Judging Juvenile Responsibility: A Social Ecological Perspective

Mark Fondacaro & Lauren Fasig. Handbook of Children, Culture, and Violence. Editor: Nancy E Dowd, Dorothy G Singer, Robin Fretwell Wilson. Sage Publication. 2008.

A separate system of juvenile justice emerged around the turn of the last century to provide child offenders with more humane treatment than they were receiving in the adult criminal justice system (Howell, 1997). Prior to this reform, the general assumption was that young children and early adolescents had not reached their full “developmental potential” for mature decision making. Accordingly, children under age 14 were generally considered too young to be held morally and legally accountable for their actions, whereas youngsters over age 14 were presumed to be competent, mature, and accountable (Tanenhaus, 2000). Thus, before the advent of the juvenile justice system, minors typically were exempted from criminal responsibility if they were under age 7, held fully accountable for their crimes if they were over age 14, and presumed exempted from criminal responsibility between the ages of 7 and 14, a presumption subject to rebuttal by contrary evidence (McCarty & Carr, 1980).

The influx of new immigrants at the beginning of the 20th century generated new pressures to extend the duration of adolescence so that these youngsters would not be exploited as a source of cheap labor and to provide them with some additional time to prepare for the assumption of adult roles and responsibilities (Grossberg, 2002). As youngsters in their later teens were increasingly seen as “works in progress,” the dividing line between “youthful immaturity” and “adult maturity” was nudged forward to late adolescence or early adulthood by juvenile justice reformers (Scott & Grisso, 1997).

The juvenile justice system was designed to assist youth who had gone astray to return to the “right path.” There was a belief that children and adolescents could be “helped” or reformed during their formative years in a way that would keep them from developing into career criminals (Howell, 1997). Blameworthiness was not at issue; a child’s need for rehabilitation drove policy. This push toward rehabilitation in juvenile justice was born in optimism. Shifting away from a focus on criminal behavior and what the child did wrong, the goal instead was to focus on the personality and identity of the child (Mack, 1909). The hope was that if character flaws were caught early, they could be “mended” or “fixed.” Presumably, rehabilitated youngsters would become better people and make morally sound choices in the future, ones that would keep them out of trouble with the law. The social sciences and mental health professions were enlisted to help develop and implement rehabilitative programs and interventions.

After nearly three-quarters of a century under this system, characterized by limited judicial oversight and little if any success at rehabilitation, observers of the juvenile justice system concluded that “nothing works” to rehabilitate juvenile offenders. (Martinson, 1974). Such pessimism among social scientists set the stage for the “due process” reforms in the juvenile justice system (In re Gault, 1967; In re Winship, 1970; Kent v. United States, 1966). Paradoxically, the push for more procedural safeguards, championed by those interested in children’s rights, ultimately led to an erosion of distinctions between the adult and juvenile justice systems. At the same time, the public blamed lax treatment in the juvenile justice system for the danger from younger and younger children who were committing more adultlike crimes, but were escaping adult sanctions and being prematurely released back into society. Ensuing calls for punishment were rooted in the view that treatment was not working, that young criminals were being coddled, and that our juvenile justice system was actually encouraging and exacerbating youth crime by failing to impose justified punishment. There was the perception that the deterrent effects of the adult system were not being utilized by the juvenile justice system. Moreover, the approach to juvenile crime was viewed as contributing to moral failure among our youth, by failing to instill a sense of personal responsibility (Wilson & Herrnstein, 1985).

In response to this critique, the juvenile justice system has become more punitive over the past few decades. This policy shift has coincided with and indeed encouraged efforts to ensure that there are procedural safeguards in juvenile justice similar to those constitutionally mandated in adult criminal trials. The current trend in juvenile justice reform is toward adopting both the procedures and substantive policy goals of the adult criminal justice system (Ainsworth, 1991; Feld, 1999). An apparent rationale for this position is the assumption that due process and rehabilitation are somehow incompatible—they cannot be promoted simultaneously. Once social scientists began to conclude that rehabilitation “didn’t work,” legal reformers and the courts began to push for what were regarded as optimal procedural safeguards—adult criminal procedural safeguards, which were considered to be synonymous with due process itself (Fondacaro, 2001).

Public debate regarding considerations of due process and critiques of the efficacy of the juvenile courts continued to grow while a new concern surfaced in the 1990s: the perception that severe adolescent crime was on the rise (Redding & Howell, 2000). Graphic media accounts of homicides and other serious offenses, school shootings, and dire predictions of the emergence of a new class of juvenile super-predators fueled changes in juvenile justice policy. Concerns about the leniency and ineffectiveness of traditional juvenile courts brought about increases in waivers and transfers to adult criminal courts and increased acceptance of incapacitation and confinement of juveniles in the juvenile justice system in an effort to take youthful offenders off the streets and promote community safety (Grisso & Schwartz, 2000).

