Unemployment, business cycles, crime, and the Canadian provinces

•We test the Cantor and Land (1985) model of unemployment and crime•Multiple measures of the state of the economy are used, including unemployment•We find support for the short- and long-run effects of the economy of crime•Different state of the economy variables have nuanced relationships with crime

PurposeTest the Cantor and Land (1985) model using multiple measures of the state of the economy.MethodsA panel data set of the 10 Canadian provinces, 1981 – 2009, is analyzed using a hybrid modeling approach called a decomposition model. Rather than one economy-related model, four are included in the analysis: gross provincial product, gross provincial product per capita, unemployment rate, and low income.ResultsAll economy-related variables matter for property and violent crime, but the sign and magnitude of the estimated parameters vary based on context.ConclusionsThe relationship between the economy and crime is complex. Only including one economy-related variable appears to result in omitted variable bias. As such, any evaluation of the relationship between the economy and crime must consider multiple measures of the economy.