’4‘

IIIIIIIIIIIIII'IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

3 1293 01812 9381

LIBRARY
Michigan State
University

 

 

 

 

This is to certify that the

thesis entitled

BOOT CAMP: EFFECTIVE SOLUTION FOR THE CRIMINAL
JUSTICE SYSTEM?

presented by

RacheI A. PeTta

has been accepted towards fulfillment
of the requirements for

M.S. _ Criminal Justice
degree in

MM

Major professcfl y

 

 

Date ¥A7Ié9

0-7 639 MS U is an Affirmative Action/Equal Opportunity Institution

w uh—F,

 

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINE return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

MI?) I a 2.002

 

 

0 ‘W"? ’I €005.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ma animus-p.14

BOOT CAMP: EFFECTIVE SOLUTION FOR THE CRIMINAL JUSTICE SYSTEM?
By

Rachel A. Pelta

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE
School of Criminal Justice

1999

ABSTRACT
BOOT CAMP: EFFECTIVE SOLUTION FOR THE CRIMINAL JUSTICE SYSTEM?
By

Rachel A. Pelta

In recent years, the criminal justice system has had to look for solutions to case its
overburdened prisons. One popular solution has been the use of boot camps. They
include military and rehabilitative aspects that are designed to reduce costs and
overcrowding of the traditional prisons and be harsh enough to deter young offenders
from recidivating. This thesis looks at the Louisiana Department of Public Safety and
Corrections boot camp program IMPACT (Intensive Motivational Program of Alternative
Correctional Treatment). Three groups are examined over a twelve-month period
following release to compare recidivism rates. In addition to recidivism, informal social
bonds of individual subjects are examined. Those with these bonds present in their lives
are expected to have lower recidivism rates as well. Using analysis of variance and
multivariate regression, it was determined that individuals who participated in the boot
camp program did not have statistically significant lower rates of recidivism. Individuals
with informal social bonds present in their lives did not have statistically significant
better performance rates than those without social bonds during the follow-up period.
Additionally, boot camp subjects did not have a statistically significant rate of more
social bonds. As structured, IMPACT seems unable to achieve the goals of rehabilitation

and recidivism.

ACKNOWLEDGMENTS

This was a long time coming, but well worth the wait. While it seemed like this
was never going to be completed, thanks to a great number of people who supported me,
it is (finally!) complete. Without their assistance, guidance, and advice, I probably never

would have found my way here.

iii

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................... v

INTRODUCTION ............................................................................................................... 1

CHAPTER 1

PRIOR RESEARCH ........................................................................................................... 3
The Structure of Boot Camps ................................................................................. 3
Goals of Boot Camp Programs ............................................................................... 5
Recidivism .............................................................................................................. 6
Rehabilitation .......................................................................................................... 8
Informal Social Bonds ............................................................................................ 9
Aftercare ............................................................................................................... 10
Criticisms .............................................................................................................. 12

CHAPTER 2

ANALYSIS ....................................................................................................................... 14

CHAPTER 3

RESULTS ......................................................................................................................... 21

CHAPTER 4

CONCLUSIONS AND SUMMARY ................................................................................ 3O

BIBLIOGRAPHY .............................................................................................................. 35

iv

LIST OF TABLES

Table l - Descriptive Statistics of Sample Groups: Shock, Prison and Probation .......... 22

Table 2 - ANOVA Statistics for Group and Technical Violations and Group and
Rearrests ............................................................................................................................ 23

Table 3 - ANOVA Statistics for Technical Violations and the Social Bonding

Variables ........................................................................................................................... 24
Table 4 - ANOVA Statistics for Rearrests and the Social Bonding Variables ............... 25
Table 5 - ANOVA Statistics for Group and the Social Bonding Variables .................... 26
Table 6 - Multivariate Linear Regression for Technical Violations ............................... 27
Table 7 - Multivariate Linear Regression for Rearrests .................................................. 28

Table 8 - Multivariate Linear Regression for Technical Violations Excluding Marital
Status ................................................................................................................................. 29

Table 9 - Multivariate Linear Regression for Rearrests Excluding Marital Status ......... 29

INTRODUCTION

Alternatives to traditional incarceration have been explored in recent years in an
attempt to find solutions for the problems of the criminal justice system. One such
alternative is shock incarceration, more commonly known as boot camp. Boot camp
programs are used because officials believe that boot camps are cheaper than traditional
incarceration and help reduce prison crowding. The public finds these programs
attractive because they appear tough on offenders and satisfy the need to punish them
(Bourque, Han and Hill 1996a).

Boot camps share a common structure. They were first used for younger
offenders convicted of their first offense, usually nonviolent (Benekos 1995). The basis
of these programs is a strong sense of military discipline that promotes physical exercise
and teamwork. It is hoped that an emphasis on these aspects will teach participants skills
that they have been lacking, such as taking responsibility for their actions. In addition,
the intensity and severity of the program may deter them from re-offending.

These programs may also include rehabilitative components: educational
programs, vocational training, treatment programs for drug and alcohol addictions, stress
management and nonviolent conflict resolution (MacKenzie and Souryal 1991). Some
programs do not offer rehabilitative programs due to the short duration of the program.
There is not a sufficient amount of time to achieve a successful rehabilitation of
participants and including rehabilitative components is considered a waste of time and

resources in these types of camps.

Most programs also lack significant aftercare and follow up services for
participants who graduate from the program. While most programs offer some type of
aftercare service to graduates, they do not last long enough, are not intensive enough, or
do not offer enough services to graduates to help them through the difficult transition
from inmate to citizen. The lack of aftercare and follow-up services affects the
formation of conventional informal social bonds that can be important to the success of a
boot camp graduate. These informal social bonds include a stable marriage, steady
employment or the completion or continuation of a phase of education (Sampson and
Laub 1990). If these bonds have been established, they can provide support to the
individual and exert social controls over the individual that can influence him or her to
resist criminal activities (Sampson and Laub 1990).

