IIHIIIWIIHIHHIHIWHIHMIIWIIIIHIHIIIHIWIIHII

  

138
128

  

..TH$

\ " I
(A

m
Rx
< (J
L.)
(1)
CC‘)
\J
v

This is to certify that the
thesis entitled

A TEST OF REINTEGRATIVE SHAMING THEORY’S
CONCEPTS OF INTERDEPENDENCE AND EXPRESSED
SHAME IN RESTORATIVE JUSTICE CONFERENCING

presented by
GINA MARIE HENDRIX

has been accepted towards fulfillment
of the requirements for the

MS. degree in Criminal Justice

 

    
   

//

 

 

l .L ‘f /. .
Major Professor’s Signature

Jar/[y 28, 2Q 2

Date

 

MSU is an Affinnative Action/Equal Opportunity Institution

 

0-02.-.-C-0-0-0-.-O-C-O-I-C-

.n...—.-.A;-.-o-.-.-a-n-a-n‘.-.-.----A‘O-one-n-l—wm'-'-'.

 

LIBRARY
Michigan State
University

 

 

 

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

 

 

 

DATE DUE DATE DUE DATE DUE
OCT 2 7 2007 "5’5 11 25 W
:0“ 9 33' ,0 7.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DateDuepes-ols

._____ ...___—._

A TEST OF REINTEGRATIVE SHAMING THEORY’S CONCEPTS OF
INTERDEPENDENCE AND EXPRESSED SHAME 1N RESTORATIVE JUSTICE
CONFERENCING
By

Gina Marie Hendrix

A THESIS

Submitted to
Michigan State University
In partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE
School of Criminal Justice

2004

ABSTRACT
A TEST OF REINTEGRATIVE SHAMING THEORY’S CONCEPTS OF
INTERDEPENDENCE AND EXPRESSED SHAME IN RESTORATIVE JUSTICE
CONFERENCING
By

Gina Marie Hendrix

This thesis examines Braithwaite’s (1989) reintegrative shaming theory’s
concepts of interdependence and expressed shame. Interdependence is operationalized
through an adaptation of Hirschi’s (1969) social control theory, specifically attachment to
mother, commitment to conventional institutions of church and school, and involvement
in conventional activities. Data from the Indianapolis Restorative Justice Experiment are
used. Bivariate analysis was employed to examine the relationship between diversion
group assignment and levels of interdependency and expressed shame. Multivariate
analysis was employed to examine the relationship of interdependency and expressed
shame with re-offending. Results showed youths that completed restorative justice
conferencing had higher levels of interdependency and expressed shame than control
youths, but interdependency and expressed shame were not predictive of re-offending.

Limitations and implications are discussed.

To my Mom and Dad for their eternal support and love.

iii

ACKNOWLEDGMENTS

I would like to thank my committee members, Dr. Ed McGarrell, Dr. Cynthia
Perez-McCluskey, and Dr. Christina DeJong, for all their guidance. I appreciate all the
time and support they provided me in this endeavor.

I would also like to thank Carol Rapp Zimmermann for taking me under her wing

and providing me with invaluable insight and support.

iv

TABLE OF CONTENTS

LIST OF TABLES .............................................................................................................. vi

LIST OF FIGURES ................................................................................. vii

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

CHAPTER 1

DEFINITION OF RESTORATIVE JUSTICE ................................................... 4
What is Restorative Justice ................................................................. 4
Restorative Justice Models ................................................................. 5

CHAPTER 2

THEORETICAL FRAMEWORK ................................................................. 8
Reintegrative Shaming ..................................................................... 8
Shame .......................................................... 10
Interdependency............................................................... ............ 11
Social Control Theory ..................................................................... 13

CHAPTER 3

DATA AND METHODS ......................................................................... 17
Current Study ............................................................................... 17
The Indianapolis Restorative Justice Experiment ..................................... 17
Sample Description ........................................................................ 19
Research Questions ........................................................................ 22
Variables .................................................................................... 24
Measures of Interdependency and Expressed Shame ................................. 25
Dependent Variable ....................................................................... 29
Control Variables .......................................................................... 30

CHAPTER 4

RESULTS ........................................................................................... 32
Bivan'ate Analysis: Interdependency and Expressed Shame ........................ 32
Bivariate Analysis: Re-Offending ...................................................... 34
Multivariate Analysis ...................................................................... 35

CHAPTER 5

DISCUSSION ...................................................................................... 38
Conclusion and Implications ............................................................. 38
Limitations .................................................................................. 4l

BIBLIOGRAPHY .................................................................................. 44

LIST OF TABLES

Table l Diversion Program Breakdown for Survey Youth .................................... 21
Table 2 Charge Breakdown ............................................ V ............................ 22
Table 3 Components Loading for PCA Extraction .............................................. 26
Table 4 Measures of Interdependency and Expressed Shame ................................. 29
Table 5 Descriptive Statistics for Control Variables ........................................... 31

Table 6 Differences in Interdependency and Expressed Shame by Diversion Group

Assignment ................................................................................. 33
Table 7 Cross-Tabulation of Diversion Group Assignment and Re-Offending ............ 35
Table 8 Logistic Regression Model Predicting Odds of Re-Offending ...................... 36

vi

LIST OF FIGURES

Figure 1 Timing of Research Questions ......................................................... 24

vii

INTRODUCTION

Over the last two hundred years, the United States’ justice system has
principally operated under a retribution model, in which punishments reflect the
mentality of “an eye for an eye.” Justice is viewed in terms of blame and punishment
(Zehr, 1990). The way in which the United States punishes criminals is by removing
them from society (and placing them in prison) for a designated period of time,
which, in theory, should correspond to the severity of the crime committed.
Offenders, upon completion of their sentence, are viewed as having “paid their debt
to society” and are hoped to have “learned their lesson” so as to not commit a
criminal act in the future. However, the retribution model is a zerocsum game with
the offender always being the loser (Wright, 2000). As a result, it has been found that
punishment does not necessarily result in the desired behavior of conformity to
society’s laws, but instead may increase the likelihood of criminal offending
(Gendreau, 1996). This belief in retribution can be seen by the increased demand for
incarceration in the United States. The incarceration rate in the United States has
more than tripled since 1980, and in the year 2002 the prison population grew by 2.6
percent (Bureau of Justice Statistics, 2000; Harrison & Beck, 2003). In addition,
between 1987 and 1996, court-ordered residential placements for juveniles increased
by fifty-one percent.

This increase in residential placements for juveniles has corresponded with a
forty-nine percent increase in the volume of cases handled by the court system in the
United States (Snyder & Sickmund, 1999). The type of cases handled by the juvenile

court system has changed as well, with an increase in cases dealing with violent,

person, weapon, and drug offenses (Snyder & Sickmund, 1999). These increases
have placed a strain on the court system’s resources and programs and a growing
consensus that the current juvenile justice system is not working effectively.

In light of this reality, practitioners, scholars, and the public have begun to
search for alternatives to the current justice system (Perry, 2002). While some have
suggested a more punitive approach, another alternative that has gained ground in
recent years is the restorative justice model. According to Johnstone (2002), a
restorative justice model aims to “meet victim’s needs” and “ensure that the offender
is fully aware of the damage they have caused to people and of their ability to repair
that damage” instead of focusing only on the punishment of the offender as with the
, retributive model (p. 1).

In the United States, the restorative justice model has primarily been touted as
a successful intervention for juvenile offenders. Research shows that youth who are
known by the juvenile justice system by an early age are more likely to have long and
violent criminal careers (Loeber & Fanington, 2000; Snyder & Sickmund, 1999). In
addition, it has been calculated that the cost to society for these criminal careers is
between 1.7 and 2.3 million dollars (Cohen, 1998). An effective intervention for
young offenders is needed to reduce these costs to society. It is hoped that a
restorative justice framework will be that effective intervention. In fact, recently
many state juvenile justice agencies have adopted restorative justice ideals.
According to Snyder and Sickmund (1999), “recently, states have been attempting to
strike a balance in their juvenile justice systems among system and offender

accountability, offender competency development, and community protection” (p.

89). AS a result, many states have begun to use restorative justice language in their
mission and vision statements. For example, Wisconsin’s vision statement states,
“the Division of Juvenile Corrections will reduce delinquent behavior and restore a
sense of safety to victims and the community” (Wisconsin Division of Juvenile
Corrections, 2003). However, while juvenile justice agencies have begun to embrace

the concept of restorative justice, it is important to know what that concept entails.

