Reflection on “A Revised Strain Theory of Delinquency”

My 1985 article presented a revised strain theory, which stated that delinquency results from the blockage of pain-avoidance behavior as well as the blockage of goal-seeking behavior. The article had some success, laying the groundwork for my “general strain theory,” now one of the leading explanations of crime and delinquency (Agnew 1992, 2007). This reflection describes how the article revised strain theory, how I built on the article, and the research inspired by the article.

Most strain theories state that delinquency results when individuals are unable to achieve their goals through legitimate channels. These theories focus on the goal of monetary success or the somewhat broader goal of middle-class status (Merton 1938; Cloward and Ohlin 1960; Cohen 1955). Lower class individuals are said to have special trouble achieving these goals because they are less well prepared for school, attend inferior schools and lack the means for advanced educations. The frustration that results from this goal blockage increases the likelihood of crime; individuals may try to obtain money through acts such as theft and drug selling, may strike out at others in their anger or may focus on alternative goals that are conducive to crime (e.g., masculinity goals emphasizing toughness). Strain theories were the leading theories of crime in the 50s and 60s and they had a major effect on public policy, being an inspiration for the War on Poverty. But by the 1980s strain theories had come under serious attack, they had little effect on crime research, and several prominent criminologists were arguing that they should be abandoned (Agnew 1985).

Strain theories were attacked for several reasons (Agnew 1985). Among other things, they were unable to explain why crime rates peak during adolescence; if anything, the strain caused by the inability to achieve monetary and status goals should be more consequential for adults. Also, research using self-report measures of crime revealed that the relationship between social class and delinquency is weaker than previously thought, with some studies finding little or no relationship (Agnew 1985). The most prominent attack, however, was based on the assertion that strain theories predict that crime should be highest among those who do not expect to achieve their educational and occupational aspirations. Research instead found that crime is highest among those with both low educational and occupational expectations and aspirations, a finding interpreted in terms of control theory (such individuals do not expect or desire much, and so have little to lose by engaging in crime). Several strain theorists responded by arguing that individuals pursue a range of goals beyond monetary and status goals, with many of these goals being more immediate in nature, such as good relations with parents and spouses. But the research here was not very supportive (Agnew 1985).

My 1985 article revised strain theory by arguing that crime is caused not so much by the inability to achieve positively valued goals, but by the inability to escape from painful or aversive conditions. This argument was said to be especially relevant to the explanation of delinquency because juveniles are “compelled to live with their family in a certain neighborhood; to go to a certain school; and, within limits, to interact with the same group of peers and neighbors” (Agnew 1985:156). There is little that juveniles can do to legally escape if they are mistreated by others in these settings. Juveniles who experience aversive treatment, however, may engage in delinquency to escape from or reduce such treatment (e.g., running away from home, truancy, fighting to end peer harassment). They may also become angry and strike out at the source of their aversive treatment or related targets. I briefly noted, however, that several factors influence whether juveniles respond to the blockage of pain-avoidance behavior with delinquency, including their beliefs regarding delinquency, their level of association with delinquent peers, the likelihood of sanction and the perceived injustice of the aversive treatment.

Certain of these arguments were tested with data from a national sample of adolescent boys, and the results indicated that delinquency is more likely among those subject to harsh, demeaning and unfair treatment by parents and teachers, as well as those who find school boring and a “waste of time.” Further, the effect of these variables on delinquency is partly mediated by a measure of anger. Subsequent research verified these findings with longitudinal data (Agnew 1989). The revised theory helps explain why crime rates peak among adolescents. It can also explain middle-class delinquency, since middle-class adolescents also encounter aversive situations from which they cannot legally escape. The revised theory represented a significant departure from prior strain theories; rather than focusing on what have been called “non-events” or the inability to achieve one's goals, the theory focuses on negative events or mistreatment by others.

While the revised theory attracted some attention in and of itself, it was important largely because it laid the foundation for my general strain theory (GST) of crime and delinquency (Agnew 1992, 2007). GST attempted to merge the revised theory with prior strain theories, and it drew heavily on the stress, emotions and justice literatures. GST states that there are three major types of strains, with strains defined as events and conditions disliked by individuals. The first type involves the inability to achieve one's goals. This strain is, of course, derived from prior strain theories, although GST states that the failure to achieve several goals is conducive to crime, including goals involving thrills/excitement, high levels of autonomy, masculine status and the desire for much money in a short period of time. This focus on goal blockage represents a break from the revised strain theory, which de-emphasized this type of strain based on data suggesting that goal blockage is unrelated to delinquency. I came to believe that the studies challenging the role of goal blockage were flawed. Among other things, these studies focused on educational and occupational goals, and they measured goal blockage in terms of the disjunction between expectations and aspirations or ideal goals. I argued that the inability to achieve ideal goals may not prompt much frustration and that goal blockage is better measured in terms of the disjunction between actual achievements and expected goals. Also, the inability to achieve educational or occupational goals is not conducive to crime, because those with high educational/occupational goals have some commitment to conventional society. Further, certain qualitative and recent quantitative research suggest that the types of goal blockage I identified do increase crime (Agnew 2007).

