Research Proposal

E M P I R I C A L R E S E A R C H

Incarceration in the Household: Academic Outcomes of Adolescents with an Incarcerated Household Member

Emily Bever Nichols • Ann Booker Loper

Received: 14 March 2012 / Accepted: 6 June 2012 / Published online: 20 June 2012

� Springer Science+Business Media, LLC 2012

Abstract The United States has the highest incarceration

rate in the world, yet there is relatively little information on

how the removal of these adults from households impacts

the youth who are left behind. This study used a child-

centered lens to examine the impact of incarceration on the

school outcomes of youth who resided with a family

member or family associate who was incarcerated prior to

the youth’s 18th birthday. We used data from 11 waves of

the National Longitudinal Survey of Youth: Child and

Young Adult (n = 3,338, 53 % female). Initial analyses

indicated that youth who experienced a household mem-

bers’ incarceration evidenced more socioeconomic chal-

lenges, more frequent home adversities, and lower

cognitive skills relative to youth who did not experience a

household members’ incarceration. Results also revealed

that youth who had experienced a household member’s

incarceration were more likely to report extended absence

from school and were less likely to graduate from high

school relative to those youth who did not experience a

household members’ incarceration. Counter to our

hypotheses, results revealed the incarceration of an exten-

ded family member being in the household was the only

relation significantly associated with worse school out-

comes. Plausibly, families who allow non-immediate

criminally involved individuals to reside in the household

are experiencing a more pervasive chaotic home environ-

ment than those with a parent or sibling incarcerated. Our

study suggests that efforts to address the needs of children

with incarcerated parents need to be widened to those who

experience the loss of any household member due to

incarceration.

Keywords Incarceration � Academic outcomes � Household members � At risk youth

Introduction

The United States has the highest incarceration rate in the

world. While researchers have spent the last two decades

beginning to understand the effects that mass incarceration

policies have on the development of prisoners’ children,

they have overlooked the impact that the imprisonment of

any household or family member may have on a young

person’s life. In a recent summary of the current literature

on parental incarceration, Phillips (2010) calls researchers

to transition from a parent-centered perspective to a child-

centered perspective. She argues that by focusing narrowly

on parental incarceration, we have missed the larger impact

that mass incarceration policies have made on children’s

households, beyond the parents, and the resulting devel-

opmental risks that may accompany these transitions

(Phillips 2010). Only one study to date has attempted to

look at how youth’s outcomes are influenced by other

family members’ imprisonment, beyond parental incar-

ceration. Farrington et al. (2001) examined the relationship

between family members’ incarceration and youth crimi-

nality in a sample of male youth in Pittsburgh. They found

that the incarceration of any family member, other than

grandmothers, resulted in a greater likelihood of a boy’s

arrest (O.R. = 2.4–4.7). The findings imply that the

negative effects of incarceration extend beyond the

E. B. Nichols (&) � A. B. Loper Curry Programs in Clinical and School Psychology,

Curry School of Education, University of Virginia,

Charlottesville, VA 22901, USA

e-mail: edb5eb@virginia.edu

A. B. Loper

e-mail: abl2x@virginia.edu

123

J Youth Adolescence (2012) 41:1455–1471

DOI 10.1007/s10964-012-9780-9

parent–child relationship to include associated kin and

need to be further investigated.

Theoretical Perspectives on Household Member

Incarceration

The influence of a household member’s incarceration on

development can be understood best from a developmental

bioecological model (Brofenbrenner and Ceci 1994). This

theory stresses that an individual’s development is influ-

enced by proximal interactions within immediate contexts

such as the home, which originally was referred to as the

microsystem, but later as proximal processes (Brofen-

brenner and Ceci 1994). Essentially, development is not

limited to the mother-child relationship, but is influenced

by any ongoing relationships within the youth’s immediate

context. This would include any adult living within the

household. Additionally, incarceration can influence the

child’s exosystem, which includes the larger social context

such as the family’s poverty, stress, social support and

caregiver characteristics. Previously, the bioecological

developmental model has been cited as a key theoretical

aid in conceptualizing the impact of parental incarceration

on children’s adjustment, and specifically their academic

development (Dallaire et al. 2010; Poehlmann et al. 2010).

However, in order to capture more completely the proximal

processes influenced by incarceration, researchers need to

look at the incarceration of other adult household members

that constitute a child’s family.

The definition of families and caregivers has expanded

in the United States (Brown et al. 2002; Sarkisian et al.

2006), and it is important that research reflects this

expansion. Consistent with the bioecological develop-

mental model, extended family members play essential

roles in the development and support of youth in their

kinship network (Pallock and Lamborn 2006; Sarkisian

et al. 2006; Wilson 1989). Although most extended kin

studies focus on grandmothers, a study by Pallock and

Lamborn (2006) found that higher levels of perceived

support in all extended kinship networks were associated

with higher teacher bonding and stronger work orientation

in both Black and White tenth graders. Expanding the

research to include youth who experience the loss of any

household member to incarceration is not only more the-

oretically valid but also more culturally relevant when

examining the effect that mass incarceration policies has on

the outcomes of adolescence.

Although a bioecological model provides an overarch-

ing structure for understanding the importance of house-

hold members in a youth’s development, there are multiple

theories that explain how incarceration may impact both

the microsystem and the exosystem within the child’s

ecosystem. Criminologists cite four theories to explain the

negative impact of parental incarceration on youth’s

adjustment: strain, attachment, social control, and stigma

(see Hagan and Dinovitzer 1999; Murray and Farrington

2008 for review). These theories also can apply to the

incarceration of any household member. Strain theory

(Hagan and Dinovitzer 1999; Sirin 2005) suggests that the

removal of financial and social opportunities negatively

impacts the youth’s development. Parents are no longer the

sole contributors of social and economic support in most

households, especially if they don’t live with the child.

Most adult members of a household contribute in some

way to the family’s income or social resources, and

therefore their arrest would result in increased financial,

social and emotional strain for the entire household. This

theory will be discussed in further detail in relationship to

school outcomes below.

Second, attachment theory (Poehlmann 2005; Murray

and Murray 2010), suggests that incarceration can disrupt

the youth’s attachment organization, and their belief that

their world is safe and predictable, which is crucial for

healthy emotional development. The arrest and trial pro-

cedure is often a time of extreme distress and uncertainty

for all family members. Secure attachment bonds can be

disrupted by a household member’s incarceration on two

levels. Primarily, if the youth has a significant attachment

to the actual prisoner, then the family member’s arrest and

separation threatens the support the youth receives from

that relationship. In addition, if the youth is not securely

attached to the household member, it is likely that their

primary caregiver does have a significant relationship with

the prisoner (e.g. the arrest of a mother’s boyfriend). The

caregiver’s emotional distress and grief in response to

being separated from the prisoner likely will make them

less responsive to the youth’s needs, impacting the

attachment between the caregiver and child.

Thirdly, social control theory (Hagan and Dinovitzer

1999; Sampson and Laub 1993) suggests that effective

discipline and parental monitoring are necessary to prevent

delinquency in adolescence. It is likely that youth have less

supervision and quality discipline after the incarceration of

a household member due to the indirect effect of increased

strain on their caregiver. The increased financial and social

strain and emotional distress resulting from the removal of

the household member likely increases the amount of stress

on the caregiver, and decreases their focus on parenting.

The parent has less time and energy to devote to providing

proper supervision to ensure that the youth is engaging in

pro-social behavior and avoiding antisocial behavior.

Finally, stigmatization (Murray 2007) suggests that

shame and guilt associated with incarceration is spread to

the prisoner’s family members. This stigma may result in

the youth feeling rejected by school staff and peers and

disconnected from the school environment as a whole.

1456 J Youth Adolescence (2012) 41:1455–1471

123

While stigma is likely greater if a parent is incarcerated,

this does not mean that children do not feel shame and

judgment from peers when someone else from their home

is arrested. Although we are unable to untangle these

complicated theories in this study, we hope to provide

some insight into the possibility that these risks are present

and need to be investigated further.

Adolescents may be impacted uniquely by the incar-

ceration of a household member, as throughout this

developmental period they are experiencing significant

changes in cognitive, social and emotional abilities while

having more opportunities for engaging in risky behaviors

(Schlafer and Poehlmann 2011). The incarceration of a

loved one may disrupt or alter the accomplishment of

important developmental tasks, one of the most significant

being high school graduation. The failure to graduate high

school is a significant indicator for poor life adjustment. It

is related to lower lifetime incomes, and it increases

chances of being unemployed, welfare-dependent, and

incarcerated (NCES 2011). Incarceration in the family may

affect an adolescents’ relationships with peers, teachers, as

well as influence their academic motivation, achievement,

and behaviors (Schlafer and Poehlmann 2011). In the

present study, we use data from a national survey in order

to examine the impact that a household member’s incar-

ceration has on adolescents engaging in academic risk

behavior—specifically extended school absence and failing

to graduate high school—in order to shed more light on this

relatively invisible at-risk population.

