1
High-Risk Families: Probationers and Their Children
Richard Burkhart, Doug Holbrook, and Sandy Pearce
Office of Research and Planning
N. C. Department of Correction
July 2001
Introduction
In North Carolina, on June 30, 2000, 107, 256 adults were under probation
supervision. These probationers live the community; work to pay fines, fees and
restitution; perform community service; submit to drug tests and curfews; attend
substance abuse counseling; and, sometimes, raise children. Research indicates
that the children of probationers are at risk of becoming the criminals of tomorrow.
Research also tells us that probationers with positive family relationships may have a
lower chance of committing subsequent crime.
Adult probation officers’ role is to monitor and enforce the offender’s compliance with
punishment and treatment ordered by the court. Though probationers may have
children, officers pay limited attention to parenting skills or family relationships. Yet
according to research on both adult and juvenile offenders, improving parenting skills
and family relationships may reduce the incidence of juvenile delinquency and adult
crime.
Automated information about the children of probationers or the parents of juvenile
offenders in North Carolina is not readily available. The Department of Correction’s
automated information system, Offender Population Unified System (OPUS),
contains extensive data about offenders under the supervision of the Department,
but minimal family information. The information may be located in individual offender
case files, but since it is not in OPUS, the data cannot be aggregated and analyzed.
The Department of Juvenile Justice and Delinquency Prevention (DJJDP) does not
currently have an automated information system, though one is under development.
Aggregate information on the probation history/status of the parents of juvenile
offenders is not accessible at this time. When a new risk and needs assessment
developed by DJJDP is automated, descriptive information on the criminal
status/history of parents of juvenile offenders will become available.
This lack of information regarding the children of probationers and the probation
history/status of the parents of juvenile offenders is the impetus for this descriptive
study. The children of probationers and juvenile offenders with parents on probation
represent high-risk families. In order to consider pro-family prevention and
intervention programs to reduce the incidence of juvenile delinquency and adult
crime in North Carolina, policy makers need explanatory information on the number
and type of probationers who have children.
2
Literature Review
High-Risk Families
Mumola (2000) indicates that in 1999 an estimated 721,500 (55%) of state prisoners
were parents of minor children. Of these parents, nearly half (48.5%) were on
probation or parole at the time of their arrest. Approximately 24% of the inmates
were on probation and 25% were on parole when they were arrested for the crime
that resulted in the most recent incarceration. No comparable federal data is
available to describe the proportion of probationers with minor children or the
number of juvenile offenders whose parents are or have been on adult probation.
At-Risk Children
Experts suggest various theories to explain why people become delinquents or
criminals, many of them focusing on childhood development and experiences
(Loeber,1982,1990). Researchers have developed profiles of at-risk children in
order to direct prevention efforts where they can have the most impact in deterring
criminal behavior.
According to OJJDP (1999), risk factors increase the chances that a child will
engage in delinquent behavior. Common factors in the at-risk profile are learning
deficits, family conflict, lack of commitment to school, and peers who engage in
problem behavior. Researchers also cite having parents who commit criminal acts
as a factor predisposing children to criminality (McCord 1979). OJJDP (1999) states
that protective factors either reduce the risks of delinquency or change how a child
responds to risks. Protective factors such as close relationships with family,
teachers, other supportive adults and peers and beliefs and standards that promote
school success and rejection of drugs and crime can increase positive behavior and
protect the child from becoming delinquent.
A recent meta-analysis of 66 studies (Lipsey and Derzon, 1998) examined the
strength of a variety of risk factors to predict serious juvenile delinquency. These
factors included antisocial behavior, personal characteristics, family characteristics,
and social factors. Family characteristics include the following: antisocial parents,
abusive parents, broken home, parent-child relations, and family socioeconomic
stratum. The researchers concluded that antisocial parents, which includes criminal
parents, parent psychopathology, and violent parents, is the fifth strongest predictor
of subsequent criminal behavior among children aged 6-11. The strongest four
predictors are criminal behavior prior to age 11, child substance use prior to age 11,
being male, and family socioeconomic stratum. Among older children, ages 12 to
14, social ties and antisocial peers were more predictive than family characteristics.
The same study indicates that, in addition to having criminal parents, other family
factors do contribute to juvenile delinquency. The studies included in the meta-analysis
found significant associations for parent-child relations and abusive parents.