This shift in juvenile justice policy toward punishment and confinement versus rehabilitation led behavioral scientists to consider the role of human development in juvenile criminal activity. In an effort to protect children from severe consequences such as the death penalty and lengthy incarcerations, these researchers drew on basic scientific findings about cognitive development to argue that youth, especially those younger than age 15, generally lack the social cognitive skills and mature judgment to be held fully accountable under a culpability-based adult system rooted in a retributive rather than a rehabilitative justice philosophy. However, the behavioral scientists and legal scholars who advocated this position did not challenge the fundamental premise of the adult criminal justice system: that the legal consequences for crime should be based largely on a determination of the offender’s “mental state” at the time of the crime. Instead, they argued that children and younger adolescents lacked full psychological capacity to form adult-like mental states. Their primary focus, however, remained on the internal, psychological functioning or “mental state” of the juvenile offender.

In this chapter, we will review behavioral science research suggesting that this relatively narrow focus on internal, psychological functioning as the basis for determining legal sanctions is incomplete and misguided, especially for juveniles. We will argue that there are both legal and empirical justifications for broadening the focus to include contextual influences on juvenile crime including family, peer, school, neighborhood, and media influences. We begin by reviewing the traditional model for judging criminal responsibility and the similarities and differences between the juvenile and adult criminal justice systems. We conclude by proposing and considering the implications of a more contextual, social ecological analysis of juvenile responsibility.

The Traditional Model for Judging Criminal Culpability

Although the adult and juvenile justice systems are still distinct in at least some respects, there is one sense in which both are, and always have been, similar: their underlying conception of human nature and behavior. Both are rooted in liberal philosophy regarding the capacity of individuals for free choice. In fact, it is this presumption of individual free choice that serves as the basis for holding people morally blameworthy for their intentional conduct, justifying a system of retribution and punishment. Although the criminal and juvenile justice systems do not consider children to have the same capacity for free choice as adults, juveniles are considered to be “works in progress,” on their way to developing a mature capacity for free choice. Thus, our legal system presumes the capacity for free choice in adults and presumes that children gradually develop this capacity as they mature.

In the adult criminal justice system, the severity of criminal sanctions turns largely on a determination of the defendant’s mental state at the time of the event. The law generally presumes that people make conscious choices about whether to engage in illegal behavior and it is the capacity, if not the exercise, of choice that provides the moral justification for retribution-based punishment. Whether or not a person has a “guilty mind” is generally based on a social judgment of the degree to which the person consciously chose to engage in criminal behavior or to ignore a substantial risk. Jurors and judges in a criminal proceeding are asked to engage in retrospective mind reading. However, the weight of behavioral science research contradicts the presumption that people act based on the exercise of consciously motivated rational and “free choice.” Rather, human behavior is a function of both personal and situational influences (Bronfenbrenner, 1979; Moos, 1973).

As a matter of common sense and experience, people, including judges and jurors, attempt to understand human nature and to predict behavior based largely on judgments about a person’s psychological attributes or personality characteristics. Here again, there is a very well-established body of behavioral science research that indicates that people tend to overestimate the importance of personal factors in guiding and directing behavior, a phenomenon referred to in the social psychological literature as the “fundamental attribution error” (Nisbett & Ross, 1980). When people are confronted by the unusual behavior of others, they tend to attribute that behavior to an aspect of the person’s psychological disposition rather than objective aspects of the situation facing the person or the person’s subjective appraisal of the situation (Nisbett & Ross, 1980). This fundamental attribution error is most pronounced when we are judging the behavior of others— criminal defendants, for example. In our adult criminal justice system, the assignment of the highest degree of criminal culpability is based on a retrospective and potentially erroneous social judgment that the person acted with intent, deliberation, and purpose.

Thus, the presumption of the human capacity for “rational choice,” which provides the moral and legal justification for retribution-based punishment, is arguably flawed. Moreover, the retribution-based model of criminal justice does nothing to advance the well-being of the offender. It serves, at best, to satisfy the public’s quest for retribution, revenge, and indirectly, incapacitad on, by providing an apparent justification for incarcerating offenders.

Duty to Juveniles

Some may argue that society owes no greater duty to adult offenders than retribution-inspired criminal justice objectives. Even if one advocates the value of retributive justice principles (which we do not), there is still a strong argument that society and the state have duties to child offenders beyond those reflected in retributive criminal justice policy objectives. The state has some duty to educate and protect children (Meyer v. Nebraska, 1923; Pierce v. Society of Sisters, 1925; Prince v. Massachusetts, 1944). Moreover, in those cases where punishment is sought, there is some obligation to ensure that the punishment has a positive effect on the child, rather than serving only as an emotional release for the punisher. Historically, a juvenile justice system separate from the adult criminal system has been justified via the state’s parens patriae power (In re Gault, 1967; Schall v. Martin, 1984). Parens patriae is the principle that the state must care for those who cannot care for themselves. It is the basis for many of the Supreme Court’s rulings designed to promote the welfare of children (Tanenhaus, 2000). The state’s authority to promote public welfare and safety under its police power also has served as a basis for protecting children’s rights (Prince v. Massachusetts, 1944). Overall, the state’s special duties to children suggest that the state must do more for children who break the law than simply “make them pay.”