Current Research on Children of Incarcerated Parents

Existing research on children with incarcerated parents

gives the best insight into the trends that may be observed

in the larger group of youth losing a household member to

incarceration. As a group, children of incarcerated parents

are at risk for delinquency as well as mental illness (Far-

rington et al. 2001; Murray and Farrington 2008; Phillips

et al. 2002). Generally, the state of the literature on chil-

dren of incarcerated parents provides substantial evidence

for the increased occurrence of negative behavioral out-

comes, but there is still much to learn about the academic

outcomes of these youth. As delinquent behavior is asso-

ciated with higher risks of high school drop out (Battin-

Pearson et al. 2000), it is easy to assume an increase in

academic failure and drop out rates in youth dealing with

parental incarceration; however, such a finding has not

been documented consistently.

Academic Outcomes

Several studies report that children of incarcerated parents

have higher rates of school failure, dropout, and lower

school achievement than the general population (Dallaire

2007a; Johnston 1995; Murray and Farrington 2008;

Stanton 1980; Trice and Brewster 2004). However, this

assertion is largely based on a small sample of studies and

has yet to be confirmed in any general reviews or meta-

analyses (Murray and Farrington 2008; Murray et al. 2012).

Murray and Farrington’s (2008) Cambridge Study in

Delinquent Development followed a sample of males from

1953 to 2008. They compared 23 boys with incarcerated

parents to three control groups: boys with parents incar-

cerated before birth, boys separated for other reasons (e.g.,

hospitalization or death), and those who were not separated

from parents. In their sample, parental incarceration was a

strong predictor for school failure. Specifically, sixty-eight

percent of the boys with incarcerated parents had failed out

of school by age 14, compared to 19 to 33 percent of the

control groups. Trice and Brewster (2004) compared

school failure and drop out rates between 58 adolescents

with incarcerated mothers and their best friends. Similar to

Murray and Farrington, they found that children with

incarcerated mothers were significantly more likely than

their friends to receive a failing grade on a report card and

to drop out of school during adolescence.

Cho (2011) found that adolescents whose mothers were

incarcerated in Cook County prison were at greater risk for

school drop out during the years of their mother’s incar-

ceration, especially if the imprisoned mother lost guard-

ianship in the process. However, Cho (2009a, b) also found

that children’s academic achievement was not impacted

significantly by maternal incarceration when compared to

other at-risk children. Cho examined a significantly larger

sample of over 4,000 children of mothers incarcerated for

one month or more in Cook County prison in Chicago,

Illinois and found that they had significantly lower rates of

grade retention when they were compared to children of

mothers who were jailed for one week or less. Cho’s

explanation for these findings was that youth were at

greatest risk when their home environment was continually

disrupted. Children of jailed mothers also may experience

more frequent disruptions at home, and they may continue

to live in situations where they are being exposed to

criminal activity. Youth with a mother in and out of jail

experience more transitions than those whose mother was

removed more permanently from the home and residing in

a state prison. These studies suggest that the presence or

absence of familial incarceration alone may not be enough

to explain poor school patterns. Rather, the levels, degree

and frequency of transition as well as the chaos that pro-

ceeds and ensues with incarceration may be more predic-

tive of school problems. Consistent with Cho’s results, a

recent meta-analysis by Murray et al. (2012) that included

unpublished manuscripts revealed that, among statistically

rigorous studies that controlled for covariates, there were

J Youth Adolescence (2012) 41:1455–1471 1457

123

no associations between parental incarceration and educa-

tional outcomes when they were compared to similarly

at-risk youth. Thus, there are conflicting findings on the

impact of incarceration on youth’s school experience.

To date, no study has examined the academic outcomes

of youth affected by the incarceration of a family member

beyond a biological parent. However, it is reasonable to

hypothesize that the incarceration of any household mem-

ber will have an effect on the academic outcomes of

children living within the household, based on the proba-

bility of greater economic strain and accumulated contex-

tual risk associated with the population. Although we

previously discussed the additional influence of attach-

ment, social control, and stigma, this study will focus on

how the strain and risk within the youth’s microsystem and

exosystem are associated with risky school behaviors.

Strain

As previously mentioned, household member incarceration

may result in the loss of a household income or an increase

in household expenses as additional dependents enter into a

household (Travis and Waul 2003). The association

between economic strain and academic achievement has

been well documented (Sirin 2005). In an examination of

early school drop out among tenth grade students, low

socio-economic status was correlated positively with

school dropout, over and above its effects on a child’s

academic achievement (Battin-Pearson et al. 2000). There

are multiple explanations for this observed relationship.

Older youth living in poverty may choose to drop out of

school in order to work and help provide for their families.

Increased economic strain has also been associated with

increased distress and authoritarian and/or erratic parenting

styles in caregivers, which in turn may affect children’s

performance and motivation in school settings (Gutman

et al. 2005). Economic strain results in increased respon-

sibility and decreased parental supervision, which increases

the likelihood that adolescents will become disengaged

from the school context. Youth experiencing a household

member’s incarceration also are likely living in economi-

cally strained homes, and may be vulnerable to feeling

disconnected from school.

Accumulated Contextual Risk

Sameroff et al. (1993) suggest that multiple environmental

risks, such as poverty, in a child’s environment can lead to

maladaptive behaviors including delinquency and school

failure. This model of accumulated risk may be one

explanation for the poor adjustment of youth with incar-

cerated parents (Dallaire 2007a; Murray and Farrington

2005), and can be applied to youth with a household

member incarcerated. Families of prisoners frequently

experience high rates of risk prior to incarceration, such as

minority status, low income, drug exposure, crime, and

mental illness (Dallaire 2007a). As minority populations

are overrepresented in the incarcerated population, a dis-

proportionate number of children dealing with a household

member’s incarceration are also likely minorities (Glaze

and Maruschak 2008). In 2007, Black children were seven

and a half times, and Hispanic children were two and a half

times, more likely than White children to have a parent

incarcerated. About 35 % of those in state prisons lived

below the national poverty line before incarceration (Glaze

and Maruschak 2008), which suggests that other members

of the same household would be subject to the same pov-

erty. About 70 % of inmates who lived with their children

prior to incarceration reported a history of substance abuse

and about half (48 %) of the incarcerated parents did

not have their high school diploma or GED (Glaze and

Maruschak 2008). Rates of drug use, mental illness, and

history of abuse are prevalent in the general incarcerated

population. Therefore, if an inmate was living in the

youth’s household, it is likely that the youth and the

remaining household members would be exposed to these

negative influences. In addition, if a parent allows an adult

to live in the household with their children while the adult

is involved actively in crime and/or drug abuse, it is pos-

sible that the house is highly disorganized. Economic strain

and the presence (or absence) of accumulated risk are just

two of the multiple pathways that link having a household

member incarcerated and the academic outcomes of youth.

It is beyond the scope of this study to incorporate all of the

reviewed mechanisms of risk (attachment, social control,

stigma), but we will consider their influence on the rela-

tionship between various outcomes and household incar-

ceration in our analyses and discussion of findings.

The Current Study

The purpose of our study was to evaluate the relationship

between household incarceration and two academic out-

comes (failure to graduate high school and extended school

absence), while controlling for economic strain. First, we

examined whether youth with incarcerated household

members experienced more social and economic adversity

and worse school outcomes than the rest of the sample.

Second, we tested whether household member incarcera-

tion accounted for differences in school outcomes, above

and beyond the measured economic adversities. Finally, we

explored the influence of the specific relation (parent,

sibling, other household member) of the prisoner on the

youth’s school outcomes. Using the National Longitudinal

Survey of Youth, Child and Youth survey (NLSY 2010),

1458 J Youth Adolescence (2012) 41:1455–1471

123

we employed weighted hierarchical logistic regression

analyses to test our hypotheses. We anticipated that

household incarceration would relate to both school out-

comes and that, consistent with previous literature, parental

incarceration would be associated with the greatest levels

of risk compared to sibling incarceration and another

household member’s incarceration.

Method

Participants

Participants included women and their children surveyed

through the National Longitudinal Survey of Youth. In the

initial design, the Department of Labor Statistics used

multi-stage stratified area probability to collect a nationally

representative sample of 12,686 men and women (aged

14–21 in 1979) from dwellings in the US, to follow

throughout their lives, focusing on education, employment,

training, and family experiences. In 1986 the survey was

expanded to a new longitudinal study, which followed the

offspring of female respondents to the initial survey

(NLSY79). This survey was comprised of NLSY admin-

istrator observations, parental reports from the original

female respondents (mothers), as well as self-reports from

their children. Beginning in 1994, a separate Young Adult

survey was created for offspring of female respondents

once they turned 15, in order to biannually collect data

relevant to entering adulthood. For the present study, data

regarding children’s characteristics and behaviors came

from the Young Adult (1995–2008) dataset and the Child

(1986–2008) dataset, while variables on maternal charac-

teristics came from the NLSY 1979 (1979–2008) dataset.

Data from these three sets was merged using the unique

Child and Mother identification codes. Data was obtained

from the National Longitudinal Survey of Youth public

access files (http://www.bls.gov/nls/nlsy79.html).

