Sibling Influences

B u l l e t i n S e r i e s

U.S. Department of Justice

Office of Justice Programs

Office of Juvenile Justice and Delinquency Prevention

J. Robert Flores, Administrator April 2003

Sparked by high-profile cases involving children who commit violent crimes, pub- lic concerns regarding child delinquents have escalated. Compared with juveniles whose delinquent behavior begins later in adolescence, child delinquents (offenders younger than age 13) face a greater risk of becoming serious, violent, and chronic juvenile offenders. OJJDP formed the Study Group on Very Young Offenders to examine the prevalence and frequency of offending by children younger than 13. This Study Group identified particular risk and protective factors that are crucial to developing effective early intervention and protection programs for very young offenders.

This Bulletin is part of OJJDP’s Child Delinquency Series, which presents the findings of the Study Group on Very Young Offenders. This series offers the latest information about child delinquency, in- cluding analyses of child delinquency sta- tistics, insights into the origins of very young offending, and descriptions of early intervention programs and approaches that work to prevent the development of delinquent behavior by focusing on risk and protective factors.

Some aspects of children’s behaviors, such as temperament, are established during the first 5 years of life. This foun- dation, coupled with children’s exposure to certain risk and protective factors, influences the likelihood of children becoming delinquent at a young age. However, the identification of these multiple risk and protective factors has proven to be a difficult task. Although no magic solutions exist for preventing or correcting child delinquency, identify- ing risk and protective factors remains essential to developing interventions to prevent child delinquency from escalat- ing into chronic criminality.

According to the Study Group on Very Young Offenders, a group of 39 experts on child delinquency and child psy- chopathology convened by the Office of Juvenile Justice and Delinquency Prevention (OJJDP), risk factors for child delinquency operate in several domains: the individual child, the child’s family, the child’s peer group, the child’s school, the child’s neighbor- hood, and the media. Most profession- als agree that no single risk factor leads a young child to delinquency. Rather,

Preventing children from engaging in delinquent behavior is one of OJJDP’s primary goals. Early inter- vention is crucial to achieving this goal, and understanding the factors related to child delinquency is essen- tial to effective early childhood inter- vention. As part of its effort to under- stand and respond to these needs, OJJDP formed the Study Group on Very Young Offenders.

This Bulletin, part of OJJDP’s Child Delinquency Series, focuses on four types of risk and protective factors: individual, family, peer, and school and community. It is derived from the chapters devoted to these critical areas for prevention and intervention in the Study Group’s final report, Child Delinquents: Development, Intervention, and Service Needs.

To succeed, intervention methods designed to prevent child delinquency from escalating into serious and vio- lent juvenile offending must address a range of risk and protective factors. In addition to the factors addressed in this Bulletin, OJJDP is pursuing research to examine the role of reli- gious traditions and training as pro- tective factors in the life of a child.

Preventing delinquency early in a child’s life can pay significant divi- dends by reducing crime rates and decreasing crime-related expendi- tures of tax dollars. More important, it can help children avoid the conse- quences of delinquent behavior by increasing their chances of leading law-abiding and productive lives.

Risk and Protective Factors of Child Delinquency Gail A. Wasserman, Kate Keenan, Richard E. Tremblay, John D. Coie, Todd I. Herrenkohl, Rolf Loeber, and David Petechuk

Access OJJDP publications online at


the likelihood of early juvenile offending increases as the number of risk factors and risk factor domains increases.

Although some risk factors are common to many child delinquents, the patterns and particular combination of risk fac- tors vary from child to child. Profes- sionals have learned a great deal about which risk and protective factors are relevant for screening and intervention. For example, most professionals agree that early on in a child’s life, the most important risks stem from individual factors (e.g., birth complications, hyper- activity, sensation seeking, temperamen- tal difficulties) and family factors (e.g., parental antisocial or criminal behavior, substance abuse, and poor child-rearing practices). As the child grows older and becomes integrated into society, new risk factors related to peer influences, the school, and the community begin to play a larger role.