If juvenile justice policy must do more than feed the public’s appetite for punishment, then our efforts to understand and address juvenile crime must do more than assign blame based on inadequate assumptions about the causes and consequences of human behavior. We need to draw on theory and research in the behavioral sciences that is aimed at educating, socializing, protecting, and effectively punishing children who break the law. Theory and research advanced by ecologically oriented behavioral scientists who study the influence of social contexts on human behavior hold considerable promise in advancing this wider range of juvenile justice policy objectives.

Behavioral Science Research on the Determinants of Human Behavior

Conceiving of adult human beings as autonomous decision makers is grossly out of line with most of the systematic behavioral science research conducted over the past century. Early in the 20th century, social psychologists such as Kurt Lewin (1936) suggested that in order to fully understand human behavior, one needs to know something about personal characteristics and the social environment in which the individual is embedded. This ecological model of human behavior is also reflected in the work of interpersonally oriented psychiatrists such as Harry Stack Sullivan (1953) as well as behavioral scientists such as Rudolf Moos (1973) and Urie Bronfenbrenner (1979). Moreover, a long line of experimental studies clearly demonstrates that human behavior is strongly influenced by social and contextual factors (Ross &Nisbett, 1991).

Although we may have a natural tendency to view behavior as the result of free choice and illegal behavior in particular as the result of bad moral choices, the overwhelming weight of evidence suggests that contextual factors have a strong influence on the nature and meaning of human behavior. Under most circumstances, any human behavior, including that which is judged to be illegal, is influenced as much or more by contextual factors as by aspects of personal choice (Park, 1999). Unfortunately, people have difficulty understanding the importance of contextual influences on behavior, particularly when they are evaluating the behavior of others rather than their own conduct (Nisbett & Ross, 1980).

This is particularly salient in the area of juvenile justice reform because there is some evidence that children and adolescents are even more influenced by contextual factors than are adults (Woolard, Fondacaro, & Slobogin, 2001). A contextual understanding of human behavior clearly has important implications for juvenile justice reform. To the extent that behavior is seen as the result of both personal and environmental influences, it becomes questionable whether a moral judgment model of criminal responsibility is just, moral, or developmentally sound. For example, if the justification for punishment and retribution is based on the assumption that individuals freely choose to engage in illegal behavior, and this assumption is false, or only partly true, then the legitimacy, fairness, and morality of a backward-looking, punishment-oriented juvenile justice system is called into question.

Treating and Preventing Delinquent Behavior

The inaugural and guiding philosophy of the juvenile justice system was to shift the focus toward who the child was as a person rather than what the child did that was wrong. This misguided focus and orientation to rehabilitation contributed in part to early failures to demonstrate that treatment interventions could be effective with juvenile offenders. Clinicians and case workers focused on trying to change the youngster’s character or intrapsychic structure rather than focusing on changing behavior per se, or the situational and environmental factors that contributed to and maintained delinquent conduct. This approach was rooted in psychodynamic principles that have since been shown to be relatively ineffective in promoting behavior change, particularly among children who are acting out rather than holding in their anger and frustrations (Davison, Neale, & Knng, 2004).

The intrapsychic focus tied to individual psychotherapy proved to be a dismal failure with juveniles for several reasons. To begin with, this orientation was first developed as an approach for understanding and treating neuroses rather than conduct problems. In fact, psychodynamic therapy may not be particularly effective even with the target populations for which it was designed— young, attractive, verbal, intelligent, and successful or “YAVIS” clients. Many such clients tend to get better without professional help (Davison et al., 2004). Even if psychodynamic therapies were effective with such clients, the typical delinquent child does not fit the YAVIS profile, particularly with regard to verbal competence and the tendency toward introspection (Wrightsman, Greene, Nietzel, & Fortune, 2002). Moreover, to the extent that all human behavior is a function of personal and environmental factors, an intervention approach that focuses only on the personal side of the equation is doomed to fail. Even in those cases where some youngsters have been able to benefit from insight-oriented therapy while in institutional settings, these youngsters tend to recidivate once they return to the same social environments that contributed to the development of their initial delinquent conduct (Wiebush, McNulty, & Le, 2000). Therefore, it is no surprise that initial efforts to evaluate the efficacy of treatment and preventive interventions with juvenile offenders uniformly concluded that nothing seemed to help. Typically, recidivism rates were near 70% (Borduin, Mann, Cone, & Henggeler, 1995).

As ecological models of human development gained ascendancy in psychology and interventions began to focus on modifying maladaptive behaviors rather than restructuring personality, the treatment and prevention literature started to take on a more optimistic tone. In fact, intervention strategies have been improved to the point that modern multisystemic interventions aimed at risk management have reduced recidivism risk from around 70% to slightly above 20% (Borduin et al., 1995). In sum, a fair appraisal of “state-of-the-art intervention strategies” suggests that ecologically oriented, cognitive-behavioral interventions aimed at the multiple life contexts in which juveniles exist (family, peer, school, neighborhood) can be both clinically and cost-effective. Given the availability of some promising intervention strategies, the next task is to ensure that “state-of-the-art” interventions are adequately implemented and incorporated into the structure of juvenile justice administration (Soler, 1994). Along these lines, multisystemic interventions lend themselves nicely to a risk management model, where the emphasis is on changing dynamic risk factors in various life domains that are associated with delinquent behavior.