The sample for the present study consisted of 3,338

offspring of female NLSY79 participants who took part in

the Young Adult survey. This subset of respondents was

selected from an initial pool of 11,495 offspring but 5,298

were not included, either due to dropout (n = 2,881) or

because they were still under the age of 15 (n = 2,417).

Another 2,786 participants were excluded because they did

not answer the question on household incarceration

(n = 2,760), or because they failed to give information

regarding their age at the time of the household member’s

incarceration (n = 26). Participants who were born after

1989 were excluded because they had not had an oppor-

tunity to complete their secondary education, and would

have incomplete information on their academic history

(n = 85). The final sample was 47 % male. The average

age of the sample was 26.5 years (SD = 3.36) at the time

of the 2008 survey. Minorities were overrepresented in the

sample, with about a fifth of the sample self-identified as

Hispanic (22.7 %) and a third identified as Black (38.9 %).

In the overall sample, 14 % reported extended school

absence and 13 % reported failing to graduate high school.

Out of the sample, 585 met criteria for the household

incarceration status, to be compared to 2,753 individuals

who did not experience household incarceration. There are

no national statistics available on the percentage of US

youth with incarcerated household members, therefore, it is

not possible to determine whether our total sample report

of 17 % household member incarceration is consistent with

national trends. However, our sample included a 6.7 %

parental incarceration rate (cumulative across multiple

years in childhood), and was consistent with other pub-

lished studies using national datasets (Hagan and Foster

2012). Of the household member incarceration group, 226

individuals reported only parental incarceration (39 % of

the household incarceration group) 183 reported only sib-

ling incarceration (31 %), and 170 reported only other

household member incarceration (18.6 %). There were 64

participants who reported the incarceration of more than

one type of family member (10.9 %). Finally, 50 of the

participants met the criteria for inclusion in household

incarceration status, but did not report their specific rela-

tionship to the incarcerated household member and,

therefore, were not included in the second set of analyses.

Measures

Household Member’s Incarceration Status

We identified the subsample of youth experiencing a

household member’s incarceration (HMI) by two responses

from either the 2006 or 2008 young adult survey: (1) the

youth responded positively to the question ‘‘Any adult

living with the respondent gone to prison/jail since

respondent was 10 years old?’’. (2) The youth reported

their own age at the time of the incarceration of the HMI as

based upon three separate items: (a) age of respondent

(R) when household member was first imprisoned; (b) most

recently imprisoned; (c) imprisoned. Only youth who

responded to both of these questions were included (see

‘‘Appendix’’ for the exact wording of the survey). Addi-

tionally, the incarceration occurred before the youth’s

eighteenth birthday. We chose 18 years old as the cut off

age because we wanted to ensure that the incarceration

occurred during the youth’s school years and in order to

cast the largest net of youth influenced by incarceration. To

ensure that the inclusion of older teenagers (16–17) would

not change our analyses, we ran parallel analyses to those

presented, but with the exclusion of the older cohort. The

J Youth Adolescence (2012) 41:1455–1471 1459

123

patterns of relationship were not affected and thus we

present data with the full cohort. We indicated household

incarceration with a single, dichotomous variable.

In order to examine patterns associated with particular

household member relationships, we also created three

independent variables that reflected whether or not indi-

viduals indicated experiencing the incarceration of a par-

ent, sibling, or other family relation (e.g. grandparent, aunt,

uncle, cousin, non-relative). Each variable was coded

dichotomously. The three variables were independently

coded, and it was possible for youth to endorse any com-

bination of the three HMI relationships.

Demographic Information

We examined survey information regarding race, sex,

socioeconomic status, and maternal education level as

potential controls for education outcomes. Race was coded

by the survey home interviewer who indicated whether the

youth was or was not Black as well as whether the youth

was or was not Hispanic. We estimated poverty status from

participants’ reports of their family’s participation in fed-

eral aid—either Medicaid or Federal food stamp programs

for more than one year—before the youth’s eighteenth

birthday. Mother’s education was a dichotomous variable,

based on mother reports of high school graduation from the

original NLSY79 dataset.

Cognitive Ability

Cognitive ability was estimated by the Peabody Picture

Vocabulary Test Revised (PPVT-R) intelligence score

(IQ). The PPVT-R was administered to children ages 3–15

to assess the child’s hearing vocabulary of the English

language using 175 vocabulary items of increasing diffi-

culty. The administrator said the item, and the child

selected one of four pictures that best represented the word.

The PPVT-R provided an estimate of verbal ability using

national age-based norms (M = 100, SD = 15). We

included the most recent administration of the PPVT-R

score to have the best estimate of the subjects’ cognitive

abilities entering high school. The PPVT-R is a widely

used measure, which has extensive standardization (Dunn

and Dunn 1981), and a median correlation coefficient of

0.77 across reliability studies (Bochner 1978).

Home Environment Quality

We included two measures of childhood home environ-

ment quality, using scores from the Home Observation

Measurement of the Environment-Short Form (HOME-SF)

(Caldwell and Bradley 2003; Ferron et al. 1994). HOME-

SF is a measure of children’s cognitive stimulation

(HOME-C), emotional support (HOME-E), and overall

quality of their home environment. The HOME-SF consists

of 53 items, based on maternal reports and trained NLSY

interviewer observations. The survey provides an overall

score for home quality, along with subscale scores for

cognitive stimulation and emotional support. Although

there were multiple answers for each survey item, NLSY

recoded the HOME-SF to create dichotomous variables for

each item in the NLSY database, and then created stan-

dardized norms for the survey population. The score per-

centiles were based on internal normalization procedures

from the NLSY sample. We included the most recent

measure of the subject’s emotional and cognitive stan-

dardized scores.

Academic Outcome: Extended Absence

Extended school absence was measured by one question on

the NSLY Young adult survey. Participants who reported

dropping out and returning after one month at least once on

any of the surveys between 1992 and 2008 were coded

positive for the dichotomous variable.

Academic Outcome: Failure to Graduate from HS

Failure to graduate high school was composed of two self-

report questions on the NLSY young adult survey. Subjects

were coded positive for the ‘‘Failure to graduate from HS’’

variable if they did not report receiving a high school

diploma. Subjects who reported earning a GED, but did not

endorse receiving a high school diploma, were coded as

failing to graduate high school. About 18 % (n = 628) of

the sample did not specify whether they received their

GED or high school diploma on the survey. These indi-

viduals were excluded from regression models regarding

high school completion.

Plan of Analyses

Data was analyzed using IBM SPSS Complex Samples 19,

a statistical software program that was designed for com-

plex weighted sample designs. We utilized the custom

weights in accordance with technical materials provided by

NLSY (Center for Human Resource Research 2009).

Lifetime occurrence of extended absence and failure to

graduate high school was measured across all years of data

collection (1994–2008), and corresponding weights were

utilized for all analyses. We first ran a series of Chi-square

and independent t tests to determine whether significant

differences existed between those with and without incar-

cerated household members on key demographic, socio-

economic, and other adversity variables (sex, ethnicity,

poverty status, mothers educational attainment, cognitive

1460 J Youth Adolescence (2012) 41:1455–1471

123

ability, and home environment quality). We treated these

variables as control covariates in all analyses. Three

logistic regression models were run for each school out-

come. The first regression for all of the outcomes included

the respondent’s demographic characteristics (IQ, Ethnic-

ity-Hispanic, Ethnicity-Black, SES, HOME-Cognitive,

HOME-Emotion, Mother education). The household

dichotomous incarceration status variable was then entered

in the second model simultaneously with the control vari-

ables. We conducted a Chi-Square test of differences to

determine whether a household member’s incarceration

status added prediction beyond that offered by the control

variables. We then repeated similar logistic regression

analyses, with parental, sibling, and other household

member’s incarceration status as independent variables

instead of the global household member incarceration sta-

tus variable. We report results of weighted cross-tabs,

it tests, and logistic regression analyses, including logistic

odds ratios and confidence intervals (95 %). All signifi-

cance tests were based on the criterion of p \ 0.05. Odds ratios were interpreted in accordance to Cohen et al.’s

(2010) guidelines for small (OR = 1.46), medium

(OR = 2.49), and large (OR = 4.14) effect sizes when

predicted outcomes are present in 10 % of the general

population.

Results

Preliminary Descriptive Analyses

A summary of the weighted Chi-square and independent

t tests of key descriptive variables is provided in Table 1.

For both academic measures, HMI youth had significantly

worse academic outcomes in comparison to youth who

did not have a household member incarcerated. However,

the HMI group also differed significantly on key demo-

graphic variables that are associated with academic

failure (see Table 2). Specifically, youth with an incar-

cerated household member were disproportionately male

(v2 = 6.64, p \ 0.01), were more likely to be receiving assistance associated with poverty funding (v2 = 57.86, p \ .001), and were more likely to have mothers without high school degrees (v2 = 17.98, p \ .001). They also had significantly lower HOME-C scores (t = 4.77,

p \ .001), HOME-E scores (t = 3.91, p \ .001) and PPVT scores (t = 4.06, p \ .001). In order to understand the specific role of household incarceration, we ran a

series of weighted logistic regression analyses controlling

for IQ, SES, mother’s education and cognitive and emo-

tional quality ratings of the home environment (HOME-C

& HOME-E).

