Should Criminal Fines be Income-Dependent? Theory, and Evidence from Finnish Speeding Fines

• In Event: Consumption Taxes, Fines, and Social Safety Nets

Abstract

Should criminal fines be income-dependent? We evaluate this question both theoretically and empirically with three motives in-mind: (1) externality reduction, (2) redistribution, and (3) retribution. With regards to (1) and (2), fines should be income-dependent if one of the following holds: externalities vary with the income of the offender or preferences for committing crime are correlated with earnings ability. Fine elasticities govern the relative importance of these two forces. Regarding (3), a retributive social planner does not fully acknowledge the economic surplus of committing crime; a planner may therefore use income-dependent fines in order to equalize the “utility cost” of committing crime across the income distribution. We evaluate the empirical relevance of these motives within the context of the Finnish speeding fine system, which is partially income-dependent. We draw on linked accident report, income-tax return, and criminal history data to assess whether speeding externalities vary with income; a key finding is that, conditional on an accident, driver income is positively correlated with the number of resulting injuries/deaths. To assess the role of preference heterogeneity, we build on the optimal commodity taxation literature, which shows that across-income preference heterogeneity can be identified by differencing out the causal earnings effect on speeding from observed (cross-sectional) speeding consumption. We measure this earnings effect using several sources of variation, including within-individual earnings changes, spousal earnings changes, and inheritance shocks; preliminary results indicate that most across-income differences in speeding behavior stem from earnings effects. Finally, to quantify retributive/fairness motives, we build a contingent-valuation survey.

Author

• Aaron Payne, The Wharton School at the University of Pennsylvania