Parent-child relations include the following: poor parent-child relationships, punitive
or mixed discipline practices, low parent involvement, low supervision, low warmth,
negative attitude to child, poor parental practices, and severity in child training.
3
Hawkins et al. (1998) conducted a meta-analysis of longitudinal studies that
identified predictors of youth violence. They describe results of specific studies that
examined salient predictors of youth violence at various developmental stages. The
researchers list five studies which conclude that parental criminality is a predictor of
youth violence. They catalog seven studies that correlate poor family management
(e.g. lax or punitive discipline, poor or authoritarian supervision style) to violence in
youth. Three studies associate low parent-child involvement and interaction with
youth violence and seven studies relate poor family bonding to subsequent violence
in youth.
Capaldi and Patterson (1997) found that parental failure to set clear expectations for
children’s behavior, poor parental monitoring and supervision of children, and
excessively severe and inconsistent parental discipline of children are predictors of
later substance abuse and delinquency. Karen Heimer (1997) suggests that
parenting practices may, in fact, be a crucial factor in the strong relationship between
violent delinquency and family socioeconomic stratum (SES). Her research
indicated that higher social class parents provided more supervision for their children
than lower social class parents, and that supervision is a key factor in reducing
violent delinquency. Heimer refers to other studies that found that lower social class
parents use more coercive child-rearing methods than their higher social class
counterparts. Jobs in the lower economic stratum tend to emphasize obedience to
authority rather than independence, and workers may generalize these feelings to
their parenting. Thus, they use less moral reasoning and more physical punishment,
threats, and restriction of privileges. These parenting behaviors are associated with
subsequent violent delinquency.
Adult Offender Risk Reduction
Andrews (1996) states that until the mid -1970’s, rehabilitation was the goal of adult
correctional supervision. However, due to a narrow interpretation of the now famous
Martinson research, rehabilitation fell into disrepute and was replaced by the goal of
public protection, and was implemented by tougher and longer sentences. In recent
years, a more balanced approach of rehabilitation and public protection has
emerged, built on the concepts of identifying risks and needs, and reducing the risk
of recidivism through addressing criminogenic (dynamic risk factors) needs in a
responsive manner. Andrews concludes the following in regard to effective
correctional interventions:
 Official punishment without the introduction of correctional
treatment services does not work;
 Providing correctional treatment services that are inconsistent with
the principles of risk, need, and responsivity does not work.
 What works is the delivery of clinically and psychologically
appropriate correctional treatment service, under a variety of
setting conditions that may be established by the criminal sanction.
 The delivery of appropriate correctional treatment service is
dependent upon assessments that are sensitive to risk, need and
responsivity.
4
Andrews, Zinger, Hoge, Bonta, Gendreau and Cullen (1990) conducted a meta-analysis
of outcome research on effective correctional interventions. The analysis
showed a 43%-58% reduction in recidivism in the comparison group (no correctional
intervention) to the treatment group (provided correctional intervention). Andrews
(1996) indicates that Lipsey’s 1990 meta-analysis documented treatment variables
associated with reduced recidivism. One of Lipsey’s findings was that effective
treatment attends to extra-personal circumstances, including the family. Based on
his research and the research of others, Andrews (1996) suggests that familial
issues are one of several dynamic needs that can be addressed to reduce risk in
adult offenders.
Andrews (1996) suggests a host of promising targets for changing offenders. These
targets include changing antisocial attitudes, promoting familial affection and
communication, promoting identification/association with anti-criminal role models,
leaning pro-social skills, learning to identify and avoid risky situations, and reducing
chemical dependencies and substance abuse.
Based on Andrews’ research, Gendreau developed the Level of Service Inventory-R
(LSI-R), a quantitative survey of attributes of offenders and their situations relevant
to the decision regarding level of service. The LSI-R is composed of 54 items based
on a response of either satisfactory (needs no improvement), relatively satisfactory
(needs some improvement), or very unsatisfactory (strong need for improvement).
The survey includes questions about the adult offender’s family and marital situation
– whether it is rewarding or not, and whether there are other family members
involved in crime.
