The Challenge of Change: Australian Cities and Urban Planning in the New Millennium



This paper reviews recent research on the changing spatial structure of Australia's major cities from the early 1990s, concentrating on (a) the location of employment and journey to work patterns, (b) the changing nature of housing, and (c) patterns of residential differentiation and disadvantage. The paper argues that the 1990s was a watershed decade during which some taken-for-granted aspects of Australian urban character experienced significant change. It then examines the latest generation of strategic planning documents for these major metropolitan areas, all published between 2002 and 2005, and argues that there is a mismatch between the strategies’ consensus view of desirable future urban structure, based on containment, consolidation and centres, and the complex realities of the evolving urban structures. In particular, the current metropolitan strategies do not come to terms with the dispersed, suburbanised nature of much economic activity and employment and the environmental and social issues that flow from that, and they are unconvincing in their approaches to the emerging issues of housing affordability and new, finer-grained patterns of suburban inequality and disadvantage. Overall, the paper contends that current metropolitan planning strategies suggest an inflexible, over-neat vision for the future that is at odds with the picture of increasing geographical complexity that emerges from recent research on the changing internal structure of our major cities.

Introduction: a turbulent decade?

There is a growing body of evidence that Australia's major cities are undergoing some fundamental changes in character and structure as we move into the twenty first century.

Something new has happened to the structure of our cities over the last two decades or so that can be seen to represent a threshold between earlier phases of urbanisation and what we might, for want of a better term, call 21st century Australian cities. (Randolph, 2004, 482–483)

Each of Australia's five major capital cities (Adelaide, Brisbane, Melbourne, Perth and Sydney) has its own distinctive character but, by the late twentieth century, they shared some long-standing features that reflected common elements in their relatively short histories (Forster, 2004). They were all highly suburbanised and low density by world standards, with high levels of home ownership and automobile dependence. All were ‘doughnut cities’, with rapidly growing new outer suburbs surrounding older areas of population decline or stagnation – albeit with significant gentrification in the inner suburbs. And, though less socially polarised than cities in many other countries, they exhibited significant internal contrasts in levels of wellbeing.

The above characteristics largely persist today. But some significant changes do seem to be under way. Since the early 1990s, rates of home ownership have fallen (particularly for young people), residential densities have increased, and there are signs that levels of car dependence might at least have stopped rising. Despite economic growth, many researchers argue that levels of social polarisation and exclusion have worsened. Patterns of residential differentiation and access to employment opportunities have certainly become more complex. And areas of population growth and decline can now be found in the inner, middle and outer suburbs.

Overall the 1990s and beyond have been characterised by economic growth and prosperity (O’Neill and McGuirk, 2002). But the economic restructuring that shaped and accompanied the growth may, together with demographic, technological and political changes, be transforming some of the taken-for-granted aspects of Australian urban character. This paper reviews recent research concerning changes in the internal spatial structure of Australia's major cities, focusing on the changing geography of employment, housing and residential differentiation, and assesses the significance of those changes for current approaches to urban planning strategy.

Employment location and urban structure

Research based on the 1991 Census (Gipps et al., 1997) confirmed that the location of employment within each of Australia's large cities was highly suburbanised. Only 25–30% of total jobs were located in the CBDs and surrounding core regions. The remainder occupied a multiplicity of locations scattered throughout the middle and outer suburbs. Journey to work patterns were correspondingly dispersed. To generalise, in each of the five major cities approximately a quarter of the employed population had a job in the central core region. A further quarter lived and worked in the same suburban local government area (LGA). The remainder, just over half of all workers, lived in one suburban LGA and worked in another, and therefore had to undertake a cross-suburban journey to work each day. Such a structure led to – in fact, depended on – a high level of automobile use (Gipps et al., 1997). In 1991 the majority of workers, ranging from 58% in Sydney to 73% in Adelaide, travelled to work by self-driven car (Forster, 2004, 67).

