ArticlePDF Available

Comparison of Correctional Services for Youth Incarcerated in Adult and Juvenile Facilities in Michigan


Abstract and Figures

Few studies have compared correctional service experiences of youth in adult prisons and juvenile facilities. This study compared 47 youth in juvenile facilities and 49 youth in adult prisons in Michigan. Controlling for offence history, socioeconomic background, and demographic characteristics, juvenile placement was found to be associated with more positive responses for counseling, medical services, and staff quality. Experiences with the quantity of education and work programs were similar, although the quality of education and work might have differed. Poor correctional services in adult prisons might be a reason for less effective rehabilitation, leading to lower deterrence.
Content may be subject to copyright.
The Prison Journal
92(4) 460 –483
© 2012 SAGE Publications
Reprints and permission:
DOI: 10.1177/0032885512457547
512457547The Prison JournalNg et al.
© 2012 SAGE Publications
Reprints and permission:
1National University of Singapore, Singapore
2University of Michigan, Ann Arbor, MI, USA
3University of Pittsburgh, Pittsburgh, PA, USA
Corresponding Author:
Irene Y. H. Ng, Department of Social Work, National University of Singapore, AS3 Level 4, 3
Arts Link, Singapore 117570, Singapore
Comparison of
Correctional Services
for Youth Incarcerated
in Adult and Juvenile
Facilities in Michigan
Irene Y. H. Ng1, Rosemary C. Sarri2,
Jeffrey J. Shook3, and Elizabeth Stoffregen2
Few studies have compared correctional service experiences of youth in
adult prisons and juvenile facilities. This study compared 47 youth in juvenile
facilities and 49 youth in adult prisons in Michigan. Controlling for offence
history, socioeconomic background, and demographic characteristics, juve-
nile placement was found to be associated with more positive responses for
counseling, medical services, and staff quality. Experiences with the quan-
tity of education and work programs were similar, although the quality of
education and work might have differed. Poor correctional services in adult
prisons might be a reason for less effective rehabilitation, leading to lower
correctional services, youth, juvenile facility, adult prison, Michigan
Ng et al. 461
Between 1988 and 2000, the majority of states in the United States revised
their criminal justice statutes to require the commitment of juveniles who
were convicted of selected crimes in adult court to be sentenced to adult
prisons. Michigan was one of these states, and it significantly increased the
numbers of juveniles, some as young as 14 years, committed to adult prisons.
No provision was made for their placement in segregated housing in the
prisons, nor was there any provision for special programming. These changes
also occurred during a period of the greatest prison overcrowding in
Michigan prisons (Shook & Sarri, 2008). In Michigan, the numbers of juve-
niles sentenced below the age of 17 peaked in 1997, but the total numbers
continue to be above prelegislations rates in 1988. Moreover, the crime rate
by juveniles has continued to decline in 2000 to the present (Michigan State
Police, 2011).
Among the arguments against the incarceration of juveniles in adult pris-
ons are their lack of rehabilitative programs, especially those that target the
developmental needs of youth (Fagan, 2008; Redding, 2003). If their asser-
tions are correct, there is cause for concern given that most offenders will be
released back to the community without having received educational, mental
health and employment services that they need if they are not to recidivate
(Goebel, 2005; Strom, 2000).
Are correctional services in juvenile facilities necessarily better and more
targeted for youth? This is a moot question. While some researchers assert
that juvenile facilities are more rehabilitative because they are usually gov-
erned by more requirements for educational programming and counseling as
well as more individualized and nonpunitive treatment (Bishop & Frazier,
2000; Redding, 2003), others argue that adult prisons often have targeted
programs and separate housing for youthful offenders (Kupchik, 2007).
However, Austin, Johnson, and Gregoriou (2000) found few prisons where
juveniles were housed separately. Instead most juvenile offenders are housed
in the regular prison system (see also Deitch et al., 2009) with offenders of all
ages. Further, they found little evidence of efforts to customize programs for
the needs of young offenders in adult prisons. Yet Redding (2003) also noted
that overcrowding in both juvenile and adult facilities resulted in inadequate
services and treatment in both juvenile and adult facilities.
In Michigan, overcrowding in prisons and budget constraints had led to
innovations in correctional programs (Michigan Department of Corrections
[MDOC], 1998, 2005). For example, Computer Assisted Instruction (CAI)
shifted routine instruction from human teachers to computers (MDOC, 1998,
462 The Prison Journal 92(4)
p. 35). CAI was slated mainly for adult basic education, with no college
education provided unless it is court-mandated (MDOC, 1998, pp. 82).
Consequently, with savings on human resources in MDOC, none of the pris-
ons claimed to have their own trained treatment or education staff (MDOC,
various years). By comparison, each public juvenile facility in Michigan had
professionally trained teachers and social workers (Dixon, 2007). Past law-
suits in Michigan over health and mental health services in prisons also signal
ongoing problems in the Michigan criminal system (e.g., Clark-Murphy v.
Foreback, 2006; Hadix v. Caruso, 2009).
Given the above backdrop of innovations in the midst of budget constraints,
what are the experiences of youth incarcerated in adult correctional institu-
tions compared to juvenile facilities, as voiced by the youth themselves? This
article compares the experiences of 49 youth who were interviewed in adult
prisons and 47 youth interviewed in juvenile facilities in Michigan, on their
access to work and education programs, counseling, medical services as well
as their ratings of staff quality. All respondents were committed before age 17
(age of adult jurisdiction in Michigan), but were unmatched otherwise.
Although results from this small study of unmatched respondents in Michigan
cannot be generalized, it provides further corroboration of findings from sev-
eral studies which compared these populations. In addition, studies reflecting
the voices of inmates are important because programs, as established, might
look very different from programs as experienced by inmates.
Prior Research on Services
and Treatment in Penal Institutions
Few studies have compared the correctional service experiences of youth in
juvenile and adult incarcerations directly. In their search for answers to
understanding the impact of transfers on young people, Bishop and Frazier
(2000) opined that most helpful were two studies that compared the views of
adolescents who remained in the juvenile system and those who were trans-
ferred. The two studies cited were Forst, Fagan, and Vivona (1989) and a
report by Bishop et al. (1998) which was later published as Lane, Lanza-
Kaduce, Frazier, and Bishop (2002). Our search since then unearthed only
two more comparative studies. All except one found experiences in juvenile
facilities more positive.
In Forst et al. (1989), 59 juveniles in juvenile facilities were compared
with 81 juveniles transferred to prisons in four states. Respondents in juve-
nile institutions reported that they received more counseling, found treat-
ment more helpful, and rated staff more highly. Lane et al. (2002) compared
Ng et al. 463
how 144 juveniles rated the impact of their dispositions on recidivism. Half
of the respondents received juvenile dispositions and half adult sentences.
Adult sanctions—which included imprisonment (majority), probation, and a
mixture of both probation and imprisonment—were preferred to “low-end”
juvenile sanctions such as probation and short-term and low-supervision
residential placements. The reasons were negative, that adult sanction would
deter them from reoffending because of how “horrible” (p. 444) prison was.
In contrast, “deep-end” juvenile sanctions (e.g. closed institutions) were the
most preferred, and for positive reasons, that they “gained life skills, got
counseling, and were given hope” (p. 450). LaFree (2002) compared the
services available to violent juvenile offenders in New Mexico adult facili-
ties (mostly prisons) with those available to violent offenders in juvenile
institutions and those available to violent offenders on probation. The
imprisoned juveniles were much less likely to be enrolled in an educational
program or a counseling program. However, they had higher rates of work
program participation.
One study, however, did find that young people in adult prisons reported
higher levels of service availability in adult prisons (Kupchik, 2007). This
study included a comparison of 95 youth in juvenile and adult facilities along
three dimensions—institutional service availability, staff mentoring, and fair-
ness. Controlling for confounding factors such as demographics and prior his-
tory with crime and justice, the study found that inmates in adult prisons gave
more favorable ratings for service availability while those in juvenile facilities
rated staff mentoring more highly. One reason given for this finding was that
young inmates in adult prisons rated higher service availability because edu-
cational and work opportunity programs in prisons might be concentrated on
young offenders, with the rest of the adult prisoners receiving minimal ser-
vices. Summing up the findings from the four articles reviewed above that
have directly compared juveniles in adult and juvenile placements, only
Kupchik made comparisons controlled for possible confounders, and only
Kupchik found more positive ratings overall by juveniles in adult prisons.
