Crime Trends and Police Expenditures

All previous investigations into the effect of crime trends on appropriations for police protection have employed either inappropriate cross-sectional data or unrealistic singular lag structures. The model developed here involves a more reasonable lag structure and is estimated with time series data. The expenditure for police appears as a function of the crime rate with a lag structure that begins at two years, that contains geometrically declin ing coefficients, and that has a mean lag of 3.9 years.