CHAPTER I
DEFINITION OF RESTORATIVE JUSTICE
What is Restorative Justice?

The term “restorative justice” encompasses many different approaches and/or
interventions, and various scholars have given it different meanings (Van Ness &
Strong, 2002). Zehr (1990) states “crime is a violation of people and relationships. It
creates obligations to make things right. Justice involves the victim, the offender, and
the community in a search for solutions which promote repair, reconciliation, and
reassurance” (p. 181). This is in direct contrast to the retributive model, which
emphasizes crime as lawbreaking only. It is this inclusion of the victim and offender
that makes it different than how crime has been traditionally handled by the criminal
justice system.

Van Ness and Strong (2002) build upon Zehr’s description of restorative
justice by stating,

Restorative justice focuses on repairing the harm caused by crime

and reducing the likelihood of future harm. It does this by

encouraging offenders to take responsibility for their actions and for

the harm they have caused, by providing redress for victims, and by

promoting reintegration of both within the community.

Communities and the government accomplish this through a

cooperative effort (p. 49).
These definitions, however, do not pinpoint how to identify restorative justice
practices. Zehr and Mika (2003) suggest that restorative justice processes can be
identified when the following occurs: the harms of the offenses are focused upon,

there is involvement of victims, offenders, and the community; restoration of victims

is a priority; offenders are supported and encouraged to follow through on obligations

for reparations; offender reparation obligations are achievable and not unduly
hardships; opportunities for dialogue between the victim and the offender are
available when appropriate; collaboration and reintegration are encouraged instead of
coercion and isolation; and respect is given to all parties involved in the process.
These guidelines for identifying restorative justice processes show that the concept of
restorative justice can incorporate many different practices; all of which- could be

described as falling under the restorative justice model.

Restorative Justice Models

Restorative justice practices fall generally under one of four different models:
community reparative boards, circle sentencing, victim-offender mediation, and
family group (or restorative justice conferencing). While each of these models views
crime as “a harm to people and relationships” and aims to involve the offenders,
victims, and community in reconcile and repair that harm, they each have slightly
different approaches and procedures to dealing with criminal offenses (Zehr, 1990).
Each of the models is described below.

Community reparative boards have also been called youth panels or
neighborhood boards, and have been in use since the 19908 (Bazemore & Umbreit,
2001). Reparative boards consist of a group of trained community members, which
conduct face-to-face meetings with offenders. In these meetings, board members
discuss the crime and the harm it caused. The offender and the board agree on what
actions need to be taken within a specific time period in order for the offender to

make reparation for the crime. The board makes a report to the court regarding the

offender’s compliance with the agreement after the stipulated time limit has passed, at
which point the board’s involvement in the case ends (Bazemore & Umbreit, 2001).

Circle sentencing (also known as peacemaking circles) is an updated version
of the traditional sanctioning processes of aboriginal societies (Melton, 1995). The
sentencing circle consists of the offender, victims, families, community members, and
justice system members. Offenders must apply to participate in the circle process.
Circle sentencing usually consists of a healing circle for the victim, a healing circle
for the offender, a sentencing circle to develop a sentencing plan, and follow-up
circles to monitor the progress of the offender (Bazemore & Umbreit, 2001). Within
the circles, participants discuss the crime and convey the harm of the offense as well
as identify how to heal all the affected parties and prevent future criminal acts
(Bazemore & Umbreit, 2001). The process may vary from community to community,
however, as each sentencing circle is designed to fit local needs.

Victim-offender mediation has been in operation since the 19703 in the United
States, Canada, and Europe (Bazemore & Umbreit, 2001). In victim-offender
mediation, the victim and offender sit down with a trained mediator. This allows the
victim the opportunity to communicate to the offender the crime’s impact on his or
her life, ask questions regarding the crime and the offender, and be involved in the
development of a reparation plan for the crime (Bazemore & Umbreit, 2001). The
exact process of victim-offender mediation depends largely on where it is conducted,
as each program is tailored to the needs of the jurisdiction in which it resides

(Umbreit, Greenwood, Fercello, & Umbreit, 2000).

Restorative justice conferencing (also known as family group conferencing)
began in the late 19803 in New Zealand. It allows youthful offenders to sit down with
the victim, the victim’s supporters, and the youth’s supporters with a trained
facilitator to discuss the crime and the resulting harm that was caused. An agreement
is reached at the end of the conference, which states how the youth can make amends
to the victim (Bazemore & Umbreit, 2001). Most participants in restorative justice
conferencing report high levels of satisfaction with the process and outcomes (Daly,
2001; Hayes & Daly, 2003; Latimer, Dowden, & Muise, 2001; Marshall, 1999;
Miers, Maguire, Goldie, Sharpe, Hale, Netten, Uglow, Doolin, Hallam, Enterkin, &
Newburn, 2001; McGarrell, 2001; McGarrell, Olivares, Crawford, & Kroovand,
2000; Umbreit & Fercello, 1998). In addition, several researchers have demonstrated
that the use of restorative justice conferencing leads to reductions in re-offending
compared to other traditional, and in some cases other diversion program,
interventions (Braithwaite, 1999; Hines, 2000; Luke & Lind, 2002; Latimer, Dowden,
& Muise, 2001; McGarrell, 2001; McGarrell, Olivares, Crawford, & Kroovand,
2000). On the other hand, some have found a difference in re-offending only for those
charged with serious (person) offenses, but not in less serious offenses, such as
property offenses (McCold & Wachtel, 2001; Sherman, Strang, & Woods, 2000).
Other studies have examined differences in re-offending among restorative justice
conference participants and have found conference factors (such as an offender
feeling remorseful and agreement with the outcome) to reduce re-offending (Hayes &
Daly, 2003; Maxwell & Morris, 2001). Restorative justice conferencing is the

model that will be examined in this research.

CHAPTER II
THEORETICAL FRAMEWORK

Reintegrative Shaming

Reintegrative shaming has been the theoretical framework largely associated
with restorative justice conferencing (Bazemore & Dooley, 2001). Braithwaite’s
(1989) theory of reintegrative shaming incorporates and builds upon control and
deterrence theory. Reintegrative shaming builds upon deterrence theory in that it
increases the costs of offending to the point that offending becomes unthinkable, and
builds upon control theory by utilizing social bonding as a mechanism to facilitate
reintegrative shaming (Braithwaite, 1989).

Braithwaite (1989) makes a clear distinction between Shaming processes that
are reintegrative (like restorative justice conferencing) and those that are
disintegrative or stigmatizing (like our traditional justice system). Braithwaite states
“reintegrative shaming means the expressions of community disapproval. . .are
followed by gestures of reacceptance into the community of law-abiding citizens.
...Disintegrative shaming (stigmatization), in contrast, divides the community by
creating a class of outcasts” (1989, p. 55). Thus, reintegrative shaming uses fear of
social disapproval and an attack on the conscience as reasons why individuals will
fail to commit unlawful acts. Reintegrative shaming prevents an individual from
developing a deviant career when the Shaming maintains the bonds of love or respect
throughout the shaming process, is directed at wrongfulness of the act as opposed to
the wrongfulness of the person, is expressed in terms of conveying general social

approval, and involves gestures or ceremonies of approval (Braithwaite, 1989).

Braithwaite (1989) recommends restorative justice conferencing as a ceremony,
which invokes reintegrative shaming to prevent re-offending. Stigmatizing shaming,
on the other hand, cannot adequately prevent an individual from adopting a deviant
master status.

Stigmatizing Shaming labels offenders as deviant and as a reaction to the
label, criminal subcultures form (Braithwaite, 1989, Cloward & Ohlin, 1960).
Braithwaite (1989) states that subcultural groups cluster sets of rationalizations and
conduct norms together to support criminal behavior. These subcultural groups
“provide systematic social support for crime in any of a number of ways - supplying
members with criminal opportunities, criminal values, attitudes which weaken
conventional values of law abidingness, or techniques of neutralizing conventional
values” (Braithwaite, 1989, p. 101).

