Integrating subjectively-derived choice sets to expand offender decision-making

ABSTRACT

Rational choice and offender decision-making are premised upon each individual’s weighing of their subjectively perceived behavioral options. However, most applications have failed to account for the heterogeneity in the options individuals perceive to have available to them within their choice set. We leveraged interdisciplinary scholarship from the fields of neuroscience and psychology to develop a strategy to capture the choice set of perceived options. Using a community sample of adults randomly assigned to one of two vignettes, we asked respondents to list the options they perceived to have available to resolve the given vignette. We then classified those options into one of four categories: pure conformist, abstain, pure criminal, or hybrid wherein an option was simultaneously criminal and prosocial. We found extensive heterogeneity in both the options and choice sets arising from each vignette. Most individuals did not note any criminal options while hybrid options appeared to a non-trivial degree in only one of the two vignettes. Our findings highlight the pitfalls associated with providing respondents with options as compared to having them subjectively construct their own options, while also pointing to how choice sets can provide a more descriptively accurate assessment of decision-making processes.

KEYWORDS:

• Rational choice theory
• offender decision-making
• choice set

Acknowledgments

We would like to thank the anonymous reviewers for their thoughtful insight that strengthened the contributions of this paper. We also thank David Hureau, Tom Loughran, Justin Pickett, and Greg Pogarsky who each saw promise in this project and provided thoughtful, critical, and constructive feedback on earlier drafts of this manuscript. We owe Samantha Penta a debt of gratitude for her valuable insight and assistance in helping us organize the qualitative nature of our data. This work was presented at the Annual Meeting of the American Society of Criminology (ASC) in San Francisco in 2019. We thank the ASC attendees who offered us guidance for advancing this project.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. While not at the stage of the choice set of options, Bouffard (2002a, 2002b) similarly suggests that by providing respondents with consequences to offending scholars’ research designs artificially provided respondents with consequences that they did not consider and bypassed the consequences that they did consider.

2. The measure available to capture the choice set was a Likert-type scale questioning respondents if they agreed or disagreed with obtaining facts and considering alternative behaviors when making a decision (Paternoster and Pogarsky 2009; Paternoster, Pogarsky, and Zimmerman 2011).

3. Wikström (2006) developed a ‘Situational Action Theory of Crime Causation’ where one’s ‘moral filter’ causes one to abstain from crime (Wikström and Svensson 2010). He proposed ‘action alternatives’ to argue how morality eliminates deviant alternatives from consideration. Our study is distinct from Wikström because we provide a novel measure of the choice set (‘action alternatives’) rather than assume it. Future studies should examine the influence morality has on the consideration of certain options with the measure our study provides.

4. The sample sizes for individual demographic measures in Table 1 vary from the analytic sample size of 406 due to item non-response and ‘breaking off.’ These were anticipated, which is why the primary questions of interest were placed at the beginning of the survey per advice of Dillman, Smyth, and Christian (2014).

5. While taking a dropped wallet and/or its contents when one knows it was lost or mislaid may not be illegal in some states, in the state where this data was collected the act is illegal and is legally defined as larceny.

6. Mutz, Pemantle, and Pham (2019) sample balance test of randomization suggests that the samples are not different from one another – or, in other words, randomization was successful. Thus, any differences in choice sets across vignettes are a product of the scenario and are not driven by differences between individuals randomly assigned to these vignettes.

7. Leder et al. (2018, 2586) note that they ‘support the hypothesis put forward by Kalis, Kaiser, and Mojzisch (2013) that option generation is a distinct construct, being different from long-term memory and the ability to come up with original ideas as one facet of divergent thinking.’

8. A helpful reviewer raised the issue of whether an Uber ride constituted a unique option as compared to catching a taxi. The reviewer noted that such options likely only reflect a difference in technological sophistication. We have retained these two options as separate options because of the mechanism identified by the reviewer and elaborate to note these two options reflect different processes on the part of the respondent that are liable to be correlated with important age-graded factors, socio-economic status, and gender. While well beyond the scope of the current manuscript, we are intrigued at the prospect of further delving into the choice set with a new lens toward technological- and age-graded differences in choice sets, and how those differences can be linked with desistance. Nevertheless, for the purposes of the analyses conducted in this paper, our primary takeaways are not sensitive to this classification decision.

9. In the online supporting information, we provide a similar breakdown of the choice sets for 20 randomly chosen respondents. Those employing the choice set in their own research may benefit from this additional coding breakdown.

Additional information

Notes on contributors

Matthew C. Kijowski

Matthew C. Kijowski is a PhD candidate in the School of Criminal Justice at the University at Albany, SUNY. His research interest includes offender decision-making, control theories of crime, and quantitative methods.

Theodore Wilson

Theodore Wilson is an assistant professor in the School of Criminal Justice at the University at Albany, SUNY. His research interests include offender decision-making, courtroom actor decision-making, and quantitative methods.