Behavioral Science Research with Adolescents

As suggested, much of the early research on human behavior focused primarily on intrapersonal characteristics such as psychological attributes and personality factors. This approach advanced our understanding only so far, usually accounting for only about 10% of the variance in human behavior (Mischel, 1968). Increasingly, behavioral scientists began to recognize, and study, the importance of situational factors and the joint and interactive influences of both intrapersonal and social factors on human behavior. This shift in perspective resulted not only in explanatory models that accounted for a greater share of human behavior, but also more effective intervention strategies. Traditional models of criminal justice, however, dovetail with the emphasis on exclusively intrapersonal explanations of human behavior. In fact, many behavioral scientists continue to do valuable work that addresses primarily if not exclusively intrapersonal aspects of behavior, while recognizing that they are accounting for only a small percentage of variance in human behavior— an awareness that is lacking for most legal scholars and laypersons.

Many of the researchers seeking to inform analyses of juvenile responsibility have concentrated on intrapersonal aspects of adolescent decision making. This focus has been more or less dictated by the legal system’s emphasis on the “guilty mind” of the presumed rational actor in choosing to engage in criminal conduct. With the recent shift in juvenile justice policy toward this “evildoer” theory of crime, the degree to which a child’s criminal actions are seen as the result of a rational and deliberate choice to do wrong has become the central factor in judging juvenile responsibility (Ainsworth, 1991; Gnsso, 1996; Steinberg, 2003). Notwithstanding this focus on atomistic and intrapsychic explanations for complex human behavior, behavioral science research with this focus is relevant to discussions about why juvenile justice should maintain its distinct focus on rehabilitation and crime prevention rather than retribution.

Reviews of empirical research and developmental theory suggest, for example, that adolescent decision making may differ from adult decision making in ways that are relevant to judgments of responsibility. Adolescents, especially those under 15 years of age, have been shown to have a foreshortened time perspective, a greater proclivity for risk behavior, changing estimates of risk likelihood, a greater propensity to be influenced by peers, reduced social responsibility or “stake in life” (Woolard et al, 2001), attention to different aspects of information used in decision making, and different subjective values placed on perceived consequences (Scott, Reppucci, & Woolard, 1995). However, the research supporting this view includes ambiguous findings, especially in regard to adolescents older than age 14, in part due to the complex nature of the question under scrutiny (Scott et al., 1995). Not surprisingly, the failure of developmental researchers to identify systematic differences between adults and adolescents older than 15 on measures of various cognitive processes underlying decision making has fueled challenges to the view that adolescents should be regarded as less culpable than adults. In response, some proponents of the developmental perspective have shifted their attention to potential maturity differences in other capacities that impinge on decision making, namely, psychosocial factors or judgment factors (Steinberg & Scott, 2003).

Recent work examining comparative psychosocial maturity finds some differences between adolescents and adults. In one such study, adolescents were more likely than adults to express antisocial inclinations in response to hypothetical situations (Cauffman & Steinberg, 2000). The authors also found consistent age differences on measures of responsibility, perspective (future orientation), and temperance (impulsivity, situational evaluation). Finally, this research indicated that participants who demonstrated more psychosocial maturity were more likely to make socially responsible decisions in the hypothetical situations than were participants who demonstrated less psychosocial maturity. Further investigation of adolescent psychosocial development may provide additional support for the developmental differences position.

Scientists arguing the developmental perspective also have focused on neurobiological growth during adolescence. Studies utilizing recent advances in imaging technology are providing new information on brain development, suggesting that adolescent brains are less well-developed than previously believed. These studies focus on development of the brain via age comparisons and patterns of activation of the brain while the participant undertakes specific tasks. Results indicate that the frontal lobe undergoes significant change during adolescence, and is the last part of the brain to develop (Singer, 2005; Sowell, Thompson, Holems, Jernigan, & Toga, 1999). Frontal lobe maturation has been linked to the control of aggression and other impulses, and correlates with measures of cognitive functioning, such as long-term planning and abstract thinking (American Bar Association, 2004; Giedd, Blumenthal, Jeffries, Castellanos, Lui, et al., 1999). Other neurobiological evidence indicates that changes in the limbic system around puberty correspond to novelty-seeking and risk-taking behavior (Dahl, 2001). However, several researchers caution that it is difficult for neuroscientists to bring imaging research into real-life contexts, and that no current research connects the specific brain activations of adolescents to behavioral problems (Bower, 2004).

Although these lines of developmental research provide useful insights into the process of adolescent decision making, knowledge regarding immaturity of decision-making abilities, judgment, and other psychosocial capacities provides an incomplete picture of juvenile offending. To fully understand the causes and consequences of any human behavior, we need to understand the dynamic interplay between the personal and situational factors associated with the behavior—a view captured by ecological models of human behavior.