This thesis examines whether boot camps are able to achieve the goals of
rehabilitation (changing offender behavior) and reduced recidivism (not committing
future crimes). If they can, boot camp graduates would have lower recidivism rates than
comparison groups. Two variables will be examined as a measure of rehabilitation and
recidivism to find out if boot camp graduates are less likely to re-offend upon release
implying that they have been rehabilitated The first variable that will be examined is the
number of rearrests. Subjects who are rearrested demonstrate a lack of rehabilitation and
a failure of the criminal justice system to deter their recidivism. The second variable to
be examined is technical violations, which include drinking alcohol and missing
appointments with the parole or probation agent. The average citizen is allowed to do

these things without fear of sanctions from the criminal justice system, but parolees and

probationers face this possibility. Committing a technical violation implies that the
individual may have been rehabilitated but may be having problems adjusting to free life.
In addition to rehabilitation and recidivism, informal social bonds will be
examined to ascertain if individuals who form these bonds perform better than those who
do not. Boot camp graduates will be examined to specifically determine if they are
forming more of these informal social bonds. Marriage, education, work and living with
family will be examined to find out if informal social bonds are being formed. It is
hypothesized that if the bonds are being formed, those who have these bonds present in
their lives have lower recidivism rates than those who lack these bonds. Problems
adjusting to free life may be exacerbated by the lack of informal social bonds leading to a

lack of social controls, which may make the individual more likely to re-offend.

PRIOR RESEARCH

The Structure of Boot Camps
The purpose of boot camps is strongly linked to traditions in American culture
and correctional history (MacKenzie and Parent 1992). It is thought that individuals
living in dysfunctional environments that lack positive organization, such as a stable
family structure and a lack of “personal discipline” commit crimes. Boot camps are
designed to create inner control over undesirable behaviors in the offender (Benda,
Toombs and Whiteside 1996; Smykla and Selke 1995).

Programs can last from six weeks to six months (MacKenzie, Shaw and Gowdy

1993). Daily life is intentionally intense and demanding. Inmates participate in physical
labor and military drill, and they are expected to maintain their living quarters and
comply with authority. A structured lifestyle is believed to be necessary to give the
offender the basic skills they need to survive in a law-abiding society and to assume
responsibility for their actions (Mack 1992). Offenders are thought to benefit from this
environment by teaching them obedience to authority and respect for others. Participants
accept responsibility for their actions and learn the self-discipline they are lacking
(Austin et al.1993; Parent 1994).

Beyond the physical aspects of boot camp programs, some programs place a
strong emphasis on treatment components that can include education, counseling and
vocational training (MacKenzie 1990). The treatment received in boot camp is intended
to promote change and to keep the participant from engaging in criminal behavior.
Exposure to education, vocational training, treatment programs and counseling will
instill positive, law abiding values in participants and will aid graduates in finding legal
employment upon release (Austin et al. 1993). The disciplined lifestyle of boot camp
will help enhance a positive self image with counseling and education. The result of the
combination is hoped to be rehabilitative for the offender and help prevent future
recidivism (Salerno 1994).

Participants volunteer for these programs after being selected from a pool of
qualified offenders (MacKenzie 1990). Selection criteria vary but generally participants
must be first time nonviolent offenders. They must be able to participate physically and

mentally. The definition of “first time offender” varies from state to state and can

include first state felony incarceration to first felony conviction (MacKenzie and Parent

1992).

Goals of Boot Camp Programs

Early reports regarding boot camps claimed they worked well, and more
programs appeared as these claims continued (MacKenzie and Shaw 1993; Parent 1994).
The lure of these programs was fueled by the notion that traditional prisons do not
rehabilitate offenders (Wamock and Hunzeker 1991). The public desire for offenders to
be held accountable for their actions through serious and severe punishment was
enhanced by the media images of a program that provided a quick fix solution through
harsh punishment and discipline (Benekos and Merlo 1995; Parent 1994;).

While the public and politicians may support boot camps, every program is
operated differently and the key components of effective treatment have not been
identified or examined. (Cowles et al. 1995; Parent 1994; Simon 1995). Goals are
vaguely defined and thus difficult to determine if they are being accomplished
(MacKenzie, Corbett and Petersilia 1994). Precise definition of these goals is necessary
to design a program that is effective and worthwhile to discover if boot camps are
achieving stated goals (Parent 1994; MacKenzie and Parent 1992).

Some goals are common to all programs. There are two primary goals of boot
camp programs. One is specific deterrence: the use of punishment after a crime is
committed to discourage the offender from committing it again. The second goal is

general deterrence, or the prevention of crimes before the act through threats or example.

Other goals include rehabilitation, punishment, incapacitation, controlling overcrowding
and reducing the costs of incarceration (Finckenhauer 1982; Mack 1992). These goals
can be difficult to set or define. Some programs do not consider punishment an
important goal but stress deterrence or reduction of prison crowding (MacKenzie 1994;
MacKenzie and Souryal 1991). Prediction of the effectiveness of deterrence is difficult
because actual deterrence is also dependent on how much risk an offender calculates

before committing a crime, a difficult idea to measure (Finckenauer 1982).

Recidivism

Limited evaluative research exists on the overall effectiveness of boot camp
programs and more is needed to find out if program goals are being accomplished
(Cowles, et al. 1995; Souryal and MacKenzie 1995). The few studies that have been
conducted have not used an experimental design using random assignment to shock
incarceration versus prison terms. This lack of random assignment threatens the validity
of the studies (Benda et a1 1996; Souryal and MacKenzie 1995) and does not allow the
researchers to determine if differences in recidivism can be attributed to boot camp
(Benda et al 1996).

Currently, no evidence exists that boot camps significantly reduce the recidivism
rates of program graduates. No state has reported a statistically significant decrease in
recidivism rates of boot camp graduates when compared with similar offenders who
served sentences that included community service, jail or prison terms or probation

sentences (MacKenzie and Souryal 1991). Boot camp graduates generally have

recidivism rates comparable to similar offenders who spent time in prison (MacKenzie
1990; Parent 1994). One problem with attempting to measure the achievement of this
goal is that recidivism is a difficult concept to measure (US GAO 1993). Studies reach
different results based on definitions of recidivism or the methods used to measure it.
Since program definitions and participant composition vary from state to state, results are
not comparable creating difficulty in drawing meaningful conclusions among the states
(MacKenzie 1990).

The few studies conducted to study recidivism rates of boot camp graduates have
been disappointing (MacKenzie and Shaw 1993; Sechrest 1989). Groups of boot camp
graduates compared with inmates who served traditional prison sentences have only
marginally lower recidivism rates that tend to diminish over time (US GAO 1993).
MacKenzie and Shaw (1993) found no significant difference in arrests, returns to prison
or failures on parole when groups of boot camp graduates were compared with groups of
inmates who served prison sentences. Return to prison rates were higher for boot camp
graduates for technical violations than new felony convictions.

