Keeping It In‐House: Households Versus Population as Alternative Proxies for Local Government Output

Abstract

Forced amalgamation has been used as a policy instrument in local government by numerous regulatory authorities across the world. A common presumption underlying municipal mergers holds that larger local councils will experience greater economies of scale. However, the empirical evidence on this question is mixed. Part of the reason for this could lie in the frequent use of population as a proxy for local government output in the empirical literature. This paper examines the use of alternative proxies, particularly the number of households but also the addition of business unit data. We demonstrate that household data represents a more accurate proxy of Australian local government output compared to population size. In addition, the paper employs experimental data, conceptual considerations on population, and household dynamics to establish that the number of households represents the most appropriate measure of local government size for both empirical and public policy purposes.

Council type and the determinants of municipal expenditure: A semiparametric analysis of South Australian local government

“…Secondly, at least in the institutional milieu of Australian local government, with its emphasis on ‘services to property’, the use of population size as the proxy for local government output cannot be justified. In their study of Victorian local government, Drew and Dollery (, p.245/246) found that ‘in Australia and the numerous other countries in which the role of local government focuses predominantly on the provision of “services to property”, both the number of households and the number of households plus the number of business entities represent superior proxies for the quantum of local government output’ to population size per se. Thus, using this variable as an output provides a more concise and less volatile measure in the performance analysis of local councils.…”

“…This measure was constructed to capture major non‐discretionary expenditure on local roads. It is noted that ‘per household’ includes business units and represents a more precise and less volatile measure of output; local business have access to the same local services as households (see Drew & Dollery, ; Drew et al., ). This means that types of expenditure are divided by the total number of households and business units before taking logs of these computations.…”

Less Haste, More Speed: TheFit for the FutureReform Program in New South Wales Local Government

“…Finally, considerable doubt has been created as to whether population size is a suitable proxy for local government output in Australia (Drew and Dollery ). The number of households and businesses aligns far better with the unit of actual service provision and it is less volatile and more accurate in inter‐censal periods.…”

“…The number of households and businesses aligns far better with the unit of actual service provision and it is less volatile and more accurate in inter‐censal periods. Thus the OLG may well be conducting its structural reform agenda on an entirely fallacious unit of scale (Drew and Dollery ).…”

“…This presents relevance problems given that the OLG has decided to calculate real operating expenditure per capita for the 2013 and 2014 financial years: they have attempted to resolve this problem by using projected 2013 population estimates with an effective one year lag . Secondly, ABS () population data in inter‐censal periods are simply an estimate: a recent study by the ABS identified errors in inter‐censal estimates ranging from 15.2% (for statistical areas with less than 2000) to 2.4% error (in statistical areas with populations greater than 20000, Drew and Dollery ). Finally, serious doubt has been cast on the practice of using population as a proxy for local government size in service provision given that ‘services to property’ (i.e.…”