Abstract: In the 1990s in the United States, crime fell in all categories and all parts of the country. This thesis examines whether or not this phenomenon occurred in Canada. Moreover, using annual data at the provincial level, empirical relationship between crime rates and the socioeconomic factors in Canada is investigated with a regression analysis.
Provincial data are collected from CANSIM for the 1990‐2009 period is utilized in the analysis. The following provinces are included in the sample: Nova Scotia, New Brunswick, Quebec, Ontario, Manitoba, Saskatchewan, Alberta, and British Columbia. Dependent variables consist of the overall crime rate, as well as violent and property crime rates. In terms of explanatory variables, factors that relate to population demographics (age, education and immigration), economic conditions (per capita income, unemployment, poverty rate, social assistance) and law enforcement (number of convicts and police expenditures) are included in the specification. Fixed effects models are adopted to control for time-‐invariant, province specific unobserved heterogeneity.