Extended Absence

A series of weighted logistic regression analyses evaluated

the relationship between a household members’ incarcera-

tions with youth’s self-reported extended absence for at least

30 days followed by a return to school. The first weighted

logistic regression evaluated the effects of key demographic

variables (sex, race, IQ, home quality, poverty status and

mother’s educational attainment) on extended school

absence. The combination of covariates was significant for

the model of extended absence (Wald v2 = 139.92, df = 8, p \ .001, Nagelkerke R2 = .087). There was a significant positive relationship between youth self-reported extended

absence and receiving federal aid (Wald v2 = 39.66, df = 1, p \ .001), maternal failure to graduate from high school (Wald v2 = 16.24, df = 1, p \ .001), and a significant negative relationship with HOME-Cognitive standard score

(Wald v2 = 6.41, df = 1, p = .01). A second weighted regression analysis included all of

these previous predictors with the addition of the house-

hold incarceration variable. The addition of the household

member’s incarceration variable significantly improved the

predictive power of the model (v2 diff (1) = 30.46, p = 0.001; Model Wald v2 = 170.20, df = 9, p \ .001, Nagelkerke R2 = 0.10). Independent from the covariates,

there was a significant additional effect for having a

household member incarcerated (Wald v2 = 16.77, df = 1, p \ .001). Youth who experienced at least one household member incarceration had a 1.9 fold increase in the odds of

being absent from school for a period of 30 days or more

and returning, compared to youth who did not have a

household member incarcerated. See Table 2 for results.

A third regression evaluated whether the relation of the

prisoner to the youth (parent, sibling, other) was related

uniquely to extended absence from school. This model

significantly predicted extended school absence (Wald

v2 = 168.93, df = 11, p \ .001, Nagelkerke R2 = 0.10). There was a small, but significant relationship between

extended absence and other family member incarceration

(Wald v2 = 6.13, df = 1, p = .01; OR = 1.87), as well as a trend effect for parental incarceration (Wald v2 = 3.47, df = 1, p = .06; OR = 1.52). Sibling incarceration was

not significant (Wald v2 = 1.44, df = 1, p = .23; OR = 1.42). Results are summarized in Table 3.

Failure to Graduate High School

We ran the same series of logistic regression analyses in

order to examine the relationship between having a

household member incarcerated and the failure to graduate

high school. The first model consisted of the same covar-

iates from the previous regression analyses and signifi-

cantly predicted the failure to graduate high school (Wald

J Youth Adolescence (2012) 41:1455–1471 1461

123

v2 = 154.23, df = 8, p \ .001, Nagelkerke R2 = .113). There was a significant positive relationship between the

failure to graduate high school and sex (male) (Wald

v2 = 13.61, df = 1, p \ .001), maternal high school graduation (Wald v2 = 4.85, df = 1, p = .03), and federal aid participation for more than one year (Wald v2 = 57.03,

Table 1 Weighted sample demographic and school outcome characteristics

Dichotomous variables Household member incarcerated?

Yes (weighted n = 1,313,615) (%) No (weighted n = 6,447,737) (%) v2 (df)

Male 43.0 49.1 7.10 (3,337)*

Ethnicity

African American 30.2 26.4 3.42 (3,337)

Hispanic 12.5 10.4 2.15 (3,337)

Receive federal aid (1 year?) 43.4 24.1 87.73 (3,337)***

Maternal education 75.5 83.6 20.83 (3,337)***

School outcomes

Extended absence 21.2 10.7 47.84 (3,337)***

Failure to graduate high school 24.9 13.0 39.5 (3,337)***

Continuous variables M (SE) t (df)

HOME (cognitive) 93.3 (0.7) 98.0 (0.3) 5.92 (3,308)***

HOME (emotion) 95.2 (0.7) 98.6 (0.3) 4.09 (3,300)***

PPVT (standard score) 94.0 (0.4) 90.7 (0.8) 3.68 (3,279)***

HMI = household member incarcerated. Maternal education = mother endorsed obtaining high school degree or GED. Extended absen-

ce = youth endorsing dropping out of school for a period of 30 days or more, and returning, at least once during school career.

PPVT = Peabody Picture Vocabulary Test; higher scores indicate greater vocabulary skills. HOME (cognitive) = Home Observation Mea- surement of the Environment: Cognitive Stimulation Standardized Score; higher scores indicate greater cognitive stimulation. HOME (emo- tion) = Home Observation Measurement of the Environment Standardized: Emotional Support Standardized Score; higher scores indicate greater emotional support. Value increments for standard errors are as follows: PPVT = 19 points; HOME-E = 16 points; HOME-C = 16

points

* p \ .05; ** p \ .01; *** p \ .001

Table 2 Final logistical regression of NLSY youth demographic characteristics and household member incarceration on self-report of academic outcomes

Extended absence (weighted n = 7,761,352) Failure to graduate (weighted n = 6,438,397) Odds ratio (95 % CI) Odds ratio (95 % CI)

Sex (male) 1.12 (0.87–1.45) 1.65 (1.27–2.14)***

African American 0.77 (0.57–1.04) 0.89 (0.65–1.21)

Hispanic 1.29 (0.95–1.76) 1.02 (0.73–1.44)

PPVT a

1.06 (0.93–1.22) 1.00 (0.87–1.16)

Federal aid 1 year? 2.20 (1.66–2.91)*** 2.84 (2.11–3.79)***

HOME-E a

1.10 (0.99–1.24) 1.12 (0.99–1.28)

HOME-C a

1.17 (1.03–1.33)** 1.34 (1.17–1.54)***

Maternal education 1.74 (1.32–2.28)*** 1.43 (1.04–1.97)*

Household member incarceration 1.87 (1.38–2.52)*** 1.65 (1.20–2.27)**

Pseudo Nagelkerke R2 0.10*** 0.12**

HMI = household member incarcerated. Maternal Education = mother endorsing obtaining high school degree or GED. Extended absen-

ce = youth endorsing dropping out of school for a period of 30 days or more, and returning, at least once during school career.

PPVT = Peabody Picture Vocabulary Test; higher scores indicate greater vocabulary skills. HOME (cognitive) = Home Observation Mea- surement of the Environment: Cognitive Stimulation Standardized Score; higher scores indicate greater cognitive stimulation. HOME (emo- tion) = Home Observation Measurement of the Environment Standardized: Emotional Support Standardized Score; higher scores indicate greater emotional support

* p \ .05; ** p \ .01; *** p \ .001 a

Odds ratios for continuous variables based on the standard deviation of each score within the dataset. Value increments as follows: PPVT = 19

points; HOME-E = 16 points; HOME-C = 16 points

1462 J Youth Adolescence (2012) 41:1455–1471

123

df = 1, p \ .001). There was a significant negative rela- tionship between the failure to graduate high school and the

HOME-Cognitive Standard Score (Wald v2 = 19.34, df = 1, p \ .001). The addition of the household member incarceration variable significantly improved the model (v2

diff = 10.56, p \ 0.01; Model Wald v2 = 164.79, df = 9, p \ 0.001, Nagelkerke R2 = 0.12). Household member incarceration was associated significantly with failing to

drop out from school, above and beyond the contribution of

the control variables (Wald v2 = 9.60, df = 1, p = 0.002). Youth who had a household member incarcerated were

more likely to fail to graduate high school than those

without an incarcerated household member (OR = 1.65).

See Table 2 for details.

As expected, the third model, which included the three

types of familial relationships to incarcerated household

members, significantly predicted self-report of failure to

graduate (Wald v2 = 169.30, df = 11, p \ .001; Nage- lkerke R2 = .121). Having an extended household member

incarcerated was the only relationship significantly asso-

ciated to the failure to graduate high school (Wald

v2 = 6.07, df = 1, p = .01, OR = 1.98), and resulted in an almost twofold increase in the likelihood of failing.

There was a trend effect for having a sibling incarcerated

(Wald v2 = 3.21, p = 0.07, OR = 1.62). See Table 3 for details.

Discussion

There are significant ripple effects to the United States’

mass incarceration policy and one of the most under-

examined is how incarceration in the family is impacting

youth development. While studies have begun to under-

stand the negative impact of parental incarceration, few

have considered that prisoners are also children’s siblings,

uncles or close family friends, and that their removal also

could change the trajectory of a youth’s life. The removal

of any household member changes a youth’s ecological

context (Dallaire et al. 2010; Poehlmann et al. 2010), and

can produce economic strain, break important social bonds,

reduce effective monitoring and social control, and expose

youth to the stigma of having a loved one imprisoned

(Hagan and Dinovitzer 1999; Murray and Farrington

2008). These mechanisms of risk have explained the link

between parental incarceration and maladjustment, but,

given the changing composition of the family, these

mechanisms also should be considered as an explanation

for the risk posed by having any household member

incarceration. Specifically, school outcomes are important

to examine as they have multiple implications on the tra-

jectory of the transition to adulthood (e.g., employment,

substance use, an incarceration). Previous studies have

found that parental incarceration is associated with poor

school outcomes (see Murray et al. 2012, for details), and

therefore it is likely that having any household member

incarcerated could have similar detrimental influences.

