Differential social organization, collective action, and crime Ross L. Matsueda # Springer Science + Business Media B.V. 2006 Abstract This article elaborates and extends Sutherland’s [Principles of criminology (4th ed.), Lippincott, Philadelphia, Sutherland (1947)] concept of differential social organiza- tion, the sociological counterpart to his social psychological theory of differential association. Differential social organization contains a static structural component, which explains crime rates across groups, as well as a dynamic collective action component, which explains changes in crime rates over time. I argue that by drawing on George Herbert Mead’s [Mind, self, and society. University of Chicago Press, Chicago, Mead (1934)] theories of symbolic interaction and social control, we can conceptualize organization in favor of, and against, crime as collective behavior. We can then integrate theoretical mechanisms of models of collective behavior, including social network ties, collective action frames, and threshold models of collective action. I illustrate the integrated theory using examples of social movements against crime, neighborhood collective efficacy, and the code of the street. Differential social organization, collective action, and crime The theory of differential association, along with the concept of white collar crime, was probably Edwin Sutherland’s greatest legacy. It is well known that the theory explains individual criminality with a social psychological process of learning crime within interaction with social groups. Criminal behavior, according to Sutherland [71], is the result of learning an excess of definitions favorable to crime. This is the differential association process. Less well known is Sutherland’s attempt to explain aggregate crime rates across groups and societies. Here, he specified the theory of differential social Crime Law Soc Change DOI 10.1007/s10611-006-9045-1 DO9045; No of Pages A portion of this chapter was presented at the Annual Meetings of the American Society of Criminology, Toronto, Canada, November 16–19. The research upon which this paper was based was supported in part by grants from the National Science Foundation (SES-0004323) and the National Institute on Drug Abuse (R01DA18148). The funding agencies bear no responsibility for the analyses and interpretations drawn here. James F. Short, Jr. and Joachim Savelsberg generously provided comments on an earlier draft. R. L. Matsueda (*) Department of Sociology, University of Washington, 202 Savery Hall, Box 353340, Seattle, WA 98195, USA e-mail: matsueda@u.washington.edu organization to explain rates of crime with an organizational process of group dynamics. The crime rate of a group is determined by the extent to which the group is organized in favor of crime versus organized against crime. Moreover, the explanation of crime rates, differential social organization, is consistent with the explanation of individual acts of crime, differential association. Sutherland spent considerable time refining his individual-level mechanism of differential association, which subsequently received much attention from the criminological community, as researchers tested (e.g., [46]) and revised the theory (e.g., [1, 13]). He devoted less time to his more sociological theory of differential social organization, which consequently never progressed beyond its original rudimentary form. Although Sutherland never explicitly specified the concrete content of organization in favor of crime and organization against crime, he did imply two components: (1) a static component, in which social structure and group organization explained differences in crime rates across social groups; and (2) a dynamic component, in which group processes explained changes in crime rates over time. In this paper, I will focus on the second component, conceptualizing differential social organization as collective action both for and against crime, and integrate theoretical mechanisms from developmental psychology, social networks, and collective action. To integrate these diverse mechanisms, I will use George Herbert Mead’s [52, 53] theory of social control as a general framework for the study of joint behavior among two or more persons. I then draw on developmental psychologists’ stages of moral reasoning [43], which extends Mead’s [52] stages of play and the game, and which may vary by the density of network ties. I also draw upon sociologists’ theories of social networks and weak ties to impose a communication structure on group behavior, as well as threshold models and collective action frames to account for the direction of collective behavior. Finally, I illustrate the integrated model with examples from social movements against crime, neighborhood informal social control, and collective acts of street crime. Differential association theory Sutherland stated differential association theory as a set of nine propositions, which introduced three concepts – normative conflict, differential association, and differential group organization – that explain crime at the levels of the society, the individual, and the group [22]. Normative conflict: the root cause of crime in society At the level of the society, crime in society is rooted in normative conflict. For Sutherland, primitive, undifferentiated societies are characterized by harmony, solidarity, and consensus over basic values and beliefs. Such societies have little conflict over appropriate behaviors. They also have little crime. With the industrial revolution, however, modern industrial societies developed, with advanced divisions of labor, market economies, and increased conflict. Such societies become segmented into groups that conflict over interests, values, and behavior patterns. These societies are characterized by specialization rather than similarity, conflict rather than harmony, coercion rather than consensus. Moreover, they tend to have high rates of crime. From these observations, Sutherland hypothesized that high crime rates are rooted in normative conflict, which he defined as a condition in which society is segmented into groups that conflict over the definition of appropriate behavior. R.L. Matsueda More precisely, normative conflict refers to conflict over the appropriateness of the law: some groups define the law as a set of rules to be followed under virtually all circumstances, while others define the law as a set of rules to be violated under certain circumstances. Therefore, when normative conflict is absent in a society, crime rates will be low; when normative conflict is high, societal crime rates will be high. In this way, crime is ultimately rooted in normative conflict. Differential association process: explanation of individual criminal acts At the level of the individual, the process of differential association provides a social psychological explanation of how normative conflict in society translates into individual criminal acts. According to differential association, criminal behavior is learned in a process of communication in intimate groups. The content of learning includes two important elements. First are the requisite skills and techniques for committing crime, which can range from complicated, specialized skills of computer fraud, insider trading, and confidence games to the simple, readily available skills of assault, purse-snatching, and drunk driving. Such techniques are necessary but insufficient to produce crime. Second are definitions favorable and unfavorable to crime. These definitions are motives, verbal- izations, or rationalizations that make crime justified or unjustified, and include Sykes and Matza’s [75] “techniques of neutralization” used by delinquents, and Cressey’s [21] “verbalizations” used by embezzlers. For example, definitions favorable to income tax fraud include “everyone cheats on their taxes,” “it’s not a real crime,” and “the government has no right to tax its citizens.” Definitions favorable to drunk driving include, “I can drive fine after a few beers,” “I only have a couple of miles to drive home,” and “I won’t get caught this time.” Definitions favorable to violence include, “if your manhood is threatened, you have to fight back,” “you got to have the backs of your boys, even if it means violence,” “to maintain respect, you can never back down from a fight.” These definitions favorable to crime help organize and justify a criminal line of action in a particular situation. They are offset by definitions unfavorable to crime, such as “Income tax fraud is wrong and immoral,” “Tax fraud deprives Americans of important programs that benefit the commonwealth,” “All fraud and theft is immoral,” “Turn the other cheek,” “Friends don’t let friends drink and drive,” “Any violation of the law is wrong.” These examples illustrate several points about definitions of crime. First, some definitions pertain to specific offenses only, such as “Friends don’t let friends drink and drive,” whereas others refer to a class of offenses, such as “All fraud and theft is immoral,” and others refer to virtually all law violation, such as “Any violation of the law is wrong.” Second, each definition serves to justify or motivate either committing criminal acts or refraining from criminal acts. Third, these definitions are not merely ex-post facto rationalizations of crime, but rather operate to cause criminal behavior. Sutherland recognized that definitions favorable to crime can be offset by definitions unfavorable to crime, and therefore, he hypothesized that criminal behavior is determined by the ratio of definitions favorable to crime versus unfavorable to crime. Furthermore, he recognized that definitions are not all equal; some are more important. Sutherland identified at least four dimensions (or modalities) on which definitions vary in importance or weight: frequency (the number of times a definition is presented), duration (the length of time a person is exposed to a definition), priority (the earlier a definition is presented in a person’s life), and intensity (the more intense relationship or prestigious the person presenting the definition). Differential social organization, collective action, and crime Therefore, the individual-level hypothesis of differential association theory states that a person will engage in criminal behavior if the following three conditions are met [47]. (1) The person has learned the requisite skills and techniques for committing crime. (2) The person has learned an excess of definitions favorable to crime over unfavorable to crime. (3) The person has the objective opportunity to carry out the crime. According to Sutherland, if all three conditions are present and crime does not occur, or a crime occurs in the absence of any one condition, the theory would be wrong and in need of revision. Thus, in principle, the theory can be falsified. The process of differential association with definitions favorable and unfavorable to crime does not occur in a vacuum, but is structured by the broader social organization in which individuals are embedded. This includes the structures and organization of families, neighborhoods, schools, and labor markets. This organization is captured by the concept of differential social organization. Differential social organization: explanation of group rates of crime At the level of the group or society, differential social organization provides an organiza- tional explanation of how normative conflict in society translates into specific group rates of crime. According to differential social organization, the crime rate of a group or society is determined by the extent to which that group or society is organized against crime versus organized in favor of crime. In modern industrial societies, the two forms of organization exist side by side – and indeed are sometimes interwoven in complex ways, such as when police take bribes and participate in organized extortion, or baseball players take steroids in full view of teammates. Sutherland hypothesized that the relative strength of organization in favor of crime versus organization against crime could explain the crime rate of any group or society. Thus, compared to suburban neighborhoods, inner-city neighborhoods are weakly organized against street crimes and strongly organized in favor of such crimes. Compared to other groups, the Mafia is strongly organized in favor of crime and weakly organized against crime. Compared to the US, Japan is strongly organized against crime, and weakly organized in favor of crime. Moreover, the group-level theory of differential social organization is linked to the individual-level theory of differential association. Groups that are strongly organized in favor of crime display numerous and intense definitions favorable to crime. Conversely, groups that are strongly organized against crime display numerous and intense definitions unfavorable to crime. It follows that differential social organization determines group crime rates by influencing the availability of definitions favorable and unfavorable to crime within a group [47]. When groups are strongly organized in favor of crime and weakly organized against crime, they will present an abundance of definitions favorable to crime and few definitions unfavorable to crime. Thus, individuals in such a group have a high probability of learning an excess of definitions of crime. Whether they do depends on their actual learning. Even in high crime communities, some residents are isolated from the abundant criminal definitions and exposed to the few anti-criminal definitions in the community. According to the theory, they will refrain from crime because of an excess of definitions unfavorable to crime. The opposite also holds. In low crime communities, some residents are exposed to the few criminal definitions in the community, and isolated from the abundant anti-criminal definitions. Given the opportunity and skills, they will engage in crime because of an excess of definitions favorable to crime. R.L. Matsueda Insights into Sutherland’s concept of differential social organization Sutherland [70] initially specified the root cause of crime to be social disorganization, a term he “borrowed from Shaw and McKay,” and viewed culture conflict as a special instance of general social disorganization [72:21]. Later, at the suggestion of Albert K. Cohen, he changed the concept to “differential social organization,” “with organization for criminal activities on one side and organization against criminal activities on the other side,” because “the organization of the delinquent group, which is often very complex, is social disorganization only from an ethical or some other particularistic point of view.” Although Sutherland did not specify the precise processes making up differential social organization, we can infer several processes from his writings. Shaw