Writing Assignment

CHAPTER

Sociological Theories of CriDle II: CriDle and Social Processes

This chapter might well have been titled "Learning to be Criminal," because it focuses on theories of crime that emphasize the role of social learning or socialization in the development of criminal behavior. Sociologists define socialization as the process of social interaction through which a society's culture is taught and learned and human personalities are developed (Renzetti & Curran, 2000). Although we typically associ- ate socialization with early childhood, it is actually an ongoing process that continues throughout an individual's life. The fact that socialization is a process of social interac- tion tells us that it occurs through communication with other people; itis not something we do on our own, in isolation. What is taught-that is, the content of socialization- varies across societies, communities, and social groups. Those who do the socializing; whom sociologists call agents of socialization, also vary. Agents of socialization influence us over the course of our lives: they are individuals, groups, and institutions that have as one of their primary functions the socialization of members of a society by providing explicit instruction in or modeling of social expectations (Renzetti & Cur- ran, 2000).

Criminologists who emphasize the importance of socialization in the etiology of crime study how various agents of socialization-especially the family, the school, the peer group, and, more recently, the media-affect an individual's likelihood of pursu- ing criminal or noncriminal activities. These theorists maintain that what distinguishes a criminal from a noncriminal is not physiology, genetics, mental disorder, race, sex, or even social class, but rather socialization experiences. The first criminologist to forward this argument as a systematic theory of crime was Edwin H. Sutherland. Thus, our exploration of the relationship between crime and social processes begins with his work.

'.

Sutherland's Differential Association Theory

Edwin H. Sutherland (1883-1950) is generally regarded as the leading criminologist of his generation (Martin et aI., 1990). Sutherland is best known for his study of white- collar crime (1949), his life-history analysis of a professional thief (193 7), and his devel- opment of the theory ofdifferential association, which is the focus of our discussion here.

135

136 CHAPTER 5

Sutherland made his first formal statement of the theory of differential associa- tion in 1939 in the third edition of his textbook, Principles ofCriminolog;y. For the fourth edition, published in 1947, he revised the theory slightly as a result of his own rethink- ing and in response to the criticisms and suggestions of his colleagues. 1 By this time, however, the theory was "considered to be one of the best known and most systematic and influential of the interpersonal theories" (Martin et al., 1990, p. 155). According to Matsueda (1988), "The theory was instrumental in bringing the perspective of sociol- ogy to the forefront of criminology" (p. 277).

Sutherland was critical of biological and psychiatric theories of crime, but he was also dissatisfied with the eclectic and disorganized nature of the prevalent sociological explanations of the time, which took a multi-factor approach to crime causation. In developing his own perspective, Sutherland drew on the work ofa variety of scholars. In our examination of differential association theory we will see the influence of Gabriel Tarde (1843-1904), whose "laws ofimitation" included the postulate that people "imi- tate one another in proportion as they are in close contact" (1912, p. 326). It has been argued, too, that the work of John Dewey was a source of inspiration for Sutherland (Martin et al., 1990).

Clearly, however, Sutherland was most strongly and directly influenced by the writings and research of his friends, colleagues, and associates at the University of Chicago. First, we will recognize in differential association theory the idea of cultural transmission (see Chapter 4). Second, we will find that the work of the symbolic inter- actionists George Herbert Mead, W 1. Thomas, and others was influential. As VoId and Bernard (1986) explain, the symbolic interactionists argue that:

people construct relatively permanent "definitions" of their situation out of the mean- ings they derive from their experiences. That is, they derive particular meanings from particular experiences but then generalize them so that they become a set way of look- ing at things. On the basis of those different definitions, two people may act toward similar situations in very different ways. (p. 211)

Although we will discuss symbolic interactionism later in the chapter when we consider labeling theory, suffice it to say here that the interactionists' focus on how individuals construct social reality through communication with one another is also a concern that underlies the theory of differential association.

Finally, the notion of culture conflict is a theme in differential association theory. Sociologist Thorsten Sellin argued in the late 1930s that crime is an outcome of a clash between cultures. According to Sellin (1938), in a homogeneous society the "conduct norms" that are codified into law represent a consensus of the society's members. But in a heterogeneous society that contains many diverse subcultures, the law represents the conduct norms of the dominant culture only, and members of various subcultures may violate the law when they follow their groups' indigenous conduct norms. From these ideas, Sutherland developed the concepts of differential social organization and dif- ferential group organization. He utilized these concepts to explain variations in crime rates across countries, cities, and groups (Cressey, 1960). Let us turn, now, to the the- ory of differential association itself.

Sociological Theories of Crime IT: Crime and Social Processes 137

Sutherland's Nine Propositions

Students usually have little difficulty learning differential association because Suther- land presented it in the form of nine, fairly straightforward propositions, each followed by a brief explanatory statement. These nine propositions are:

1. Criminal behavior is learned. 2. Criminal behavior is learned in interaction with other persons in a process of

communication. 3. The principal part of the learning of criminal behavior occurs within intimate

personal groups. 4. When criminal behavior is learned, the learning includes (a) techniques of com-

mitting the crime, which are sometimes very complicated and sometimes very simple; (b) the specific direction of motives, drives, rationalizations, and attitudes.

5. The specific direction of motives and drives is learned from definitions of legal codes as favorable and unfavorable.

6. A person becomes delinquent because of an excess of definitions favorable to vio- lation of law over definitions unfavorable to violation of law.

7. Differential association may vary in frequency, duration, priority, and intensity. 8. The process oflearning criminal behavior by association with criminal and anti-

criminal patterns incorporates all the mechanisms that are involved in any other learning.

9. Although criminal behavior is an expression of general needs and values, it is not explained by those general needs and values since noncriminal behavior is an expression of the same needs and values. (Sutherland, 1947, pp. 6-8)

From the outset: Sutherland makes it clear that criminality is not inherited. Rather, it is learned in the same way that any other behavior is learned: through inter- personal communication and social interaction in intimate groups-what sociologists call primary groups (including family and friends). What is learned through this process includes particular attitudes and motivations as well as techniques for committing crimes. However, being exposed to criminal attitudes and motivations and even know- ing how to commit a crime does not mean that a person will engage in criminal activ- ity. Many people who desperately need money and who know various illegal ways to obtain it nevertheless persevere in solving their financial problems through entirely legal means. Indeed, to become a criminal or a delinquent, one also must learn specific situational meanings or definitions. To quote Sutherland, "A person becomes delin- quent because of an excess of definitions favorable to violation of law over definitions unfavorable to ~olation of law." Sutherland called the process of social interaction by which such definitions are acquired differential association.

Sutherland chose the term differential association to emphasize that, "In any soci- ety, the two kinds of definitions of what is desirable in reference to legal codes exist side by side, and a person might present contradictory definitions to another person at dif- ferent times and in different situations" (Cressey, 1960, p. 2). All associations are not equal. In fact, Sutherland specified that associations vary in frequency, duration, priority,

· 138 CHAPTER 5

and intensity. In other words, associations that occur often (frequency) and are long- lasting (duration) will have a greater impact on an individual than brief, chance encounters. Associations that occur early in a person's life, especially in early child- hood, are more important than those that occur later on (priority). And associations with prestigious people or with those one holds in high esteem will be more influen- tial than associations with those for whom one has little regard or who are socially dis- tant in one's life (intensity).

The theory of differential association is an explanation of how individuals become criminal or delinquent. It is important to keep in mind, however, that Suther- land wished to explain not only differences in individuals' participation in criminal activity, but also group and societal variations in crime rates. To go beyond the indi- vidual level, Sutherland originally utilized the concepts of culture conflict and social dis- organization (see Chapter 4). In his 1947 statement of the theory, though, he used the terms differential social organization and differential group organization instead of social disorganization. Sutherland's objective in making this revision was to point out that areas with high crime rates are not unorganized, but rather are composed of various groups with divergent standards of conduct, which increases the probability that mem- bers of some groups living there will learn definitions favorable to law violation. As Cressey (1960) explains, "In a multi-group type of social organization ... there are alternative educational processes in operation, varying with groups, so that a person may be educated in either conventional or criminal means of achieving success" (p. 2). Put somewhat differently, "Sutherland's theory, then, states that in a situation of differential social organization and [culture] conflict, differences in behavior, includ- ing criminal behaviors, arise because of differential associations" (Void & Bernard, 1986, p. 213).2

Differential association theory is valuable. It addresses questions that the strain, subcultural, and opportunity theories left unanswered. For instance, how can we account for the fact that individuals who have equal opportunities to commit crimes do not all engage in criminal activity? Why is it that individuals who are equally pressured toward nonconformity by factors such as poverty do not all become nonconformists? And why do some individuals who appear to have all their material needs met-mem- bers of the upper- and upper-middle classes-nevertheless embezzle business funds, defraud consumers, and participate in price fixing schemes and insider stock trading, as well as other criminal practices? For Sutherland, the answer was clear: differential association. 3 Other criminologists, however, were less certain, and Sutherland's work became-and to some extent, remains-at the center of controversy.

Strengths and Weaknesses of Differential Association Theory

You will recall from Chapter 1 that one essential criterion for determining the strength of a theory is the extent to which it is supported by empirical testing. Conse- quently, perhaps the most damaging criticism of differential association theory is that it is untestable (see, for example, Adams, 1974; Glueck, 1956; Hirschi, 1969; Korn-

Sociological Theories of Crime IT: Crime and Social Processes 139

hauser, 1978). Sutherland (1947) argued that ideally the propositions of the theory could be "stated in quantitative form and a mathematical ratio [of an individual's expo-

. sure to weighted definitions favorable and unfavorable to law violation] be reached" (p. 7). However, he acknowledged that developing such a formula would be extremely difficult.

Matsueda (1980) argues, however, that specific hypotheses, propositions, and empirical implications of the theory are testable, and he as well as others have under- taken such tests. For example, DeFleur and Quinney (1966) used the mathematical model of set theory to determine if empirically verifiable hypotheses can be derived from the theory. They claim that, "Those who have declared Sutherland's theory to be incapable of generating testable hypotheses appear to have underestimated it. In fact, it can generate more hypotheses than could be adequately studied in several lifetimes" (DeFleur & Quinney, 1966, p. 20; see also Orcutt, 1987). The question remains, how- ever, as to how one can measure or observe an excess ofdefinitions favorable to law viola- tion. It appears that, despite DeFleur and Quinney's claim, the key variable in the theory is difficult to operationalize. In fact, Cressey considered this to be one of the most serious weaknesses in differential association theory (Akers, 1996).

Still, criminologists have developed various ways to empirically test differential association theory. One popular method is to ask a sample of juveniles or adults not only about their own values and behavior, but also those of their friends (for example, "How many of your friends have been arrested in the past year?" "To what extent do you think your friends approve of [a specific deviant behavior]? "How many of your friends have done any of the following [deviant behaviors] in the past year?"). The underlying assumption of this approach is that an individual will most likely learn delinquency or criminality from friends who approve of delinquent or criminal behav- ior and who engage in such behavior themselves. This, then, may serve as an indirect measure of the acquisition of definitions favorable to law violation. Various measures of the frequency, duration, priority, and intensity of peer associations also are often used in these studies.

Overall, this research lends support to the theory of differential association, showing a strong correlation between individuals' associations with delinquent or criminal peers and their own likelihood of engaging in delinquent or criminal activi- ties (Cheung & Ng, 1988; Matsueda & Heimer, 1987; Orcutt, 1987; Tittle et al., 1986). In fact, "the best predictor of the extent of an adolescent's involvement in delin- quent behavior is ... the number of the youth's delinquent associations" (Johnson et al., 1987). Such findings, however, do not demonstrate the validity of the theory. "In other words, the data indicate a relationship between delinquent [or criminal] behav- ior and interaction with others, but they do not demonstrate that it is in fact the con- tact with these 'certain others that causes the behavior. Causal sequence is not established nor are other potential factors ruled out" (Martin et al., 1990, p. 166; see also Costello & Vowell, 1999). We cannot pinpoint through such research whether delinquent or criminal values were actually transmitted from delinquent or criminal peers. It may also be the case that those who already hold delinquent or criminal val- ues seek out peers like themselves or, as Sheldon and Eleanor Glueck put it, "birds of

140 - C HAP T E R 5

a feather flock together." The problem of temporal sequence is made worse by the fact that most of these studies rely on cross-sectional data, which do not allow researchers to decipher the temporal sequence of the learning/action process.

One recent study, conducted by Warr and Stafford (1991), addresses the issue of temporal sequence utilizing data from the National Youth Survey, a five-year panel study of a national probability sample of youths who were aged eleven to seventeen in 1976. These researchers uncovered both positive and negative evidence relative to dif- ferential association theory. They found, as Sutherland had argued, that "The attitudes of adolescents are influenced by the attitudes ... of their peers, and those attitudes in turn affect delinquency" (p. 162). They also discovered, however, that the behavior of friends has a strong effect on adolescents' behavior independent of attitudes, indicat- ing that Sutherland's theory may be incomplete because of its emphasis on the trans- mission of definitions favorable to law violation. More specifically, Warr and Stafford (1991) found that:

First, the effect of friends' attitudes and friends' behavior is in fact enhanced when the two are consistent. Friends who behave as well as think in a delinquent fashion produce the most delinquent associates. However, when the attitudes and behavior of peers are inconsistent, the behavior of peers appears to outweigh or override the attitudes of peers. The actions of peers, it seems, speak louder than their attitudes. (pp. 859-860)

Importandy, these findings were obtained when the data were analyzed longitudinally as well as cross-sectionally. Costello and Vowell (1999) also found that friends' delin- quent behavior had a greater effect on research subjects' own delinquent behavior than did definitions favorable to law violation.

In another study, Johnson and his associates (1987) examined factors that influ- ence adolescents' use of drugs. They, too, found support for differential association, but at the same time, they identified an intervening variable. Johnson et al. report that parents' use of drugs and parents' prodrug definitions have relatively litde impact on adolescents' drug use. "By the time a child reaches adolescence, all of the direct parental influences seem to play only minor roles in determining his or her drug use" (p. 333). Instead, the most significant factor influencing adolescents' drug use is the proportion of the adolescent's best friends who use drugs. This variable itself had a fairly strong effect on friends' prodrug definitions which, in tum, had a moderate influence on drug use. However, regardless of friends' prodrug definitions, friends' actual use of drugs had the strongest effect of all the variables tested.

Johnson et al. (1987) explain this finding by arguing that it is situation pressures to use drugs, not I?eers' prodrug definitions, that play the dominant mediating role in adolescents' drug use. "In other words, most of the impact from friends' drug behav- ior to personal behavior seems to bypass the definitions or attitudes variable. It is not so much that adolescents use drugs because the drug use of their friends makes drug use seem right or safe; rather, they apparently use drugs simply because their friends do" (p. 333, authors' emphasis).

Differential association theory has been criticized for several other reasons as wel1.4 One ongoing debate, for instance, centers on the role of the media in crime cau-

sation. Sutherland (1947) argued that crime is learned within intimate personal groups and that "impersonal agencies of communication [such as the media], playa relatively unimportant part in the genesis of criminal behavior" (p. 6). Of course, when Suther- land wrote this statement, the media, especially television, were not very influential in adults' and children's everyday lives. However, as we see in Box 5.1, contemporary researchers disagree about the relationship between media consumption and crime, a topic that is widely researched and highly controversial today.

Sociological Theeries of Crime II: Crime and Social Processes

(continued)

141

(Media Report to Women [MRTW] , 1993). In contrast, in the United States, little action has been taken to curb media violence. In 1993, the chief executives of the major national networks agreed to air parental advisories before programs the networks considered violent, and in 1997, the networks instituted a. six-category ratings system to advise viewers fifteen seconds before the start of a program as to appropriate viewer ages for the program. However, research shows that in the following year, 1998, the number of violent programs increased (Mifflin, 1998).