There is little evidence to support the idea that boot camps can lower the
recidivism rates of program graduates. Boot camps may increase the rate of recidivism
because of the increased control over graduates during parole (Benda, et al. 1996).
While this may seem discouraging, the results thus far are not conclusive and this does
not indicate that boot camp programs have failed (Parent 1994). It may be too soon to
tell how effective these programs are, and this suggests the need for more research and

long term evaluation (Cowles et al. 1995; US GAO 1993).

Recidivism can still be considered an indicator of the ability of the program to
achieve the goals of rehabilitation and specific deterrence (Coyle 1990). Careful study
should be conducted on the types of offenders selected to participate in the programs to
see if those currently being chosen are selected based on the perception that they will
succeed (Souryal and MacKenzie 1995). If participants are being chosen based on a “gut
instinct” that the individual will succeed, selection criteria should be reviewed before

boot camps can be considered a successful alternative to incarceration.

Rehabilitation

Achievement of the rehabilitation goal is linked to decreased recidivism. The
model of boot camp assumes that exercise and drill will assist in rehabilitating
participants (Wamock and Hunzeker 1991). These activities are assumed to give
participants self-esteem, self-discipline and responsibility (all things they are thought to
lack) which would give them a sense of achievement (Wamock and Hunzeker 1991).
Once they have these skills, they may carry this over to their free lives, reducing the rate
of recidivism. States that do include rehabilitative components in their programs focus
on having participants deal with substance abuse, job skills, illiteracy, inadequate social
skills, vocational training, health instruction, and AIDS awareness (Benekos 1995).

Rehabilitation is achieved in two ways, through transference and treatment (Mack
1992). If rehabilitation through transference occurs, the disciplined and structured
lifestyle of boot camp will be carried over to the outside world and continued to be used

by the graduate (Mack 1992). Treatment deals with substance abuse problems, hostile or

violent feelings and improving job skills (Osler 1991 ). By focusing on these issues,
treatment allows participants to face and resolve any of these issues (Osler 1991)
Successful rehabilitation through either of these methods will lower recidivism rates of
graduates. However, no data exist to measure rehabilitation as a separate concept to

determine if rehabilitation without recidivism is being achieved (Mack 1992).

Informal Social Bonds

An additional factor can affect recidivism and it should be considered in program
design: the informal social bond. Informal social bonds are what link individual
members of society to one another and to other social institutions such as work, family
and school (Laub and Sampson 1993). This “web” of informal bonds is considered the
“conventional” bond to society (Laub and Sampson 1993; Needels 1996).

Interaction with social institutions exerts informal social control over individuals,
decreasing crime and deviance (Sampson and Laub 1990). These informal social
controls modify the path toward criminality and deviance (Laub and Sampson 1993). In
addition to creating an inhibition against criminal activities, individuals also find
themselves in more structured and routine activities. As involvement in conventional
activities increases, involvement in deviant activities decreases (Horney, et al. 1995). If
these social bonds are weakened or broken, an individual becomes more likely to engage
in deviant activities (Laub and Sampson 1993).

Marital attachment is one of the strongest predictors of adult criminality.

Individual who are married are far less likely to engage in criminal activities (Horney, et

al. 1995). Similarly, individuals attending school are also less likely to become involved
in crime and commitment to conventional educational and occupational goals has a
similar effect (Homey, et al. 1995). Job stability can also effect future criminal
behaviors (Sampson and Laub 1990).

Sampson and Laub (1990) examined the influence of social bonds on individuals
and their criminal behavior. They found that “social bonds to the adult institutions of
work, education and the family exert...influence on adult crime...” Individuals with
strong bonds to job, education, family and spouse were less likely to engage in deviant
behaviors. Those with weaker bonds were more likely to engage in these activities.
Homey, Osgood and Marshall (1995) found similar results. Individuals were less likely
to engage in criminal activities when they had formed bonds to a wife, were in school or

working.

Aftercare

Critics have observed that boot camps seem unable to achieve many of the stated
program goals, and they claim that this is because of the lack of significant aftercare
services for graduates. It seems unlikely that the short boot camp experience, coupled
with the lack of aftercare, could cause a complete change in the offender (Souryal and
MacKenzie 1995). Without bridge services (afiercare programs), the rate of recidivism
among boot camp graduates may not decrease if the Skills and techniques participants
learned are not carried over into their lives as citizens. (Osler 1991). Change for

participants can only occur when treatment and education are offered as part of the

10

program and for extensive periods afier the offender’s release, something programs are
less likely to provide (Kappeler, Blumberg and Potter 1996).

Many graduates leave the program with the same disadvantages they faced before
incarceration, such as lack of work skills or history, and inadequate education. After
release they have the added burden of a felony conviction (Kappeler et al 1996). Leaving
the boot camp program means an abrupt end to the services, training and treatment they
received and participants find themselves having a hard time readjusting to life as a
citizen (Osler 1991). The sudden lack of support upon release leaves the graduate in an
uncertain situation and can eliminate the possibility of success (Osler 1991). These
“reentry” difficulties occur because the graduate moves from a highly structured routine
and controlled life to a life without these controls in which choices must be made
without guidance. The inability to cope with this sudden change combined with a lack of
support from either the family or outside sources can cause problems for the graduate
(Kappeler et a1 1996). Since the offender is returning to same environment with the
same conditions that influenced him or her in the first place, an extended period of
supportive services and opportunities must be provided to the offender to aid in the
transition to a law-abiding lifestyle (MacKenzie and Parent 1995).

To ensure that the changes a participant started to make during the program are
continued upon release, aftercare services should be provided for at least nine months to
one year following release (Osler 1991). These aftercare services are essential for the
graduate to succeed in the future. Giving graduates the opportunity to continue learning

and using what has been learned in the program will increase their chances for success in

11

the future. It is important that graduates have resources available to assist them with
areas such as education and employment. By providing graduates with an avenue for
success (e. g., job skills) it is arguable that they will be less likely to commit crimes in the

future.