While direct comparisons of juvenile and adult facilities are few, one
might glean insights from studies on conditions in each of the systems sepa-
rately. Studies assessing programs in juvenile institutions generally conclude
that conditions and services are inadequate. For example, in a national survey
of youth in residential placement (SYRP) by the Office of Juvenile Justice
and Delinquency Prevention (OJJDP), incarcerated youth were found to have
higher risks of mental health, health, and educational needs than the general
population, but services were found inadequate relative to the risks and needs
of these youth (Sedlak & McPherson, 2010).
464 The Prison Journal 92(4)
In terms of education, Foley (2001) reviewed the literature and suggested
that education for incarcerated youth in juvenile facilities is generally avail-
able because of state law requirements regarding school attendance (Foley,
2001). However, the large increases in the incarcerated population over the
last several decades have resulted in a decline in education programs (Zweig,
2003). Further, the needs of youth in incarceration are often greater than the
educational programming being provided in penal institutions. In the SYRP,
the rate of learning disability among respondents was seven times the rate
among the general population. Although the majority of SYRP respondents
attended school while in custody, only 46% of those with a diagnosed learn-
ing disability received special education (Sedlak & McPherson, 2010).
Consequently, it is apparent that educational services in juvenile facilities are
not necessarily meeting the needs of many youth and that improvements are
needed to both increase completion rates and special needs education.
In terms of counseling and mental health programs, the SYRP findings
indicate that although mental health services were available in all facilities
surveyed, only 47% of youths were in facilities that provided these services
to all residents. Given the report by Grisso, Vincent, and Seagrave (2005) that
two of three youth in juvenile facilities have a diagnosable mental health
disorder, this indicates a substantial gap. They also noted that the likelihood
of receiving counseling did not depend upon proper diagnosis. Further, the
literature suggests that the effectiveness of programs depends upon the kind
of treatment provided. Cognitive-based therapy that focused on changing
behavior was found to reduce misconducts in prisons and postprison. On the
other hand, nondirective and unstructured counseling was not effective
(French & Gendreau, 2006; Mackenzie, 2000). Physical health care was gen-
erally available in juvenile facilities, but usually that at a basic level. Needs
in terms of dental, vision and hearing care was often inadequate (Sedlak &
McPherson, 2010).
With respect to services to youth in adult prisons, educational programs are
closely tied to work programs that are necessary not only for reintegration, but
also for institutional maintenance (e.g., building maintenance, cooking, grounds
care). An Urban Institute report, for example, studied training and employment
programs in prisons, which included education, vocational training, prison
industry, employment/transition training (Lawrence, Mears, Dubin, & Travis,
2002). It posited that prison programming of all types decreased as prison pop-
ulations increased in the 1990s. Because the educational levels of the prison
population trailed those of the general population, education programs were
identified as the primary programs for rehabilitation. Yet these programs
were deemed to be inadequate due to prison crowding and participation rates
Ng et al. 465
were low and varied across states (ranging from 6.4% in Illinois to 58.4% in
Ohio). The Urban Institute study revealed that for the four different types of
employment programs, too much emphasis was placed on occupying pris-
oners’ time and too little on developing skills for postrelease employment.
Furthermore, another study, on a National Supported Work Demonstration
Project, found that work programs did not reduce reoffending among individu-
als in their teens and early 20s (Uggen, 2000). These programs only reduced
crime for those aged 27 and above. Findings such as these might provide justi-
fication for focusing on education and not work with young offenders.
The inadequacy and lack of health care in prisons has provoked a large
number of class actions suits. For example, Plata v. Davis (2003), Hadix v.
Caruso (2009), Clark-Murphy v. Foreback (2006) have resulted in a number
of decisions mandating improved health care in the respective states.
Noteworthy for the present study is that the latter two cases were in Michigan.
In addition, existing research suggests that health care in prisons is insuffi-
cient because prisoners tend to be a population with more medical problems
and therefore greater healthcare needs (Anderson et al., 1998; Lindquist &
Lindquist, 1999). The medical needs of prisoners were also highlighted from
the perspective of the specific needs of certain populations (e.g., older or
women prisoners), the more prevalent types of health needs (e.g., mental
health, HIV, TB), and the fact that many prisoners originated from disadvan-
taged communities that lacked medical facilities. From the perspectives of
these specific groups, transitional medical attention for reentry to community
was found wanting (Hammett, Gaiter, & Crawford, 1998).
In addition to the provision of services, the quality and professionalization
of staff is key to shaping the quality of services offered and the relationships
that they form with staff. Redding (2003) and Forst et al. (1989) suggest that
a caring and fair staff is especially important to young offenders. However,
this is one area where the different goals of the juvenile and criminal justice
systems can be expected to produce contrasting differences in staff treatment
of inmates. Bishop and Frazier (2000) proposed that while juvenile facilities
are supposed to foster messages of caring and to be treatment oriented, a
prime concern of prisons is security. In their comparison of preadjudicatory
confinement, Bishop & Frazier found that generally, respondents in both jails
and juvenile detention centers viewed staffs as indifferent. However, many
youths who repeatedly cycled through detention centers reported forming
significant attachments to at least one staff member. In contrast, respondents
who went through jails reported no such attachments. Instead, they reported
that jail officers did not differentiate between them and the chronic and vio-
lent adult offenders with whom they were housed.
466 The Prison Journal 92(4)
The review in this section has highlighted that services provided to young
people in correctional facilities is lacking, but the problem appears greater in
the adult prisons than in juvenile facilities despite the high needs of these
youth. As a substantial number of youths are placed in adult facilities, the
implications of sending them to adult prisons needs to be better understood.
Yet, the research base for direct comparisons is fairly limited and there is a
need for more research documenting the differences between experiences in
juvenile and adult facilities. This is particularly important in light of research
that shows that transferred youth are more likely to recidivate than those
retained in juvenile facilities. Several recent reviews have examined this
research and have documented these differences in recidivism, attributing
some of the differences to the conditions youth experience in the adult system
(Fagan, 2008; Hartney, 2006; Redding, 2008).
As the only study besides Kupchik (2007) to provide a multivariate com-
parison of correctional services in juvenile and adult incarceration, the find-
ings in this study will add much needed evidence to the current limited pool
of knowledge. While LaFree (2002) and Lane et al. (2002) made no claims to
matching their samples of juvenile and adult respondents, Forst et al. (1989)
matched on the same offence criteria. Further, Bishop and Frazier (2000)
cautioned that facilities in the jurisdictions that they and Forst et al. studied
may not be representative. While the samples in our study were also not
matched aside from the fact that all had committed serious crimes before age
17, they included youths placed in a range of institutions in Michigan, thereby
accounting for different types of services offered in a range of institutions.
This study compares the responses of youth in juvenile and adult placements
on four types of correctional services: formal education, work and employ-
ment, counseling, healthcare, and staff quality. The questions on services were
collected as part of a larger study of juveniles committed to adult prisons in
Michigan. Interviews were conducted with a sample of 47 youth in the
Department of Human Services (DHS), which had jurisdiction over juvenile
facilities, and 49 youth from the Department of Corrections (DOC), which had
jurisdiction over adult facilities. Cases were selected by administrators in DHS
and DOC along two main criteria. As one aim of the larger study was to learn
how the youth were being prepared for reentry and reintegration back to their
home communities, all cases were identified as meeting the criteria of being
Ng et al. 467
released from their respective facilities within 6 months. The more important
criterion was that all selected cases had to be committed for crimes that were
committed below the age of 17, the age of adult court jurisdiction in Michigan.
As such, both samples entered a closed juvenile facility or an adult prison at
the median of 16 years and were getting ready for release into the community.