and McKay’s [61] concept of social disorganization – the breakdown in community institutions resulting in loss of control over youth – becomes weak organization against crime, and their concept of cultural transmission – older street gangs in disorganized areas transmit a delinquent tradition to groups of younger boys, resulting in high rates of delinquency across generations – becomes strong organization in favor of crime. Tannenbaum’s [76] concepts of conflict between spontaneous youth play groups and the adult community (which derives from [78]), the subsequent dramatization of evil, and the hardening process in which sporadic delinquency is transformed into delinquent careers is a key component of organization in favor of crime. Finally, seminar conversations between Sutherland, Alfred Lindesmith, Henry McKay, and Frank Sweetser about neighborhood delinquency rates contributed to differential social organization. Sweetser went on to write his PhD dissertation on personal neighborhoods, which drew parallels with differential social organization (Gaylord & Galliher, 1988). Analyzing network ties in a single block, Sweetser [74] found that residents’ networks of acquaintances and associates varied widely, that “personal neighborhoods” differed for most residents, and therefore, that person-to- person networks within spatially defined neighborhoods should replace family units or neighborhoods-as-primary-groups as neighborhood units of analysis. This explains why “many boys in the most delinquent areas fail to absorb the delinquent ‘tradition’ and remain law abiding”: the “culture of the delinquency area is a spatial interpenetration of a delinquent and a law-abiding tradition, perpetuated by differential acquaintance and association among neighbors” [74]). In other words, the neighborhood is differentially organized for and against crime. In laying out the theory, Sutherland never explicitly identified the content of diffe- rential social organization. His most explicit writings on the subject appear in a 1943 essay in Ogburn’s collection, American Society in Wartime, titled “Crime,” where he hypothesizes that increasing rates of theft during war may be due to differential group organization. War increases theft by undermining organization against crime, consisting of five processes [73:124–125]: First, the external opportunities for thefts increased because the owners of property spent a larger part of their time away from home, because supervision in industrial and commercial establishments decreased, because the railways carried increased loads of commodities without increase in the number of guards, because a large number of persons, not selected on the basis of trustworthiness, were placed in positions of responsibility, and because the police force decreased in number and training. Second, the efforts to train children and adults in opposition to stealing were relaxed Differential social organization, collective action, and crime or abandoned because parents were away from home and because schools and churches were closed or diverted to other purposes. Third, many persons, especially younger children, were left in complete idleness with no provision for supervised legal activities. Fourth, many of the poor people developed increased hostility toward some property owners who were regarded as hoarders or owners of goods by other illegal methods. Fifth, the meaning of property ownership and of property rights was confused by governmental appropriation of private property, by radical departures from the previous system of determining values and distributing property, and by general use of public property with little attention to its ownership. Clearly organization against crime is a multifaceted process that anticipates future criminological theories. Sutherland anticipated routine activities theory by positing that theft is increased when capable guardians are removed or undercut. He drew on social disorganization to emphasize the role of families, schools, and churches in controlling children through training and socialization against crime. He anticipated Hirschi’s [37] concept of involvement in conventional activities by specifying that involving children in legitimate activities prevented idleness, which reduced crime. He anticipated Sykes and Matza’s [75] technique of neutralization, “denial of victim,” by specifying a mechanism by which theft becomes increasingly justifiable when property owners are themselves viewed as criminals or hoarders. More generally, he maintains that when the sanctity of private property is weakened by government responses to war, laws protecting property are undermined. For Sutherland [73:125], war also increases theft by fostering organization in favor of crime, of which he identifies four mechanism: Sixth, contacts with criminal patterns were increased because of the passage of large numbers of children and women from the sheltered environment of the home to the heterogeneous environment of the factory, shop, and store, because of the great increase in the mobility of people, and because persons who had been stealing previously now stole with increased frequency and thus more effectively presented the patterns of theft to the non-thieves. Seventh, many public and private employees, especially railway employees, were in collusion with the thieves. Eighth, many persons who were not in economic distress were engaged systematically in stealing for the black markets. Ninth, a situation had occurred which was appropriate for theft according to the cultural definition which had been held for a long time and somewhat generally, namely, the situation in which theft is the only apparent alternative to starvation. Organization in favor of crime, then, is also multi-faceted. Sutherland shows that mobility can increase crime by moving law-abiding individuals (such as children and women) out of the protected private context of the home into heterogeneous public places, thereby increasing contacts with criminal patterns. He argues that when the frequency rate of theft (as opposed to the participation rate – see [12]) increases, an epidemic can result because of a sudden spike in criminal patterns made available to non-criminals. Black markets flourished by enticing those who were not in dire poverty and by colluding with employees – the latter anticipates Cloward and Ohlin’s [15] concept of bonds between conventional and criminal elements (the fence, fix, bail bondsmen) as a key structural element that help subcultures flourish. Finally, Sutherland observes that definitions of situations calling for crime interact with objective situations: the cultural belief that theft is justified when the alternative is starvation – an example of what Sykes and Matza [75] later R.L. Matsueda termed “appeal to higher loyalties” – will increase rates of theft only when individuals face (or perceive to face) starvation. Sutherland concludes that war is associated with higher rates of theft because the above organizational processes weaken social control of theft and strengthen patterns of theft. This implies that differential social organization is specific to forms of crime, such as theft, which increased during the war, and sex offenses and assaults, which did not. He also notes that this explanation of crime rates does not explain why organization changes. In his application, however, he is assuming that a fairly exogenous factor–war–changes local organization, which in turn explains changes in crime. Furthermore, Sutherland is assuming that crime rates are the result of both criminal patterns or organization and the reactions to crime or organization against crime. Thus, societal reactions or what later became known as labeling, is endemic to differential organization. Finally, and perhaps most importantly, Sutherland [73:126] argues that both forms of organization consist of two principal elements – “consensus in regard to objectives and in implementation for the realization of objectives.” In other words, such social organization is the result of collective action and entails building consensus over a problematic situation, and then translating that consensus into action. This problem can be addressed using Mead’s [52] theory of interaction and social control. Mead’s theory of symbolic interaction and social control In 1910, as a graduate student at the University of Chicago, Sutherland enrolled in Mead’s course on social psychology, and later wrote an article, “The Biological and Sociological Processes,” which essentially summarized Mead’s theories of role taking and the meaning of objects [69]. Nevertheless, Mead’s influence on Sutherland’s theorizing was, in all likelihood, mainly indirect, through the writings and teachings of W.I. Thomas [30:32]. Indeed, Mead’s concepts do not appear explicitly in the principles of differential association or differential social organization. But if we want to develop the dynamic portion of differential social organization, and link it to research on collective behavior, a return to Mead’s social psychology is useful. Mead begins with three assumptions. First is a methodological holism, in which “the whole (society) is prior to the part (the individual), not the part to the whole; and the part is explained in terms of the whole, not the whole in terms of the part or parts” [52:7]. Second is a social process model within which society, selves, and cognition, arise, and which can be studied by using the abstract concept of the social act, a cooperative act between two or more individuals. Third is an organic or functionalist social psychology, in which social acts are viewed as a “dynamic whole” (rather than as aggregations of stimulus–response sets), in which the component elements are analyzed in terms of their functions [52:7]. Taking the role of the other The key concept in Mead’s social psychological writings is role-taking, which occurs in social acts.1 Within an ongoing social process, social acts are built up by participants adjusting their responses to each other within an ongoing social process. When adjustments 1 Elsewhere, Karen Heimer and I have described Mead’s concept of role-taking and social control and applied it to delinquency; here I draw liberally from that discussion (see [48, 49: 169–170]). Differential social organization, collective action, and crime are smooth and routine, situations are non-problematic, and behavior non-reflective. When, however, an ongoing response or impulse is temporarily blocked, the situation becomes problematic, and individuals engage in role-taking, seeking solutions to the problematic situation by taking the role of others, viewing themselves as objects from the standpoint of others, and considering alternative lines of action from the standpoint of others [52]. Specifically, when an impulse is blocked by a physical or social barrier, an emotion – such as anger, fear, sadness, or happiness – is released, and the impulse is transformed into an image, which includes a plan of action and the anticipated reactions of others to the plan. The impulse is reacted to by another impulse, which follows the plan to overt behavior, combines the plan with another, or blocks the plan, causing the situation to remain problematic – in which case, the individual again takes the role of the other, forms a self as an object, and considers new alternatives from the standpoint of others. This process – the serial process of cognition – continues until the problem is solved or the social act fades. Mead [52] termed the image the “me” and the impulse the “I,” and specified them as two phases of the self – the self as an object drawn from the past, and the acting self responding in the present. By “solution” Mead meant not that the problem is resolved in an optimal way, but rather that it is solved for the practical purposes at hand, which means the blocked or conflicting impulses are freed and the social act is allowed to resume. Moreover, when similar problematic situations are repeatedly solved in functionally equivalent ways, they become less problematic, and behavior becomes habitual or non- reflective. In highly institutionalized settings, with strong norms, most behaviors are non- reflective, actors are not self-conscious, and stimuli lead directly to responses. At times, however, even normative behavior is interrupted by unanticipated or unconventional exigencies, and behaviors become reflective, actors take themselves as objects, and stimuli are mediated or interpreted by cognitive processes. Mead’s [52] theory of cognition consists of this dialectical inner-dialogue of the “I” and the “me,” which, in form and substance, resembles the “conversation of gestures” occurring between two individuals, except that it occurs between phases of the self in the mind. The response of the “I” occurs in the present, but only insofar as it has been called out by the “me” (a past) in terms of a specific anticipated future. Moreover, the “I” – or more precisely, the dialectical unfolding of the “I” and the “me” – contains an element of novelty or emergence, which stems from being in multiple perspectives simultaneously. Mead used the term “sociality” to refer to the ability to be in multiple spatio-temporal perspectives simultaneously – a prerequisite for role- taking, which is made possible by the use of significant symbols, which call out functionally identical responses in oneself as well as in others [54]. Mead’s analysis of the social act – an interaction among two or more persons, with an emerging goal, and division of labor – provides a framework for examining collective behavior within an interactionist perspective consistent with Sutherland’s differential social organization. Three key points emerge here. First, consensus over the goal of the act is an ongoing accomplishment possible through taking the role of others and using significant symbols to call out functionally equivalent responses, which constitute shared meanings. Such intersubjectivity is a practical accomplishment, and not an ontological imperative [40]. Second, individual decisions are practical solutions to problematic situations – and not optimal solutions – in that they are simply the first solution considered that removes the obstacle that blocked the ongoing act, which is arrived at through a trial-and-error cognitive process or imaginative rehearsal. The solutions considered derive from the information gleaned not only from the concrete actors within the social act, but also the rules and norms governing the larger social group within which the act is embedded. Behavior, then, is R.L. Matsueda better characterized as “satisficing” than utility maximizing [64]. Third, consensus over goals and means for attaining goals cannot be assumed a priori, but is often built up or constituted through reciprocal role-taking. In other words, the “I” can suggest new alternatives (“me’s”) and vice-versa. This parallels Dewey’s [23] theory of “ends in view,” in which ends are always in the present and are reciprocally related to means. Once a problematic situation is solved, and conflicting impulses are resolved through role-taking, a reconstruction of the situation has occurred, and a new self emerges from the old self: “Solution is reached by the construction of a new world harmonizing the conflicting interests into which enters the new self” [53:149]. The social structure of the self and stages of development of the self The self, then, arises in social interaction as an object, and thus, is socially constituted (given meaning) as an object in the same way other physical objects are constituted. For Mead [53:141], the organized society is prior to the individual, and the self has a definite social structure, which derives from the larger society in which the individual participates: “Inner consciousness is socially organized by the importation of the social organization of the outer world.” That structure is revealed when Mead describes the process of acquiring a mature self using his well-known analogy of “play” and “the game.” Early in life, children learn to play roles by taking the role of concrete others independently: they play policeman and arrest themselves; they play parent and scold themselves [52:150]. During this period, the child becomes aware of his or her body, learns to identify with the body (that is, draw a connection between the self and the body), and differentiate the body from the rest of the world.2 Later, having developed a sense of the body, and a rudimentary or compartmen- talized self, children diversify the self by learning the game, in which they take the role of the entire group or “generalized other,” including the norms, rules, and expectations governing various positions and roles of the group, community, or society. They learn to relate the rules, expectations, and obligations of their own roles to those of others within the organized system. This process of taking the role of the generalized other is the most effective form of social control because organized groups and institutionalized norms enter individual behavior. Stages of moral reasoning and linguistic codes For Mead, the movement from the stage of play to the game illustrates not only the development of the self and social control, but also the development of moral reasoning. Beyond the rudimental beginning of the game stage, expanding the scope of the generalized other increases one’s facility to solve moral problems by transcending conventional moral thought. The social structure and its conventional standards, which are unimpaired by the problem, allow individuals to use reason to abstract from morally problematic values to a more universal discourse containing a solution. The ideal is a universal society, in which perfected intelligence allows each member to take the role of others and arrive at the same 2 For Mead [52], play is organized around collective reciprocal role-taking, occurring around age four, whereas Piaget saw pretend play as “nonserious symbolic gestures, and occurring around age two [16:412]. Differential social organization, collective action, and crime meanings – that is, a universal understanding of others [52]. Joas [41:137] summarizes Mead’s “rational resolution of moral problem-situations”: It consists in taking into account and understanding all values which appear in a situation. This does not mean that one merely juxtaposes these values in a relativistic fashion; rather, it means questioning the claim to validity of each of these values, from the standpoint of bringing about a universal community based on communication and cooperation among its members. Comprehensive communication with one’s partners in a moral situation and orientation to the realization of this ideal society are, then, two rules for the solution of moral problems. Kohlberg’s stages of moral reasoning The developmental psychologists, Piaget and Kohlberg, have expanded this notion of stages in cognition and role-taking and applied them to moral reasoning. For Piaget, children initially experience, with their parents, a relationship of one-sided authority, constraint, and respect. Consequently, they are unable to understand the rationale behind rules, and thus, see rules as “external and unchangeable” [14:191]. Morality is one of constraint by authority figures and their rules. Later, however, children’s interactions with peers are as equals, and therefore, they must learn to cooperate by taking the role of the other and understanding the perspectives of others. Empirical research showing problems are solved better by groups of children than individuals is consistent with Piaget, as well as Mead [41:163–164]. The child’s egocentrism allows it to tolerate contradictory perceptions in isolation; however, in groups, presentations of contradictory perceptions by different individuals cannot be dismissed and, consequently, stimulates the group to resolve the contradiction through role-taking. The emergent rules are based on mutual agreements and therefore are not sacred, but can be changed by mutual agreement. A morality of cooperation is less the application of moral rules or the use of moral principles, and more a process of resolving moral conflicts by coordinating perspectives [14]. At the heart of cooperation is role-taking. For Kohlberg [44:134], stages of moral development “represent successive modes of taking the role of others in social situations,” which, he notes, is the position of Mead, Dewey, and others.3 Building on Piaget’s work, Kohlberg conducted empirical research on children and adolescents, in which he presents them with moral dilemmas, asks them their opinion of the dilemma, and most importantly, probes them for the kind of reasoning they use to arrive at the opinion. An example, which we use to illustrate Kohlberg’s stages, called the “Heinz dilemma,” involves a legal dilemma: “Should Heinz steal a drug to save his dying wife if the only druggist able to provide the drug insists on a high price that Heinz cannot afford to pay?” [17:9]. Based on answers to such scenarios, Kohlberg [43] identified six stages of moral reasoning, which fell into three ordered levels: (1) pre-conventional morality, (2) conventional morality, and (3) post-conventional morality (see Table 1). The first stage, “obedience and punishment,” is characterized by unquestioned obedience to authority, avoidance of punishment (which proves that disobedience is wrong), and an egocentric point of view that ignores the interests of others. Here the child would argue that 3 Joas [41:163–166] has argued convincingly that stage theories of Piaget, Kohlberg, and others fail to develop Mead’s theory of the constitution of physical objects through practical intersubjectivity; instead, a focus on presocial cognitive development precludes a “truly sociological research on socialization.” R.L. Matsueda Heinz should not steal the drug because it is against the law, is bad, or will result in punishment. The second stage, “individualism and exchange,” recognizes that individuals have different viewpoints and interests, individuals act in their own self-interest, what is right is fair exchange, and ultimately what is right is not absolute, but is relative. Here the child would recognize that stealing would be in the interest of Heinz but not the druggist. Those in stages 1 and 2 are in a pre-conventional level because they act as individuals rather than members of a collectivity. They are rooted in Mead’s play stage, able only to take role of concrete others serially and individually. In contrast, those who have progressed to level II, conventional morality, recognize that they are members of a larger group or society. The third stage, “good interpersonal relationships,” emphasizes being good, which means having good motives, showing concern for significant others, and maintaining trust and loyalty in close relationships. Shared agreements and expectations take precedence over individual interests. Here, youth can put themselves in someone else’s shoes; they would view Heinz’s motives (looking out for his wife) as good, and those of the druggist (overcharging) as bad, selfish, and greedy. The fourth stage considers the social system as a whole: youth here take the role of the entire group, recognizing the rules and laws that govern relationships among roles. Right is defined as contributing to society and abiding by laws unless a law conflicts with other social duties. This corresponds to Mead’s stage of the game. Here, youth might say they un- derstand Heinz’s predicament, feel his motives are good, but cannot condone theft because if everyone did it, the social order would break down. Both stages three and four operate at level II, conventional morality because neither critiques current conventional institutions. Those who have progressed to level III, post-conventional morality, are able to take a “prior to society perspective” in which they imagine what an ideal society might look like [17:22]. In the fifth stage, “social contract and individual rights,” young adults recognize the validity of the social contract, which protects personal and property rights, Bentham’s utility principle – “the greatest good for the greatest number” – and the need to integrate perspectives or conflicting values through legitimate mechanisms such as agreements, contracts, and due process. Here, in judging Heinz’s behavior, they might reason that life is more important than property, and therefore, Heinz’s actions are justified. In stage six, “universal ethical principles,” adults recognize that the validity of a law rests on underlying universal principles of justice, such as respect for the dignity and worth of individuals, and the equality of all persons. If a law violates the universal principle, the principle should be followed. To implement stage six, all individuals involved would take the role of all others to arrive at an impartial judgment. Thus, in the example, the druggist would take the role of the wife and agree with the others that her life trumps his profit. Thus, in stage six, a consensual decision is a likely outcome. Table 1 Kohlberg’s stages of moral reasoning Levels Stages Level I Pre-conventional morality Stage 1. Obedience and punishment orientation Stage 2. Individualism, exchange, and instrumentalism Level II Conventional morality Stage 3. Good relationships and mutual expectations Stage 4. Law and order Level III Post-conventional morality Stage 5. Social contract and individual rights Stage 6. Universal ethical principles Differential social organization, collective action, and crime It is important to note that different stages of moral reasoning can lead to the identical outcome. In the example above, a child at stage one may disagree with Heinz’s theft for the simple reason that it is against the law, while a youth at stage four may also disagree, but for the reason that protecting the social order takes precedence over Heinz’s good intentions. Nevertheless, we would expect that higher levels of moral judgment will be associated with moral behavior. If we assume that conventional laws, norms, and morals tend to approximate the ideal of the social contract, we would expect individuals operating at a high level of moral reasoning to solve their problematic situations in ways consistent with laws and norms. Whether one used the reasoning of social contract or individual rights or of universal ethical principles, the result will likely be consistent with conventional norms. Empirical research supports this view: On average, individuals operating at higher stages of moral reasoning tend to engage in moral behavior, such as being honest, resisting temptation, and engaging in prosocial behavior [9, 16]. But the relationship between moral reasoning and behavior is likely complex and context dependent. Criminal laws and norms always contain a political element, favoring some classes or groups over others – which makes the social contract imagery at best an abstract ideal, and at worst an ideological tool of obfuscation. For example, Garland [29] argues that “the key to understanding criminal law in class terms is to appreciate the ways in which particular interests are interwoven with general ones,” and the ways in which the protection of class interests are disguised as protection of universal interests. If this statement is true and transparent to members of society, we would expect that, at least at times, higher stages of moral reasoning will be associated with behavior at odds with conventional norms and laws. Citizens recognizing the class bias in particular laws will be sympathetic to the plight of those (perhaps themselves) at the losing end. Thus, the legitimacy of laws would be drawn into question. More broadly, stages of moral reasoning point to the complicated issue of applying general abstract norms to specific concrete situations. There is a strong tendency to slant such applications in ways that reflect one’s self-interest, which may affect all but those of stage six. For example, the principle of stealing drugs for a dying wife can be easily transformed to stealing illicit drugs for a needy girlfriend. Such vocabularies of motive, such as embezzler’s rationalization that they are “merely borrowing their firm’s money,” [21] are precisely the definitions favorable to crime that Sutherland [71] felt were the key determinants of crime. Even the application of post-conventional moral reasoning requires accurate information, untainted by the bias of in-group or individual interests, with which to make moral judgments that transcend the individual or narrow group. We know, however, that information and knowledge is slanted, biased, and framed in ways that tend to reflect the interests of individuals. Thus, the structure of communication networks may be crucial for understanding how such moral reasoning translates into behavior. Moral reasoning and social networks Moral reasoning may vary directly with social networks. For example, Coser [20] argues that in simple role-sets, characteristic of a Gemeinshaft, interactions will be restricted to similar others, behavioral expectations are predictable, and the intentions and meanings of others is transparent. Intentions of others can be taken for granted, a morality of constraint dominates over a morality of cooperation, and individuals have little incentive to develop higher stages of moral reasoning. Rigid norms govern behavior which everyone understands. Consequently, in simple role sets, speech patterns devolve into what Bernstein R.L. Matsueda [8] terms restricted codes, which are particularistic in meaning, predictable, and rigidly organized around a narrow range of syntactic alternatives. Restricted