Ecological Models of Human Behavior

Significant empirical research conducted over the past century demonstrates the powerful influences that situational factors have on guiding and directing individual behavior (e.g., Darley & Batson, 1973; Milgram, 1963; Mischel, 1968). A person does not act in a vacuum. Evidence demonstrates that factors extrinsic to the individual and embedded in multiple levels of context play a significant role in the actions of every actor. Theorists adopting this ecological view have focused on the dynamic interaction between the individual and aspects of his or her social environment. According to this view, understanding the ongoing relationship between the individual and aspects of his or her social environment is essential to understanding the actor’s motivation and behavior on any particular occasion (Fondacaro, 2000). Bronfenbrenner (1979, 1989) applied this approach to human development, theorizing that the developing child is situated within structured, interconnected systems including family members, peers, the child’s school environment and local community, as well as more macro-level influences such as the legal system and the national economy. Factors linked to any of these systems may influence the child’s thoughts, feelings, and behaviors in a particular social context, even when the child is not in direct contact with these systems (Fondacaro & Jackson, 1999). The child and the child’s conduct are embedded in many different contexts. Some of these relevant contexts, including families and peer groups, have been investigated more systematically than others. This has contributed to a growing body of evidence across the various domains that begins to inform a more comprehensive ecological analysis of juvenile responsibility in relation to delinquent behavior.

Individual Factors

Behavioral science research traditionally has addressed the individual factors involved in delinquent behavior. Developmentally oriented investigators especially have focused on aspects of adolescent decision making, including cognitive processes, proclivity for risk taking, reasoning abilities, and impulsivity (Lynam, Caspi, Moffitt, Wilström, Loeber, et al, 2000; White, Moffitt, Caspi, Bartusch, Needles, et al., 1994). Most of this research has examined these factors in total or partial isolation from the other contexts bearing on the actor and from the situation in which the juvenile acts.

Family Factors

Family factors are among the strongest predictors of risk for delinquent behavior (Gorman-Smith, Tolan, & Henry, 2000). Poor parental monitoring, including inadequate direct supervision and knowledge of the child’s whereabouts and activities greatly increases the risk of delinquency (Laird, Petit, Bates, & Dodge, 2003). Moreover, Laird et al. (2003) found evidence for reciprocal influences between parental monitoring and delinquent behaviors.

Other parenting practices, including physical punishment and inconsistency of discipline, are related to increased risk for delinquency. Relational factors such as parent-child communication, emotional warmth, and parental involvement all have been found to have independent effects on the risk for delinquent behavior (Gorman-Smith et al., 2000; Loeber, Farrington, Stouthamer-Loeber, Moffitt, & Caspi, 2001; Scaramella, Conger, Spoth, & Simons, 2002; Wiesner & Windle, 2004). Personal characteristics of the parents themselves (e.g., antisocial behavior, substance abuse, psychopathology) also are related to increased risk for delinquency (Loeber, Farrington, & Petechuk, 2003).

Peer Influences

Peer influences play an important role in adolescent crime (Reiss & Farrington, 1991). Peers may exert direct influence through pressures to conform, coercion (Berndt, 1979; Scaramella et al, 2002), modeling, and social comparison (Adler & Adler, 1998; Kiesner, Cadinu, Poulin, & Bucci, 2002). Lhey also may exert more indirect influences through their impact on the approval-seeking motives of the at-risk child (Loeber et al., 2003; Scaramella et al., 2002).

Peer interactions occur in almost all of the contexts in which adolescents find themselves—schools, neighborhoods, and social activities. Lhrough daily interaction in these contexts, peers form networks and create behavioral expectations within these networks and contexts. Social comparison and conformity pressures continually shape ongoing behavior (Fagan, 2000). Peers also serve as an audience in daily interactions. “Witnesses are part of the landscape of social interactions and they influence adolescents’ decisions on how to conduct social relations and which behaviors to value” (Fagan, 2000, p. 373). Empirical evidence confirms that the majority of adolescent antisocial behavior occurs in groups (McCord & Conway, 2002), and that individuals take more chances when they are with peers than when they are alone (Steinberg, 2003).


Norms among peer groups and peer networks develop in neighborhoods (Bronfenbrenner, 1979). Opportunities and social controls in a neighborhood shape an individual’s behaviors and, to some extent, circumscribe the possible range of behaviors. Over the last 20 years, sociological research has attempted to identify community characteristics that are most influential in the development of crime. Level of poverty and socioeconomic status has emerged as the neighborhood structural variable that best predicts a range of negative outcomes including low school achievement, psychopathology, and delinquent behavior (Leventhal & Brooks-Gunn, 2000). In general, living in a high-poverty or low-socioeconomic status neighborhood has consistently been linked to delinquency (Loeber et al., 2001; Sampson, Raudenbush, & Earls, 1997).

Several studies have examined the relationship between racial/ethnic diversity and adolescent offending and have found that greater ethnic diversity is positively related to criminal activity (Loeber et al., 2001; Sampson & Groves, 1989, as cited in Leventhal & Brooks-Gunn, 2000). Lhis same research also found that high rates of residential instability are related to increases in juvenile crime. In addition, the Pittsburg Youth Study (Loeber et al., 2001) demonstrated that adolescents of an unemployed father were at increased risk for delinquency and violence.


