Federal Prison Inmates: Rehabilitative Needs and Program Participation

November 19, 2012 (R41525)

Contents

Tables

Appendixes

Summary

The stated mission of the Bureau of Prisons (BOP) is "to protect society by confining offenders in the controlled environments of prisons and community-based facilities that are safe, humane, cost-efficient, and appropriately secure, and that provide work and other self-improvement opportunities to assist offenders in becoming law-abiding citizens." In support of this mission, BOP offers a variety of rehabilitative programs, including work opportunities through the Federal Prison Industries (FPI), occupational education programs, literacy/GED courses, and a variety of drug abuse treatment programs.

CRS used data from the 1997 and 2004 Survey of Inmates in State and Federal Correctional Facilities to analyze (1) whether there has been an increase in the proportion of federal inmates in need of rehabilitative programming, (2) if inmates with a need for rehabilitative programming were more likely than other inmates to participate in a program to address their need, and (3) whether there was an increase between 1997 and 2004 in the probability that inmates with a reported need participated in rehabilitative programming.

Some of the key observations based on the CRS analysis include the following: (1) data from the 1997 and 2004 surveys show that it was likely that there was increased need for drug abuse treatment programs, but the need for literacy/GED and occupational education programs along with FPI work assignments remained flat; (2) with the exception of inmates who were unemployed before being arrested, inmates who had a rehabilitative need were significantly more likely than other inmates to participate in a rehabilitative program to address their need; (3) there was not a significant difference in the likelihood that inmates in 1997 and 2004 with rehabilitative needs participated in programming; and (4) the probability that a typical inmate with a rehabilitative need participated in a program to address that need was, in most cases, less than 1 in 2, and the probability that the highest-participating inmates were involved in rehabilitative programs was 3 in 5 or lower.

Potential issues relevant to this analysis include the amount of resources available to BOP to carry out its mission to provide rehabilitative programming to federal inmates and the structure of incentives for inmates to participate in rehabilitative programs.


Federal Prison Inmates: Rehabilitative Needs and Program Participation

Introduction

The stated mission of the Bureau of Prisons (BOP) is "to protect society by confining offenders in the controlled environments of prisons and community-based facilities that are safe, humane, cost-efficient, and appropriately secure, and that provide work and other self-improvement opportunities to assist offenders in becoming law-abiding citizens" (emphasis added).1 Rehabilitative programs are designed not only to prevent inmate idleness, but to also form a part of an effective reentry strategy for inmates.2

CRS used data from the 1997 and 2004 Survey of Inmates in State and Federal Correctional Facilities to analyze (1) whether there has been an increase in the proportion of federal inmates in need of rehabilitative programming, (2) if inmates with a need for rehabilitative programming were more likely than other inmates to participate in a program to address their need, and (3) whether there was an increase between 1997 and 2004 in the probability that inmates with a reported need participated in rehabilitative programming.

Potential issues relevant to this analysis include the amount of resources available to BOP to carry out its mission to provide rehabilitative programming to federal inmates and the structure of incentives for inmates to participate in rehabilitative programs.

The report starts with a brief overview of BOP, followed by a description of BOP's main rehabilitative programs. Next, there is a short review of the research on the effectiveness of rehabilitative programming. The report then turns to the analysis described above. The report concludes with a discussion of selected issues for Congress.

The Bureau of Prisons (BOP)

BOP is a component of the Department of Justice (DOJ), and it is responsible for the custody and care of all federal prisoners.3 BOP was established in 1930 to house federal inmates, to professionalize the prison service, and to ensure consistent and centralized administration of the federal prison system.4 At the end of 1930, BOP operated 14 facilities for just over 13,000 inmates.5 By 1940, BOP opened 10 additional prisons while the federal prison population grew by approximately 11,000 inmates. According to BOP, the number of inmates did not change significantly between 1940 and 1980.6 However, during this time period the number of prisons operated by BOP almost doubled as it moved from operating large prisons confining inmates of many security levels to operating smaller prisons that confined inmates with similar security needs.

BOP currently operates 117 correctional facilities, housing approximately 178,000 inmates.7 BOP reports that approximately 82% of inmates under its jurisdiction are housed in BOP-operated facilities; the remainder are housed in contract facilities (i.e., privately managed secure or community-based facilities or local jails).8 All BOP correctional facilities are classified according to one of five different security levels: minimum, low, medium, high, or administrative. A facility's security level is based on the facility's features, which include the presence of external patrols, towers, security barriers, or detection devices; the type of housing within the institution; internal security features; and the staff-to-inmate ratio.9

BOP also contracts with Residential Re-entry Centers (RRCs) (i.e., halfway houses) to provide assistance to inmates nearing release.10 RRCs are intended to provide inmates with a structured and supervised environment along with employment counseling, job placement services, financial management assistance, and other programs and services.11 RRCs facilitate inmates' efforts at reestablishing ties to the community while allowing correctional staff to supervise inmates' activities.12 According to BOP, its goal is to place inmates in RRCs for the amount of time necessary to provide the greatest likelihood of successful reentry into the community.

BOP's Rehabilitative Programs

BOP provides a variety of rehabilitative programs for federal prison inmates, three of which are the focus of this report: work assignments through the Federal Prison Industries (FPI),13 educational programs, and substance abuse treatment. Each program is discussed in more detail below.

Federal Prison Industries (FPI) 14

It is BOP's policy that all sentenced inmates are required to work if they are medically able. According to BOP, work programs assist with inmate management by reducing idleness and they teach inmates skills and the importance of a good work ethic. Inmates can work in either institutional work assignments or FPI factories.15 For both institutional work assignments and FPI factory jobs, a high school diploma or a General Equivalency Diploma (GED) is required for all work assignments above entry level.

Inmates working in FPI jobs are paid between $0.23 and $1.15 per hour. All inmates or detainees may be considered for an FPI job unless the inmate is a pretrial inmate or is currently under an order of deportation, exclusion, or removal. Inmates working in FPI factories are also eligible, as earnings permit, to participate in training that relates directly to the inmate's job assignment. Inmates selected to participate in FPI-funded job training must have enough time left on their sentences to allow them to complete the training. FPI also provides scholarships to allow inmates to begin, or continue, postsecondary business and industry courses or vocational training. According to BOP, 9.0% of inmates had an FPI work assignment in FY2010.

Education Programs

BOP provides inmates with a variety of education programs, including the Literacy/GED program, the Occupational Education program, the English-as-a-Second Language program, and the Postsecondary Education program. This report focuses on the Literacy/GED and Occupational Education programs because they are the two programs inmates are the most likely to participate in.

Literacy/GED16

Under current law, BOP is required to both provide a functional literacy program for all mentally capable inmates who are not functionally literate and to offer literacy/GED programs for inmates who have not earned a high school diploma or its equivalent.17 In addition, federal inmates are required to make satisfactory progress toward earning a high school diploma or a GED in order to earn their full allotment of good time credit.18

BOP's literacy program is designed to help inmates develop reading, math, and writing skills, and to prepare inmates for earning a GED credential. With few exceptions (e.g., pretrial inmates, inmates committed for the purpose of observation, inmates determined by the staff to be temporarily unable to participate in the literacy program due to circumstances beyond their control, and sentenced deportable aliens), all sentenced inmates without a high school diploma or a GED are required to enroll in the adult literacy program for a minimum of 240 hours or until a GED is earned, whichever comes first.

