Testing Deterrence Theory with Offenders: The Empirical Validity of Stafford and Warr's Model

Abstract

Stafford and Warr (1993) reconceptualized general and specific deterrence into a single theory in which individuals' propensities to engage in crime are based on a combination of personal and vicarious experiences with being punished and avoiding punishment. The current study extends prior tests of this conceptualization of deterrence by expanding the definition of vicarious experience, analyzing extralegal as well as legal consequences, examining multiple types of offending behavior, and drawing its data from a sample of work release facility inmates. The results fail to support legal deterrence as an explanation of offending for this sample but suggest the importance of extralegal consequences.

Notes

¹The possible penalties for each crime were reviewed by a practicing criminal defense attorney. The attorney agreed that in the State of Florida the punishments were realistic for each crime, bolstering the realism of the scenarios.

²Weekly income and the number of times in the past five years that respondents had committed various offenses (driven drunk; purchased illegal drugs; taken something from a store without paying for it; hurt or threatened to hurt someone with a weapon; acted loud, rowdy, or unruly in a public place; driven a car without the owner's permission; broken into a house or building; vandalized someone's property; driven without a license; or started a fistfight) contained extreme outliers (e.g., having driven drunk 1,000,000 times in the past five years). Because these values had marked effects on our initial empirical analyses, each outlier was recoded to the 95th percentile.

†p ≤ .10, *p ≤ .05, **p ≤ .01.

³Because the dependent variables were positively skewed, this study also examined the data using tobit regressions (Breen 1996). None of the variables were affected.

⁴An anonymous reviewer also noted that our data are not longitudinal. Although longitudinal data would provide an alternative approach, we are not convinced that our cross-sectional data are necessarily inferior. As Grasmick and Bursik (1990) have pointed out, longitudinal deterrence studies risk misestimation of deterrent effects due to changes in respondents' perceptions of sanction threat over time. Our approach permits us to examine the effects of past experiences and present perceptions of punishment certainty on respondents' estimates of whether they will commit an offense in the future.

Additional information

Notes on contributors

Alicia H. Sitren

ALICIA H. SITREN is an Assistant Professor at the University of North Florida in the Department of Criminology and Criminal Justice. Her published work has been devoted to punishment philosophies and revisions of deterrence theory. Her research agenda includes studies in corrections and white collar crime as well as continuing to examine central elements of deterrence theory using data drawn from convicted offenders.

Brandon K. Applegate

BRANDON K. APPLEGATE is an Associate Professor and Chair the University of South Carolina in the Department of Criminology and Criminal Justice. He has published more than forty articles on punishment and rehabilitation policy, correctional treatment, juvenile justice, public views of correctional policies, jail issues, drunk driving, and decision making among criminal justice professionals. He is also co-author of Offender Rehabilitation: Effective Correctional Intervention.