According to Palmer (1996), in general, behavioral, cognitive-behavioral, life skills,
and family intervention programs are most effective. Family intervention with
behavioral programming is correlated with reduced recidivism. Gendreau (1996)
identifies a variety of programs that are effective correctional interventions based on
research data. These include family systems therapy programs for juveniles and
cognitive behavior programs for adult offenders. Effective programs include
intensive services that provide positive behavioral reinforcers.
This growing body of research promotes public protection and rehabilitation by
addressing crime-producing risks and needs of adult offenders. Researchers agree
that a family relationship is one of the dynamic risks in an offender’s life that, if
appropriately addressed, may reduce recidivism.
Parent- and Family-Focused Interventions
James Q. Wilson (1998) states that there are several recurring themes in the
research on serious and violent delinquency. These themes are:
 The precursors of crime occur very early in the lives of many
serious offenders.
 These precursors involve a complex array of relationships between
child and parents.
5
 Changing those precursors requires a complex intervention that
affects parents as well as children and that best occurs early in the
life of the youngsters.
He advocates trying policies and programs which rebuild weak families as a
plausible antidote to the problem of serious delinquency.
Loeber and Farrington (1998) state that delinquency can be significantly reduced by
prevention. Their studies indicate that by using different program models,
delinquent behavior can be averted at every point through childhood; this finding
provided the underlying theme of the research, and title of the report, Never Too
Early, Never Too Late. Based on literature reviews, Loeber and Farrington (1998)
state that effective intervention programs include interpersonal skills training,
behavioral contracting, cognitive-behavioral treatment, or teaching family homes
programs.
Several researchers (Kazdin 1987; Webster-Stratton 1984) promote parent
management training as a method to change the family environment so that children
are positively reinforced for their pro-social behaviors and are ignored or punished
for anti-social behaviors. Parents participate individually or in groups in a clinical or
school setting. Wasserman and Miller (1998) describe such programs as
opportunities to learn and practice parenting strategies such as communicating
clear expectations about positive and negative behavior in children, identifying
positive and negative behavior in children, identifying antecedents of problem
behavior, and providing positive and negative consequences. Evaluations of
parenting management training by various researchers (Kazdin, 1987; Webster-
Stratton, 1991) indicate substantial changes in parent and child behavior. However,
Webster-Stratton (1991) found in a follow-up study of families receiving this training
that 25%-40% of children continue to have significant behavior problems after the
intervention.
The Office of Juvenile Justice and Delinquency Prevention (1995) identified a
number of interrelated social problems that contribute to delinquency – substance
abuse, child abuse and neglect, family violence, transience, gun availability, gangs,
uneducated and undereducated children and youth, teen parents, latchkey children
and poor parenting. Kumpfer and Alvarado (1998) note that many of these social
problems are connected to the weakness of families to care for their children. They
prepared a bulletin for OJJDP that examines effective family-strengthening
intervention programs. The three family intervention strategies effective in reducing
risk factors and increasing protective factors are behavioral parent training, family
skills training, and family therapy. There was not sufficient research to include
parent education characterized by didactic, knowledge-only approaches or affect-based
parent training as an effective intervention. Examples of behavioral parent
training include Parents and Children Training Series (video-based parenting
program), Treatment Foster Care Program (for foster parents), and Helping the
Noncompliant Child (behavior training for parents which includes time for practice
with their children under supervision). Examples of family skills training include
Kumpfer’s 14-session Strengthening Families Program, the 7-session Iowa
6
Strengthening Families Program, the 33-session Focus on Families, the 14-session
Nurturing Program, Families and Schools Together, and Family Effectiveness
Training. According to Kumpfer and Alvarado (1998), family-focused programs are
more effective than programs that focus solely on the child or the parents.
Walker (1999), in the Journal of the American Probation and Parole Association
Magazine (Perspectives), discusses the craft of prevention as it relates to a
probation officer’s role. He describes key principles of prevention programs and
provides details on the award-winning Parents as Teachers program, an early
childhood family education and support program designed to help parents improve
their children’s emotional, social, intellectual, language and physical development.
Walker summarizes the research on the effectiveness of the program. Findings
indicate advances in the children’s language, problem solving and social
development, and higher scores on standardized math, reading and achievement
tests. Parents were more knowledgeable about child development and more
confident in their parenting skills. Child aggression was reduced among boys from
single parent homes, and these families showed a lower incidence of child abuse
and welfare dependency.