Recent research on job location and the journey to work (Healy and O’Connor, 2001; O’Connor and Rapson, 2003; Parolin and Kamara, 2003; Forster, 2004, 57–74) has revealed the increasing complexity of the patterns and the processes shaping them. Four features characterise current patterns. The first is that the central cores, particularly in Sydney and Melbourne, are now increasing or at worst maintaining their share of total metropolitan employment. This reflects strong growth in finance, business and other categories of employment associated with the producer services sector of the economy, and favouring central locations. Similar job growth has occurred in adjacent high-status suburban areas such as Sydney's North Shore and Melbourne's inner eastern suburbs. In Sydney the term Global Arc has been applied to the belt of employment growth now extending from Botany in the south, through the CBD and North Sydney, to St Leonards, Chatswood and Ryde (New South Wales Government, 2005). As the label implies, the phenomenon is seen as springing from Sydney's developing role as Australia's global city (Connell, 2000) and main beneficiary from the globalisation process. But in all five major cities there is an increasing tendency for such regions, described variously as cones of wealth (Badcock, 1997), new economy core cities[and]advantaged near inner suburbs (Baum et al., 2005) and advantaged global economies (O’Connor et al., 2001, 169) to grow and prosper, in high-status residential development as well as employment, and to appear increasingly distinct economically from the rest of their metropolitan areas (O’Connor and Rapson, 2003).

Beyond the prospering cores and high-status regions, the second feature of current patterns is that most suburban employment continues to be attracted to a complex mix of dispersed locations and specialised clusters, rather than to neatly-planned centres, ‘edge cities’ or corridors (Freestone and Murphy, 1998; Pfister et al., 2000; O’Connor and Rapson, 2003). The third feature is that the development of self-contained sub-regions or urban realms within which people both live and work may be occurring to some extent (O’Connor and Healy, 2004), but the process remains patchy and incomplete. Fagan and Dowling's (2005) study of Western Sydney shows that, by 2001, the region contained 80 jobs for every 100 resident workers, but that employment was strongly concentrated in the manufacturing sector and deficient in the finance and business categories. Overall, 70% of resident workers had jobs within the broad region, but most commuted from one local government area to another – generally by car. Moreover, some sub-areas within the region where jobs greatly outnumbered resident workers still had very high unemployment rates. As has long been recognised (Forster, 1974), a balance between the number of jobs and workers in a suburban region, even in the unlikely event that the types of job match precisely the nature of the local workforce, is no guarantee that people will find work locally. The real world is much less simple than that, and this is reflected in the fourth feature of the geography of employment and travel to work: a preponderance of cross-suburban journeys between dispersed origins and destinations, leading inevitably to continued and very high levels of automobile dependence (though the 2001 Census suggests that the proportion of trips by self-driven car may, at least, no longer be rising (Forster, 2004, 67)).

Throughout the 1990s, debates similar to those in the USA (Cervero, 1995; Gordon and Richardson, 1995; 1996) focused on the issues of economic efficiency, environmental sustainability and social equity arising from the dispersed suburbanisation of employment (Forster, 1999). Newman and Kenworthy (1989; 1992; Kenworthy and Newman, 1990) and others (Gollner, 1996) argued that such a structure was inherently unsustainable, and that planners needed to combat automobile dependence by concentrating employment in suburban regional centres, transport corridors and mixed-use ‘urban villages’. Brotchie (1992), O’Connor (1992) and Brotchie et al. (1995) countered that the dispersed suburbanisation of employment provided a flexibility of location for businesses that was essential for the global competitiveness of Australian cities. They also argued that, as in the USA (Gordon et al., 1991), journey to work durations had stabilised or even fallen during the 1980s – the so-called ‘commuting paradox’– as the number of jobs in suburban regions grew to match the number of resident workers and a significant degree of self containment developed. However Mees (1994) contended that some of the apparent stabilisation of trip durations in Australian cities was due to flaws in census journey to work data. And, even if journeys were no longer increasing in duration, they had continued to grow in length (Gipps et al., 1997), largely as a result of switching from public transport to self-driven cars.

Detailed research on changes in journey to work duration, length and mode since the mid 1990s is yet to be published. But we do know that the spatial structure of our cities, as represented by employment location, continues to grow in complexity. And how people participate in employment has also become more complex and variable (O’Connor et al., 2001, 47–51; Forster, 2004, 54–57): more part-time work; more casual work; longer hours of work for some; fewer secure long-term jobs – a process that will be intensified by the Federal government's current workplace deregulation policies. As a result, the travel patterns of most suburban households are likely to remain highly dependent on the automobile.