The juvenile sample included 33 youth who were briefly residing in five
halfway houses who were responsible for preparing the youth for release
from the DHS plus an additional 14 who were referred directly from two
state training schools. All of these youth had spent 3 or more years in the state
training schools in Michigan. The youth were interviewed in private rooms or
corners at the respective facilities. The 49 youth in adult prisons were identi-
fied by the DOC as individuals who had been included in the larger sample
of 2,240 youth sentenced as juveniles to the Department of Corrections
between 1985 and 2004. The interviews were conducted in eight different
prisons in Michigan, and all took place in private rooms with no staff present.
The median length of stay for those in adult prisons was 5 years, but some of
these youth had been released on parole earlier, reoffended and returned to
prison and were ready for discharge at the time of the interviews.
A limitation of the sample of youth in adult facilities is that it did not
include youth who had life or very long sentences because they were ineli-
gible for release. For example those convicted of first degree murder were
excluded whereas the juvenile sample did include youth who had been placed
in facilities for very serious felonies, including murder. Also excluded were
individuals over the age of 30 who had been imprisoned as juveniles so as to
have closer age comparability of both samples. Another limitation is that the
sampling strategy did not match specific offense and offender characteristics
across the two samples. Having a matched sample would have been impos-
sible given the focus of the larger study to examine the preparation of the
youth for reentry. The court systems varied for these two populations espe-
cially with respect to sentencing and discharge decisions. Although we con-
trolled for some of these offense factors in the multiple regressions, it is
possible that any differences in service receipt may be a result of differences
in a individual youth’s characteristics.
While recognizing that the profiles of these two groups are not closely
comparable, a major strength of the sample is that youth were drawn from
a range of state run correctional institutions in both the juvenile and adult
justice systems. Thus, our findings are not particular to a specific institution
or group of offenders, but represent the experiences of individuals across
those two systems as many of these youth were likely to be in a number of
different institutions during their time in the systems.
468 The Prison Journal 92(4)
The procedures and instruments for this study were all approved by the
Behavioral Science Institution Review Board (IRB) of the University of
Michigan on April 12, 2006, followed by periodic reviews until the study was
completed in 2009. A privacy certificate was obtained from the National
Institute of Child Health and Human Development on July 28, 2006 to assure
proper procedures and to protect the research staff in the event of subpoena.
Our analysis applied multiple regression, where the dichotomous variable of
incarceration in an adult corrections institution (DOC = 1) as opposed to
placement in a juvenile facility (DOC = 0) was regressed on services vari-
ables as dependent variables. The dependent variables captured various fac-
tors identified to aid rehabilitation and reintegration: work and education,
counseling, medical services, and staff quality. In total, the 10 services
variables defined below were analyzed
1. Work and education were measured with four variables: (a) cur-
rently assigned a work order (yes = 1); (b) number of hours per day
at work, with those not in a work program excluded; (c) currently in
an education program (yes = 1); and (d) number of hours per week
attending class, with those not in an education program excluded.
2. Counseling was analyzed using three variables: (a) whether coun-
seling was required (yes = 1); (b) number of counseling hours
received per week only for those who answered that they received
counseling; and (c) whether respondent preferred to see a counselor
more (yes = 1).
3. Medical services had two variables: (a) frequency of seeing medi-
cal doctor or nurse (1 = never to 4 = once a month or more); and
(b) whether respondent ever wanted to see a doctor but was not able
to (yes = 1).
4. To measure the quality of staff, a scale was created from the sum-
mation of ten questions asking respondents to rate from 1 for no
staff to 4 for most staff the proportion of staff who met the descrip-
tions below. All 10 items loaded well together, and had a high
consistency value of α = .89. The items included (a) show little
concern for my well-being (reverse-coded); (b) treat me unfairly
(reverse-coded); (c) treat me with respect; (d) emphasize education;
(e) emphasize treatment; (f) are positive role models; (g) punish
us for things we did not do (reverse-coded); (h) are unclear about
Ng et al. 469
expectations (reverse-coded); (i) help us with family problems; and
(j) help us get along with each other.
Among the independent variables, the main variable of interest was
whether the respondent was incarcerated under the Department of Corrections
(DOC = 1) for adult incarceration or the Department of Human Services
(DOC = 0) for juvenile incarceration. The other variables served as con-
trols to “DOC.”
First, the types of services received are likely to be related to the serious-
ness and persistence of offending history. Risk assessments are often con-
ducted to evaluate the types of programs inmates need for discipline while
incarcerated and rehabilitation towards reintegration. Offence history was
measured in three ways. First, a dichotomous variable was created from five
types of offenses for which the current placements were meted out. Person or
weapons offences were considered more serious and given the value 1.
Property, drug, or status offences were considered less serious and took the
value 0. The second and third offense history variables identified a persistent
offender by the age at which the respondent was arrested for the first offence
and the number of times the respondent has ever been arrested. Age of first
arrest was rated as a ranked variable from below 10 years old to 17 years old,
with one year increment in between. Each of the three measures are limited
reflections of offence history, but three imperfect measures together provided
relatively good controls for offender characteristics that might have led to the
type of programming received by respondents.
Receipt of services might also be related to family background factors
such as being poor or a foster child. Individuals from impoverished families
or those that grew up in the foster care system are often in more need of ser-
vices. In addition, they might be less forthcoming or capable at accessing
in-prison services, or, particularly in the case of those who were in the foster
care system, they might be skeptical of services. Consequently, we included
two dummy variables, one for whether one’s family was on public assistance
while growing up (proxy for poverty) and one for having been in foster care.
Whether one is referred for services should depend also on one’s need for
such services, e.g. due to mental health problems or substance abuse prob-
lems. Prison populations have been identified to have high prevalence of
both (Abram, Teplin, McClelland, & Dulcan, 2003; Grisso et al., 2005). One
indicator of mental health in the survey was a set of 20 self-reported ques-
tions from the Centre for Epidemiologic Studies–Depression Scale (CES-D,
Radloff, 1977). Each answer in the CES-D scale ranged from 0 to 3, so that
after summing the 20 individual items, those with scores above 16 were
470 The Prison Journal 92(4)
classified as depressed. Substance abuse was also added as a control, where
respondents were classified as abusers of substances if their lifetime use of
(a) marijuana or hashish, (b) prescription drugs without doctor’s prescription,
or (c) cocaine, crack, heroin, or methamphetamine, were forty or more times.
Finally, services might be offered and accepted differentially by inmates
of different ethnicity, sex, and age. In this analysis, age was expressed in
years; ethnicity as a dummy for white respondents; and sex as a dummy for
female respondents. The base group was therefore minority male youths.
Analytical Strategy
Empirical studies are currently reverting from logistic regression to linear
regressions for binary dependent variables, after critical research such as
Mood (2010) which pointed out the inaccuracy of estimates derived from
logistic regression. Due to the functional form of the logistic model, its esti-
mates reflect not only the effects of the independent variable of interest, but
also unobserved heterogeneity in the empirical model. Therefore, whether
reported as coefficients, odds ratios, or probability changes, estimates from
logistic regression are not comparable across different empirical models with
different sets of variables. Linear probability models (linear regression with
binary dependent variables) have been proposed as a more accurate alterna-
tive to logistic models. Although a linear model is a wrong functional form
for binary variables, the probability measures are felt to be less problematic
unless many predicted values fall beyond the 0 to 1 range.
In light of the above empirical development, Ordinary Least Squares
(OLS) regression was used as the empirical model for our analysis of all ten
dependent variables. This includes the binary variables (whether working,
whether in education program, whether counseling required, whether wants
more counseling, whether not able to see doctor) and the rank-ordered vari-
able (Frequency of seeing doctor).
The regressions were done after multiple imputation of the data, due to
missing observations in some variables. Table 1 lists the number of cases in
DOC and DHS for each variable. Missing data was particularly high for
hours of work and education (12%-17%), the counseling variables for the
DOC sample (4%-16%), age of first arrest for the DHS sample (14.9%), and
number of arrests for DHS (25.5%).