After completing the required 240 hours of instructional time in the literacy program, an inmate may elect to discontinue participation in the literacy program.19 However, if an inmate chooses to opt out of the literacy program after completing 240 hours of instructional time, it can affect the inmate's ability to earn the full allotment of good time credit. Also, not continuing in the literacy program, or not earning a GED, can limit the type of work assignments an inmate can receive. According to BOP, 12.3% of federal inmates participated in a literacy/GED course in FY2010.

Occupational Education Program20

BOP provides an occupational education program at nearly all of its correctional facilities. To participate in an occupational education program, an inmate must have a high school diploma, have earned a GED, or be concurrently enrolled in a GED program.

Occupational education programs may include the following types of training:

Inmates may be required to make a co-payment to enroll and participate in an occupational education program. Co-payments are limited to the cost of books for the courses. According to BOP, 6.4% of federal inmates participated in an occupational education course in FY2010.

Drug Abuse Treatment Programs21

Current law (18 U.S.C. §3621) requires BOP to provide, subject to appropriations, residential substance abuse treatment22 and appropriate aftercare23 for all eligible prisoners.24 Prisoners who are convicted of nonviolent crimes who successfully complete a residential substance abuse treatment program can have their sentence reduced by not more than one year.25

BOP's policy is to interview all new inmates to screen them for drug abuse problems. BOP also reviews each inmate's records for evidence of drug abuse. Based on the interview and record review, BOP staff makes a drug education/treatment referral. BOP offers four types of drug abuse treatment to federal inmates: (1) a drug abuse education course, (2) non-residential drug abuse treatment, (3) residential drug abuse treatment, and (4) community transition drug abuse treatment. Each program is discussed in more detail below.

Drug Abuse Education Course

A drug abuse education course is offered at all BOP institutions. Inmates are required to participate in a drug abuse education course when they have been sentenced or returned to custody because of a parole or supervised release violation after September 30, 1991, and it is determined that (1) there is evidence in the presentence investigation report that alcohol or other drug abuse contributed to the offense; (2) alcohol or other drug use was the reason for the violation of supervised release, parole, or conditions of RRC placement or home confinement; or (3) the inmate was recommended for drug abuse treatment programming by the sentencing judge.

The course informs inmates about the relationship between drug abuse and crime and the psychological, physical, and social impact of abusing drugs and alcohol. It is also designed to motivate inmates who need additional drug abuse treatment to participate in non-residential or residential drug abuse treatment. Inmates not required to participate in the drug abuse education course may volunteer to participate when space is available. According to BOP, 27.6% of inmates participated in a drug abuse education treatment program in FY2010.

Non-residential Drug Abuse Treatment

Like the drug abuse education course, non-residential drug abuse treatment is available at all BOP institutions. Participation in the institution's non-residential drug abuse treatment program is voluntary. According to BOP, non-residential drug abuse treatment is designed with the flexibility necessary to meet the treatment needs of the inmates at the particular institution.

Specific populations targeted for non-residential drug abuse treatment include the following:

According to BOP, 8.4% of federal inmates participated in a non-residential drug abuse treatment program in FY2010.

Residential Drug Abuse Treatment (RDAP)

BOP offers residential drug abuse treatment in 61 institutions. To be eligible for residential treatment, an inmate must (1) have a verifiable and documented drug abuse problem, (2) have no serious mental impairment that would substantially interfere with or preclude full participation in the program, (3) sign an agreement acknowledging the inmate's responsibilities while participating in the program, and (4) volunteer for the program. If an inmate is housed at an institution that does not operate an RDAP, the inmate can be transferred to an institution with an RDAP.

Inmates in residential drug abuse treatment programs are housed in separate units reserved for drug abuse treatment. Residential drug abuse treatment programs last 6 to 12 months (a minimum of 500 hours). Inmates in the program spend half of their day in treatment and the other half in education, work skills training, and/or other inmate programs. According to BOP, 10.9% of federal inmates participated in residential drug abuse treatment in FY2010.

Community Transition Drug Abuse Treatment

Inmates who participate in the residential drug abuse treatment program must also complete a course of follow-up treatment with a community-based treatment provider when transferred to RRC. This community transition program must be completed if an inmate, who is otherwise eligible, is to receive an early release. According to BOP, 9.8% of federal inmates participated in community transition drug abuse treatment in FY2010.

Research on the Effectiveness of Rehabilitative Programs

One researcher concluded that, in general, the body of literature on the effectiveness of rehabilitative programming suggests that inmates who participated in Adult Basic Education (ABE), GED, postsecondary education, vocational education, correctional industries, and incarceration-based drug abuse treatment programs were significantly less likely to recidivate than those who did not participate.26 However, in general, research on the effectiveness of rehabilitative programming has limitations, including the following:

One analysis suggests that some of the methodological shortcomings of research in the field might affect whether a treatment program is deemed to be effective. This analysis evaluated the relationship between a study's research methodology and the likelihood that the study found a treatment effect.27 The study found that less methodologically rigorous studies were more likely to find a treatment effect. Moreover, less-rigorous studies were more likely to report a larger effect size than more-rigorous studies. This trend held when the researchers only compared randomized studies (i.e., studies in which subjects were randomly assigned to either a treatment or control group) to the highest-quality non-randomized studies (i.e., quasi-experimental, but the researchers attempted to statistically control for difference between subjects in the treatment and control groups). The researchers also found that when they only considered the results of studies where the findings were statistically significant, less-rigorous studies were still more likely to find a positive effect. The researchers concluded that their findings suggested that a study's research design does have an impact on a finding of effectiveness and, while their results should be seen as a preliminary step in understanding how research design affects study outcomes in criminal justice, it suggests that reviews of what works in criminal justice might be biased when non-randomized studies are included.

Rehabilitative Needs of Federal Inmates

This section of the report provides an analysis of whether the need for rehabilitative programming changed between 1997 and 2004. The analysis uses data from the Bureau of Justice Statistics' Survey of Inmates in State and Federal Correctional Facilities (hereinafter, "survey") for 1997 and 2004 to determine whether federal inmates in 2004 were more or less likely than inmates in 1997 to have indicators of the need for rehabilitative programming.28 The selected indicators include the following:

These variables were chosen because they provide an indication of what type of rehabilitative programming an inmate might need.

Data from the 1997 and 2004 surveys show that it was likely that there was increased need for drug abuse treatment programs, but the need for literacy/GED and occupational education programs along with FPI work assignments remained flat. As shown in Table 1, in both 1997 and 2004 approximately one-quarter of federal inmates reported not having a GED or a high school diploma and approximately 30% of inmates in both years reported that they were unemployed before being arrested. While there was no significant increase in the percentage of inmates who reported not having a GED or a high school diploma or being unemployed before being arrested, inmates in 2004 were significantly more likely than inmates in 1997 to report that they had regularly used drugs and that they used drugs before being arrested.