The debate over the effects of violent viewing is periodically fueled by incidents involving viewers acting out what they have seen in a program or film. In one case, for example, a five-year-old Ohio boy set fire to his family's mobile home after watching the MTV cartoon Beavis and Butthead, in which the characters depicted setting fires as fun. The boy's two- year-old sister died in the fire. The films Natural Born Killers, The Program, Colors, and Taxi Dri- ver have also been implicated in murders and other violent crimes (Mifflin, 1998). Still, despite literally thousands of studies, we are no closer to definitively answering the question, "Does vio- lent viewing cause violent behavior in viewers?" To summarize this voluminous research, it can be said that there is a strong correlation between violent viewing and violent behavior, but a corre- lation between two variables does not necessarily mean that one causes the other.

CONTROVERSY AND DEBATE Do We Learn to Behave Violently by Watching Violent Media?

BOX 5.1

In the 1930s, when Sutherland was about to pub- lish his theory of differential association for the first time, it was argued by some that motion pic- tures were a major contributing factor to delin- quency. During the late 1940s, when Sutherland was revising his theory, a controversy raged over whether comic books contributed to delin- quency. In both cases, claims were made that these media graphically depicted crime, violence, and sex in such ways that viewers or readers, especially the impressionable young, could be led into crime or "sexual deviation" by imitating the behavior of their movie or comic strip heroes or heroines. In each case, a multitude of studies

(

was undertaken, and the bulk of the research indicated that there was no empirical evidence supporting a causal relationship between media depictions of crime and violence and people's actual behavior. Even today, however, the media's role in the etiology of crime and anti- social behavior continues to be debated.

In 1993, the Canadian Radio-Television and Telecommunications Commission issued strict rules regulating the broadcast of violent programming. These rules include a ban on the depiction of gratuitous violence, a limitation of the time that adult programming (including ads and promotions) containing violence can be broadcast (between 9 P.M. and 6 A.M.), and a total ban on any violent depictions in children's pro- grams that minimize the effects of violence or that encourage or promote imitation of violence

142 CHAPTER 5

BOX 5.1 Continued There are three major explanations of this

relationship (Vivian, 1993). One emphasizes the cathartic effect of violent viewing, stating that viewing violence can actually reduce the violent drives of viewers because watching allows them to fantasize about violence, thereby releasing tensions that may lead to real-life aggression. It . has also been argued that this catharsis may lead viewers to take positive rather than violent action to remedy the problem. For instance, Vivian (1993) reports that following the broad- cast of the television movie, The Burning Bed, in which a severely abused woman ultimately kills her batterer-husband by setting fire to his bed while he sleeps, domestic violence agency hot- lines were flooded with calls from battered women seeking help.

Vivian also notes, however, that The Burning Bed may have also inspired some people to take violent action. One man, for instance, set his estranged wife on fire and another severely beat his wife, both claiming'Lhey were motivated by the movie. Such acts of direct imitation are at the heart of a second explanation that focuses on the modeling effect of violent viewing. Put simply, this explanation maintains that media violence teaches viewers to behave violently through imi- tation or modeling, a concept that we will dis- cuss in greater detail later in this chapter. Suffice it to say here that despite the sensationalism sur- rounding individual acts of direct imitation, they are very rare. Moreover, there are several inter- vening factors that influence whether a specific act will be imitated. These include the model's and the learner's relative age and sex, the model's objective status and her or his status in

the eyes of the learner, and whether the model is rewarded or punished for engaging in the behavior in question.

These and other factors are considered by researchers who propose a third explanation that emphasizes the catalytic effect of violent viewing. This position says that if certain conditions are present, viewing violence may prompt real-life violence. The emphasis is on probabilistic causation rather than direct causation. The violent viewing "primes" the viewer for violent behavior; it increases the risk of violent behavior just like cig- arette smoking increases the risk of developing cancer. If the violence is portrayed as realistic or exciting, if the violence succeeds in righting a wrong, if the program or film contains characters or situations that are similar to those the viewer actually knows or has experienced, and if the viewer's media exposure is heavy, the probability of the viewer behaving violently increases (Bok, 1998; Mifflin, 1998; Vivian, 1993).

It is doubtful that the federal government will enact legislation to curb violent program- ming any time soon. Previous government attempts to regulate broadcast hours in order to prevent children from viewing programs or films with adult themes have been struck down by the courts as a violation of the First Amend- ment (see, for example, Lewis, 1993). What is more certain, however, is that the majority of U.S. households will continue to fulfill the last condition of the catalytic effect: frequency of viewing. It is estimated that by the time a thild leaves elementary school, he or she will have watched 8,000 murders and more than 100,000 violent acts on television (Bok, 1998).

A final criticism of differential association theory comes primarily from those interested in the psychological underpinnings of human behavior. These critics main- tain that Sutherland's conception of learning is too simplistic. Although Sutherland said that the process of learning both criminality and law abiding behavior involves all the mechanisms of learning, his propositions only vaguely outline how learning occurs. This criticism is significant not only because it highlights an area that Sutherland neglected, but also because it prompted a number of criminologists to develop new

Sociological Theories of Crime II: Crime and Social Processes 143

positions that utilize many of the principles of differential association while incorpo- rating more complex work of various learning theorists. Let's examine, then, some of the theoretical offshoots of differential association theory.

Sutherland's Legacy

Akers's Social Learning Theory

First developed in the 1960s by Robert L. Burgess and Ronald L. Akers (1966) and later elaborated by Akers (1973; 1998), social learning theory is a revision of Sutherland's work that utilizes the central concepts and principles of modern behaviorism. 5 According to Akers (1994, p. 101), the theory is a "general processual explanation of all criminal and delinquent behavior." Like Sutherland, Akers maintains that criminal behavior is learned. However, the way it is learned, he argues, is through direct operant condition- ing and imitation or modeling of others.

The principle of operant conditioning is probably familiar to you; most of us have heard of Pavlov's dogs who were trained or conditioned to salivate when they heard a bell ring. In this kind of conditioning-ealled classical conditioning-the behav- ioral response is elicited by a prior stimulus. According to Akers, though, the form the behavior takes and its frequency of recurrence depend on instrumental conditioning; that is, the behavior is learned or conditioned as a result of the effects, outcomes, or consequences it has on an individual's environment.

Operants are not automatic responses to eliciting stimuli; instead, they are capable of developing a functional relationship with stimulus events. They are developed, main- tained, and strengthened (or conversely are repressed or fail to develop), depending on the feedback received or produced from the environment. (Akers, 1985, p. 42)

There are two major processes involved in instrumental conditioning-reinforce- ment and punishment, and each of these may take two forms (see Table 5.1). Behavior is reinforced when the consequences it has or the reactions of others encourage an

TABLE 5.1 Akers's Social Learning Perspective of Deviant Behavior

Stimulus

+ '.

Behavior increases- reinforcement

Positive reinforcement (reward received)

Negative reinforcement (punisher removed or avoided)

Behavior decreases- punishment

Positive punishment (punisher received)

Negative punishment (reward removed or lost)

Source: Ronald L. Akers (1985). Deviant Behavior: A Social Learning Approach (3rd ed.). Belmont, CA: Wadsworth, p. 45. Reprinted with permission of the author.

144 CHAPTER 5

individual to do the same thing again when confronted with similar circumstances. In other words, reinforcement causes a behavior to increase in frequency. Sometimes this occurs by rewarding the behavior (positive reinftrcement). However, a behavior may also be reinforced if engaging in it allows a person to prevent or avoid an unpleasant or painful stimulus (negative reinftrcement). Punishment also may be positive or negative but, unlike reinforcement, the goal of punishment is to weaken a behavior or to extin- guish it altogether. When an unpleasant or painful response (such as a slap) follows a behavior, the punishment is considered positive punishment. If a privilege or reward is taken away in response to a behavior, this is negative punishment.

Besides direct instrumental conditioning, we noted that behavior may also be developed or extinguished through imitation or modeling. Models may be real or fic- titious, and observers may be passive onlookers or active participants in activities with the models but, Akers (1985) cautions, modeling is "a more complicated process than 'monkey see, monkey do'" (p. 46). A number of factors influence the modeling process. For instance, one tends to imitate those one likes, respects, or admires. Imitation is also more likely if the observer sees the model being reinforced, if the model displays plea- sure or enjoyment, or if imitating the model in itself is being rewarded. An observer, though, may do the reverse or opposite of what a model does if he or she dislikes the model, sees the model punished, or if imitation of the model is being punished.

According to Akers and his colleagues (1979), "Whether deviant or conforming behavior is acquired and persists depends on past and present rewards or punishments for the behavior and the rewards and punishments attached to alternative behavior" (p. 638). This is the principle of differential reinforcement. "Differential reinforcement operates when both acts are similar and both are rewarded, but one is more highly rewarded. But differential learning of this kind is most dramatic and effective when the alternatives are incompatible and one is rewarded while the other is unrewarded" (Akers, 1985, p. 47).

Differential reinforcement is largely a social process; it takes place primarily in the context of interaction with others. Here Akers utilizes Sutherland's concept of differ- ential association. Those with whom one has the greatest contact-those who rein- force or punish a person the most-will have the greatest influence over that individual. Typically, these will be a person's family and friends, but may also include media personalities and institutional agents, such as school perS01ll1el, employers or co- workers, and government and law enforcement. These sources of differential rein- forcement also provide definitions of or give normative meanings to behaviors as either right or wrong. "Therefore, deviant behavior can be expected to the extent that it has been differentially reinforced over alternative behavior ... and is defined as desirable or justified" (AkeJ.;s et al., 1979, p. 638).

Social learning theory has the advantage that it is more readily testable than dif- ferential association theory. In fact, Akers (1998) presents an impressive array of stud- ies that have used field research methods and surveys to test hypotheses derived from the theory. The results of these studies are supportive of social learning theory. For example, Akers and his colleagues (1979) surveyed over 3,000 male and female teenagers from seven communities in the Midwest about a common form of adolescent deviance: the use of alcohol and drugs. They found, as the theory predicts, that the

Sociological Theories of Crime II: Crime and Social Processes 145

teenagers in their sample used drugs or alcohol to the extent that s.uch behavior was reinforced by their peers (especially those peers they most admired) and was defined by peers as desirable, or at least as justified as abstinence. However, the relationship between the differential reinforcement-differential association variables and alcohol and drug abuse were considerably weaker. Similar results were obtained in a longitudi- ' nal study of cigarette smoking among adolescents (Spear & Akers, 1988) and a study of drinking among the elderly (Akers et al., 1989).

These findings, however, raise one potential weakness in social learning theory. Specifically, most attempts to test the theory have examined relatively minor forms of deviation and offending. Some criminologists question whether the theory will be sup- ported in studies of serious criminal offending. Boeringer (1992) did test the theory in his study of rape and sexual coercion by male college students and obtained supportive findings (see also Boeringer et al., 1991). Nevertheless, it remains to be seen if the the- ory can adequately explain other forms of serious criminal offending.

Social learning theory has been praised for its practical implications in the areas of counseling and corrections. For example, many correctional facilities have adopted behavior modification treatment programs based on operant conditioning principles for at least some types of offenders. The difficulty here, however, is that evidence regarding the effectiveness of such "treatments" is contradictory, and some observers have raised serious ethical concerns about a number of these programs.6

One final criticism of social learning theory is that it does not address the ques- tion of how or where criminal or deviant definitions and labels originate. Akers (1985) admits, "The theory is ... incapable of accounting for why anyone or anything is socially defined as undesirable.... The theory does not say how or why the culture, structure, and social patterning of society sets up and implements certain sets and schedules of reactions to given behavior and characteristics" (p. 43). Although the the- ory recognizes that some reinforcers exert greater influence on individuals' behavior than others, it nevertheless overlooks the differential access of certain groups to a soci- ety's resources and rewards, as well as their differential power to escape punishment, to punish others, and to label others criminal or deviant. This, in fact, is a criticism that may be leveled against a majority of the theories we have discussed so far in this text, and it is an issue that will be raised again in this chapter and especially in Chapter 6. However, it must also be noted that Akers (1994, 1999) is optimistic about the possi- bility of integrating theories that do address these issues (e.g., social disorganization, anomie, and conflict theories) with social learning theory.

Differential Identification and Differential Anticipation,

The notion of differential identification, developed by Daniel Glaser (1956; 1973), derives not only from the principle of modeling, but also from reference group theory. People belong to and orient themselves toward many different groups. The groups with whom they identify are their reference groups, whether they are actually members of these groups or not. One may, for instance, aspire to membership in a group with higher social status than the groups to which one belongs, or identify with the lifestyle

146 CHAPTER 5

of a group portrayed in the media. In any event, individuals tend to judge themselves relative to the norms and values of these groups and try to emulate or model their behavior after those group members whom they most respect or admire. Criminality or deviance, then, results when an individual develops greater identification with mem- bers of criminal or deviant groups than with members of conformist groups.

The theory of differential identification is appealing for several reasons. First, it recognizes that people can learn from one another without having direct intimate con- tact or association. In addition, it concurs with our personal experiences and observ-a- tions. We know that people generally adopt particular images and incorporate them into their everyday lives. We are likely to recognize, for example, an "executive look" and a "grunge look"; each signifies a set of values and norms for behavior to which the individuals who adopt these models try hard to conform. At the same time, however, the theory is too simplistic in its depiction of identification; to reiterate Akers (1985), modeling is "a more complicated process than 'monkey see, monkey do'" (p. 46). Why are some groups more appealing to certain individuals than other groups? In what ways are one's choice of models limited or constrained? Why do some people who are repeatedly exposed to deviant or criminal images reject crime and deviance in favor of conformity?

To an extent, Glaser (1978) answered at least two of these questions by forward- ing a second theory, differential anticipation theory. In this approach, Glaser argues as Akers does, that people are likely to engage in behaviors from which they expect to obtain the greatest rewards and the least punishment. These expectations derive from three sources: differential learning, perceived opportunities, and social bonds. Differ- entiallearning refers to the process by which one develops tastes, skills, and rational- izations about whether he or she can best gratify himself or herself through criminal or noncriminal activities. Perceived opportunities reflect an individual's evaluation of his or her circumstances as well as the advantages and risks of engaging in criminal or alter- native activities. Social bonds, "both anticriminal and procriminal ... create stakes in conforming to the conduct standards of others so as to please rather than alienate them" (Glaser, 1978, p. 126). According to Glaser (1978), "Differential anticipation theory assumes that a person will try to commit a crime wherever and whenever the expectations of gratifications from it-as a result of social bonds, differential learning, and perceptions of opportunities-exceed the unfavorable anticipations from these sources" (p. 127).

The major weakness in differential anticipation theory is the same one that plagues differential association theory: It is difficult, at best, to test it. More specifically, how does a researcher measure differential anticipation? The theory implies that one cim add up an individual's anticipations unfavorable to law violation and subtract them from all the indiVidual's anticipations favorable to law violation and, if the result is pos- itive, a crime will be committed. Needless to say, this is impossible and, consequently, the theory is tautologiGal.

The development of differential anticipation theory was an attempt to integrate the central ideas of a variety of criminological perspectives into one general theory of crime. We recognize in it, for example, aspects of the strain theories that we discussed in Chapter 4, the principles of operant conditioning that we reviewed in this chapter,

Sociological Theories of Crime II: Crime and Social Processes 147

and elements of rational choice theory which was presented in Chapter 1. The empha- sis on social bonds comes from a theory that has enjoyed tremendous popularity in criminology: control theory.

Control Theory

Although criminologists historically have been interested in the issue of social control, specific theories of social control became especially popular during the 1940s and again in the 1970s and 1980s. There are, in fact, numerous theories that may be classified as control theories (see, for example, Briar & Pilliavin, 1965, Nye; 1958; Reiss, 1951; Toby, 1957).7

Virtually all of the theories we have discussed so far in this text focus on answer- ing the question, "Why do some people commit crimes?" In contrast, control theorists adopt a Hobbesian view of human nature; to them, everyone is basically criminal at heart. Everyone is equally motivated to commit crimes because fulfilling one's desires usually can be done most effectively, efficiently, and pleasurably by violating the law. Unlike subcultural and differential association theorists who focus on the problem of culture or normative conflict among diverse groups, control theorists assume that soci- ety is characterized by a single, conventional moral order. To the control theorist, then, the question criminologists must answer is "Why do people obey the rules oftheir soci- ety?" In answering this question, control theorists argue that it is a person's ties-or, depending on the individual theorist, a person's links, attachments, binds, or bonds- to conventional social institutions, such as family and school, that inhibit him or her from acting on criminal motivations (Liska & Reed, 1985).

