AN EMPIRICAL APPLICATION OF ADDITIVE MODELLING TO SINGAPORE CRIME STATISTICS

Abstract

The objective of this thesis is to develop an additive model to analyse the annual variations of Singapore's crime rate from 1981 to 1996. A fundamental assumption underlying this study is that offenders exercise a high degree of rationality when selecting crime targets and they are motivated to engage in illegal activities because of the perceived high returns and low expected costs. In addition, factors that would enhance individuals' exposure and proximity to the motivated criminals are also considered. The Alternating Conditional Expectations (ACE) technique is applied to obtain optimal transformations for the model and its forecasting power is compared with that of several other models. Empirical results show that all the motivational and victimisational factors are significant in explaining the annual variations of Singapore's crime rate. The ACE model's explanatory power and forecast accuracy are also better than that of the other models considered.