Date Available


Year of Publication


Degree Name

Doctor of Philosophy (PhD)

Document Type

Doctoral Dissertation


Arts and Sciences



First Advisor

Dr. Carrie B. Oser


Agnew’s (1992; 2006) general strain theory (GST) has become one of the foremost theories to explain crime in contemporary criminology. While it has undergone several empirical tests over the years, there remain many understudied aspects of the theory. The current study addresses some of these aspects by longitudinally exploring the relationship between multiple types of strain and drug and non-drug crime among a sample of African American women.

Data for this study were collected as part of a larger study on how drug use and criminality are related to health disparities, particularly HIV, and service utilization among African American drug-using and non-drug-using women across justice system status—prison, probation, and community, no supervision. The overall sample comprised of 462 women who completed structured interviews at four time intervals with a response rate of 87 percent.

The study was guided by three specific research aims. First, this study examined whether various types of strain―economic hardship, criminal victimization, and gendered racism―were conducive to different types of crime. Second, this study examined whether certain negative emotional states―anger, depression, and anxiety―mediated the effects on the strain-crime relationships among the women. And third, this study examined whether certain factors―social support, coping skills, and spiritual well-being―moderated the strain-crime relationships among the women.

Findings revealed all three forms of strain had statistically significant effects on involvement in crimes unrelated to drug use, such as check fraud or burglary, among the women in the sample, while only economic hardship and victimization had significant effects on drug crime. In addition, partial mediation was found between economic hardship, anger, and non-drug crime and complete mediation occurred between gendered racism, anger, and non-drug crime. That is, women in the sample became angry after experiencing these types of strain and responded to that feeling by engaging in crimes unrelated to drugs. Lastly, moderation was only found in the logistic regression model 3 examining gendered racism, social support, and non-drug crime. In other words, women with perceived high social support were less likely to commit non-drug crimes than those with low social support except when their gendered racism experiences became extremely high. This study’s findings will make significant contributions to the scholarship across multiple disciplines, as well as potentially inform practice and policy. Drawbacks and directions for future research are discussed.

Digital Object Identifier (DOI)