These findings have policy and program implications for probation agencies. They
suggest that interventions that improve parents’ child rearing skills are beneficial in
terms of reducing violence and other criminality in children of probationers. Although
criminal justice involvement of parents is a static factor in the life of a child, parental
criminality can be reduced through proper supervision and effective rehabilitation.
Targeting probationers with children for parenting management training is one
method of targeting at-risk children for intervention services.
The Present Research
This research is the precursor to developing policies and services targeted at high
risk families – probation families. The purpose of this research is to describe the
proportion of adult probationers who have minor children and to identify
characteristics of the children and parents, based on available data in the N. C.
Department of Correction’s automated offender system (OPUS). In this study,
probation officers are asked to document basic demographic data about the children
of a sample of probationers in the Department’s Offender Population Unified System
(OPUS). With this data, the Office of Research and Planning can produce profiles of
the proportion of probationers who have children, and can analyze the age
distribution of the children.
Methodology
The primary activity of the High-Risk Families grant is to produce a descriptive data
set of adult probationers, on probation on August 1, 2000, who have minor children.
Staff of the Office of Research and Planning selected a sample of adult probationers
to determine whether they have children, and if so, those children’s demographics.
Staff chose a sample size of 3,000 offenders, half male and half female. Females
were oversampled relative to their representation in the overall probation population
7
(about 20%) because the researchers believe that female offenders are more likely
to be living in the same household as their children.
The Department’s information system, OPUS, has a screen (the Relative Reference
Screen) that allows entry of family information, but it has not been routinely used as
a part of intake for probation cases. It can capture the following information about
the offender’s family members:
- Type of relationship (son, daughter, brother, sister, etc.)
- Relative’s age
- Relative’s name
- Relative’s social security number
- Relative’s county of residence
- Quality of the relationship
- Whether the relative was a victim of the offender’s crime
- Relative’s criminal history (yes, no, or unknown), and
- Whether the offender will reside with the relative after release from prison.
Research and Planning staff used the Probation Client Profile in OPUS to randomly
select offenders for the study. Only active offenders, on probation on August 1,
2000, were chosen; absconders and offenders supervised out of state were not
included in the sampling frame. The selected offenders were sorted by supervising
officer, and lists of offenders were sent to Chief Probation and Parole Officers.
Research and Planning staff asked the Chiefs to distribute the list to their officers,
and instruct the officers to determine for each offender on the list whether they had
minor children, and to enter the information about those children into the Relative
Reference Screen. The Director of the Division of Community Corrections sent a
memorandum to the Chief Probation and Parole Officers explaining the intent and
importance of the study.
The list of offenders was distributed by mail to Chiefs on August 10, 2000. Officers
were asked to enter the information by October 1, 2000. Research staff monitored
the entry of data and followed up to determine the response rate. The Relative
Reference Screen does not allow the officer to identify offenders with no children,
therefore there is no way for the researchers to tell, for a given offender, if he/she
has no children or if the officer did not enter the requested information. To improve
the quality of the data, Research staff asked officers to return the list of offenders to
the Office of Research and Planning when they completed the data entry. Based on
this process, ninety-two percent of the officers responded to the request for
information. Research staff asked probation officers how they obtained the
information to enter into OPUS; they reported basing the entry on information in
existing case files, asking the offender for the information, or using their previous
knowledge of the offender.
Research staff made an important observation during this data collection effort - the
transitory nature of the probation population. Though the data entry period was
relatively short (just over six weeks) many of the officers responded on their list that
offenders had absconded, or been transferred to another county or had been
revoked in the one-week period between the selection of the sample and the day the
lists arrived by courier at the probation offices. This may be the result of a data entry
8
lag, as well as the transitory population. Research staff deliberately selected a
larger sample than was required for statistical significance to allow for some
population transience, so this did not pose serious problems in conducting analysis.
Research staff also intended to seek ways to merge the information collected in this
process with data available from the Department of Juvenile Justice and
Delinquency Prevention. By identifying offenders with children, and conversely
identifying adjudicated juvenile delinquents whose parents had a criminal history, an
important programming link could be forged. As of this writing, the DJJDP does not
have an automated system that would enable these linkages. Nevertheless, staff
from the DOC and DJJDP have met and have established a working relationship to
assure that data will be shared as it becomes available.