codes are context- dependent and known only among in-group members who share a common history, interests, and folk knowledge. By contrast, in complex role-sets, characteristic of a Gesellschaft, interactions include dissimilar others, behavioral expectations are unpredictable, and intentions and meanings are opaque [20]. To understand the intentions, meanings, and perspectives of others requires the mental work of role-taking, in which the perspectives of others are identified by locating them within a larger structure of roles (generalized other). A morality of cooperation is necessary, which provides an incentive to develop higher stages of moral reasoning. Norms are fluid and open to reinterpretation within interaction. Thus, in complex role sets, elaborate role-taking exposes individuals to new organized groups, and intellectual flexibility and individualism result. Speech patterns evolve into elaborated codes, which are organized flexibly around a wide range of syntactic alternatives, such that meanings are more universal and context-free [8:77–80]. Moreover, codes are structured by class. The lower classes tend to form small insular peer groups, use restricted codes, and follow rigid rules; upper classes tend toward heterogeneous groups tied together by weak bridging ties, use elaborated codes, and use higher levels of moral reasoning in applying flexible rules. Code switching, then, is a useful skill because “the ability to switch codes controls the ability to switch roles” [8:129]. Given that elaborated codes are prevalent among middle classes and within conventional institutions, like schools, members of lower classes are handicapped not only by their lack of resources but also their limited access to elaborated codes, which fosters intellectual flexibility. Because codes are linked to class through social ties, an important question is how are network ties structured? Network structure, collective action frames, and thresholds For Mead, like W.I. Thomas and Sutherland, social structure consists of the stable compo- nents of organized groups. At an abstract level, we may consider such structures as existing independently of content – that is, independent of the specific purposive action of actors who constitute and reconstitute those structures.4 For Mead, social organization is constituted through social interaction, and therefore, the critical elements of social structure are communication networks [62]. Individuals’ generalized others, which constitute their selves and informs their practical activity, are limited to those with whom they interact. But to whom do they interact? We can explore the answer to this question using recent work on structural ties. Social networks and social capital Let me begin with Coleman’s [19] concept of social capital and social ties. Although Coleman had a different project – beginning with a utilitarian model of rational action and linking preferences and information to structural ties in the form of social capital – we can draw upon his discussion of social ties to help answer our question of who interacts with whom. For Coleman, closed network structures enable greater social capital and social 4 In reality, of course, structures arise because of the historically specific purposes of actors at a specific conjunction of history and spatial context. Differential social organization, collective action, and crime control. Figure 1a depicts an open structure in which A and B are linked to C, but not linked to each other. A and B can independently and additively influence C by using individual sanctions, developing trust, establishing norms, using moral persuasion, and the like. But they cannot engage in joint behavior because they lack social ties. Figure 1b depicts a closed structure in which A and B are linked not only to C, but also to each other. They can not only influence C independently, but – because they interact with each other – they can jointly (multiplicatively) influence C by developing coordinated strategies, simultaneous sanctions, similar rhetorical arguments, and the like. In other words, they can each take the role of C, as well as each other, in developing a collective strategy to influence C. Coleman gives a second example of parents and children. Figure 2a gives an open structure in which child B is friends with child C, but their parents, A and D do not know each other, and therefore, can only influence their own child’s behavior indepen- dently. Figure 2b provides a closed structure in which the parents are linked. Again, the parents can now work collectively to control or influence their children, monitoring each others’ children and calling each other, coordinating their punitive strategies and rhetorical arguments, etc. Given common everyday knowledge of the way in which these structures operate, individuals can use the structures strategically, seeking out closed structures to increase control over members, and perhaps working to maintain open structures, such as 1a, to insure that A and B are unable to conspire against C, in a kind of divide and conquer strategy. Strategic parents will befriend parents of their children’s friends, and develop joint strategies when possible. Of course, other structural conditions will affect whether this is possible, including job flexibility, income and other resources, (which provide free time to engage in such strategies), neighborhood residential stability (which provides a stable set of parents who care about the community and its children), and the like. Similarly, strategic peers will befriend the friends of their friends – and develop cohesive cliques and peer a A B A B a a. Open Neighborhood Structure b b b. Closed Neighborhood Structure Figure 2 Open and closed neighborhood structures. C A B A B C a. Open Structure b. Closed Structure Figure 1 Open and closed social networks. R.L. Matsueda groups – to enable joint activity, including coordinating goals and actions, developing trust and norms, sanctioning rule violators, and so on. The strength of weak ties Closed structures tend to form dense networks of like-minded actors. Granovetter [32, 34] argues that assortative matching is typically based on homophily, which creates close relationships among similar individuals. We seek out and form close attachments with individuals like ourselves. Strong ties within a homogenous group not only encourage conformity but also lead to the circulation and recirculation of similar ideas. Such groups will tend to be stable over time, and have strong internal social control – through shared information, consensus over goals, and strong norms and sanctions – particularly when members identify with the group. The group’s homogeneity and closed structure will cause it to be rigid, lack cognitive flexibility, and have difficulty adjusting to changes in the environment. Innovation will come slowly if at all. As Coser [20:242] argues, “there may be a distinct weakness in strong ties,” because “the Gemeinschaft may prevent individuals from articulating their roles in relation to the complexities of the outside world.” In contrast, groups that are not entirely closed, but have weak ties to other groups, will benefit from information flows between groups. Thus, Figure 3 depicts two groups, 1 and 2, with a single tie or bridge between C and E, which links the two groups. The information flowing across the bridge will expose members of each group to novel ideas and information, since it is coming from a set of comparatively dissimilar individuals. Granovetter [32] argues that weak ties provide group members with information on the latest ideas, fashions, and job openings, as well as increasing the likelihood of members being organized into social movements. Conversely, the absence of weak ties not only isolates members, but presents obstacles to building a critical mass necessary to produce a political movement or goal-oriented social organization. Moreover, echoing Sweetser’s [74] suggestion for studying the person-to-person contacts within a neighborhoods, Fine and Kleinman [25:9] argue for linking the concept of subculture to cultural spread through interlocking group networks (see also [62]). B A C E G F Group 1 Strongly Tied Group 2 Strongly Tied D H Local Bridge (Weak Tie) Figure 3 Local bridge (weak tie) linking two distinct closed networks. Differential social organization, collective action, and crime This conception is consistent with Mead’s [52] notion that creativity, innovation, and intellectual flexibility for solving novel problems is increased when one participates in multiple disparate groups, taking the role of those groups, and considering problems from their unique perspective. Expanding the generalized other toward the theoretical limit of a “universal other” maximizes a person’s ability to solve novel and complex problems by considering a multitude of perspectives, and even combining those perspectives. Such structures may be linked to classes. For example, following Blau [10], Granovetter [34] suggests that strong ties may be more frequent in the lower classes because of the tendency toward homophily and the pyramid-shaped class structure of modern societies, with few at the top and the masses at the bottom. The result is that lower class individuals will tend to form strong ties because there are so many lower class individuals with whom to form such ties, while upper class individuals will form a large number of weak ties – many of which are to the numerous lower class individuals – because there are so few upper class individuals with whom to form strong ties. This purely structural argument would lead to a fragmented lower class, with small isolated pockets of strongly tied groups, which require more time and energy to maintain, and an upper class characterized by diffuse weak ties, requiring less time, but yielding greater information and intellectual flexibility. To overcome their small number, members of upper classes must strategically form strong ties to one another through elite clubs and schools. Moreover, the poor and disenfranchised, isolated in parochial groups, perceive few alternatives to their problematic situations and adapt by investing in strong reciprocal ties to kinship networks [34:213]. Such structures provide a basis for theorizing about organization against crime and organization in favor of crime. For residents of affluent neighborhoods, who enjoy regular employment, good incomes, and sufficient time and resources to address local problems (such as delinquency), a mix of strong and weak ties is empowering. Strong ties enable such residents to reach consensus about shared problems, agree on promising solutions, and work collectively to try out such solutions [e.g., 58]. Weak ties to outsiders enable them to introduce innovative solutions by providing fresh ideas and information, and to draw directly on ties to outside social agencies. For residents of disadvantaged neighborhoods, high rates of residential mobility, poverty, and lack of time and resources undermines their ability to reach consensus about crime control beyond the kinship network, identify novel ways of controlling crime, link to other agencies, and act collectively. Moreover, in such neighborhoods, disadvantaged youth, who have high rates of school failure and bleak labor market trajectories, have a strong incentive to develop alternate ways of gaining status, perhaps in illicit ways [e.g., 2, 15]. Such innovation may be more likely when the group of disadvantaged youth have weak ties to other disadvantaged groups who share the same objective situation. To link these structures to instrumental action, we turn to theories of collective action. Collective action, frame analysis, and individual thresholds Let us begin by using Sutherland’s terms, “organization in favor of crime,” and “organization against crime,” but with the caveat that these are analytical abstractions, rather than concrete phenomena. It is only a slight exaggeration to say that organization lies inert, waiting to be actualized by actors’ instrumental action; it is an exaggeration because in reality, organization and the instrumental actions it may facilitate interact. Like social capital, the same social organization used for legitimate ends can be appropriated by actors for criminal ends [e.g., 18:S108]. In reality, the two are related in complex ways, at times R.L. Matsueda causally related – as when expanding criminal patterns cause conventional groups to organize against crime – and at times intertwined within a single fabric – as when violence is generally condemned, except when committed against outsiders. Again for analytic purposes, we can distinguish two polar opposite forms of social organization. The first, which may be termed “institutional action,” are those acts which occur in institutional contexts, entail routinized, formal, and norm-following behavior, and typically are found in formal organizations. The extreme case for us would be responses to crime that are carried out by the criminal justice system, a highly bureaucratized organization. The second, which we term “collective action,” are those acts commonly defined as occurring outside of institutional contexts in informal groups or gatherings, tending to be more spontaneous and creative, and requiring the building of coalitions and consensus in the absence of a strong normative system. I will focus on collective action, rather than institutional action, for three reasons. First, much research has been done on the formal role of the legal system, schools, and families on crime and delinquency. Second, with a few exceptions, such as the Mafia and drug networks, the threat of formal sanction by the state has consistently kept institutional action from becoming a prominent feature of most criminal activity. Third, the model of collective action among informal groups entails dynamic mechanisms that are relevant to institutional action within formal organizations when a novel problem arises, and rules and routines break down. In addition, institutional action typically originates in collective behavior, which then becomes useful for solving local problems, and is then routinized and institutitionalized [e.g., 7]. Of course, not all criminal acts or reactions to crime are collective acts of more than one individual. But, as Mead [52] argued, the prototypical behavioral act is thoroughly social, the individual is an abstraction from the group, and we can explain isolated individual acts with the same model.5 Symbolic interactionism played an important role in studies of collective action during the 1960s and 1970s with theories of spontaneity and emergent norms [42]. By the 1980s, however, such ideas were on the decline, replaced by resource mobilization theory, which argued that rational actors seeking to maximize utility draw upon social networks, pre- existing