Although there are many control theories, the one that undoubtedly has enjoyed the greatest popularity and had the greatest influence is that developed by Travis Hirschi in 1969. Consequently, it is to that brand of control theory that we will devote our attention here.

Hirschi's Control Theory

In his book, Causes ofDelinquency, Travis Hirschi (1969) presented his own version of control theory, along with an analysis of the empirical data he had gathered to test it. Like other control theorists, Hirschi begins with the assumption that "delinquent acts result when an individual's bond to society is weak or broken" (p. 18). Conversely, individuals with strong social bonds are unlikely to engage in delinquency. Hirschi specifies four elements of the social bond: attachment, commitment, involvement, and belief. '

The most important element of the bond is attachment. Attachment refers to an individual's sensitivity to the feelings of others. "If a person does not care about the wishes and expectations of other people-that is, if he [sic] is insensitive to the opin- ion of others-then he is to that extent not bound by the norms. He is free to deviate" (p. 18). Thus, attachment to others facilitates the internalization of society's norms and the development of a conscience.

148 C HAP T E R 5

Hirschi refers to the second element of the bond, commitment, as the "rational component in conformity." The underlying idea of commitment is that people develop a stake in playing by the rules. They invest their time, energy, money, emotions, and so on in pursuing a specific activity (such as getting an education, building a career, and establishing themselves as respected members of their communities). When consider- ing whether to commit a crime, individuals must factor in what they stand to lose if they get caught. "Most people, simply by the process of living in an organized society, acquire goods, reputations, prospects that they do not want to risk losing. These accu- mulations are society's insurance that they will abide by the rules" (p. 21).

Involvement is an opportunity element of the bond. The premise underlying involvement is straightforward: If a person is engrossed in conventional activities (such as studying, working, or playing a sport), he or she simply will not have time to partic- ipate in deviant or criminal activities. This is a commonly held view and one that pro- vides the rationale for many recreation-oriented delinquency prevention programs, such as the Police Athletic League (PAL).

The final element of the bond identified by Hirschi, belief, refers to the extent to which an individual believes he or she should obey the rules of society. As noted pre- viously, control theorists do not recognize variations in normative belief systems among different groups in society. Hirschi, like other control theorists, maintains that all individuals are socialized into a common value system. What Hirschi argues, how- ever, is that there is variation in belief in the moral validity of social rules. The less a person believes a rule should be obeyed-the lower the person's belief in the moral validity of the rule-the greater the likelihood that he or she will violate that rule. 8

In addition, Hirschi recognized that the four elements of the social bond are interrelated. Thus, an individual who is strongly attached to his or her parents and cares about their feelings will also be likely to express a strong belief in the moral validity of social rules. Likewise, an individual who has a high stake in conformity (that is, a high level of commitment) is also likely to be actively involved in conventional activities. Still, each element of the bond is analytically distinct and "should affect deviance uniquely and additively" (Matsueda, 1989, p. 430). Thus, if we hold three of the elements constant, the remaining element should, by itself, inhibit delinquent or criminal activity.

The majority of Causes ofDelinquency is devoted to Hirschi's empirical test of his theory. Hirschi surveyed a sample of more than 4,000 junior and senior high school boys in Contra Costa County in the San Francisco Bay area of California. Included in the questionnaire were items that measured the youths' relationships with their par- ents, teachers, and peers; their attitudes toward school; how frequently they engaged in such activities as working, studying, reading books, and dating; and whether during the past year they had stolen anything worth less than $2, stolen anything worth between $2 and $50, stolen anything worth more than $50, taken a car without the owner's permission, damaged or destroyed another person's property, or had beat up or deliberately hurt someone other than a sibling. These last six items together formed Hirschi's index of delinquency.

Most of Hirschi's findings support his control theory. In particular, he found that youths who had strong attachments to their parents and who cared about their teach-

Sociological Theories of Crime II: Crime and Social Processes 149

ers' opinions were less likely to be delinquent than youths without such ties, irrespec- tive of the delinquent activities of their friends. Indeed, Hirschi argued that attachment itself appears to be critical; those boys with strong attachments to their friends were less likely to be delinquent than unattached boys, even if their friends were delinquent. This does not mean that association with delinquent peers would never lead to delin- quency, but rather that such associations per se are not sufficient to cause delinquency. Delinquent boys, Hirschi found, have only weak and distant relations with others, including their peers.

Hirschi also found that the more committed a youth was and the stronger his belief in conventional morality and the legitimacy of law, the less likely he was to be delinquent. However, Hirschi's data were equivocal with respect to the relationship between involvement in conventional activities and delinquency. Specifically, Hirschi found that the boys who reported a high frequency of working, dating, reading books, watching television, and playing games were also more likely to be delinquent. Accord- ing to Hirschi (1969), the data indicate that control theory "overestimated the signifi- cance of involvement in conventional activities" and, further, did not take into account how some delinquent activities may contribute to an individual's self-concept or self- esteem, an issue to which we will return later in the chapter (pp. 230-231).

Among Hirschi's most important findings were that there was no relationship between social class an~ reported delinquency, and there was only slight variation in reported delinquency by race, despite the significant racial disparity in official arrest records. The attachment relationships he discovered held regardless of the boys' race or social class, and he found no evidence of a lower class subculture (see Chapter 4). Instead, he found that academic achievement was related to belief in conventional val... ues: Those boys who did well academically, whatever their objective socioeconomic position, held what Hirschi labeled "middle-class" values, whereas those who per- formed poorly academically held values that previously were identified as part of a "lower-class subculture." Concluded Hirschi (1969):

[T]he values in question are available to all members of American society more or less equally; they are accepted or rejected to the extent they are consistent or inconsistent with one's realistic position in that society. They are not, in other words, "class" values in the sense that they are transmitted by class culture.

In short, the data suggest, there are no groups of substantial proportions in American society that positively encourage crime in the sense that those belonging to the groups in question would prefer their children to follow their own rather than a conventional way of life. In fact, on the basis of the data presented here, it appears there are no groups of substantial proportions in American society whose values are neutral with respect to crime. (p. 230)

'.

These findings were good news to criminologists who objected to the class biases and racism of many sociological theories of crime. This is not to say, though, that Hirschi's control theory, as well as his research to test it, were accepted uncriti- cally in the criminological community. As noted earlier, Hirschi's work has drawn more ~ttention, both positive and negative, than any of the other perspectives that fall into the category of control theories. It certainly has been the most extensively tested

150 CHAPTER 5

(Stitt & Giacopassi, 1992). Let's turn our attention now to some of this empirical research.

Involvement, Belief, and Delinquency. Recall that Hirschi's hypothesis about the relationship between involvement in conventional activities and delinquency was not supported by his data. The young men in Hirschi's study who reported high levels of involvement also reported high levels of delinquent activity. One recent study under- took a reexamination of the involvement-delinquency relationship by focusing on ado- lescents' leisure activities (Agnew & Petersen, 1989). Leisure may indirectly influence all four of the bonds identified by Hirschi (that is, participating in enjoyable leisure activities with parents may strengthen adolescents' attachment to their parents). How- ever, it has its most obvious and direct impact on involvement. Adolescents whose free time is consumed by legitimate leisure activities will not have much chance to engage in delinquency.

Recognizing that some leisure activities may be considered more pleasurable than others by adolescents and that some leisure activities, rather than inhibiting delin- quency, may actually promote it, Agnew and Petersen considered not only types of leisure activities (that is, sports versus hobbies versus work or chores, and organized versus unsupervised activities), but also with whom the adolescents engaged in each activity (parents or peers) and the extent to which they liked each activity. They also examined the relationship between these variables and serious (stole a car) and minor (stole from a store) delinquency.

Among Agnew and Petersen's (1989) findings were that organized leisure activ- ities, passive entertainment, and noncompetitive sports were negatively related to delin-· quency, whereas "hanging out" with friends and unsupervised social activities with peers were positively related to delinquency (the former with total and serious delin- quency, the latter with total and minor delinquency). They also found that time spent in most favorite leisure activities with parents was unrelated to delinquency, but time spent in least favorite leisure activities with parents was positively related to delin- quency. In general, however, it appeared that the extent to which a leisure activity was liked was unrelated to delinquency.

Agnew and Peterson (1989) report that, "Overall, the leisure variables explain approximately 6 percent of the variance in total and minor delinquency, and 4 percent in serious delinquency" (p. 347). In other words, the leisure variables account for only a small amount of delinquent activity. Although the researchers argue that these figures "are comparable to the effects of variables measuring other institutional spheres, such as family and school ... [and that] the effect of certain leisure variables ... is as large or larger than t:1¥t of many traditional predictors of delinquency" (p. 347), such claims do not necessarily bolster their position. Looking at the glass half empty instead of half full, one might just as aggressively argue that most of the variables traditionally examined-school, family, leisure, or other measures-.are all rather poor predictors of delinquency. 9

Other researchers have concentrated on the belief component of the social bond. 'What is the empirical relationship between belief in the moral validity of conventional social rules and delinquent activity? Typically, belief is measured by asking a sample of

Sociological Theories of Crime II: Crime and Social Processes 151

respondents questions about honesty (such as "Is it a good thing to always tell the truth even though it may hurt oneself or others?").

Using such measures, several studies have found support for Hirschi's hypothe- sis that the stronger one's belief in the moral legitimacy of social norms, the lower one's participation in delinquent or criminal activities (Costello & Vowell, 1999; Krohn & Massey, 1980; Minor, 1984; Wiatrowski et al., 1981). Matsueda (1989) has criticized this research, however, because the samples from which the data were col- lected were usually small in size, limited to a single geographical area, and drawn from a "captive" population such as the student body of a school. Moreover, the research designs were cross-sectional, looking at attitudes and activities at one point in time; thus, they were unable to specify the causal ordering of any observed relationships. In other words, if the participants in such a study who were higWy delinquent were also low on measures of belief, the argument that involvement in delinquency led them to alter their support for conventional rules is at least as plausible as Hirschi's position. In any event, cross-sectional studies cannot demonstrate otherwise.

In his examination of the relationship between moral beliefs and delinquency, Matsueda (1989) developed a complex research design that utilized longitudinal data collected over a period of eight years from a national probability sample of boys who were in the tenth grade in the fall of 1966. Because other studies had shown that con- trol theory explains minor offenses better than it explains serious crimes-a point to which we will return shortly-Matsueda, using questions about honesty as indicators of moral belief, examined the incidence of five nonserious forms of youthful deviation: being suspended or expelled from school; skipping a day of school; running away from home; staying out past curfew; and fighting with parents. As Matsueda (1989) reports, "The results fail to replicate previous research which found support for social control theory's stipulation of the relationship between belief and deviance. Contrary to pre- vious results of cross-sectional studies, the effect of belief on deviance is relatively small and dwarfed by the effect of deviance on belief" (p. 428).10

Matsueda believes that participation in deviant activity may affect belief through two social psychological processes: cognitive dissonance reduction and self-perception for- mation. First, social psychologists tells us that when we act in a way that goes against an internalized belief, 'we experience psychological discomfort called cognitive disso- nance. One way to reduce this discomfort is to change the belief-or our support for the legitimacy of the belief-so that our belief system is consistent with our behavior. In self-perception formation, our behavior provides us with clues about who we are- about the elements of our self-identities, including our individual belief systems. "In short," Matsueda maintains, "belief and deviance should be specified as reciprocally causally related in a dynamic causal model" (p. 434).

Similarly, Agnew (1985) also reports that in his longitudinal research, all of the control variables (involvement, belief, commitment, and attachment measures) together explained' only 1 to 2 percent of the variance in future delinquency (see also Agnew, 1991; 1993; Paternoster et al., 1983). He attributes this, at least in part, to the impact that delinquency has on these variables, in particular on involvement, school attachment, and belief. Delinquency did not impact parental attachment, however-a finding supported by additional research, as we will soon see.

152 CHAPTER 5

School, Family, and Delinquency. The two remaining elements of the social bond are attachment and commitment. These elements are typically studied in terms of individuals' relationships with family members, teachers, and peers, as well as in terms of attitudes toward education. We will discuss school variables first and then take a look at research incorporating the more complex array of family variables.

It is a well-established fact that failure in school is related to delinquency and crime, but the precise nature of this relationship is not clearly understood (see also Chapter 2). Children who do poorly in school are more likely to engage in delin- quency and to be arrested as adults than those who do well in school (Jassim, 1989;, Rosenbaum & Lasley, 1990; Thornberry et al., 1985). Moreover, a good deal of crime, including serious violent crime, takes place at school (Applebome, 1996; Chandler et al., 1998; Kozol, 1991; Lively, 1997; Schwartz & DeKeseredy, 1997). Does early expo- sure to and involvement in delinquency interfere with children's learning activities, thereby lowering their motivation and achievement? Or, do children turn to delin- quency and crime when they reap few rewards in the classroom and grow increasingly alienated from school?

In Hirschi's model, children who care about what their teachers think of them and who value their teachers' opinions (indicating high attachment) are less likely to engage in delinquency. In addition, Hirschi (1969) postulates that children who have high educational aspirations, who value good grades, and who say they work hard in school (indicating high commitment) also are unlikely to pursue delinquent activities. While Hirschi himself found support for these hypotheses in his own research, have other criminologists replicated his results?

In a recent study ofJapanese youth, Tanioka and Glaser (1991) found fairly strong support for Hirschi's position. Japanese students who stated that they liked school had much lower delinquency rates than students who disliked school. However, the extent to which students cared about what their teachers thought of them was related only to rates of minor status offenses, not more serious crimes, and the relationship was weak. Nevertheless, when T anioka and Glaser examined all of the school attachment variables simultaneously, they explained 14 percent of the variation in delinquency rates, twice the rate explained by parental attachment variables. Similarly, the school commitment variables showed a strong inverse relationship with delinquency. The higher the stu- dents' educational aspirations and scores on an educational "Achievement Index," the lower their rates of self-reported delinquency (see also Tanioka, 1992).

There are, of course, particular features ofjapanese society that make it signifi- cantly different from the United States. Among these are a stronger educational sys- tem, greater government support of education in terms of both regulations and funding, a greater certainty of employment for students once they complete their schooling, a sub~bintially lower poverty rate, and a much higher frequency of extended family households in which not just two adults, but often three or four adults reside with the children (Upham, 1987; White, 1987). Such factors alone or in combination might account for greater attachment and commitment to school and low rates of devi- ation among Japanese youth.

In the United States, where control theory originated, the research findings on school attachment and commitment and delinquency have been somewhat less affirm-

Sociological Theories· of Crime II: Crime and Social Processes 153-

ing. For instance, in their analysis of self-report data from 1,508 high school students, Rosenbaum and Lasley (1990) discovered that increases in positive attitudes toward school and educational achievement resulted in significant reductions in delinquent activity only fOr the males in their sample. Although we will examine the issue of sex dif- ferences more carefully in the next section, suffice it to say here that in Rosenbaum and Lasley's research, a commitment to success in school had only a minimal impact on female delinquency.

In a study of a school-based delinquency prevention program, Gottfredson (1986) examined the effects of several factors that should have led to increased 'school attachment and commitment among participating students. The factors included: involvement of school staff, students, and community members in planning and imple- menting a school improvement initiative; changed disciplinary procedures; an enhanced educational program that included activities designed to raise achievement and create a more positive atmosphere in the school; and special services for high-risk students designed to improve their self-concepts, increase their success experiences, and strengthen their bond to school. The program slightly reduced delinquency and misconduct among the general student body at participating schools, but not among high-risk students. The findings show that high-risk students' commitment to educa- tion-as measured by dropout, retention, and graduation rates, as well as by standard- ized achievement test scores-did increase; however, their rates of delinquency and misconduct simply did not decline.

