compared against the
Non-RIO cohort, are one-fifth as likely to recidivate in their first year after
release and less than one-fourth as likely in their first three years. The PreRelease RIO cohort produced superior results to the Texas Workforce
Commission Post-Release RIO cohort, but both groups have a recidivism rate
that is less than half of the Non-RIO cohort throughout the first three postrelease years. [See Table 15.]
Table 15

Recidivism Rates for Project RIO
Recidivism Rate

2001

2003

2005

TWC Post-Release Rio

7.6%

11.7%

Not yet available

Pre-Post RIO

2.1%

6.0%

Not yet available

Categorical Label

Pre-Release RIO

4.6%

9.4%

Not yet available

Non-RIO

10.3%

25.4%

Not yet available

7.6%

16.1%

Not yet available

Totals

Of all the populations served by the Texas workforce system, ex-offenders
have the greatest barriers to employment:
ƒ

Most ex-offenders have low levels of educational attainment and work
experience.

ƒ

Dozens of occupations have statutory prohibitions against the employment
of ex-offenders.

ƒ

There is a strong social stigma attached to ex-offenders.

The longitudinal data clearly demonstrates that the most effective response to
the unique obstacles facing ex-offenders involves the intensive and
specialized range of services provided through Project RIO.
An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 45

Outcome Evaluation of Offenders Released from the SAFP and IPTC Programs
in FY 2004, January 2007, The Community Justice Assistance Division,
Department of Criminal Justice.

According to this study, the state budget shortfall in 2003 resulted in the
following:
ƒ

The length of In Prison Therapeutic Community (IPTC) and Substance
Abuse Felony Punishment (SAFP) in-patient facility programs was
reduced from nine months to six months. (Special-needs offender
programs remained at nine months.)

ƒ

Some SAFP facilities were converted to state jail/transfer facilities.

ƒ

Post-release residential treatment capacity in Transitional Treatment
Centers was reduced.

ƒ

Other substance abuse treatment services were reduced.

ƒ

Waiting lists for both IPTC and SAFP programs and delays in release to
Transitional Treatment Centers may have contributed to prison capacity
issues by:
Œ

Delaying releases of offenders required to complete the IPTC program
due to a waiting list.

Œ

Reducing sanction options for probationers due to an SAFP waiting
list.

Œ

Delaying releases of IPTC and SAFP participants due to lack of
Transitional Treatment Center capacity.

ƒ

Outcomes for the IPTC and SAFP programs were slightly worse than the
previous study of fiscal year 2000 IPTC and SAFP releases, although the
differences were not statistically significant.

ƒ

The IPTC program continues to demonstrate a significant reduction in
recidivism rates for participants.

ƒ

The SAFP program does not demonstrate a reduction in recidivism due to
the high percentage of technical revocations when compared to the IPTC
program.

ƒ

Offenders who complete treatment in IPTC and SAFP programs have
significantly lower recidivism rates than offenders who do not complete
treatment.

ƒ

The use of SAFP re-entry drug courts in Dallas appears to significantly
reduce the revocation rate of offenders who participate in that program.

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 46

ƒ

ƒ

The waiting list for the IPTC program can delay the release of some
offenders who are required to complete the program before they can be
released from prison. The numbers of male and female offenders on the
IPTC waiting list as of December 2006 were:
Œ

Males: 109

Œ

Females: 65

In December 2006, 389 offenders in the SAFP and IPTC programs were
waiting to be placed in a Transitional Treatment Center due to capacity
limitations. Because there is a waiting list to be placed in a Transitional
Treatment Center, beds in the in-prison portion of the program are being
occupied by offenders who have completed the in-prison portion, causing
the waiting list for admission into the Transitional Treatment Centers
program to increase.

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 47

Appendix 5

Recent State Auditor’s Office Work
Recent SAO Work
Number

Product Name

Release Date

06-022

An Audit of the Criminal Justice Information System

February 2006

07-005

An Audit Report on Performance Measures at Five State Agencies

December 2006

An Audit Report on Selected Rehabilitation Programs at the Department of Criminal Justice
SAO Report No. 07-026
March 2007
Page 48

Copies of this report have been distributed to the following:

Legislative Audit Committee
The Honorable David Dewhurst, Lieutenant Governor, Joint Chair
The Honorable Tom Craddick, Speaker of the House, Joint Chair
The Honorable Steve Ogden, Senate Finance Committee
The Honorable Thomas “Tommy” Williams, Member, Texas Senate
The Honorable Warren Chisum, House Appropriations Committee
The Honorable Jim Keffer, House Ways and Means Committee

Office of the Governor
The Honorable Rick Perry, Governor

Department of Criminal Justice
Members of the Board of Criminal Justice
Ms. Christina Melton Crain, Chairman
Mr. Pierce Miller, Vice-chair
Ms. Patricia A. Day, Secretary
Mr. Adrian A. Arriaga
Mr. Oliver J. Bell
Mr. Gregory S. Coleman
Pastor Charles Lewis Jackson
Mr. Tom Mechler
Mr. Leopoldo Vasquez III
Mr. Brad Livingston, Executive Director

This document is not copyrighted. Readers may make additional copies of this report as
needed. In addition, most State Auditor’s Office reports may be downloaded from our Web
site: www.sao.state.tx.us.
In compliance with the Americans with Disabilities Act, this document may also be requested
in alternative formats. To do so, contact our report request line at (512) 936-9880 (Voice),
(512) 936-9400 (FAX), 1-800-RELAY-TX (TDD), or visit the Robert E. Johnson Building, 1501
North Congress Avenue, Suite 4.224, Austin, Texas 78701.
The State Auditor’s Office is an equal opportunity employer and does not discriminate on the
basis of race, color, religion, sex, national origin, age, or disability in employment or in the
provision of services, programs, or activities.
To report waste, fraud, or abuse in state government call the SAO Hotline: 1-800-TX-AUDIT.