The resulting findings of Jensen’s ” Parents, Peers, and Delinquent Action: A test of the Differential Association Perspective” study (1972) assesses the arguments “using differential association theory to encompass certain “known” relationships, and consistent with previous research the number of delinquent friends, the perception of “trouble” in the neighborhood, and the variable acceptance of attitudes and beliefs favorable to the violation of legal codes were significantly related to involvement in delinquent action”. However, Sutherland and Cressey’s argument was failed to be supported by the findings of this study, which “cast doubt on their arguments concerning the home and family in relation to delinquent behavior”, but noted that what goes on in the family situation seems to have significance of its own which is not included by Sutherland and Cressey’s differential association perspective (Jensen, 1972). Still, Jensen study did reveal that some interaction existed when definitions favorable to the violation of the law were controlled. Finally, Jensen’s study (1972) states that “the theory of differential association stresses “definitions” and “cultural” variables (values, norms, and beliefs) to such an extent that processes shaping human behavior other than internalization of normative standards tend to be slighted” (Jensen, 1972). However, contrary to Jensen’s analyses, Matsueda’s interpretation upon a reexamination of the findings of the Richmond Youth study provides a different conclusion from those results.
“Testing Control Theory and Differential Association: A Causal Modeling Approach”, by Ross L. Matsueda at the University of California, Santa Barbara, is another second which investigates the validity of Sutherland’s differential association theory. Matsueda’s study considers the findings of both the original Richmond Youth study and the findings of Jensen’s 1972 study. The major difference between Sutherland’s differential association theory and the control theory is that Sutherland stresses that definitions of the legal code mediate the effects of structural factors on crime. On the other hand, the control theory states that definitions of the legal codes reflect the degree of belief in the moral order and therefore do not mediate attachment, commitment, or involvement. One of the most detrimental criticisms about the differential association theory is that the primary variable of the theory, that a person’s ratio learned behavior patterns, cannot be precisely determined, therefore the theory of differential association cannot be subjected to empirical verification. However, Matsueda (1982) theorizes that “by explicitly modeling the measurement error structure of the ratio of learned behavior patterns favorable and unfavorable to the violation of legal codes, the critical variable of Sutherland’s theory can be operationalized”.
The methodology or design behind Matsueda’s ideas follows from the ideas of both Sutherland and Cressey. Although it is argued that this variable cannot be operationalized, Matsueda hypothesis contradicts this argument. Matsueda’s (1982) interpretation hypothesizes that “when the ratio of behavior patterns favorable and unfavorable to delinquency are treated as a “latent” or “unobservable” variable, it does have operational implications for the relationships among variables that can be observed, which in turn can be specified as indicators of the underlying theoretical construct”. Therefore, Matsueda states that Sutherland’s theory of differential association can and should be weighed empirically. “The oft-cited difficulty of operationalizing the ratio of learned behavior patterns can be conceptualized as a problem of measurement error. The most practical way to address this problem is by explicitly modeling the measurement error structure of items tapping the underlying definitions construct, therefore, allowing the testing of specific hypotheses derived from the differential association theory” (Matsueda, 1982).
The resulting findings of Matsueda’s study (1982) supported and favored Sutherland’s differential association theory, but could only be used to predict a general index of delinquent behaviors for general indicators of the definitions of legal code. Concluding, Matsueda (1982) emphasizes that “Sutherland’s differential association remains as an important principle that focuses attention on the fact that all behavior is oriented by norms, and it integrates and explains variations in crime rates. Yet, not all of the studies explore this aspect of Sutherland’s differential association theory.
“Doing What Simple Simon Says?: Estimating the Underlying Causal Structures of Delinquent Associations, Attitudes, and Serious Theft”, a study by Mark D. Reed and Dina R. Rose published in June 1998, investigates one of the more fundamental and stronger points of Sutherland’s differential association theory. Their study attempts to investigate the “relative contributions of socialization, group pressure, social selection and rationalization processes in explaining the relationships between peer-group associations, attitudes and serious theft” (Reed & Rose, 1998). By estimating a nonrecursive model, Reed and Rose (1998) argue that (delinquency) serious theft is as likely to affect delinquent associations and attitudes as associations and attitudes are to affect (delinquency) serious theft. Reed and Rose’s study (1998) attempts to improve upon earlier studies examining differential association and social learning theories in two distinct ways. “First, this research uses panel data that allows us to model the reciprocal relations of delinquent associations, delinquent attitudes, and serious theft, thereby enabling a test of the relative contribution of several different social processes on serious theft”. Second, presently the majority of cross-sectional and longitudinal studies lack the inclusion of a measurement model in the substantive model, thus, rendering the increased feasibility that the structural parameter calculations are biased due to measurement error. The research of this study examined by Reed and Rose (1998) estimates a measurement model, correcting for the potential biasing effects of correlated measurement error.
