The Distribution of Crime Victimisation in the Population

Abstract

This paper addresses the question of how best to determine the appropriate theoretical model for explaining the frequency distribution typically observed in self-report crime victimisation surveys of general adult household populations. The contemporary, prevailing approach is characterised as a ‘double-hurdle model’ of exposure to victimisation risk mat focuses, separately, upon the transition initially from a non-victim to a victim state (the ‘lifestyle-exposure’ hypothesis), and thence upon the transition to a subsequent, specific level of risk (the ‘repeat victimisation hypothesis’). An alternative model — the ‘immunity hypothesis’ — is proposed with the aim of addressing some of the theoretical and empirical difficulties identified in the current approach. This model takes the form of a compound-Poisson generalisation of the Negative Binomial statistical distribution. Its chief difference from the current approach is its assumption of a general tendency in the population towards ‘immunity from’ rather than ‘exposure to’ crime victimisation risk. An outflow table of data on household property crime victimisation from a longitudinal panel survey is analysed. The results provide support for the hypotheses derived from the immunity model.