Liska and Reed's (1985) research may help to explain such anomalous findings. Liska and Reed hypothesized that there is a reciprocal relationship between attach- ment and delinquency; that is, they argued that delinquency is as likely to affect attach-: ment as attachment is to affect delinquency (see also Matseuda & Anderson, 1998). In fact, their findings showed that most of the negative relationship between school attachment and delinquency that other studies have found is the result of delinquency's effect on school attachment, not vice versa. At the same time, however, their findings did support Hirschi's hypothesized relationship between parental attachment and delinquency: "most of the observed negative relationship between parental attachment and delinquency comes about because of the effect of parental attachment on delin- quency" (p. 537).

Liska and Reed explain these findings by noting first that attachment between parents and children is less conditional on the behavior of either party than attachment between teachers/school personnel and students. Moreover, as we have already noted, much juvenile misconduct, including criminal violations, takes place in school or on school grounds. "This leads to reactions by teachers and school administrators, which in turn decrease,school attachment" (p. 557). Also, in response to a question we raised at the outset of this section, Liska and Reed argue that "adolescents involved in delin- quency simply have less time for school; thus, delinquency, independent of teacher reactions, may decrease school attachment" (p. 557).

Liska and Reed rightly conclude that, "Generally, it is all too clear that the causal structure underlying the relationship between social attachment and delinquency is not as simple as implied in theories of social control" (p. 559). Their findings indicate that it is strong attachment to parents that lowers delinquency, but that delinquency

154 CHAPTER 5

lo~ers school attachment which, in turn, may weaken parental attachment. Their find- ings point to parents, not schools, as the major institutional sources of social control. Do other researchers agree with them?

In examining how children's relationships with their parents might affect their delinquency rates, one must also consider a variety of methodological issues. Studies that attempt to measure parental attachments often use adolescents' reports of their parents' behaviors or attitudes. Can we be certain that children correctly perceive, accurately recall, and honestly report their parents' behavior or attitudes (McCord, 1991)? In addition, researchers have operationally defined parental attachments in a wide variety of ways, including questions about "affection and love, interest and con- cern, support and help, trust, encouragement, lack of rejection, desire for physical closeness, amount of interaction or positive communication, and 'identification,' " as well as with variables measuring direct controls, such as monitoring or supervision and punishment or disciplinary techniques (Rankin & Wells, 1990, p. 142). The question, of course, arises as to whether all of these variables are actually measures of parental attachment. Nevertheless, the correlation between parent-ehild relationships and crime is one of the most extensively researched in the criminological literature.

Rankin and Wells (1990) have examined the effects of both indirect and direct parental controls on the behavior of male children. Indirect control through parental attachment was measured by two indices: (1) identification, which included questions regarding how much the youth likes his mother and father, how close he feels to each parent, how much he wants to be like each parent, and how much time he spends with his father; and (2) positive communication, which included items about how much influence the youth has in family decisions, how often his parents listen to his side of. things in arguments, how often his parents talk over important family matters with him, and how often his parents respond fairly and reasonably to his requests. Direct control of the youth by parents was measured in terms of: supervision (the extent to which parents determined their child's friends and activities-with whom he socializes, where he may go, and what he may do); strictness; contingency of punishment (how often the parents completely ignore instead of punish the child's misbehavior); and strength of punishment (how often parents use yelling, slapping, threats of slapping, and withdrawal of privileges as punishments). Their sample included 1,886 boys who had participated in the Youth in Transition Study, the same data source used in the Liska and Reed (1985) study we discussed earlier.

In general, Rankin and Wells's findings confirm that both direct controls and indirect controls (that is, attachment) are negatively related to delinquency (see also Wiatrowski et al., 1981). However, these researchers found that as parental discipline increases, delinquency does not necessarily decrease. "Punishment that is too strict, frequent, or severe can lead to a greater probability of delinquency regardless of parental attachments. That is, a strong parent-ehild bond will not lessen the adverse impact of punishment that is too harsh" (p. 163). Rankin and Wells (1990) found that while punishment that is consistent is negatively related to delinquency, punishment that is severe is positively related to delinquency. In terms of strictness, there was a curvilinear relationship; medium levels of parental strictness were most effective in

Sociological Theories of Crime IT: Crime and Social"Processes 155

lowering delinquency, whereas both low and high levels of parental strictness increased delinquency (see also Straus & Sugarman, 1997).

Similar studies that have focused on child-rearing techniques, monitoring and other forms of direct parental control over children's behavior, and quality of family interaction, have arrived at similar findings. McCord (1991), for instance, reports that mothers judged to be competent (self-confident, affectionate, consistently nonpunitive in their disciplinary style, and providing leadership to their children) seemed to be able to insulate their children against criminogenic influences even in high-crime neigh- borhoods. Boys who had competent mothers and who grew up in households in which there were high family expectations (children were expected to do well in school) had low rates of juvenile delinquency which, McCord argues, reduces their probability of adult criminality (see also Larzelere & Patterson, 1990; Laub & Sampson, 1988). However, McCord also found that, compared with mothers' influences, fathers' influ- ences are not significant early on, but increase as boys grow older. According to McCord (1991):

Fathers who interact with their wives in ways exhibiting high mutual esteem, who are not highly aggressive, and who generally get along well with their wives provide models for socialized behavior. Conversely, fathers who undermine their wives, who fight with the family, and who are aggressive provide models of antisocial behavior. Both types of fathers, it seems, teach their sons how to behave when they become adults. (p. 412)

Thus, fathers, unlike mothers, appear to have more of a direct effect on adult crimi- nality than on juvenile delinquency, at least among males. This, McCord maintains, indicates that the causes of crime are not identical across age groups, a point that will be raised again later in this chapter.

Before moving on to other issues accounted for (or overlooked) by control the- ory, one final dimension of parental attachment deserves our attention: family struc- ture. The "broken home hypothesis" is the notion that children from single-parent homes are more likely to become involved in delinquent activities than children from tWo-parent homes. The common rationale offered for this hypothesis is that one par- ent is simply less effective in monitoring children's behavior than two parents. A sec- ond rationale, however, derived from the 1965 Moynihan Report in which it was argued that delinquency rates are higher among Black youth than among White youth because of "a tangle of pathology" growing out of life in matriarchal (female-headed) house- holds, high rates of births to unwed mothers, high unemployment, and differential socialization. Thus, this perspective attempted to explain "the joint relationships between race, broken homes and delinquency" (Matseuda & Heimer, 1987, p. 826).11

In general, it can be said that there is little evidence in support of the broken home hypothesis, although it has managed to repeatedly find its way into public pol- icy debates over the past several decades (see, for example, "OJJDP Model Programs 1990," 1992). Although official statistics show that those youth most likely to be processed through the criminal justice system do come from broken homes, findings from self-report studies indicate little or no relationship between family intactness and

156 CHAPTER 5

delinquency (Johnson, 1996). Given that single-parent, female-headed households are one of the fastest growing types of family structures in the United States, this is a sig- nificant finding.

There are, however, several factors that complicate the family structure-delin- quency relationship. Van Voorhis and her colleagues (1988), for instance, found that family structure seemed to be related only to incidence of status offenses andwas unre- lated to overall home quality. Overall home quality, though, regardless of family struc- ture, was a stro~g predictor of delinquency, but the analysis failed to identify specific aspects of family functioning that contributed most to delinquent activity.

Matsueda and Heimer (1987) have addressed the race-broken homes-delin- quency relationship. They found that although broken homes have a greater effect on delinquency rates among Black youth than among White youth, this effect, along with the effects of attachment to parents and peers, were mediated by learning definitions favorable to delinquency. In other words, their work appears to support differential association theory more than control theory. Matsueda and Heimer argue, however, that differences in crime rates by race should be examined in the context of the "his- torical emergence of social and economic structures that give rise to distinct racial pat- terns of social organization" and that reflect a history of racial discrimination (p. 837)-a point that we will take up again later in this chapter and in Chapter 6.

It is also important to note that there are different types of intact families as well as different types of single-parent families. Many researchers, for example, have docu- mented negative effects on children when they live in homes where parents are con- stantly arguing (Barber & Eccles, 1992; Hetherington et al., 1989; Wallerstein & Blakeslee, 1989). In addition, Johnson (1986) found that family structure was unrelated to frequency or seriousness of delinquency, except in mother/stepfather homes where boys had an unusually high involvement in delinquent activity. Johnson, however, found some important sex differences: Although boys, but not girls, in mother/stepfather house- holds reported high levels of delinquency, officials were more likely to respond to the misbehavior of children-especially female children-living in mother-only families. Thus, the myth of the broken home-delinquency relationship appears prevalent among law enforcers, which may account, at least in part, for the higher official delin- quency rates of both boys and girls from single-parent female-headed households.

Johnson's research highlights the importance of considering sex differences when evaluating any theory of crime. Most tests of control theory have utilized all-male samples, as Hirschi's (1969) study did. In addition to sex, there are several other factors worth considering in our assessment of control theory, such as age and seriousness of offense.

'. Sex, Age, and Other Factors Affecting Deviation. As we have already noted, Hirschi (1969) tested his theory by surveying a sample of high school boys, reporting in a foot- note that in the analysis, "the girls disappear" (p. 36). Several subsequent tests of social control theory have included females in their samples and, like Johnson (1986) cited previously, these studies reveal significant sex differences in their results. It has been argued by some that control theory actually does a better job of explaining female delinquency than male delinquency (see Jensen, 1990; Krohn & Massey, 1980). Cem-

Sociological Theories of Crime IT: Crime and Social Processes 157

kovich and Giordano (1987), though, also report that the extent to which control the- ory accounts for sex differences in delinquency depends to some degree on how the elements of the social bond are operationalized. For instance, they used a multidimen- sional measure of attachment and found that "although the total explained variance is similar among males and females, the relative importance of the variables is not" (p. 315). For males, the dimensions that were most strongly associated with delin- quencywere control and supervision, intimate communication, and instrumental com- munication, whereas for girls, the most important dimensions were identity support, conflict, instrumental communication,· and parental disapproval of peers. According to Cernkovich and Giordano (1987), "This seems to suggest that while family attachment is important in inhibiting delinquency among all adolescents, the various dimensions of this bond operate somewhat differently among males and females" (p. 315).

Researchers have also found that attachment to peers operates differently for males and females. Hirschi (1969) made the controversial argument that attachment to peers, even if peers were delinquent, inhibited delinquency. Others' research has not suP?orted this position. Giordano and her colleagues (1986), for instance, found that boys in peer groups feel considerable pressure from their friends to engage in risk- taking behavior, including delinquency. Girls' groups, in contrast, function differently; girls interact with friends in ways that encourage self-disclosure and foster intimacy. Thus, while boys' attachment to peers may promote delinquency, girls' attachment to peers may inhibit it (see also Fordham, 1996; Gilligan et a1., 1995; Matsueda & Ander- son, 1998).

Farnworth (1984) has found that among African American youth, problems in school predict delinquency better than family problems for girls; but for boys, family problems seem to be better predictors than school problems. Rosenbaum and Lasley r (1990) also examined the school-delinquency relationship and found several significant sex differences. They report, as we noted earlier, that positive attitudes toward school and school achievement inhibit delinquency more for boys than girls. Boys, they argue, appear to have a greater stake in future success predicated on strong school perfor- mance than girls do. However, among those girls who did very well in school, "atti- tudes toward achievement and school seem to have the same insulating effect from delinquency as for males" (p. 510). Rosenbaum and Lasley also found that increased involvement in school activities, as well as positive attitudes toward teachers, inhibited delinquency more for girls than boys. Finally, Rosenbaum and Lasley found social class to be an important intervening variable, with differences in social class producing more significant changes in the school-delinquency relationship for females than for males. They conclude on the basis of this finding that "school conformity will be instilled socially in fem~les and in middle/upper class youths more strongly than in males and in youth from the lower class" (p. 511).

Several researchers have argued that findings such as these, which show differ- ential access to opportunities and rewards on the basis of sex and social class, point to the need to examine power differences between the sexes and classes when attempting to explain crime and delinquency (see Chesney-Lind, 1997). This will be a major theme in Chapters 6 and 7. For now, however, let us conclude our evaluation of con- trol theory by considering two other factors: age and seriousness of offense.

158 CHAPTER 5

Hirschi concentrated on minor delinquent acts committed by male high school students, and most subsequent tests of his theory have taken a similar approach. A number of criminologists, however, have argued that this focus is too limited. Can con- trol theory account for offending by adults as well as by juveniles? And is control the- ory, as Matsueda (1989, p. 432) has argued, best at explaining only "trivial impulsive deviant acts, such as status offenses ... rather than serious event-like offenses, such as crimes against persons"? "Whitehead and Boggs (1990) offer an answer to both ques- tions. They studied recidivism among a sample of adult felony probationers in New Jersey, operationalizing attachment by marital status and offender'sliving arrange- ments; commitment by years of schooling and percentage of time employed during the two years prior to the current offense; and involvement by whether the offender was either in school or employed at the time of the offense. 12 Whitehead and Boggs report that their "most dramatic finding is the lack of impact of most of the control theory variables.... This analysis was hard pressed to find significant effects of control the- ory variables" (p. 4). Only two control theory variables-those used to measure com- mitment-were significant in explaining recidivism by the adult felony probationers. Most of the recidivism was accounted for instead by legal variables (previous convic- tions) and by demographic variables (race and age).

Other studies have obtained support for the applicability of control theory to adults for certain types of offenses, such as white-color crimes, and when the adult offenders were misdemeanants, not felons. Agnew's (1985) research further shows that apart from the age of the offender, control theory appears to apply only to minor forms of delinquency; "the explanatory power of the theory diminishes as we focus on more serious forms of delinquency" (p. 58).

The wealmesses of control theory that we have discussed here and in previous sections are serious, but not fatal. The theory, in fact, may be a very good one, but with limited utility. Certainly, the research conducted thus far has done more to fuel this debate than to settle it. In the meantime, several criminologists, including Hirschi himself, have developed revisions of control theory with wider applicability. Let's look at these theories now.

Tittle's Control Balance Theory

In 1995, Charles Tittle published Control Balance, in which he offers an important revi- sion of traditional control theory. Tittle accepts Hirschi's proposition that control is the major component of conformity, but he argues that it is not control per se that counts, but rather maintaining a balance between the amount of control one is subject to at the hands of others and the amount of control one can exercise over others. This relationship can be expressed as a ratio, where the numerator is the control to which one is subject and the denominator is the control one exercises. If the numerator exceeds the denominator, or vice versa, a control imbalance occurs. Control imbalances can result in deviant behavior, including in some cases crime.

Tittle identifies two types of control imbalances. The first type of control imbal- ance is a control deficit, whereby the amount of control to which one is subjected exceeds the amount of control one can exercise over others. A control deficit, Tittle believes,

; ~ !

Sociological Theories of Crime II: Crime and Social Processes 159

I

i J

I

t I

I

~

I

produces repressive deviance. There are three forms of repressive deviance, each of which helps individuals escape control deficits and restore balance to their control ratios. The first type of repressive deviance is predation, which involves physical vio- lence and is intended to harm others. Predation includes many criminal acts, such as sexual assault and robbery, but also includes many types of property crime, such as theft. The second type of repressive deviance is defiance, or deviation that challenges dominant norms but typically does not inflict harm on others. Included in the category of defiance are behaviors such as truancy, having sex with multiple partners simulta- neously, and vandalism. The third type of repressive deviance is submission, which involves "passive, unthinking, slavish obedience to the expectations, commands, or anticipated desires of others" (Tittle, 1995, p. 13 9). Tittle includes in this category repressing other people to please someone perceived to be more powerful (e.g., writ- ing racial slurs on the dorm room door of a Latino student to curry favor with mem- bers of the all-White, elite fraternity on campus). This category also includes allowing oneself to be physically abused, humiliated, or degraded (e.g., being tied up and whipped for the sexual pleasure of another person).

The second type of control imbalance, which Tittle conceptualizes as a control surplus, occurs when the amount of control a person exercises over others exceeds the amount of control others impose on her or him. According to Tittle, a control surplus leads to a different kind of deviance, autonomous deviance, which helps the individ- ual further extend control over others and thereby increase her or his control surplus. There are three kinds of autonomous deviance. The first kind is exploitation, which is an indirect form of predation that includes such behaviors as hiring someone to injure a rival before a competition. The second kind of autonomous deviance is plunder, which is engaged in by individuals or organizations who want to further their own goals while ignoring or trampling the rights and safety of others. For example, plun- der occurs when a corporation knowingly sells a faulty product overseas after it was banned for sale in the United States just so it can extend its markets and reap profits. Tittle also includes genocide under the heading plunder. The third type of autonomous deviance is decadence, irrational acts engaged in on a whim or the spur of the moment, such as humiliating another person for one's own pleasure.