Analysis
Children
The most critical piece of information gleaned from the data set is the raw number of
children with parents on probation. Based on this data collection effort, 35.33% of
the sample group had at least one child. This number is lower than the expected
proportion; several probation field staff had estimated during the course of the
project that 66-75% of the probation population had children. Nevertheless, applying
this proportion to the overall probation population (107,256 on June 30, 2000) would
suggest that 37,894 probationers in North Carolina have children. Twenty-one
percent of the population in the sample had multiple children. This finding suggests
that there may be over 50,000 children in the state whose parents are convicted
offenders receiving probation supervision in the community.
As several studies make clear, the risk factors that pre-dispose children to
delinquency are different depending on the child’s age. Because the sample of
probationers was drawn across all ages, including probationers as old as 77, 23% of
the children identified in this study were actually over the age of 18. Children 18 and
under totaled 1,653 in the sample group. The average age of the children in the
sample is 12.77 years. In the interest of developing program strategies to divert
children from delinquency, it is critical to examine the ages of these children. A
breakdown of the children of probationers by age is displayed in the following table.
Table 1: Children of Adult Probationers in NC on June 30,
2000
Age of Children Percent
0-5 24.21
6-10 24.86
11-15 20.19
16+ 30.74
Almost 25% of the children included in this study are five and under, and another
25% are between six and 10. Almost 70% are under 16, the age at which North
Carolina deems people as adults for purposes of criminal court proceedings.
9
Among the children, 48.33% are white and 48.47% are black. Sixty-one percent of
the children were females, and males accounted for 39 percent; the sample itself
was chosen so that males and females each accounted for 50% of the group. Other
demographic factors tend to be less clear in OPUS. For instance, the information
system provides a variable for the offenders’ socioeconomic level, but this is not a
mandatory field for data entry, and thus may not be a robust variable. For the
offenders who had data in this field, the majority (39%) were categorized as “Low,”
but less than 25% of the offenders had data in the field. Eighty-seven percent of the
children in the sample were recorded as having a “Good” relationship with their
parent on probation. However, because this field is not mandatory, and not explicitly
defined in policy, this finding is not especially meaningful. The finding that the
children were predominantly female is inexplicable. It may suggest that program
resources should be targeted at girls, but most research indicates that boys are at
higher risk for delinquency than girls.
Parents
The Department’s information system has a wealth of information about the offender.
The sample was pulled in order to oversample women, who were felt to be more
likely to have children in their custody; the sample was half male, half female. The
sample was fairly evenly split along racial lines, with 47.67% of the offenders
selected in the sample being white, and 48.33% black. In the overall sample, the
most common crime types were drug-related (19%), breaking/entering/larceny
(17%), and Driving while Impaired (16%). Of the men in the sample, 28% reported
having children, while 42% of the women had children in their household.
All probationers receive a risk assessment when they begin their sentence. This
instrument, the Offender Trait Inventory or OTI, contains important information about
the offender’s criminal history, financial situation, marital status, and substance
abuse history. The OTI has been validated for the North Carolina probation
population to predict likelihood of revocation, either for subsequent crime or technical
reasons. In either case, it is a predictor of compliance with probation. A copy of the
instrument itself is attached to this report. Each item on the OTI is discussed with
respect to parents in the study below.
Current or Past Convictions
The OTI assigns points based on convictions for certain types of crimes that are
highly correlated with subsequent criminality. Current or prior convictions of Driving
while Impaired receive one point. Breaking and Entering, Burglary, Larceny,
Robbery, and related crimes receive two points, and Forgery, Worthless Checks,
and Fraud receive three points. Thirty-eight percent of the parents in the sample
had no points on this item, although other information does indicate that Larceny is
the most common offense, either current or prior, among the sample (8%). There
were also a large number of drug-related offenses in the sample. The findings for
parents do not differ significantly from the probationers in the sample who did not
have children.
Financial Situation
Financial Situation is self-reported information and determines whether offenders
can support themselves and handle their own finances. Thirty-two percent of the
10
parents identified themselves as self-sufficient, and 52% classified themselves as
having “No Known Difficulties.” Only 16% were designated by their officer as having
“Severe Difficulties in Meeting Obligations.” This pattern was essentially the same
for non-parents.