organization, and political opportunities to bring about collective action. In the past two decades, students of collective action have returned to interactionist ideas using the concept of frame alignment [6]. The argument is that the structure of network ties, political opportunities, and institutional support helps explain opportunity structures for collective action, but has little to say about the moment-to-moment dynamics of an emerging social movement, and in particular, how the framing of grievances may foster social movements. Drawing on Goffman’s [31] frame analysis, Snow et al. [68] view individuals as “signifying agents,” who actively produce, maintain, and fight for meanings about issues they find important, and “struggle to have certain meanings and under- standings gain ascendance over others” [67:587]. They produce “collective action frames,” which are emergent beliefs and meanings that foster social movements by framing a problematic situation as calling for an action-oriented solution. In other words, they help build, in Sutherland’s [73:126] terms, “consensus in regard to objectives and in implementation for the realization of objectives.” Framing is the process of constructing reality in ways that foster collective action and “entails agency in the sense that what evolves is the product of joint action by movement participants in encounters with antagonists and targets,” and is also “contentious in the 5 This is unlikely to apply to rare instance of behaviors that are the direct result of mental impairments. Differential social organization, collective action, and crime sense that it generates alternative interpretation schemes that may challenge existing frames” [67:587]. Snow et al. [68:478] argue that the process of frame alignment – linking the interpretive frameworks of individuals and social movement organizations – is the key task for social movement organizers. This includes frame bridging (linking ideologically congruent but structurally unconnected frames), frame amplification (invigorating frames by tying them to core values and beliefs), frame extension (enlarging a frame’s boundaries to incorporate views or interests salient to others), and frame transformation (radically changing a frame to secure participation and support). In addition, frames can be made more resonant by presenting them as salient and credible, and strategically diffused over time and across space [6]. Moreover, frames vary by Bernstein’s [8] restricted versus elaborated codes: restricted frames are narrow, closed, and difficult to extend or amplify, whereas elaborated frames are syntactically flexible, easily elaborated, and thus, more inclusive of disparate aggrieved groups [66]. The process of framing is rooted in Mead’s analysis of the social act. Using Mead’s terms, frames are constellations of meanings of objects rooted in the organization of groups. Moreover, they can be called up to solve problematic situations in the present, whereby they are modified and jointly constituted within social acts, calling out specific responses in self and other. Thus, frames are ways of interpreting a world out there, which helps marshal collective solutions to problematic situations. As individuals buy into the frame, participate in the social movement, and commit resources to the cause, they develop a personal identity from the activity, and when acting together, develop collective identities – that is, a shared sense of selves acting jointly to accomplish a common goal [e.g., 79]. Successful social movements, of course, require objective opportunities, including “the amount of resources at the discretion of potential constituents, the degree of previously existing organization among potential constituents, the configuration of allies and enemies, the social control policies of authorities, the strategy and tactics of organizers, and – overall – the structure of the political economy constraining mobilization and wielding of resources” [26:10]. What makes a frame more effective in mobilizing others to action? Gamson [28:155] argues that effective collective action frames contain three elements: “(a) it defines the root of the problem and its solution collectively rather than individually; (b) it defines the antagonists – ‘us’ and ‘them’; and (c) it defines an injustice that can be corrected through a challenger’s action.” Thirty years earlier, Cloward and Ohlin [15] made a similar argument about lower class boys who face structural barriers to success, and consequently a problem of social status. If they attribute their failure to an unjust system – “them” as opposed to “us” – they will seek out collective solutions to their status problems, such as innovating illegal ways of attaining success. Individual Thresholds and Collective Action Theoretically, there is reason to believe that beyond the antecedents of collective action discussed so far – the structure of network ties, political opportunities, and institutional support, as well as the successful situational framing of a grievance – we need to consider an additional situational mechanism in determining whether an individual joins an emerging collective action. This is Granovetter’s [33] notion of individual thresholds of collective behavior. Drawing on Schelling’s [55] threshold model of residential segregation, Granovetter [33] posits that each individual has a threshold for joining the collective action, which is defined as the proportion of the group that has already joined. Once that threshold has been exceeded, the individual will join, all else being equal. Thus, if one has a threshold of 0.6 and only half of the group has joined a riot, one will stay away from the riot. Granovetter [33] interprets this R.L. Matsueda effect within a rational choice framework: the individual with a high threshold either acts on a belief in safety in numbers (the certainty of arrest declines with number of rioters) or norms of fairness. Thus, he assumes that for individuals, “given their goals and preferences, and their perception of their situations, they act so as to maximize their utility” [33:1422]. But what explains individual differences in thresholds? Granovetter is rather vague on this issue, suggesting they result from background characteristics, personality, identities, available alternatives, sensitivity to the opinions of others, and preferences such as beliefs or norms. From the standpoint of Mead, the most important determinant would be a conception of self or identity – formed from the standpoint of significant others – followed by availability of alternatives, and sensitivity to opinions of others. Thus, those with an image of self (from the standpoint of others) as a “rabble-rouser” will have a low threshold for rioting, whereas those with an identity as a “conservative” will have a high threshold. Moreover, identities that are embedded in the emerging organization of collective action – the generalized other with interrelated roles – will be particularly salient. A rabble-rouser who has instigated riots within the group at hand will have a very low threshold. Also consistent with Mead [52], thresholds, like identities, are not completely stable, but can change within the course of a situation, such as when police arrive on the scene, increasing the costs of rioting. Thresholds also tend to be specific to forms of collective action, such as participating in a riot, breaking up a fight among kids, or being a good Samaritan. Threshold models reveal surprising macro-level outcomes from assumptions of initial distributions of thresholds. A simple example from Granovetter [33] illustrates this point well. Begin with a street corner in which 100 people milling around have a uniform distribution of thresholds to riot ranging from zero to 99. The instigator, with a threshold of zero (meaning he will riot in isolation) throws a brick through a window, followed by the person with threshold 1, then 2, and so on until all 100 individuals have rioted. This is a cascade or bandwagon effect, with equilibrium value of 100. Now, repeat the scenario but replace the person with threshold 1 with a threshold of 2. The instigator breaks a window, but no one else joins. The distributions of thresholds in the two populations are nearly identical, but the outcomes are radically different. Of course, all members of a group are not equal – some may be friends and have more influence on one’s decisions. Granovetter shows in this simple example that the first equilibrium of 100 rioters is very unstable in the face of small changes in friendship structure, whereas the second equilibrium of one rioter is very robust to such changes. The important point here is that collective action is not merely a result of access to resources, successful framing (or definition of the situation), or the absence of alternatives, but also has a dynamic temporal dimension that emerges in the situation, which is consistent with the social process model of Mead and other pragmatists. Social organization against crime We can apply collective action frames and individual thresholds to social organization against crime. I will do this by example, first describing a few social movements against crime and then providing a more detailed description of mechanisms by which collective action frames and individual thresholds translate social capital and collective efficacy into collective action against crime. In each of these examples, I focus on the collective action dynamic, and take the constellation of institutions – the static element of differential social organization – as a given. Differential social organization, collective action, and crime Social movements against crime A number of social movement studies have examined collective action that organizes groups against various crimes. For example, Becker [5] argued that the creation of deviance and crime begins with the process of rule creation, which is often instigated by moral crusaders, who are typically from upper classes, motivated by humanitarian concerns, and preoccupied with the substantive ends (rather than the logistical means) of their crusades. Moral crusaders bring attention to the problem using the mass media, and marshal support from various interest groups that may have disparate interests in seeing the rule passed. Moral crusades tend to have a natural history, beginning with a broad set of values – like self-determination or the protestant work ethic – deriving a specific rule based on those values, and then creating a bureaucratic system to enforce those rules. The job of enforcing rules falls to “rule enforcers” – police, prison guards, security – who tend to be more concerned with the bureaucratic imperatives of enforcement than with the actual substantive content of the rules. Becker [5] applies this framework to the passage of the Marijuana Tax Act of 1937, arguing that the Federal Bureau of Narcotics (FBN) used the news media to create a national panic over marijuana use, compelling Congress to enact legislation.6 McCarthy [51] argues that during the 1960s the “auto safety frame,” in which auto accidents are attributed to unsafe design features of automobiles gained ascendance under the leadership of Ralph Nader, and led to the creation of the National Highway Traffic and Safety Administration (NHTSA). The “drunk driving frame,” in which auto accidents are attributed to drinking and driving remained in the background until a report delivered to Congress in 1968 highlighted the number of crashes and deaths resulting from drinking and driving, attributed the problem to personal responsibility of the “drunk driver,” and pro- posed legal solutions to the problem. NHTSA committed to the frame, waged a “war on drunk driving,” and funded DUI law enforcement grants and demonstration programs. But, it was not until the late 1970s that a citizen’s movement brought drunk driving to the public consciousness. The movement included national organizations with local chapters, such as Mothers Against Drunk Driving (MADD) and Remove Intoxicated Drivers (RID), suc- ceeded in getting their message to the media because of their raw emotional message, emphasis on victimization, and local chapters in close proximity to local media [51]. Nevertheless, autonomous agents of the state played a key role in subsequently legitimating, dissemi- nating, and embedding the frame in law enforcement and judicial practices. Moreover, once the drunk driving frame became widespread, it provided a “collective identity opportunity” for other victims of drunk driving to conceive of themselves as victims [51:157]. Jenness and Broad [39] argue that hate crime legislation resulted from the intersection of four social movements. The civil rights movement provided a model for addressing discrimination and intergroup violence using government policy. The women’s movement altered the discourse on violence against women from a private personal problem dealt with in the privacy of the home to a public social problem involving a specific social category of victims – namely women. The gay and lesbian movement broadened the definition of what constitutes legitimate subjects of discrimination and reshaped discourse on sexual politics and violence. Finally, the crime victims movement pushed to articulate and expand the 6 Others, however, have argued that the act was either rooted in racist fears among law enforcement officials that marijuana fueled violence by Mexicans, rooted in an attempt by the FBN to expand its authority and scope, or rooted in a fear that it was associated with dropping out of society (see Galliher and Walker, 1977; Himmelstein, 1983). R.L. Matsueda rights of victims of crime, which are assumed inadequately addressed by the criminal justice system. Civil rights organizations such as the NAACP began documenting instances of violence against blacks and other minorities, as part of the racial justice movement. A coalition of organizations, using frames of racism and hate, pushed for legislation to address bias-motivated crimes. Jenness and Broad [35] show how the gay and lesbian movement appropriated frames from the women’s movement, such as “sexual terrorism,” to include violence against gays and lesbians as instances of hate crime. In a study of the diffusion of hate crime laws across states, Grattet et al. [35] find that early adapters of hate crime laws experimented with a variety of legal strategies (e.g., penalty enhancement, civil rights violation, ethnic intimidation), but later adapters coalesced around a couple (penalty enhancement and ethnic intimidation). They also find that the domain of hate crime expanded over time, as more classes of victims were included (sexual orientation and disability) and more forms of violence were included. Finally, Jenness and Grattet [40] examine enforcement of hate crime laws, and find that