Criticisms

While past studies have shown that black arrestees are more likely to recidivate
than white arrestees (DeJong 1997; Smith and Akers 1993; Uggen and Kruttschnitt 1998)
studies that examine boot camps have not examined outcome performance along racial
lines. Official records show that more blacks than whites are involved in crime (Siegel
and Senna 1991) and are consistently over-represented in arrest data (Harris 1991).
Several explanations exist for these findings. One explanation is that blacks are over
represented in the criminal justice system due to biases of the system (police, courts,
juries, etc.) (Siegel and Sienna 1991). This possible discrimination may lead to
prejudices in choosing which behaviors are criminally sanctioned and which are not
(Harris 1991).

Another explanation is that the disparities are not a result of bias but of
ecological differences that create breeding grounds for crime (Siegel and Senna 1991).
More blacks than whites reside in some areas characterized by a lack or absence of legal
employment and where poverty and limited wealth are the norm (Siegel and Senna
1991). Other explanations exist for why more blacks than whites are involved in the

criminal justice system and none are conclusive (Siegel and Senna 1991). There are no

12

data that can directly explain the positive correlation between race and arrest (Hanis
1991)

Additional criticisms exist regarding boot camp programs. Some critics feel that
the only reason these programs continue to grow is because legislatures believe that
military training can cause a lasting behavioral change in participants (Parent 1994).
Other critics argue that while most programs require that participation be voluntary,
participants are actually coerced and no one can actually “volunteer” for the program
(Salerno 1994). There is a growing awareness about the shortcomings of the military
model - a model motivated by punishment and retribution rather than rehabilitation
(Sharp 1995).

Prisoner abuse is another criticism of boot camps. Guards working in these
programs have broad discretion and are generally untrained (Osler 1991). This leaves
open the possibility of the abuse of authority and may bring out the “dark side” of the
officers (Benekos 1995; Sechrest 1989). Boot camp prisons might attract the type of
staff that might use their position of power and control to take advantage of the inmates
and abuse them (MacKenzie and Parent 1992). The use of negative leadership to tear
down and then build up inmates teaches them to react aggressively (Morash and Rucker
1990). Each instructor is different, and inmates will encounter different officers
throughout the program. The unpredictable style of each guard can lead to stress and
cause confusion from inconsistent standards that can create the potential of abuse by the
guards and negative outcomes for the participants (Benekos 1995; Morash and Rucker

1990).

13

Perhaps the most serious criticism is that the definitions of acceptable and
unacceptable behavior is based on the middle class idea of punishment (Sechrest 1989).
Neither boot camps nor traditional prison sentences will erase the social conditions or
problems parolees must face upon release (Sechrest 1989). There will be no long-term
deterrent effect if no satisfaction can be found through a legitimate lifestyle (Sechrest
1989). Boot camps were developed as a quick fix solution to a growing problem that
satisfied the public need to be “tough on crime” (Osler 1991; Benekos 1995). However,
this “well intentioned...legacy of failure”(Benekos 1995) postpones the “real” issue of

having to confront the problem of what to do about crime and criminals menekos 1995).

ANALYSIS

To better determine the effects of boot camp programs on recidivism rates of
graduates and whether subjects with informal social bonds will perform better upon
release, secondary data analysis will be conducted The data were collected by Doris
MacKenzie, James Shaw and Voncile Gowdy on the Louisiana Department of Public
Safety and Corrections (LDPSC) boot camp program IMPACT (Intensive Motivational
Program of Alternative Correctional Treatment), over two years (1987-1989).
Participants in the program spent 90 to 180 days in a medium security prison engaging in
traditional boot camp activities: drill, exercise and labor. The program did not include
any rehabilitative aspects such as job training or education. Those who successfully

completed the program were released to Intensive Community Supervision.

14

The researchers examined five offender samples: those who completed the
program (program completes), those who started but dropped out of the program
(program dropouts), offenders sentenced to probation and not the program (non-
participant probationers), parolees who did not participate in the program (non-
participant parolees) and incarcerated inmates who were not selected for the program
(non-participant incarcerated). The original five groups were eventually collapsed into
three groups: non-participants, program completes and program dropouts. The three non-
participant samples were matched as closely as possible to the program participant
sample by selecting for comparison only subjects who would have been eligible to enter
the program.

Samples in this study are somewhat arbitrary. While MacKenzie et al. (1987)
used a quasi-experimental design to select and match offenders from each of the sample
groups, inclusion in these groups was originally decided by a pre-sentence
recommendation by the prosecutor and a judge who chose to follow or disregard the
recommendation. While both the sentencing recommendation and judge’s decision were
based on a number of legal factors, the possibility exists that both the original
recommendation and final sentence were influenced by other factors. These factors may
have influenced which offender was sentenced to the IMPACT program, probation or
incarceration, thus affecting the original sample pool and data samples.

The subjects were followed for one year after release from the IMPACT program
while on Intensive Community Supervision, while on parole or during probation. During

this time they did not have access to continuing care. A total of 116 subjects completed

15

the boot camp program, 108 individuals were in the probation sample, 74 individuals
were in the parole sample, and 98 individuals were in the incarcerated sample.

Participants in the study completed a self-report questionnaire that asked various
questions about their boot camp experience and their adjustment to free life one or more
times during the follow-up. These self-reported data were combined with information
collected from official records of LDPSC and questionnaires filled out by the
parole/probation agent during the follow-up. Information from LDPSC consisted of
demographic information and sentence information such as sentence length, release date
and criminal history. Parole performance, as evaluated by the agent, was completed
each month for the entire 12 month follow-up, (excluding rearrests which were only
followed for the first four months of the follow-up), and included items relating to
performance on community supervision, performance at work and in school, substance
abuse counseling, and further contact with the justice system.

The use of self-report data to measure variables presents problems with the
accuracy of reported information. The questionnaire asked respondents personal
questions regarding their past activities. Some respondents may have feared that the
information would be released to the authorities, so they may have lied about their
unreported criminal history. Another problem is that these data were collected 12 years
ago. Changes in program structure or participant selection processes may have changed
since these data were collected. Consequently, any conclusions that are drawn may be
accurate based on the data but inaccurate based on current operating procedures. It is

possible that the conclusions and recommendations made in this study may have already

16

been implemented by the IMPACT program.

