A Panel Data Approach

We use aggregate datafrom 42policeforce areas to examine the relationship between economicfactors and property crime. The change in unemployment is seen to impact positively on four types of criminal activity examined. A -widening of the distribution of weekly earnings of full-time manual men (a measure of wage inequality) and increases in cars per capita (a measure of the availability of thievable property) lead to increases in crime. We also find evidence that increases in the size of the police force reduce property crime. The past decade has seen a dramatic change in the number of notifiable offences recorded by the police in England and Wales. For example, theft and handling stolen goods per 100,000 of the population was 4,014 in 1986, 3,844 in 1988, 5,622 in 1992 and 4,600 in 1996 (Criminal Statistics 1996: 45, table 2.3). It was also a decade that saw considerable fluctuations in the unemployment rate. The UK male unemployment rate fell from 13.7 per cent in 1986 to 7.5 per cent in 1990. However, between 1990 and 1993 it increased back to 14.0 per cent, but fell thereafter to 10.3 per cent in 1996. On a theoretical level the link between unemployment and crime may go either way. One hypothesis would be that high unemployment implies a restriction on the availability of legal activities and thus reduces the opportunity cost of engaging in crime. Thus a positive correlation would be predicted between unemployment and crime. On the other hand unemployment may be viewed as not only an indicator of the number of individuals who are out of work, but also as a more general measure of economic activity (e.g. amount of goods and services produced). In this case, unemployment may also have a negative effect on crime, since it could increase guardianship of property (more people home when economic activity is low) and reduce the availability of property (e.g. fewer goods in circulation). Cantor and Land (1985) argue that with this possible negative effect it is difficult to identify empirically both the positive and negative effects that might be present and that this may account for the null effect overall observed in previous analyses. The aim of this paper is twofold. First, we use police force level data in England and Wales to establish the impact of unemployment on crime. In doing this, we attempt to control for the availability of thievable property that is correlated with economic activity.