within the courts, appellate cases gradually arrived at a consensus over what constitutes a hate crime (settling on the broad definition of bias-motivated crime over a narrower conception requiring animus), and this definition was then followed by prosecutors and police. In this way, the definition was institutionalized. Collective action against crime: informal social control and collective efficacy Perhaps the most powerful deterrent of crime is informal social control, in which families, neighborhoods, schools, and other institutions seek to control crime informally without resorting to the legal system. Morenoff et al. [55] have used the concept of collective efficacy as a neighborhood-level concept defined as the “willingness of local residents to intervene for the common good,” which is largely dependent on “conditions of mutual trust and solidarity among neighbors” (Sampson & Raudenbush, 1997:919). It is a collective counterpart to self-efficacy, and like the latter, is “relative to specific tasks such as maintaining public order” [57:108]. Concretely, collective efficacy is measured by neighbors’ willingness to intervene when groups of children are hanging out, fighting, vandalizing, and the like. Moreover, Sampson et al. [58] link collective efficacy to the concept of social capital, arguing that intergenerational closure (ties between the parents of different children in the neighborhood) and reciprocated exchange (exchange of advice, favors, goods among neighbors) provide the “resource potential of personal and organizational networks” for children, which is realized in collective efficacy. In turn, the density of social ties are conditioned by neighborhood disadvantage versus affluence, including socioeconomic status, racial composition, and residential turnover. But the relationship between structural ties and the capacity of neighborhoods to solve problems may be more complex. Insular neighborhoods with dense networks may be effective in addressing some problems but not others. Uncomplicated problems – such as youth outsiders coming into the neighborhood and assaulting local youth – for which prior consensus exists may be readily solved, as solutions are known and straightforward (e.g., activating a neighborhood watch or calling the police). But more complicated problems – such as resource-poor local residents of a housing project seeking social control of violence – may be more difficult to solve. Such problems may require innovative ideas, intellectual flexibility, and high stages of moral reasoning characteristic of affluent neighborhoods containing networks with numerous weak ties to dissimilar groups. Thus, it is not only the density of ties within a neighborhood, but the distribution of weak ties that Differential social organization, collective action, and crime determines social capital. In general, strong ties may be more important for maintaining trust and cooperation (as in a Gemeinschaft), but weak ties more important for disseminating new information (as in a Gesellschaft). Moreover, this characterization of neighborhood control does not describe the dynamics of informal control. That is, given they are more or less embedded in local networks, how are residents able to act collectively to reduce crime and disorder in the neighborhood? I argue that moral, rational, and pragmatic persuasion through conversation is a key element of collective efficacy, and that collective action frames are a valuable resource to help create consensus over the specific forms of behavior requiring control and over the specific ways of achieving that control. In other words, there is a direct parallel between the dynamics of social movements, such as movements to control hate crime, and local collective behavior, such as achieving neighborhood or familial control over youth. In each case, consensus must be built up before acting collectively. In the case of neighborhood collective efficacy, residents use a “neighborhood frame,” in which the residents draw upon values of a safe and clean neighborhood, appeal to neighborhood pride, and create collective identities as neighbors [e.g., 65], and an “anti- crime frame,” which emphasizes the evils of delinquency, drugs, and misbehavior. Thus, a neighborhood with collective efficacy is one in which residents are able to use neighborhood and anti-crime frames to mobilize one another to achieve consensus over objectives, such as insuring a safe neighborhood, and procedures to attain objectives, such as supervising children of other families, being alert to unusual people or events in the neighborhood, and keeping youth away from bad elements and bad places (e.g., staging grounds for fights or parties for doing drugs). Elaborated, rather than restricted, neighborhood frames may be more effective in appealing to a broad set of residents. Individual residents will vary in the value they place on safe neighborhoods, and more importantly, in their own ability to persuade others. In this context, we define social efficacy as an individual’s ability to create consensus over group (neighborhood) objectives and procedures, and translate the procedures into action.7 Such individuals use higher stages of moral reasoning to consider not merely the parochial issues that affect their own self- interest, but also those that affect the community as a whole, including the way in which various roles operate within the neighborhood and operate between the neighborhood and relevant institutions. They would likely be proficient in elaborated speech, but in the case of mixed neighborhoods, be capable of code switching. Such individuals would correspond to indigenous or natural leaders described by Thomas and Znaniecki [77], who help reorganize a disorganized group by discovering new schemes of behavior, lobbying the group to accept the new schemes, rules and institutions, and overcoming the objections of defenders of the traditional system.8 One might be tempted to hypothesize that the greater aggregate or average social efficacy, the greater the collective efficacy in the neighborhood. But this simple hypothesis glosses over important heterogeneity within neighborhoods. For 8 In a different context, efficacious residents correspond to “ institutional entreprenuers,” organizational actors who create new institutions or organizational forms when they have sufficient resources and “see in them an opportunity to realize interests that they value highly” (DiMaggio 1988, p. 14) 7 This differs from Sampson’s notion of collective efficacy, which is a property of groups or neighborhoods, rather than individuals. It is a more specific application of Bandura’s [3:71] concept of self-efficacy, which refers to “people’s beliefs about their capabilities to produce designated levels of performance that exercise influence over events that affect their lives.” Social efficacy refers to an objective ability to organize social groups to realize a common goal, rather than a perceived belief about one’s capability to produce general effects important to one’s life. R.L. Matsueda example, neighborhoods may need only a small number of efficacious individuals, and beyond a threshold, there may be diminishing returns to such individuals. Moreover, the way in which an efficacious individual is embedded in a neighborhood’s relations may be critical for collective efficacy. For example, in a network highly centralized around a well-connected hub (a node with high degree centrality and between- ness centrality), if the hub is a socially efficacious individual (who values the neighborhood), the neighborhood’s structure is conducive to collective efficacy. With social ties to nearly all residents, the socially efficacious hub is in a position to mobilize residents to improve the neighborhood (see Figure 4). In contrast, in an identical neighborhood network structure, but with an inefficacious hub, the presence of two efficacious residents on the network’s periphery is unable to compensate for the B A C E G F J is an Efficacious Resident D H I J Figure 4 Structure conducive to collective efficacy. B A C E G F C and D are Efficacious Residents D H I J Figure 5 Structure not conducive to collective efficacy. Differential social organization, collective action, and crime inefficacious hub (Figure 5). This is because their structural location limits their ability to mobilize their neighbors. Of course, if they were truly efficacious, they might be able to use their agency to overcome their structural limitations and, for example, make a concerted effort to get to know their neighbors. But such efforts face an uphill battle compared to the naturally linked efficacious case (natural leader): residents may look with a wary eye on a neighbor looking to get to know them for instrumental reasons. Beyond a neighborhood’s structural ties among roles, the occupants of those roles are important for collective action. Specifically, a neighborhoods’ distribution of individual thresholds among role-occupants will determine whether cascades occur; in turn, the thresholds will be a function of the extent to which their personal identities include being a good resident of the neighborhood, the degree to which they have internalized neighborhood frames, the value they place on safe neighborhoods, and the degree to which they view delinquency as negative. The ways in which the distribution of individual thresholds interact with the distribution of structural ties, the distribution of socially efficacious residents, and the use of neighborhood frames will vary by the kind of problem facing the neighborhood. A simple problem, such as a couple of local children vandalizing houses, may require only a couple of parents to remedy. Clearly, neither elaborate framing, nor natural leaders, nor low thresholds are needed to resolve the problem. But a more complex problem, such as an alleged crack house being set up on the neighborhood’s periphery and enticing young residents into the business, requires a much greater degree of coordination and organization. Here, strong ties among residents including efficacious leaders with low thresholds, participation in voluntary associations such as a neighborhood watch, and convincing framing by those most affected may be necessary to mobilize most of the neighborhood. Moreover, neighborhood ties to outside organizations, such as law enforcement and local housing authority, may be needed. Complicating matters, such collective action may be impeded by criminal elements enmeshed in the neighborhood, who have an interest in maintaining their activity unencumbered by local residents. This discussion assumes that the composition of the neighborhood – the collection of individuals giving rise to collective efficacy – is exogenous. That is, the residents of a neighborhood are not changing in ways that influence crime. There is, however, good reason to expect exogeneity not to hold. Residential mobility in urban areas is typically a bit under 20% per year, and varies by renters (over 35%) versus home-owners (less than 10%), and youth (40% for those younger than 35) versus elder (11% for those older than 35). Thus, the composition of neighborhoods is changing sufficiently to question whether this mover–stayer process is in equilibrium. If the propensity of potential new residents to move into the neighborhood and that of old residents to move out of the neighborhood is orthogonal to the neighborhood’s crime rate, selectivity into and out of the neighborhood can be safely ignored. In the treatment effects literature, this would qualify as satisfying the “ignorability” criterion. Unfortunately, access to good schools and neighborhood crime rates are two of the most important considerations in housing decisions [e.g., 24]. Therefore, residential mobility decisions – which change the composition of neighborhoods – may confound research on the effects of neighborhood composition, such as collective efficacy. We can give two stylized examples to illustrate this point. At one extreme, suppose an affluent neighborhood has a long tradition of safety. The absence of crime may be due to the distant history of collective efficacy of the neighborhood, but that history has been altered by residential succession. The low crime rate will attract new residents who are less-tolerant of crime, wealthy socially efficacious people, but possibly also wealthy R.L. Matsueda isolates, looking to reap the rewards of the neighborhoods’ legacy of safety or free ride on the work of a few other residents who are highly socially efficacious. At the other extreme is a neighborhood with a long history of high crime rates. Residents who move in are those unable to afford safer neighborhoods, may be more tolerant of crime, or in the case of minorities, face housing discrimination elsewhere. Those residents who are able to move to better neighborhoods do so, leaving behind the less-efficacious and more impoverished. It could be that the correlation between collective efficacy and crime is confounded by these historical changes in residential composition. In the extreme case, neighborhood crime rates are due primarily to historical processes, or state dependence, which also explains the composition of the neighborhood, including social ties and perceptions of collective efficacy (respondents infer collective efficacy from the low crime rates). The most important point here is not that models of neighborhood collective efficacy may yield biased estimates, but rather that neighborhood residential mobility is an important part of the process by which neighborhoods affect crime. Examining both residential moves and collective action may uncover unanticipated results. Research could, for example, specify an individual threshold model of residential preference based on the number of criminals in the neighborhood [e.g., 60]. Such a threshold is likely to correlate with a collective action threshold for interceding in neighborhood problems. Combining the two threshold models would likely yield interactions between the two dynamic composition effects and other non-linear effects. Social organization in favor of crime Crime as Collective Action Crime rarely coalesces into a social movement – unless one would include revolution as an instance of crime. As noted by Sutherland [73], this is principally because states use legitimated violence to quell criminal movements, such as riots and collective violence, before they become widespread. Crime, however, is more often than not committed in groups of two or more persons. Therefore, if we follow Blumer’s [11] classic definition of collective behavior as essentially non-institutional group forms of spontaneous action, most crime falls under the rubric of collective action. We may speak of “collective criminal behaviors” to separate them from isolated acts of individuals. To explain collective