EMPIRICAL analyses in the private sector have disclosed significant economies of scale (e.g., Adams (1967) and Johnston (1960)). Interest naturally turned to an examination of the possibilities for scale economies in the public sector. This issue is particularly important as local governments encounter greater difficulties in financing the services demanded by their residents.1 The increased financial pressures arise from a larger number of residents demanding services of a better quality as well as the limited ability of fragmented governmental units to raise revenue. Likewise, inflation has taken a heavy toll on municipal budgets (Walzer 1971). Municipal services are largely of a personal nature and significant increases in productivity may not be easily achieved (Baumol 1967). This is especially important in police protection where wages and salaries account for almost 90 per cent of total expenditures. The labor-intensiveness of police services has led researchers to doubt the presence of significant scale economies. The claim has been made that extra costs of maintaining substations and precinct headquarters will offset cost reductions from mass purchases, etc.2 However, a large police department can concentrate more resources on a particular neighborhood during critical periods such as during expected criminal activity. Also, neighborhoods can be sealed off after a crime has been committed and research has suggested that immediate investigation by police officers substantially increases the likelihood of apprehending the offender (Challenge of Crime, 1967). Since lengthy investigations are time-consuming and thus very expensive, large police departments should have a decided cost advantage. Larger departments are also able to employ computerized data storage-retrieval systems and specialized personnel to a greater extent. On an a priori basis, one might expect specialized investigators to solve cases more quickly than would relatively inexperienced officers. This paper reviews some of the earlier findings on scale economies in police protection, suggests a new measure of scale, and reports the results of an empirical investigation on a sample of Illinois cities. Finally, an expenditure per capita and population analysis is conducted on the same sample to determine whether the observed declining average cost curve results from sample differences or variations in measurement techniques.

Evidence on cost savings from reorganizing township low-volume, rural-road systems into larger units is presented in the study of technical efficiencies in producing local government services. Data are from a sample of midwestern townships. Heterogeneity of surface types is accounted for by specifying a multiple output translog cost function. Examination of economies of size suggest that cost savings could be realized by reorganizing townships into larger units. The presence of economies of scope suggest that jurisdictions should not specialize in maintenance responsibilities due to the joint use of inputs. Copyright 1988 by MIT Press.