Clearly, Tittle's control balance theory is designed to cover a wide variety of behaviors, not all of which are criminal. In fact, Tittle is careful to point out that a con- trol imbalance does not inevitably produce deviance of any kind. Tittle draws on a number of theories to specify which conditions are likely to result in deviance when a control imbalance occurs. Whether a control imbalance produces a deviant outcome depends on an individual's predispositional motivations, situational motivations, con- straint, and opportunity. Each of these, like control itself, exists along a continuum, varying in form, frequency, and intensity. Tittle conceives of predispositional motiva- tions as natural, the products of one's innate physical and psychological needs and desires, including what he characterizes as the "almost universal" desire for autonomy. Even if one is predisposed to crime or deviance, one must become aware of the con- trol imbalance, and this awareness develops from situational provocations, such as being turned down for a date, fired from a job, or insulted by a friend, relative, or stranger. Although Tittle sees predispositional and situational motivations as being

160 C HAPTE R 5

strong precipitators of deviance, he nevertheless believes that deviance still may not occur if the individual faces a high internal or external level of constraint. A quick- tempered person who has just been cut off on the highway by another driver may be motivated to run that fellow driver off the road, but the police car in the rearview mir- ror constrains him or her. "[C]onstraint refers to the actual probability that potentially controlling reactions will be forthcoming" (Tittle, 1995, p. 167). Of course, even if one is motivated and constraint is low, deviance still may not occur if there is no opportu- nity to deviate. Realistically, though, Tittle recognizes that the opportunity for some kind of deviation is almost always available.

Tittle's control balance theory gives us much food for thought. One strength of the theory is that, unlike many of the theories we have discussed so far in this text, con- trol balance theory does not concentrate on traditional street crime, but explains as well "hidden offending;" such as dating violence, white-collar crimes (e.g., fraud hy computer, insider stock trading), and organizational and governmental crimes (e.g., sale of unsafe or banned products, human rights violations). Like all theories, however, control balance theory must be subjected to the test of empirical research to determine its validity, a task made all the more difficult by a fact about which Tittle (1997) him- self cautions us: There are no secondary data sets currently available that will allow us to calculate anyone's control ratio. Tests of control balance theory will have to be undertaken "from scratch," with researchers designing appropriate research instru- ments and systematically collecting and analyzing original data. At least one study has already made an attempt, though limiting itself to a test of two types of repressive deviance (predation and defiance) using a questionnaire administered to a sample of college students (piquero & Hickman, 1999). In this study, Piquero and Hickman found support for control balance theory, but contrary to Tittle's predictions, both control deficits and control surpluses appeared to lead to predation and defiance. Piquero and Hickman suggest that perhaps control balance theory is incorrect in pre- dicting that one type of imbalance will lead to a specific type of deviation, but correct in predicting that an imbalance can lead to deviation.

A number of criminologists have raised concerns about the basic propositions of control balance theory, ranging from the difficulty in specifying cut-offs for small, medium, and large imbalances, to overlap in the types of deviance specified in the the- ory, to the theory's neglect of the fact that specific behaviors are evaluated differently by different individuals and groups of people (see Braithwaite, 1997; Jensen, 1999; Savelsberg, 1999; for responses, see Tittle, 1997; 1999). Given the newness of the the- ory, we hope this debate will motivate criminologists to develop useful strategies for empirically testing the control balance perspective (see, for example, Curry, 1999).

'.

Self-Control and Crime

We have already noted that one of the most consistent findings in the research litera- ture is that relationships between parents and children are strongly associated with delinquent and criminal behavior. Unfortunately, the precise nature of this association remains unclear. Family structure appears to be far less important than emotional ties

between members of all types of families. Various child-rearing methods, such as the degree and form of discipline that parents use in responding to their children's misbe- havior, also seem to playa prominent role.

In their book, A General Theory of Crime, Michael Gottfredson and Travis Hirschi (1990) zero in on ineffective child rearing as the primary cause of all types of deviance, from smoking to victimization to unwanted pregnancy to white-collar crime. Most of the theories we have examined so far in this chapter maintain that crime is learned, but Gottfredson and Hirschi argue that crime is a product of a lack of social- ization or learning. They accept the classical assumption that "crime is the natural con- sequence of unrestrained human tendencies to seek pleasure and avoid pain" (p. xiv). Although they state that their theory also incorporates aspects of modern positivism, especially positivistic research on the role of the family in crime causation, Gottfred- son and Hirschi place themselves squarely in the tradition of the classical school of criminology and the rational-choice model.

To fully understand Gottfredson and Hirschi's (1990) position, we must con- sidEr their definition of crime. Crimes, they tell us, are "acts of force or fraud under- taken in pursuit of self-interest" (p. 15). According to Gottfredson and Hirschi, all crimes share certain common characteristics: They provide easy and immediate grat- ification of desires; they are exciting, risky, and thrilling; they offer few, if any, l~:>ng­ term benefits; they require little skill, planning, or specialized knowledge; and they often cause pain or discomfort for the victims. What kind of person, then; would engage in such activities? According to Gottfredson and Hirschi, it would be a person with low self-control.

People who deviate-whether that deviation is drinking too much or driving recklessly, assaulting someone or embezzling from an employer-lack self-control. Compare Gottfredson and Hirschi's list of the characteristics of crime presented pre- viously with their list of the characteristics of people with low self-control. People lack- ing self-control: have a concrete "here and now" orientation and have difficulty deferring gratification; tend to lack diligence, tenacity, or persistence in a course of action; are adventuresome, active, and physical; tend to have unstable marriages, friendships, and employment histories, and are uninterested in or unprepared for long- term occupational pursuits; neither possess nor value cognitive or academic skills, nor do they necessarily have good manual skills; tend to be self-centered, and indifferent or insensitive to the needs of others. When individuals low in self-control are presented with opportunities to commit crimes, they more likely than not will commit the crime.

Low self-control, as we have already implied, is a result of ineffective or inade- quate socialization. Gottfredson and Hirschi maintain that in order for effective social- ization to occur and, consequently, for strong self-control to develop, someone who cares about the child must be responsible for meeting three basic conditions: (1) mon- itoring the child's behavior; (2) recognizing when the child deviates; and (3) punishing the deviation. Since parents or guardians are typically a child's first socializers, the two theorists lay the blame for inadequate socialization with them. It's not that parents or guardians prefer their children to be unsocialized or to lack self-control; Gottfredson and Hirschi (1990) "rule out in advance the possibility of positive socialization to unso- cialized behavior (as cultural or subcultural deviance theories suggest)" (p. 98). Rather,

1 I

Sociological Theories of Crime II: Crime and Social Processes 161

162 C HAP T E R 5

certain factors may inhibit or prevent parents or guardians from sufficiently socializing their children:

First, the parents may not care for the child (in which case none of the other conditions would be met); second, the parents, even if they care, may not have the time or energy to monitor the child's behavior; third, the parents, even if they care and monitor, may not see anything wrong with the child's behavior; finally, even if everything else is in place, the parents may not have the inclination or the means to punish the child. (Gott- fredson & Hirschi, 1990, p. 98)

Gottfredson and Hirschi cite in support of their position various studies that show, for example, that parents who are hostile or indifferent toward their children are more likely to have children who become delinquent. They further point out that crime appears to be concentrated in certain families not because of heredity or explicit parental encouragement, but because parents who are themselves criminal-that is, who themselves lack self-control-are not adept at instilling self-control in their chil- dren. These parents may not even recognize criminal behavior in their children and tend to be lax in discipline or to use punishments that are "easy, short-term, and insen- sitive" (Gottfredson & Hirschi, 1990, p. 101).

Ineffective socialization may be especially likely to occur in families with large numbers of children, in single-parent and stepparent families, and in families in which the mother works outside the home. In the first instance, parents often do not have the time or energy to adequately monitor and discipline their children. The single parent shares this problem, but it is compounded by the fact that the single parent has less psychological and social support than co-parents have. The problem in stepfamilies is that stepparents are less likely to have "parental feelings" toward their stepchildren. In households where the mother works outside the home, adequate supervision is again the concern.

Of course, it might be argued that if parents are inadequate socializers, teachers and school personnel can serve as substitutes, especially since the school is a social institution officially charged with socializing children. Unfortunately, according to Gottfredson and Hirschi (1990), "The evidence suggests ... that in contemporary American society the school has a difficult time teaching self-control" (pp. 105-106). They attribute this not to a lack of adequate educational resources or to poorly paid and often poorly trained teachers, but rather to a lack of cooperation and support from parents who have already failed in their socialization duties. Although the school may have some positive impact on some students, "self-control differences seem primarily attributable to family socialization practices. It is difficult for subsequent institutions to make up for de.ficiencies, but socialization is a task that, once successfully accom- plished, appears to be largely irreversible" (Gottfredson & Hirschi, 1990, p. 107).

Does this mean then that if a child is not adequately socialized by his or her par- ents, all is lost, that he or she will be deviance-prone and there is nothing anyone can do to prevent or halt the behavior? What are the implications of Gottfredson and Hirschi's position for public policy? First it should be said that Gottfredson and Hirschi do not claim that crime is an inevitable outcome of low self-control. While they maintain that criminality is stable over time-that is, there is little or no change among

Sociological Theories of Crime II: Crime and Social Processes 163

1 f

\

r I f

individuals from high self-control to low self-control-they also point out that social- ization is ongoing throughout an individual's life. Consequently, the number of offenders or deviants declines as a cohort ages. "Even the most active offenders burn out with time, and the documented number of 'late-comers' to crime, or 'good boys gone bad,' is sufficiently small to suggest that they may be accounted for in large part by misidentification or measurement error'" (Gottfredson & Hirschi, 1990, pp. 107- 108). A decrease in the number of crimes committed by people with low self-control occurs as they get older in large part because of the exigencies of aging and, to a much lesser extent, because of ongoing socialization.

In terms of public policy, Gottfredson and Hirschi eschew law enforcement and crime control programs that involve building more prisons, employing more police, or enacting gun control laws. Rather they argue that crime prevention efforts should be targeted at parents and other adults with responsibility for raising children and should concentrate on teaching them how to be alert to and recognize signs of low self-control and how to punish children when they display these signs. As Gottfredson and Hirschi (1990) put it:

We offer an alternative view, a view in which the state is neither the cause nor the solu- tion to crime. In our view, the origins of criminality or low self-control are to be found in the first six to eight years of life, during which time the child remains under the con- trol and supervision of the family or a familial institution. Apart from the limited ben- efits that can be achieved by making specific criminal acts more difficult, policies directed toward enhancement of the ability of familial institutions to socialize children are the only realistic long-term policies with potential for substantial crime reduction. (pp.272-273) .

Gottfredson and Hirschi's argument is a seductive one in a society such as ours that places a great deal of faith in the power of socialization to bring about social change. But to what extent has the theory been empirically verified?

A General Theory of Crime or a Limited One?

Recall that Gottfredson and Hirschi maintain that theirs is a general theory of crime, capable of explaining all forms of deviation, not just criminal offenses. All crimes are pretty much the same, they tell us; crimes are spontaneous acts that require no spe- cialized knowledge and that yield short-term, simple gratifications. Much of Part ill of their book is devoted to reviewing research that supports their claims (see also Hirschi & Gottfredson, 2000; LeBlanc & Kaspy, 1998), and this is where much of the criticism of Gottfreds(~mand Hirschi's argument has centered. The most frequent criticisms of Gottfredson and Hirschi's work claim that they tailored the facts of crime to fit their theory and that they selectively overlooked abundant evidence that does not support their position (Geis, 2000).

Polk (1991), for instance, takes issue with their characterization of homicide as being of two basic varieties: (1) those that result from a heated argument that goes too far, involving two people who know one another and who have argued frequently in the past; and (2) those that occur during a robbery or, less often, during a miscalculated

II I1

Ii

164 CHAPTER 5

burglary in which the victim turns out to be home. According to Polk's own extensive research in this area, however, homicides are far more varied: Many are carefully planned or premeditated; others include "parents starving a child to death in a belief that fasting would cure the child's cold, young girls who cannot face the reality of preg- nancy and so the infant dies at birth from neglect," those who kill because "voices" tell them they must, and "criminals who kill a friend because of the possibility that the friend may give vital evidence against them in a forthcoming trial" (polk, 1991, p. 577). In fact, Polk's research leads him to conclude that there probably is no "typical" homi- cide. Some homicides are like those characterized as typical by Gottfredson a.Q.d Hirschi but, as Polk (1991) points out, "Gottfredson and Hirschi do not offer us a general the- ory of some crime, it is a general theory of crime. Ifthe actual empirical nature of crime is not as Gottfredson and Hirschi describe it, then the theory must collapse" (p. 577).

Similar critiques have been made with respect to Gottfredson and HirsdV's analysis of white-collar crime. The two theorists focus on white-collar crime because they correctly note that positivistic theories have failed to adequately explain it and that many criminologists have come to see it as a unique form of offending that is different from other types of crimes. In contrast, they see white-collar crime as relatively uncommon in occurrence, but as conforming to the same age and race distributions as other crimes, and sharing other crimes' characteristics: spontaneity, quickness, requir- ing no specialized knowledge, yielding limited profits for offenders. They rely on FBI Crime Report data (see Appendix, pp. 243-245) to specify the types and incidence of offenses that constitute the category white-collar crime, and thus focus on embezzle- ment, fraud, and forgery.

There are several major wealu"1esses with such an analysis, not the least of which stems from their operationalization of white-collar crime. The crimes they have cho- sen do readily fit their definition of crime in general, but they hardly encompass the full array of offenses that constitute white-collar crime. In particular, they overlook organizational, corporate, and governmental offending. The VCR data are biased toward minor, low-level offenses, particularly because they reflect arrests; simple offenses are more likely to be detected and result in successful prosecution, whereas more sophisticated crimes are likely to be underrepresented in the VCR. Such crimes include terrorism for political goals, securities fraud, antitrust violations of pollution laws. Numerous studies show that these types of crimes are complex, involve a high level of technical detail, and are difficult to detect and prosecute (Calavita & Pontell, 1983; Geis, 2000; Reed & Yeager, 1996). Reed and Yeager (1996) further point out that a number of sociolegal processes operate in such a way that corporate offenses are screened out of the criminal justice system. These include interagency disagreements and rivalries, qut especially the ability of powerful corporations and governments to influence the very definition of lawlessness and compliance.

One may also question Gottfredson and Hirschi's characterization of white- collar crime on at least six other major points, all of which revolve around the inclusion of corporate and business offenses in the definition of this type of crime. First, is white- collar crime really uncommon or as infrequent as these theorists claim? Although it is difficult at best to know for certain how many white-collar offenses are committed each

r I

Sociological Theories of Crime IT: Crime and Social Processes 165

year, given the difficulties with regard to detection raised previously, some researchers argue that such offenses are at least as common, and perhaps more common, than con- ventional street crime (Reed & Yeager, 1996).

Second, are white-collar offenses quick and spontaneous acts? Again, available evidence indicates that, to the contrary, some corporate offenses are carefully planned and executed over an extended period of time in the interests of continued or future viability of a single business or an entire industry. Long-term price-fixing conspiracies illustrate this point well. These offenses are usually motivated by the rational pursuit of corporate goals rather than by an impulsive pursuit of immediate self-gratification by an undersocialized individual (Reed & Yeager, 1996).

This raises a third issue: Do white-collar crimes typically yield low profits for offenders? If one examines only the offenses Gottfredson and Hirschi consider to be white-collar crimes, the answer is yes. In contrast, if one includes corporate, business, and government offenses, the answer is an unequivocal no. Consider, for example, the savings and loan fraud of the 1980s, which is expected to cost taxpayers between $300 billion and $473 billion by 2021 (Calavita & Pontell, 1993).