Although focusing on risk factors is important, examining protective factors that reduce the risk of delinquency is as important for identifying interven- tions that are likely to work. For exam- ple, some common protective factors against child delinquency and disrup- tive behavior are female gender, proso- cial behavior (such as empathy) during the preschool years, and good cognitive performance (for example, appropriate language development and good aca- demic performance). The proportion of protective factors to risk factors has a significant influence on child delinquen- cy, and protective factors may offset the influence of children’s exposure to mul- tiple risk factors.

This Bulletin is based on four chapters from the Study Group’s final report, Child Delinquents: Development, Inter- vention, and Service Needs (Loeber and Farrington, 2001): “Individual Risk and Protective Factors,” “Family Risk and Protective Factors,” “Peer Factors and Interventions,” and “School and Com- munity Risk Factors and Interventions.”

The risk factors for child delinquency discussed in this Bulletin are categorized into four groups: (1) individual, (2) fami- ly, (3) peer, and (4) school and commu- nity. A greater understanding of these risk and protective factors could serve as the basis for future social policies designed to prevent and control delin- quency (see Burns et al., in press, anoth- er OJJDP Bulletin in this series).

Individual Risk Factors Children’s behavior is the result of genetic, social, and environmental fac- tors. In relation to child delinquency, the Study Group defined individual risk and protective factors as an individual’s genetic, emotional, cognitive, physical, and social characteristics. These fac- tors are frequently interrelated, yet the underlying mechanism of how this occurs is not fully understood.

Antisocial Behavior Early antisocial behavior may be the best predictor of later delinquency. Anti- social behaviors generally include vari- ous forms of oppositional rule violation and aggression, such as theft, physical fighting, and vandalism. In fact, early aggression appears to be the most signif- icant social behavior characteristic to predict delinquent behavior before age 13. In one study, physical aggression in kindergarten was the best and only pre- dictor of later involvement in property crimes (Haapasalo and Tremblay, 1994; Tremblay et al., 1994). In contrast, proso- cial behavior (such as helping, sharing, and cooperation), as rated by teachers, appeared to be a protective factor, specif- ically for those who have risk factors for committing violent and property crimes before age 13.

Studies conducted in Canada, England, New Zealand, Sweden, and the United

Child Delinquency Research: An Overview

Historically, delinquency studies have focused on later adolescence, the time when delinquency usually peaks. This was particularly true in the 1990s, when most re- searchers studied chronic juvenile offenders because they committed a dispropor- tionately large amount of crime. Research conducted during this period by OJJDP’s Study Group on Serious and Violent Juvenile Offenders concluded that youth re- ferred to juvenile court for their first delinquent offense before age 13 are far more likely to become chronic offenders than youth first referred to court at a later age. To better understand the implications of this finding, OJJDP convened the Study Group on Very Young Offenders in 1998. Its charge was to analyze existing data and to address key issues that had not previously been studied in the literature. Consist- ing of 16 primary study group members and 23 coauthors who are experts on child delinquency and psychopathology, the Study Group found evidence that some young children engage in very serious antisocial behavior and that, in some cases, this behavior foreshadows early delinquency. The Study Group also identified sev- eral important risk factors that, when combined, may be related to the onset of early offending. The Study Group report concluded with a review of preventive and reme- dial interventions relevant to child delinquency.

The Child Delinquency Bulletin Series is drawn from the Study Group’s final report, which was completed in 2001 under grant number 95–JD–FX–0018 and subsequent- ly published by Sage Publications as Child Delinquents: Development, Intervention, and Service Needs (edited by Rolf Loeber and David P. Farrington). OJJDP encour- ages parents, educators, and the juvenile justice community to use this information to address the needs of young offenders by planning and implementing more effec- tive interventions.