To our knowledge, this study is among the first to find

that the incarceration of any household member may

influence youth’s school outcomes. Counter to our

hypotheses, only the incarceration of an extended house-

hold member was associated with extended school absence

and the failure to graduate high school. Neither parental

nor sibling incarceration was associated with significant

variance in the likelihood of either school outcomes. These

findings are consistent with Murray et al.’s (2012) most

recent meta-analysis, in which they suggest that there is not

strong enough support in the current literature to suggest an

association between parental incarceration and school

outcomes in all populations.

Risks Associated with General Household Incarceration

The present study used data from a national, longitudinal

survey to examine the influence of having a household

member incarcerated on educational outcomes, above and

beyond the contribution of social and economic adversity.

Even while controlling for influential variables such as a

Table 3 Logistical regression of parental, sibling, and other household member incarceration on self-report of academic outcomes

Extended absence (weighted n = 7,761,352) Failure to graduate (weighted n = 6,438,397) Odds ratio (95 % CI) Odds ratio (95 % CI)

Parent incarceration 1.51 (0.98–2.34) a

0.80 (0.46–1.39)

Sibling incarceration 1.41 (0.81–2.50) 1.63 (0.96–2.78) a

Other household member incarceration 1.88 (1.14–3.11)** 1.98 (1.15–3.40)**

Pseudo Nagelkerke R 2

0.10*** 0.12***

The same covariates as provided in Table 2 were initially entered in this model. See Table 2 for details on the covariates

* p \ .05; ** p \ .01; *** p \ .001 a

Observed trend effect p \ .10

J Youth Adolescence (2012) 41:1455–1471 1463

123

mother’s educational attainment, home environment qual-

ity, and poverty, youth who had a household member

incarcerated before their 18th birthday had a greater like-

lihood of reporting an extended absence from school (30?

days) and failing to graduate high school than those who

did not report having a household member incarcerated.

This suggests that youth who have any household member

incarcerated may be experiencing similar school related

risks than those with an incarcerated parent (Murray and

Farrington 2008; Stanton 1980; Trice and Brewster 2004).

While our analyses were limited in their ability to test

the mechanisms of risk (social bonding, social control,

stigma, and strain), markers of general economic strain and

contextual risk were measured. Preliminary analyses

revealed that children who grew up with an incarcerated

household member were likely to experience considerably

more adversity then the general population. Specifically,

the household incarceration group reported significantly

higher rates of poverty and lower rates of maternal edu-

cation, home quality, and vocabulary skills. Combined,

these adversities create a picture of an economically

strained household, with myriad interconnected risk fac-

tors. These multiple environmental risks often result in

maladaptive behaviors, such as truancy and drop out, due

to increased strain placed upon an individual during sig-

nificant developmental periods (Sameroff et al. 1993).

Consequently, the present study examined whether these

adverse economic and social variables were sufficient to

explain any observed school-related risk factors among

youth with incarcerated household members, or whether

there was something unique about experiencing a house-

hold member’s incarceration that elevated the risk of this

already vulnerable population.

Our main analyses revealed that household incarceration

had a small but significant relationship with youth’s reports

of extended absence and failure to graduate high school,

above and beyond factors commonly found among incar-

ceration-affected families. Experiencing a household

member’s incarceration elevated the odds ratios of expe-

riencing an extended absence from school (30? days) by

1.9, and likewise increased the risk of failing to graduate by

1.6. These observed odds ratios are consistent, but overall

smaller, with those observed in smaller samples of children

with incarcerated parents (Murray and Farrington 2008;

Trice and Brewster 2004). This suggests that the risks

observed previously in parent-centered research may also

exist to a smaller extent in a larger population of adoles-

cents with any household member incarcerated.

Extended School Absence

When a parent or another household member is incarcer-

ated, it is likely that the youth experiences a decrease in

supervision due to the increased strain on the household.

Consistent with both strain and social control theory, this

would result in the youth acting out, perhaps by failing to

attend school for extended periods of time. Whether their

primary caregiver was incarcerated or left responsible for

dealing with the repercussions of the incarceration, the

youth may feel free to skip school without any conse-

quences. The gap in school attendance also may be due to

the shifting of households that typically occurs after

incarceration. There may be days or weeks in which the

child is unsure where they will live and, as a result, unsure

of which school they should be attending. The uncertainty

associated with living and educational arrangements high-

lights the greater concerns a youth may have regarding how

they will have their basic needs met, threats to their sense

of security, and beliefs that adults in their lives are able to

care for them. The higher rates of extended school absence

among youth with incarcerated household member may

reflect the serious household transition associated with a

family member’s imprisonment and resulting school diffi-

culties may be due to a compromised sense of attachment

and safety.

Failure to Graduate High School

In this study, the percentage of youth failing to graduate

high school in the household incarceration group (23.6 %)

was almost double the national average (12.8 %) during the

span of the survey (NCES 2011). While economic strain

partially accounts for the association between the failure to

graduate high school and household incarceration in the

current study, it did not completely explain the relation-

ship. Future studies should explore the following possible

explanations for this observed association.

Different theoretical perspectives may elucidate further

the association between academic outcomes and the

incarceration of members of one’s household. From an

attachment approach, the association between failure to

graduate high school and household incarceration may be

caused by the relational disruption that results from the

removal of the household member (Murray and Farrington

2008; Murray and Murray 2010). Relational disruption can

lead to later internalizing and externalizing problems and

adjustment problems which can interfere with school

completion (Murray and Farrington 2008; Murray and

Murray 2010). According to social control and modeling

perspectives, youth may imitate the behaviors of their

incarcerated family members, and these behaviors could be

reinforced by receiving approval from valued friends and

family members who are demonstrating similar anti-social

behaviors and values. A youth with an incarcerated older

brother, for example, may emulate and seek to imitate

their brother’s poor academic attitudes. Another possible

1464 J Youth Adolescence (2012) 41:1455–1471

123

explanation for the link between household member

incarceration and failures to graduate high school may be

that social stigmatization of families who experience

household member incarceration may contribute to nega-

tive perceptions of youth by teachers and other students.

Youth affected by incarceration may seem less capable or

potentially more dangerous, leading to less beneficial

attachments with school staff and other students (Dallaire

et al. 2010).

All of these perspectives offer possible explanations for

how household incarceration may indirectly influence high

school graduation rates. It is likely that a combination of

the proposed explanations account for differences in indi-

viduals’ experience in the school system. Given that the

failure to graduate high school is considered a serious

indicator for poor life adjustment (e.g., related to lower

lifetime income, increased chances of being unemployed,

welfare-dependent, and incarcerated) (NCES 2011), future

studies should explore the aforementioned theoretical per-

spectives for the observed association between the incar-

ceration of a household member and higher rates of failing

to graduate high school.

Specific Relation of Incarceration Household Member

Parental Incarceration

Secondary analyses examined the outcomes associated

with specific relationships to the incarcerated household

member (parental incarceration, sibling incarceration, and

extended family member incarnation). We anticipated that,

in accordance with previous studies, parental incarceration

would have the largest impact on youth’s school outcomes.

However, while a trend towards association between

parental incarceration and increased reports of dropout-

return was observed, there was no apparent relationship

between parental incarceration and graduating from high

school. These seemingly inconsistent findings may be

explained when the outcomes are examined individually.

The variable used to capture extended school absence

includes any extended period of time (one month or more)

in which the youth did not attend school, but then returned

to school. The variable may thus capture a possible tran-

sitional period that many children experience after the

parent’s incarceration. Children of incarcerated parents

may experience a change in households if the incarcerated

parent was the primary caregiver (Dallaire 2007a). Such a

disruptive home transition may be less apparent if a sibling

or extended relative is imprisoned. Therefore, the associ-

ation of parental incarceration with extended school

absence and return may be a reflection of a chaotic

household transition associated with the incarcerated par-

ent’s unpredictable presence in the child’s life. As social

services are increasingly involved when a parent in the

household is removed due to imprisonment, it is possible

that they were able to mediate the transition period and

ensure that the children were settled and back in school in a

more timely manner than if the family had to coordinate

the transition on their own.

There was a lack of relationship between parental

incarceration and the failure to graduate high school, which

was inconsistent with previous literature (Murray and

Farrington 2008; Stanton 1980; Trice and Brewster 2004).

It is possible that removal of the parent in some cases may

result in the child moving into a more stable household.

Jaffee et al. (2003) suggest that children adjust best when

they spend as little time as possible with a parent involved

in anti-social behaviors. Furthermore, social services and

school-based supports may provide forms of intervention

for children with incarcerated parents, which also help

buffer from the negative impact of parental incarceration.