Table 1. Indicators of Rehabilitative Need Among Federal Inmates, 1997 and 2004

 

1997

2004

Did not have a GED or a high school diploma

26.3%

26.8%

Unemployed before being arrested

28.2%

29.9%

Regularly used drugs

57.3%

64.6%

Used drugs before being arrested

44.8%

50.5%

Source: CRS analysis of data from the 1997 and 2004 Survey of Inmates in State and Federal Correctional Facilities.

Note: "Regularly used drugs" is defined as using any drug once a week or more for at least a month. "Used drugs before being arrested" is defined as using any drug in the month before being arrested for the crime(s) for which the inmate was incarcerated at the time of the interview.

Participation in Rehabilitative Programs

This section of the report provides an analysis of whether federal inmates with rehabilitative needs are participating in rehabilitative programming. Table 2 provides data on the overall percentage of inmates who reported currently having an FPI work assignment and the percentage of inmates who reported participating in an occupational education, literacy/GED, or some type of drug abuse treatment program during their current incarceration. Overall, between 1997 and 2004 there was a significant decrease in the percentage of inmates who reported currently having an FPI work assignment. On the other hand, there was a significant increase in the percentage of inmates who reported participating in a drug education program. There was no significant increase or decrease in the percentage of inmates who reported participating in any of the other rehabilitative programs.

Table 2. Participation in Rehabilitative Programs, 1997 and 2004

 

1997

2004

FPI work assignment

17.5%

12.0%

Occupational education

28.7%

31.7%

Literacy/GED program

24.0%

25.1%

Residential drug abuse treatment

7.2%

5.6%

Non-residential drug abuse treatment

3.8%

4.9%

Drug education program

16.7%

20.6%

Source: CRS analysis of data from the 1997 and 2004 Survey of Inmates in State and Federal Correctional Facilities.

The analysis used logistic regression to test whether inmates with rehabilitative needs were more likely than other inmates to participate in rehabilitative programming (see the Appendix). The results of the analysis suggest that in both 1997 and 2004, inmates who reported being unemployed before being arrested were not significantly more likely than other inmates to report that they had an FPI work assignment; however, in both years inmates who reported being unemployed before being arrested were significantly less likely than other inmates to have participated in an occupational education program. In addition, in both 1997 and 2004 inmates who reported not having a GED or a high school diploma were significantly more likely than other inmates to report that they participated in a literacy/GED program. Also, in both years inmates who reported either regular drug use or using drugs before being arrested were significantly more likely than other inmates to have participated in residential and non-residential drug abuse treatment and a drug abuse education course. Yet, the analysis also suggests that there was not a significant difference in the likelihood that inmates in 1997 and 2004 with rehabilitative needs participated in programming. This could suggest that BOP was not able to expand opportunities for rehabilitative programming between 1997 and 2004 or that inmates' desires to participate in rehabilitative programming did not increase between 1997 and 2004.

Even though inmates with rehabilitative needs were, in most cases, more likely than inmates without a need to participate in rehabilitative programs, the analysis suggests that it was less than certain that inmates with rehabilitative needs participated in programs to address their needs. The results of the analysis suggest that the probability that a typical inmate29 with a rehabilitative need participated in a program to address that need was, in most cases, less than 1 in 2. For example, the estimated probability that a typical inmate who was unemployed before being arrested had an FPI work assignment was about 1 in 5 for both 1997 and 2004, and the probability that a typical inmate who was unemployed before being arrested participated in an occupational education course was about 2 in 5 for both years. In addition, the estimated probability that a typical inmate who either reported regular drug use or used drugs before being arrested participated in residential or non-residential drug abuse treatment in 1997 or 2004 was about 1 in 10. The estimated probability that a typical inmate who either reported regular drug use or used drugs before being arrested participated in a drug abuse education course was approximately 1 in 5 in 1997 and 3 in 10 in 2004. The results of the analysis suggest that a typical inmate without either a GED or a high school diploma was more likely than not (an approximately 1 in 2 chance in 1997 and a 3 in 5 chance in 2004) to have participated in a literacy/GED course.

The results of the analysis also suggest that the probability that the highest-participating inmates participated in rehabilitative programs was 3 in 5 or lower. For example, the probability that the highest-participating inmates who reported being unemployed before being arrested had an FPI work assignment in 1997 was approximately 3 in 5, while in 2004 it was approximately 3 in 10. Moreover, the probability that the highest-participating inmates who reported being unemployed before being arrested participated in an occupational education course was approximately 1 in 2 in 1997 and approximately 3 in 5 in 2004. The estimated probability that the highest participating inmates who reported either regular drug use or using drugs before being arrested participated in a residential drug abuse program was approximately 3 in 10 in 1997 and 2 in 5 in 2004, while the estimated probability that these inmates participated in a non-residential drug abuse program was approximately 1 in 5 in 1997 and 3 in 10 in 2004. The analysis suggests that the probability that the highest-participating inmates who reported either regularly using drugs or using drugs before being arrested participated in a drug abuse education course was approximately 1 in 2 in both 1997 and 2004. Also, the results of the analysis suggest that the probability that the highest-participating inmates who reported not having a GED or a high school diploma participated in a literacy/GED program was approximately 3 in 5 in both 1997 and 2004.

One shortcoming of the analysis is the age of the data used to estimate whether inmates with rehabilitative needs are participating in programs to address their needs. As noted above, the most recent data are from 2004, which is more than eight years ago. Could the probability of participation in rehabilitative programming have increased since the last survey was conducted? Data from BOP on the percentage of inmates who participated in rehabilitative programming from FY2000 to FY2011 suggest that this might not be the case. As shown in Table 3, the total proportion of federal inmates who participated in rehabilitative programming each fiscal year since FY2000 has remained relatively steady, with the exception of the proportion of inmates who had an FPI work assignment, which decreased between FY2000 and FY2011. Also, the proportion of inmates in FY2010 and FY2011 who participated in the drug abuse education course was higher than it was in FY2000-FY2009. However, as noted above, this is the most basic form of drug abuse treatment offered by BOP. Hence, unless inmates with rehabilitative needs accounted for a greater share of the inmates who participated in rehabilitative programming after FY2004, the data from BOP suggest that the trends observed in the 1997 and 2004 survey data could be applicable to current federal inmates.

Table 3. Percentage of Inmates that Participated in BOP Rehabilitative Programs, FY2000-FY2011

 

 

Education Programs

Drug Abuse Treatment Programs

FY

FPI Work Assignment

Literacy/ GED

Occupational Education

Drug Abuse Education Course

Non-residential

Residential

Community Transition

2000

17.3%

N.A.

N.A.

12.5%

6.3%

10.0%

6.7%

2001

17.3%

N.A.

N.A.

13.2%

8.3%

11.8%

8.7%

2002

15.6%

N.A.

N.A.

13.0%

8.4%

11.8%

9.5%

2003

13.9%

14.2%

6.8%

14.3%

8.2%

12.0%

10.3%

2004

12.7%

14.5%

6.9%

14.5%

8.5%

12.0%

10.8%

2005

12.4%

14.2%

6.2%

14.3%

8.9%

11.3%

10.3%

2006

13.0%

13.8%

6.1%

14.2%

8.4%

10.7%

9.5%

2007

13.8%

12.7%

6.4%

14.1%

8.6%

10.5%

9.2%

2008

13.2%

12.5%

6.7%

14.0%

8.6%

10.6%

9.3%

2009

11.0%

12.5%

6.5%

17.8%

8.5%

10.9%

9.4%

2010

9.0%

12.3%

6.4%

27.6%

8.4%

10.9%

9.8%

2011

8.0%

11.8%

6.4%

23.2%

8.5%

10.4%

9.5%

Source: CRS presentation of data provided by U.S. Department of Justice, Bureau of Prisons.