There have been few empirical studies regarding Braithwaite’s reintegrative
shaming theory, despite its relative popularity (Hay, 1998; Hay, 2001; Houts, 1995;
Makkai & Braithwaite, 1994; Vagg, 1998; Wong, 1999; Zhang, 1995). In fact, many
have applied reintegrative shaming as the theoretical framework for restorative justice
conferencing without testing the theory empirically (Bazemore & Dooley, 2001;
Braithwaite & Mugford, 1994; McCold & Wachtel, 1998; Sherman, Strang, Barnes,
Braithwaite,1pken, & The, 1998). In addition, findings from most empirical studies
on reintegrative shaming theory have been inconclusive (Hay, 2001). Hay (1998)
argues that the lack of empirical tests of Braithwaite’s reintegrative shaming theory is

due to “the lack of theoretical clarity in Braithwaite’s statement of the theory” (p.

421). However, there are two clear concepts, shame and interdependency, which will

be examined.

§h_am_e

Shaming is an integral part of Braithwaite’s (1989) reintegrative shaming
theory. As stated previously, Braithwaite (1989) differentiates shaming into shaming
that is reintegrative and shaming that is stigmatizing. Reintegrative shaming is
conducive to reducing re-offending, while stigmatizing shaming is not (Braithwaite,
1989). In fact, Braithwaite (1989) argues shaming is the process that makes criminal
behavior “abhorrent” to individuals through invoking the punishers of “social
disapproval and pangs of conscience” (p. 7, 75). Other researchers have given this
notion credence, arguing informal consequences of being caught engaging in criminal
behavior (such as shaming) may be as important or more important than formal
consequences (Grasmick & Bursik, 1990; Paternoster & Iovanni, 1986). In addition,
Braithwaite (1989) argues individuals who do not feel shame are more likely to
engage in criminal behavior because they do not have a fear of shame for themselves
or in regards to intimates.

These assertions and findings lead to two issues regarding shame and
reintegrative shaming, whether shame needs to be acknowledged and whether
shaming is more effective when an individual expresses Shame that intimates found
out about the criminal behavior. Researchers have delineated between
unacknowledged and acknowledged shame. In acknowledged shame, an individual

accepts feelings of shame and believes he or she has acted in a socially undesirable

10

way, which allows an individual to express culpability for the behavior and make
reparations for harm caused (Ahmed, Harris, Braithwaite, & Braithwaite, 2001;
Lewis, 1971; Wicker, Payne, & Morgan, 1983). In unacknowledged shame, the
opposite occurs and an individual is more likely to resist expressing culpability and
making amends (Ahmed, Harris, Braithwaite, & Braithwaite, 2001). It is also
unacknowledged shame that leads to crime (Katz, 1988; Lansky, 1995). There is
some evidence that offenders who expressed shame were less likely to re-offend than
those who did not (Jensen & Gibbons, 2002). In fact, Jenson and Gibbons (2002)
found offenders are more likely to accept culpability for the offense when they
expressed shame. Therefore, expressed Shame is similar to acknowledged shame.
Reintegrative Shaming facilitates acknowledged Shame through exposing
criminal behavior to one’s intimates (Braithwaite, 1989; Grasrnick & Bursik, 1990,
Jenson & Gibbons, 2002; Williams & Hawkins, 1986). Thus, restorative justice
interventions that are based on reintegrative shaming principles are believed to
increase levels of acknowledged shame, and as a result, reduce further criminal
behavior (Ahmed, Harris, Braithwaite, & Braithwaite, 2002). Therefore,
acknowledged shame is assumed to occur in interventions that invoke reintegrative

shaming. This study will examine this assumption.

Interdependency
Another concept that is clear in Braithwaite’s (1989) theory is that of
interdependency. Individuals that “have personal obligations to others within a

community” are said to be interdependent (Braithwaite, 1989, p. 85). Individuals are

11

more susceptible to shaming when they are interdependent with those who are
shaming them (Braithwaite, 1989). In addition, Braithwaite (1989) argues
interdependency is linked with decreased criminal activity (although perhaps
indirectly).

Braithwaite (1989) outlines several predictors of interdependency. He
identifies males, those aged between 15 and 25, the unmarried, and unemployed as
being less likely to have interdependent relationships (Braithwaite, 1989). In
addition, Braithwaite (1989) makes a direct connection with interdependency and
Hirschi’s (1969) social control theory. Braithwaite (1989) argues that the concept of
“interdependency is approximately equivalent to the social bonding, attachment and
commitment of control theory” (p. 100). Thus, interdependency and/or control theory
acts as the mechanism through which reintegrative Shaming operates.

Makkai and Braithwaite (1994) and Hay (2001) examined the relationship of
interdependency and reintegrative shaming. Makkai and Braithwaite (1994)
examined nursing home compliance with the law. Interdependency was measured by
whether nursing directOrs reported they knew some of the inspectors prior to their
regulatory encounter (Makkai & Braithwaite, 1994). Results showed no reintegrative
effect when directors knew none of the inspectors, that is, where there was low
interdependency (Makkai & Braithwaite, 1994).

Hay (2001) also examined interdependency. Hay (2001) examined juvenile
perceptions of reintegration and shaming used by their parents. Interdependency was
only measured through attachment to parents. Hay (2001) found a significant

relationship between interdependency and shaming, but it was not as strong as the

12

relationship found between interdependency and parent use of reintegration. Hay
(2001) argued, “reintegrative sanctioning of children [may] lead to high parent-child
interdependency rather than vice versa” (p. 147). Therefore, interventions that use
reintegrative shaming should lead to higher interdependency.

While these two studies provide some insight into interdependency, there are
still many issues left to be addressed. This study aims to address some of these
issues, specifically broadening the definition of interdependency, examining the
relationship between interdependency and re-offending, and assessing whether

interventions based on reintegrative shaming leads to higher interdependency.

Social Control Theory

An important part of Braithwaite’s (1989) concept of interdependency is its
link with Hirschi’s (1969) social control theory. Hirschi ( 1969) argues that
individuals conform to societal rules because they are bonded to society. Hirschi
(1969) argues that there are four elements to social bonding: attachment,
commitment, involvement, and belief.

Attachment is sensitivity to others and is the element that allows the
internalization of norms and values. Hirschi (1969) argues that attachment to parents
and school strengthens an individual’s bond to society. Hirschi (1969) specifically
found (through a self-report study) that boys who were more closely attached to their
parents were less likely to commit delinquent acts than those who were less closely
attached. Youth also are more likely to commit delinquent acts when they have low

attachment to school, exhibiting characteristics such as poor school performance, a

13‘

dislike for school, and a rejection of a school’s authority (Hirschi, 1969). Hirschi

( 1969) argues that attachment to peers can be problematic when discussing
strengthening bonds to society. He argues that youth, who have a stake in
conformity, already have strong bonds to society, will not align themselves with
delinquent peers (Hirschi, 1969). However, youth with weak bonds already are more
likely to align themselves with delinquent peers and to commit delinquent acts
(Hirschi, 1969).

Commitment is the investment an individual has in conformity and the fear of
the consequences of a delinquent act. Hirschi (1969) states that commitment is the
“rational component in conformity,” in which an individual will weigh the costs and
benefits before committing a delinquent act (p. 20). Thus, an individual with a greater
“stake in conformity” will be less likely to commit a delinquent act due to the
negative consequences of such an act (Toby, 1959). Hirschi (1969) specifically
examined commitment to education and occupation. Hirschi (1969) found that
educational and occupational aspirations and expectations of delinquents were lower
than non-delinquents. In addition, Hirschi (1969) states, “those boys whose stated
aspirations exceed their expectations are less likely to be delinquent than boys with
equivalent expectations and lower aspirations” (p. 185).

Involvement in conventional activities is based on the belief that opportunities
for delinquent activities is restricted by “being busy”. Hirschi (1969), however,
admits that research has failed to produce evidence that involvement in conventional

activities curbs delinquency. He further goes on to state, “in the common sense, the

14

idea of involvement remains central to much thinking about the causation and
prevention of delinquency” (Hirschi, 1969, p. 188).

The element of belief covers the concept that individuals will be less
likely to commit delinquent acts if they believe they should obey the rules or
laws of society. Hirschi (1969) clarifies this idea by stating; “delinquency is
not caused by beliefs that require delinquency but is rather made possible by
the absence of (effective) beliefs that forbid delinquency” (p. 198). Hirschi
(1969) specifically looks at values relating to defining delinquent behavior
and acts as “good” or “bad”.