States have confirmed that early anti- social behavior tends to be the best predictor of early-onset delinquency for boys. For example, in a study by Patterson and colleagues, antisocial behavior was the best predictor of age at first arrest when compared with family social disadvantage, parental monitoring, and parental discipline. Long-term results also indicated that those with an early arrest (before age 13) were most likely to be chronic offenders by age 18 (Patterson, Crosby, and Vuchinich, 1992; Patterson et al., 1998). Likewise, the Cambridge Study in Delinquent Development in London, England, showed that one of the strongest predictors of a conviction between ages 10 and 13 was trouble- some behavior between the ages of 8 and 10, as rated by teachers and peers (Farrington, 1986).

In another study, the two best predic- tors of later antisocial behavior were mothers’ ratings of their children as difficult to manage at 3 years of age and parents’ ratings of behavior problems at 5 years of age (White et al., 1990). Most children whose caregivers perceived them as difficult to manage at age 3 did not become delinquents before age 13. However, most children who became delinquents before age 13 had behavior problems that had emerged in the first years of life.

Emotional Factors Although early aggressive behavior is the most apparent and best predictor of later delinquency, other individual fac- tors may contribute to later antisocial

behaviors. By the end of the third year of life, children can express the entire range of human emotions, including anger, pride, shame, and guilt. Parents, teachers, and even peers affect chil- dren’s socialization of emotional expres- sion and help them learn to manage negative emotions constructively. Thus, how children express emotions, espe- cially anger, early in life may contribute to or reduce their risk for delinquency.

Many studies of delinquency have focused on the concepts of behavioral inhibition and behavioral activation. Behavioral inhibition (in response to a new stimulus or punishment) includes fearfulness, anxiety, timidity, and shyness. Behavioral activation includes novelty and sensation seeking, impulsivity, hyperactivity, and predatory aggression. The Study Group found evidence that high levels of behavioral activation and low levels of behavioral inhibition are risk factors for antisocial behavior. For example, high levels of daring behavior at ages 8–10 predicted convictions and self-reported delinquency before age 21, whereas measures of anxiety and guilt did not (Farrington, 1998). Overall, stud- ies have shown that impulsive, not anx- ious, boys are more likely to commit delinquent acts at 12 to 13 years of age. More studies are needed to determine whether emotional characteristics in childhood are causes of or simply corre- lates of later antisocial behavior.

Cognitive Development Emotional and cognitive development appear to be associated with children’s ability to control social behavior within the first 2 years of life. Evidence sug- gests that these factors play an impor- tant role in the development of early delinquency and may affect the learning of social rules. In addition to traditional measures such as IQ, the Study Group considered cognitive development in terms of language development, social cognition, academic achievement, and neuropsychological function.

Childhood Risk Factors for Child Delinquency and Later Violent Juvenile Offending

The following risk factors are discussed in this Bulletin.

Individual factors

● Early antisocial behavior

● Emotional factors such as high behavioral activation and low behavioral inhibition

● Poor cognitive development

● Low intelligence

● Hyperactivity

Family factors

● Parenting

● Maltreatment

● Family violence

● Divorce

● Parental psychopathology

● Familial antisocial behaviors

● Teenage parenthood

● Family structure

● Large family size

Peer factors

● Association with deviant peers

● Peer rejection

School and community factors

● Failure to bond to school

● Poor academic performance

● Low academic aspirations

● Living in a poor family

● Neighborhood disadvantage

● Disorganized neighborhoods

● Concentration of delinquent peer groups

● Access to weapons

Source: This list is largely based on R. Loeber and D.P. Farrington, eds. 2001. Child Delinquents: Development, Intervention, and Service Needs. Thousand Oaks, CA: Sage Publications, Inc.


Poor cognitive development and behav- ior problems during early childhood could explain the association between academic achievement and delinquency. For example, numerous studies have shown that delinquents’ verbal IQs tend to be lower than their nonverbal IQs (e.g., Moffitt, 1993). Delinquents also have lower mean global IQs and lower school achievement rates compared with nondelinquents (e.g., Fergusson and Horwood, 1995; Maguin and Loeber, 1996).