Note: BOP was not able to provide data on the number of inmates that participated in the GED/Literacy and Occupational Training programs prior to FY2003. Percentages were calculated as the number of inmates that participated in the program divided by the institutional population for that fiscal year.

Select Issues for Congress

The analysis above suggests either that BOP is not offering rehabilitative programming to all inmates with a need for it, or that inmates are choosing not to participate in rehabilitative programming even though they indicate a need for it. There may be multiple explanations for both findings. Moreover, as discussed earlier, there is not necessarily conclusive evidence that rehabilitative programs for inmates are effective. This section of the report discusses two potential issues Congress might choose to consider.

Providing Access to Rehabilitative Programming

A potential issue for Congress is whether BOP has the resources it needs to carry out its mission to provide rehabilitative programming to federal inmates. According to BOP, its biggest challenge is "managing the ever increasing federal inmate population, and providing for their care and safety, as well as the safety of BOP staff and surrounding communities, within budgeted levels."30 In its FY2011 congressional budget submission, BOP notes that it has "stretched resources, streamlined operations, improved program efficiencies, and reduced costs to operate as efficiently and effectively as possible."31 Even though total appropriations for BOP increased every fiscal year since FY2000, BOP has dedicated a growing percentage of its base budget to cover mandatory requirements, including inmate medical care, food, utilities, fuel, and correctional staffing. BOP also reports that the inmate-to-staff ratio, which includes staff to provide rehabilitative programming, has increased over the past several fiscal years.32 Data provided by BOP indicate that the total number of BOP staff per 1,000 inmates decreased from 242.0 in FY2000 to 202.2 in FY2011. As Congress continues its oversight of BOP, it might consider whether staffing levels at BOP are adequate to allow BOP to carry out its mission to rehabilitate inmates.

A related issue is whether BOP has enough facilities to manage the federal prison population. BOP opened 50 new prisons between FY1991 and FY2012 and additional prisons mean more classrooms and factories in which BOP can offer rehabilitative programs. In addition to providing more classrooms and factories, additional prisons decrease crowding. Less crowding could increase inmates' access to rehabilitative programming by decreasing the amount of time inmates have to wait for admission into programs. However, crowding in BOP institutions averaged 37% over rated capacity each fiscal year between FY2005 and FY2012. During this time period, BOP had a net increase of two new prisons.33 BOP plans to add nearly 12,000 new bedspaces to its current capacity between FY2012 and FY2018, but even with the additional capacity BOP still projects prison crowding to increase.34 Congress might consider whether to increase funding to BOP so it might expand its capacity and reduce crowding. Congress might also consider whether BOP should place more low, and possibly medium, security offenders in privately operated prisons.

Incentives for Participating in Rehabilitative Programs

Another potential issue before Congress is the structure of incentives for inmates to participate in rehabilitative programs. The Comprehensive Crime Control Act of 1984 (P.L. 98-473) abolished parole for federal inmates and modified how much good time credit an inmate could earn.35 There was not a single event that led Congress to abolish parole for federal offenders; rather, it was the culmination of several critiques and concerns about sentencing policy that began in the 1970s.36 Some viewed the discretion exercised by the judiciary and parole boards over the length of an offender's sentence as arbitrary and unfair.37 Others thought that allowing parole boards to determine when an inmate could be released based on the inmate's behavior while incarcerated along with the goal of rehabilitation was tantamount to coddling criminals.38 In the late 1970s, states, and eventually the federal government, started to limit judicial discretion by replacing indeterminate sentences with determinate sentences (e.g., sentencing guidelines).39 There was also a movement to limit judicial and parole board discretion by enacting "truth-in-sentencing" laws, which required offenders sentenced for certain crimes (usually violent crimes) to serve no less than 85% of their sentence before being eligible for release.40

By eliminating parole and modifying the amount of good time credit an inmate could earn, Congress decreased some of the incentives inmates had to participate in rehabilitative programs. Prior to the act, when considering whether to grant an inmate parole the U.S. Parole Board considered whether an inmate had followed institutional rules while incarcerated and whether, in light of the nature and circumstances of the offense and the history and characteristics of the inmate, granting parole would depreciate the seriousness of the offense and promote disrespect of the law and jeopardize public welfare.41 As such, inmates had an incentive to participate in rehabilitative programs in order to prove to the parole board that they would not be a threat to the community if paroled and that they had attempted to atone for their crimes. In addition, prior to the act inmates could earn up to three days of industrial good time credit per month for the first year of participation in an FPI program and up to five days of industrial good time credit per month for any succeeding year.42 The law also permitted BOP to award the same amount of industrial good time credit to "prisoner[s] performing exceptionally meritorious service or performing duties of outstanding importance in connection with institutional operations."43 Industrial good time credit was in addition to good time credit inmates could earn for following the institution's rules.44 The previous good time credit system provided an incentive for inmates to participate in work programs by allowing them to earn up to an additional 36 days of industrial good time credit in the first year and up to 60 days of industrial good time credit for each additional year.

The current good time credit structure only provides inmates with an incentive to participate in basic education programs and drug abuse treatment. Under current law, inmates are not eligible for their full allotment of good time credit if it is determined that they are not making satisfactory progress on earning a high school diploma or a GED.45 However, even if an inmate earns the full allotment of good time credit, the inmate will serve about 85% of the imposed sentence, which in most cases would be longer than the amount of time an inmate would serve under the previous system. In addition to the amount of good time credit an inmate can earn, BOP is allowed to reduce a nonviolent inmate's sentence by up to one year if the inmate participates in residential substance abuse treatment.46 Policymakers could consider whether providing additional good time credit or longer placement in an RRC might encourage more inmates to participate in rehabilitative programs.47 For example, current law could be amended to allow inmates who earn a vocational degree to receive additional good time credit. On the other hand, providing additional good time credit for inmates to participate in rehabilitative programming might also provide an incentive for inmates to participate for the sole purpose of reducing the time they have to serve in prison rather than participating because they wish to prepare themselves for reentry into the community. This might mean that inmates who would benefit from participating in rehabilitative programming would not be able to due to resource restrictions. In addition, it might also mean that inmates who are not fully rehabilitated would be released from prison earlier than they otherwise would have been.

Appendix. Methodology

This appendix provides an overview of the data used, methods used, and results of the analysis CRS conducted for this report.

Dataset

The analysis was conducted using data from the 1997 and 2004 editions of the Bureau of Justice Statistics' (BJS's) Survey of Inmates in State and Federal Correctional Facilities (hereinafter referred to as "the survey"). The survey provides nationally representative data on inmates held in state and federal prisons. The survey collects data on inmates' current offense and sentence; criminal history; family background and personal characteristics; prior drug and alcohol use and treatment programs; gun possession and use; and prison activities, programs, and services. Interviews for the 1997 survey were conducted between June and October of 1997 while interviews for the 2004 survey were conducted between October 2003 and May 2004.