Poor academic performance is related to the prevalence, onset, and seriousness of delinquency (Brewer, Hawkins, Catalano, & Neckerman, 1995; Wiesner & Windle, 2004). In addition, low school commitment, low educational goals, and poor motivation place children at risk for offending (Hawkins, Herrenkohl, Farrington, Brewer, Catalano, etal., 1998). School characteristics that have been linked to antisocial behavior include low levels of teacher satisfaction, lack of teacher cooperation, poor student-teacher relations, the prevalence of norms that support antisocial behavior, poorly defined rules and expectations for conduct, and inadequate rule- enforcement behavior (Herrenkohl, Hawkins, Chung, Hill, & Battin-Pearson, 2001).

Social Controls

Social controls on behavior include internalized restraints such as learned behavioral expectations and a continuum of external restraints ranging from informal social rules and regulatory processes to formal rules attached to specific locations (Duncan & Raudenbush, 1999; Fagan, 2000). “Social cohesion among individuals in the setting influences the strength of these regulatory processes” (Fagan, 2000, p. 375). Weak social controls and lack of community structure (disorganization) allow delinquent behavior to go on unchecked (Sampson etal., 1997). Adolescent problem behavior is negatively related to informal social controls (Elliot, Wilson, Huizinga, Sampson, Elliott, et al., 1996). Furthermore, strong social controls counter the impact on risk for delinquent behavior of negative neighborhood effects found in the poorest urban neighborhoods (Gorman-Smith et al, 2000).


Behavioral science data clearly demonstrate the influence of electronic visual media on viewers, both positive and negative. Over the past 40 years, research examining the relationship between viewing violent television and movies and the development of aggressive behavior has included experimental studies; observational studies; and more recently, longitudinal studies. Consortia of researchers, several different professional groups, and experts commissioned by the U.S. government have conducted numerous reviews of the literature dating as far back as the 1950s. The conclusions have been uniform, for the most part, and have only grown stronger as the breadth and depth of the research have increased. This literature base provides “unequivocal evidence that media violence increases the likelihood of aggressive and violent behavior in both immediate and long-term contexts” (Anderson, Berkowitz, Donnerstein, Huesmann, Johnson, et al., 2003, p. 81). Although theories regarding how this effect occurs and the differences between short-term effects and long-term effects are still developing, scientists investigating this topic agree on the following: Severe aggressive behavior usually occurs asa result of a convergence of multiple predisposing and precipitating factors, and exposure to media violence acts as one such factor (Huesmann, Moise-Titus, Podolski, & Eron, 2003).

Investigations in the field of media effects on child development recently have examined moderators of the influence of media violence on aggressive behavior. Such research demonstrates that viewer characteristics function as moderators. Specifically, highly aggressive individuals show greater effects of exposure to media violence than less aggressive individuals, although relatively nonaggressive children are also affected by violent media material (Bushman, 1995; Bushman & Geen, 1990, as cited in Anderson et al., 2003). Research also has shown that children’s perceptions of the violence as “lifelike” or “real” and their identification with aggressive characters is positively related to aggressive behavior (Huesmann et al., 2003). Characteristics of the televised material, such as the traits of the aggressive perpetrator and the justifications and consequences of the aggression, moderate viewers’ behavior (Anderson et al., 2003; Anderson and Gentile, 2005; Kunkel, 2005).

Unfortunately, little research has examined the social environment’s ability to moderate the impact of media violence. Some studies show that parents’ control of their children’s viewing of violence, as well as what parents do when their children view violence, mitigates the effects of the televised violence (Anderson et al., 2003). Although researchers have begun to consider the influence of culture, neighborhood, socioeconomic status, and peers on the impact of media violence, the results are somewhat equivocal and the studies are too few to draw definitive conclusions (Anderson et al., 2003). However, the development of an integrated ecological approach to this issue is likely to provide insights useful in designing effective interventions aimed at reducing violent behavior.

Situational Influences

Researchers have firmly established that people behave differently in different settings, even when experiencing similar motivations and emotions (Fagan, 2000). Features of the setting of antisocial acts interact with the actor to influence the outcome of the situation. Such features of the setting include the opportunities for crime; the people present; behavioral norms attached to specific aspects of the setting; provocations, such as alcohol or drug use; other criminal activity; and social controls (Fagan, 2000). The composition of a setting, such as the number and types of people present and the characteristics of those people (Stark, 1987); its physical location; and physical attributes such as lighting, noise, and décor (Felson, Baccagglini, & Gmelch, 1986) all influence behavior in that setting. Normative patterns of behavior within a setting, or beliefs about appropriate behavior in the setting, also influence current behavior—examples include beliefs about alcohol use in bars and at social gatherings (Fagan, 2000; Felson et al., 1986) compared to quiet, “mannerly” behavior in a courtroom or attentive behavior in a classroom.

Factor Integration

In recognition of the importance of ecological context in the determination of behavior, a growing body of work brings two or more factors together to investigate multiple contexts of delinquency. A study by Lynam et al. (2000) examined how the individual characteristic of impulsivity and juvenile offending differed as a function of neighborhood variation. The authors found that impulsive boys were at great risk for offending in poor neighborhoods (as defined by the percentage of families below the poverty line, rates of male unemployment, number of single-parent households, median income, number of households utilizing public assistance, and the percentage of African Americans in the census track). However, impulsivity posed little risk for delinquency for boys in better-off neighborhoods.

