Agency and outcome expectancies for crime desistance: measuring offenders' personal beliefs about change†

Abstract

While the majority of offenders eventually desist from crime, the internal psychological mechanisms hypothesized to drive the process of desistance and offender change have not been systematically measured. This study developed scales for assessing intention to change, or offenders' beliefs regarding their perceived ability to stay crime-free (agency) and expected outcomes for crime and desistance (expectancies). Incarcerated offenders (N=142) endorsed these beliefs in a way that is consistent with theories of offender change. The structure of beliefs suggests offenders with positive expectancies for desistance and negative expectancies for crime also endorse a higher sense of personal agency to desist. Outcome expectancies for desistance were unrelated to static risk variables, suggesting these measures may be complementary to risk assessment. Overall, the scales developed for this research showed high internal consistency and evidence for concurrent and construct validity. Refining the measurement methods and assessing recidivism outcome post-release should further advance this avenue of research.

Keywords:

• crime desistance
• offender change
• agency
• outcome expectancies
• antisocial attitudes

Acknowledgement

Caleb D. Lloyd acknowledges financial awards from the Social Sciences and Humanities Research Council of Canada and the Ontario Graduate Scholarship during the course of this research. We are also grateful to two anonymous reviewers for constructive and helpful direction on an earlier draft of this paper.

Notes

1. Offender treatment targeting cognition through cognitive–behavioural techniques is effective for reducing recidivism among treatment completers (e.g. McGuire et al., 2008), but there is scant research effectively demonstrating that changes in cognitive variables drive the observed reduction in risk.

2. However, in addition to building strengths, crime desistance requires utilizing self-control in situations where criminal actions would achieve immediate goals. Indeed, individuals do not always behave with rational regard for the consequences of their behaviour, especially long-term consequences (Brezina, 2002; Burnett & Maruna, 2004).

3. We invite all readers interested in viewing the items developed for this research to contact the first author for copies of the questionnaires.

4. All analyses were conducted after transforming the variables to correct for skew. However, results did not appreciably differ when compared with analyses using untransformed variables; thus, the results reported in this paper employ the untransformed variables.

5. Moulden, Marshall, and Marshall (2005) found that a preparatory rehabilitation program for sex offenders significantly increased their sense of agency, as measured by both the Hope and State Hope scales. This study provides normative data on the scales for a specific offender population. Participants scored an average of 12.2 out of 16 points on the pathways subscale of the Hope Scale and 9.5 out of 16 points on the agency subscale (Moulden et al., 2005). On the State Hope Scale, participants scored an average of 30.0 points pre-treatment and 34.5 points post-treatment (with possible scores ranging 8–64; Moulden et al., 2005). In this study, average scores on the State Hope Scale were higher compared to the sex offender sample (M=39.6, SD = 7.5). Average scores on the Hope Scale in this sample were similar, but higher on the pathways subscale (M=13.1, SD = 2.2) and especially higher on the agency subscale (M=12.4, SD = 2.4).

6. Testing whether these items form a single factor using principal components analysis would require a larger sample size than is available in this study.

7. We present simple bivariate correlations in this section due to low statistical power for analyses using more complex statistical models.