There have been a large number of empirical studies on Hirschi’s
(1969) theory of social control. These studies have largely concluded that
there is support for two of the four elements of social bonding: attachment
and commitment (Costello & Vowell, 1999; Krohn & Massey, 1980). On
the other hand, studies examining involvement in conventional activities
have been less than clear and others have found the variable of conventional
beliefs to be problematic (Kempf, 1993; Vold, Bernard, & Snipes, 2002).
Some studies have, however, found support for involvement through
involvement in extracurricular school activities (Agnew, 1985; Wiatrowski,
Griswold, & Robbins, 1981). Others have used an adapted form of Hirschi’s
(1969) social control theory, in which attachment to and commitment and
involvement in conventional activities restricts the propensity to be
delinquent (Nagin & Paternoster, 1994; Sampson & Laub, 1991, 1992,

1993). Thus, according to this adaptation, attachment and commitment to

15

and involvement in conventional activities and/or institutions (for example,
school, extracurricular activities, and church) reduce the likelihood of
delinquency. This study will utilize this adaptation of Hirschi’s (1969)
social control theory, which examines attachment, commitment to

conventional institutions, and involvement in conventional activities.

16

CHAPTER III
DATA AND METHODS

Current Studv

Braithwaite (1989) specifically draws upon Hirschi’s (1969) social
control theory to define his concept of interdependency. This study will use
the adapted form of Hirschi’s (1969) social control theory put forth by
Paternoster and Nagin (1994) and Sampson and Laub (1991, 1992, 1993) to
evaluate attachment, involvement in conventional activities, and
commitment to conventional activities as measures of interdependency. In
addition, this study will examine the role of expressed shame in reintegrative
shaming. This study will test Braithwaite’s (1989) reintegrative shaming
theory as an appropriate theoretical framework; specifically it’s concept of
interdependency and assumption of expressed shame, for the intervention of

restorative justice conferencing.

The Indianapolis Restorative Justice Exmriment

This study will utilize data obtained from the Indianapolis Restorative Justice
Experiment, which began in 1997 by a team of researchers from Indiana University
and the Hudson Institute (see McGarrell, 2001; McGarrell, Olivares, Crawford, &
Kroovand, 2000). The Indianapolis Restorative Justice Experiment involved the
implementation of restorative justice conferencing as an alternative for very young
offenders. The initial study involved an experimental design. Youth were randomly

assigned to either restorative justice conferencing or to one of the other diversion

17

programs in Marion County, Indiana. To be eligible for assignment to a restorative
justice conference, youth had to meet specific criteria. The criteria for admission
included:

0 Youth was a first time offender.

0 Youth was not charged with a serious or violent offense.

0 Youth had no other charges pending.

0 Youth was no older than 14 years of age.

0 Youth admitted responsibility.

Seven hundred and eighty-two youthful offenders participated in the experiment
from September 1, 1997 through September 30, 1999. Four hundred youth were
assigned to the Restorative Justice Conferencing group, while 382 were assigned to
the control group. In the experiment, 61% (233) control youth successfully
completed the assigned diversion program, while 80.5% (322) of the experimental
youth successfully completed the assigned restorative justice conferencing. The
difference is statistically significant (x2 = 36.08, p < .01). The experiment utilized
several outcome measures including: post-conference/diversion surveys with
offending youths, victims, victim supporter(s), and offender’s parents or supporters; a
one year follow-up interview with offending youth; a one year follow-up interview
with the victim; and an initial six month recidivism analysis. The researchers report
plans to follow the sample for a 24-month follow-up period (McGarrell, 2001;
McGarrell, et. al, 2000). Survey questions were answered using a binary scale or
scale for strength (usually a Likert-scale). Recidivism was measured by whether a

youth was re-arrested (McGarrell, 2001; McGarrell, et. al,'2000).

18

Sample Description

This study utilizes data from a sub-sample of the original data from the
Indianapolis Restorative Justice Experiment: youthful offenders that completed the
post-conference survey. Post-conference and diversion program interviews were
attempted with all youths who entered the study on the basis of an original offense
that occurred between August 1998 and June 2000. A total of 458 youths entered the
study during that time period. These included 233 youths in the experimental group
(restorative justice conference referral) and 225 youths in the control group (diversion
program referral). Of these youths, 269 interviews were successfully completed.

This included 156 completed surveys of experimental group youths (67% response
rate) and 113 completed surveys of control group youths (50% response rate).1 The
control group response rate appears to have been reduced by the difficulty in reaching
control group youths who failed to complete their diversion program. This introduces
a potential source of bias in comparing the two groups, as the control group appears
to under-represent youths who had difficulty with compliance with court-ordered
programs.

Three youth were eliminated from the sub-sample due to discrepancies when
recidivism data were matched with youth interview data. Therefore, the entire sub-
sample used in this research included 266 youths (153 experimental and 113 control).
Eight additional youth were excluded from this study for failure to complete the
assigned diversion program or restorative conferencing program, 4.4% of the control

group and 2.0% of the restorative justice group (X2 = 1.33, p > .05). This rate of

 

' UnfOrtunately, the number of attempts to reach youth were not recorded neither were the reasons for
not reaching particular youth. Therefore, the implications cannot be fully discussed.

19

completion is much lower than that of the original sample, in which the differences
between the two groups were significantly different. Of the resulting 257 subjects,
149 (58%) were in the Restorative Justice conferencing (experimental) group. The
control group consisted of the remaining 108 youths, which were assigned to
diversion programs other than the Restorative Justice Conferencing. Breakdown of

youths into diversion programs for the sub-sample is illustrated in Table l.

20

Table 1. Diversion Program Breakdown for Survey Youth

 

 

Diversion Program N Percent
Exmrimental Group:
Restorative Justice Conferencing 149 58.0%
Control Group:
Shoplifting Education 41 16.0%
Garden Project (community service) 13 5.1%
Victim Offender Mediation 2 .8%
IPD Drug Education 20 7.8%
Volunteer Services 3 1.2%
Paint It Clean (community service) 16 6.2%
Tutoring Program 3 1.2%
Essay 3 1.2%
National Correctional Training Institute 1 j .4%
Operation Kids Can 5 1.9%
Boys Vision 1 .4%
Total 257 100.0%

 

The mean age for the sample was 12.69 years. There were more males

' (61.1%) than females (38.9%), and slightly more whites than non-whites (50.2% and
49.8% respectively). In addition, a majority of the youths were charged with either
shoplifting or battery (37.9% and 35.2% respectively). Breakdown of the sub-sample

by charge is illustrated in Table 2.

21

Table 2. Charge Breakdown‘

 

Charge RJ Control Total
Conversion (Shoplifting) 39.2% 36.1% 37.9%
Battery 37.1% 32.4% 35.2%
Theft 9.5% 7.4% 8.6%
Criminal Mischief 6.7% 8.3% 7.4%
Disorderly Conduct 2.7% 5.6% 3.9%
Trespass 2.0% 4.6% 3.1%
Public Intoxication 0.0% 2.9% 1.2%
Other 1.4% .9% 1.2%
Intimidation 0.0% .9% .4%
Resisting Arrest .7% 0.0% .4%
Residential Entry .7% 0.0% .4%
Possession of Marijuana 0.0% .9% .4%
Total 100% 100% 100%
N = 148 N=108 N = 256

 

' Differences not significant (.r2 = 13.09, p > 0.5).

Research Ouestiorg

The study proposes several research questions to test the applicability of
Braithwaite’s (1989) concept of interdependency, conceptualized by Hirschi’s (1969)

control theory adaptation and Braithwaite’s (1989) assumption of expressed shame.

22

Braithwaite (1989) indicates that reintegrative shaming is more conducive for those
who are interdependent and for those who acknowledge or express shame during an
intervention that follows reintegrative shaming principles. Therefore, one of the
research questions for this Study is whether there is a difference between the
restorative justice conferencing group and the control group regarding
interdependency and expressed shame after the intervention. Braithwaite (1989) also
states that interventions based on reintegrative shaming will reduce re—offending,
while those interventions utilizing stigmatizing Shaming (or traditional interventions)
will not. Therefore, another research question for this study is whether there is a
difference in re-offending between the restorative justice conferencing group and the
control group. Braithwaite (1989) asserts that individuals who are interdependent and
express shame are less likely to re-offend. Thus, the final research question is
whether there is a relationship between re-offending at 24 months and reported levels
of interdependency and expressed shame (as recorded in post-conference surveys).
Figure 1 shows the timing of when the survey and re-offending occurred, as well as
when each research question was addressed.