Reed and Rose’s methodology and design for this study (1998) came from the longitudinal data, collected from the NYS (National Youth Survey), which are grounded on a probability sample of households using a multistage, cluster sampling design. “Five annual waves of data for 1976 to 1980 are available; the present analysis examines the first three waves. Of the 2,360 youths (ages ranged from 11 to 17) asked to participate in the original study, 1,725 (73%) agreed. Respondent loss over the first three waves was 6%, leaving 1,626 participants in the study, which had no effect on the studies results” ( Reed and Rose, 1998).
The results of Reed and Rose’s study (1998) showed “that the socialization effects are negligible and insignificant in the reciprocal estimations: Peers influence attitudes, but attitudes are not the primary determinants of behavior. More important, however, the research established that the “group pressure effects are substantially smaller in the nonrecursive models compared to those commonly reported in recursive studies”(Reed and Rose, 1998). Yet, perhaps the most fundamental findings of the research from Reed and Rose’s study (1998) “indicates that there are several social processes operating simultaneously. Thus, the adherence to a single social process will likely result in the oversimplification of the causal mechanisms underlying delinquent behavior”.
While Sutherland’s differential association theory has been examined to explain nearly all types of crime, overall, these studies demonstrate the broad complexity of Sutherland’s theory. But to what extent if any is the validity of Sutherland’s theory of differential association given as a result of these studies? The results of the first study, “Parents, Peers, and Delinquent Action: A test of the Differential Association Perspective”, by Gary F. Jensen failed to find validity in Sutherland’s theory. Particularly, the data of Jensen’s study disagreed with the arguments of Sutherland and Cressey concerning the home and family in relation to delinquent behavior. However, the second study, “Testing Control Theory and Differential Association: A Causal Modeling Approach”, conducted by Ross L. Matsueda found that Sutherland’s differential association theory held a great validity than the social control theories. Importantly, Matsueda had noted that most commonly criticized problem of measurement error could be corrected by explicitly modeling the measurement error structure of items tapping the underlying definitions construct, which then gives greater validity to Sutherland’s theory over that of social conflict theories. The third and final study discussed was, “Doing What Simple Simon Says?: Estimating the Underlying Causal Structures of Delinquent Associations, Attitudes, and Serious Theft”, by Mark D. Reed and Dina R. Rose. Although, Reed and Rose’s examination of Sutherland’s differential association theory are a little more indirect than that of the other two studies, it still explores the validity of Sutherland’s theory. Reed and Rose’s study does give favor to Sutherland’s theory, they also incorporated the ideas of other theories, with would tend to make a better case that Sutherland’s theory does not hold true for all types of crimes and criminal behavior.
Throughout his career Sutherland attempted to clarify his ideas in efforts to postulate a concise general theory of crime and criminal behavior, however he never managed to do so. “Despite his many critics, little doubt exists that Edwin Sutherland was a true “pioneer” in the field of criminology, not in spite of his differential association theory but because of it” (Jones, 1986). The universal acceptance of Sutherland’s differential association theory due to its less complex and more coherent approach to crime causation, in addition to the reality of it being a definition of individual criminal behavior. Perhaps in part because of Sutherland’s differential association theory is supported by much evidence, a large quantity of modern theory and research in the field of criminology can be attributed to the work of Sutherland’s original composition. Although, Sutherland has often been misinterpreted for his ideas on the associations with criminals leading to criminal behavior, what he meant was that the contents of patterns in association would differ from individual to individual. “Differential association is a theory based on the social environment and its surrounding individuals and the values those individuals gain from significant others in their social environment” (DeMelo, 1994). Crime was viewed by Edwin Sutherland as a consequence of conflicting values. Moreover, the basic premises behind the Sutherland’s combined and final differential association theory is that criminal behavior is learned and that in a “situation of differential organization and normative conflict, differences in behavior, including criminal behaviors, arise because of differential associations (Vold, 1998).
Overall, Edwin Sutherland’s differential association theory does not best explain any one specific type of crime or criminal behavior. In fact, perhaps the reason why Sutherland’s theory has sustained its significance is because it encompasses a general formula and the fundamental ideas or way of examining crime and criminal behavior. Conceivably, a better question and challenge that might be raised instead of what types of crime does it best explain, would be what types of crime or criminal behavior could you not apply to Sutherland’s differential association theory. Therefore there is no suggestion that Sutherland’s theory is not valid, merely that the application of Sutherland’s theory must follow in accordance with Sutherland’s own calculations and
370