Structural reform in the guise of enforced local council mergers has a very long history in Australian local government (Vince 1997). In an analysis of the debates that have surrounding various episodes of council consolidation, Dollery et al. (2006b, pp.139–55) have provided a detailed evaluation of both the theoretical and empirical arguments typically advanced for local government amalgamation in Australia. They identified seven main conceptual considerations that have been brought to bear on the problem: ‘Optimum community size’; economies of scale; economies of scope; local government capacity; administration and compliance costs; the ‘coincidence of municipal and natural boundaries’; and public-choice arguments. This taxonomy provides a useful framework for considering the merits or otherwise of the conceptual case for amalgamation presented in the Final Report of the Queensland Reform Commission. We will thus briefly consider each of these factors in turn.
As we have seen, the Reform Commission placed a good deal of weight on the importance of ‘regional communities of interest’ especially in economic development, planning and other matters with a comparatively large spatial focus. This demonstrates that the Commission considered the ‘region’ to represent an optimal ‘community size’, rather than smaller local government areas. However, the Commission seems completely unaware that notions of optimum community size are not simply plucked from the air but derive from the theory of fiscal federalism pioneered by Oates (1972) and subsequently refined by legions of economists (Mueller 2003). The theory of fiscal federalism is centrally concerned with the question: which levels of government (national, state or local) should provide specific categories of public goods? The theoretical answer follows from Oates’ (1972) correspondence principle: the size of a government should reflect the area of benefit of the goods it provides to its constituents. Each public good should thus be provided by the smallest (that is, lowest-level) government where there are no spatial externalities affecting adjacent areas.
This principle has direct implications for local government amalgamation. In the first place, Oates (1972) has demonstrated through the decentralisation theorem that if local preferences determine the composition of local service provision, then welfare gains accrue to society because preferences are never spatially uniform. Local service provision should thus be decided at the local level, implying the retention of small local councils, at least insofar as deciding the composition of local services. Secondly, if spillover effects exist contingent on the exercise of local choice, then either subsidies or taxes must compensate neighbouring jurisdictions, or decisions over the services generating externalities should be taken at a higher level of government. Demand-side considerations of this kind are precisely the reason that various alternative models of local government prescribe that services with a regional impact should be decided at the regional level, while supply-side considerations, like scale economies, lead to an analogous conclusion in these models.[7]
Economies of scale refer to a decrease in per-unit cost of production as the quantity of output increases. As we have seen, in its Final Report the Commission made frequent appeal to scale economies by arguing that larger amalgamated councils would provide services more cheaply. This claim cannot be sustained for several reasons (Dollery et al. 2006b). Firstly, despite the recent expansion of Australian local government services to include more ‘services to people’ (Dollery et al. 2006c), Australian local councils still have a strong ‘services to property’ orientation in terms of the services they provide. However, despite the relatively narrow range of service provision, Australian municipalities nonetheless deliver an extensive range of services. Because these services are produced using entirely different processes, there is no a priori reason for different services to display the same cost characteristics. It follows that while council mergers will secure scale economies for some services, they will also capture diseconomies of scale in other services. Sancton (2000, p.74) has summarised this argument by observing that ‘there is no functionally optimal size for municipal governments because different municipal activities have quite different optimal areas’.
Secondly, it is well known that scale economies are not relevant if service provision can be separated from service production through the ‘purchaser-provider split’, since scale economies only occur during the production phase (Oakerson 1999). Thus local councils too small to achieve scale economies can still reap these cost advantages by outsourcing to private firms, Regional Organizations of Councils, area integration models, and the like. In other words, council size need bear no relationship to scale economies.
Thirdly, in its adoption of the ‘big is beautiful’ perspective of local councils, the Commission alludes to the cost ‘dividends’ attendant upon scale economies in the proposed new larger councils. The Commission thus implicitly endorsed inter alia the findings of Stephen Soul (2000) in his influential doctoral thesis, which examined the effect of council size (as measured by population) on gross expenditure per capita, and concluded that increasing population yields a lower level of gross expenditure per capita up to a council size somewhere between 100,000 and 316,000 people, at which point ‘scale diseconomies’ begin. But the theoretical basis of this study has been shown to be badly flawed on the basis of pioneering work by Boyne (1995) ignored by Soul (2000) (Dollery et al. 2006b). In essence, Boyne (1995) has demonstrated that council size (as proxied by population) bears no relationship to scale economies, since population is linked to numerous other variables affecting expenditure.
Finally, the Commission apparently takes for granted that substantial scale economies exist in Australian local government. This presumption is unwarranted and ignores both Australian empirical evidence on economies of scale in local government (see, for instance, Byrnes and Dollery 2002) as well as empirical evidence abroad (see, for example, Bish 1971; 2000; Boyne 1998a; Duncombe and Yinger 1993; Hirsch 1968; and Rouse and Putterill 2005), which points to the fact that scale economies cease for many municipal functions for populations above 50,000 residents and many labour-intensive services exhibit diseconomies of scale. With regard to Australian local government, Byrnes and Dollery (2002, p.405) conclude that ‘the lack of rigorous evidence of significant economies of scale in municipal service provision casts considerable doubt on using this as the basis for amalgamations’.
The Commission makes no reference at all to scope economies. Given the complete neglect of the literature outside of Queensland in the Final Report, this omission is unsurprising. However, given the potential significance of scope economies, this oversight is most unfortunate.