Fourth, can anyone, without any specialized knowledge, commit a white-collar offense? The answer, of course, is that it depends. No specialized knowledge is needed to write a bad check or to take money out of a cash register. However, it certainly can- not be claimed that specialized knowledge is not required to commit such crimes as computer fraud or insider stock trading. It seems equally unlikely that such white- collar offenders are interchangeable in this sense with street offenders-that is, that "today's burglar is yesterday's insider trader and tomorrow's rapist." Rather, such white-collar offenders engage almost solely in financial crimes. 13 .

Fifth, are the age and race distributions of white-collar offenders the same as the age and race distributions of more conventional street offenders? Polk (1991) points out that if one uses Gottfredson and Hirschi's operational definition of white-collar crime, more Blacks are arrested for such offenses than Whites. However, given the underrepresentation of African Americans and other racial minorities in the upper echelons of corporate America, it is highly unlikely that they are engaged in insider trading, price fixing, military contract fraud, or similar crimes. If newspaper photos may be trusted, such offenders are invariably White. Moreover, given the educational and experiential requirements of their positions, it is also likely that they are older than conventional offenders (Steffensmeier, 1989).

Finally, in light of the characteristics of major white-collar crimes and criminals outlined here, would crime prevention programs aimed at teaching parents and other adults how to spot and punish low self-control in their children be successful? One wonders to what extent such programs would be effective in preventing even most con- ventional street crime. To support such a claim requires us to discount the roles that factors such as poverty, unemployment, homelessness, and institutional discrimination play in promoting criminal activity. As we will see in Chapter 6, available data do not permit us to make such a quantum leap of faith.

Gottfredson and Hirschi (2000) .have responded to some of the criticisms of self-control theory and continue to cite studies that support it. Nevertheless, even in

III II

Ii II

166 CHAPTER 5

studies that affirm the theory, the findings are modestly supportive at best, explaining just 3-11 percent of the variation in deviance and criminal offending (Longshore, 1998). As Polk (1991) concludes, "too much crime falls outside the boundaries of [Gott- fredson and Hirschi's] definition for this general theory to be of much use" (p. 579).

I

II I!

I·

I I I I i I I

,i ,

I:I II Ii:

'Ii I, I

Neutralization Theory

The theories we have discussed so far have highlighted differences between criminals and noncriminals, and delinquents and nondelinquents. A different approach was pro- posed by Gresham Sykes and David Matza (1957; Matza, 1964). Sykes and Matza pointed out that if traditional positivist theories of crime are correct, some individuals would be criminal all the time, whereas others would never deviate. They observed, however, that even the most active delinquents spend most of their time in noncrimi- nal pursuits (see also Brunson & Miller, 2001). They were especially critical of the notion of the delinquent subculture which, as we noted earlier, depicts offenders as having a value system at odds with that of the dominant culture. Were this the case, Sykes and Matza argue, delinquents would show no remorse for their behavior and would not view it as wrong. When detected and apprehended, though, delinquents typically exhibit guilt and shame over their behavior. Although some skeptics might see such expressions as an attempt to appease those in authority, Sykes and Matza take these youthful offenders at their word. 14

Instead of seeing delinquency as a rejection of societal norms, Sykes and Matza view it as the endproduct of a process they call neutralization. To understand neutral- ization, one must also understand the concept of drift. According to Sykes and Matza, adolescent behavior runs along a continuum, with total freedom at one end and total constraint at the other. Rather than locate themselves consistently at one pole or the other, adolescents vacillate between these two extremes. "The delinquent transiently exists in a limbo between convention and crime, responding in turn to the demands of each, flirting now with one, now the other, but postponing commitment, evading deci- sion. Thus, he [sic] drifts between criminal and conventional action" (Matza, 1964, p. 28, author's emphasis). The drift into delinquency is facilitated by learning justifica- tions or rationalizations that neutralize the constraint of society's norms of behavior and thus legitimate deviation. Sykes and Matza called these justifications or rational- izations techniques of neutralization.

Sykes and Matza identified five basic types of techniques of neutralization:

denial oj responsibility-The deviant disavows personal responsibility for the offense, claiming that it was not his or her fault. Barterers, for example, fre- quently deny responsibility for an abusive incident by claiming they were drunk.

denial ofinjury-The deviant maintains that an offense didn't really occur because no one was harmed by his or her actions. Thus, individuals arrested for illegal gambling will sometimes maintain their innocence on the ground that "nobody gets hurt" from what they do.

Sociological Theories of Crime IT: Crime and Social Processes 167

denial ofvictim-By maintaining that their victims deserve what happens to them, offenders may also justify or legitimate their offending. A defense attorney at a recent rape trial, for instance, argued that his client should be acquitted because the woman he raped was dressed provocatively and, therefore, "deserved it."

condemnation ofcondemners-An offender legitimates his or her behavior by claim- ing that his or her accusers, judges, or others in authority are corrupt and, there- fore, also guilty. For example, a child whose parent catches him or her smoking marijuana may argue, "Why shouldn't I smoke pot? You drink, and everybody lmows alcohol is worse for you than marijuana."

appeal to higher loyalties-In this case, the offender disavows personal benefit or gain from his or her behavior, claiming that he or she didn't do it for themselves, but for others. A fraternity member, for instance, was recently apprehended while breaking into a professor's office in whose class he was not even enrolled. During questioning, he admitted that he was planning to steal an exam, but that he was doing this not for his own benefit; but for "his brothers."

In considering these techniques of neutralization, two important points must be made. First, although many of the illustrations offered here are after-the-fact justifica- tions for misbehavior, Sykes and Matza maintain that techniques of neutralization occur prior to the commission of a deviant act. They serve to motivate or facilitate devi- ation by loosening moral constraints on individuals. Second, we can see here a close resemblance between techniques of neutralization and what is defined in law as miti- gating circumstances (including self-defense, accident, and insanity). This is no comci- dence. Because delinquents understand and (usually) adhere to the society's normative value system, they also understand and concur with the law, which allows for extenu- ating circumstances that negate an offense. Although they agree that people should be held responsible for their actions, they also know that there are conditions under which infractions are excusable and sometimes even permissible. Techniques of neutraliza- tion are expansions and distortions of the same conditions that excuSe the accused in law (Matza, 1964).

Once the restraints of social norms are temporarily neutralized, individuals are free to drift into delinquency. They may, of course, be diverted, but because they now feel they have no control over their circumstances, that what lies ahead is destiny, they are motivated to act or to make something happen. It is this sense of desperation that provides the will to commit new infractions. "The will to repeat old infractions requires nothing very dramatic or forceful. Once the bind of the law has been neutralized and the delinqueI\t put into drift, all that seems necessary to provide the will to repeat old infractions is preparation" (Matza, 1964, p. 184, emphasis added).

Before moving to an assessment of neutralization theory, one final question needs to be addressed. As Sykes and Matza pointed out, "This approach to delin- quency centers its attention on how an impetus to engage in delinquent behavior is translated into action. But it leaves unanswered a serious question: What makes delin- quency attractive in the first place?" (Matza & Sykes, 1961, p. 712). Their emphasis on the similarity between delinquent values and the values of the dominant culture

j ,

I I I I

jl I'

II .: I ;, '

I 'll! : I" II·

I I

168 CHAPTER 5

informs their response. According to Sykes and Matza, coexisting with the explicit or official values of society are a set of subterranean values-"values, that is to say, which are in conflict with other deeply held values but which are still recognized and accepted by many" (Matza & Sykes, 1961, p. 716). These are not the conflicting val- ues of two opposing groups, but rather they exist within a single individual. Subter- ranean values include the element of adventure (displays of daring and the search for excitement and thrills), the desire for a "soft" job where one earns money as quickly and painlessly as possible, the pursuit of conspicuous consumption, and an acceptance of aggression and violence (the ability to "take it and hand it out," to defend one's rights and one's reputation with force, and "to prove one's manhood with hardness and physical courage").

Sykes and Matza maintain, then, that these subterranean values are widely held in U.S. society, but that their manifestations are usually confined to certain circum- stances deemed "appropriate" or "proper" (such as sporting events, conventions, or "the big night on the town"). Delinquent youth conform to these values, and fre- quently accentuate them. Trouble often arises not only because of this accentuation, but also because young people are notoriously poor judges of appropriate times and situations.

In short, we are arguing that the delinquent may not stand as an alien in the body of society, but may represent instead a disturbing reflection or a caricature. His [sic] vocabulary is different to be sure, but kick, big-time spending, and rep have immediate counterparts in the value system of the law-abiding. (Sykes & MatL;a, 1961, p. 717)

1 I, I ,

i I : I

I , ! I, I . I , I I

I II

..II it

Evaluating Neutralization Theory

The validity of neutralization theory has been challenged by a number of criminolo- gists working from various perspectives. One question that has been raised, for exam- ple, is if delinquents and nondelinquents are no different, can variations in delinquency rates among youths be attributable only to some individuals' greater capability to neu- tralize or to the fact that they just frequently happen to be in situations or with people who promote the will to deviate? How does one explain persistent serious offending and youths who grow up to pursue criminal careers as adults?

Sykes and Matza acknowledge the problem of the "hardcore" delinquent and the persistent offender, but they maintain that their numbers are small and that it may be necessary to explain their behavior in other ways (such as social and personal isolation). Most young people are pretty conventional and most "age out" of delinquency. They come to learn the appropriate times and places to pursue subterranean values and, as they grow older and acquire greater responsibilities and lose a good deal of leisure time, offending is harder to justify. To some critics, though, this response is unsatis- factory since it appears to undermine the theory: The theorists are distinguishing types of delinquents when their goal is to show that there is really no difference between delinquents and nondelinquents (see, for example, Taylor et aI., 1973).

Sociological Theories of Crime IT: Crime and Social Processes 169

Research on the process of neutralization has produced inconsistent findings, although it must also be noted that many of these studies have serious methodological problems, such as the use of small, unrepresentative samples (see, for example, Ball, 1966; Hindeglang, 1970; Minor, 1980, 1981; Regoli & Poole, 1978; see also the Appendix). Hamlin (1988) has persuasively argued that the techniques of neutralization do not precede deviant behavior, but rather follow it. According to Hamlin:

Motives are utilized and changed in the process of legitimating social action and have very little to do with the actual cause of the action. We generate motive in response to a "question situation" and, through our (read: white, Western, male) perceived progres- sive linear time, logically put the motive prior to the action. This prior sequencing is a fallacy. Motives are a product of social action. It is not until after a social action, or more precisely, not until action needs to be legitimized, that motives are produced. (p. 431)

Hamlin's critique raises questions not only about the process of neutralization, but also about techniques of neutralization themselves. Some critics point out that one weakness in neutralization theory is the implicit idea that all techniques are equal; they all neutralize the bind of social norms in the same way. Some techniques, how- ever, do not just extend conventional morality, but rather challenge it. An example will make the point clearer: "A homosexual who says he cannot help being a homo- sexual because he is sick is very different from a homosexual who denies the fact of harm to the victim, who declares 'gay is good' and that his partner agrees" (Taylor et al., 1973, p. 184).

Of course, that various techniques of neutralization may be valued differentlyby different individuals or social groups at different times and in different circumstances does not negate the importance of Sykes and Matza's work in identifying the existence of these techniques in the first place. There is evidence, in fact, that techniques of neu- tralization may be more widely used than Sykes and Matza first proposed and that their typology of techniques should be expanded. Coleman (1987), for example, found six techniques of neutralization that are commonly used by white-collar offenders, and he notes that most white-collar employees report that their workplace culture is imbued with a set of expectations that encourage unethical and even criminal business prac- tices. Hagan and his colleagues (1998) believe that these techniques of neutralization grow directly out of the cultural value placed on individualized competition for mate- rial success-or what they call hierarchic self-interest-that is inherent in market soci- eties. In fact, they hypothesize that societies experiencing rapid economic change to a capitalist market economy are especially likely to feel the effects of hierarchic self- interest expre~sed through techniques of neutralization that encourage criminal and delinquent behavior. Hagan et al.'s (1998) research in Germany supports this argu- ment. Box 5.2 looks at another possible site for testing their hypothesis.

Box 5.2 is also important because it highlights the gendered nature of crime and deviance. Like most traditional criminological theories, Sykes and Matza's neutraliza- tion theory was developed through research that included only males. Few tests of the theory have looked at female offending (for an exception, see Ball, 1977). However,

170 CHAPTER 5

given research indicating that females may use a different vocabulary of motives than males (Gilligan, 1982), studies designed to test the accuracy and pervasiveness of neu- tralization theory must include women and girls. IS

While the parents' motive for selling their children is arguably understandable on some level, how do most customers rationalize their behavior? One rationalization offered by many customers is that they are helping the children (denial-of injury). They argue that they are pro- viding them with much-needed money for their families and preventing them from having to work at even more dangerous or menial occupa- tions. The customers also often rationalize that children from impoverished countries become sexually active at earlier ages anyway.

Another motivation of customers seeking out young children is the belief that child pros- titutes are less likely than adult prostitutes to be infected with HN, the virus that causes AIDS (Sherry et al., 1995). According to international health experts, AIDS is spreading rapidly among prostitutes in many countries, especially in Asia. It is also spreading from country to country because of international trafficking in prostitutes and because of travelers who contract the dis- ease abroad and bring it home with them. Pros- titutes report that few of their customers wear condoms, and the younger the prostitute, the more powerless she is to insist that a condom be worn. However, rationalizations for crime and deviance are often not grounded in fact. Accord- ing to health experts, child prostitutes are at g;reatest risk of contracting HN because of their age. A child's vagina or anus is more easily torn from intercourse, causing open cuts, sores, and bleeding that facilitate HIV transmission (Lim, 1998). In fact, health experts expect the inci- dence of AIDS in Asia as well as Latin America to continue to rise, with children making up an increasing percentage of those who become infected and eventually die from the disease.

HOW THE WORLD SEES IT Sex Tourism

BOX 5.2

According to children's rights advocates, an increasing number of children in countries such as Brazil, the Philippines, and Cambodia are being kidnapped and forced into prostitution or sold to pimps by their parents, most of whom are desperate for income. The children may be locked in the brothels if they are considered likely to try to escape, but usually such measures are unnecessary; beatings and threats are usually enough to convince the children to stay (Lim, 1998). Accurate estimates of the number of child prostitutes are difficult to come by, with some experts setting the lower limit in the tens of thousands and others saying it is at least one mil- lion (Lim, 1998). The children involved, the vast majority of whom are girls, are as young as six and as old as fifteen (the age of consent in most

. countries is sixteen) (Goering, 1996; Kristof, 1996; Sherry et al., 1995).

Who are the customers of these child prostitutes? Some are local men, neighbors of the children, to whom the children are "rented out" by their parents. Parents often rationalize their behavior by appealing to higher loyalties: Selling their children's bodies provides desper- ately needed income for the entire family. Even greater financial gains can be had, however, if the children are sold to foreign businessmen and tourists. Some of these men are individual trav- elers, but others travel on organized sex tours. The tours, which first began in Japan, are now sold in countries such' as Great Britain, South Korea, and Taiwan. If a child is a virgin, the fee may be as much as $500, but immediately fol- lowing the loss of her virginity, a young girl may be hired for anywhere from $2 to $10, depend- ing on her age and experience (Kristof, 1996; Lim, 1998; Sherry et al., 1995).

Ii

II,,.

I I I I I

Sociological Theories of Crime II: Crime and Social Processes

Self-Esteem and Crime

171

It is a widely accepted belief that how we behave depends, in large part, on whether we think positively or negatively about ourselves. Ifwe are self-confident and feel capable and in control of our lives-that is, if we have high self-esteem-we will probably behave responsibly and treat others respectfully. Conversely, if we see ourselves as losers or as failures-that is, if we have low self-esteem-we are likely to become with- drawn or to do something to try to pick ourselves up, if only for a short time; we may even engage in some form of self-destructive behavior. Moreover, most of us also know the kinds of experiences that lower self-esteem: "constant failures and a constant bombardment with the message that one does not count as a person or with others" (Smelser, 1989, p. 7).

