These five critical perspectives define and locate the problems in different ways. Some place more emphasis on structural inequalities among the spheres of government, others focus on the inefficiencies of our complex system, and others on the lack of clear responsibilities for improving outcomes for citizens and business. While the diagnoses and prescriptions are diverse, there is a more fundamental distinction that can be drawn between those who support incremental models for change (adjust and improve current arrangements) and those who support radical models (fundamental reconstruction of governance systems including abolition of the current array of states).

The radical case for abolishing the States and building up the local or regional level of government has been proposed from two main viewpoints, which I will term the ‘top-down rationalist’ approach and the ‘regional-responsiveness’ approach. The top-down rationalist model for change is largely shared by both political centralists and the advocates of economic efficiency. Both argue for abolishing the States in order to cut wasteful duplication, cut costs of ‘surplus’ politicians and bureaucrats, and promote national standards. Massive benefits (up to $30 billion annually) are suggested (Drummond 2001b), of which half arises from lower public sector running costs (simplified machinery of government), and half from boosting economic productivity (lower taxes and reduced regulatory costs for business). Former Brisbane Lord Mayor, Jim Soorley, claimed some years ago that the costs saved by abolishing the States (and thereby removing ‘duplication, bureaucratic red tape and waste’ between the three levels of government), could be used to eliminate Australia’s foreign debt within 20 years (Soorley 1994).

As with most top-down models of change, proponents have given little consideration to the full range of issues. One problem is the lack of attention to likely costs as well as benefits – for example, the medium-term organisational costs of transitional arrangements, the difficulties of maintaining and building legitimacy and support while large-scale changes are underway, and the practicalities of leadership and change management where there are dispersed winners and losers. A second problem is the lack of attention to ensuring that the alleged deficiencies of states (fragmentation, inefficiency, appropriate powers, etc.) would not be repeated, and indeed on a much wider scale, among the new regional entities. It is very difficult to ensure that major changes do not create unintended dis-benefits.

A third set of problems is the lack of clarity about how many local/regional authorities are required to achieve the right balance between organisational capacity or effectiveness and democratic responsiveness – obviously somewhere between the current number of States/Territories (8) and the current number of local authorities (700). The answer has varied widely. Soorley (1994) claimed that ‘about 20’ new regional governments would suffice. On the other hand, a federal report on regional development identified 67 regions from the viewpoint of economic development characteristics (Kelty 1993), fewer than the 80 regions proposed much earlier by the Department of Urban and Regional Development (DURD 1973). Two Federal Labor MPs (Tanner and Snow) who proposed abolition of the States, in Federal parliamentary resolutions tabled in 1993-94, proposed to substitute ‘less than 100’ new local/regional bodies. Drummond (2000, 2001a) has more recently proposed ‘between 40 and 60’ as the optimal number with a preference for the larger number; around 30 of these would be based on metropolitan and provincial cities.

An alternative to the top-down systemic approach is the regional responsiveness approach, which draws its strength from a sensitivity to regional identity, interests and sense of place (see Berwick this volume). Regional champions argue that ‘communities of interest’ are not sufficiently recognised in the current arrangements for state-level politics and administration. This sentiment underlies the ‘New States’ movements (discussed in Brown this volume) that have emerged in some areas of rural and remote Australia, especially in the larger States – Queensland, New South Wales and Western Australia. Here, a distrust of metropolitan governments is closely linked with arguments that the historical boundaries of the large states are quite inappropriate. Support for greater devolution of powers to the regional level is stronger in such areas. This is a case-by-case approach, generated by local and regional sentiments and local leadership rather than by systems theory.

In terms of the ideal scale for coherent and effective regional entities, it might seem plausible to assume that some existing smaller jurisdictions such as Tasmania and the Australian Capital Territory are perhaps about the optimal size in terms of population and coherence. However, it is an open empirical question as to whether they are actually seen as more legitimate in eyes of their citizens than the larger jurisdictions.

In terms of the appropriate powers required by regional entities, there is no agreement among the advocates of two-tier federalism about whether new regional governments should have the same powers as the existing States (and thus the same conflicts with the Federal Government over roles and responsibilities) or should become largely service delivery arms of the national government. There is a distinct possibility in two-tier federalism that centralisation could be even greater, with many of the more important state powers flowing to Federal Government and with regional governments being more fragmented and weaker than the current States.