crimes, we begin with the same framework we used to explain collective acts against crime: objective opportunities, including access to resources such as physical ability, weapons, and suitable targets, degree of prior organization among participants, alliances and enemies, and in particular, the social control policies of political authorities. The threat of formal sanction clearly separates crime from other forms of collective action. Conditional on these opportunities, then, the framing of criminal behavior is key in persuading others to join in. I will use two classic ethnographic studies to illustrate this explanation, Becker’s [4, 5] study of learning to be a marijuana user, and Anderson’s [2] study of violence and the code of the street. Becker [5] found that novice smokers must learn how to smoke marijuana, including how to inhale and hold the smoke in the lungs, how to recognize the effects of being high, and how to define the effects as pleasurable. In this way, an inherently ambiguous physiological experience – dizziness, nauseous, euphoric, or comical – is transformed and redefined into a collective action frame, in which a physiological state is defined as being “high,” and more importantly being pleasurable. The specific elements of the “weed Differential social organization, collective action, and crime smoking frame” are built up in interaction in groups, as other experienced members help demonstrate how to smoke properly, how to recognize the feeling of being high (including having the “munchies”), and how to interpret the high feeling as pleasurable and even euphoric. Thus, “marihuana acquires meaning for the user as an object which can be used for pleasure” and with repeated experiences of this sort, “there grows a stable set of categories for experiencing the drug’s effects” [5:56]. In this way, “deviant motives actually develop in the course of experience with the deviant activity” [5:42]. Here vocabularies of motive are constructed, such as “everyone is doing it, it’s cool,” and “no one gets hurt.” Because marijuana is illegal, whether the beginner progresses to an occasional user and then to a regular user depends on how he or she adapts to social control attempts to limit supply of the drug, detect drug users, and define the behavior as immoral. Through interaction, regular users develop contacts with drug dealers, learn verbalizations that neutralize definitions of the behavior as immoral, and deal with the possibility of being caught by segregating acquaintances into users versus nonusers, withdrawing into groups who condone marijuana, or concluding that detection would not be so bad. Through these processes, regular users adopt a stable self- concept as a marijuana smoker. Moreover, in a concrete situation of a party, the distribution of individual thresholds may determine whether marijuana smoking becomes joint behavior. If an instigator lights up, but within the group, all thresholds are above 1, the lone smoker might feel compelled to put the cigarette out. However, if the instigator provides a compelling frame, a threshold might be lowered to a 1, causing a cascading effect if other thresholds are, for example, clustered around three. If a novice user defines the situation of his first puff as aversive rather than pleasurable, and expresses displeasure, it may be sufficient to raise those thresholds clustered around 1 or 2 to stop the cascade. Becker’s discussion ignored social processes causing individuals to select or be selected into the party. Thus, objective opportunities are again important for collective behavior. Those opportunities include social networks to other partiers who have connections to drug dealers and can afford to purchase marijuana and other drugs. A second example is that of inner-city youth violence. A long history of research in criminology has found violence in impoverished inner-city neighborhoods is bound up with cultural responses to structural disadvantage. Facing structural barriers to gaining status and respect within the educational system and labor market, disadvantaged youth turn to the streets to gain a semblance of respect among other disadvantaged youth. For example, Cloward and Ohlin [14] argue that neighborhoods with high rates of gang violence are characterized by restricted legitimate opportunity structures and relatively prevalent illegitimate opportunity structures, consisting of prospects for gaining status through violence presented by turf gangs. The two structures, legitimate and illegitimate, and the cultures they produce, are intertwined. Such structures may lead to isolated networks of strong ties among similarly situated disadvantaged individuals. Isolated from interactions with dissimilar others (like members of upper classes), members of lower classes may experience fewer incentives to take the perspectives of others within complex role sets, resolve moral conflicts by coordinating perspectives, learn elaborated speech to convey particularistic perspectives in a universal medium, and develop the capacity for intellectual flexibility. Moreover, we can conceptualize the cultural components as “cultural frames,” which are used to define and make sense of situations [31]. In the abstract, legitimate and illegitimate frames are contradictory; at times that contradiction is starkly revealed in concrete situations. For example, using the term “code” rather than frame, Horowitz identified two cultural codes that structure behavior in Latino neighborhoods. The legitimate frame, the R.L. Matsueda instrumental code of the American Dream, is organized around economic success, espoused by community members, but in conflict with the reality of lower class schools and available jobs. The illegitimate frame – the code of honor on the streets – is organized around respect, manhood, and deference, and is espoused by young men. Violations of the code of honor – “actions that challenge a person’s right to deferential treatment in public” – can be “interpreted as an insult and a potential threat to manhood” [38:81]. In response, men must restore their honor, respect, and dignity by “being a man,” which calls for physical acts of violence and intimidation. Street identities of young men are shaped by their responses to challenges, negotiations of threats to manhood, and ability to maintain honor. Latino youth, according to Horowitz, must balance the instrumental frame, which requires being “decent,” against the honor code of the street, which requires being “street” or “bad.” We can speak of a “street frame” that is used to make sense of situations on the streets of inner-city impoverished neighborhoods. Like other frames, street frames contain vocabu- laries of motive, rules, and tacit sanctions for violating rules. Anderson [2] has identified the dimensions of rules or norms within the street frame. The most fundamental norm is “never back down from a fight.” Violations of this rule will result in a loss of street credibility and social standing, loss of self-esteem, and an increase in the likelihood of being preyed upon in the future. Status on the street is achieved by demonstrating “nerve” – a willingness to express disrespect for other males by getting in their face, throwing the first punch, pulling the trigger, messing with their women – which builds a reputation for “being a man.” Moreover, the phrase, “I got your back,” implies that a street youth will protect his friends and loved ones from insult, disrespect, or attack from others. Indeed, an insult or assault on one’s “crew” calls for revenge or payback. Underlying the notion of vengeance and never backing down is a norm of reciprocity, in which one is expected to respond in kind when disrespected by challenges, name calling, and violence. One of Anderson’s [2] key points is that not just street youth, but decent youth as well have an incentive to learn the tenets of the code of the street. Ignorance of the code may provoke a violent confrontation by staring too long at a street youth, stepping on someone’s toe, or failing to project a look of someone not to be messed with. The street frame is available on the streets to use instrumentally to incite collective action, maintain a sense of honor, and gain respect and status. For example, knowing the tenets of the frame, youth in search of a reputation seek to increase their status by “campaigning for respect,” by challenging, humiliating, or assaulting others, and disrespecting them by stealing their material possessions or girlfriends. When a member of a group is disrespected or assaulted, other members need only invoke the street frame, with its attendant rules, motives, and sanctions, to mobilize the group to exact payback. Short and Strodtbeck [63] found that in a situation of a gang fight, alternatives have narrowed to join the action or remain aloof – presumably through the street frame of a gang fight. Using subjective expected utility theory, Short and Strodtbeck [63:254] show that a gang leader calculates the probability and values of losing status by remaining aloof against that of being arrested for joining the action. We can add that gang leaders likely have low thresholds for joining the action; all else being equal, a non-leader may require a larger proportion of the gang to fight before joining. As in the case of collective action in general, the distribution of thresholds in the group may be important for collective acts of payback. Individual thresholds will be a function of the extent to which individuals are committed to the street frame, the degree to which their identities and self-esteem are tied to notions of manhood and nerve, and their role within disparate law-violating and law-abiding groups. Thus, most street youth will have low Differential social organization, collective action, and crime thresholds, whereas decent youth will have high thresholds. Interestingly, the equilibrium point of a group may be suboptimal for most group members [33]. Suppose that members of a “decent youth group” abhor violence and feel it is wrong, but are each unwilling to admit this to the group for fear of losing status, and being labeled a coward, woman, or “punk.” Each member believes that the others want to commit violence and exact payback, resulting in a “shared misunderstanding” to use Matza’s [50] terms. Their thresholds will be low, causing them to follow the street frame and engage in violence if a member has been disrespected. Thus, the violent collective outcome, resulting from the equilibrium point of the distribution of thresholds will be suboptimal for each individual, who would prefer to avoid violence [33]. Conclusion The importance of Sutherland’s abstract statement of differential social organization lies in its use of macro-level concepts of social organization to explain variations in crime rates and its acknowledgement that social organization can both foster and impede crime rates. Unfortunately, Sutherland did not explicate the precise mechanisms by which social organization affects crime, instead choosing to focus attention on the individual level explanation of differential association, and rely on examples from the literature to fill in the content of differential social organization. Focusing on the dynamic component of differential social organization, I have tried to identify theoretical mechanisms by which individuals mobilize others into collective behavior for crime and against crime, including closure in social networks, weak ties, collective action frames, and individual collective action thresholds. This initial foray into specifying some dynamic mechanisms of differential social organization raises additional questions and creates some new puzzles. First, I have treated organization in favor of crime separately from organization against crime, only touching on the relationship between them. In fact, the two are at times causally related, as Sutherland [73] recognized: increasing organization in favor of crime eventually elicits a strong response from conventional society, which organizes against crime. At other times, the two may interact in their effects on crime rates. Moreover, the two may be interwoven in very complicated ways that emerge through history. Several recent studies provide examples. In his study of public housing, Venkatesh [80] found that residents fought for a safe and secure neighborhood, battling with the housing authority, public service bureaucracies, and other city officials, and eventually turning to the very street gangs that dealt drugs in the neighborhood to help secure social order and provide services. In her study of a black middle class neighborhood juxtaposed to an impoverished high crime neighborhood, Patillo-McCoy [56] finds that strong neighborhood networks included drug selling gang members, who provided residents with protection, social control, and resources in exchange for latitude in enforcing their drug selling. Lyons [45] finds that rates of anti-black (but not anti-white) hate crimes are higher in neighborhoods with high collective efficacy, as organized neighborhoods seek to defend their territory against outsiders who pose a symbolic threat. Second, I have given short shrift to the important role of local institutions – families, schools, law enforcement, churches, voluntary associations – as well as broader social structures in organization against crime (as well as the rarer role in organization in favor of crime). Much has been written on such relationships, but the argument presented here is incomplete without specifying how broader structures and institutions exert social control R.L. Matsueda directly and also affect the very possibility of collective action. Clearly, the legal system looms in the background in every crime situation. How the legal system interacts with mechanisms of social organization is of paramount importance. Third, I have side-stepped the question of causes of individual criminal acts. Of course, all of the mechanisms I have identified have implications for the probability that criminal acts will occur for a given group in a given situation. But I have not tried to re-state Sutherland’s social psychological theory of differential association using the mechanisms identified, including social networks, stages of moral reasoning, collective action frames, identities and norms, and threshold models. Such a task would certainly be worthwhile, but is well beyond the scope of this paper. References 1. Akers, R. L. (1998). Social learning and social structure: A general theory of crime and deviance. Boston: Northeastern University Press. 2. Anderson, E. (1999). Code of the street: Decency, violence and the moral life of the inner city. New York: Norton. 3. Bandura, A. (1994). Self efficacy. In V. S. Ramachaudran (Ed.), Encyclopedia of human behavior, vol. 4 (pp. 71–81). New York: Academic. 