Given the intuitive appeal of these ideas, it is not surprising that several social sci- entists have postulated a causal relationship between self-esteem and criminal behav- ior, and numerous crime prevention and rehabilitation programs are premised on the notion that deviance is a direct outgrowth of the devalued or disvalued self (pollack, 1998). However, it is Howard B. Kaplan's (1975, 1980) formulation that is considered by many criminologists to be the most comprehensive and most widely tested theoret- ical statement of the self-esteem model of crime and delinquency, and so we will con- centrate primarily on Kaplan's work in this section.

Kaplan begins with the fundamental sociological obserVation that we develop our sense of self through interaction with others in the groups to which we belong (our family, our peers). We learn to place a particular value on ourselves as persons and on our-behavior through others' reactions to us. Over time, these others need not even be present; we internalize their responses so that our mere imagination of them influences our self-concepts, or "self-attitudes" as Kaplan puts it. According to Kaplan (1975), "persons who in the course of their group experiences have developed relatively nega- tive self-attitudes are significantly more likely to adopt deviant response patterns in a specified future period than persons who in the course of their group experiences have developed relatively positive self-attitudes" (p. 51). For individuals with low self- esteem, crime and delinquency may come to be viewed as self-enhancing opportuni- ties. This is not to say that all individuals with low self-esteem will commit crimes; rather, their low self-esteem predisposes them toward deviant activity. "Whether deviant behavior is adopted and which type is chosen depend on circumstances-what kinds of deviant activity are visible and available, the perceived attractiveness of these opportunities, and so on. Whether the involvement is continued ... depends in turn on the extent to which deviant activity is in fact felt to be self-enhancing or self- derogating" (Scheff et al., 1989, p. 171).

It is disappointing to report that, despite its widespread intuitive support- indeed, Wells (1989) argues that the idea of a causal link between self-esteem and crime is so well accepted that it seems a "truism"-the empirical support for Kaplan's theory has been weak and often contradictory. Kaplan's (1980) own test of the the- ory-a longitudinal survey of more than 3,000 seventh-grade male and female students who were questioned each year for three years-initially yielded supportive findings. In this study, those adolescents with initially low levels of self-esteem, as well as those

I I,. I, I:

11

11.:....! I",

III,': II'i I I .. 'I

11,1

172 CHAPTER 5

who subsequently experienced increases in self-rejecting attitudes, were more likely than the other adolescents surveyed to later engage in deviant activity. Additional lon- gitudinal tests of other data by other researchers, however, failed to confirm the self- enhancing effects of delinquent behavior and showed negligible or nonexistent direct effects of self-esteem on deviance (see, for example, BYllller et al., 1981; McCarthy & Roge, 1984; Wells & Rankin, 1983).

In subsequent analyses, Kaplan and colleagues (1986, 1987) respecified some of the variables and elaborated the self-enhancement model to take into account addi- tional factors, such as "early involvement in deviant activities" (reported during the first round of questioning) and "deviant peer associations" (as reported during the sec- ond round of questioning). These studies, as well as those of others, have shown that the relationship between self-esteem and deviance may be more complex than Kaplan's theory originally conveyed. In some of the research, low self-esteem had both a posi- tive and a negative effect on subsequent deviance. Individuals with pathologically low levels of self-esteem seem to experience self-enhancement through deviant activity. Others, whose self-esteem levels are within the normal range of low-high variation, experience lowered feelings of efficacy as a result of their deviation which, in turn, increases their need to conform and inhibits future deviance. Paradoxically, however, some studies also show that individuals with exceptionally high levels of self-esteem may experience self-enhancement through deviance-a finding that cannot be explained by Kaplan's theory since it says nothing about motivational dynamics at the extreme upper-end of the self-esteem continuum (Evans et aI., 1991; Pollack, 1998; Wells, 1989). Pollack (1998), for example, reports that female inmates whom she inter- viewed explained their offending in terms of high self-esteem. As one woman told,the interviewer, "I love myself, that's why I did this. I wanted money, that's why I did this" (quoted in Pollack, 1998, p. 5, author's emphasis).

Pollack (1998) points out another serious weakness in self-esteem theory: Focus- ing on low self-esteem as the cause of crime individualizes the crime problem and decontextualizes offenders from the social, political, and economic constraints they face in their everyday lives. Offending becomes a psychological problem, while oppres- sion in the form of classism, racism, and sexism are ignored as contributing factors. "[T]he problem, and the solution to the problem, lie within the individual" (pollack, 1998, p. 3).

We will return to the question of how oppression contributes to crime in the next chapter. Now, however, we will take another look at the relationship between low self- esteem and crime. Individuals may engage in a deviant act for any number of reasons, but if their deviance is detected and elicits a negative reaction, they may internalize the stigma, develop a negative self-concept, and engage in future deviation because they have come to see themselves as deviants. This idea is one of the central tenets of the final theoretical perspective we will discuss in this chapter, labeling theory.

Crime and Stigma: The Labeling Perspective

Labeling theory, or social reaction theory as it is sometimes called, was developed in rebel- lion against the dominance of the positivist paradigm in criminology. At the same

r Sociological Theories of Crime II: Crime and Social Processes 173 time, however, although one of its goals was to assert the rational element of deviation, it also would be inappropriate to consider it a descendant of the classical school. Label- ing theory does not fall within the boundaries of the Marxist paradigm either, but it is perhaps best to characterize it, as some observers have, as a bridge or link to the radi- cal criminological theories we will examine in the next chapter (Martin et al., 1990).16

Labeling theory constituted a bold new approach to explaining crime. It gained popularity understandably during a period of sweeping social change-the 1960s and 1970s-when questioning authority and the status quo was widespread (see Chapter 1). Whereas most criminological theories focused on criminal behavior, labeling theorists struck out in a new direction; their emphasis was on how certain behaviors come to be defined as criminal and the consequences of these definitions for individuals found to be engaging in such activities. To understand their approach more clearly, let's begin with a discussion of labeling theorists' views of crime itself.

The Relativity of Crime

Up to this point, the theories we have discussed have held an absolutist view of crime; that is, crime is behavior that violates a law, an agreed-upon rule. From this perspec- tive, there are some behaviors whose characteristics inherently make them criminal; rape and homicide are two frequently cited examples. If one accepts this position, then the logical course of action is to identify those who break the law and try to discover what it is about them (biological and psychological theories) or about their environ- ments, life conditions, or circumstances (sociological theories) that would lead them to commit crimes.

In contrast, labeling theorists see crime from a relativist point of view. An act becomes criminal or deviant only when it is defined as such by a group of observers. As Howard Becker (1973) put it in a much-quoted passage from his book, Outsiders:

[S]ocial groups create deviance by making the rules whose infraction constitutes deviance, and by applying those rules to particular people and labeling them as outsiders. From this point of view, deviance is not a quality of the act the person commits, but rather a con- sequence of the application by others of rules or sanctions to an "offender." The deviant is one to whom that label has been successfully applied; deviant behavior is behavior that people so label. (p. 9; author's emphasis)

There are several aspects of this view of crime that deserve to be highlighted. First, labeling theorists are pointing out that what is defined as criminal or deviant depends on a number of factors, including the situational and historical contexts in which the behavior occurs, the characteristics of the individual engaged in the behav- ior, and the cnaracteristics of the definers. Say, for example, someone sees your lips moving as if you are having a conversation, but there is no one visibly present to whom you could be speaking. The observer asks you what you are doing, and you reply that you are talking to God. Ifyou are in a church, synagogue, or other official place of wor- ship, such a response probably would not be considered deviant. If, however, you are at the neighborhood deli eating lunch, it might be viewed at least as unusual or odd. Similarly, many behaviors formerly against the law are now legal and vice versa. In the nineteenth century, for instance, there were state and local laws that prohibited women

I, I I

j",I l

I

ii,

174 CHAPTER 5

from phoning men for dates, undressing in front of a photograph of a man, and appear- ing on a public highway wearing a bathing suit (unless they were accompanied by at least two officers or anned with clubs). Surely, today we would consider these laws, not the behaviors they prohibit, deviant.

The characteristics of the actors and definers are also significant. Labeling theo- rists are quick to point to the preoccupation among criminologists and law enforce- ment agents with street crimes, while overlooking or downplaying most forms of high-level white-collar crime, including corporate and governmental deviance. There are, of course, significant differences, including racial and social class differences, between street offenders and corporate rulebreakers. The relative powerlessness of the former makes it highly likely that they will be apprehended and processed through the criminal justice system, while the status of the latter allows them to escape being iden- tified and labeled criminal (Becker, 1970). Some groups-what Becker (1973) refers to as moral entrepreneurs-are also better able to get their interests represented in law, thus ensuring that certain behaviors (read: not their own) get defined as criminal, while others do not.

This brings us to a second major point of labeling theory's view of crime: It sees crime as the product of social interaction. What is crucial is not that an individual vio- lates a rule or a law, but rather that others respond to that individual's behavior, label- ing him or her a criminal or deviant. This may be done informally, but of greater significance to labeling theorists is when this process takes place in what they refer to as public status dep;radation ceremonies, such as court hearings or trials (Garfinkel, 1965). With the label attached, the individual undergoes a fundamental change in identity. Indeed, the labels criminal and deviant constitute a master status-a status that takes precedence over all other statuses or characteristics of the individual. Others, who have deeply ingrained, proconceived ideas of what a criminal is like-untrustworthy, unpredictable, sinister-begin to structure their interactions with the labeled individ- ual on the basis of these stereotypes (for example, they stop doing business or socializ- ing with him or her). They may even redefine past behavior on the part of this individual so that it conforms to stereotypes attached to the deviant label (for example, the routine hug upon greeting an old friend is seen in a new light after the friend comes out as gay or lesbian).

If we imagine for a moment what such experiences may be like for the labeled person, it is not difficult to understand that eventually he or she may come to accept the label and begin to alter his or her behavior to confonn to it (Crocker et al., 1998). What occurs is a self-fulfilling prophecy: expectations about how the individual will behave are fulfilled, not so much because the person is truly "bad" or "abnonnal," but because both the person and others have come to believe he or she is "bad" or "abnor- mal" and they act accordingly.16

In short, labeling theorists see social reaction as the key element in crime and deviance causation. As Edwin Lemert (1967), whose work is part of the foundation of labeling theory, put it:

This is a large turn away from older sociology which tended to rest heavily upon the idea that deviance leads to social control. I have come to believe that the reverse idea,

r Sociological Theories of Crime IT: Crime and Social Processes 175 i.e., that social control leads to deviance, is equally tenable and the potentially richer premise for studying deviance in modern society. (p. v)

Lemert's controversial statements raise the issue of the consequences of social reaction for the labeled individual, a point that we have already touched upon, but which war- rants closer consideration.

Social Reaction and Conunitment to a Deviant Career

A useful way of understanding labeling theory's view of the impact of social reaction on the labeled individual is to consider Lemert's (1951) distinction between primary devi- ation and secondary deviation. Primary deviation is simply rule breaking. An individ- ual may engage in primary deviation for any number of reasons that may be social, economic, or political. He or she may be acting on a hedonistic impulse or out of des- perate need for money to buy food or to satisfy a drug habit. Whatever the specific causal factors that give rise to it, primary deviation is of little concern to labeling the- orists unless it is detected and elicits a reaction.

Secondary deviation is deviation that results from societal reaction. As we noted previously, when an individual's deviation elicits, in particular, a formal, public reac- tion, the reaction process can lead to a total reorientation of the individual's self- perceptions. As Cohen (1966) wrote:

The label-the name of the role-does more than signify one who has committed such-and-such a deviant act. Each label evokes a characteristic imagery. It suggests someone who is normally or habitually given to certain kinds of deviance; who may be expected to behave in this way; who is literally a bundle of odious or sinister qualities. It activates sentiments and calls out responses in others: rejection, contempt, suspicion, withdrawal, fear, hatred. (p. 24, author's emphasis)

In other words, others' reactions may close off legitimate or nondeviant opportunities and interactions for the labeled individual. The alcoholic, for example, may not be invited to parties or to friends' homes anymore, thus adding to his or her isolation. The ex-convict may face tremendous difficulty in getting a good-paying job.

This closing off of legitimate opportunities and interactions, coupled with the destruction of one's public image and character, may leave the labeled individual with little choice but to seek out deviant associations and to pursue deviant or criminal opportunities. The reaction process, then, may cause the resocialization of the labeled individual toward acceptance of and conformity to the role attached to the deviant label. That is: the labeled individual becomes committed to a deviant identity and embarks on a deviant career. This secondary deviation is essentially a defensive and adaptive strategy on the part of the labeled individual; it is an effort to survive and "a means of sustaining a 'social self' in the face of exclusion and stigmatization" (Taylor etal., 1973,p.151;seealsoLemert, 1951).

The relationship between primary and secondary deviation is represented sche- matically in Figure 5.1. Notice here that typically primary deviation goes undetected

"I

176 CHAPTER 5

(a) primary deviation ----'~~ no social reaction ~ no secondary deviation

(b) primary deviation ~ mild social reaction ~ no secondary deviation

(c) primary deviation ~ strong and/or public social reaction -+- secondary deviation

(d) primary deviation X. secondary deviation

FIGURE 5.1 The Relationship between Primary Deviation and Secondary Deviation

iI ~ !

I'

"

Ii.

I·.

(a). Even if it is detected, the response may be mild or rather limited, leading to no fur- ther deviation (b). The more dramatic the response, though, the more likely that sec- ondary deviation will be the outcome (c). However, there can be no secondary deviation without social reaction (d).

It is not difficult to see why the labeling perspective ignited a fierce debate among criminologists and polieymakers alike. Let's examine first some of the empirical research that has attempted to test the basic tenets of the theory, and then move on to consider the policy implications of this perspective.

The Empirical Validity of Labeling Theory

The majority of the research on labeling theory has focused on the phenomenon of secondary deviation and the extent to which the social reaction process affects or alters the labeled individual's self-concept and behavior. Much of the early criticism of the perspective focused on its determinism. A5 Akers (1962) expressed it, "One sometimes gets the impression from reading this literature that people go about minding their own business, and then-'wham'-bad society comes along and slaps them with a stig- matized label" (p. 465). To such a critique, Becker (1973) has replied:

[T]he act oflabelling [sic], as carried out by moral entrepreneurs, while important, can- not possibly be conceived as the sole explanation of what alleged deviants actually do. It would be foolish to propose that stick-up men stick people up simply because some- one has labelled them stick-up men, or that everything a homosexual does results from someone having called him [sic] homosexual. Nevertheless, one of the most important contributions of this approach has been to focus attention on the way labelling places the actor in circumstances which make it harder for him to continue the normal rou- tines of everyday life and thus provoke him [sic] to "abnormal" actions ... The degree to which labelling has such effects is, however, an empirical one, to be settled by research into specific cases rather than by theoretical fiat. (p. 179)

'.

The question remains, then, has the notion of secondary deviation been empirically verified?

The answer to this question is largely no. For one thing, many studies have failed to show that being labeled criminal or delinquent leads to a negative self-image on the part of those so labeled (Evans et a1., 1991; Martin, 1985; Shoemaker, 1984). At best, researchers have found only weak support for a relationship between labels and sec-

Sociological Theories of Crime II: Crime and Social Processes 177

ondary deviance (Gold, 1970; Gold & Williams, 1969; McEachern, 1968, Smith & Paternoster, 1990). A number of researchers have even reported that informal labels, imposed early in one's life, especially by parents, have a far greater impact on individ- uals' self-concepts than formal public "degradation ceremonies" that may not occur until adolescence (paternoster & Triplett, 1988).

A second difficulty with this perspective highlighted by the research is that label- ing theorists depict individuals as being rather passive during the labeling process, when, in fact, the imposition of a deviant label is often strenuously resisted by those being labeled. Moreover, there is evidence that deviant labels are not indelible. Indi- viduals can, and often do, overcome stigma (prus, 1975; Rogers & Buffalo, 1974).