The pattern seems to be that the group with more negative answers had
more missing data. The DOC sample reported worse service outcomes and
had more nonresponse for the education and counseling variables; the DHS
sample had more negative offender characteristics and had more missing
Table 1. Descriptive Statistics of Services Variables, Covariates, and Auxiliary Variables used in Imputation.
Adult incarceration (N = 49) Juvenile incarceration (N = 47)
Mean SD Range % NMean SD Range % N
Dependent variables
Percent with current work detail 66.67 48 76.60 47
Mean hours of work per day* 4.14 2.56 0-8 29 3.49 3.23 .5-10 28
Percent currently in education program 63.64 44 80.85 47
Mean hours of education program per week* 8.98 8.30 0-35 26 18.56 13.38 0-40 32
Percent who were required to attend counseling 63.41 41 76.60 47
Mean hours of counseling per week* 1.50 1.22 .5-5 36 6.50 5.15 .5-20 45
Percent who would like to see counselor more often 25.58 43 15.22 46
Mean frequency per year of seeing doctor or nurse (1-4)* 2
or less)
49 3
(A few
Percent ever wanted to see a doctor but not able to 41.67 48 28.26 46
Mean staff quality index (0-40)* 18.71 0-37 49 28.72 0-40 47
aDifference between respondents in adult and juvenile incarceration significantly different at 5%.
472 The Prison Journal 92(4)
arrest information. The data is therefore not missing at random (NMAR).
This is problematic because multiple imputation requires an assumption of
missing at random, especially since it is the dependent variables that were
missing. Imputing dependent variables might result in overestimation.
However, with good auxiliary variables (i.e., variables used in imputation but
not in the analysis) and a correctly specified model for the missing data
mechanism, multiple imputation can provide unbiased and efficient estimates
due to the random component generated by multiple imputation. While it is
possible to identify good auxiliary variables (those with moderate to high
correlation with the variables having missing data), there is no statistic to
help identify whether a missing data model is appropriate. One approach is to
try different imputation models and test whether estimates are sensitive to the
model (see Allison, 2009, for explanation of imputation when missing cases
are NMAR or reside in dependent variables).
The current data had a few good candidates for auxiliary variables, and dif-
ferent imputation models changed the significant results minimally. The final
imputation model used three auxiliary variables and yielded conservative esti-
mates that were not more significant than an analysis with pairwise deletion.
“Whether respondents attended counseling” had complete data and was natu-
rally highly correlated to the counseling variables. It also had moderate correla-
tion with the other services variables and a few independent variables, such as
age of first arrest. “Frequency of family visits” was a rating from 1 for never to
6 for weekly. It had low correlation with the dependent variables but high cor-
relation with some independent variables, namely frequency of arrest, current
age, and foster care. Being from Wayne County was an interesting auxiliary
variable. This County is probably best known for the city of Detroit, which hails
the state’s highest crime and poverty rates. For juvenile justice, on the other
hand, Detroit has implemented creative diversionary programs for juveniles and
also sentenced very few youth to adult facilities because there was a provision in
the transfer laws that allowed them to try a juvenile as an adult, but then place
them in a juvenile facility and exonerate them at 21 years if that had not commit-
ted another violation. Therefore, juveniles from Wayne County have a higher
likelihood of juvenile placements. As a result, the variable correlates relatively
well with both the dependent and independent variables. In particular, since the
population of Detroit is large and made up of majority African Americans, the
change in the system in Wayne County had led to lower rates of African
American youth in adult prisons than in juvenile facilities.
Multiple imputation was done using the ICE command in STATA with
five data sets. All the dependent, independent, and auxiliary variables were
used in the all the imputations, except for the work and education variables.
Ng et al. 473
As hours of work was derived from whether the respondent was in a work
program, hours of work was excluded from the imputation of being in a work
program and vice versa. This was also the case for the two education vari-
ables. The other variables were relatively independent.
As the purpose of imputation was to generate predicted values of the
imputed variables, and not to analyze coefficient estimates of regressions, the
standard regressions were applied to each type of imputed variables: linear
regressions for continuous variables, logistic regressions for binary variables,
and ordered logistic regression for the rank-ordered variable.
Sample Profile
The descriptive statistics of the variables indicate that the youth in adult
incarceration rated lower on services but that the youth in juvenile place-
ments had more disadvantaged backgrounds (Table 1). Youth in adult pris-
ons had more negative responses to all the service variables except hours at
work per day. However, only the differences in hours of education, counsel-
ing attendance, hours of counseling, frequency of medical visits, and quality
of staff were significant. The youth in DHS, on the other hand, had commit-
ted more serious offences, experienced their first arrest at a younger age, and
were more likely poor. However, they reported lower numbers of arrests.
They were also younger and more likely to be females.
Indeed, one limitation of the data is the overrepresentation of certain pro-
files of respondents. There were only two females in the adult prison sample.1
The respondents in DOC were also much older, even though they had all
been initially committed to DOC before their 17th birthday (age of adult
jurisdiction). The mean (and median) age at the time of the interview was
24 years for the DOC group compared to 18 years for the DHS group. The
above limitations in getting more closely matched samples were due to the
difficulty of interviewing respondents in the various facilities scattered
around Michigan, based on lists generated by headquarter administration.
The descriptive statistics of the auxiliary variables for the multiple imputa-
tion are provided in Table 2. A higher percent of youth in juvenile placements
received counseling and originated from Wayne County, but there were no
significant differences in the frequency of visits from family members.
Multivariate Results
Table 3 presents the unit and linear probability changes of adult versus juve-
nile incarceration on the various services, controlling for possible confound-
ing factors. In each case where there was a significant difference, respondents
Table 2. Descriptive Statistics of Services Variables, Covariates, and Auxiliary Variables used in Imputation
Adult incarceration (N = 49) Juvenile incarceration (N = 47)
Mean SD Range % NMean SD Range % N
Median age of first arresta14
2.10 <10-17 46 13
1.45 <10-17 40
Mean number of arrestsa9.73 7.54 1-37 47 5.54 5.25 0-30 35
Percent whose family was on public
assistance while growing upa
47.92 48 76.09 46
Percent who had been in foster care 34.69 49 44.44 45
Mean current agea24 1.32 21-29 49 18.15 1.02 16-20 47
Percent who are White 47.92 48 38.30 47
Percent who are femalea4.08 49 21.28 47
Percent who are depressed 51.02 49 44.68 47
Percent who abused substancesa81.63 49 46.68 47
Auxiliary Variables
Percent who attended counselinga77.55 49 95.74 47
Mean frequency of seeing family (1-6) 3 (Every
48 4
Percent from Wayne Countya12.24 49 34.04 47
Note: Range in parenthesis.
aDifference between respondents in adult and juvenile incarceration significantly different at 5%.
Ng et al. 475
reported receiving more or better services in the juvenile system. Across the
five domains—work, education, counseling, health, and staff quality—edu-
cation was the only area where juvenile respondents did not report receiving
more services.
As the table shows, respondents in adult prisons reported lower probabil-
ity of working than those in the juvenile system. In fact, the probability of
working in the adult system was 63 percentage points lower than that in a
juvenile placement. However, there were no significant differences in the
number of hours of work that respondents reported. This indicates that,
accounting for other factors, juvenile facilities offered more work opportuni-
ties to respondents but the number of hours worked across these two systems
were similar.
As noted above, no significant differences were found with regard to
receipt of education services. However, the multivariate results mask differ-
ences in terms of types of education received. More respondents in the juve-
nile sample were in high school and college programs, whereas more
respondents in the adult sample were in GED and vocational programs (Table 4).
Furthermore, the qualitative components of our interviews suggest that in the
adult system, GED education was largely by self-study and that the voca-
tional programs were tied to work that inmates were assigned to. Hence,
while the quantity of education received was not found to be different, the
quality of education might well be.