The survey sample was selected using a two-stage stratified sampling design. A sample of federal prisons was chosen in the first stage of sampling. In 1997, one male and two female prisons were selected with certainty, while in 2004 one female and two male prisons were selected with certainty. The remaining male prisons were grouped into five strata based on the prison's security level (administrative, high, medium, low, and minimum), while the remaining female prisons were grouped into two strata (minimum and all other security levels). Within each stratum, facilities were ordered by the size of the prison's population, and a sample was selected with probability proportional to size. For the 1997 survey, 32 out of 113 male prisons and 8 out of 22 female prisons were selected. For the 2004 survey, 32 out of 131 male prisons and 8 out of 17 female prisons were selected.

In the second stage of sampling, a sample of inmates was selected from the prisons selected in the first stage of sampling. BOP generated a list of inmates held in the sampled prisons, and inmates were selected from the list by using a randomly selected starting point and a predetermined skip interval. However, because drug offenders are overrepresented in the federal prison population, a systematic sample would have resulted in too few non-drug offenders; hence, the federal prison population was initially oversampled so that enough non-drug offenders in the sample could be analyzed. From the initial list of inmates, one out of every three drug offenders, along with all selected non-drug offenders, were retained for the final sample. For the 1997 survey, a total of 4,479 inmates were selected for inclusion in the sample, of which 4,041 were interviewed (a 90.2% response rate). For the 2004 survey, a total of 4,253 inmates were selected for inclusion in the sample, of which 3,686 were interviewed (a 86.7% response rate). The 4,041 inmates interviewed in 1997 represented 89,072 federal inmates, while the 3,686 inmates interviewed in 2004 represented 129,299 federal inmates.

The survey data are available for download from the Inter-university Consortium for Political and Social Research's website (http://www.icpsr.umich.edu/icpsrweb/ICPSR/). However, certain variables are redacted from the publicly available dataset to protect the privacy of inmates who responded to the survey. Two of the redacted variables—the variables that identified which strata and prison each inmate in the sample was in—were needed to adjust estimated standard errors to compensate for the complex sample design. As such, CRS had to enter into a restricted data use agreement with the National Archive of Criminal Justice Data (NACJD) in order to gain access to the redacted variables. Per the agreement with the NACJD, the restricted dataset was destroyed after this report was completed. A copy of the restricted data use agreement is on file with the author.

Methods

Logistic regression models were used to test the relationship between six dependent variables—whether the inmate had (1) an FPI work assignment at the time of the interview, (2) participated in a VoTech program, (3) participated in a literacy or GED course, (4) participated in the residential drug abuse treatment program, (5) participated in the non-residential drug abuse treatment program, and (6) participated in the drug abuse education course—and the four indicators of rehabilitative need: (1) not possessing a GED or a high school diploma, (2) being unemployed before being arrested, (3) using drugs regularly, and (4) using drugs before being arrested. Each model includes a series of demographic variables to control for any differences in program participation rates that may exist between different groups of inmates. All variables included in the analysis are dummy coded. All models were estimated using the SURVEYLOGISTIC procedure in SAS 9.2. The SUVEYLOGISTIC procedure includes variables that SAS can use to adjust standard errors to account for the complex sampling design (see description above).

The analysis excludes inmates who were held in prison but not incarcerated (i.e., pre-trial inmates) because these inmates are not cleared by BOP to participate in rehabilitative programs.

Missing data were first "logically" imputed by reviewing the survey instrument to see if data were missing because the inmate was not asked the question due to the skip pattern in the instrument. For example, data might be missing from the questions asking about whether the inmate used drugs before being arrested because in a previous series of questions the inmate responded that he or she has not used drugs. In these instances, the missing values could be replaced with "no." Data that could not be logically imputed were imputed using the MI procedure in SAS 9.2.

Each model was estimated using the 1997 data and the 2004 data. The models estimated using the 1997 data are presented in Table A-1 and the models estimated using the 2004 data are presented in Table A-2. Significance tests of the difference between coefficients estimated using each set of data were conducted using the methods specified in Raymond Paternoster et al.'s 1998 article.48 The estimated effect that a variable had on the probability that an inmate reported participating in a rehabilitative program was calculated using the formula specified by Paul Allison.49

The following formula, along with the estimated coefficients presented in Table A-1 and Table A-2, can be used to calculate the probability that inmates in a particular group participated in rehabilitative programming. The formula allows the reader to compare the probability that inmates in different groups within the same year or inmates in the same group between the years participated in rehabilitative programming.

Where "α" is the intercept of the estimated logistic regression model, "β" is the estimated coefficient, "x" is the value for the independent variable, and "e" is base e (roughly equal to 2.718).

Limitations of the Data

The data used in this analysis represent the two most recent surveys conducted by BJS. However, even the most recent data are more than six years old. Hence, even though the data from 1997 and 2004 suggest some trends in terms of the rehabilitative needs of federal inmates and their participation in rehabilitative programming, it is possible that those trends have not continued. Because data in the survey are self-reported, there is the possibility of measurement error because interviewed inmates could provide inaccurate answers due to poor recollection or because they could choose to exaggerate or minimize their responses. Also, even though the survey provides a rich source of data on federal inmates, it did not collect data on every variable that might have been useful for this analysis. For example, the survey did not ask inmates if they were eligible for parole and the other data collected did not allow CRS to determine if a given inmate was eligible for parole. As such, the coefficients for the variables included in the model might be biased to the extent that omitted variables are correlated with the variables included in the models. Therefore, there must be some degree of caution when considering how much more likely inmates with rehabilitative needs were than other inmates to participate in rehabilitative programs. Finally, the variables "regularly used drugs" and "used drugs before arrest" are admittedly rough proxies for an actual diagnosis of drug abuse. It might well be that inmates who responded that they either regularly used drugs or that they used drugs before being arrested would not be diagnosed as a drug abuser.

Table A-1. Estimated Logistic Regression Models Using Data from the 1997 Survey

 

 

Dependent Variables

 

 

FPI Work Assignment

Occupational Education

Literacy/ GED Program

Residential Drug Abuse Treatment

Non-residential Drug Abuse Treatment

Drug Abuse Education Course

Residential Drug Abuse Treatment

Non-residential Drug Abuse Treatment

Drug Abuse Education Course

Ages 31-40a

β

0.1602

0.0721

-0.0858

0.2773n

0.4194

0.1448

0.3452l

0.4927n

0.2007n

 

SE

(0.1213)

(0.1099)

(0.0984)

(0.1297)

(0.2234)

(0.0977)

(0.1278)

(0.2231)

(0.0920)

Ages 41-50a

β

0.3419m

-0.1670

-0.1296

0.2160

0.1446

0.0118

0.3023

0.2104

0.0744

 

SE

(0.1226)

(0.1171)

(0.0967)

(0.1580)

(0.2769)

(0.1464)

(0.1572)

(0.2763)

(0.1409)

Ages 51 or oldera

β

0.5014l

-0.6092l

-0.1794

-0.0362

0.3273

-0.5610m

-0.0393

0.3288

-0.5532m

 

SE

(0.1517)

(0.1487)

(0.1084)

(0.2968)

(0.3334)