Third, in some cases, the deviant label is valued and even sought after. Akers (1968) uses the example of gang members whose deviant identityis formed before they are ever officially labeled. We might also consider political terrorists and rebels, who adopt a deviant identity to set themselves apart from the authorities they challenge. The process of officially being labeled deviant by these authorities is often to them a -symbol of accomplishment. Mankoff (1971), in particular, has argued that one of the most significant weaknesses of labeling theory is its failure to recognize that individu- als may choose to embark on a criminal or deviant career without ever having experi- enced social reaction to their behavior.

Among the other criticisms that have been leveled against labeling theory are those that have to do with its position on who is most likely to get labeled deviant or criminal. Siegel (1992), for example, argues that while labeling theory leads us to believe that it is the poor and powerless who are most likely to get (unfairly) labeled deviant or criminal, the justice system is not consistently unfair and biased against the poor or in favor of the rich. However, Taylor, Walton, and Young (1973) are critical of labeling theory because of its failure to more fully examine "the way in which deviance and criminality are shaped by society's larger structure of power and institu- tions" (p. 169).

This disagreement forms the core of a debate that we will take up again in Chap- ter 6. To conclude the present chapter, however, we will examine the policy implica- tions of the labeling perspective. It is in this area that some of the strongest criticisms of the theory have been centered.

What Is to Be Done?

Labeling theorists' response to the question that serves as the title of this section would undoubtedly be "nothing," or at most, "very little." Keeping in mind that labeling the- ory sees deviance as being amplified and promoted by labeling and that processing through the ctiminal justice system, as one of the strongest and most public social reac- tions, generates secondary or career deviance, it is no wonder that labeling theorists argue, as Wright (1991) puts it, that less is best when it comes to punishing offenders. Labeling theorists favor, for example, the legalization of "victimless crimes," such as prostitution and drug offenses, pointing out that such laws are incredibly difficult to enforce and, rather than reducing crime, simply engender greater deviation (including bribery and police corruption; public disrespect for the law, given widespread demand

1 :,1.

1

'j

!

,", : .j" ,

I I

'II. 11 ,',

I !: II " ', "

178 CHAPTER 5

for particular illegal goods and services; and secondary offending, such as when a drug addict takes to stealing to support his or her habit) (see, for example, Schur, 1965).

However, the most hotly debated policy recommendation of the labeling theo- rists is noninteroention. This position was promoted most strongly by Edwin Schur (1973) in his discussion of the treatment of juvenile offenders. He argued that the state should legalize many acts currently considered delinquent and ignore most others. Only the most serious infractions should reach the attention of the courts, but in any event, even these offenders should not be committed to correctional facilities. Indeed, Schur advocates abolishing juvenile correctional institutions and replacing them with noninstitutional programs that are largely voluntary. In Schur's (1973) own words, the tenet that should guide polieymakers and those working in the juvenile justice system is "leave kids alone wherever possible" (p. 155).

We have already reviewed a considerable amount of research that shows that support for the proposition that labeling causes secondary deviation is rather weak and inconsistent. In addition, there is a body of evidence that indicates that offenders who are formally processed through the criminal justice system are actually less likely to recidivate than those who are treated informally or who are simply ignored (Wright, 1991).

Wright (1991) also makes the point that social reaction in the form of official processing serves two important crime-reduction purposes: (1) incapacitation-if offenders are incarcerated in correctional facilities they Carlnot engage in secondary deviant activities; and (2) general deterrence-although some offenders may be "hard- ened"by labeling and incarceration, others in the society at large may see through their example that "crime doesn't pay" and be inhibited from offending. Although there is ample data that call into question the effectiveness of both the incapacitation and deterrence functions of imprisonment, Wright's critical assessment of nonintervention policies that have been implemented gives one reason to pause and carefully reconsider the wisdom and long-term efficacy of such an approach. Thus, even if social reaction produces some secondary deviation, the question arises as to whether it may actually reduce crime more than it creates it.

More recently, some researchers have tried to identify those conditions under which social reaction may be beneficial to offenders. Australian criminologist John Braithwaite (1989) argues, for example, that the key to beneficial social reaction lies in the process of shaming. Braithwaite (1989) defines shaming as social disapproval with the "intention or effect of invoking remorse in the person being shamed and/or con- demnation by others who become aware of the shaming" (p. 100). According to Braith- waite, "Societies with low crime rates are those that shame potently and judiously" (p. 1). What about labeling theorists' contentions that shaming-or, as they would put it, stigmatizatioh-pushes offenders into criminal or deviant subcultures and launches them on a criminal or deviant career? In response, Braithwaite makes a distinction between disintegrative and reintegrative shaming.

Disintegrative shaming is counterproductive and may lead to further criminality because it separates offenders from the community and treats them as outcasts. In such cases, "punishment erects barriers between the offender and the punisher through transforming the relationship into one of power assertion and injury" (Braithwaite,

r Sociological Theories of Crime II: Crime and Social Processes 179 1989, p. 73). It should come as no surprise, therefore, that offenders who experience disintegrative shaming are highly likely to recidivate and to become members of crim- inal or deviant subcultures, since within these groups they at least receive some social support and self-validation.

In contrast, reintegrative shaming is first of all an expression of disappointment in the individual who has done wrong. Rather than treating the offender as a "bad per- son," reiIltegrative shaming actually reaffirms the offender's morality; the disappoint- ment stems from the fact that a "good person" would do something wrong. "Reintegrative shaming means that expressions of community disapproval, which may range from mild rebuke to degradation ceremonies, are followed by gestures of reac- ceptance into the community of law-abiding citizens" (Braithwaite, 1989, p. 55). Importantly, the offender is encouraged to assume the role of repentant.

Braithwaite maintains that reintegrative shaming is most likely to be found in societies characterized by a strong sense of community and a high level of interdepen- dency among its members. Braithwaite believes that contemporary Japanese society fits this description, although he recognizes that Japanese culture. and traditions cannot simply be transposed onto other societies. However, among Braithwaite's policy rec- ommendations for all societies are an increased use of informal institutions of social control that can best employ techniques of reintegrative shaming; the integration of the repentant role (similar to that used at Alcoholics Anonymous and other twelve-step programs) into rehabilitation programs; and more media coverage not only of wrong- doing by individuals and corporations, but also of corporate and individual offenders who can be held up as models of reform following their wrongdoing.

Braithwaite's idea of reintegrative shaming understandably has drawn consider- able attention from criminologists (see, for example, Hay, 1998; see also Chapter 6). Undoubtedly, it will be the subject of debate and the focus of extensive research in the years to come, much the same way labeling theory was during the 1960s and 1970s.

Summary and Conclusion

The focus of this chapter has been on theories that emphasize the role of learning or socialization in the etiology of crime. Using only the length of the chapter as a gauge, it is easy to perceive the popularity of this approach Nevertheless, we have seen that there is tremendous diversity among the many perspectives that may be considered learning theories of crime.

The commonalities of the theories, and the intuitive appeal of many of them, have led some criminologists to develop integrative perspectives that combine elements of two or more ofthe theories, such as differential association and social control, or social control and labeling (see, for example, Elliott et aI., 1985; Triplett, 1990). Other criminologists, however, maintain that despite their promise of providing a truly social analysis of crime causation, these theories fall far short of that mark by ultimately drawing one's attention back to characteristics of individual offenders and their social psychological development. Such a focus is inevitable as long as one operates within a positivist framework, as many social learning theorists do. From the point of view of

180 CHAPTER 5

critical criminologists, then, social theories of crime must be developed within a com- pletely different framework and, for many of them, this framework is the Marxist par- adigm. We take up their work in the next chapter of this text.

KEY TERMS

agents of socialization-individuals, groups, and institutions that have as one of their primary functions the socialization of members of a soci- ety by providing explicit instruction in or mod- eling of social expectations.

autonomous deviance-deviance produced by a con- trol surplus, which helps an individual further extend control over others and thereby increase his or her control surplus.

differential anticipation theory-the view that peo- ple are likely to engage in behaviors from which they expect to obtain the greatest rewards and the least punishment.

differential association-the process of social interac- tion by which individuals acquire definitions favorable and unfavorable to law violation.

differential identification-the process by which criminality develops because of an individual's greater identification with members of criminal or deviant groups as opposed to members of conformist groups.

differential reinforcement-the process by which deviant or conforming behavior is acquired and internalized through past and present rewards and punishments attached to one form ofbehav-

SUGGESTED READINGS

Akers, R. L. (1998). Social learning and social structure: A general theory of crime and deviance. Boston: Northeastern University Press. Akers's most recent, and most extensive, discussion of his social learning theory.

Braithwaite, J. (1989). Crime, shame, and reintegration. Cambridge: Cambridge University Press. Braithwaite introduces his idea of reintegrative shaming and examines contemporary Japanese society in support of his position.

Ermann, M. D., & Lundman, R. J. (Eds.). (1996). Cor- porate and governmental deviance. New York: Oxford University Press. A collection of articles

ior relative to those attached to the alternative behavior.

drift-the process of vacillation along the behav- ioral continuum of total freedom versus total constraint.

master status-a status that takes precedence over all other characteristics or statuses of an individual.

primary deviation-rule breaking, which may be caused by any number of individual, social, eco- nomic, or political factors.

repressive deviance-deviance produced by a control deficit as an individual attempts to escape the deficit and restore balance to his or her control ratios.

secondary deviation---<l.eviation that results from the process of being labeled.

self-fulfilling prophecy-expectations that are fulfilled not because of true causes, but because people believe them to be true and act accordingly.

socialization-the process of social interaction through which a society's culture is taught and learned and human personalities are developed.

techniques of neutralization-a priori justifications or rationalizations that legitimate and, there- fore, facilitate deviation.

that examines various forms of business and political crime, from the Challenger disaster to police use of force in the Rodney King case. Read it and decide for yourself how well these data support some of the theories discussed in this chapter.

Meier, R. F., & Geis, G. (1997). Victimless crime? Los Angeles: Roxbury. An analysis of four topics (prostitution, drugs, homosexuality, and abor- tion) that have been hotly debated with regard to whether the behaviors associated with them should be considered criminal and how the legal system should deal with them, if at all.

Sociological Theories of Crime II: Crime and Social Processes

NOTES

1. Sutherland's Principles of Criminology was first published in 1924. Mer Sutherland's death in 1950, his close friend and leading proponent, Donald Cressey, who also became a prominent criminologist, took the textbook through six subsequent editions; the tenth was published in 1978. In these editions and other publications, Cressey offered what he considered clarifica-· tions, not revisions of the theory. Following Cressey's death, David Luckenbill revised the text, but not the theory, for an eleventh edition of Principles of Criminology, published in 1992.

2. VoId and Bernard (1986) actually use the term normative conflict instead of· culture conflict because after Sutherland's death, Cressey made this substirotion. Cressey made the change, he said, to bring greater clarity and specificity to the theory. "Culture" is a broad term that encompasses not only a group's or society's norms, but also many other things, material (objects) and nonmaterial (beliefs). Norms, on the other hand, are simply rUles of behavior. Normative conflict occurs when various groups in a social setting hold divergent views about what is "correct" or appropriate behavior in a given siroation. According to Cressey, the term normative conflict more accurately conveys the meaning of Sutherland's position.

3. Sutherland was highly critical of those theorists who associated the causes of crime with such factors as poverty and "broken homes." He argued that criminal behavior is found among all social classes; only its forms vary across class groups. He frequently cited his study of white- collar crime (1949) to support this position.

4. Some criminologists argue that many of these criticisms are based on misinterpretations of the theory and misreadings of Sutherland's and others' work. See, for example, the exchanges between Akers (1996) and Hirschi (1996), and between Matsueda (1997) and Costello (1998). When the theory was first published, it also received extensive acclaim and criticism. For a detailed discussion of these early critiques as well as a careful response to them, see Cressey, 1960.

5. For a more detailed discussion of behaviorism, see Akers, 1998; Bandura, 1977; Lefton, 2000; and Skinner, 1953. Akers's social learning the- ory was originally called differential association- reinfOrcement theory, an indication of its strong

181

affiliation with Sutherland's differential associa- tion theory.

6. For instance, the use of particular behavioral modification techniques in corrections and counseling has been challenged in the courts. See Mackey v. Procunier (1973) and Kamowitz v. Michigan Department ofMental Health (1973).

7. Walter Reckless's (1961) containment theory is also a form of control theory. In addition, some· theorists consider Sykes and Matza's (1957) drift theory, to be discussed later in this chapter, a form of control theory;

8. Some observers have likened Hirschi's beliefs element of the social bond to Sutherland's notion of definitions favorable or unfavorable to law violation. However, Costello and Vowell (1999) make an important distinction between the two ideas by pointing out that Hirschi and Sutherland specified different learning processes. Sutherland maintained that defini- tions regarding law violation were learned through interaction with others, indicating that the individual is integrated into the social group. In contrast, Hirschi's control theory makes the claim that "tolerant attirodes toward law viola- tion are more reflective of a lack of social inte- gration than the result of learning definitions through integration into deviant groups" (Costello & Vowell, 1999, p. 834).

9. One might also take issue with Agnew and Petersen's (1989) distinction between serious and minor delinquency. Although it is obvious that using a knife or gun to get something should be weighted more in terms of seriousness than stealing an expensive car part, the differ- ence between hurting someone badly (weighted four in terms of seriousness) and serious fighting (weighted two) is less clear. Moreover, "hit mother," "hit father," and "hit instructor or supervisor" are all weighted two without taking into account the degree of harm done. Can we assume that adolescents who completed the self- report delinquency scale on which these items appeared would report such hitting as "hurt someone badly" if injury had occurred during the incident?

10. However, Costello and Vowell (1999, p. 834) argue that "because most srodies of delinquency are conducted with respondents old enough to have already established attachments, commit- ments, involvements, and beliefs, it is unlikely

182 CHAPTER 5

that even longitudinal studies can untangle the true temporal priority of each dimension of the bond. Thus, it makes sense to conceptualize these dimensions as contemporaneous rather than try to specify one element as a cause of another."

11. Importantly, Hirschi (1969) hypothesized that single-parem households have rates of delin- quency comparable to two-parent households because, all things being equal, one parent can as . effectively socialize children as two parents can. However, as Matsueda and Heimer (1987) point out, all things are rarely, if ever, equal in single- parent households, especially those headed by women, relative to two-parent households. This does not mean, however, that the absence of one of two parents from the home causes delinquency or that delinquency is an inevitable outcome of growing up in asingle-parent family (see also Demo & Acock, 1992). For a discussion of the difficulties faced by single-parent families, par,- ticularly those headed by women, see Renzetti and Curran, 1999.

12. Whitehead and Boggs (1990) decided not to measure the belief element of the social bond in their study because their sample was composed of people who had already violated the law: known felony offenders.

13. Several researchers point out that even "ordi- nary" street offenders often specialize in partic- ular types of crimes. See, for example, Benson and Moore, 1992, and Wright et al., 1995.

14. Matza, in particular, is committed to a method- ological approach he calls naturalism, which he defines as remaining true to that which one is studying. Matza believes that most researchers'

'.

explanations of deviance distort or contradict the explanations deviants themselves would give of their behavior. Researchers, Matza says, should let deviants speak for themselves. See Matza (1964) for a fuller treatment of natural- ism. See Taylor et al. (1973) for a sympathetic critique of this approach.

15. Hagan and his colleagues (1998) suggest that their concept of hierarchic self-interest may be helpful in explaining gender differences in offending. They hypothesize that males are more susceptible to hierarchic self-interest than females. Their research in Germany offers strong support for this hypothesis and will hope- fully encourage additional research on this topic.

16. Many of those who are identified as labeling theorists resist reference to labeling as a theory, preferring instead to call it "a way of looking at a general area of human activity; a perspective whose value will appear, if at all, in increased understanding of things formerly obscure" (Becker, 1973, p. 181). Becker, who is one of the founders of labeling, has also argued that he prefers to call it an interactionist theory of deviance, but conventionally it continues to be referred to as labeling theory or social reaction theory.

17. For one classic analysis of the self-fulfilling prophecy in education, see Rosenthal andJacob- son (1968). However, the application of the self- fulfilling prophecy to crime and deviance is not a unique contribution of contemporary labeling theorists. It was flctually raised in the 1930s by Frank Tannenbaum (1938), who warned against the negative consequences of what he called "the dramatization of evil."