The differences in counseling, medical attention, and staff quality were
significant. Although there was no significant difference in counseling being
required in the two types of incarceration, DOC respondents received seven
fewer hours of counseling and their probability of wanting more counseling
were 58 percentage points more than the DHS respondents. For medical ser-
vices, DOC respondents rated one point lower in the ranked frequency of
doctor visits than DHS respondents, and their probability of not being able to
see a doctor when needed was 57 percentage points more than their counter-
parts in DHS. DOC respondents also rated staff quality 12 points lower than
DHS respondents.
For the control variables, older respondents had fewer hours of counseling
and lower frequency of seeing a doctor. The rest of the covariates were sig-
nificantly associated with only one of the ten service variables, or none at all.
Overall, it is striking that the effects of the control variables are mostly insig-
nificant, yet the negative effects of adult incarceration on correctional ser-
vices held for six out of the ten models.
Table 3. Linear Probability/OLS Effects of Youth in Adult Prisons on Services, Results by Multiple Imputation
(2) (3) (4) (5) (6) (7) (8) (9)
Hours of
work Education
Hours of
Hrs of
Wants more
Frequency of
seeing doctor
Not able to
see doctor Staff quality
Adult prison −0.63 −0.77 0.15 −7.61 −0.21 −6.77 0.58 −1.08 0.57 −12.60
(0.26)* (3.05) (0.28) (9.63) (0.28) (2.31)** (0.24)* (0.45)* (0.26)* (4.47)**
Serious offence −0.12 1.75 −0.080 −3.47 0.071 −0.033 0.15 0.18 −0.049 −0.58
(0.11) (1.04) (0.11) (3.90) (0.12) (0.99) (0.096) (0.18) (0.11) (1.74)
Age of first
−0.035 0.19 −0.026 −0.68 −0.013 0.20 0.072 −0.001 −0.041 0.42
(0.032) (0.31) (0.032) (1.45) (0.054) (0.32) (0.040) (0.055) (0.033) (0.56)
Number of
−0.002 0.051 0.005 −0.32 −0.010 0.089 −0.006 0.012 0.005 0.32
Arrested (0.009) (0.12) (0.009) (0.39) (0.008) (0.085) (0.007) (0.015) (0.009) (0.19)
Family received −0.14 −0.09 0.077 −1.95 −0.045 −0.86 0.14 0.092 0.050 1.39
(0.11) (0.99) (0.12) (3.93) (0.12) (0.97) (0.12) (0.19) (0.11) (1.92)
Was in foster
−0.043 0.37 −0.21 −1.93 0.003 −0.20 0.033 −0.19 −0.16 −4.17
(0.10) (1.11) (0.11) (4.30) (0.11) (0.97) (0.094) (0.18) (0.10) (1.79)*
Depressed 0.087 0.44 −0.048 −0.29 0.014 0.071 −0.067 0.050 −0.073 0.22
(0.041)* (0.52) (0.046) (1.57) (0.044) (0.36) (0.037) (0.071) (0.042) (0.69)
0.25 −0.084 −0.003 3.95 0.14 −0.042 0.028 −0.095 −0.20 −1.04
(0.10)* (0.93) (0.10) (3.57) (0.11) (0.95) (0.099) (0.17) (0.10) (1.71)
Current age −0.085 −1.01 0.023 6.54 −0.20 −5.58 0.068 0.65 −0.038 −3.77
(0.16) (1.61) (0.16) (4.72) (0.16) (1.33)** (0.14) (0.27)* (0.16) (2.63)
White −0.051 −0.42 −0.043 −1.72 −0.020 −0.008 0.11 0.26 0.25 −2.05
(0.097) (0.83) (0.10) (3.18) (0.11) (0.87) (0.092) (0.17) (0.099)* (1.64)
Female −0.18 −1.03 −0.15 1.15 0.028 −0.91 0.040 0.22 −0.042 −2.00
(0.12) (1.11) (0.12) (4.85) (0.14) (1.09) (0.11) (0.21) (0.12) (2.00)
Constant −0.14 −8.49 2.16 36.09 0.72 4.59 0.17 1.64 2.15 21.77
(0.84) (9.08) (0.90)* (32.14) (0.98) (7.82) (0.91) (1.46) (0.85)* (14.73)
N96 68 96 70 96 83 96 96 96 96
R20.17 0.20 0.15 0.30 0.13 0.50 0.20 0.35 0.25 0.44
Note: Standard errors in parentheses.
*significant at 5%. **significant at 1%.
Ng et al. 477
Controlling for factors that might influence service use, including offence
history, receipt of public assistance, foster care, depression, substance abuse,
age, race, and sex, the conclusion is clear: youth sent to adult prison as juve-
niles in this Michigan study generally experienced fewer services than youth
in juvenile facilities and reported considerably lower staff quality. This con-
clusion concurs with a few other studies that have compared these two popu-
lations (Forst et al., 1989; LaFree, 2002, Lane et al., 2002). However, it
differs from Kupchik (2007), who found that adult inmates reported better
institutional services. Kupchik’s institutional service index averaged affir-
mative answers to whether the following were available: library, religious
services, counseling, family care/parenting classes, job training, education,
drug treatment, and other. It would have been interesting to see whether the
10 individual services in its index were significantly different for the two
groups of respondents. In the present analysis, while the Michigan adult-
placed inmates did not have less access to education, they did report receiv-
ing less counseling, wanting more counseling, seeing a doctor less frequently,
and wanting to see a doctor more. In both studies, staff was rated more
negatively by adult-placed inmates. The agreement in the findings on staff
quality, in particular, might reflect the more caring orientation of the juvenile
system in contrast to the more security orientation of the prison system.
What accounts for the differences in the findings for the other services
variables? It might be that the adult sample in this study was older.2 It is pos-
sible that more services might have been offered when they were younger.
However, a number of studies cited earlier (e.g., Austin et al., 2000; Deitch et al.,
2009) have found that there is an overall lack of separate services for youth
in adult prisons. While the age gap in the juvenile and adult samples in our
study is wider than that in Kupchik, 24 is still young relative to the rest of the
prison population. Youth at this age still crucially require education, work,
medical, and other services that are important as investments towards reinte-
gration in the community.
Table 4. Types of Education Program Received in the Juvenile and Adult Systems
DOC N = 49 DHS N = 47
GED 36.73% 18 25.53% 12
High school 6.12% 3 44.68% 21
College 6.12% 3 21.28% 10
Vocational training 22.45% 11 12.77% 6
478 The Prison Journal 92(4)
Besides the age gap between our samples from the adult and juvenile sys-
tems (24 years old for the DOC sample and 18 years old for the DHS sample),
the DHS sample also consisted of more serious offenders. Indeed, a limita-
tion of this study is that the two samples were not matched. The main basis
for comparison was that if not for tougher legislation and decision making
practices at the county level, both groups would have remained in the juve-
nile system. Yet, as noted previously, the strength of the study is that the
sample included young people placed across a range of institutions in the
juvenile and adult justice systems and tapped into services being provided
prior to a youth transitioning back to the community. Consequently, its find-
ings are not limited to a particular institution or type of offender, but, instead,
account for the services offered in a range of institutions. This is important
because young people in the justice systems vary considerably and their
experiences often include spending time at different facilities. Further, it is
important because services delivered prior to a youth transitioning back to the
community can be instrumental to increasing the likelihood of a successful
The differences that we found across these systems are important for two
of reasons. First, both samples of juvenile and adult-placed respondents in
our study were being prepared for discharge. Yet, the older group in DOC
was being released into the community at an older age but, potentially, more
disadvantaged in terms of rehabilitation. Hence, despite the limited compara-
bility in terms of age and severity of offences committed by respondents, the
findings in this study indicate the need for considering service quality as
another factor that might undo the intended deterrent effect of a policy to
punish juveniles more harshly by transferring them to the adult system. If
services for rehabilitation and reintegration are inferior in adult prisons, spe-
cific deterrence will be less effective.
Second, despite decreases in violent and serious juvenile crime, more
juveniles in Michigan continue to be committed to adult prisons than prior to
the legislative reforms of the last several decades (Shook & Sarri, 2008).