Marital Status
Marital Status of the offender is correlated with success on probation, and may be
related to the offender’s overall family involvement. Thirty-seven percent of the
parents in the study reported being married or widowed, and the same proportion
(37%) reported being single. Twenty-five percent were divorced or separated.
Based on this finding, about 62% of the parents may be raising children alone. Sixty
percent of the non-parents were single, and 17% were divorced or separated.
Attitude
The Probation Officer has an opportunity to make a determination of the offender’s
attitude, or motivation to change. This item on the OTI is highly subjective, and has
two values: Motivated to Change and Negative. The officer deemed ninety-two
percent of the parents in the sample as Motivated to Change. The findings are the
same for non-parents.
Past History of Drug Addiction
The Probation Officer records whether an offender has a history of addiction, based
on the offender’s self-report; the offender receives five points if this item is true.
Twenty-one percent of the parents claimed to have battled drug addiction, while 22%
of the non-parents said they had an addiction history.
Employed less than 7 months out of last 12
The offender receives four risk points if he was employed for less than seven months
in the preceding year. Sixty-six percent of the parents had maintained employment
for seven months or more during the preceding year, which differs only slightly from
the non-parents (63%).
Unemployed/Unstable Employment
Employment stability is a good predictor of risk, because offenders with stable jobs
are less likely to engage in ongoing criminal behavior. Seventy-two percent of the
parents in the study have stable employment, as do the probationers in the non-parent
group.
High School Dropout
Having finished high school makes a difference in likelihood of revocation; high
school graduates typically have an easier time finding jobs. Sixty-five percent of the
parents in the sample completed high school. Among the non-parents, 62% had
graduated.
Conclusions
From this research, the Department of Correction can conclude that there are a
sufficient number of probationers with children to merit serious attention in terms of
program offerings and policy development. Given the scope of the issue, the
11
Division of Community Corrections can take several approaches to help probationers
become better parents and to prevent children from becoming delinquents.
This research project has been a first step for the agency in increasing awareness
among probation officers about the family circumstances of their probationers.
Officers have become familiar with using the OPUS screen necessary to capture
information about the children of the offender, and the MIS section has made this
screen a routine component of initial case setup. As a result, the base of family
information available about the probation population will continually grow and
improve.
The effort to increase officers’ awareness of children in probationer households will
expand with policy development and planning. The Office of Research and
Planning will work closely with the administration of the Division of Community
Corrections to promote training probation officers to be more aware of the children in
their probationers’ homes. There may be federal funding to implement this type of
training, as well as to develop program models for offering parenting skills training to
probationers. The Office of Research and Planning will seek funding to develop in-service
training for officers in identifying at-risk behaviors in probationers’ children,
so that the officer can discuss these issues with the probationer and work with them
to get the child’s needs addressed.
Based on the finding that almost one third of the probation population has children,
there is a critical mass of probationers who need parenting classes, and whose
children may need prevention services. Based on this research, the Office of
Research and Planning will promote the concept of parenting interventions for
probationers since there is evidence that improving offenders’ parenting skills can
reduce their children’s likelihood of becoming criminally involved later. If funding is
made available, pilot sites could be set up across the state to experiment with
delivering parenting skills interventions.
In the future, different program curricula could be used in the four sites to determine
if different curricula have different effects with our probation population. To the
extent possible, these programs should meet standards of research-proven effective
interventions. Implementation could be accomplished through existing community
resources already working with probationers, such as Day Reporting Centers.
Based on the findings of this study, it would be beneficial to build a collaborative
relationship with the Department of Juvenile Justice and Delinquency Prevention
(DJJDP). The DJJDP works with delinquent juveniles and also provides prevention
services across the state. By partnering with DJJDP, the Department of Correction
can identify at-risk children in offender families and engage them in prevention
programs offered by the juvenile justice system. Furthermore, the two departments
can share information about the offenders in their care, identifying families where
both parent and child are court-involved. There is currently no way to determine if
adult probationers have children who have been adjudicated delinquent, or if juvenile
delinquents’ parents are currently on adult probation. During the course of this
research project, DOC staff has met with DJJDP staff to determine if there are ways
of sharing information, and have secured agreement from the juvenile justice staff
12
that, as their information system is developed, efforts will be made to assure that
information about parents, including criminal justice status, will be captured for
delinquent juveniles.