Scope economies arise where production functions facilitate the joint production of two or more services simultaneously. Under increasing returns to scope, joint production by one organization generates more output than separate production by two different organizations using the same quantity of input. Within Australian local government service provision, then, it is possible to identify four potential sources of scope economies and diseconomies in local council services: decreasing returns to inputs; jointness in inputs; jointness in outputs; and interactions between the processes of service provision. According to (Dollery and Fleming 2006), the most likely source of scope economies in Australian local government derives from jointness in inputs, which occurs where one input can be used in the production of more than one service. Municipal administrative functions, where the same functions can be used in more than one sphere of activity, are easy to identify. For instance, in the event of council amalgamation or council resource-sharing, centralized administrative inputs can be used to support various activities, thereby reducing costs. Despite the promise offered by scope economies, no empirical studies have yet investigated the phenomenon in the Australian local government context (Dollery and Fleming 2006).
As we have seen, in its Final Report the Commission set great store on enhanced local council capacity as a positive consequence of its forced amalgamation recommendations. In particular, the Commission (State of Queensland 2007, p.39) held that local government capacity could be expected to improve in four main areas: better asset and infrastructure management; increased ability to ‘attract and retain quality staff in key positions’; superior ‘risk management and compliance with financial and other reporting requirements’; and improved growth management. While no conceptual or empirical evidence at all is presented in support of this claim, we contend that this assertion seems reasonable, especially in the context of small remote and rural councils.
In Australian Local Government Economics, Dollery et al. (2006b, pp.145–6) consider precisely the same argument:
A proposition sometimes advanced in the Australian debate over amalgamation is that larger councils tend to possess greater levels of administrative and other expertise, in part due to the fact that their size permits the employment of specialist skills that cannot be acquired readily by smaller municipalities. Given the increasing burden placed on Australian local government by its state and federal counterparts, through cost shifting and other activities, it is held that this confers a significant advantage on larger municipal units because it enables them to accomplish a wider and more complex range of tasks in a more efficient manner.
While adding the caveat that no empirical work has been undertaken on the issue in Australian, Dollery et al. (2006b, p.146) nonetheless argue that ‘there seems to be considerable merit in this argument’ since ‘small regional and rural councils do struggle in terms of expertise and cannot always use consultants in an effective and prudent way’. However, they add that many alternative models to amalgamation can achieve the same outcomes since they too can ‘pool their resources to acquire the skills in question, at no greater cost than to single and larger councils’.
Dollery et al. (2006b, p.148) observed that:
An additional argument often put forward in support of local government amalgamation is that larger consolidated councils economize on their direct costs of administration and the compliance costs imposed on individuals who participate in the municipal political process, [where] administrative costs include the compensation paid to elected and appointed officials and staff and the overheads (buildings, supplies, utilities, etc.) required to support those officials [and] compliance costs include the costs incurred by municipal voters to keep informed on issues and candidate positions and the potential cash and time of registering an opinion by participating in hearings, meetings, voting, and other activities.
In a nutshell, the argument holds that fewer local councils results in lower administration and compliance costs. This argument is analogous to claiming that scale economies exist in both administration and compliance.
With respect to administration costs, in its dismissal of alternatives to amalgamation in Chapter 4, the Commission asserted that large amalgamated councils would involve ‘less administration and bureaucracy’ and avoid ‘complexity and delays’ inevitable in cooperative arrangements between smaller councils. It did not address a comparison of administrative costs between consolidated councils and existing councils for non-shared functions. This is unfortunate because in the public administration literature a wealth of evidence exists that larger bureaucracies are less effective since (a) longer administrative hierarchies inhibit the efficacious transformation of policy decisions into policy action and (b) decision-making is further removed from situational knowledge and thus is less well informed (Hood 2000). The Commission also ignored a second possibility: reduced oversight costs incurred by the Queensland DLGPS&R arising from far fewer local councils.
The recent debate in New South Wales over council mergers in that state witnessed the development of an argument in favour of much larger local government areas that has come to be called ‘eco-civic regionalisation’ (Brunckhorst and Reeve 2006). In essence, this view is based on the notion that while ‘ecology is forever’, human settlement is transitory and therefore regional management strategies should ensure that ecologically homogeneous geographic areas should form the basis for the design of local government areas, rather than current and future patterns of human settlement.
In section 3.3 of Chapter 3 of its Final Report, the Reform Commission outlined the factors it took into account in determining the boundaries of local government areas, which included ‘environmental’ factors and ‘community of interest’. These issues were considered exclusively in terms of human factors, such as economic development, growth management, mineral extraction, population migration, and the like, with the single exception of forestry as a potential ‘greenhouse sink’. In other words, the Commission did not incorporate eco-civic regionalisation into its deliberations at all. This is fortunate since the extreme notion of eco-civic regionalisation not only completely ignores the costs of the massive restructuring it would entail, but cannot demonstrate why state government agencies cannot adequately manage environmental issues (Dollery and Crase 2004).
The final category in the Dollery et al. (2006b) classification involves public-choice considerations. These arguments have been articulated by numerous scholars of local government, perhaps most notably Bailey (1999) and Boyne (1998a). In effect, they amount to a conceptual rebuttal of the notion that ‘big is beautiful’ by maintaining that large councils are less accountable and transparent and more complex than their smaller counterparts and thus less easily monitored by voters, who have less contact with elected representatives. It is also argued that smaller municipalities are much closer to constituents and thereby better informed than large councils. A second empirical string to the public-choice bow is that ‘bigger is not better’, since considerable evidence has demonstrated that small councils deliver services more cheaply (see, for example, Boyne 1992; 1998b).
As we have seen, the Final Report adopts precisely the opposite view to these public-choice considerations at both the conceptual ‘big is beautiful’ level and the empirical ‘bigger is better’ level. Apart from the fact that the authors of the Final Report cite no literature outside of Queensland, this can also perhaps be partially explained by the fact that the public-choice perspective examines local government effectiveness from a citizen’s perspective rather than an organisational perspective.