Patrol shift staffing levels and community crime rates: Analyzing the impact of reduced per-shift staffing levels on reported crime rates

Abstract

The purpose of this case study was to test the hypothesis: Police staffing deployment levels inversely affect reported crime rates. There is a growing body of research that looks at the issue of police staffing levels and their relationship to reported crime. The findings are conflicting, with some showing a moderate negative correlation. These studies use panel data from several large agencies and tend to center on the relationship of crime rates to the total number of officers employed for a given period. This method could lead to errors due to the varying structures of law enforcement agencies across the nation, which utilizes non-sworn staff. A longitudinal study was designed and implemented to examine the effect of actual fielded staff on reported crime rates. This study examined 5 years of daily patrol staffing levels for a medium size agency in south central Kansas, and compared it to the annual reported crime during the same period. The results of various statistical models showed a significantly high negative correlation between the mean number of officers on daily patrol and the amount of crime reported. The correlation was further analyzed to determine if there was the presence, and if so the direction of predictive causality. A Granger Causality Test was applied to this data and in 3 of 4 lags an there was an indications a likelihood that low staffing deployment levels caused an increase in crime. The same test was performed reversing the hypothesis and each of the 4 lags indicated a strong likelihood that reported crime did not affect shift deployment levels.