These juveniles include many less serious offenders who will be released
from prison at relatively young ages (Shook & Sarri, 2008). Consequently,
there is a need to consider the access to and quality of services that these
transferred offenders are receiving in the criminal justice system. There is a
need to consider whether these services can be better delivered in the juvenile
system and whether we need to rethink transfer policies and practices.
The budgetary constraints outlined in the beginning of this article might
have affected programs in adult facilities more than juvenile facilities, and
might have contributed to the differential service quality found in this study.
Ng et al. 479
Since 2008, Michigan has experienced more cuts to corrections budgets.
However, as crime rates have also declined, leading to smaller incarcerated
populations in both adult and juvenile facilities, there have been no measur-
able changes in correctional or juvenile services, except in two areas. First, a
decrease in educational programs in prisons has been noted, and this is con-
sistent with the findings in this study. Second, many adult and juvenile pro-
grams (such as health and food services) have been privatized. However,
little is known of service quality after privatization (Evans, 2011; Michigan
House of Representatives Fiscal Agency, 2011).
More research on the service experiences of juveniles is needed. Further
research could compare more and finer dimensions of services, and be devel-
oped into measures of the quality of correctional services received. This
study was interested in services as the outcome. It did not consider whether
services were effective towards end outcomes such as recidivism or
postrelease employment. For example, counseling was queried generally
without differentiating the type of counseling (e.g., CBT or drug related).
Further, although we found no differences in terms of quantity of education,
our data suggests that the types of education were different. Even the delivery
of these different types of education might be different. From the qualitative
components of our interviews, while the juvenile educational programs have
qualified teachers and regular high school curriculum, GED programming in
adult prisons was often simply a self-study program. Vocational training was
also often a result of work orders needed for institutional maintenance.
More research comparing correctional services for juveniles in juvenile and
adult facilities can contribute to filling the knowledge gap of what happens
between an adult sentence and reoffending. Might inferior services in adult pris-
ons be the reason? The findings in this study suggest that it might be the case.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research,
authorship, and/or publication of this article.
The authors disclosed receipt of the following financial support for the research,
authorship, and/or publication of this article: This research was supported by funding
from the W. K. Kellogg Foundation (#007494).
1. Only 4% of juvenile females were sentenced to adult prisons at the time of this
study, so this small number reflects their actual representation in the population.
480 The Prison Journal 92(4)
2. In Kupchik, adult inmates were 1 to 2 years older than juvenile inmates. Hence,
Kupchik and our studies compared adult-placed and juvenile-placed inmates on
different dimensions. While Kupchik maintained comparability by age, this study
compared only youth who would have been considered juvenile at the time of
Abram, K. M., Teplin, L. A., McClelland, G. M., & Dulcan, M. K. (2003). Comorbid
psychiatric disorders in youth in juvenile detention. Archives of General Psychia-
try, 60, 1097-1108.
Allison, P. D. (2009). Missing data. In R. E. Millsap, & A. Maydeu-Olivares (Eds.),
The SAGE handbook of quantitative methods in psychology (pp. 72-89). Thousand
Oaks, CA: Sage.
Anderson, B., & Farrow, J. A. (1998). Incarcerated adolescents in Washington state:
Health services and utilization. Journal of Adolescent Health, 22, 363-367.
Austin, J., Johnson, K. D., & Gregoriou, M. (2000). Juveniles in adult prisons and
jails: A national assessment. Rockville, MD: National Council on Crime and
Delinquency, Institute on Crime, Justice and Corrections.
Bishop, D., & Frazier, C. (2000). Consequences of transfer. In J. Fagan & F. E.
Zimring (Eds.), The Changing Borders of Juvenile Justice: Transfer of Adoles-
cents to the Criminal Court. Chicago: Chicago University Press.
Bishop, D. M., Frazier, C. E., Laza-Kaduce, L., & White, H. G. (1998). Juvenile
Transfers to Criminal Court Study: Phase I Final Report. Washington: Office of
Juvenile Justice and Delinquency Prevention.
Clark-Murphy v. Foreback, 439 F.3d 280 (6th Cir. 2006).
Deitch, M., Barstow, A., Lukens, L., & Reyna, R. (2009). From time out to hard time:
Young children in the adult criminal justice system. Austin, TX: Lyndon B Johnson
School of Public Affairs, University of Texas at Austin.
Dixon, L. (2007). Building safe and healthy communities. Legislative Testimony.
Lansing, MI: Michigan Department of Human Services.
Evans, J. (2011, February 14). Juvenile Justice Services in Michigan. Lansing, MI:
Dept. of Human Services, Bureau of Juvenile Justice.
Fagan, J. (2008). Juvenile crime and criminal justice: Resolving border disputes.
Future of Children, 18(2), 81-118.
Foley, R. M. (2001). Academic characteristics of incarcerated youth and correctional
educational programs: A literature review. Journal of Emotional and Behavioral
Disorders, 9, 248-259.
Forst, M., Fagan, J., & Vivona, T. S. (1989). Youth in prisons and training schools:
Perceptions and consequences of the treatment-custody dichotomy. Juvenile and
Family Court Journal, 40, 1-40.
Ng et al. 481
French, S. A., & Gendreau, P. (2006). Reducing prison misconducts: What works!
Criminal Justice and Behavior, 33, 185-218.
Goebel, K. (2005). Re-entry and corrections education. Retrieved from http://www
Grisso, T., Vincent, G., & Seagrave, D. (2005). Mental health screening and assess-
ment in juvenile justice. New York, NY: Guilford Press.
Hadix v. Caruso, 2009 U.S. Dist. LEXIS 52884 (W.D. Mich. 2009).
Hammett, T. M., Gaiter, J. L., & Crawford, C. (1998). Reaching seriously at-risk
populations: Health interventions in criminal justice settings. Health Education
& Behavior, 25(1), 99-120.
Hartney, C. (2006). Youth under age 18 in the adult criminal justice system (Fact
sheet. Views from the National Council on Crime and Delinquency). Oakland,
CA: National Council on Crime and Delinquency. Retrieved from http://www
Kupchik, A. (2007). The correctional experiences of youth in adult and juvenile prisons.
Justice Quarterly, 24, 247-270.
LaFree, G. (2002). New Mexico—Mixing juveniles with adults. In Violent juvenile
offenders in New Mexico adult and juvenile systems. New Mexico Criminal Jus-
tice Statistical Analysis Center. Retrieved from
Lane, J., Lanza-Kaduce, L., Frazier, C. E., & Bishop, D. M. (2002). Adult versus juve-
nile sanctions: Voices of incarcerated youths. Crime & Delinquency, 48, 431-455.
Lawrence, S., Mears, D. P., Dubin, G., & Travis, J. (2002, May). The practice and
promise of prison programming (Research Report). Washington, DC: Urban
Institute Justice Policy Center.
Lindquist, C. H., & Lindquist, C. A. (1999). Health behind bars: Utilization and evalua-
tion of medical care among jail inmates. Journal of Community Health, 24, 285-303.
MacKenzie, D. L. (2000). Evidence-based corrections: Identifying what works. Crime
& Delinquency, 46, 457-471.
Michigan Department of Corrections. (1998). Annual report. Retrieved from http://
Michigan Department of Corrections. (2005). Annual report. Retrieved from http://
Michigan Department of Corrections. (various years). Statistical report. Retrieved
Michigan House of Representatives Fiscal Agency. (2011). Appropriations for 2011-2012
(pp. 47-72). Lansing, MI: House Fiscal Agency.
Michigan State Police. (2011). Uniform crime reports, crime statistics, offence and
arrest reports. Retrieved from,1607,7-123-1645
482 The Prison Journal 92(4)
Mood, C. (2010). Logistic regression: why we cannot do what we think we can do,
and what we can do about it. European Sociological Review, 26(1), 67-82.
Plata v. Davis, 329 F.3d 1101 (9th Cir. 2003).
Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in
the general population. Applied Psychological Measurement, 1, 385-401.
Redding, R. E. (2003). The effects of adjudicating and sentencing juveniles as adults:
Research and policy implications. Youth Violence and Juvenile Justice, 1, 128-155.