Mild neuropsychological deficits pres- ent at birth can snowball into serious behavior problems by affecting an infant’s temperament (Moffitt, 1993). These deficits can affect children’s control of behaviors such as language, aggression, oppositional behavior, at- tention, and hyperactivity. Basic cogni- tive deficits may also be associated with impaired social cognitive processes, such as failure to attend to appropriate social cues (e.g., adults’ instructions, peers’ social initiations).

Hyperactivity Studies have shown that restless, squirmy, and fidgety children are more likely to be involved in later delinquent

behavior (e.g., Farrington, Loeber, and Van Kammen, 1990; Lynam, 1997). Clin- ical studies of hyperactive children have shown that they also are at high risk of delinquency (e.g., Loeber et al., 1995). For example, motor restlessness (hyperactive or hyperkinetic behavior), as rated by kindergarten teachers, was a better predictor of delinquency between ages 10 and 13 than lack of prosocial behavior and low anxiety (Tremblay et al., 1994). Another study concluded that hyperactivity leads to delinquency only when it occurs with physical aggression or oppositional behavior (Lahey, McBurnett, and Loeber, 2000).

Family Risk Factors Children and their families defy narrow descriptions. Social, environmental, and family risk factors tend to cluster, and any number of them can occur together within the same family. Understanding the role and influence of each of these factors is a difficult task. For example, early child offending may develop through several pathways. For some children, the primary risk factor may be a family risk factor such as lack of parental super- vision; for others, it may be an individ- ual risk factor such as a diagnosis of attention deficit hyperactivity disorder (Cicchetti and Rogosch, 1996).

The Terrible Twos

The Study Group identified evidence linking behavior problems around age 3 with delinquency by age 13. Antisocial behaviors, such as anger and physical aggression, can appear during the first year of life but often peak at the end of the second year after birth. Thus, before age 3, most children engage in behavior that would be considered antisocial at a later age, including physical aggres- sion. However, most children out- grow early problem behavior. The ones who do not outgrow such behavior are of concern here be- cause of the increased risk that they may become child delinquents.

A Question About Biological Factors

All behavior, including delinquency, is influenced by biological factors. These fac- tors include not only physical strength but also brain functioning, such as neuro- transmitters that pass signals to the brain. Serotonin receptors, for instance, are neurotransmitters that have been associated with impulsive behavior (Goldman, Lappalainen, and Ozaki, 1996). Other biological factors have also been associated with delinquency. Compared to nondelinquents, delinquents tend to have a lower heart rate and a lower skin response (Raine, 1993), which are measures of autonomic nervous activity. Another line of research has concentrated on hormones, including testosterone. However, a high level of testosterone during the elementary school years is not known to predict later delinquency. Currently, research on genes has come as far as the identification of proteins associated with neurotransmitters, but it is unlikely to shed light on complex processes such as delinquency (Rowe, 2002). In summary, it is far from clear to what extent biological processes determine delin- quency at a young age.


Studies have shown that inadequate child-rearing practices, home discord, and child maltreatment are associated with early-onset delinquency (e.g., Derzon and Lipsey, 2000). In addition, the strongest predictors of early-onset violence include family size and parental antisocial history. Early temperamental difficulties in the child coupled with parental deficiencies that interfere with proactive parenting are also thought to be important in the development of early- onset behavior problems.

In looking at the clustering of family risk factors, one goal is to identify which combinations of risk factors promote early misbehavior because, more than likely, early misbehavior is the result of an accumulation of a number of factors. The number of risk factors and stres- sors and the length of exposure to them have a strong impact on child behavior (e.g., Tiet et al., 1998; Williams et al., 1990).

A number of social adversities in fami- lies can affect children’s delinquency. These factors include parenting, mal- treatment, family violence, divorce, parental psychopathology, familial anti- social behaviors, teenage parenthood, family structure, and family size.

