Racial Discrimination and Hirschi's Criminological Classic: A Chapter in the Sociology of Knowledge

Abstract

In Causes of Delinquency, Travis Hirschi attempted to falsify the strain theory claim that racial discrimination might contribute to the delinquency of African American youths. A reanalysis of the Richmond Youth Project data used in his classic study, however, reveals that perceived racial discrimination is a robust predictor of delinquent involvement. This finding suggests that Hirschi missed a historic opportunity to focus the attention of a generation of criminologists on how the unique experiences of African Americans may shape their criminality. Given the salience of perceived racial bias in the lives of many African Americans, the subsequent neglect by scholars of discrimination as a potential source of crime is a remarkable omission—so much so that it constitutes a significant and as yet untold chapter in the sociology of knowledge.

Keywords:

• perceived racial discrimination
• social bond theory
• sociology of knowledge
• racial discrimination
• control theory

Notes

1. An analysis (using the EM option within SPSS) of the missing data showed that Little's (1988) missing completely at random test was statistically significant (p = .000). Consequently, we employed two strategies to investigate whether the statistically significant relationship between perceived racial discrimination and delinquency reported in Model 1 of Table 1 results from biases introduced by the listwise deletion of the missing cases. First, using the EM algorithm in LISREL, we generated a data set with imputed values. All of the variables in the analysis including those used to create the delinquency index and those used to create the independent variable scales (e.g., Achievement, Adult Activities) were used to impute the missing values. All totaled, 48 variables were used to generate the imputed values for the missing data. This data set included 1440 African American youths. As reported above, an analysis of this data set showed that perceived racial discrimination significantly predicted delinquent involvement. Second, using the MI and MIANALYZE procedures within SAS, we generated a combined set of regression coefficients for Perceived Discrimination while controlling for the other covariates (based on five data sets with imputed values). The combined estimate of the Perceived Discrimination coefficient was .03, its standard error was .007, and the 95% confidence interval was .018–.045. Also, the p‐value (.000) indicates that the hypothesis that the Perceived Discrimination parameter is equal to 0 can be rejected. Together, these results indicate that perceived discrimination is a significant predictor of delinquency among the African American youth included in the Richmond Youth Project data.

2. Following previous research on GST, we also explored whether the relationship between perceived racial discrimination and delinquency would be mediated by measures of anger, depression, and racial identity. These measures did not substantively reduce the relationship reported in Table 1. The results are available upon request.

3. We caution that the finding that African American offending is related to attending a school that they perceive to be racially antagonistic needs further exploration and elaboration. We were able to replicate this finding using the data set with imputed values for the missing data (we used LISREL's EM algorithm to generate this data set) (Beta = .05, p = .03, n = 1440). (The results are available upon request.) However, the regression equation that generated this finding only included the covariates that were statistically significant and it did not include the measure of whether African American youths believed that they were treated badly because of their race. The bivariate correlation between delinquency and the Perceived School Discrimination scale, using the data set with imputed values for the missing data, was .18 (p = .000, n = 1440) and .18 (p = .000, n = 1205) with missing data deleted. Further elaborations of this finding may wish to examine whether racially hostile schools are criminogenic for African American youth because it weakens their educational commitment and involvement and/or whether it heightens their likelihood of being personally denigrated because of their race within the school setting.