Economically disadvantaged juvenile offenders tried in adult court are perceived as less able to understand their actions, but more guilty

Abstract

We investigated the influence of a juvenile defendant's socioeconomic status (SES) on mock jurors’ perceptions of a juvenile tried in adult court. As predicted, participants convicted the low SES juvenile defendant of felony murder significantly more than the middle or high SES juvenile defendant. Yet, participants also rated the low SES juvenile as less mature than the middle or high SES juvenile – a belief that past research shows predicts leniency in verdicts (i.e., not guilty judgments). Finally, stereotypes about the criminality of low SES juvenile defendants, not a lack of perceived similarity, partially mediated the effect of SES on guilt.

Keywords:

• stereotypes
• guilty and responsibility
• juror decision-making
• juvenile offenders
• socioeconomic status

Acknowledgments

We thank Dr. John Lakey for helpful comments on earlier drafts of this research. This research was awarded the 2010 First Place Senior Thesis Award and supported by grants to the first author from the University of Evansville.

Notes

1. Thirteen participants suspected that the study was examining the influence of defendant SES on case judgments and were excluded from analyses, reducing the total sample of 128 to 115.

2. Although most participants were middle to high SES, a small minority (12%) were of low SES and therefore might perceive themselves as more similar to the low SES defendant. Thus, one could argue that these participants should be excluded. To test this possibility, we conducted an ANOVA to examine whether the SES of a participant (either low or middle/high) would interact with defendant SES on how similar participants felt to the defendant. For this analysis, we categorized participants according to how similar their own total family income was to either the low or middle/high defendant SES condition. Those participants with total combined incomes of less than $30,000 were labeled as ‘low SES,’ while those with total combined incomes of more than $30,000 were labeled as ‘middle/high SES.’ An ANOVA revealed no significant interaction of participant SES and defendant SES on how similar participants felt to defendant, F(2, 101) = 0.60, ns. Because there was no relationship or interaction between participant and defendant SES on perceived defendant similarity, we included all participants in analyses. We also replicated all analyses excluding all low SES participants (12% of participants) and found no differences in the results, further supporting our decision to include all participants in analyses to help preserve statistical power.

3. One limitation of this item is that it references both maturity and responsibility (not maturity alone).

4. Twenty-five participants missed the manipulation check regarding parental occupation. We conducted all analyses excluding these participants and found few differences in results. Surprisingly, there were slightly fewer effects of SES when excluding these participants. Specifically, excluding these participants reduced the significant effect of defendant SES on degree-of-guilt for murder to marginally significant, F(1, 95) = 0.12, p<0.10, and the marginally significant effect of defendant SES on guilt score to not significant, F(1, 100) = 0.36, ns. There were no other differences in the results. Because there were so few differences in results, and when we excluded these participants the effects of defendant SES were weaker, we present results including all participants, thereby maintaining the integrity of our sample and preserving statistical power.

Ten participants missed the manipulation check regarding defendant age and believed the defendant to be an adult. We conducted all analyses excluding these participants and found no differences in the results. Because there were no differences in results, we present results including all participants.

5. There are different degrees of freedom within the analyses because some participants chose to skip certain questions, and therefore, some data are missing.

6. Indeed, our research confirms that these judgments are pro-defense judgments. Specifically, greater perceived maturity and greater ability to comprehend risks both predicted greater degree-of-guilt judgments for robbery and for murder (all βs > 0.20, all ps < 0.05).

7. We also conducted mediation analyses using the dichotomous murder verdict as the dependent variable. The results were almost identical to the mediation analysis using the continuous degree-of-guilt variable (Sobel z = − 1.69, p=0.09).

8. All analyses were replicated controlling for participant race (coded as White vs. non-White) and revealed no differences in the results as compared to not controlling for race.