Though crime and the fear of crime have a deep negative impact on the well-being of individuals and societies, there are several aspects of this phenomenon that are not adequately understood at present within economic models. For instance, individuals' perceptions of the probability of punishment (for a given crime) differ widely within and across societal groups, and these perceptions in turn differ from the actual probability of punishment. Also, even after controlling for various economic and deterrence variables, crime rates in different societal groups have often been observed to be strongly correlated with background variables like location, age structure and ethnicity.
One of the points of departure of this paper concerns the treatment of individuals' perceptions. In the literature based on Becker's seminal study, an individual's perceived probability of punishment is typically treated as an exogenous parameter that is the same for all individuals and equal to the actual probability of punishment. The present paper begins by developing a model in which an individual's perceptions and the resulting choice are endogenously determined. This model deals with some basic economic features of the information available to individuals, with how this information is generated within the economy, and with some of the consequences of the endogeneity of their perceptions.
This model of individual perception and choice, when aggregated to the economy-wide level, yields dynamic relationships linking certain features of the (economy including past crime rates), to individuals' current choices, and then to the current crime rates in different societal groups. These relationships suggest a possible way of understanding some observed patterns. They also suggest insights concerning, for instance, how criminality might evolve over time in a society, why two societal groups with roughly similar economic parameters might exhibit different crime rates, and why crime might spill over across different societal groups.