These research questions are addressed in several ways. The first question is
addressed through bivariate analysis, specifically independent t-tests, for each of the
measures of interdependency and that of expressed shame. The relationship between
re-offending and diversion group assignment was examined through the chi-square
statistic. Binary logistic regression was employed to answer the third research

question.

23

Figure 1. Timing of Research Questions.

T1 T2 T3
Intervention Survey Re-Offend
Ql RJ/Control Interdep. &
Shaming
02 RJ/Control Interdep. & Re-offend
Shaming
Q3 RJ/Control g Re-offend

 

Variables

The concepts of interdependency and expressed shame were the independent
variables for multivariate analysis and were the dependent variables for bivariate
analysis. The concept of interdependency was examined through three measures:
attachment to mother (attachment), commitment to conventional institutions of
church and school (commitment), and involvement in conventional activities
(involvement). There was only one measure of expressed shame. Diversion group
assignment was used as the independent variable for bivariate analysis and the
multivariate analysis. Re-offending was used as the dependent variable for

multivariate analysis. Race, age, and gender were controls for multivariate analysis.

24

Measures of Interdep_endency and Expressed Shame

As noted previously, Braithwaite (1989) and Hirschi (1969) would expect
experimental youths (restorative justice conference referral) to have higher
interdependency and expressed shame. Therefore, two predictor measures were used
in this study: interdependency and expressed shame. Measures of the concept of
interdependency will be discussed first.

As a first step, I used factor analysis to explore and develop two measures of
interdependency. Those two measures included attachment to mother and
commitment to conventional institutions of church and school. Principal component
analysis (PCA), through factor analysis, using oblique rotation was used as a data
reduction technique to gain these two measures. The principal component analysis
applied the standard criteria of requiring an eigenvalue of one as an indication of
which components’ factors scores should be loaded. PCA results in z scores (called
factor scores) for each case. A factor score greater or less than +/- 1.96 is considered
an extreme score. Table 3 shows the factor loadings for each of the variables that

resulted from the PCA.

25

Table 3. Components Loading for PCA Extraction

 

 

 

Component Loading
Variable Factor 1 Factor 2
Grades important .871 .1 18
Like school .789 .296
Attend religious services .789
Close to mother .918
Gotten along well with mother .327 .920

 

The measure attachment to mother is the combination of two survey questions
using a likert-scale response: “I feel close to my mother,” and “Throughout my life, I
have gotten along well with my mother” (1 = strongly agree, 5 = strongly disagree).
It is important to note that the factor loadings for the two measures were high (greater
than .9), which indicates that both survey questions are measuring the same thing
(Johnson & Wichern, 2002). Therefore, only one survey question, “I feel close to my
mother”, was used as the measure for attachment to mother. The mean answer for this
survey question was 1.97, indicating that most of the sample agreed that they felt
close to their mother. Table 4 shows the descriptive Statistics for this measure.

It is also important to note here that attachment to father was not used as a
measure. Observation data from the restorative justice conferences showed that of
the 215 conferences observed, 194 (90.2%) had at least one parent attending.

Unfortunately, the data did not distinguish between .whether it was the mother or the

26

father attending. Previous research has shown that attachment to mother or female
caregivers is more indicative of reductions in re-offending (Allen, Moore,
Kuperminc, & Bell, 1998; Benda & Corwyn, 2002; Simons, Patemite, & Shore,
2001). In addition, the U. S. Census Bureau (2002) reports that in the 2000 census,
62.8% of households in Marion County, Indiana are single-female headed
householders with children under the age of 18. Thus, only attachment to mother was
used.

The measure commitment to conventional institutions of church and school is
a combination of three survey questions using a likert-scale response: “In general, do
you like or dislike school” (1 = like it, 3 = dislike it); “How important is getting good
grades to you personally” (1 = very important, 3 = not important); and “How often do
you attend religious services” (1 = several times a month or more, 3 = hardly ever or
never). By viewing the correlation matrix, it was determined that factor scores for
commitment to conventional institutions of church and school, that are below zero
correspond to lower levels of commitment, while scores greater than zero correspond
to higher levels of commitment.2 Commitment to conventional institutions had a
minimum factor score of -2.57 and a maximum factor score of 1.26. These measures
were determined to be reliable (cronbach’s alpha = .71).

Additive scales were used to create indicators for the third measure of
interdependency, as well as the measure for expressed shame from the survey
questions. Factor analysis could not be used as a data reduction technique in this

instance, since the survey questions had binary answers (yes or no), and as a result,

 

2 Values less than zero originally corresponded to higher levels of commitment and vice versus for
values greater than zero. Values were multiplied by a negative one to reverse this in order for easier
readability of results.

27

reliability between the survey questions are quite low. The third measure of
interdependency will be discussed first, followed by the measure of expressed shame.
The third measure of interdependency is involvement in conventional
activities. An additive scale was used to tally the responses (1 = yes, 0 = no) to
several survey questions that indicate involvement (“Are you active in athletic teams
at school;” “Are you active in cheerleading at school;” “Are you active in science
clubs at school;” “Are you active in service clubs at school;” “Are you active in band
or musical groups at school;” “Are you active in art or dance at school;” “Are you
active in student government at school;” “Are you active in honor societies at
school;” “Are you active in athletic teams outside of school;” “Are you active in

’9 ‘6

cheerleading outside of school;” “Are you active in a part-time job; Are you active
in band or musical groups outside of school;” “Are you active in art or dance outside
of school;” “Are you active in a youth group; are you active in any other activities
outside of school;” and “Do you belong to a church”). Youths reported an average
involvement in 1.23 activities (ranging from zero to six). These measures had a
cronbach’s alpha of .36.

The measure of expressed shame is an additive scale of three survey questions
(“How ashamed or proud did you feel that your family found out about the offense
you committed,” “How ashamed or proud did you feel that your friends found out
about the offense you committed,” and “How ashamed or proud did you feel when
the police found out about the offense you committed”). Responses to these

questions were on a likert-scale (1 = very ashamed, 6 = proud). These responses were

recoded as a dummy variable (0 = did not express shame, I = expressed shame).

28

These questions were dichotomized in order to simplify the difference between
expressed shame and not expressing shame. An additive scale was then used to tally
up the responses for each of these survey questions to become the measure for
expressed shame. Youths reported expressing shame from none of the three
dimensions (family, peers, and police) to all three dimensions. These measures had a
cronbach’s alpha of .54. Table 4 shows the descriptive statistics for the two of the

measures of interdependency (attachment and involvement) and for expressed shame.

Table 4. Measures of Interdependency and Expressed Shame

 

 

Measure N Mean SD Minimum Maximum
Interdependency
Attachment 254 1.97 .66 1.00 5.00
Involvement 255 1.23 1.21 .00 6.00
Expressed Shame 255 2.67 .67 .00 3.00

 

Deppndent Variable

The dependent variable used in the multivariate analysis is re-offending. Re-
offending is a dummy variable that is measured by whether the youth had been re-
arrested in the preceding 24 months since the initial court appearance (0 = no, 1 =
yes). Of the entire sub-sample of youths (257), 103 re-offended (40.1%) and 154

(59.9%) did not.

29

Control Variables

Braithwaite (1989) argues that males and those aged between 15 and 25 are
less likely to have interdependent relationships, and as such, are more likely to re-
offend. Other research has also identified minority Status as a risk factor for
criminality (Loeber & Farrington, 1998). Therefore, multivariate analysis used the
dummy variables race (white = 1) and gender (male = 1), and the continuous variable
of age (range from 7 to 16) as controls. The experimental group and control group
did not vary significantly on race, gender, or age (x2 = 1.13, p > .05; x2 = .614; p >
.05; x2 = 13.04, p > .05, respectively). Table 5 Shows the breakdown in size, percent,

and chi-square for each control variable.

30

Table 5. Descriptive Statistics for Control Variables.

