To test the norm of just deserts, the study employs ratio scale measures of crime seriousness and punishment severity and used a formula derived from both equity theory and psychophysics in model fitting. Results from a sample survey indicate dramatically strong use of the principle of just deserts by the public but less adherence to just deserts by demographically disadvantaged respondents (i.e., low income or black). A path model of the relation between aggregate and individual scores further demonstrates a fundamental tension between the two versions of normative consensus, in that the more respondents used the principle the more systematically they deviated from the group average response. The paper discusses the implications of both these findings and the new methodologies employed for understanding normative consensus and the assessment of justice norms. Tables and 38 references are supplied. (Author abstract modified)