Parenting Inadequate parenting practices are among the most powerful predictors of early antisocial behavior (e.g., Hawkins et al., 1998). Compared with families in which the children do not have con- duct problems, families of young chil- dren with conduct problems have been found to be eight times more likely to engage in conflicts involving discipline, to engage in half as many positive inter- actions, and, often unintentionally, to reinforce negative child behavior (Gard- ner, 1987; Patterson and Stouthamer- Loeber, 1984). Three specific parental practices are particularly associated with early conduct problems: (1) a high level of parent-child conflict, (2) poor monitoring, and (3) a low level of posi- tive involvement (Wasserman et al.,

1996). In the Pittsburgh Youth Study, the co-occurrence of low levels of moni- toring and high levels of punishment increased the risk of delinquency in 7- to 13-year-old boys. Conversely, attach- ments to conventional parents and to society’s institutions are hypothesized to protect against developing antisocial behavior (Hirschi, 1969).

Maltreatment Child maltreatment or abuse commonly occurs with other family risk factors associated with early-onset offending. Focusing specifically on the relationship between physical abuse and children’s aggression, one study suggests that 20 percent of abused children become delinquent before reaching adulthood (Lewis, Mallouh, and Webb, 1989). Clearly, most physically abused children do not go on to become antisocial or violent. However, one study that compared chil- dren without a history of abuse or neg- lect with children who had been abused or neglected found that the latter group accrued more juvenile and adult arrests by the age of 25 (Widom, 1989). Abused or neglected children also offended more frequently and began doing so at earlier ages.

Family Violence Each year, approximately 3.3 million children witness physical and verbal spouse abuse (Jaffe, Wolfe, and Wilson, 1990). Witnessing domestic violence has been linked to increased child behavior problems, especially for boys and younger children (Reid and Crisafulli, 1990). Little is known about the age range in which children may be most vulnerable or how long associa- tions persist. In most families, when the woman is battered, children are also battered (McKibben, De Vos, and New- berger, 1989). The co-occurrence of child abuse and witnessing domestic violence affects children’s adjustment more than twice as much as witness- ing domestic violence alone (Hughes, Parkinson, and Vargo, 1989). Other

factors that impose additional risk in violent families include a high incidence of other behavior problems (e.g., alco- hol abuse and incarceration) in male batterers. Maternal psychological dis- tress may also expose children to addi- tional indirect risks, such as the mother being emotionally unavailable to the children (e.g., Zuckerman et al., 1995).

Divorce Compared with boys whose parents remained married, boys whose parents divorced have been found to be more likely to have continuing problems with antisocial, coercive, and noncompliant behaviors through age 10 (Hetherington, 1989). As with many family factors, establishing the exact effects of divorce on children is difficult because of other co-occurring risks, such as the loss of a parent, other related negative life events (e.g., predivorce child behavior problems, family conflict, decrease in family income), and a parent’s subse- quent remarriage. When these related factors are considered, the impact of divorce itself is substantially less.

Parental Psychopathology High rates (as high as 45 percent) of parental antisocial personality disorder have been consistently reported for parents of boys (including preadoles- cents) referred for conduct problems (e.g., Lahey et al., 1988). Similar rates occurred for parental substance abuse and depression (Robins, 1966). Depressed parents show many parenting deficien- cies associated with increased antisocial behaviors in children, such as inconsis- tency, irritability, and lack of supervision (Cummings and Davies, 1994). Parental psychopathology has been linked to increased rates of psychiatric disorder among school-aged children (Costello et al., 1997). The Pittsburgh Youth Study found that the association between delinquency and parental anxiety or depression was stronger in younger than in older children (Loeber et al., 1998).


Familial Antisocial Behaviors A long history of research demonstrates that aggressive behavior and criminal- ity are more prevalent in some families than in others. For example, the Cam- bridge Study in Delinquent Development, which followed 411 families, found that offending was strongly concentrated in a small group of families and that approxi- mately 5 percent of the families account- ed for about half of the juvenile criminal convictions (West and Farrington, 1977).