Gorman-Smith et al. (2000) examined how patterns of delinquency related differently to various configurations of family risk in different types of neighborhoods. The researchers found that in families with consistent parenting and organization, and strong emotional cohesion and family orientation, the children were less likely to engage in delinquent behavior. These effects were greater in neighborhoods with greater resources. Other studies also indicate that differential parenting practices are more or less effective depending on the neighborhoods in which the families live (Leventhal & Brooks-Gunn, 2000).

Some researchers hypothesize that peers influence antisocial behavior because of a lack of community or neighborhood institutions (such as organized activities or adult supervision) to regulate behavior (Leventhal & Brooks-Gunn, 2000). Consistent with this argument, Dishion, Andrews, and Crosby (1995) found that close friends of antisocial boys live in the same neighborhood and they spend time together in unsupervised and unstructured activities.

At least one study has examined the influence, relating to delinquent behavior, of three contextual factors: parents, peers, and neighborhood. Scaramella et al. (2002) found that relationships with deviant peers predicted delinquency. Furthermore, parents influence the types of friends with whom their adolescents associate. Specifically, nurturing and involved parenting were related to less association of adolescents with deviant peers. Parents also influence their children’s peer relationships by structuring their social environments. Parents select the schools their children attend, choose the neighborhoods in which they live, and influence the extracurricular and other activities in which their children engage (Bryant, 1985).

Taken as a whole, this research provides an emerging view of the adolescent offender that is more complex and differentiated than the image portrayed in recent juvenile justice policy. Considerations of multiple levels of influence are essential in addressing juvenile crime. As the empirical evidence suggests, an ecological approach requires an analysis of human problems in context and suggests a system based on more forward-looking solutions (Slobogin & Fondacaro, 2000).

A Social Ecological Analysis of Responsibility

A social ecological analysis of juvenile responsibility would consider the personal and situational influences on the juvenile and the juvenile’s behavior to make the determination of how best to prevent additional crime while meeting our social obligations to the child and the public at large. This perspective suggests that a rational thought process often is not the principal determinant of action (Slobogin & Fondacaro, 2000). Thus, if prevention of criminal behavior via deterrence is a goal of the system, punishment may have little impact, even assuming that the legal system is effective in detecting criminal behavior and that potential offenders are aware of this. Instead, deterrent effects will be linked to changing contextual factors that affect the well-being and behavior of adolescents. To promote general deterrence, laws passed primarily with a deterrent purpose need to focus at the aggregate level on those contextual factors that have been empirically linked to reduced risk for antisocial behavior. Although this perspective is based on a broader conception of deterrence than the one suggested by some scholars (Andenaes, 1966), we deliberately adopt the broader view because of the central role of general deterrence as a legitimate and popular criminal justice policy goal. To the extent that general deterrence continues to enjoy a privileged position in the law and wide popular support, future efforts to deter adolescent crime should broaden their focus beyond symbolic threats to include the promotion of environmental setting characteristics that are associated with reduced crime rates among adolescents.

Although recent policy shifts are moving the juvenile justice system further away from its original purpose, its goal has always been to prevent children from undertaking a life of antisocial behavior. This goal presumably benefits both society and the child. As noted above, the state has adopted a stance of societal obligation for the welfare of our children. In addition to processes aimed at general deterrence, a second approach to prevention is intervention to change individual behavior. A social ecological analysis dictates consideration of the youth’s historical and current context as part of the development of individualized intervention plans aimed at preventing recidivism. Such individualized intervention plans utilize a variety of approaches, programs, and processes to enhance protective factors and reduce risk factors for antisocial behavior. While this approach may closely resemble the system’s original goals, it would be a marked improvment because it would be based on empirical evidence of how best to effectively change human behavior. The adoption of a social ecological perspective would not mean that decision-making abilities and maturity of judgment, typically the primary focus of a “rational actor” analysis, would become irrelevant to judgments of juvenile responsibility. Such intrapersonal characteristics are important in the analysis of the individual’s propensity to engage in antisocial behavior, and should be considered as one among many relevant personal and contextual factors. Overall, a juvenile justice system guided by a forward-looking social ecological perspective would be focused not only on protection of children and their future opportunities, but also on public safety and the prevention of future crimes and other harms to society (Slobogin, Fondacaro & Woolard, 1999).

The social ecological perspective suggests that sanctioning youth as rational actors who are making poor decisions neither accurately addresses the course of development of the behavior nor effectively serves to prevent future antisocial behavior. Thus, retribution-based legal policies are not likely to succeed either in preventing juvenile crime or in meeting our societal obligations to improve the welfare of children. Rather, sanctions should be informed by an understanding of how best to change behavior, including interventions aimed at rehabilitation, crime prevention, restitution, and even punishment when it is empirically linked to the constructive objectives of educating, socializing, and protecting juveniles. Such interventions serve the dual purpose of reducing criminal activity and meeting the state’s special duties to children.