Finally, the discrepancy in our findings and the general

literature may be more reflective of differences within the

sample and analytical methods. Consistent with Murray

et al.’s recent meta-analysis (2012), but contrary to the

majority of previous literature, there was a lack of rela-

tionship between parental incarceration and poor school

outcomes (Murray and Farrington 2008; Stanton 1980;

Trice and Brewster 2004). Cho’s studies (2009a, b, 2011),

which examined a larger population of women in Chicago,

had trends similar to the present findings, in that maternal

incarceration did not impact significantly grade retention or

academic achievement. Typically, previous studies had

significantly smaller sample sizes, controlled minimally for

covariates, and were limited to a single geographic area.

The differences observed by Murray and Farrington (2008)

and Trice and Brewster (2004), for example, might have

been unique to the high-risk populations examined, or the

geographic location they chose to study.

It should be noted, however, that selection procedures

used for the NSLY study may have skewed the population

in terms of the present questions regarding parental incar-

ceration. The youth sample was drawn from children of

mothers who had participated in the original NSLY survey.

Solicited survey youth whose own mothers were incar-

cerated may not have participated to the same degree as

other solicited youth, as the key contact (the mother) may

have been unavailable due to her own incarceration or

transition issues particular to incarceration. Previous

research has suggested that incarceration of a mother may

be particularly disadvantageous to a child as it is more

likely to result in transfers of custody and poor attachment

patterns (Dallaire 2007a, b; Novero et al. 2011; Poehlmann

2005). For the current study, a trend effect (p = .07) was

observed for parental incarceration and extended absence

from school. Plausibly, if such un-surveyed high-risk youth

J Youth Adolescence (2012) 41:1455–1471 1465

123

with incarcerated mothers were to be included, a significant

effect may have been present. Although the NLSY survey

afforded the opportunity to examine the issue in a large

national sample, the use of an alternative data set with

more conventional selection procedures would be more

appropriate to examine the effects of parental incarceration

specifically.

Sibling Incarceration

Sibling incarceration was unrelated to either of the exam-

ined outcomes, although there was a trend effect for the

failure to graduate high school. It is important to note that

the survey specifically asked about the imprisonment of an

adult living in the household (See ‘‘Appendix’’for survey

questions), and, as respondents had to report the incarcer-

ation at some point between their 10th and 17th birthdays,

it can be assumed that all of the incarcerated siblings were

older than the respondents. While the incarcerated sibling

may have been contributing financially, it is unlikely that

their incarceration would have resulted in a change in

households and caregivers, and therefore would not have

resulted in an extended school absence. Additionally, the

incarceration of the sibling may actually have been a wake-

up call for both the parent and the respondent. Social

control and modeling theories suggest that delinquent

behavior, such as truancy, occurs when youth do not have

adequate discipline and monitoring (social control), or see

the antisocial behavior as rewarding (social modeling).

Seeing their offspring imprisoned for criminal activity may

motivate the parent to seek out ways to provide more

effective monitoring and discipline for their younger chil-

dren. As the parent’s social control increases, the oppor-

tunities for the younger siblings to skip school would

decrease. The removal of the older criminal sibling also

may remove the sibling’s negative influence, making the

younger siblings less susceptible to model their anti-social

behaviors. The incarceration of a sibling appears to have no

influence on extended absence, possibly because it does not

result in a transition of caregivers, and may make parents

more involved in their remaining children’s daily behav-

iors, such as attending school.

The observed trend effect for sibling incarceration and

increased likelihood for failing to graduate high school

begs explanation. According to economic strain and

accumulated contextual risk perspectives, the youth’s

school failure and their sibling’s incarceration could both

be due to their shared exposure to environmental risks,

limited resources, and negative social influences. Those

with an incarcerated sibling may be at greater risk for long-

term negative academic outcomes in comparison to youth

with a parent incarcerated because it is less likely that the

incarceration of a sibling would result in a transition of

households and caregivers. The youth is left in the same

home environment, which may have contributed to their

sibling’s delinquency. While parents may increase their

monitoring of the youth, such as attending school, general

parenting competency and strain would not change.

Additional services that a youth with an incarcerated parent

may receive, such as social services and school based

prevention programs, are often not provided to youth who

have an incarcerated adult sibling. Therefore, it is possible

that youth who lose a parent to incarceration experience

less financial and social strain than youth who lose a

sibling.

Dallaire et al. (2010) found that teachers were more

likely to rate hypothetical school children as less competent

if they were thought to have parents in prison. This ste-

reotype may be even more generalized to those with

incarcerated siblings, especially if the siblings were pre-

viously in the same school system. Within the school set-

ting, the stigma of having a sibling incarcerated may be

equivalent to or stronger than the stigma of having a parent

incarcerated. Overall, the stigma and stereotypes experi-

enced in school may result in the youth feeling less

attached to school, less motivated to attend, and less likely

to attain a high school degree. Those with incarcerated

siblings do not have higher rates of extended absence

because they become so disconnected from school that they

drop out and do not return. However, it is important to

remember that only a trend was observed, and it is likely

that many youth with incarcerated siblings go on to grad-

uate high school like many of their peers.

Other Household Member Incarceration

Surprisingly, children reporting the incarceration of

‘‘other’’ household members had a greater chance, in

comparison to those reporting sibling or parent incarcera-

tion, of reporting a school dropout-return and a failure to

graduate from high school. Other household members

included aunts, uncles, cousins, grandparents, and non-

related household members such as family friends or

romantic partners. The presence of a criminally involved

extended family member in the home may be an indication

of an unstable home environment. Academic outcomes

may be particularly sensitive to household instability, as

the structures needed to ensure that students regularly

attend school, do homework, and engage in academic work

may be lacking (Dallaire et al. 2010). The parent may not

realize the negative influence that the household member’s

behavior has on the child, or the family may be in such

need of financial or social support that the caregiver

overlooks the consequences to the child’s development.

This would suggest a pre-existing situation in which there

is economic strain and likely inconsistent monitoring and

1466 J Youth Adolescence (2012) 41:1455–1471

123

discipline. Unlike the context for parental incarceration, the

removal of a household member such as an uncle or par-

ent’s romantic partner likely would provide minimal

change in the unstable environment and parenting skills.

Overall, an extended household member’s incarceration

may result in the greatest amount of risk because a child

would remain in the disruptive environment that was

originally open to the addition of delinquent individuals.

Economic strain may also explain the association

between the incarceration of an extended household

member and school outcomes. The incarceration of a

household member there may result in increased financial

strain on the adolescent’s caregivers. The loss of a house-

hold member may also result in the loss of income that was

used to help provide for the children in the home. Unlike

the situation for a family with an incarcerated parent, the

loss of an extended household member rarely qualifies a

family for increased federal aid, social assistance, or school

services. In addition to potential loss of financial support,

the household also may lose some emotional support. The

caregiver must manage their own emotional response to the

imprisonment, making them less emotionally available for

their children and disrupting the parent–child connection.

The incarceration would increase the amount of strain on

the household, both financial and emotional, but without

necessarily removing anti-social influences (Jaffee et al.

2003) or initiating specialized support for dealing with

home disruptions that undermine academic objectives.

Limitations

Our study was limited to providing a broad descriptive

analysis of a previously unexamined group of at risk youth.

The purpose of this study was to determine whether a

household member’s incarceration was associated signifi-

cantly with a youth’s lifetime report of academic outcomes.

Our findings do not assert a causal relationship between a

household member’s incarceration and a student’s exten-

ded absence or failure to graduate high school. The scope

of our study was limited by the nature of the data. The

outcome data used self-reports, and participants may have

minimized or exaggerated their school failures and

achievements. It is also possible that youth were misin-

formed about the nature of a household member’s absence,

and therefore did not correctly report the relation of the

incarcerated individual, or how old they were when the

incarceration occurred.

As participants did not report consistently the earliest

age of incarceration, it was impossible to control for aca-

demic and household data prior to the household member’s

incarceration without drastically reducing the sample size.

We also were unable to determine whether the difficulty in

academics, specifically grade retention and extended

absence, occurred before or after the household member

was incarcerated. This failure in establishing a develop-

mental model with the data is a common limitation in the

literature on children of incarcerated parents (see review in

Murray et al. 2012), due to the nature of the current lon-

gitudinal data on this subject. Additionally, our study did

not differentiate between those incarcerated in prison and

those incarcerated in jail, due to the limitation of the nature

of the dataset. While the length of separation and severity

of the accused criminal activity varies dramatically across

the two populations, we chose to follow the definition of

incarceration as defined by previous researchers on this

topic (Huebner and Gustafson 2007; Murray and Farring-

ton 2008; Murray et al. 2012), which includes any form of

incarceration.

It is important to note that nearly 19 % of the sample did

not respond to the graduate high school item, potentially

undermining the value of the variable for the secondary

analyses regarding specific relationships. Further research,

possibly with other large-scale, nationally representative

samples, may clarify this issue. It would be useful to

determine the efficacy of existing services meant to protect

youth from school failure and whether these interventions

can be applied to youth with the unique risks associated

with having a sibling or another household member in jail.