The structure and tenure of housing in Australia's major cities conformed to a fairly standard formula in 1991 (Forster, 2004, 80). About 75% of the stock consisted of separate houses, with very little high density housing except in Sydney. Around 70% of households owned or were buying their dwellings, 5% rented from a public housing authority (except in Adelaide, where the figure was 10%) and 20–25% rented from a private landlord. This pattern had changed very little over the previous 30 years. It had evolved during the so-called Long Boom of the 1950s and 1960s, largely as a result of Federal government policies that encouraged home ownership, provided funding for public housing systems in each State, and neglected the private rental sector (Paris, 1986; 1993). Around 80% of households achieved, at some stage in their lives, the Great Australian Dream of owning their own separate house in suburbia. Other types of tenure and dwelling structure, though comprising 40% or more of the housing stock at any given time, tended to be seen as inferior, temporary or marginalised forms of accommodation (Ferber et al., 1994).

Since the late 1980s a range of policy changes has affected both housing tenure and structure. All the metropolitan planning authorities adopted policies favouring urban consolidation: trying to cut the rate of suburban expansion by encouraging the construction of new dwellings, usually at higher densities, within the existing built-up areas of the cities. As a result, an increasing percentage of new dwellings in recent years have been apartments, townhouses, duplexes or other forms of medium density housing. Because houses last a long time, the total dwelling stock remains dominated by housing built in earlier periods and therefore has not changed radically in type since 1991. In Melbourne, for example, the overall proportion of dwellings that consist of separate houses only fell from 77% to 74% between 1991 and 2001, even though multi-unit housing made up over 40% of new dwellings constructed over the same period (Buxton and Tieman, 2005). But residential landscapes are changing. Particularly in Sydney, though to some extent in all the major cities, high and medium density dwellings are becoming increasingly noticeable.

Housing tenure has also changed. Despite rising economic prosperity and – by recent Australian standards – low interest rates, housing affordability has declined since 1991 with the rise in property prices outstripping the rise in incomes (O’Neill and McGuirk, 2002). Again, overall rates of home ownership have not changed significantly because the system remains dominated by older households who bought their first homes in easier times. But Yates (2000) shows that the percentage of persons aged 30–34 who owned or were buying their own home fell from 69% in 1974 to 55% in 1994. Debate continues over the significance of this trend. It may simply represent a postponement of home ownership, as young adults take longer to get established in permanent careers and aim to have children later in life (the average age at which Australian women have their first child is now 30). But it may mean that in future an increasing proportion of households in Australian cities – especially Sydney, where property prices are highest – will never achieve the Great Australian Dream of owning their own home (Badcock and Beer, 2000).

The public housing sector in Australian cities has also undergone major changes since 1991. Successive Federal governments have cut direct funding to State public housing authorities (Badcock, 1999) and placed emphasis on more tenure-neutral housing assistance, particularly extending rent rebates to welfare dependent households in the private rental sector (Yates, 1997). Consequently, little new public housing has been built in Australian cities in recent years. In addition, many of the estates that were built during the 1950s and 1960s have been redeveloped, involving demolition of some of the old rental stock and its replacement by privately owned dwellings (Arthurson, 1998; Randolph and Judd, 2000). As a result the public rental sector has declined in importance in all the major cities and only households in acute housing need now have a realistic prospect of being allocated a public housing property. Low-income households, unable to gain access either to home ownership or to public housing, are becoming increasingly dependent on the private rental sector (Wulff, 1997; Berry, 2003; Yates and Wulff, 2005).

The changes to housing have largely been driven by economic forces and government policies affecting the supply side of the market. There is little sign that fewer households are aspiring to own their own homes. It has simply become financially harder to do so. Similarly, there is scant evidence to suggest that Australians have generally become more enthusiastic about high-density housing (McDonald and Moyle, 1995). Planners have argued for some time that housing demand ought to be changing. Falling fertility rates and the deferral of marriage and childbearing, longer life expectancy, and high rates of marital breakdown have combined to reduce the size of the average Australian household from 3.8 in 1947 to 2.6 in 2001 (Hugo, 2005), and to increase the diversity of household types. Couples with children now make up only a third of the households in Australia's major cities (Forster, 2004, 89). Almost half the households contain only one or two persons, and that proportion is increasing.