 

 

 

RJ Control
Variable N % N % Total N
White 79 61.2% 50 38.8% 129
Male 88 56.1% 69 43.9% 157
Age 257
7 0 0.0% 1 0.9%
8 4 2.7% 1 0.9%
9 6 4.0% 0 0.0%
10 4 2.7% 4 3.7%
11 16 10.7% 11 10.2%
12 28 18.8% 17 15.7%
13 46 30.9% 31 28.7%
14 43 28.9% 36 33.3%
15 2 1.3% 5 4.6%
16 0 0.0% 2 1.9%

 

' Differences not significant (x2 = 13.03, p > .05; x2 = 1.13, p> .05; and x? = .61, p >
.05 respectively).

31

CHAPTER IV

RESULTS

Bivariate Analysis: Interdeppndency and Expressed Sham;

Independent T-tests were employed to examine whether there were
differences between youths assigned to restorative justice conferencing and youths
assigned to other diversion programs in the measures of interdependency and
expressed shame. Restorative justice conferencing youths were significantly different
from control youths in two of the three measures of interdependency (attachment to
mother and involvement in conventional activities), as well as in expressed shame.
Table 6 shows the sample Size and mean of restorative justice conferencing youths
and control youths for each measure of interdependency and expressed shame, as well
as the t statistic.

Results were consistent with Braithwaite’s (1989) reintegrative shaming
theory, which argues that youth participation in a reintegrative shaming intervention
will increase interdependency and will increase expressions of shame. Results,
however, were not completely consistent with the adopted form of Hirschi’s (1969)
social control theory. Only two of the three measures used as indicators for
interdependency had significant differences between the experimental and control
groups: attachment to mother and involvement in conventional activities. The control
and experimental group did not vary significantly on commitment to conventional
institutions of church and school. Results show youths assigned to restorative justice

conferencing were involved in more conventional activities (t = 3.44, p <.05), were

32

more attached to their mother (t = -5.28, p < .05), and were more likely to express

shame (t = 7.39, p < .05) than youths assigned to the control (other diversion

programs).

Table 6. Differences in Interdependency and Expressed Shame by Diversion Group

Assignment

Measure N M t

 

Interdeppndency

 

Attachment 7. 13
R] 147 1.74
Control 107 2.29

Commitment -1. 13
R] 147 -.08
Control 107 .06

Involvement -4.76‘
R] 147 1 .52
Control 108 .83

Expressed Shame -7.39'

R] 149 2.92
Control 106 2.30

 

° Results significant one-tail test, p < .05

33

Bivaflte Analysis: Re-Offending

Bivariate analysis using the chi-square statistic was employed to examine
whether diversion group assignment affected re-offending. The initial six-month
recidivism analysis of the entire sample in the Indianapolis Restorative Justice
Experiment found that the experimental youths (restorative justice conference
referral) re~offended at a significantly lower rate than control group youths, a
reduction of 40% (McGarrell, 2001; McGarrell, Olivares, Crawford, & Kroovand,
2000). By the 12-month recidivism analysis, there was a decrease to a reduction of
25%, which was a significant decrease (McGarrell, 2001; McGarrell, Olivares,
Crawford, & Kroovand, 2000). This Study of the sub-sample examined recidivism at
24-months and did not find a significant difference in re-offending between the
control group (other diversion program assignment) and assignment to the restorative
justice conferencing group (it2 = .918, p > .05). Therefore, diversion group
assignment does not affect re-offending at 24 months, at least for the sub-sample of
youths included in the survey. This finding also does not support Braithwaite’s
(1989) reintegrative shaming theory. Braithwaite (1989) argued participants of an
intervention utilizing reintegrative shaming principlesvwould have decreased rates of
re-offending. While restorative justice conference youths re-offended at a lower rate
than control group youth, the differences were not significant. Table 7 shows the

cross-tabulation table of diversion program assignment and recidivism.

34

Table 7. Cross-Tabulation of Diversion Group Assignment and Re-offending’

 

 

Control RJ Total
Re-offend 43.5% 37.6% 40.1%
Did not re-offend 56.5% 62.4% 59.9%
Total 100% 100% 100%
N=108 N=149 N=257

 

'Results not significant (1:2 = .918. p > .05).

Multivariate Analysis

Binary logistic regression was employed to predict the odds that a youth
would recidivate. The predictor indicators were the youth’s age, gender, race,
diversion group assignment,3 attachment to mother, commitment to conventional
institutions of church and school, involvement in conventional activities, and
expresses shame. A test of the full model versus a model with the intercept only was
not significant (x2 = 9.02, p > .05). The Hosmer and Lemeshow test indicated that the
logistic model’s estimates fit the data at an acceptable level, that is the model was a
good fit (x2 = 11.67, p > .05). The Hosmer and Lemeshow test assesses how closely
the observed and predicted probabilities match, not whether the model’s predictors

were significant (Johnson & Wichern, 2002). The model was able to correctly classify

 

3 Diversion group assignment was originally added in a second model. However, it did not make a
significant difference, so the second model was dropped.

35

88.1% of those who did not re-offend and 17.0% of those who did, for an overall
success rate of only 59.8%.

Table 8 shows the logistic regression coefficients, significance levels, and
odds ratios for each of the predictors. None of the predictors in the model were

statistically significant, employing a .05 criterion of statistical significance.

Table 8. Logistic Regression Model Predicting Odds of Re-offending

 

 

Predictor B p Odds Ratio
Male -.22 .122 .80
Age .1 1 .224 1.1 1
White .12 .147 . 1.22
RJ .18 .329 1.19
Attachment .12 .602 1.13
Commitment -.04 .775 .96
Involvement .14 .215 1.15
Expressed Shame .07 .794 1.07

 

Although not statistically significant, the predictor with the largest effect on
the model was being white. The odds of a white person re-offending were 1.22 times
greater than for a non-white person. In addition, while not statistically significant,
results showed that being male decreased the odds of re-offendin g by 20%, when

holding all others predictors in the model constant. These two findings are

36

interesting, even though they are not statistically significant. Research has
consistently shown that males are more likely to re-offend and whites are less likely
to re-offend (Loeber & Farrington, 1998). The results of this study, although not
Statistically significant, are inconsistent with the convention.

The rest of the predictors had smaller affects. Results showed that the
measures of interdependency and the measure of expressed shame did not
Significantly impact whether a youth re-offended. This is in direct contrast to

Braithwaite’s' (1989) reintegrative shaming theory.

37

CHAPTER V

DISCUSSION
Conclusion and Implications

This study was designed as a partial test of Braithwaite’s (1989) reintegrative

shaming theory by using a sub-sample of youths from the Indianapolis Restorative
Justice Experiment. In accordance with the theory, youths completing restorative
justice conferencing programs should have increased interdependency and higher
levels of expressed shame than youth who completed other diversion programs
(Braithwaite, 1989; Hay, 2001). This study mostly confirms this assertion. Bivariate
analysis showed that youths assigned to restorative justice conferencing were more
attached to their mother, were more involved in conventional activities, and expressed
more shame. However, interdependency was defined through an adoption of
Hirschi’s (1969) social control theory, which combined attachment, commitment to
conventional institutions, and involvement in conventional activities. This study
failed to find a significant difference between youths assigned to restorative justice
conferencing and youths assigned to other diversion programs in terms of
commitment. This is surprising, since commitment is one of two of the elements of
the social bond (attachment being the other) that has the most empirical support
(Costello & Vowel], 1999;'Krohn & Massey, 1980). Further research may aim to
combine commitment and involvement into one measure, such as commitment to
conventional activities and institutions, to examine whether these measures differed

between the two groups.

38

Interestingly, this study failed to find a reduction in re-offending by youths
that completed restorative justice conferencing. Most researchers have found
restorative justice conferencing impacts recidivism in some way (Braithwaite, 1999;
Hayes & Daly, 2003; Hines, 2000; Luke & Lind, 2002; Latimer, Dowden, & Muise,
2001; Maxwell & Morris, 2001; McCold & Wachtel, 1998; McGarrell, 2001;
McGarrell, Olivares, Crawford, & Kroovand, 2000; Sherman, Strang, & Woods,
2000). The results are especially surprising when they are compared to the six and
12-month re-offending results from the Indianapolis Restorative Justice Experiment,
the experiment upon which this study was based. McGarrell, Olivares, Crawford, &
Kroovand (2000) found a 40% reduction in re-offending at the 6-month re-offending
analysis, but by the 12-month re-offending analysis it decreased to 25% (both of
which are significant differences in re-offending between the control and
experimental group)" In this study, there was only a 5.6% difference in re-offending
between groups, which was not significant. This difference could perhaps be linked
to the length of time re-offending were measured, 24-months following the initial
court appearance. Therefore, these results suggest that the reduction in re-offending
will decrease with time. In addition, these findings, taken into consideration with the
finding that the restorative justice group and the control group differed significantly
on interdependency and expressed shame, suggest that restorative justice
conferencing has an immediate reduction in re—offending, but no longer term
reduction. Therefore, prolonged exposure to the intervention may be needed to

impact re-offendin g for the long-terrn. For instance, it may be beneficial to have

 

‘ The re-offending analysis did not take into account the entire sample, due to many youths not
reaching the 12-month mark by publication.

