A replication study of moral hazard in bodily injury liability auto insurance claims filing decisions

Abstract

Cummins and Tennyson identified the presence of moral hazard in bodily injury liability auto insurance claims by adding to their model a survey response variable indicating attitudes towards dishonest behavior. The attitudinal variable was a proxy for the psychic costs a claimant considers when deciding whether to file a fraudulent insurance claim. Cummins and Tennyson found a significant relationship between the acceptance of fraudulent behaviors and the frequency of bodily injury liability auto insurance claims. This paper is a conceptual replication of the Cummins and Tennyson study utilizing a different set of data that allows for a more direct measure of the decision to file a claim and a different research methodology more appropriate for the new data. The results largely support the original results found by Cummins and Tennyson except the variable used to account for the psychic costs of initiating a fraudulent claim is not significant across all versions of the variable as indicated by the original research.