The validity and reliability of this research may be threatened by several factors.
Because this study relies heavily on official records as a measurement tool, there is a
threat to the reliability of the research. Using official records is a problem because
much of the data regarding technical violations may be tainted. Since it is the parole or
probation agent who decides whether to sanction a technical violation, two individuals
may receive two different sanctions. No two agents may use the same criteria to decide
if the same technical violation merits a formal sanction versus and informal sanction.
This may inflate the rate of technical violations for any group.

Construct validity is threatened because the empirical measures used may be
incomplete. Strictly measuring the numerical results of how many graduates fail on
parole (and how quickly they fail) does not reveal internal changes that may have
occurred in a participant. Release requirements may have been violated when a positive
change did occur but a lack of support existed. The instrument used here also does not
measure the strength of a social bond. While a bond may have been formed by the
subject with one of the variables, this data does not measure how strongly the subject
may have bonded with it. Strength of a social bond may affect how well an individual
does or does not perform during post-release.

A threat to the external validity of this study is generalizability. While the results
of this study may apply to other boot camp programs designed in a similar manner, the
results may not apply to all programs. All boot camp programs are not designed alike

and do not offer all the same services to participants. This difference in design could

l7

stem from a subculture that exists in Louisiana and nowhere else. Attitudes and
perceptions of criminal behavior and how to deal with that behavior would be different
in one subculture to the next, affecting how that group of people choose to deal with
criminals. The difference in attitudes could affect program design, in turn affecting
individual experiences in the boot camp.

A third, unanticipated variable may have also affected the results of this study.
This third variable is known as the “aging out” phenomena that has been well
documented in criminology literature (Harris 1991). According to the literature, when
aging out occurs, criminal involvement decreases with age. Generally the peak age for
offending is twenty then drops off as the person ages (Harris 1991). This finding will
usually hold true across historical eras, geography and types of crimes (Harris 1991).
This implies that the reason offenders stop committing crimes is not due to the deterrent
effects of any punishment experienced by the offender but simply because they got “too
old” (Harris 1991).

The most significant reason that criminals age out of crime is credited to the
acquisition of meaningful and conventional bonds such as gainful employment,
continued education, the completion of a phase of education and marriage (Steffensmeir
and Streifel 1991). Employment must be meaningful and include financial security and
the possibility of advancement. Once steady employment is found, the offender’s
attention will shift from criminal activity to law-abiding activity. Their daily routine will
change leading that person away from illegal activities and toward legal ones.

Establishing and maintaining these bonds orient the offender to conventional society, and

18

decreases the opportunities for committing criminal acts (Steffensmeir and Streifel
1991)

It is hypothesized that boot camp graduates will not have less technical violations
and rearrests than comparison parole and probation groups during the follow—up period.
Specifically, boot camp graduates are expected to have similar technical violation and
rearrest rates to the parole and probation groups during the follow-up period.
Alternatively, if this boot camp program is successful in achieving the goals of
rehabilitation and reduced recidivism, the boot camp group will have smaller rates of
technical violations and rearrests. It is also hypothesized that subjects who do not form
informal social bonds during the follow-up period will not perform well during the
follow-up period. Subjects who have these bonds present in their lives are expected to
have less technical violations and rearrests. Boot camp graduates are not expected to
have more social bonds in their lives.

Each of the three groups of offenders (shock, prison, probation), will be
compared to determine successes and failures, rearrested (yes or no), and technical
violations (yes or no). Rearrests and technical violations are used as a measure of
recidivism and rehabilitation. Technical violations are defined as positive drug tests,
curfew violations, failure to tell the agent of their whereabouts, and missing
appointments. These measure will be examined by race (black and white), to determine
if black subjects and white subjects recidivate at Similar rates.

Four variables will be used to examine the formation of informal social bonds

and their possible effects on subjects’ post-release performance: employment (working or

19

not working), education (completed a GED, did not complete a GED), living
arrangements (family or not family), and marital status (married, not married). The
presence of informal social bonds in subjects with technical violations and rearrests will
be examined to determine if the bonds are not as prevalent in those with technical
violations and rearrests. It is hypothesized that a subject who recidivates will lack those
bonds. Those who have these bonds present in their lives should perform better with
regards to technical violations and rearrests.

In the original data set, there is information available for each subject for each
month of the [2-month follow-up (rearrests were only followed for the first four months
of the follow-up). The data consist of two types of responses. The first is a positive or
negative response for the month in which an event occurred. For example, the variable
completing a GED was measured as a “yes/no” response for every month (did the subject
complete a GED this month?) The variable indicating with whom the subject lived with
was measured as a “family/not family” response for each month (did the subject live with
family this month?) The second type of variable indicates how many months a subject
performed positively on a particular variable. The work (employment) variable was
recorded as a numeric value. For example, a value of eight would mean a subject worked
for months one through eight of the follow-up period. A value of five would mean a
subject worked during the first five months of the follow-up.

Three groups will be examined: those who had been through the shock program
and graduated (shock), those who had served traditional prison sentences and are on

parole (parole), and those who were sentenced to probation and serving that sentence

20

(probation). The original data sets were collapsed into one aggregated data set for
analysis purposes. The original codings remained intact but were difficult to analyze
correctly to answer the questions in this study. For analysis purposes, some variables
were recoded into yes/no questions. The work (employment) variable was recoded to ask
“Did the subject work full-time at all during the follow-up?” Recoding the data in the
manner makes it possible to look at subject and group performance for the entire 12

months at once instead of having to examine the data month by month.

RESULTS

As indicated in Table 1, the majority of cases are in the probation group (42.6%).
A majority of sample members are black (64.0%) while 34.5% are white (1.6% of the
sample is classified as “missing” for race). The average age of subjects is 23; however,
58.1% of the sample’s age is missing. Using self-reported information at the time of
their arrest, 22.8% of the subjects reported they were employed either full or part time.
Only 13.2% of the subjects had graduated from high school or completed a GED, and
15.5% had completed some high school. Only 5.8% were married at the time of their
arrest.

Two of the variables in this study had Significant reporting problems. The age
variable will not be used in this analysis because nearly 150 subjects did not have an age
recorded. No reason is given for this in the literature provided with the data set. Marital

status has a similar problem. Almost 61% of the subjects did not have a marital status

21

recorded. Again, no explanation is given for this. The marital status variable is used in
some models and omitted in others to include a greater number of cases in the analysis.
Both types of models are included and analyzed in this thesis to see if significant
differences emerge.