39

offenders repeat the restorative justice conferencing to reinforce the benefits gained
during the initial conference.

Also, future research should aim at analyzing time to failure, which may show
more robust differences in the rates of re-offending by the two groups. Future
research should also aim at conducting surveys prior to the intervention, as well as
repeating the surveys at six, 12, and 24 months to examine whether interdependency
and expressed shame change with the passage of time and to draw a better conclusion
on the impact of restorative justice conferencing on interdependency and expressed
shame.

In addition, multivariate analysis failed to show any significant findings
regarding the impact on interdependency and expressed shame on predicting re-
offending. This is in direct contrast to Braithwaite (1989), who argues there should
be a reduction in re-offending for those who are more interdependent and express
shame. However, Braithwaite (1989) does state the effect of interdependency on re-
offending may be indirect. Future research is needed to determine whether there is a
direct effect or indirect effect for interdependency and re-offending.

On the other hand, Braithwaite (1989) makes it quite clear that shame must be
acknowledged (or expressed) in order for re-offending to be impacted. In fact,
acknowledged shame is viewed as an outcome of reintegrative shaming (Ahmed,
Harris, Braithwaite, & Braithwaite, 2001; Braithwaite, 1989). However, this study
failed to find that expressed shame predicts re-offending. This finding may be due to

the limited scope of the indicators used for the expressed shame measure. Future

40

research Should broaden this measure to include whether the youth reported feeling
ashamed of his or her actions.

Two other interesting findings, although not statistically significant, were that
being male decreased the odds in re-offendin g, while being white increased the odds
of re-offending. Future research should examine why this occurs, and determine
whether the intervention itself has different impacts on males versus females, and
whites versus non-whites.

While interdependency and expressed shame or diversion group assignment
did not predict re-offending in the sub-sample, program completion may. Adding
program completion to an analysis may be beneficial in understanding what reduces

re-offending.

Limitations

This study has several limitations. One limitation is the robustness of its
results. While the Indianapolis Restorative Justice Experiment employed random
sampling, the sub-sample of survey youths did not. As a result, the sub-sample may
not be a good representation of the full sample (Bachman & Paternoster, 1997). This
can be seen in the completion rate of the diversion program. Only eight youths (3%)
of the sub-sample failed to complete the assigned program. This is in stark contrast to
the full sample, in which 29% of the youths failed to complete the assigned program.
This could greatly affect re-offending rates, in which program completion is
predictive of re-offending. In fact, for the full sample, youths that completed either

program were more likely to not re-offend (X2 = 60.07, p < .05). These findings could

41

be interpreted that the sub-sample was not as representative as it could have been.
Future research Should examine the relationship between program completion and re-
offending.

Another limitation is that the results may not be generalizable to other
populations outside of Indianapolis. This is again due in part to the sampling method
employed for the surveyed youths. In addition, the measures of interdependency and
expressed Shame used in this study may not be complete reflections of these concepts.
As stated previously, interdependency can be measured in a variety of ways, as can
expressed shame. The way in which concepts are operationalized can greatly affect .
results (Bachman & Paternoster, 1997). Therefore, if the concepts of
interdependency and expressed shame are measured with different indicatorss, the
results could be affected. Future research should examine other indictors for these
measures.

Restorative justice conferencing may have other benefits that are not
discussed in this study. For example, a cost-analysis should be done to determine
whether restorative justice conferencing is more cost effective than traditional
methods of dealing with juvenile offenders, as well as how it compares with other
diversion programs. In addition, while re-offending rates were not statistically
different between restorative justice conferencing and other diversion programs, there
may be a difference in re-offending between those handled in a traditional way and

those who went through a diversion program.

 

5 Such as the addition of other survey questions, which could add different dimensions to the overall
concept.

42

Overall, findings suggest that the intervention, restorative justice
conferencing, increased interdependency and expressed Shame among this sample of
delinquent youths in comparison to those youths that participated in other diversion
programs. While this did not impact re-offending at 24 months, the intervention

should not be discounted.

43

BIBLIOGRAPHY

Agnew, R. (1985). Social control theory and delinquency: a longitudinal test.
Criminology, 23(1): 47-61.

Ahmed, E., Harris, N., Braithwaite, J., & Braithwaite, V. (2001). Shame management
through reintegration. Cambridge: Cambridge University Press.

Allen, J. D., Moore, C., Kuperminc, G., Bell, K. (1998). Attachment and adolescent
psychosocial functioning. Child Development, 69, 1406-1419.

Bachman, R., ‘& Paternoster, R. (1997). Statistical methods for criminology and
criminal justice. New York: McGraw-Hill.

Bazemore, G., & Dooley, M. (2001). Restorative justice and the offender: the
challenge of reintegration. In G. Bazemore & M. Schiff (Eds), Restorative

community justice: repairing the harm and transforming communities
(pp.101-126). Cincinnati, OH: Anderson Publishing.

Bazemore, G., & Umbreit, M. S. (2001). A comparison of four restorative
conferencing models. Washington DC: U. S. Department of Justice, Office of
Juvenile Justice and Delinquency Prevention.

Benda, B. B., & Corwyn, R. F. (2002). The effect of abuse in childhood and in
adolescence on violence among adolescents. Youth & Society, 33(3), 339-365.

Braithwaite, J. (1989). Crime, Shame and Reintegration. Cambridge: Cambridge
University Press.

Braithwaite, J. (1999). Restorative justice: Assessing optimistic and pessimistic
accounts. Crime and Justice: A Review of Research, 25, 1-127.

Braithwaite, J., & Mugford, S. (1994). Conditions of successful reintegration
ceremonies. British Journal of Criminology, 34, 139-171.

Bureau of Justice Statistics. (2000). Correctional populations in the United States,
1997. Washington DC: U. S. Department of Justice, Office of Justice
Programs, Bureau of Justice Statistics.

Cloward, R., & Ohlin, L. (1960). Delinquency and opportunity: a theory of
delinquent gangs. Glencoe, IL: Free Press.

Cohen, M. (1998). The monetary value of saving a high-risk youth. Journal of
Quantitative Criminology, 14(1), 5-33.

Costello, B. J. & Vowell, P. R. (1999). Testing control theory and differential
association. Criminology, 37(4), 815-842.

Daly, K. (2001). Conferencing in Australia and New Zealand: Variations, research
findings, and prospects. In A. Morris & G. Maxwell (Eds.), Restorative justice
for juveniles: Conferencing, Mediation and circles (pp. 59-83). Oxford: Hart.

Gendreau, P. (1996). Principles of effective intervention with offenders. In A. T.
Harland (Ed.), Choosing correctional options that work. Thousand Oaks, CA:
Sage.

Grasmick, H. G., & Bursik, R. J. (1990). Conscience, significant others, and rational
choice: Extending the deterrence model. Law and Society Review, 24(3), 837-
861.

Hay, C. (1998). Parental sanctions and delinquent behavior: toward clarification of
Braithwaite’s theory of reintegrative shaming. Theoretical Criminology, 2,
419-443.

Hay, C. (2001). An exploratory test of Braithwaite’s reintegrative shaming theory.
Journal of Research in Crime and Delinquency, 38(2), 132-153.

Harrison, P. M., & Beck, A. J. (2003). Prisoners in 2002. Washington DC: U. S.
Department of Justice, Office of Justice Programs, Bureau of Justice
Statistics.

Hayes, H., & Daly, K. (2003). Youth justice conferencing and reoffending. Justice
Quarterly, 20(4), 725-764.

Hirschi, T. (1969). Causes of Delinquency. Berkeley, CA: University of California
Press.