Redding, R. E. (2008). Juvenile transfer laws: An effective deterrent to delinquency.
Juvenile Justice Bulletin, August, 1-12.
Sedlak, A. J., & McPherson, K. S. (2010, April 22). Youth’s needs and services: Find-
ings from the survey of youth in residential placement. Juvenile Justice Bulletin.
Retrieved from
Shook, Jeffrey J., & Sarri.R. (2008). Trends in the commitment of juveniles to adult
prisons: Toward an increased willingness to treat juveniles as adults? The Wayne
Law Review, 54(4), 1725-1765.
Strom, K. J. (2000). Profile of state prisoners under age 18, 1985-97 (Bureau of
Justice Statistics Special Report No. NCJ176989). Washington, DC: U.S. Depart-
ment of Justice. Office of Justice Programs. Retrieved from http://bjs.ojp.usdoj.
Uggen, C. (2000). Work as a turning point in the life course of criminals: A duration
model of age, employment, and recidivism. American Sociological Review, 65,
Zweig, J. M. (2003). Vulnerable youth: Identifying their need for alternative educa-
tional settings. Washington DC: Urban Institute.
Irene Y. H. Ng is assistant professor of social work in the National University of
Singapore and executive editor of Asia Pacific Journal of Social Work and
Development. Her research interests include poverty and inequality, youth crime, and
social welfare policy. She serves on the National Youth Council, Chinese Development
Assistance Council, and the Singapore Family Research Network.
Rosemary C. Sarri is professor emerita of social work and social research at the
Institute of Social Research, University of Michigan. She has authored many books
and journal articles related to social policy, child welfare, and criminal justice. She is
currently studying child welfare careers and the “drift” of child welfare youth to the
justice system.
Jeffrey J. Shook is associate professor of social work and law at the University of
Pittsburgh. His research focuses on intersections of law, policy, and practice in the
Ng et al. 483
lives of children and youth. Published articles and book chapters are on issues related
to juvenile justice and the justice system involvement of youth who age out of the
child welfare system.
Elizabeth Stoffregen is a research assistant in the Center for Political Studies at the
University of Michigan. Prior work includes program evaluation, research, and exten-
sive work in juvenile residential treatment centers in Detroit and Indiana. Her
research interests include juvenile justice, juveniles incarcerated as adults, and the
child welfare system.
... Additionally, juveniles incarcerated in adult correctional facilities are exposed to a socially deprivative environment. During incarceration in an adult facility, juveniles often cannot participate in programming, religious services, and vocational activities, as well as have diminished access to visitation and common areas (e.g., Lane & Lanza-Kaduce, 2018;Ng et al., 2012). Coinciding with major life-course perspectives (Gottfredson & Hirschi, 1990;Moffitt, 1993;Sampson & Laub, 1993), the experiences unique to confinement in adult institutions can alter the behavioral and psychological trajectory of adolescents as they age to adulthood. ...
... In particular, while the prison environment deprives all incarcerated individuals of basic freedoms, juveniles incarcerated in adult correctional facilities are exposed to a variety of social deprivations that do not exist in the general population, juvenile correctional facilities, and the community (Lane & Lanza-Kaduce, 2018). These social deprivations include limited contact with community members and family, limited opportunities to partake in religious programming or vocational training, and limited movement around the correctional facility (Lane & Lanza-Kaduce, 2018;Ng et al., 2012). Exposure to these extreme social deprivations during adolescence might encourage formerly incarcerated juveniles to engage in a variety of behaviors to re-establish peer networks upon reentry (Kopetz et al., 2019;Shapiro et al., 2010). ...
Full-text available
Introduction: Recent trends in reproductive rights have contributed to lasting concerns about adolescent childrearing in American society. Beyond being generally unprepared when raising a child, having a child during adolescence is associated with a variety of environmental, social, and psychological consequences for both the parents and the child. It is important to understand the factors contributing to adolescent childrearing. Although research has identified many factors that contribute to adolescent childrearing, a notable gap remains when considering the role of the correctional system and, in particular, the age-specific effects of confining adolescents in adult correctional facilities. Methods: The current study examined the age-specific effects of time spent in adult correctional facilities from 13 to 34 years of age on childrearing between 14 and 35 years of age using the National Longitudinal Survey of Youth-1997 (NLSY97). The NLSY97 is a nationally representative sample of Males (51%) and Females (49%) born in the United States. Respondents of the NLSY97 were interviewed about life events beginning at age 7 and continued to participate in the study as recently as 2021. Results: The results of the lagged growth curve models suggest that the time spent incarcerated between 13 and 17 years of age heightens the risk of childrearing between 14 and 18 years of age, an effect that is not observed during adulthood. Conclusion: Overall, the results suggest that the conditions adolescents are exposed to during incarceration in an adult correctional facility could contribute to a heightened likelihood of adolescent childrearing.
... Juvenile offenders face a variety of hardships in the criminal justice system, particularly if placed in adult prisons. 1 Compared to youth detained in the juvenile justice system, juvenile offenders processed and sanctioned in the criminal justice system are more insulated from rehabilitation programs (Bishop & Frazier, 2000;Ng et al., 2012;Redding, 2003) and suffer greater developmental disruptions than youth retained in the juvenile justice system (Mulvey & Schubert, 2012). Juveniles placed in the criminal justice system have higher rates of depression (Ng et al., 2011), suicide (Holman & Ziedenberg, 2006;Ruch et al., 2019), risk of victimization (Ahlin & Hummer, 2019;Mulvey & Schubert, 2012), and post-incarceration recidivism compared to those retained in the juvenile justice system and in juvenile correctional facilities (Bishop et al., 1996;Johnson et al., 2011;Lanza-Kaduce et al., 2005;Myers, 2003;Winner et al., 1997;Zane et al., 2016). ...
This study examines patterns of violent institutional misconduct among a cohort of serious juvenile offenders who were incarcerated in state-level juvenile correctional facilities and then state adult prison institutions. Within the cohort, one group of offenders was immediately transferred to the adult prison system following their incarceration in a juvenile correctional facility. The other portion of the cohort was first released to the community, and then later incarcerated in adult prisons. Results of the analysis indicate that being a younger prison inmate with a history of childhood trauma and considered a high-rate violent institutional misconduct perpetrator as a juvenile ward were significant predictors of engaging in violent misconduct in adult prisons. Implications for theory, research, and practice are discussed.
... Another problem of lowering the age of criminal majority may be the negative consequences of accommodating younger offenders in adult correction facilities (Cauffman et al., 2017). Compared to juvenile facilities, there tend to be fewer services available for rehabilitation in adult correctional facilities (Kupchik, 2007;Ng et al., 2012). This may be problematic because the lack of support for reintegration is noted as one of the possible reasons for the increased likelihood of reoffending by juvenile transfer (Redding, 2008). ...
Full-text available
Japanese youth justice has experienced several reforms to date. Currently, a radical revision is under consideration: to lower the age of criminal majority from 20 years to 18 years. Japanese scholars and practitioners have since been engaged in debates over this proposal. Drawing on existing empirical research on youth offending and juvenile justice, the purpose of this paper is to advance a critical analysis on (in)appropriateness of lowering the age of criminal majority. By focusing on its potential consequences, we also discuss what the future of youth justice in Japan would look like. We conclude with offering research implications.
... In a review of the services provided for youth in correctional facilitates, Gagnon and Barber (2010) found that, within correctional facilities, approaches to youth behavior are limited and punitive, calling for new program models that use school-wise positive behavioral interventions and/or cognitive behavior therapy as ways to promote improved outcomes for young people. Placement is also a powerful influence in the experiences of incarcerated youth: young people in juvenile detention facilities receive better counseling and health services than their peers in adult correctional facilities, and report better staff quality in the juvenile detention centers, including educational staff (Ng et al., 2012). Foley and Gao (2002) also note the challenges of delivering quality education programs for incarcerated youth, including the diversity of the student population (accounting for a mostly non-White group of young people identified as having dis/abilities), the lack of attention to educational services as a priority in correctional facilities, and the constant movement of youth between facilities. ...