It seems logical that smaller households might find apartments in accessible locations more appealing than the conventional separate house in outer suburbia, and major booms in apartment building, particularly in inner Sydney and Melbourne, have been based on this premise (and/or on the attractions of housing as an investment). But these apartments are largely occupied by young adults (Vipond et al., 1998), including niche groups such as overseas students and other temporary migrants (Hugo, 2005). ‘Mainstream’ households still express a long-term preference for low-density housing (Yates, 2001; Troy, 2003). The question, as posed by Davison (1994), is whether planning and housing policies will deny them that preference on both environmental and economic grounds.

Residential differentiation

As recently as the late 1980s the familiar factorial ecology model (Murdie, 1969; Stimson, 1982) still provided a useful starting point for understanding patterns of residential differentiation in Australian cities. Each had sectors of contrasting socio-economic status. Family status, measured by age structure, fertility and household composition, exhibited a concentric pattern, with the proportion of families with young children rising as distance from the city centre increased. Ethnic composition, measured by birthplace, religion, ancestry and language, formed a pattern of clusters (Forster, 1995).

Today this model has palpably exceeded its use-by date. A new, more complex and fine-grained, pattern of differentiation is evolving (Gleeson, 2003; Forster, 2004, Chapter 5; Randolph, 2004). In the inner suburbs, the process of gentrification has become almost complete (Badcock, 1995; 2001), and decades of population decline have been replaced by growth. Even the city cores have experienced startling rates of population increase since 1996, albeit from a very low base, as high rise apartments proliferate in the CBDs and adjacent areas such as Melbourne's Southbank–Docklands and Sydney's Ultimo-Pyrmont. Old industrial buildings have also been reborn as upmarket residential or commercial developments. We have not quite seen ‘the end to inner city poverty’ (Randolph, 2004, 486) but the remaining low-income people are increasingly marginalised, on high rise public housing estates, in boarding houses or charitable institutions, or sleeping rough in parks, vacant buildings and doorways. The rise of the prosperous inner regions is generally associated with the economic restructuring that, as discussed earlier, has produced an increasing concentration of professional, managerial and administrative employment in the inner city and adjacent areas.

A number of writers have gone so far as to suggest a ‘two cities’ thesis: that the core regions, epitomised by Sydney's Global Arc, are increasingly separate entities, distinguished from the rest of their metropolitan area not only by having different employment structures, population compositions and housing markets (O’Connor and Rapson, 2003), but also by having different political and cultural values. They are characterised by a high percentage of residents with tertiary qualifications, only a small proportion of traditional nuclear family households, a tolerance of non-mainstream lifestyles and small-l liberal values. They were the main areas to support Australia becoming a republic at the 1999 referendum, were the areas that most strongly rejected Pauline Hanson's populist right wing One Nation Party in the late 1990s (Stimson, 2001), and contained most of the electorates where the Labor Party and the Australian Green Party did well in the 2004 Federal election. Residents of many middle and outer suburbs, on the other hand, resent the perceived cultural elitism of the urban core regions and often see themselves as victims of globalisation rather than its beneficiaries. They rejected the republic, were tempted by One Nation (Badcock, 1998) and, in 2004, even in traditional Labor heartland seats, swung towards the Liberal-National Party Coalition as well as to the new conservative force of the Family First Party.

