Asymmetry in process-based model relationships: A longitudinal study of adjudicated adolescents*

The asymmetry hypothesis predicts that negative police encounters matter more than positive ones for legitimacy, suggesting that officers may get little credit for using procedural justice. We tested the asymmetry hypothesis and extended it to other process-based model relationships by estimating asymmetric fixed effects models with longitudinal data from adjudicated adolescents. By utilizing within-individual variability and decomposing accumulated positive and negative changes in the predictors, these models pushed beyond the limits of existing research. Prior studies of asymmetric effects in policing either focused on the impact of a single encounter, often one that was both hypothetical and vicarious, or were unable to control for all time-invariant confounders. Our findings reveal that positive and negative changes in perceived procedural justice are both related to changes in legitimacy. The relationship is symmetric for global perceptions of procedural justice but asymmetric for encounter-specific perceptions. Both positive and negative changes in legitimacy are related to changes in offending variety, but the relationship is symmetric. The implications are that police do get credit for procedural justice, and that they must work to maintain legitimacy once they have it, because it is something that can be lost, and its loss has consequences for offending.