The analysis indicates that plausible assumptions about the way in which unemployment affects participation in property crimes lead one to expect highly lagged effects. These assumptions may explain why studies of the relationship between unemployment and crime produce such inconsistent results. A simple Poisson model of offender populations is developed to explore the effects on offender populations, and therefore aggregate crime rates, of changes in the propensity to become involved in crime and changes in the propensity to persist in crime. Consequences of assuming that unemployment increases the rate at which individuals become involved in offending are explored, and implications of the analysis for future research on the relationship between unemployment and crime are discussed. The authors indicate that, even with large instantaneous increases in unemployment, the model shows a very slow increase in the criminal participation rate. 26 references and 3 figures