(0.1689)

(0.2711)

(0.3261)

(0.1725)

Blackb

β

0.3093m

0.3763l

0.2624n

-0.3247n

0.0092

-0.4045l

-0.3029n

0.0275

-0.3889l

 

SE

(0.0936)

(0.0810)

(0.1162)

(0.1332)

(0.1633)

(0.0982)

(0.1324)

(0.1631)

(0.0961)

Hispanicb

β

0.2159

0.3803l

0.3698m

-0.1893

-0.0853

-0.7254l

-0.1993

-0.1129

-0.7370l

 

SE

(0.1215)

(0.0882)

(0.1328)

(0.1818)

(0.2530)

(0.1781)

(0.1837)

(0.2492)

(0.1723)

Other raceb

β

0.0787

0.2312

0.1411

-0.1807

-0.5743

-0.5490l

-0.1849

-0.5870

-0.5522m

 

SE

(0.1337)

(0.1516)

(0.2259)

(0.2053)

(0.3941)

(0.1614)

(0.2161)

(0.3941)

(0.1716)

Violent offensec

β

1.1544l

0.7700l

0.6233l

0.4239n

0.4415n

0.8580l

0.5049m

0.5064m

0.9062l

 

SE

(0.1389)

(0.1077)

(0.1577)

(0.1808)

(0.1998)

(0.1428)

(0.1843)

(0.1930)

(0.1427)

Drug offensec

β

0.5801l

0.6148l

0.7141l

0.5070n

-0.2079

0.8610l

0.5551n

-0.1633

0.8835l

 

SE

(0.1514)

(0.0904)

(0.1542)

(0.2298)

(0.2724)

(0.1192)

(0.2366)

(0.2705)

(0.1210)

Public order offensec

β

0.5469l

0.4255l

0.3661n

0.2095

-0.2881

0.5917l

0.2886

-0.2292

0.6427l

 

SE

(0.1406)

(0.0868)

(0.1647)

(0.1715)

(0.2278)

(0.1390)

(0.1743)

(0.2277)

(0.1403)

Non-U.S. citizend

β

0.7697l

-0.1085

0.0658

-0.2356

-1.9892m

0.1687

-0.3312

-2.0569m

0.0950

 

SE

(0.0899)

(0.1135)

(0.1480)

(0.3766)

(0.6239)

(0.2025)

(0.3648)

(0.6392)

(0.1892)

Femalee

β

-0.1803

0.2815m

0.2274

0.2035

0.5593n

0.0249

0.1926

0.5598n

0.0216

 

SE

(0.2017)

(0.1025)

(0.1567)

(0.2174)

(0.2747)

(0.1520)

(0.2115)

(0.2757)

(0.1476)

Recidivistf

β

0.1080

0.1440

0.1416

0.2667

0.4162n

0.1676

0.3465

0.4933n

0.2324n

 

SE

(0.0926)

(0.0964)

(0.1109)

(0.1811)

(0.1957)

(0.1136)

(0.1898)

(0.2051)

(0.1102)

One year or less to serveg

β

-0.6422l

-0.1270

-0.0989

0.8002l

-0.1189

0.3823m

0.8074l

-0.1234

0.3849l

 

SE

(0.1498)

(0.1131)

(0.1078)

(0.1927)

(0.1369)

(0.1158)

(0.1888)

(0.1423)

(0.1151)

Unemployed before arresth

β

0.0561

-0.2770m

 

SE

(0.0843)

(0.0905)

 

 

 

 

 

 

 

No GED or high school diplomai

β

-0.3996m

-0.7761l

1.6651l

 

SE

(0.1207)

(0.0986)

(0.0910)

 

 

 

 

 

 

Regularly used drugsj

β

1.4804l

1.1067l

1.0454l

 

SE

 

 

 

(0.2038)

(0.2024)

(0.0979)

 

 

 

Used drugs before arrestk

β

1.1348l

0.8546l

0.8596l

 

SE

 

 

 

 

 

 

(0.1591)

(0.1717)

(0.0917)

Intercept

β

-2.5573l

-1.4649l

-2.5739l

-4.4760l

-4.3257l

-2.9922l

-4.2090l

-4.1595l

-2.8477l

 

SE

(0.2136)

(0.1870)

(0.1930)

(0.3108)

(0.3223)

(0.1818)

(0.2366)

(0.2841)

(0.1776)

Source: CRS presentation of results of analysis conducted using data from the 1997 Survey of Inmates in State and Federal Correctional Facilities.

a. Reference group is inmates between the ages of 18 and 30.

b. Reference group is white inmates.

c. Reference group is inmates convicted of property offenses.

d. Reference group is inmates who are U.S. citizens.

e. Reference group is male inmates.

f. Reference group is inmates who do not have a prior conviction.

g. Reference group is inmates who have more than a year left on their sentences.

h. Reference group is inmates who were employed before being arrested.

i. Reference group is inmates who have a GED or a high school diploma.

j. Reference group is inmates who reported that they did not regularly use drugs.

k. Reference group is inmates who reported that they were not using drugs before being arrested.

l. Significant at the p < 0.001 level.

m. Significant at the p < 0.01 level.

n. Significant at the p < 0.05 level.

Table A-2. Estimated Logistic Regression Models Using Data from the 2004 Survey

 

 

Dependent Variables

 

 

FPI Work Assignment

Occupational Education

Literacy/ GED Program

Residential Drug Abuse Treatment

Non-residential Drug Abuse Treatment

Drug Abuse Education Course

Residential Drug Abuse Treatment

Non-residential Drug Abuse Treatment

Drug Abuse Education Course

Ages 31-40a

β

0.3493n

0.1182

-0.0235

0.3601n

0.0728

-0.0828

0.3667n

0.0670

-0.0759

 

SE

(0.1650)

(0.1027)

(0.0960)

(0.1764)

(0.1964)

(0.1140)

(0.1794)

(0.1966)

(0.1144)

Ages 41-50a

β

0.3454m

-0.1305

-0.0876

0.2225

0.5678m

-0.0007

0.2465

0.5738m

0.0161

 

SE

(0.1474)

(0.1166)

(0.1295)

(0.1778)

(0.2043)

(0.1179)

(0.1874)

(0.2044)

(0.1150)

Ages 51 or oldera

β

0.0053

-0.4853l

-0.7122l

-0.0433

0.2227

-0.2835

-0.0562

0.1095

-0.3212n

 

SE

(0.1808)

(0.1283)

(0.1159)

(0.2847)

(0.3622)

(0.1558)

(0.3002)

(0.3786)

(0.1594)

Blackb

β

0.3247n

0.5541l

0.4393m

-0.3677

-0.3934

-0.0884

-0.3703

-0.4389

-0.1134

 

SE

(0.1532)

(0.0848)

(0.1381)

(0.1877)

(0.2219)

(0.1118)

(0.1988)

(0.2324)

(0.1147)

Hispanicb

β

0.0559

0.3260m

0.4773m

-0.1406

-0.9417m

-0.2183

-0.1757

-1.0189m

-0.2624

 

SE

(0.1821)

(0.1138)

(0.1785)

(0.2481)

(0.3447)

(0.1439)

(0.2553)

(0.3537)

(0.1417)