Table 1: Descriptive Statistics of Sample Groups: Shock, Prison and Probation
(n=258)

 

Total per group Percent Number
Shock 28.7% 74
Parole 28.7% 74
Probation 42.6% 1 10

Race
White 34.5% 89
Black 64.0% 165
Missing 1.6% 4

Mean Age 23 years
Missing 58.1% 150

Subjects Working at Time of Arrest 22.8%
Completed High School 13.2%

Marital Status
Married 5 . 8%
Not Married 30.7%
Missing 60.9%

 

Analysis of variance will be conducted to examine two things: if the shock group
has the same recidivism rate as the parole and probation groups and if they have more or
less social bonds. The groups are compared on the rearrest and technical violation
variables first to measure overall performance. The rearrest and technical violation

variables are compared with each of the social bonding variables to see if they were or

22

were not present when the subject recidivated The groups are then compared with the
social bonding variables to see which group had the highest mean.

Table 2 shows the rate of technical violations across group (shock, parole and
probation). The relationship between technical violations and group is significant, but
eta-squared is very low, .04. This low eta-squared suggests a very weak relationship
between the groups and technical violations. The probation group has the highest group

mean number of technical violations, 2.55, and the parole group the lowest, 1.49.

Table 2: ANOVA Statistics for Group and Technical Violations and Group and
Rea rrests‘

 

Technical Violations

Mean Significance of F n2
.01 .04
Shock 1.74
Parole 1.49
Probation 2.55
Total 2.01 (162)
Rearrests
Mean Significance of F n2
.42 .0]
Shock .42
Parole .29
Probation .38
Total .37 (77)
‘ alpha = .05

 

Comparing technical violations and the social bonding variables two variables are

significantly related, completing a GED (F = .02) and working (F = .00) (see Table 3).
23

The group mean for subjects who did complete a GED is 2.65 and eta-squared is 2.1%,
resulting in a very weak relationship. For working, the mean is higher for subjects who
did not work during the follow-up, 2.82, and eta-squared is a little higher, 3.5% but also a

very weak relationship. The other variables are not significantly related.

Table 3: ANOVA Statistics for Technical Violations and the Social Bonding
Variables‘

 

Variable Mean Significance of F n2

Compete a GED .02 .02
No 1.83
Yes 2.65

Working .00 .03
Not Working 2.82
Working 1.76

Marital Status .46 .01
Not Married 1.73
Married 1.33

Living Arrangement .15 .01
Not Family 1.53
Family 2.13

" alpha = .05

 

The social bonding variables are not significantly related to the rearrest variable (Table

4).

24

Table 4: ANOVA Statistics for Rearrests and the Social Bonding Variables‘

 

Variable Mean Significance of F 112

Complete GED .52 .00
No .35
Yes .41

Working .09 .01
Not Working .48
Working .32

Marital Status .56 .00
Not Married .47
Married .37

Living Arrangement .63 .00
Not Family .33
Family .38

’ alpha = .05

 

Each group is compared on the social bonding variables (see Table 5).
Completing a GED is the only variable significantly related to group (significance of F =
.00). The shock group had the highest group mean for completing a GED (39% of those
in the shock group completed a GED). However eta-squared is low (.07) suggesting a
weak relationship. Marital status was unavailable for every member of the probation
group. This may have affected the outcome for this comparison. There is no explanation

for the missing marital status for the probation group.

25

Table 5: ANOVA Statistics for Group and the Social Bonding Variables‘

Variable Mean Significance of F 112
Complete GED .00 .07
Shock .39
Parole .10
Probation .25
Working .74 .00
Shock .79
Parole .75
Probation .74
Married .53 .00
Shock 1.84
Parole 1.99
Probation Missing
Living with Family .61 .00
Shock .80
Parole .84
Probation .85
‘ alpha = .05

 

Multivariate linear regression is used to examine the relationship between the
technical violation and rearrest variables and the social bonding variables: working,
completing a GED, marital status and living arrangement. Technical violations are
examined first and the results are in Table 6. This model includes whether the subject
completed a GED, marital status, employment status, living arrangements and race. The
value of r-squared indicates that using these variables explain 19.2% of the variation in
the number of technical violations. Completing a GED is linearly related to technical

violations (significance of t = .03). The value of beta for this variable is small and

26

positive (.26) showing that completing a GED during the follow-up increases the number
of technical violations for those subjects, controlling for all other variables. The
remaining variables, working, samfle, living arrangement, marital status and race are not

significantly related to technical violations.

Table 6: Multivariate Linear Regression for Technical Violations‘

 

Variable Coefficient Beta Significance
Constant .49 .76
Complete GED 1.17 .26 .03
Married .13 .02 .84
Work Full-Time -. 10 -.20 .08
Live with Family .31 .06 .59
Black .51 .12 .29
Group .05 .01 .93

‘ alpha = .05, r2 = .19, significance ofF = .01

 

The linear regression model for rearrests and the social bonding variables is not
statistically significant (F = .29 -— see Table 7). Using these variables explains only

9.06% of the variation in rearrests. None of the variables are linearly related to rearrests.

27

Table 7: Multivariate Linear Regression for Rearrests‘

 

Variable Coefficient Beta Significance
Constant .56 .25
Complete GED -.05 -.O4 .74
Married -.08 -.05 .70
Work Full-Time -.03 -.22 .08
Lived with Family .27 .17 .13
Black .07 .06 .62
Group -.19 -. 14 .25

‘ alpha = .05, 12 = .09, significance of F = .29

A second set of regression models that omits marital status is conducted to
include more cases due to the large number of missing cases for marital status. The first
dependent variable analyzed is the number of technical violations (see Table 8). In this
model r-squared is weak, only 13.7% of the variation in technical violations is explained
using the variables in this model. Completing a GED is significantly related to the
number of violations. The relationship has a small, positive beta (.14) showing that
completing a GED increases the number of technical violations during the follow-up
period. Race emerged as a significant relationship. Coding for the race variable is 0 ==
white, 1 = black. This beta suggests that black offenders have more technical violations
than white offenders. Working full-time during the follow-up period also emerged as a
significant variable (.00). This variable has a small, negative beta (-.17) indicating that

subjects who work have less technical violations than subjects who do not work.

