As background to the key facts and lessons relevant to current institutional choices, it is important to confront the difficult relationship between concepts of federalism and regionalism in Australia. We need look no further than the standard international political science definition of federalism, to be reminded that federalism and regionalism are fundamentally intertwined, in theory and practice. According to this definition, federalism is ‘a system of government in which authority is constitutionally divided between central and regional governments’ (Gillespie 1994). In Australia, the constitutionally-recognised ‘regions’ of the federation are the six States, being the former British colonies as they stood in 1900. However the force given to ‘state-regionalism’ under the 1901 Constitution immediately raises a tension, because our normal understanding of a ‘region’ – in political life, in economic life, in biogeographic terms and so on – is very different. For the most part, it rarely and sometimes never aligns with our concepts of state government.

Occasionally, we find commentators trying to make Australian federalism fit the mould of the international definition, by describing the 19^th century process of colonial subdivision as one in which British political authority was fragmented between ‘six regional centres’ (Holmes and Sharman 1977: 12-14). However such descriptions are rare, because as an ex post facto justification of Australia’s current structure they are, from a historical perspective, grossly inaccurate (see Brown 2004a, b; 2005; 2006). Since at least the late 1960s, when the legitimacy of federal principles began to revive among Australian experts, much of the debate about the practical realities of federalism has, consequently, resembled two ships passing in the night. Many experts and policy actors have based their analyses on the constitutionally-recognised assumption that ‘state-regional’ differences are the only ones that matter, when it comes to trying to make the federal system work (see Holmes and Sharman 1977: 34-101, 172-80; Galligan 1986: 245-55). In the real world of public policy and popular political culture, however, the vast bulk of citizens operate on an entrenched assumption that Australia has many more than six regions (as also recognised by Holmes and Sharman 1977: 86, 129).

Does this definitional conflict matter in practice, as opposed to theory? The answer is sometimes presumed to be ‘no’, because the concept of ‘regionalism’ in Australian public policy itself takes at least four different forms. First, the concern to map trends in globalisation sees the term ‘region’ often used in a supra-national sense, as meaning groupings of the nations of the globe (e.g. ‘the Asia-Pacific region’). Without exception, ‘regionalism’ is not used in this sense in this book, even though globalisation does have importance for the concepts of subnational regionalism here discussed. Second, as we already see above, many experts in federalism need to see regionalism expressed in direct, political, geographically-specific ways before it can potentially take on constitutional significance. This is true at subnational, supra-national and trans-national levels alike. From this view, credible movements for secession are perhaps the easiest way to identify a ‘region’ in this way (e.g. the Basque region in Spain, Scotland in Britain, or Quebec in Canada), although less militant forms of cultural regionalism are also generally recognised (e.g. regional differences within France, Italy or Switzerland, and indeed between regions such as Ticino that effectively span ‘national’ borders).

However, for reasons associated with the history of the introduction of the term into Australia, at a subnational level regionalism is also defined in a third way – as a reference to ‘administrative’ or ‘scientific’ regionalism, a top-down concept used by experts for purposes of planning, bureaucratic organisation, funding distribution, service delivery or, more recently, community engagement (see Brown 2005: 19-27). This third concept operates independently of regionalism as a bottom-up political or constitutional phenomenon, because it can be used by any government as an administrative strategy for recognising and dealing with the spatial layout of society, whatever the formal political structure. Indeed, because different public programs have different spatial objectives, economies, consumers and stakeholders even within the same community, this concept of regionalism tends to lead to multiple, overlapping definitions of what is a ‘region’ in any given area; as well as multiple, overlapping and sometimes conflicting regional institutions of various kinds.

These regions are actually more accurately described as a product of top-down ‘regionalisation’, than bottom-up ‘regionalism’ based on political self-identification and/or cultural expression (for more on the difference, see Ford 2001: 204-8; Bellamy et al 2003; Gray 2004). Just because many conceptions of ‘the region’ are generated from the top-down, however, does not mean that they do not also provide an accurate description of the social, economic, political and cultural demography of the nation. Regionalisations may be normative, such as Australia’s first official national regionalisation in 1949, setting out 97 ‘regions for development and decentralisation’ (see Brown 2005: 20); and there are important debates over the effects of the different ways in which regional boundaries are drawn, or revised, in public policy (e.g. Brunckhorst and Reeve 2006). However, many regionalisations are purely descriptive, and indeed sometimes also reflect ‘bottom-up’ political realities. From the web of regional boundaries drawn by federal and/or state and/or local governments, broad patterns emerge which confirm that for the vast majority of public purposes and programs – at all levels of governance – we operate according to agreed understandings of ‘region-regionalism’ with little or no relationship with the ‘state-regionalism’ embedded in the Constitution. Currently, key national regionalisations include:

Source: DOTARS 2005

A fourth definition of regionalism has also arisen in Australian public debate since the mid-1990s, tending to further confuse the issue. Increasingly the phrase ‘regional Australia’ has become a bastardised political synonym for ‘rural and remote regions’ – that is, all regions outside the capital cities. The term ‘rural and regional Australia’ or ‘RaRA’ has become familiar (see Pritchard and McManus 2000; Commonwealth 2001). As Brendan Gleeson argues (this volume), this co-option of the term ‘regional’ has run the risk of slanting regional policy only towards rural regions, as if metropolitan regions do not exist, or do not require federal or national policy intervention. However, while a misleading way to approach regional policy, this particular reinvention of the term does demonstrate that ‘regionalism’ is not just a top-down administrative convenience, but also a live phenomenon in electoral politics. We know this because the renewed political interest in ‘regional Australia’ has arisen in response to a particular phase of political restiveness, or electoral instability, in rural regions – and indeed outer-metropolitan ones. Accordingly this bastardised definition reflects something of a hybrid between top-down and bottom-up concepts of regionalism. The response has also extended to a new suite of administrative initiatives in community engagement and place management, often targeted to less advantaged urban and peri-urban communities in addition to rural community renewal.

Rather than seeking a definitive reconciliation of these definitions, it is more important to note here that they exist, and that, on any of them, the place of the region in Australian federalism now matters enormously. It not only matters in immediate political and administrative terms, but raises questions about the evolution of institutional structures over the medium-term and into the future – especially if one accepts the following facts and lessons.