The above picture is clearly based on very broad generalisations, particularly about the expanses of middle and outer suburbia where the great majority of urban Australian households continue to live. Up to the mid 1990s, the main trend in residential differentiation and variation in well being was a widening disparity between high income and low-income sectors (Forster, 1991, Murphy and Watson, 1994). But since then the patterns have become more complex. In traditional low-income sectors, such as the western suburbs of Sydney, Melbourne and Adelaide, public housing estates are being broken up by urban regeneration projects (Arthurson, 1998; Randolph and Judd, 2000), while segregated master-planned communities cater nearby for successful ‘aspirational’ households – with the aid of large mortgages (Gleeson, 2003; Gwyther, 2005). As previously discussed, the remaining public housing tenants are increasingly characterised by welfare dependence. Low income households unable to afford entry to the master-planned communities and unable to gain access to the shrinking stock of public rental housing find themselves increasingly restricted to pockets of low quality housing in the private rental sector (Randolph and Holloway, 2005). Across town, the more comfortable higher status middle and outer suburbs have experienced less change, though property values are appreciating more slowly than in the inner suburbs and in some areas, even in some outer suburbs, both population numbers and income levels have fallen as households progress through their life cycles into the ‘empty nest’ stage and beyond into retirement (Forster, 2004, 119).

The geography of disadvantage has therefore become more complex, with once-working class sectors now containing clearly demarcated areas of rich and poor, sometimes separated only by an arterial road or rail line. Gleeson (2002; 2003) fears the increasing disintegration of urban society into self interested ‘Privatopias’ on the one hand and residual enclaves of disadvantage on the other. Healy and Birrell (2003) also claim that there is an ethnic component to this process in western Sydney, since the residents moving to the new master-planned communities are largely Australian-born, English-speaking and Christian (Gwyther, 2005), whereas many of the low income households ‘left behind’ are of non-English-speaking origin.

Latham (2002; 2003) has argued that rising socio-economic diversity within regions like western Sydney is a good thing, breaking down stereotyped working class images and enabling upwardly mobile ‘aspirational’ local residents to remain living in the region. But, in March 2005, a violent confrontation between police and local residents in the public housing area of Macquarie Fields in Sydney's outer south west drew national attention to serious issues of concentrated disadvantage and alienation in the increasingly marginalised low-income suburbs. Dodson (2005) makes the familiar point (Hamnett, 1979) that urban planning policies can do little to address issues of poverty that have deeper social and economic causes. But the consequences of the changing spatial patterns of poverty and inequality – some the result of past housing and planning policies – still need to be better understood than they are at present.

Metropolitan planning strategy

The planning authorities responsible for the future development of Australia's five major cities have all recently published new strategic plans or draft plans (Government of South Australia, 2005; New South Wales Department of Infrastructure, Planning and Natural Resources, 2004; Queensland Government, Office of Urban Management, 2004; Victoria, Department of Infrastructure, 2002; Western Australian Planning Commission, 2004). The cities, though they have different growth rates and economic prospects, are all growing more slowly than during the boom years of the 1950s and 1960s, and this is especially the case in Adelaide, Melbourne and Sydney. But, because households are getting smaller, household formation rates – the driving force behind housing demand in each city – remain high (Hugo, 2005). The strategic plans therefore anticipate that, over the next 20–25 years, each city will need to accommodate significant numbers of additional dwellings (Table 1).

Table 1.  Anticipated need for additional dwellings in the next 20–25 years, Australian major metropolitan regions.
Metropolitan RegionNumber of additional dwellingsTime period
  1. Sources: Government of South Australia, 2005; New South Wales Department of Infrastructure, Planning and Natural Resources, 2004; Queensland Government, Office of Urban Management, 2004; Victoria, Department of Infrastructure, 2002; Western Australian Planning Commission, 2004.

Melbourne600 000by 2030
Sydney including Central Coast550 000by 2026
South East Queensland550 000by 2026
Perth-Peel375 000by 2031
Adelaide69–137 000by 2031

The planning strategies all contain a commitment to the ‘triple bottom line’ of economic competitiveness, social justice and environmental sustainability. The introduction to the Draft South East Queensland Regional Plan, for example, refers to ensuring ‘that the region can grow and change in a sustainable way, generating prosperity while at the same time maintaining and enhancing quality of life and providing high levels of environmental management’ (Queensland Government, Office of Urban Management, 2004, 5). And the approaches to shaping future growth, though by no means identical, are similar enough across the five metropolitan areas to suggest a consensus view on planning policy, built on the three principles of containment, consolidation and centres.