Another limitation arises from the large time span

(between 1994 and 2008) in which the data was collected.

Consequently, there is up to a 15-year age difference

between the participants, and uncontrolled historical,

political, and cultural differences in the various age cohorts

may have influenced individuals’ school experiences.

However, any differences caused by age were minimized

by the large sample size and representation of age groups.

The selection features of the NLSY study were another

consideration. The participants were only recruited from

the children of mothers who participated in the initial

NLSY 1979 study. If the dataset also included children of

male NLSY 1979 participants, it would have been possible

to conduct a separate and more thoroughly controlled

model for parental characteristics when looking at educa-

tional outcomes. These limitations should be taken into

consideration when interpreting the implications of our

results.

Conclusion

In this study, we observed that youth with incarcerated

household members experience significantly higher rates of

economic strain, adversity, and negative school outcomes

than those without an incarcerated household member.

J Youth Adolescence (2012) 41:1455–1471 1467

123

Those who experience a household member’s incarceration

have negative school outcomes that are predictable above

and beyond their higher rates of economic and social

adversity. One explanation for this finding is that having a

household member incarcerated may be an indication of a

generally chaotic and disruptive home, which is brought

under greater stress due to a household member’s incar-

ceration. Surprisingly, those with an incarcerated non-

immediate household member were the youth who were at

greatest risk for both failing to graduate high school and

extended absences.

This study, to our knowledge, is the first to document

that the incarceration of non-immediate family members is

associated with negative school outcomes. The negative

impact of loss due to incarceration on youth’s development

is not limited to parents, and outcomes may be worse if a

non-immediate family member who is living in the home is

incarcerated than if a parent is incarcerated. There appears

to be a unique experience to living with an extended rel-

ative or unrelated household member who engages in

criminal activities, and is then imprisoned, which

researchers have yet to capture. These findings call for

further investigation into this overlooked population.

Future studies should focus on identifying a national

sample of families at risk for having a household member

incarcerated and following them longitudinally from early

childhood through adulthood, with a special emphasis on

capturing family, community, and school characteristics.

Additionally, it is important to control for individual

behaviors and household characteristics prior to incarcer-

ation, in order to establish a causal relationship between the

separation and the examined outcomes. Researchers also

should explore interactions within the school context,

specifically school characteristics, school connectedness,

student-teacher interactions, and caregiver-teacher inter-

actions to determine mediating influences of school success

or failure. These studies hopefully will provide greater

detail on how the school experiences of youth with an

incarcerated household member differ from those of their

peers, and give directions for more specific interventions.

Recently, there has been increased national focus on

children with incarcerated parents. New services, such as

the Mentoring Children of Prisoners program (Department

of Health and Human Services, 2003–2011), and federal

policy briefs (e.g., Justice Center’s Action Plan for Federal

Policy, 2009) focus on ways to assist children who have an

incarcerated parent. However, these programs only serve

youth dealing with parental incarceration. As demonstrated

by this study, youth with any incarcerated family member

may face considerable household disruption and chaos,

which has lasting effects on their academic success. It thus

appears that services need to widen their scope to include

provision of assistance on youth who have a non-imme-

diate family member living in their homes. Services pro-

vided through government agencies, social services,

community boards, and the school systems need to be

extended to this large and previously overlooked popula-

tion. Schools should make an effort to address these vul-

nerable youth’s academic and behavioral risks through

in-service teacher training, collaboration with community

mental health services, parent/ caregiver support groups,

and the addition of universal prevention programs that will

encourage academic engagement and school connection.

Researchers, policy makers, and service providers need to

broaden their perception of incarceration-related risk, and

address the presence and needs of these often overlooked

casualties of the US’s mass incarceration policies.

Appendix

See Table 4.

Table 4 Description of National Longitudinal Survey of Youth (NLSY) variables

Variables Survey question NLSY survey (YA/child/1979) Years collected

Household incarceration variables

Incarcerated household

member

Since you were 10 years old, has

an adult member of your

household (other than yourself),

that is someone who was living

in the same household as you at

the time, been sent to jail or

prison?

NLSY young adult (YA) 2006 and 2008

Relation to youth How was the person who went to

jail or prison related to you?

(Select all that apply)

NLSY YA 2006 and 2008

1468 J Youth Adolescence (2012) 41:1455–1471

123

Table 4 continued

Variables Survey question NLSY survey (YA/child/1979) Years collected

Parent incarcerated Mother NLSY YA 2006 and 2008

Father NLSY YA 2006 and 2008

Stepmother NLSY YA 2006 and 2008

Stepfather NLSY YA 2006 and 2008

Sibling incarcerated Brother NLSY YA 2006 and 2008

Sister NLSY YA 2006 and 2008

Other household member Grandmother (mothers side) NLSY YA 2006 and 2008

Grandfather (mother’s side) NLSY YA 2006 and 2008

Grandmother (father’s side) NLSY YA 2006 and 2008

Grandfather (mother’s side) NLSY YA 2006 and 2008

Aunt NLSY YA 2006 and 2008

Uncle NLSY YA 2006 and 2008

Cousin NLSY YA 2006 and 2008

Other relative NLSY YA 2006 and 2008

Other non-relative NLSY YA 2006 and 2008

Age at incarceration

(compiled information

from 3 questions)

(1) How old were you the first time

[relationship to R] ([loop

number]) was sent to jail or

prison (while you living were in

the same household)

NLSY YA 2006 and 2008

(2) How old were you the most

recent time [relationship to R]

([loop number]) was sent to jail

or prison (while you living were

in the same household)?

NLSY YA 2006 and 2008

(3) How old were you when

[relationship to R] ([loop

number]) was sent to jail or

prison?

NLSY YA 2006 and 2008

School outcomes

Extended absence Did you drop out of regular school

for at least 1 month and then

return?

NLSY YA 1994–2008

Failure to graduate high

school (compiled from 2

questions; GED = failing

to graduated high school)

(1) Do you have a high school

diploma or have you ever passed

a high school equivalency or

G.E.D test

NLSY YA 1994–2008

(2) Which do you have, a high

school diploma or G.E.D?

NLSY YA 1994–2008

Control variables

Poverty/Gov’t assistance

(compiled from 2

questions; yes to

either = 1)

(1) There is a national program

called Medicaid or Medi-Cal/

Medical Assistance/Welfare/

Medical Services) that pays for

health care for persons in need. Is

your health care now covered by

Medicaid or one of these public

assistance health care programs?

NLSY YA 1994–2008

(2) Have you received government

food stamps in any month since

Year of survey?

NLSY YA 1994–2008

Sex Sex of child (1: male 2: female) NLSY child/YA All waves

J Youth Adolescence (2012) 41:1455–1471 1469

123

References

Battin-Pearson, S., Newcomb, M., Abbott, R., Hill, K., Catalano, R.,

& Hawkins, D. (2000). Predictors of early high school dropout:

A test of five theories. Journal of Educational Psychology, 92(3), 568–582. doi:1O.1037//0O22-O663.92.3.568.

Bochner, S. (1978). Reliability of the Peabody Picture Vocabulary

Test: A review of 32 selected research studies published between

1965 and 1974. Psychology in the Schools, 15(3), 320–327. doi:10.1002/1520-6807(197807)15:3\320:AIDPITS2310150302 [3.0.

Brofenbrenner, U., & Ceci, S. (1994). Nature-nurture reconceptual-

ized in developmental perspective: A bioecological model.

Psychological Review, 101(4), 568–586. Brown, S., Cohon, D., & Wheeler, R. (2002). African American

extended families and kinship care: How relevant is the foster

care model for kinship care? Children and Youth Services Review, 24(2), 53–77. doi:10.1016/SO190-7409(01)00168-2.

Caldwell, B. M., & Bradley, R. H. (2003). Home observation for measurement of the environment: Administration manual. Tempe, AZ: Family & Human Dynamics Research Institute,

Arizona State University.

Center for Human Resource Research. (2009). National longitudinal survey of youth data. Columbus: Ohio State University.

Cho, R. M. (2009a). Impact of maternal imprisonment on children’s

probability of grade retention. Journal of Urban Economics, 65, 11–23. doi:10.1016/j.jue.2008.09.004.

Cho, R. M. (2009b). The impact of maternal imprisonment on

children’s educational achievement: Results from children in

Chicago public school. Journal of Human Resources, 44, 772–797.

Cho, R. M. (2011). Understanding the mechanism behind maternal

imprisonment and adolescent school drop out. Family Relations, 60, 272–289. doi:10.1111/j.1741-3729.2011.00649.x.

Cohen, H., Cohen, P., & Chen, S. (2010). How big is a big odds ratio?

Interpreting the magnitudes of odds ratios in epidemiological

studies. Communication in statistics: Simulation and Computa- tion., 39(4), 860–864. doi:10.1080/03610911003650383.

Dallaire, D. H. (2007a). Children with incarcerated mothers: Devel-

opmental outcomes, special challenges and recommendations.