Other raceb

β

0.3039

0.0767

0.2116

0.4252

-0.2294

0.1236

0.4025

-0.2790

0.1089

 

SE

(0.2324)

(0.2124)

(0.2221)

(0.2265)

(0.2954)

(0.2014)

(0.2092)

(0.2962)

(0.1919)

Violent offensec

β

1.0947l

0.4743m

0.6028l

-0.3978

0.6135n

0.3421

-0.3587

0.7257m

0.3715

 

SE

(0.2494)

(0.1612)

(0.1298)

(0.2875)

(0.2798)

(0.2024)

(0.2748)

(0.2615)

(0.1952)

Drug offensec

β

0.5927m

0.2611

0.6221l

0.3585

0.1338

0.4573n

0.3543

0.2545

0.4824m

 

SE

(0.2470)

(0.1503)

(0.1352)

(0.2707)

(0.2338)

(0.1903)

(0.2708)

(0.2152)

(0.1775)

Public order offensec

β

0.5955m

0.2060

0.3569n

0.3336

0.2136

0.0323

0.3705

0.3174

0.0717

 

SE

(0.2407)

(0.1597)

(0.1612)

(0.2348)

(0.3439)

(0.1815)

(0.2381)

(0.3250)

(0.1785)

Non-U.S. citizend

β

-0.1550

-0.5500l

-0.3743n

-1.3239l

-0.0458

-0.6804m

-1.3031l

-0.1577

-0.7146m

 

SE

(0.2004)

(0.1418)

(0.1456)

(0.2735)

(0.3874)

(0.2354)

(0.2770)

(0.3656)

(0.2344)

Femalee

β

-0.1227

0.2946n

-0.1507

0.3094

0.2787

-0.1409

0.2686

0.2353

-0.1739

 

SE

(0.2245)

(0.1366)

(0.1169)

(0.2176)

(0.1569)

(0.1189)

(0.2129)

(0.1545)

(0.1214)

Recidivistf

β

-0.0883

0.0837

0.2120n

0.1146

0.5400m

0.4432l

0.1657

0.6629l

0.5044l

 

SE

(0.1102)

(0.0613)

(0.0958)

(0.1745)

(0.1829)

(0.1084)

(0.1847)

(0.1868)

(0.1072)

One year or less to serveg

β

-0.5526l

-0.1287

0.0169

1.3365l

0.3612

0.5239l

1.3458l

0.3686n

0.5300l

 

SE

(0.1313)

(0.0902)

(0.1248)

(0.1658)

(0.1894)

(0.1229)

(0.1610)

(0.1850)

(0.1213)

Unemployed before arresth

β

0.0970

-0.2617m

 

SE

(0.1433)

(0.0884)

 

 

 

 

 

 

 

No GED or high school diplomai

β

-0.7189l

-0.7027l

1.6926l

 

SE

(0.1509)

(0.1078)

(0.0990)

 

 

 

 

 

 

Regularly used drugsj

β

1.7581l

1.6516l

1.0481l

 

SE

 

 

 

(0.3266)

(0.2819)

(0.1236)

 

 

 

Used drugs before arrestk

β

1.3403l

0.8138l

0.7839l

 

SE

 

 

 

 

 

 

(0.1967)

(0.1786)

(0.1169)

Intercept

β

-2.6937l

-1.0386l

-2.5174l

-5.0617l

-4.9097l

-2.6782l

-4.5737l

-4.2230l

-2.4009l

 

SE

(0.2802)

(0.2017)

(0.1857)

(0.3766)

(0.4125)

(0.2474)

(0.2640)

(0.4078)

(0.2431)

Source: CRS presentation of results of analysis conducted using data from the 2004 Survey of Inmates in State and Federal Correctional Facilities.

a. Reference group is inmates between the ages of 18 and 30.

b. Reference group is white inmates.

c. Reference group is inmates convicted of property offenses.

d. Reference group is inmates who are U.S. citizens.

e. Reference group is male inmates.

f. Reference group is inmates who do not have a prior conviction.

g. Reference group is inmates who have more than a year left on their sentences.

h. Reference group is inmates who were employed before being arrested.

i. Reference group is inmates who have a GED or a high school diploma.

j. Reference group is inmates who reported that they did not regularly use drugs.

k. Reference group is inmates who reported that they were not using drugs before being arrested.

l. Significant at the p < 0.001 level.

m. Significant at the p < 0.01 level.

n. Significant at the p < 0.05 level.

Acknowledgments

[author name scrubbed], formerly with CRS; [author name scrubbed], Section Research Manager; and Kristin M. Finklea, Specialist in Domestic Security, peer reviewed this report. [author name scrubbed], formerly of CRS, and [author name scrubbed], Analyst in Labor Policy, provided insights and comments on the regression models.

Footnotes

1.

U.S. Department of Justice, Bureau of Prisons, Mission and Vision of the Bureau of Prisons, http://www.bop.gov/about/mission.jsp.

2.

Jeremy Travis, But They All Come Back: Facing the Challenges of Prisoner Reentry (Washington, DC: Urban Institute Press, 2005), pp. 323-352.

3.

BOP is authorized in law at Chapter 303 of Title 18 of the United States Code (18 U.S.C. §4041 et seq.).

4.

U.S. Department of Justice, Bureau of Prisons, About the Bureau of Prisons, http://www.bop.gov/about/index.jsp.

5.

U.S. Department of Justice, Bureau of Prisons, A Brief History of the Bureau of Prisons, http://www.bop.gov/about/%20history.jsp.

6.

Ibid.

7.

Data provided by the U.S. Department of Justice, Bureau of Prisons. Data are on file with the author.

8.

U.S. Department of Justice, Bureau of Prisons, State of the Bureau 2008, Washington, DC, p. 2, http://www.bop.gov/news/PDFs/sob08.pdf.

9.

U.S. Department of Justice, Bureau of Prisons, Prison Types & General Information, http://www.bop.gov/locations/institutions/index.jsp.

10.

U.S. Department of Justice, Bureau of Prisons, Community Corrections, http://www.bop.gov/locations/cc/index.jsp.

11.

Ibid.

12.

U.S. Department of Justice, Bureau of Prisons, About the Federal Bureau of Prisons, July 2007, http://www.bop.gov/news/PDFs/ipaabout.pdf.

13.

For more information on FPI, see CRS Report RL32380, Federal Prison Industries, by [author name scrubbed].

14.

Information in this section was obtained from U.S. Department of Justice, Bureau of Prisons, Program Statement 5251.06, Work and Performance Pay, Inmate, http://www.bop.gov/policy/progstat/5251_006.pdf; U.S. Department of Justice, Bureau of Prisons, Program Statement 8120.02, Work Programs for Inmates—FPI, http://www.bop.gov/policy/progstat/8120_002.pdf; U.S. Department of Justice, Bureau of Prisons, Work Programs, http://www.bop.gov/inmate_programs/work_prgms.jsp; U.S. Department of Justice, Bureau of Prisons, State of the Bureau 2008, http://www.bop.gov/news/PDFs/sob08.pdf; and U.S. Department of Justice, Federal Prison Industries, Inc., UNICOR Annual Report 2009, http://www.unicor.gov/information/publications/pdfs/corporate/catar2009.pdf.