None of these principles is new. The centres concept builds on planning ideas dating back 50 years, particularly in Sydney and Melbourne (Logan, 1986). It involves encouraging the development of a multi-nuclear metropolitan structure with suburban retail and office employment concentrated into a limited number of major nodes, at locations well served by public transport. These nodes are then intended to serve as foci for high and medium density housing. Urban containment and consolidation, as noted earlier, have been promoted by State governments since the 1980s. Initially the main rationale was economic efficiency: avoiding expenditure on new infrastructure and reducing transport costs. But, during the last ten years consolidation has been seen increasingly as the key to more environmentally sustainable cities, through producing higher densities more favourable to public transport than the private car, through developing smaller dwellings and blocks that use less water and power, and through reducing the impact of urban expansion on surrounding ecosystems.

The merits of urban consolidation have been debated extensively over the past decade (for a summary see Buxton and Tieman, 2005). Academic opponents, notably Troy (1996), have queried both the desirability of consolidation and the feasibility of bringing about significant reductions in the rate of suburban expansion, while urban activist groups, led by the Melbourne organisation Save Our Suburbs (Lewis, 1999), have opposed its impact on existing suburbia, and housing industry leaders (Day, 2005) blame consolidation policies for rising house prices. But, since 1991, policies explicitly or implicitly favouring consolidation have been responsible for a significant rise in all five major cities in the number of new dwellings that are not conventional separate houses (Forster, 2004, 79–85).

The latest strategic plans reaffirm these approaches. The Melbourne, Adelaide and Perth plans incorporate formal urban growth boundaries that set limits to further development at the urban fringe, and all the plans envisage that new housing developments on the fringe will make up less than half the needed net growth in dwelling numbers. For example, the Melbourne 2030 plan (Victoria, Department of Infrastructure, 2002), anticipates that, of the 620 000 additional dwellings that will be needed by 2030, only 31% will be in new greenfield developments. Of the remainder, 28% will be housed in infill developments within existing suburbs and 41% are expected to live in medium and high density ‘strategic redevelopment areas’ associated with major activity centres and containing affordable housing. ‘A significant proportion of new development … must be affordable for households on low to moderate incomes ...’ (Victoria, Department of Infrastructure, 2002, 107). Planned suburban centres are a feature of all the planning strategies, intended to reduce automobile dependence by concentrating employment and services into nodes and corridors served by public transport and containing significant numbers of new dwellings at medium or high densities.

Policy responses to concentrations of inequality and disadvantage have focused in recent years on urban regeneration and social inclusion. Public housing estates such as ‘The Parks’ in Adelaide's west are being redeveloped with the aim of reducing the proportion of public rental dwellings and therefore developing a more diverse social mix and lower levels of welfare dependence (Arthurson, 2002; Randolph and Judd, 2000). Community development programs then seek to develop social inclusion (Lilley, 2005) and build social capital (Winter, 2000). At the same time, improvements to housing stock and infrastructure are expected, with the aid of clever marketing, to develop a more positive image. The new metropolitan planning strategies, particularly in the slower-growing cities (see for example Victoria, Department of Infrastructure, 2002, 108 and Government of South Australia, 2005, 74) incorporate and reaffirm the urban regeneration approach.

Urban regeneration and social inclusion draw much of their rhetoric from the policies pursued by the Blair government's Social Exclusion Unit in Britain (UK Cabinet Office, Social Exclusion Unit, 2000; Lilley, 2005). The approach aims to break down the stigmatisation that locks residents out of employment opportunities and inhibits full participation in society (Taylor, 1998). It will certainly succeed in diluting the largest concentrations of disadvantage. But this will be at the cost of breaking up established communities, reducing the overall stock of public housing and further marginalising the remaining public-housing tenants. Gleeson and Randolph (2002) and Randolph and Holloway (2005) also point out that social inclusion and regeneration programs neglect the private rental sector. In fact, the programs are likely to increase the dependence of low-income households on private rental housing.

Conclusions: parallel universes?