28

Table 8: Multivariate Linear Regression for Technical Violations Excluding Marital
Status‘

. Variable Coefficient Beta Significance
Constant -.58 .54
Complete GED .78 .14 .04
Work Full Time -.10 -. 18 .01
Live with Family .77 .ll .08
Black .99 .20 .00
Group .38 .06 .28

‘ alpha = .05, r2 = .14, significance ofF = .00

 

Rearrests are examined excluding marital status and the relationship is not
significant (see Table 9). Significance of F is equal to .17 and r-squared is only 14.7%
suggesting that very little of the variation in rearrests is explained by these variables.
Working was found to be linearly related to rearrests (significance of t = .03). Beta was
again small and negative showing that subjects who worked has fewer rearrests than

subjects who did not work.

Table 9: Multivariate Linear Regression for Rearrests Excluding Marital Status‘

 

Variable Coefficient Beta Significance
Constant .51 .05
Complete GED .01 .00 .96
Work Full-Time -.02 -. 16 .03
Live with Family .17 .09 .17
Black -.04 -.03 .69
Group -.08 -.O6 .41

‘ alpha = .05, r2 =03, significance ofF = .17

 

29

CONCLUSIONS AND SUMMARY

The results found here should be viewed cautiously for a number of reasons. The
type of data used in this thesis does not measure the quality of change that may have
occurred in the subjects. Performance is based strictly on numerical outcomes and
cannot measure internal changes that may have resulted from any form of punishment.
As stated earlier, no qualitative way exists to measure rehabilitation of the subjects who
may only be having problems adjusting to free life versus subjects who are not
rehabilitated.

Technical violations and rearrests are used as measures of rehabilitation and
recidivism. However, the actual determination of officially sanctioning a technical
violation may be influenced by the individual parole agent or probation officer deciding
whether a violation has occurred and if it Should be sanctioned. While there are some
clear-cut decisions, other decisions were not so easily made. In these cases, individual
decisions by probation officers and parole agents may directly affect how many subjects
received a sanction for a technical violation and these decisions would affect the
outcomes of these models.

On certain variables, one group may have performed better than another. As a
numeric measurement there is no explanation why one group may have performed better
or worse. No method exists to measure what kind of access the different groups may
have had to different resources during their incarceration or probation or how this may

have affected their access to resources in the community.

30

Using this type of data does not allow for measurement of the quality or strength
of social bonds that may have been formed during the follow-up period. Many subjects
may have formed the types of bonds measured but may not have bonded fully with them.
Without a strong social bond, the subject may not be as successful as a subject who
formed a stronger bond. A numerical count of how many subjects formed a bond does
not help explain why those with the bonds may have performed differently than those
who did not form the bond or how strongly bonded the individual was to the variable.

Overall, the shock group did not perform any better than the parole and probation
groups. Although analysis of variance revealed that the shock group did have fewer
technical violations than the probation group, the shock group had more technical
violations than the parole group. The shock group also had the highest group mean for
rearrests. Having more technical violations than at least one group and more rearrests
than both groups suggest that subjects may not have been rehabilitated subsequently not
reducing recidivism rates. This finding suggests that this boot camp is not successful at
achieving rehabilitation and recidivism goals.

Technical violations were significantly related to completing a GED and working.
Analysis of variance showed that subjects who did not work had more technical
violations than subjects who did work. This finding was not surprising. Subjects lacking
the social bond to work were expected to perform worse than subjects who did have this
social bond. The surprising result was that subjects who completed a GED during the
follow-up had more technical violations than subjects who did not. This result was also

true for GEDs and technical violations in both of the regression models: completing a

31

 

GED increased the number of technical violations. Interestingly, the shock subjects were
the most likely to have completed GEDs during the follow-up period. The reason why
shock subjects completed the most GEDs and why those subjects were more likely to
have technical violations cannot be ascertained from this data. This finding requires
further study to decide if this finding is incorrect or if somehow this social bond affected
outcomes differently than expected.

Working full-time was found to be significantly related in the regressions that
excluded marital status to technical violations and linearly related to rearrests. As would
be expected, subjects who worked during the follow-up period (formed an informal
social bond with a job) had less technical violations or rearrests.

Race emerged as significantly related to technical violations in the second set of
regression models. Linear regression suggested that blacks had more technical violations
than whites. Looking strictly at race and the rehabilitation and recidivism variables
showed that 65% of blacks (112 subjects) had a technical violation during the follow-up
compared with only 35% of whites (48 subjects). The findings are similar for rearrests:
64.8% of blacks (53 subjects) were rearrested while only 35.2% of whites (23 subjects)
were rearrested. Many explanations exist for this finding that the data cannot account for.

As structured, the Louisiana boot camp program IMPACT does not seem able to
achieve the goals of rehabilitation and reduced recidivism. Shock subjects had
comparable technical violation and rearrest rates and the social bonds examined
generally did not seem to improve performance in individual subjects. Only one social

bond, education, was significantly related to group, and this bond seemed to have the

32

opposite affect by increasing the number of technical violations of subjects who
completed a GED.

To discover if these results are accurate, a few steps could be taken. A longer
follow-up period could be used to follow subjects. Tracking subjects for longer than one
year may reveal more changes that occur over the long-terrn that are not evident in a
short term follow-up, such as this. More qualitative measures could be used to attempt to
measure the kind of change that took place in the subjects instead of strictly measuring
the numerical outcomes.

While this program seems unable to achieve the goals examined here, the
program itself as presently or previous structured may be considered successful. Many
other factors could be used to measure the success of a program. These results may also
indicate the need for more comprehensive after care services for program graduates to
help them in establishing strong bonds with society to assist them in the transition to a

law-abiding life.

33

 

BIBLIOGRAPHY

34

BIBLIOGRAPHY

Austin, James, Michael Jones and Melissa Bolyard (1993) “The Growing Use of Jail
Boot Camps: The Current State of the Art.” NIJ Research in Brief.

Benda, Brent 8., Nancy J. Toombs, and Leanne Whiteside (1996) “Recidivism Among
Boot Camp Graduates: A Comparison of Drug Offenders to Other Offenders.”
Journal of Criminal Justice 24 (3): 241-253.

Benekos, Peter J and A Merlo eds. (1995) “Shock Incarceration: The Military Model in
Corrections.” Corrections: Dilemmas and Directions, Anderson Publishing
Company, Highland Heights, KY.