The second major type of strain involves the loss of positive stimuli, such as the loss of money or property, breakup with a romantic partner and the death of a friend. The third type involves the presentation of negative stimuli, such as verbal and physical abuse. These later two types of strain deal with painful or aversive events and conditions. GST, however, does not focus on the blockage of pain-avoidance behavior, but simply on the experience of painful events/conditions. This shift was based on my reading of the stress, emotions and justice literatures, as well as certain qualitative research in criminology. These sources suggest that painful events and conditions generate negative emotions and sometimes prompt criminal coping, even when legal escape is possible. At the same time, GST does state that criminal coping is more likely when individuals lack the skills and resources to cope in a legal manner (more below).

GST further builds on the revised theory by better specifying the types of aversive treatment most likely to lead to crime (Agnew 2001, 2007). In retrospect, it became clear that not all types of aversive treatment increase crime. Some types, in fact, reduce crime, such as parental punishments that are not overly harsh and that are contingent on the juvenile's misbehavior. The most criminogenic strains or aversive events are high in magnitude (severe, frequent, of long duration, expected to continue into the future and involving central goals, needs, values, activities and/or identities). They are seen as unjust, involving the voluntary and intentional violation of relevant justice norms. They are associated with low social control. For example, that strain involving parental rejection is associated with weak bonds to parents and poor supervision. By contrast, that strain involving long study hours is associated with a strong bond to school and high grades. They are easily resolved through crime (e.g., a desperate need for money). And they involve exposure to others who reinforce crime, model crime and/or teach beliefs favorable to crime (e.g., peer abuse). GST list several specific strains with these characteristics: parental rejection; harsh, erratic, and/or excessive discipline; child abuse and neglect; negative secondary school experiences (e.g., low grades, negative relations with teachers); peer abuse; work in the secondary labor market; chronic unemployment; certain marital problems, such as verbal and physical abuse and frequent conflicts; criminal victimization; homelessness; discrimination; and residence in severely deprived communities. Research suggests that these strains increase the likelihood of crime, with some being among the most important causes of crime (Agnew 2007).

GST also builds on the revised theory by better describing why strains increase the likelihood of crime. Like the revised theory, GST emphasizes the key role played by anger. Strains, particularly major strains that are seen as unjust, are likely to make individuals angry. This anger creates pressure for corrective action, interferes with the use of certain legitimate coping strategies, such as negotiation, reduces concern for the consequences of one's behavior, and creates a desire for revenge. And data suggest that anger, particularly state anger, partly explains the effect of strains on crime (Agnew 2007). GST also focuses on other negative emotions and certain recent research is exploring the idea that different types of strain lead to different negative emotions (e.g., anger versus fear), and that different emotions are conducive to different types of crime (e.g., anger to violence, depression to drug use; e.g., Ganem 2010). Researchers have also begun to explore additional mediating mechanisms between strains and crime. Limited evidence suggests that strains may also increase crime by reducing social control, increasing association with criminal peers, fostering beliefs favorable to crime, contributing to criminogenic traits (e.g., negative emotionality) and prompting the view that crime is a cost-effective response (Agnew 2007).

Further, GST devotes much attention to those factors that may condition the effect of strains on crime, again building on the revised strain theory. A range of factors are said to be relevant here, including coping skills and resources (e.g., problem-solving skills, financial resources, self-efficacy), levels of conventional and criminal social support, social control, association with criminal others, beliefs regarding crime and exposure to situations where the costs of crime are low and the benefits high (Agnew 2007). The research here has produced mixed results. For example, some studies find that individuals with criminal peers are more likely to cope with strains through crime, while other studies do not. The reasons for these mixed results are unclear, although several possibilities have been suggested. For example, researchers usually consider one conditioning variable at a time, with other conditioning variables controlled. But it has been suggested that individuals may not engage in criminal coping unless they score high on several conditioning variables conducive to crime, especially given the strong condemnation and sanction normally associated with crime (Agnew 2007; Mazerolle and Maahs 2000).

Finally, GST has been applied to range of issues beyond the explanation of why some individuals have higher levels of crime than others – the focus of the revised strain theory. Among other things, GST has been used to explain patterns of offending over the life course of given individuals. In particular, GST has been used to explain why some individuals offend primarily during their adolescent years and others offend at high levels over much of their lives (Agnew 2007; Slocum 2010). GST has also been used to explain group differences in crime, including, gender, age, race/ethnic and class differences (e.g., Agnew 2007; Broidy and Agnew 1997; Kaufman et al. 2008). Further, GST has been used to explain community and societal differences in crime (e.g., Agnew 1999, 2007). Group and community/societal differences in crime are explained in terms of differences in the extent of strain, the types of strain and/or the factors that condition the response to strains. For example, males are said to have higher levels of crime than females because they are more likely to experience many of the strains conducive to crime, such as criminal victimization. Also, males are more likely to cope with strains through crime, particularly other-directed crime. This stems from such things as gender differences in traits such as self-control and empathy, in levels of supervision and in association with delinquent peers. Gender differences in crime are not due to differences in the level of strain; females, in fact, may experience higher levels of strain than males. Most recently, GST has been used to suggest crime-control policies, most of which focus on reducing exposure to criminogenic strains and the likelihood of criminal coping (Agnew 2010).

In sum, the revised strain theory described in the 1985 Social Forces article has itself been substantially revised and extended. The resulting general strain theory is now one of the leading theories of crime and delinquency and has inspired hundreds of studies. But the distinguishing features of GST—its focus on negative treatment and the central role it assigns to negative emotions—were first highlighted in the Social Forces article.

Notes

References