Current metropolitan strategy is clearly based on a particular vision of future urban structure. In 20–30 years time, if the plans come to fruition, our major cities will be characterised by limited suburban expansion, a strong multi-nuclear structure with high density housing around centres and transport corridors, and infill and densification throughout the current inner and middle suburbs. Residents will live closer to their work in largely self-contained suburban labour sheds, and will inhabit smaller, more energy-efficient and water-efficient houses. The percentage of trips using public transport, walking or cycling will have doubled. Regeneration programs will have broken up large concentrations of disadvantage, and a diminished public housing sector will only house welfare-dependent households in acute need. Other low-income households will be able to find affordable dwellings in accessible locations within consolidation developments.

How likely is this scenario, given what recent research has revealed about the changing structure of Australian cities? A succession of critics (O’Connor, 2003; Searle, 2004a; Birrell et al., 2005; Buxton and Tieman, 2005; Cox 2005; Recsei 2005a; 2005b) have queried both the feasibility and the desirability of current planning strategy. They argue that the imposition of urban growth boundaries and similar restrictions will further reduce the affordability of conventional housing, while forcing increasing numbers of households into higher density dwellings that they show little sign of actually desiring. Birrell et al. (2005) title their critique of the Melbourne 2030 plan Planning Rhetoric Versus Urban Reality and go so far as to assert that ‘the UGB [urban growth boundary] is really about pushing the less well-off into high density housing to satisfy the design and ideological objectives of its bureaucratic and academic supporters’ (Birrell et al., 2005, 3–17). O’Connor and Rapson (2003, 41) describe current policy thinking about centres as ‘operating in a world of wishful thinking ... remote from the real world priorities of employers and their employees’. Fagan and Dowling (2005, 79) categorise the notion of self-contained suburban journey to work regions in Sydney as an unrealistic ‘neoliberal vision’.

The above views begin to suggest the existence of parallel urban universes: one occupied by metropolitan planning authorities and their containment – consolidation – centres consensus; the other by the realities of the increasingly complex, dispersed, residentially differentiated suburban metropolitan areas most Australians live in. Recent metropolitan planning strategies have been seen as attempts to reassert a role for urban policy after a decade dominated by neoliberalism and the pre-eminence of market forces (McGuirk and O’Neill, 2002; McGuirk, 2005). Compared with the planning documents of the early 1990s the strategies emphasise the importance of environmental sustainability (Gleeson et al., 2004; Searle, 2004b), and there is no doubting the need for effective public policy in that area. Newman (2005) defends the containment – consolidation – centres approach as being critical to the future environmental, social and economic sustainability of our major cities, and criticises Recsei's (2005a) arguments in particular as being based on a fear of change. But, given what research is telling us about how the structures of Australian cities are actually changing, there are two key concerns about the direction of the current strategy consensus.

The first concern is the continued adherence to the concept of neatly structured suburban development organised around centres and self-contained urban realms. The hope is that such a structure, together with higher residential densities, will deliver significantly lower levels of dependence on the automobile and therefore greater environmental sustainability. But the research reviewed earlier suggests that patterns of economic development, labour force participation and journeys to work continue to grow in complexity and are unlikely to fit this neat pattern. Levels of car use – at least for the journey to work – may have stabilised, but they will remain high even if the planning targets for increased use of alternative modes are met. Policies such as encouraging smaller, cleaner and more fuel-efficient vehicles are needed to deal directly with this, rather than simply hoping that consolidation and centres will make the problem go away.

The second key concern is that the new metropolitan strategies may recognise the issues of housing affordability and social exclusion, but they all fall well short of addressing those issues in a convincing manner. For example the Melbourne 2030 strategy (Victoria, Department of Infrastructure, 2002, 107–108) recognises that public housing will in future be restricted to households in acute need, and that ‘the supply of affordable housing in all parts of the metropolitan area will need to be increased’. But the strategy then admits that ‘the planning system is not well equipped’ to meet this need and simply suggests that ways of addressing it ‘will be explored’. Given that all the strategies are based on restricting suburban expansion and therefore inevitably driving up the cost of land, the issues of housing costs for lower-income households and the emergence of new concentrations of disadvantage cannot simply be left to chance.

Overall, the metropolitan planning strategies suggest an inflexible, over-neat vision for the future that, however well-intended, sits dangerously at odds with the picture of increasing geographical complexity that emerges clearly from recent research on the changing internal structure of Australian cities since the early 1990s.