Journal of Applied Developmental Psychology, 28, 15–24. doi: 10.1016/j.appdev.2006.10.003.

Dallaire, D. H. (2007b). Incarcerated mothers and fathers: A

comparison of risks for children and families. Family Relations, 56(5), 440–453. doi:10.1111/j.1741-3729.2007.00472.x.

Dallaire, D., Ciccone, A., & Wilson, L. (2010). Teachers’ experiences

with and expectations of children with incarcerated parents.

Journal of Applied Developmental Psychology, 31, 281–290. Dunn, L. M., & Dunn, L. M. (1981). Peabody picture vocabulary

manual. Circle Pines, MN: American Guidance Service. Farrington, D., Jolliffe, D., Loeber, R., Southemer-Loeber, M., & Kalb,

L. (2001). The concentration of offenders in families, and family

criminality in the prediction of boys’ delinquency. Journal of Adolescence, 24, 579–596. doi:10.1006/jado.2001.0424.

Ferron, J., Ng’andu, N., & Garrett, P. (1994). Evaluating the

dimensional structure of the home observation for measurement

of the environment-short form. Educational and Psychological Measurement, 54, 535–538.

Table 4 continued

Variables Survey question NLSY survey (YA/child/1979) Years collected

Ethnicity Race of child (Mother’s racial/

ethnic cohort from screener)

1: Hispanic; 2: Black; 3 Non-

Hispanic/NonBlack Recoded into 2 dichotomous variables

NLSY child/YA All waves

Recoded Hispanic Recoded. Hispanic (1) = 1; Black (2) and Non Hispanic/NonBlack

(3) = 0

Recode

Recoded Black Recoded Black (2) = 1; Hispanic (1) and NonHispanic/NonBlack

(3) = 0

Recode

Peabody Picture

Vocabulary Test

Peabody Picture Vocabulary Test-

Revised Form Long (PPVT):

total standard score

NLSY child 1986–2008

Home observation

measurement of the

environment short form:

cognitive stimulation

HOME inventory: cognitive

stimulation standard score

NLSY child 1986–2008

Home observation

measurement of the

environment short form:

emotional support

HOME inventory: emotional

support standard score

NLSY child 1986–2008

Mother’s educational

attainment

Do you have a high school

diploma or have you ever passed

a high school equivalency or

GED test?

NLSY 1979 1979–2008

All survey information taken from the National Longitudinal Survey of Youth. https://www.nlsinfo.org/investigator, April 14, 2012

1470 J Youth Adolescence (2012) 41:1455–1471

123

Glaze, L., & Maruschak, L. (2008). Parents in prison and their minor children. US Department of Justice, Bureau of Justice Statistics, Washington, DC, NJC 222984.

Gutman, L., Mcloyd, V., & Tokoyawa, T. (2005). Financial strain,

neighborhood stress, parenting behaviors, and adolescent adjust-

ment in urban African American families. Journal of Research on Adolescence, 15(4), 425–449. doi:10.1111/j.1532-7795.2005. 00106.x.

Hagan, J., & Dinovitzer, R. (1999). Collateral consequences of

imprisonment for children, communities, and prisoners. Crime and Punishment, 26, 121–162.

Hagan, J., & Foster, H. (2012). Intergenerational educational effects

of mass imprisonment in America. Sociology of Education, 20(10), 1–28.

Huebner, B., & Gustafson, R. (2007). The effect of maternal

incarceration on adult offspring involvement in the criminal

justice system. Journal of Criminal Justice, 35, 283–296. doi: 10.1016/j.jcrimjus.2007.03.005.

Jaffee, S. R., Moffitt, T. E., Caspi, A., & Taylor, A. (2003). Life

with (or without) father: The benefits of living with two

biological parents depend on the father’s antisocial behavior.

Child Development, 74, 109–126. doi:10.1111/1467-8624.t01-1- 00524.

Johnston, D. (1995). Effects of parental incarceration. In K. Gabel &

D. Johnston (Eds.), Children of incarcerated parents (pp. 59–88). New York: Lexington Books.

Murray, J. (2007). The cycle of punishment: Social exclusion of

prisoners and their children. Criminology and Criminal Justice, 7(1), 55–81. doi:10.1177/1748895807072476.

Murray, J., & Farrington, D. (2005). Parental imprisonment: Effects

on boy’s antisocial behavior and delinquency through the life-

course. Journal of Child Psychology and Psychiatry, 46, 1269–1278. doi:10.1111/j.1469-7610.2005.01433.

Murray, J., & Farrington, D. (2008). The effects of parental

imprisonment on children. Crime and Justice, 37, 133–206. Murray, J., Farrington, D., & Sekol, I. (2012). Children’s antisocial

behavior, mental health, drug use, and educational performance

after parental incarceration: A systematic review and meta-

analysis. Psychological Bulletin, 138(2), 175–210. doi:10.1037/ a0026407.

Murray, J., & Murray, L. (2010). Parental incarceration, attachment,

and child psychopathology. Attachment and Human Develop- ment, 12, 289–309. doi:10.1080/14751790903416889.

National Longitudinal Survey of Youth. (2010). National longitudinal survey of youth: Child and young adult dataset [data file]. Retrieved from http://www.bls.gov/nls/nlsy79.html.

Novero, C., Loper, A., & Warren, J. (2011). Second generation

prisoners: Adjustment patterns for inmates with a history of

parental incarceration. Criminal Justice and Behavior, 38(8), 761–778. doi:10.1177/0093854811406637.

Pallock, L., & Lamborn, S. (2006). Beyond parenting practices:

Extended kinship support and the academic adjustment of

African American and European American teens. Journal of Adolescence, 29, 813–828.

Phillips, S. D. (2010). The past as prologue: Parental incarceration,

service planning and intervention development in context. In J. Poehlmann & M. Eddy (Eds.), Children of incarcerated parents (pp. 13–32). Washington, DC: The Urban Institute.

Phillips, S., Burns, B., Wagner, R. H., Kramer, T., & Robbins, J.

(2002). Parental incarceration among adolescents receiving

mental health services. Journal of Child and Family Studies, 11, 385–399. doi:10.1023/A:1020975106679.

Poehlmann, J. (2005). Representations of attachment relationships in

children of incarcerated mothers. Child Development, 76(3), 679–696. doi:10.1111/j.1467-8624.2005.00871.x.

Poehlmann, J., Dallaire, D., Loper, A. B., & Shear, L. D. (2010).

Children’s contact with their incarcerated parents: Research

findings and recommendations. American Psychologist, 65, 575–598. doi:10.1037/a0020279.

Sameroff, A., Seifer, R., Baldwin, A., & Baldwin, A. (1993). Stability

of intelligence from preschool to adolescence: The influence of

social and family risk factors. Child Development, 64, 80–97. Sampson, R., & Laub, J. (1993). Crime in the making pathways and

turning points through life. Boston: Harvard University Press. Sarkisian, N., Gerena, M., & Gerstel, N. (2006). Extended family ties

among Mexicans, Puerto Ricans, and Whites: Superintegration

or disintegration? Family Relations, 55, 331–344. Schlafer, R. J., & Poehlmann, J. (2011). Adolescence in the context of

parental incarceration: Family, school, and community factors.

In J. Poehlmann & M. Eddy (Eds.), Children of incarcerated parents (pp. 13–32). Washington, DC: The Urban Institute.

Sirin, S. R. (2005). Socioeconomic status and academic achievement:

A meta-analytic review of research. Review of Educational Research, 75(3), 417–453.

Stanton, A. (1980). When mothers go to jail. Massachusetts: Lexington Books.

Travis, J., & Waul, M. (2003). Prisoners once removed: The children

and families of prisoners. In J. Travis & M. Waul (Eds.),

Prisoners once removed: The children and families of prisoners (pp. 1–33). Washington, DC: The Urban Institute Press.

Trice, A., & Brewster, J. (2004). The effects of maternal incarceration

on adolescent children. Journal of Police and Criminal Psy- chology, 19(1), 27–35.

US Department of Education, National Center for Education Statis-

tics. (2011). The condition of education 2011 (NCES 2011-033). Retrieved from http://nces.ed.gov/programs/coe/tables/table-sde-

1.asp.

Wilson, M. (1989). Child development in the context of the black

extended family. American Psychologist, 44(2), 380–385.

Author Biographies

Emily Bever Nichols , M. Ed is a doctoral student at the University of Virginia, Curry School of Education’s Program in Clinical and School

Psychology. Her research focuses on youth with incarcerated parents

and family members, with a specific interest in school-related

outcomes.

Ann Booker Loper , Ph. D. is a clinical psychologist and professor at the University of Virginia’s Curry School of Education. She received

her PhD in Psychology from the University of Texas at Austin. Her

research focuses on the mental health and adjustment of prisoners and

their family members, with a particular emphasis upon the parent–

child relationships in justice-involved families.

J Youth Adolescence (2012) 41:1455–1471 1471

123

Copyright of Journal of Youth & Adolescence is the property of Springer Science & Business Media B.V. and

its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's

express written permission. However, users may print, download, or email articles for individual use.