ABSTRACT
Objective
A large literature across the social sciences examines the punishment preferences of the mass public. Although this work yields valuable insights into the correlates of punitive attitudes, differences in research design and measurement complicate comparisons across studies. We engage in a large-scale exploratory analysis of the dynamics that shape punitiveness.
Methods
We fielded two surveys in the United States with multiple measures of punishment preferences, providing a uniquely comprehensive assessment of the individual characteristics associated with these attitudes. Within the surveys, we also conducted 13 experiments examining how contextual features of crime—including offense type and victim and offender characteristics—influence punitiveness.
Results
Three general findings emerge. First, despite variation across measures, some individual characteristics—such as worry about victimization and authoritarianism—exert consistent independent effects on punishment preferences. Second, average levels of punitiveness vary with contextual features of the crime, particularly descriptions of the offense and the individuals involved. Third, individual characteristics interact with contextual features such that inferences about the predictors of punitiveness often depend on crime context.
Conclusion
We call for greater conceptual clarity and increased attention to how measurement choices shape conclusions about punitiveness. We propose a framework for organizing predictors of punitiveness, situating empirical findings within the broader literature, and guiding future research.