4. Becker, H. S. (1953). Becoming a marijuana user. American Sociological Review, 59, 235–242. 5. Becker, H. S. (1963). Outsiders: Studies in the sociology of deviance. New York: Free. 6. Benford, R. D., & Snow, D. A. (2000). Framing processes and social movements: An overview and assessment. Annual Review of Sociology, 26, 611–639. 7. Berger, P., & Luckmann, T. (1967). The social construction of reality. Garden City, New York: Doubleday. 8. Bernstein, B. (1971). Class, codes and control, vol. 1. London: Routledge & Kegan Paul. 9. Blasi, A. (1980). Bridging moral cognition and moral action: A critical review of the literature. Psychological Bulletin, 88, 1–45. 10. Blau, P. (1974). Parameters of social structure. American Sociological Review, 39, 615–635. 11. Blumer, H. (1939). Collective behavior. In R. E. Park (Ed.), Principles of sociology. New York: Barnes & Noble. 12. Blumstein, A., & Cohen, J. (1987). Characterizing criminal careers. Science, 237, 985–991. 13. Burgess, R. L., & Akers R. L. (1966). A differential association-reinforcement theory of criminal behavior. Social Problems, 19, 101–113. 14. Carpendale, J. I. M. (2000). Kohlberg and piaget on stages of moral reasoning. Developmental Review, 20, 181–205. 15. Cloward, R., & Ohlin, L. E. (1960). Delinquency and opportunity: A theory of juvenile gangs. New York: Free. 16. Colby, A., & Kohlberg, L. (Eds.) (1987). The measurement of moral judgement, vol. 1–2. New York: Cambridge University Press. 17. Colby, A., Kohlberg, L., Gibbs, J., Lieberman, M., Fisher, K., & Saltzstein, H. D. (1983). A longitudinal study of moral judgement. Monographs of the Society for Research in Child Development, 48, 1–124. 18. Coleman, J. S. (1988). Social capital in the creation of human capital. American Journal of Sociology, 94, S95–S120. 19. Coleman, J. S. (1990). Foundations of social theory. Cambridge, Massachusetts: Harvard University Press. 20. Coser, R. (1975). The complexity of roles as seedbed of individual autonomy. In L. Coser (Ed.), The idea of social structure: Essays in honor of Robert Merton. New York: Harcourt Brace Jovanovich. 21. Cressey, D. R. (1953). Other people’s money. Glencoe, Illinois: Free. 22. Cressey, D. R. (1960). Epidemiology and individual conduct: A case from criminology. Pacific Sociological Review, 3, 47–58. 23. Dewey, J. (1958). Experience and nature. New York: Dover. 24. Emerson, M. O., Yancy, G., & Chai, K. J. (2001). Does race matter in residential segregation? Exploring the preferences of white Americans. American Sociological Review, 66, 922–935. Differential social organization, collective action, and crime 25. Fine, G. A., & Kleinman, S. (1979). Rethinking subculture: An interactionist analysis. American Journal of Sociology, 85, 1–20. 26. Fireman, B., & Gamson, W. A. (1979). Utilitarian logic in the resource mobilization perspective. In M. N. Zald & J. D. McCarthy (Eds.), The dynamics of social movements (pp. 8–44). Cambridge: Winthrop. 27. Galliher, J. F., & Walker A. (1977). The Puzzle of the Social Origins of the Marihuana Tax Act of 1937. Social Problems, 24, 367–376. 28. Gamson, W. A. (1990). Strategy of social protest (2nd ed.). Belmont, California: Wadsworth. 29. Garland, D. (1990). Punishment and modern society. Chicago: University of Chicago. 30. Gaylord, M. S. & Galliher J. F. (1988). The criminology of Edwin Sutherland. New Brunswick, NJ: Transaction Publishers. 31. Goffman, E. (1974). Frame analysis. New York: Harper Colophon. 32. Granovetter, M. (1973). The strength of weak ties. American Journal of Sociology, 78, 1360–1380. 33. Granovetter, M. (1978). Threshold models of collective behavior. American Journal of Sociology, 83, 1420–1443. 34. Granovetter, M. (1983). The strength of weak ties: A network theory revisited. Sociological Theory, 1, 201–233. 35. Grattet, R., Jenness, V., & Curry, T. (1998). The homogenization and differentiation of hate crime law in the United States, 1978–1995: Innovation and diffusion in the criminalization of bigotry. American Sociological Review, 63, 286–307. 36. Himmelstein, J. L. (1983). The strange career of marihuana: politics and ideology of drug control in America. Westport, CT: Greenwood Press. 37. Hirschi, T. (1969). Causes of delinquency. Berkeley: University of California Press. 38. Horowitz, R. (1983). Honor and the American dream. New Brunswick, New Jersey: Rutgers University Press. 39. Jenness, V., & Broad, K. (1997). Hate crimes. New York: Aldine de Gruyter. 40. Jenness, V., & Grattet, R. (2001). Making hate a crime: From social movement to law enforcement. New York: Russell Sage. 41. Joas, H. (1985). George Herbert Mead: A contemporary reexamination of his thought. Cambridge: Polity (translated by R. Meyer). 42. Killian, L., & Turner, R. (1972). Collective behavior. Englewood Cliffs, NJ: Prentice-Hall. 43. Kohlberg, L. (1963). The development of children’s orientation toward a moral order: I. Sequence in the development of moral thought. Vita Humana, 6, 11–33. 44. Kohlberg, L. (1981). The philosophy of moral development. San Francisco: Harper & Row. 45. Lyons, C. J. (2006). Social (dis)organization and racially motivated crime in Chicago. Unpublished Ph.D. dissertation, Department of Sociology, University of Washington. 46. Matsueda, R. L. (1982). Testing control theory and differential association: a casual modeling approach. American sociological review, 47, 489–504. 47. Matsueda, R. L. (1988). The current state of differential association theory. Crime and Delinquency, 34, 277–306. 48. Matsueda, R. L. (2006). Criminological implications of the thought of George Herbert Mead. In M. Deflem (Ed.), Sociological theory and criminological research: views from Europe and the United States. Oxford, UK: Elsevier Science. 49. Matsueda, R. L., & Heimer, K. (1997). A symbolic interactionist theory of role transitions, role, commitments, and delinquency. In T. Thornberry (Ed.), Advances in criminology theory: developmental theories of crime and delinquency,, vol. 7 (pp. 163–214). New Brunswick, N.J.: Transaction. 50. Matza, D. (1964). Delinquency and drift. New York: Wiley. 51. McCarthy, J. D. (1994). Activists, authorities, and media framing of drunk driving. In E. Larana, H. Johnston, & J. Gusfield (Eds.), New social movements: From ideology to identity (pp. 133–167). Philadelphia: Temple University Press. 52. Mead, G. H. (1934). Mind, self, and society. Chicago: University of Chicago Press. 53. Mead, G. H. (1964). Selected writings. Chicago: University of Chicago Press (edited by A. J. Reck). 54. Miller, D. L. (1973). George Herbert Mead: Self, language, and the world. Austin, Texas: University of Texas. 55. Morenoff, J., Sampson, R. J., & Raudenbush, S. (2001). Neighborhood inequality, collective efficacy, and the spatial dynamics of urban violence. Criminology, 39, 517–560. 56. Pattillo-McCoy, M. (1999). Black picket fences. Chicago: University of Chicago Press. 57. Sampson, R. J. (2004). Neighborhood and community: Collective efficacy and community safety. New Economy, 11, 106–113. 58. Sampson, R. J., Morenoff, J., & Earls, F. (1999). Beyond social capital: Spatial dynamics of collective efficacy for children. American Sociological Review, 64, 633–660. R.L. Matsueda 59. Sampson, R. J., & Raudenbush, S. W. (1999). Systematic social observation of public spaces. American Journal of Sociology, 105, 603–651. 60. Schelling, T. (1971). Dynamic models of segregation. Journal of Mathematical Sociology, 1, 143–186. 61. Shaw, C. R., & McKay, H. D. (1969). Juvenile delinquency and urban areas. Revised edition. Chicago: University of Chicago Press. 62. Shibutani, T. (1955). Reference groups as perspectives. American Journal of Sociology, 60, 562–569. 63. Short, J. F. Jr., & Strodtbeck, F. L. (1965). Group process and gang delinquency. Chicago: University of Chicago. 64. Simon, H. A. (1957). Models of man: social and rational. New York: Wiley. 65. Small, M. L. (2002). Culture, cohorts, and social organization theory: Understanding local participation in a latino housing project. American Journal of Sociology, 108, 1–54. 66. Snow, D. A., & Benford, R. D. (1992). Master frames and cycles of protest. In A. D. Morris & C. M. Mueller (Eds.), Frontiers in social movement theory. New Haven, Connecticut: Yale University Press. 67. Snow, D. A., & Oliver, P. E. (1995). Social movements and collective behavior: Social psychological dimensions and considerations. In K. S. Cook, G. A., Fine & J. S. House (Eds.), Sociological perspectives on social psychology (pp. 571–599). 68. Snow, D. A., Rochford, E. B. Jr., Worden, S. K., & Benford, R. D. (1986). Frame alignment process, micromobilization, and movement participation. American Sociological Review, 51, 464–481. 69. Sutherland, E H., (1926). The biological and sociological processes. Papers and Proceedings of the Twentieth Annual Meeting of the American Sociological Society, 20, 58–65. 70. Sutherland, E. H. (1939). Principles of criminology (3rd ed.). Philadelphia: Lippincott. 71. Sutherland, E. H. (1947). Principles of criminology (4th ed.). Philadelphia: Lippincott. 72. Sutherland, E. H. (1973 [1942]). Development of the theory. In K. Schuessler (Ed.), Edwin H. Sutherland on analyzing crime (pp. 13–29). Chicago: University of Chicago Press. 73. Sutherland, E. H. (1973 [1943]). Wartime crime. In K. Schuessler (Ed.), Edwin H. Sutherland on analyzing crime (pp. 120–128). Chicago: University of Chicago Press. 74. Sweetser, F. L., Jr. (1942). A new emphasis for neighborhood research. American Sociological Review, 7, 525–533. 75. Sykes, G. M., & Matza, D. (1957). Techniques of neutralization: A theory of delinquency. American Sociological Review, 22, 664–670. 76. Tannenbaum, F. (1938). Crime and the community. Boston: Ginn and Co. 77. Thomas, W. I., & Znaniecki, F. (1958 [1918]). The polish peasant in Europe and America, vols.1 and 2. New York: Dover. 78. Thrasher, F. M. (1927). The gang. Chicago: University of Chicago Press. 79. Turner, R., & Killian, L. (1957). Collective behavior. Englewood Cliffs, New Jersey: Prentice-Hall. 80. Venkatesh, S. (2000). American project: The rise and fall of a Modern Ghetto. Cambridge, Massachusetts: Harvard University Press. Differential social organization, collective action, and crime Differential social organization, collective action, and crime Abstract Differential social organization, collective action, and crime Differential association theory Normative conflict: the root cause of crime in society Differential association process: explanation of individual criminal acts Differential social organization: explanation of group rates of crime Insights into Sutherland’s concept of differential social organization Mead’s theory of symbolic interaction and social control Taking the role of the other The social structure of the self and stages of development of the self Stages of moral reasoning and linguistic codes Kohlberg’s stages of moral reasoning Moral reasoning and social networks Network structure, collective action frames, and thresholds Social networks and social capital The strength of weak ties Collective action, frame analysis, and individual thresholds Social organization against crime Social movements against crime Collective action against crime: informal social control and collective efficacy Social organization in favor of crime Conclusion References << /ASCII85EncodePages false /AllowTransparency false /AutoPositionEPSFiles true /AutoRotatePages /None /Binding /Left /CalGrayProfile (Gray Gamma 2.2) /CalRGBProfile (sRGB IEC61966-2.1) /CalCMYKProfile (ISO Coated) /sRGBProfile (sRGB IEC61966-2.1) /CannotEmbedFontPolicy /Error /CompatibilityLevel 1.3 /CompressObjects /Off /CompressPages true /ConvertImagesToIndexed true /PassThroughJPEGImages true /CreateJDFFile false /CreateJobTicket false /DefaultRenderingIntent /Perceptual /DetectBlends true /DetectCurves 0.1000 /ColorConversionStrategy /sRGB /DoThumbnails true /EmbedAllFonts true /EmbedOpenType false /ParseICCProfilesInComments true /EmbedJobOptions true /DSCReportingLevel 0 /EmitDSCWarnings false /EndPage -1 /ImageMemory 1048576 /LockDistillerParams true /MaxSubsetPct 100 /Optimize true /OPM 1 /ParseDSCComments true /ParseDSCCommentsForDocInfo true /PreserveCopyPage true /PreserveDICMYKValues true /PreserveEPSInfo true /PreserveFlatness true /PreserveHalftoneInfo false /PreserveOPIComments false /PreserveOverprintSettings true /StartPage 1 /SubsetFonts false /TransferFunctionInfo /Apply /UCRandBGInfo /Preserve /UsePrologue false /ColorSettingsFile () /AlwaysEmbed [ true ] /NeverEmbed [ true ] /AntiAliasColorImages false /CropColorImages true /ColorImageMinResolution 150 /ColorImageMinResolutionPolicy /Warning /DownsampleColorImages true /ColorImageDownsampleType /Bicubic /ColorImageResolution 150 /ColorImageDepth -1 /ColorImageMinDownsampleDepth 1 /ColorImageDownsampleThreshold 1.50000 /EncodeColorImages true /ColorImageFilter /DCTEncode /AutoFilterColorImages true /ColorImageAutoFilterStrategy /JPEG /ColorACSImageDict << /QFactor 0.76 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> /ColorImageDict << /QFactor 0.76 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> /JPEG2000ColorACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 15 >> /JPEG2000ColorImageDict << /TileWidth 256 /TileHeight 256 /Quality 15 >> /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 150 /GrayImageMinResolutionPolicy /Warning /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 150 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict << /QFactor 0.76 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> /GrayImageDict << /QFactor 0.76 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> /JPEG2000GrayACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 15 >> /JPEG2000GrayImageDict << /TileWidth 256 /TileHeight 256 /Quality 15 >> /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 600 /MonoImageMinResolutionPolicy /Warning /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 600 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict << /K -1 >> /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False /Description << /CHS /CHT /DAN /DEU /ESP /FRA /ITA (Utilizzare queste impostazioni per creare documenti Adobe PDF adatti per visualizzare e stampare documenti aziendali in modo affidabile. I documenti PDF creati possono essere aperti con Acrobat e Adobe Reader 5.0 e versioni successive.) /JPN /KOR /NLD (Gebruik deze instellingen om Adobe PDF-documenten te maken waarmee zakelijke documenten betrouwbaar kunnen worden weergegeven en afgedrukt. De gemaakte PDF-documenten kunnen worden geopend met Acrobat en Adobe Reader 5.0 en hoger.) /NOR /PTB /SUO /SVE /ENU (Use these settings to create Adobe PDF documents for journal articles and eBooks for online presentation. Created PDF documents can be opened with Acrobat and Adobe Reader 5.0 and later.) >> >> setdistillerparams << /HWResolution [2400 2400] /PageSize [595.276 841.890] >> setpagedevice