This dissertation study explores the experiences of incarcerated youth, specifically the relationships between dis/ability (Baglieri& Knopf, 2004), giftedness, and incarceration in the lives of these young people. The extant literature rarely addresses these relationships through the voices of young people who have experienced incarceration first-hand; furthermore, the research pathologizes incarcerated youth and removes the challenges they encounter from their systemically oppressive contexts. This research places the unique experiences of the participants in the larger conversations of deficit-thinking, racism, ecological systems and theories of critical praxisâ??specifically, through the lenses of Dis/ability Studies in Education and Critical Race Theory. As a qualitative research study, this dissertation includes the stories and experiences of three formerly-incarcerated individuals who were held in adult correctional facilities as adolescents, as well as a non-profit leader who is deeply involved in prison abolition and movements that work to improve the lives of inmates. The focus on qualitative methodology and methods works to broaden conversations on juvenile justice and include perspectives on competence, youthâ??s strengths, and resistance to deficit-oriented research. This dissertation argues that we must challenge deficit-oriented models to develop the best pedagogical practices for exceptional students and useful methods for disrupting oppressive systems within educational spaces. Through the centering of participant voices, this study provides implications for understanding dis/ability and the giftedness of young people in exceptional learning environments, the need for close attention to the final stages of the school-to-prison pipeline, and the development of new theories of understanding.
... Yet, foster children aging out of the system can find themselves facing termination of housing, health care, psychosocial and financial support, and educational opportunities (Courtney & Heuring, 2005). Delinquent youth in juvenile facilities may be transferred to adult correctional facilities, where they are at an increased risk of abuse and distress (Levitt, 2010;Ng, Sarri, Shook, & Stoffregen, 2012). Aging out often involves a desperate hunt for life skill trainings and extended medical care, with limited information about how to access or evaluate the quality of services (McConkey, Kelly, Mannan, & Craig, 2010;Neutens, Schwanen, Witlox, & De Maeyer, 2010). ...
Full-text available
This chapter focuses on vulnerable youth, the challenges they face during their transitions to adulthood, and the adverse effects of limited support systems on those transitions. The authors offer recommendations on how adult educators can help facilitate smooth transitions into adulthood for vulnerable youth.
Given contemporary efforts to prevent adolescents from experiencing the negative consequences of incarceration, it is critical to assess the impact of juvenile transfer. Relative to a potential deterrent effect on the recidivism of transferred juveniles, empirical evidence regarding their institutional experience is lacking. Drawing on record data from those admitted to adult and juvenile systems in a Midwestern state from 2011 to 2014, this study compared the correctional experience of teenage males housed in an adult prison with young adults and teenagers housed in juvenile residential facilities. After controlling for individual profiles using the propensity score analysis, youth in adult facilities had similar or more access to institutional programs but also exhibited relatively higher involvement in misconduct based on official reports. The implications for correctional policies and practices for transferred and incarcerated youth are discussed.
This study examines the effects of juvenile transfer laws on youth charged with sex offenses and youth charged with robbery offenses. Using matched samples of youth charged in South Carolina between 1990 and 2001, we examined the effects of transfer on adjudication, incarceration, and recidivism. For youth charged with sex offenses, there were no significant effects of transfer on adjudication or incarceration. Transferred youth were more than 4 times as likely to be convicted of a new person offense as youth adjudicated in juvenile court. For youth charged with robbery offenses, transferred youth were less likely to be adjudicated, but when adjudicated, they were more likely to be incarcerated than those processed as juveniles. Transferred youth were also less likely to be arrested for or convicted of new nonperson offenses and less likely to be arrested for any new offenses. Implications for juvenile transfer policy and future research are discussed.
The experiences of juveniles in adult prisons have gained increased attention over the last several decades. This article adds to understandings of these experiences by examining the relationship between age and prison behavior among transferred juveniles (N = 763). Results reveal that juveniles committed to prisons at younger ages accumulate more misconducts than those committed at older ages. Results also indicate that African American youth, youth with mental health issues, youth with more extensive prior histories in the juvenile system, and youth committed for property and weapons offenses accumulated more prison misconducts. Policy and practice implications are discussed.
This article examines the consequences of legislative changes that eased the process of treating juveniles as adults. Overall, it finds evidence that there has been an increase in the number of juveniles sent to the criminal court and the majority either remain in the community or return to the community shortly after serving a jail or prison sentence. In light of geographic and racial/ethnic disparities in transfer, differences in the programs and services offered to youth in the juvenile and criminal justice systems, and higher rates of recidivism among transferred offenders, the article supports the contention of many scholars and advocates that there is a need for policy and practice reform.
Full-text available
Sociologists have increasingly emphasized "turning points" in explaining behavioral change over the life course. Is work a turning point in the life course of criminal offenders? If criminals are provided with jobs, are they likely to stop committing crimes? Prior research is inconclusive because work effects have been biased by selectivity and obscured by the interaction of age and employment. This study yields more refined estimates by specifying event history models to analyze assignment to, eligibility for, and current participation in a national work experiment for criminal offenders. Age is found to interact with employment to affect the rate of self-reported recidivism: Those aged 27 or older are less likely to report crime and arrest when provided with marginal employment opportunities than when such opportunities are not provided. Among young participants, those in their teens and early twenties, the experimental job treatment had little effect on crime. Work thus appears to be a turning point for older, but not younger, offenders.
Logistic regression estimates do not behave like linear regression estimates in one important respect: They are affected by omitted variables, even when these variables are unrelated to the independent variables in the model. This fact has important implications that have gone largely unnoticed by sociologists. Importantly, we cannot straightforwardly interpret log-odds ratios or odds ratios as effect measures, because they also reflect the degree of unobserved heterogeneity in the model. In addition, we cannot compare log-odds ratios or odds ratios for similar models across groups, samples, or time points, or across models with different independent variables in a sample. This article discusses these problems and possible ways of overcoming them.
This article reports findings from face-to-face interviews with youthful offenders in Florida, about half of whom had been transferred to the adult system and half of whom were retained in the juvenile system. The focus is on the youths’global assessments of the impact of their correctional experiences relevant to subsequent offending. The overall impact of each recalled correctional disposition was rated (ranging from beneficial impact to negative impact). For respondents who had experienced multiple correctional dispositions, comparisons were made about the relative impact of low-end versus deepend juvenile commitments and juvenile versus adult sanctions. Youths believed deep-end juvenile placements were most beneficial. Those programs were viewed as having provided education or life skills. When youths viewed adult sanctions as being beneficial, the benefit was linked to the time and pain of prison confinement. Those youths who attributed positive impact to prison had “skipped” deep-end juvenile placements.
Across the nation, serious and chronic juvenile offenders are increasingly being tried as adults in criminal court and incarcerated in adult correctional facilities. This trend raises important questions for policy makers. To what extent do trials in criminal courts and incarceration in adult prisons promote or inhibit community protection and the accountability and rehabilitation of juvenile offenders? This article discusses the legal consequences of adjudication in criminal court and offers a comprehensive review of research findings on the deterrent effects of transfer laws, conviction and sentencing patterns and recidivism rates in juvenile versus criminal courts, and conditions and programming in juvenile versus adult correctional facilities. The implications of these research findings for juvenile justice policies on adjudicating and sentencing juveniles as adults and directions for future research are discussed.
This article analyzes data from interviews with inmates to examine the correctional experiences of young men incarcerated through criminal (adult) courts in a large Northeastern state. The sample (N = 95) includes respondents from five correctional institutions; some of these inmates have been sentenced to adult department of corrections facilities, and some to juvenile facilities operated by the state's children's services bureau. Relative to the adult facilities, the juvenile facilities are smaller, have much lower inmate-to-staff ratios, and they place greater emphasis (in their official guidelines) on treatment, counseling, education, and mentoring of inmates. As a result, one might expect juvenile-facility inmates to report a relatively more supportive, mentoring-focused style of staff-inmate interactions than adult-facility inmates. Yet surprisingly, inmates in adult facilities report better access to education and treatment/counseling services offered in their facilities.