Purpose: Negative effects of crime encompass several different dimensions. As a result, there is no existing methodology capable of dealing with all the relevant issues simultaneously and the interpretation of the estimates currently available lacks theoretical foundation. The purpose of this paper is to provide a unified view of the meaning and relationship between the various dimensions of the welfare cost of crime and violence available in the literature.

Design/methodology/approach: The paper develops a theoretical model of crime and illustrates the different interpretations of welfare costs of crime and violence within this unified framework. This theoretical benchmark is then used as a benchmark to review the empirical literature on the topic.

Findings: The analysis suggests that the most commonly estimated dimension of the welfare cost of crime - related to the total loss associated with crime - although relevant as an illustrative tool, is not very useful from a policy perspective. The literature should therefore move closer to the idea of estimating marginal costs and benefits in order to become policy relevant.

Research limitations/implications: Policy-oriented research related to optimal law enforcement should move in the direction of estimating the marginal willingness to pay of individuals for reductions in crime. This should be compared to the marginal cost of alternative policies in order to guide public policy in the area.

Originality/value: This survey rationalizes in economic terms the estimates from the existing methodologies, highlights some of their limitations, and points out potential directions for future research. It provides one of the first unified views of the various dimensions of welfare cost of crime and violence that have been presented in the literature.