Houts, S. S. (1995). Explaining alcoholism treatment efficacy with the theory of
reintegrative shaming. Alcoholism Treatment Quarterly, 13, 25-38.

Jensen, K. D., & Gibbons, S. G. (2002). Shame and religion as factors in the
rehabilitation of serious offenders. Journal of Ofiender Rehabilitation,
35(3/4), 215-230.

Johnson, R. A., & Wichern, D. W. (2002). Applied multivariate statistical analysis
(5“ ed.). Upper Saddle River, NJ: Prentice Hall.

Johnstone, G. (2002). Restorative Justice: Ideas, Values, Debates. Portland, OR:
Willan Publishing.

Katz, J. (1988). Seductions to crime. New York: Basic Books.

45

Kempf, K. (1993). The empirical status of social control theory. In F. Adler & W. S.
Laufer (Eds.), New Directions in Criminological Theory (pp. 143-185). New
Brunswick, NJ: Transaction.

Krohn, M. D., & Massey, J. L. (1980). Social control and delinquent behavior: an
examination of the elements of the social bond. The Sociological Quarterly,
21(4), 529-543.

Lansky, M. R. (1995). Shame and the scope of psychoanalytic understanding.
American Behavioral Scientist, 38(8), 1076-1090.

Latimer, J., Dowden, C., & Muise, D. (2001). The effectiveness of restorative justice
practices: A meta-analysis. Retrieved January 20, 2004, from
http:canada.justice.gc.ca/enlps/rs/rep/meta-b-e.html.

Lewis, H. B. (1971). Shame and guilt in neurosis. New York: International University
Press.

Loeber, R., & Farrington, D. P. (Eds.). (1998). Serious and violent juvenile offenders:
Risk factors and successful interventions. Thousand Oaks, CA: Sage.

Loeber, R., & Farrington, D. P. (2000). Child Delinquents: Development,
Intervention, and Service Needs. Thousand Oaks, CA: Sage Publications.

Luke, G., & Lind, B. (2002). Reducing juvenile crime: Conferencing versus court.
Crime and Justice Bulletin, 69, 1-20.

Makkai, T. & Braithwaite, J. (1994). Reintegrative shaming and compliance with
regulatory standards. Criminology, 32(3), 361-383.

Marshall, T. (1999). Restorative justice: An overview. London: Home Office.

Maxwell, 6., & Morris, A. (2001). Family group conferences and reoffending. In A.
Morris & G. Maxwell (Eds.), Restorative justice for juveniles: Conferencing,
mediation and circles (pp. 243-263). Oxford: Hart.

Melton, A. (1995). Indigenous justice systems and tribal society. Judicature, 70(3),
126-133.

Miers, D., Maguire, M., Goldie, S., Sharpe, K., Hale, C., Netten, A., Uglow, S.,
Doolin, K., Hallam, A., Enterkin, J., & Newbum, T. (2001). An exploratory
evaluation of restorative justice schemes (Crime Reduction Research Series
Paper 9). Retrieved February 6, 2004, from

http://www.homeoffice.gov.uklrds/prgpdfslcrrsO9.pdf.

46

McCold, P., & Wachtel, P. (1998). Restorative policing experiment: the Bethleham
Pennsylvania police family group conferencing project. Pipersville, PA:
Community Service Foundation.

McGarrell, E. F. (2001). Restorative justice conferences as an early response to young
offenders. Juvenile Justice Bulletin. Washington, DC: U. S. Department of
Justice, Office of Juvenile Justice and Delinquency Prevention.

McGarrell, E. F., Olivares, K., Crawford, K., & Kroovand, N. (2000). Returning
Justice to the Community: the Indianapolis Juvenile Restorative Justice
Experiment. Indianapolis, IN: Hudson Institute, Inc.

Nagin, D. S., & Pasternoster, R. (1994). Personal capital and social control: the
deterrence implications of a theory of individual differences in criminal
offending. Criminology, 32(4), 581-606.

Paternoster, R. & Iovanni, L. (1986). The deterrent effect of perceived severity: A
reexamination. Social Forces, 64(3), 751-777.

Perry, J. (2002). Introduction: challenging the assumptions. In J. Perry (Ed.),
Restorative justice: repairing communities through restorative justice (pp. 1-
17). Lanham, MD: American Correctional Association.

Sampson, R. J., & Laub, J. H. (1991). Crime and deviance over the life course: the
salience of adult social bonds. American Sociological Review, 55, 608-627.

Sampson, R. J., & Laub, J.H. (1992). Crime and deviance in the life course. Annual
Review of Sociology, 18, 63-84.

Sampson, R. J., & Laub, J. H. (1993). Crime in the making: pathways and turning
points through life. Cambridge, MA: Harvard University Press.

Sherman, L. W., Strang, H., & Woods, D. (2000). Recidivism patterns in the
Canberra Reintegrative Shaming Experiments (RISE). Canberra: Centre for
Restorative Justice, Research School of Social Sciences, Australian National
University.

Sherman, L. W., Strang, H., Barnes, G., Braithwaite, J., kaen, N., & The, M. (1998).
Experiments in restorative policing a progress report to the national police
research unit in the Canberra Reintegrative Shaming Experiments (RISE).
Canberra: Australian Federal Police and Australian National University.

Simons, K. J., Paternite, C. E., & Shore, C. (2001). Quality of parent/adolescent

attachment and aggression in young adolescents. Journal of Early
Adolescence, 21(2), 182-203.

47

Snyder, H. N., & Sickmund, M. (1999). Juvenile Ojfenders and Victims:
1999 National Report. Washington, DC: US Department of Justice,
Office of Justice Programs, Office of Juvenile Justice and
Delinquency Prevention.

Toby, J. (1959). Review of family relationships and delinquent behavior by F. Ivan
Nye. American Sociological Review, 24, 282-283.

Umbreit, M. S., & Fercello, C. (1998). Family group conferencing program results in
client satisfaction. Juvenile Justice Update, 3(6), 3-13.

Umbreit, M. 8., Greenwood, J ., Fercello, C., & Umbreit, J. (2000). National survey of
victim-offender mediation programs in the United States; St. Paul, MN: Office
for Victims of Crime.

U. S. Census Bureau. (2002). Indiana: 2000 summary population and housing
characteristics: 2000 census of population and housing. Washington, DC:
U. S. Department of Commerce, Economics and Statistics Administration,
U. S. Census Bureau.

Vagg, J. (1998). Delinquency and shame: data from Hong Kong. British Journal of
Criminology, 38, 247-264.

Van Ness, D. W., & Strong, K. H. (2002). Restoring justice (2nd ed.). Cincinnati, OH:
Anderson Publishing.

Vold, G. 13., Bernard, T. J., & Snipes, J. B. (2002). Theoretical criminology (5th ed.).
New York, NY: Oxford University Press.

Wiatrowski, M. D., Griswold, D. B., & Robbins, M. K. (1981). Social control theory
and delinquency. American Sociological Review, 46, 525-541.

Wicker, F. W., Payner, G. C., & Morgan, R. D. (1983). Participant descriptions of
guilt and shame. Motivation and Emotion, 7, 25-39.

Williams, K., & Hawkins, R. (1986) Perceptual research on general deterrence: A
critical review. Law and Society Review, 20(4), 545-572.

Wong, D. (1999). Culturally specific causes of delinquency: implications for juvenile
justice in Hong Kong. Asia Pacific Journal of Social Work, 9, 98-113.

Wisconsin Division of Juvenile Corrections. (2003). Vision Statement. Retrieved
October 3, 2003 from www.wi-doc.com/index iuvenile.htm.

Wright, R. (2000). Nonzero: the logic of human destiny. New York, NY: Pantheon.

48

Zehr, H. (1990). Changing lenses: a new focus for crime and justice. Scottdale, PA:
Herald Press.

Zehr, H. & Mika, H. (2003). Fundamental concepts of restorative justice. In E.
McLaughlin, R. Fergusson, G. Hughes, & L. Westmoreland (Eds.),
Restorative justice: critical issues (pp. 4043). London: Sage.

Zhang, S. X. (1995). Measuring shaming in an ethnic context. British Joumal of
Criminology, 35, 348-262.

49

IIIIIIIIIIIIIIIIIIIIIIIIIIIII

willjjljllljjjjjuljljjll

   

_.__. l—.——a _—-_— ..—.. _, .