Bourque, Blair 8., Mei Han and Sarah M. Hill (1996) “A National Survey of Aftercare
Provisions for Boot Camp Graduates.” NIJ Research in Brief 15 pages. (a)

Bourque, Blair 3, Mei Han and Sarah M. Hill (1996) “An Inventory of Aftercare
Provisions for 52 Boot Camp Programs.” NIJ Research Report. (b)

Cowles, Ernest L., Thomas C. Castellano and Laura A. Gransky (1995) “Boot Camp,
Drug Treatment and Aftercare Interventions: An Evaluation Review.” NIJ
Research in Brief, 178 pages.

Coyle, Edward J. (1990) “Boot Camp Prisons: Criminal Disposition Commission,
Alternatives to Incarceration Committee.” Briefing Report Criminal Disposition
Commission, Newark NJ, 17 pages.

DeJong, Christina. (1997) “Survival Analysis and Specific Deterrence: Integrating
Theoretical and Empirical Models of Recidivism.” Criminology 35 (4): 561-575.

Finckenauer, James O. (1982) Scared Straight and the Panacea Phenomenon.
Englewood Cliffs, New Jersey - Prentice Hall.

Harris, Anthony R. (1991) “Race, Class and Crime.” Criminology Wadsworth, Inc.
Belmont CA.

Homey, Julie, D. Wayne Osgood and Ineke Haen Marshall (1995) “Criminal Careers in

the Short-Term: Inna-Individual Variability in Crime and its Relation to Local
Life Circumstances.” American Sociological Review 60: 655-673.

35

Kappeler, Victor E., Mark Blumberg and Gary W. Potter, eds. (1996) The Mythology of
Crime and Criminal Justice 2nd edition, Waveland Press, Prospect Heights IL.

Laub, John and Robert Sampson ( 1993) “Turning Points in the Life Course: Why Change
Matters to the Study of Crime.” Criminology 31 (3) 301-324.

Mack, Dennis E. (1992) “Combining Shock Incarceration and Remedial Education.”
Journal of Correctional Education 43 (3) pages.

MacKenzie, Doris Layton (1994) “Boot Camps: A National Assessment.” Overcrowded
Times 3 pages.

MacKenzie, Doris Layton (1990) “Boot Camp Prisons: Components, Evaluations and
Empirical Issues.” Federal Probation 54 (3) 44-52.

MacKenzie, Doris Layton and James Shaw (1993) “The Impact of Shock Incarceration
on Technical Violations and New Criminal Activities.” Justice Quarterly 10 (3)
463-487.

MacKenzie, Doris Layton and Claire Souryal (1991) “Rehabilitation, Recidivism
Reduction Outrank Punishment as Main Goals.” Corrections Today 53 (6): 90-
96.

MacKenzie, Doris Layton with Ronald Corbett and Joan Petersilia (1994) “Results of a
Multi-Site Study of Boot Camp Prisons.” Federal Probation 58 (2) 60-66.

MacKenzie, Doris Layton and Dale G. Parent (1992) “Boot Camp Prisons for Young
Offenders.” Smart Sentencing: The Emergence of Intermediate Sanctions eds
James M. Byme, Arthur J. Lurigio and Joan Petersilia Sage Publications, Inc.
Newbury Park.

Morash, Merry and Lila Rucker (1990) “A Critical Look at the Idea of Boot Camp as a
Correctional Reform.” Crime and Delinquency 36 (2) pages.

Needels, Karen (1996) “Go Directly to Jail and Do Not Collect? A Long-Term Study of
Recidivism, Employment, and Earnings Patterns Among Prison Releasees.”
Journal of Research in Crime and Delinquency 33 (4): 471- 496.

Osler, Mark W. (1991) “Shock Incarceration: Hard Realities and Real Possibilities.”
Federal Probation 55 (1) 34—42.

Parent, Dale G. (1994) “Boot Camps Failing to Achieve Goals.” Overcrowded Times 5
(4) 8-11.

36

 

 

Salerno, Anthony W. (1994) “Boot Camps: A Critique and A Proposed Alternative.”
Journal of Offender Rehabilitation 20 (3/4) 147-158.

Sampson, Robert J. and John H. Laub (1990) “Crime and Deviance Over the Life Course:
The Salience of Adult Social Bonds.” American Sociological Review 55 (5): 609—
627.

Sechrest, Dale K. (1989) “Prison ‘Boot Camps’ Do Not Measure Up.” Federal
Probation 53 (3): 15-20.

Sharp Deborah (1995) “Boot Camps - Punishment and Treatment.” Corrections Today
57(3): 551-558.

Siege], Larry J. and Joseph J. Senna (1991) Juvenile Delinquency: Theory, Practice and
Law. 4th ed. St. Paul: West Publishing Company 1991.

Simon, Johnathan (1995) “They Died With Their Boots On: The Boot Camp and the
Limits of Modern Penality.” Social Justice 22 (2) 25-30.

Smith, Linda G. and Ronald L. Akers (1993) “A Comparison of Recidivism of Florida’s
Community Control and Prison: A F ive-Year Survival Analysis.” Journal of
Research in Crime and Delinquency 30 (3): 267-292.

Smith, M. Dwayne, Joel A. Devine and Joseph F. Sheley (1992) “Crime and
Unemployment: Effects Across Age and Race Categories.” Sociological
Perspectives 35 (4) 551-572.

Smykla, John and William Selke eds. (1995) Intermediate Sanctions in the 90 's.
Anderson Publishing and ACJ S. Cincinnati OH.

Souryal, Claire and Doris Layton MacKenzie (1995) “Shock Incarceration and
Recidivism: An Examination of Boot Camp Programs in Four States.”
Intermediate Sanctions: Sentencing in the 1990s eds. John Ortiz Smykla and
William L. Selke; Anderson Publishing and ACJ S, Cincinnati OH.

Uggen, Christopher and Candace Kruttschnitt (1998) “Crime in the Breaking: Gender
Differences in Desistance.” Law & Society Review 32 (2): 339-366.

US GAO (1993) “Prison Boot Camps: Short-Tenn Prison Costs Reduced, But Long-Tenn
1m pact Uncertain.”

Wamock, Kae and Donna Hunzekier (1991) “Prison Boot Camps: Policy Consideration
and Options.” State Legislative Report 16 (1) 10 pages.

37

HICchnN STATE UNIV. LIBRRRIES
illllllllllllllllllllllllllllllllllllllllllllllllllllllllll
31293018129381