Intentions to Steal: How Equal is the Interplay Between Anticipated Moral Emotions and Self-Serving Justifications Across Male and Female Young People?

ABSTRACT

The present study is concerned with the interplay between anticipated moral emotions and self-serving justifications in the explanation of intentions to steal. Theoretically, it has been suggested that both moral dimensions may interact in such a way that their effects amplify each other. This amplification hypothesis has rarely been tested. The following research question is addressed: To what extent do anticipated moral emotions amplify the impact of self-serving justifications on intentions to steal? To assess the stability of the findings, the hypothesis is tested across males and females. Online survey data are obtained from a large convenience sample of young people (N = 3584). A visual vignette was used to elicit participants’ offense-specific responses to the key concepts. Evidence suggests that self-serving justifications in the specific context of stealing and anticipated shame-guilt have strong main effects on intentions to steal. A statistically significant interaction effect is found in the full sample and the female subsample, suggesting that participants reporting weak anticipated shame and guilt are most likely to form intentions to steal in the presence of self-serving justifications. However, the effect is small and not found in the male subsample. We discuss the limitations and implications for further research.

KEYWORDS:

• Intentions to steal
• anticipated shame
• anticipated guilt
• self-serving justifications
• moral norms
• interaction
• amplification
• visual scenario

Acknowledgement

The authors are very grateful to Prof. Dr. Jean-Louis van Gelder, director of the Department of Criminology at the Max Planck Institute for the Study of Crime, Security, and Law (Freiburg, Germany), for making the visual scenario available.

Disclosure statement

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

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15564886.2023.2228775

Notes

1. Throughout the text, we will consistently use the term self-serving justifications to conceptualize highly specific outcome-related personal moral norms. Moral norms are situational rules of conduct, expectations guiding what is right or wrong, and statements about what is permitted and what is not (Alexander, 1987; Opp, 2001; Turner & Stets, 2006). The concept of self-serving justifications aligns with the concept of state or situational morality as opposed to trait or dispositional morality (Van Gelder & de Vries, 2014).

2. It should be noted that studies measuring criminal intent were not included.

3. Although empirical knowledge that improves predictive accuracy for criminal conduct should not be devalued, the predictor variables of criminal conduct and their ability to influence the occurrence of criminal conduct are two distinct issues (Andrews & Bonta, 2010).

4. Evolutionary approaches to culture postulate that natural selection has favored various learning strategies in individuals, both human and non-human primates (Boyd et al., 2011; Richerson & Boyd, 2008), and provide predictions about when individuals should rely on asocial (individual learning) experience or when they should rely on social learning strategies, such as the tendency to conform to values, beliefs, practices from other members of the social group (Muthukrishna et al., 2016).

5. The term young people is used by the World Health Organization (WHO) to combine adolescents and youths. The WHO defines adolescents as those people between 10 and 19 years of age and youth as the 15–24 year age group. Although our sample combines the 12 to 25 year age group, we will also use the term young people throughout the text.

6. Young people were chosen as the target population mainly for two reasons: (1) the transition from child to adulthood is a critical period of social, psychological, and biological changes in human development. For young people, this transition involves demanding emotional challenges and important choices during which they take on new roles and responsibilities with consequences for the moral and emotional development (Lloyd, 2005). Young people have more opportunities to become moral agents in their social roles (e.g., spending time with other socializing agents (peers), smoking, drinking) (Smetana & Turiel, 2008). This may entail taking moral decisions with consequences for self and others. (2) A very consistent finding across studies in offending is the relationship between age and crime from late childhood (10-12y) till early adulthood (early 20s) (the age-crime curve) (for a review see Farrington et al., 2019). As such participants in this life span were chosen to achieve as much variation in the outcome variable as possible.

7. Although, Dillman (1978; Dillman et al., 2014) originally recommended placing the demographics at the end, rather than at the beginning of the survey instrument, the question of the placement of demographics (at the end or the beginning) is an empirical one. Green et al. (2000) tested this question in an experimental study. No statistically significant differences were found in the amount of missing data when the response characteristics of participants assigned to the demographics-at-the-beginning condition were compared with the response characteristics of participants assigned to the demographics-at-the-beginning condition (Green et al., 2000).

8. Ultimate reasons are not the same as proximate mechanisms, but a better understanding of the underlying (ultimate) principles facilitates our understanding of why people think, feel, and act as they do (Durrant & Ward, 2015).

9. We conducted the same analyses with and without prior offending. The pattern of results proved the same. The significance levels remained the same although the size of the coefficients lowered somewhat.

10. To test whether the estimated empirical relationships between the central constructs are equivalent across the independent samples, the following statistical test for the equality of regression coefficients was used $Z=\frac{b_1-b_2}{\sqrt{SE{b_1}^2+SE{b_2}^2}}$ (Paternoster et al., 1998).

11. Non-parametric independent samples Kruskall-Wallis tests were also conducted. The results did not differ from the one-way independent ANOVA tests.

12. In a third model, we examined the stability of the interaction effect by applying the approach suggested by Lubinski and Humphreys (1990) to introduce the quadratic terms of the predictor variables involved in the product term in addition to the product term itself and the individual predictors. The quadratic terms enable the estimation of the interaction effect that is not obscured by artifacts related to the non-normality of the outcome variable. After correcting for the quadratic terms of the predictors anticipated shame-guilt and personal moral norms, the interaction terms retained their significance and their sign in the full sample, and both subsamples (males and females). This finding enhances the stability of the interaction hypothesis (Hirtenlehner et al., 2015). The results can be found in Table 6 in the supplementary materials.

13. Additional analyses of linear regression models estimating the interaction between anticipated shame-guilt and intentions to steal at different levels of personal moral norms can be obtained from the first author.