15.

This report focuses on FPI work assignments because (1) they are usually considered more desirable because of the higher per-hour pay; (2) they allow inmates to learn a trade; and (3) an analysis of participation in all work programs would be invalid because all medically able inmates are required to have a prison work assignment, meaning that only a small number of inmates interviewed for the surveys did not work and the estimated coefficients in the model would be overly reliant on the characteristics of this small number of inmates.

16.

Information in this section was obtained from BOP Program Statement 5350.28, Literacy Program (GED Standard), http://www.bop.gov/policy/progstat/5350_028.pdf, and an August 29, 2007, BOP briefing on BOP's education programs.

17.

See 18 U.S.C. §3624(f)(1) and 18 U.S.C. §3624(b)(3).

18.

Each prisoner serving a term of imprisonment of more than one year, but not prisoners serving a life sentence, can receive a good time credit of up to 54 days per year to count toward serving the sentence. The amount of the credit is subject to the determination of BOP. 18 U.S.C. §3624(b).

19.

In some instances, inmates cannot opt out of the literacy program because they are required by statute to participate in it. According to BOP, inmates sentenced pursuant to the Youth Corrections Act and the Narcotics Addict Rehabilitation Act cannot opt out of the literacy program.

20.

Information in this section was obtained from U.S. Department of Justice, Bureau of Prisons, Program Statement 5353.01, Occupational Education Programs, http://www.bop.gov/policy/progstat/5353_001.pdf; U.S. Department of Justice, Bureau of Prisons, Program Statement 5300.21, Education, Training and Leisure Time Program Standards, http://www.bop.gov/policy/progstat/5300_021.pdf; and an August 29, 2007, BOP briefing on BOP's education programs.

21.

Information in this section was obtained from U.S. Department of Justice, Bureau of Prisons, Program Statement 5330.11, Psychology Treatment Program, http://www.bop.gov/policy/progstat/5330_011.pdf; U.S. Department of Justice, Bureau of Prisons, FY2011 Congressional Budget Submission; U.S. Department of Justice, Bureau of Prisons, Substance Abuse Treatment, http://www.bop.gov/inmate_programs/substance.jsp; and a conversation on August 18, 2008, with U.S. Department of Justice, Bureau of Prisons, Office of Congressional Affairs.

22.

"Residential substance abuse treatment" is defined as a course of individual and group activities and treatment, lasting at least six months, in residential treatment facilities set apart from the general prison population (which may include pharmocotherapies, where appropriate) that may extend beyond the six-month period. 18 U.S.C. §3621(e)(5)(A).

23.

"Aftercare" is defined as placement, case management, and monitoring in a community-based substance abuse treatment program when the prisoner leaves the custody of BOP. 18 U.S.C. §3621(e)(5)(C).

24.

An "eligible prisoner" is defined as a prisoner who is determined by BOP to have a substance abuse problem and to be willing to participate in a residential substance abuse treatment program. 18 U.S.C. §3621(e)(5)(B).

25.

The following categories of inmates are not eligible for early release: (1) Immigration and Customs Enforcement detainees; (2) pretrial inmates; (3) contractual boarders (for example, District of Columbia, state, or military inmates); (4) inmates who have a prior felony or misdemeanor conviction for homicide, forcible rape, robbery, or aggravated assault, or child sexual abuse offenses; (5) inmates who are not eligible for participation in a community-based program as determined by the institution's warden on the basis of his or her professional discretion; or (6) inmates whose current offense is a felony. 28 C.F.R. §550.58.

26.

Doris Layton MacKenzie, What Works in Corrections: Reducing the Criminal Activities of Offenders and Delinquents (New York: Cambridge University Press, 2006).

27.

David Weisburd, Cynthia M. Lum, and Anthony Petrosino, "Does Research Design Affect Study Outcomes in Criminal Justice," The Annals of the American Academy of Political and Social Scientist, vol. 578 (2001), p. 50.

28.

Even though the survey interviewed inmates held in both state and federal correctional facilities, the analysis only includes federal inmates. For more information, see the Appendix.

29.

A typical inmate is defined as an inmate who is black, is between the ages of 31 and 40, is convicted of a drug offense, is male, is a U.S. citizen, has a prior conviction, and has more than a year to serve on his sentence.

30.

U.S. Department of Justice, Bureau of Prisons, FY2011 Performance Budget, Congressional Submission, Salaries and Expenses, p. 5, http://www.justice.gov/jmd/2011justification/pdf/fy11-bop-se-justification.pdf.

31.

Ibid., p. 1.

32.

Ibid., p. 5.

33.

BOP reported that it operated 116 prisons in FY2005. In FY2006, BOP closed four prison camps and opened two new prisons, leaving it with a total of 114 total prisons. BOP opened a new prison in each of FY2009, FY2010, FY2011, and FY2012 leaving it with a total of 118 prisons at the end of FY2012.

34.

U.S. Department of Justice, Bureau of Prisons, FY2012 Performance Budget, Congressional Submission, Buildings and Facilities, p. 2, http://www.justice.gov/jmd/2012justification/pdf/fy12-bop-bf-justification.pdf.

35.

Anyone sentenced to incarceration for a federal crime committed after November 1, 1987, is not eligible for parole.

36.

Jeremy Travis, But They All Come Back: Facing the Challenges of Prisoner Reentry (Washington, DC: Urban Institute Press, 2005), pp. 17-20. For more information on the history of determinate and indeterminate sentencing structures, see CRS Report RL32766, Federal Sentencing Guidelines: Background, Legal Analysis, and Policy Options, by Lisa M. Seghetti and [author name scrubbed].

37.

Ibid., p. 17.

38.

Ibid.

39.

Ibid., p. 19.

40.

Ibid.

41.

18 U.S.C. §4206(a), as it was in effect before being repealed by Section 218(a) of P.L. 98-473.

42.

18 U.S.C. §4162, as it was in effect before being repealed by Section 218(a) of P.L. 98-473.

43.

Ibid.

44.

Prior to the act, inmates who were sentenced to less than a life sentence and who observed institutional rules and were not subjected to punishment could earn the following amounts of good time credit: five days for each month if the inmate was sentenced to at least six months but less than a year; six days for each month if the inmate was sentenced to at least one year but less than three years; seven days for each month if the inmate was sentenced at least three years but less than five years; eight days for each month if the inmate was sentenced to at least five years but less than 10 years; and 10 days for each month if the sentence is 10 years or more. 18 U.S.C. §4161, as it was in effect before being repealed by Section 218(a) of P.L. 98-473.

45.

Currently, inmates can earn up to 54 days of good time credit for every year served. 18 U.S.C. §3624(b)(1).

46.

18 U.S.C. §3621(e)(2)(B).

47.

Under 18 U.S.C. §3624(c), inmates can, to the extent practicable, serve up to the last 12 months of their sentences in a residential reentry center (RRC).

48.

Raymond Paternoster, Robert Brame, and Paul Mazerolle et al., "Using the Correct Statistical Test for the Equality of Regression Coefficients," Criminology, vol. 36, no. 4 (1998), pp. 859-866.

49.

Paul D. Allison, Logistic Regression Using the SAS System (Cary, NC: SAS Institute Inc., 1999), p. 30.