Antisocial adults tend to select antiso- cial partners (e.g., Farrington, Barnes, and Lambert, 1996). Overall, antisocial parents show increased levels of family conflict, exercise poorer supervision, experience more family breakdown, and direct more hostility toward their chil- dren. In addition, having an antisocial sibling also increases a child’s likeli- hood of antisocial behaviors (e.g., Far- rington, 1995). The influences of siblings are stronger when the siblings are close in age.

Teenage Parenthood Being born to a teenage mother has been found to strongly predict offending in adolescence (Conseur et al., 1997), although much of this effect may stem from the mother’s own antisocial histo- ry and involvement with antisocial part- ners (Rutter, Giller, and Hagell 1998).

Family Structure Many single parents are able to raise their children very well. However, chil- dren from single-mother households are at increased risk for poor behavioral outcome (Pearson et al., 1994; Vaden- Kiernan et al., 1995; McLanahan and Booth, 1989; Sampson, 1987), even con- trolling for the fact that single-mother households on average have fewer eco- nomic resources. Other factors could explain this relationship. Especially as compared with partnered women, sin- gle mothers report more mental health problems (e.g., Guttentag, Salasin, and

Belle, 1980), have higher levels of resi- dential mobility (McLanahan and Booth, 1989; McCormick, Workman-Daniels, and Brooks-Gunn, 1996), and have fewer resources to monitor their children’s activities and whereabouts. Each of these factors on its own contributes to increased levels of early childhood behavior problems.

Family Size The more children in a family, the greater the risk of delinquency. The Cambridge Study found that, compared with boys who had fewer siblings, boys who had four or more siblings by the age of 10 were twice as likely to offend, regardless of the parents’ socioeconomic status (West and Farrington, 1973). These associations may be related to dimin- ished supervision in larger families.

Peer Risk Factors Peer influences on child delinquency usually appear developmentally later than do individual and family influ- ences. Many children entering school, for example, already show aggressive and disruptive behaviors. Two major mechanisms associated with peer fac- tors or influences are association with deviant peers and peer rejection.

Association With Deviant Peers Association with deviant peers is related to increased co-offending and, in a minor- ity of cases, the joining of gangs. Since a 1931 report showing that 80 percent of Chicago juvenile delinquents were arrested with co-offenders, empirical evidence has supported the theory that deviant peer associations contribute to juvenile offending (Shaw and McKay, 1931). The unresolved question is whether deviant peers model and rein- force antisocial behaviors or whether the association with deviant peers is simply another manifestation of a child’s predisposition to delinquency. In other

words, do “birds of a feather flock togeth- er” or does “bad company corrupt”?

The Study Group found that a strong case could be made that deviant peers influ- ence nondelinquent juveniles to become delinquent. For example, according to data from the National Youth Survey on a representative sample of U.S. juveniles ages 11 to 17, the most frequent pattern was a child moving from association with nondelinquent peers to association with slightly deviant peers, and then on to commission of minor offenses. More frequent association with deviant peers and more serious offending followed, leading to the highest level of associa- tion with deviant peers (Elliott and Menard, 1996; Keenan et al., 1995).

Deviant peers influence juveniles who already have some history of delinquent behavior to increase the severity or frequency of their offending. A few stud- ies of children younger than 14 support this hypothesis. For example, in a study of Iowa juveniles, involvement in the juvenile justice system was highest for those who engaged in disruptive behav- ior and associated with deviant peers at a young age (Simons et al., 1994). The Study Group concluded that deviant

Sibling Influences

Based on data from the 1979 National Longitudinal Survey of Youth, a num- ber of publications have underscored the role played by siblings in influenc- ing delinquent behavior in both the domains of family and peer influence. For example, compared with teens with lower rates of offending, teens with high rates of offending were more likely to have siblings who also committed delinquent acts at a high rate. Some studies speculate that older siblings who are prone to delin- quent behavior may reinforce anti- social behavior in a younger sibling, especially when there is a close, warm relationship (Rowe and Gulley, 1992).