The Dynamic Behavior of Public Budgets: An Empirical Study of Australian Local Governments

Abstract

POPULAR criticism to government emphasizes a slowness of the bureaucracy to respond to changing demands for service and a slowness to curtail unnecessary services. There is a long intellectual tradition supporting the continuing taxpayer revolt. But this argument emphasizes a different kind of bureaucratic distortion: the long-run tendency of government institutions to expand services beyond the voter preferred level and to waste resources in the production of services. ' It may be that short-run inflexibility may be a more important source of citizen complaint than long-run bureaucratic or political distortions. The customary methodology of public sector econometrics tends to divert attention away from such issues by assuming from the start that public services are closely related only to the preferences of the median voter 2 It supposes that divergences from median preferences are rapidly eliminated by the dynamics of political competition. For local governments an alternative argument for the median voter specification is the possibility that dissatisfied voters will migrate out of the jurisdiction. This model leaves no room for institutional distortions of either the inflexibility type or the over-expansion type.3 This paper analyzes only the former of these two distortions: the proposition that government budgets are relatively inflexible. A finding of substantial inflexibility suggests that public dissatisfaction with government could be in part due to government budgets that are continually out of equilibrium with the ever-changing demands for public services.4 Specifically, this paper is confined to the narrow issue of measuring the adjustment dynamics of public budgets among Australian local governments during the period 1967-1974. In a changing world inflexibility implies that budgeted expenditures may be far away from their equilibrium levels at any particular time and that the rate of convergence is slow. Such inflexibilities may arise from at least two sources. One is bureaucratic inertia. For example, a too large agency may resist budget cuts successfully, perhaps by mobilizing interest groups benefiting from its budget. Or conversely, a too small department may experience resistance to rapid expansion of its services from other departments competing for the same total revenues. A second source of inflexibility is the existence of capital adjustment costs. These are the costs of expanding or contracting the existing stock of public capital which imply that rapid changes in capital expenditures will be avoided.5 The econometric specification hypothesizes a long-run budget target, which incorporates in an implicit way both voter preferences and long-run institutional distortions. In order to test the hypothesis that the actual budget is slow to respond to changes in the long-run target allocation, a generalized adjustment function is specified. The results are then compared with those of the more conventional assumption that all adjustments are completed within one year. This specification also permits an investigation of the dynamic response of the budget to an exogenous shock or a trend in any of the variables which determine the long-run budget target. Consider, for example, the dynamic impact of a policy shift in the level of intergovernmental grants and suppose these are conditional grants for road construction. One would expect that the spending impact of the Received for publication May 7, 1979. Revision accepted for publication September 30, 1980. * University of Utah. I wish to thank Bruce M. Hill, Frank Lad, and a referee for helpful comments and David Hamilton for computational assistance. This research was supported by a grant from the Schools of Economics and Social Sciences at La Trobe University. 1 For a useful review of this literature see Borcherdring (1977). 2 For example, see Oates (1969) or Deacon (1978). 3Pommerehne (1978) presents cross-sectional evidence that while political and bureaucratic distortions exist in the long run, they do not dominate variables connected with the preferences of the median voter. 4 There is a substantial literature on bureaucratic inflexibility; for example, see Simon (1959) or Crecine (1969). 5 There is a large literature on adjustment costs and their impact in determining the investment behavior in the private sector; for example see Gould (1968).