The economics of law enforcement: Production technology comparisons between large and small police units

Two empirical models are developed to test whether the production or cost function is stable across different sizes of police units and to consider the implications for optimal police industry structure. Recent studies have found economies of scale among police units, however, different operational environments may suggest different levels of evolvement and, therefore, that different sizes of police units may operate on different production functions. Cost measurements that assume a homogeneous production function are therefore suspect. A population oriented size criterion is developed to segregate police units into two groups of small and large . The cost function appears to be stable within each group and only needs to be compared between the two groups. Data consist of the merger of two large data bases for the year 1973. These are the Uniform Crime Reports: National Time Series Community-Level Database, and E. Ostrom\u27s Decision-Related Research on the Organization of Service Delivery Systems in Metropolitan Areas: Police Protection. These data allow for a detailed specification of police production among 223 municipal police units. Model 1 is estimated by OLS or GLS for eight crime specific clearance and crime rates where the effects of the left hand side variable are diminished by regression on aggregate endogenous variables. Model 2 is a simultaneous equation system that is estimated using a weighted-two-stage-least-squares procedure. Model 1 and Model 2 are specified exactly the same except that a necessary exclusion restriction in Model 2 is based upon the results of Model 1. This appears valid as the results of both models are generally comparable. Model 2 is regressed on aggregate crime and clearance rates in addition to the crime specific variables. Joint F-tests indicate that the production or cost function is stable across both sizes of police units. There are significant economies of scale among the larger police units. On the other hand, local control may adversely affect the efficiency and output quality of smaller police units. Implications for optimal police industry organization include merger, consolidation, or regionalization