We believe the development and adoption of a risk management model similar to the one proposed by Slobogin and colleagues (1999) would begin to move us in this direction. A risk management model would not focus on assigning blame and meting out punishment that is deemed to be proportionate to culpability. Rather, the primary goal for dealing with juveniles who engage in illegal behavior would be to reduce the likelihood of recidivism in the future. As Slobogin et al. (1999) suggest, a risk management model of juvenile justice would be comprehensive and multisystemic, focusing on the various social systems that affect the child (family, peers, schools, neighborhood) and not just the child in isolation. What the child may or may not have been thinking at the time he or she offended loses its central importance and determinative influence under a risk management model.

Justice would be truly individualized, but not in the narrow sense that it would be tailored to the unique intrapsychic characteristics of the child. Rather, each juvenile offender would have an Individual Risk Management Plan (IRMP)—a plan that focuses on addressing dynamic (i.e., changeable) risk factors in the interconnected personal and social domains of the child’s life (e.g., poor impulse control, social incompetence, parental deviance and incompetence, affiliation with deviant peers, academic and vocational underperformance). Each child’s IRMP would be managed by a risk management team (RMT) headed by a caseworker who would seek input from the child and other relevant team members including school authorities, juvenile justice officials, mental health experts, vocational counseling professionals, and other interested parties. The RMT would oversee the implementation and ongoing evaluation of the IRMP, which would be subject to modification based on feedback and relevant empirical data on new and promising interventions that have been shown to work with juveniles with similar risk profiles. This would require the establishment of a juvenile justice record-keeping system capable of monitoring offender characteristics and risk profiles and tracking the efficacy of various intervention strategies to guide ongoing reform at the level of specific interventions and programs. Finally, the juvenile court or some organizational body with the competence to implement and evaluate juvenile justice policy should provide oversight of the risk management system and associated record keeping, which would provide important information to guide reform at the policy level.


One of the key decisions facing policy makers and the American public in the 21st century is whether to maintain a separate juvenile justice system. Virtually everyone agrees that we must judge and address deviant behavior by very young children in a manner that is quite different from the way we presently judge and punish adult offenders. Even before the era of juvenile justice reform in the United States around the turn of the last century, children under age 14 were typically presumed not to be responsible for their criminal conduct. However, in the context of the prevailing political rhetoric that champions personal responsibility, we are willing to treat children in this age range who break the law even more harshly than they would have been treated before the creation of the original juvenile justice system. Now, children who are 12 years old and younger are being tried, convicted, and punished severely as murderers. This would not have happened even back in the frontier days.

Shortcomings over the last century in implementing the ideals of juvenile justice reform seem paradoxically to have brought us to a point where children are treated even less humanely than they were during periods of our history when child exploitation was widely ignored if not openly encouraged. The fact that we have reached this point should give us pause and signal to the general public that we are doing something seriously wrong. Punishing children harshly under the rhetorical banner of personal responsibility is not working, is not fair, and costs too much.

Some behavioral scientists have attempted to soften our harsh treatment of juveniles by arguing and marshalling some modest evidence that children “think differently” than adults and therefore should be held less accountable. One implication of this line of reasoning is that if children thought like adults, then it would be appropriate to treat them in the same way we currently treat adult criminals. Juveniles are considered “developmentally immature,” on their way to adultlike autonomous decision making, but not yet there. We believe this premise itself is flawed but will leave to another day a full discussion of criminal responsibility across the entire lifespan. With respect to juveniles, we believe that focusing primarily on internal psychological characteristics of the child (i.e., mens rea) to judge criminal responsibility, as we do with adults, leads us to erroneous judgments. Arguably, at least from the standpoint of a moral judgment model, continuing to knowingly embrace the tenets of mens rea analysis in the face of compelling evidence that situational factors contribute as much or more to criminal behavior as conscious, albeit immature, decision making by juveniles is not only misguided—it is immoral.

Over time, juveniles are increasingly facing sanctions that look more and more like adult punishment. Apparently, we have gotten right back to where we started—some would say, we have even taken a step further back. Developmentalists and forensic psychologists working within the fundamental premises of a retributivist framework for criminal justice have provided some evidence that children and younger adolescents on average do not make decisions and exercise judgment in a manner that is as competent as decisions made by adults. They then suggest that this should lead to a presumption that children and younger adolescents who break the law should not be held fully responsible for their crimes and that a separate juvenile justice system is necessary for these youngsters. However, this approach abandons many older adolescents (and adults whose decision making and judgment may be indistinguishable from that of the typical younger adolescent) to an increasingly retributivist and punitive adult criminal justice system.

We believe that behavioral scientists and other informed citizens must challenge and change the narrow and outdated criteria for judging juvenile responsibility that presently tie harsh legal sanctions almost exclusively to what a judge or jury might believe the child was thinking at the moment he or she broke the law. The causes of human behavior are much more complex. This “immature evildoer” theory of crime does not take full account of what we have learned over the past century about the interrelated social and personal causes and consequences of human behavior. We need a framework for judging and addressing juvenile crime that takes this complexity into account.