Abstract
Many suburban areas in Australian cities built between the 1930s and 1960s are facing major challenges from socio-demographic change, ageing housing, long-term disinvestment and entrenched pockets of social disadvantage. Yet despite emerging relative disadvantages, these middle ring or ‘third city’ suburbs are experiencing locally generated piecemeal market-led reinvestment and renewal. This paper explores housing, household and housing investment characteristics and trends in the older parts of Sydney’s western suburbs drawing on analyses of socio-demographic and development application data. Policy options to achieve more appropriate urban renewal outcomes under broad social sustainability criteria are canvassed. Similarities and differences to the American ‘first suburb’ phenomenon are noted.
Introduction
Until recently, the simplistic dichotomy of a dense, mixed, historically struggling but resurgent inner city versus a more homogeneously affluent low-density residential zone has pervaded many cultural critiques, planning theories and urban policy related to suburban development. The reality, of course, is not, and indeed never has been, so neat. Suburban diversity has become the new orthodoxy (Nicolaides and Wiese, 2006). Until the 1970s, the residential differentiation of many Western cities when factorially analysed had crystallised broadly into a geometry of socioeconomic sectors, socio-demographic crescents and ethnic clusters. A more complexly patterned polyglot residential mosaic has since emerged, fragmented along class, geography, amenity and multicultural lines. Cyclical waves of investment, disinvestment and reinvestment have been expressed through processes of greenfield development, ageing and decay, and brownfield redevelopment and urban renewal, with a pervasive drift to densification.
The evolving geography of Australian cities parallels broader Western experience and models (Forster, 2004). Perhaps more than any other country, Australia is seen as a nation of suburbs (Davison, 1995; Forster, 2004). The traditional expansion of the city through low-density single house lot sub-division on the urban fringe, predominantly for homeownership, was the standard model of middle-class suburban growth in Australian cities for much of the 20th century (Troy, 2004). The principal housing form of suburban Australia has been the owner-occupied detached house (Davison et al., 1995; Badcock and Beer, 2000). Even at the latest 2006 census, 75 per cent of the national population still resided in a detached house, the vast majority located in the suburbs of the largest cities. As is happening globally, that picture is changing and becoming much more complex as suburban development becomes more differentiated, diverse and polarised. The old factorial ecology has been ruptured although not erased by new forces of residential mobility, renewal, changing household structures and globalisation (Flood, 2000). There is evidence of both change and continuity across Australian cities that have been characterised as defining a new era of post-suburbanity (Anderson et al., 2006).
In the Australian context, within these new patterns of social differentiation, the band of ageing suburbs built between the 1940s and 1960s now seems recognisably distinctive in housing and population terms with mixed evidence of both decline and renewal. These are the extensive tracts located between the increasingly gentrified inner city on the one hand and, on the other, the more modern post-1970s outer suburbs. They loom as key areas for understanding the economic and social dynamics of broader metropolitan change, yet have received comparatively little academic or policy attention (Ruming et al., 2010).
This paper uses the notion of ‘third-city suburbs’ to explore these middle suburbs. The term is adapted from Latham (2003) who caricatured Sydney as having three distinctive ‘arcs’. First was the ‘global’ arc defining international Sydney—the central areas, the high-density inner suburbs, the high-amenity beachside suburbs, the international airport and the major concentration of producer services employment. The second was the ‘outer’ arc of mainly new housing developments with a socioeconomic profile of relative affluence, high rates of home purchase, low unemployment, higher education levels, car dependence and single-family dwellings. The third major and intermediate zone was the ‘middle’ arc of older suburbs mainly in western and south-western Sydney with a predominance of working-class and ethnic communities, and facing issues of high unemployment and crime. The notion has intuitive appeal and a resonance in other Australian cities (Forster, 2004). It suggests the idea of a ‘third city’ of ageing low-density suburbs differentiated from the ‘first city’ comprising the higher-value and heterogeneous inner zone and the ‘second city’ denoting relatively recent fringe suburban growth accommodating an upwardly mobile suburban population (Randolph, 2002).
Our concern is with the social differentiation and physical renewal of middle-ring suburbs. There are four major objectives. The first objective is to identify the distinctiveness of Australia’s middle suburbs from the literature in terms of socioeconomic status, housing quality and processes of renewal. The second objective is to construct a socio-demographic profile of these places in Sydney using census data and explore the evidence for them as places of increasing social and economic disadvantage. This shifts the focus towards western Sydney which sits outside the glitter of the global arc (Mee, 2002). The third objective is to recover evidence of the physical changes to the housing stock through reinvestment that are reshaping some of these areas and which may be acting to counter the stereotype of disadvantage. The fourth objective is to consider the broader planning and housing policy implications arising from these largely unplanned renewal processes. The structure of the paper follows this sequence.
The paper seeks to convey the overall heterogeneity of social and physical change in Sydney’s middle suburbs as illustrative of broader processes evident in Australian cities. The evidence pulls in two directions. One is towards a picture of relative socioeconomic disadvantage compared with metropolitan norms. Yet complicating the picture of universal decline is evidence of significant private reinvestment in the housing market. This latter process points to positive resilience within what otherwise present as less affluent suburban communities, but its piecemeal nature may impede more strategic opportunities for co-ordinated renewal. Our analysis of these twin-edged phenomena concludes by considering the similarities and differences between the Australian ‘third city’ phenomenon and the seemingly comparable benchmark of the struggling American ‘first suburbs’.
Australian Middle-ring Suburbs
The middle-ring suburbs are predominantly low to moderate income suburbs created between the late 1930s and the late 1960s, and mostly through the post-Second-World-War housing boom. They came to represent the Australian housing dream, typified by a single-family house on the iconic ‘quarter acre block’. However, this role has changed significantly in the intervening years. Recognition that these suburbs constituted a distinctive suburban zone dates from the 1980s when their imminent social trajectory was noted by Maher (1982, p. 18) who identified the former suburban growth frontier as “showing signs of maturity”. Symptoms of decline have become more evident in the past decade. Forster (2004) depicts these locations as a belt ‘strongly affected by ageing in place’ and suffering significant population loss in some locations. Darcy (2000) describes the middle-ring suburbs as feeling the brunt of selective urban consolidation policies favouring new medium-density housing, a process which has become highly contested in more affluent localities (Lewis, 1999).
Gleeson (2006) caricatured these in-between suburbs as the ‘heartlands’ of Australia experiencing a variety of stresses. They had become the netherworlds of the private rental market, studded with decrepit housing stock and generally wearing the mantle of public neglect (Gleeson, 2006, p. 46).
Newton (2010) refers to these regions as the ‘greyfields’ of Australian metropolitan areas, a usage which can be distinguished from the new urbanist appellation applied more specifically to the extensive and wasteful car parks attached to suburban retail malls (Calthorpe and Fulton, 2001). He also depicts these swathes of greyfield as sliding into physical, technological and environmental obsolescence.
Suggesting direct parallels to the challenges faced by American ‘first suburbs’ (Hanlon et al., 2010), Stilwell (2005) comments upon middle-ring suburbs as being ‘in limbo’ and contributing to a long list of other contemporary urban problems. He characterises three sets of issues: infrastructure stress on ageing schools, hospitals, water and sewerage systems; a local rates base shrinking relative to more affluent localities; and a failure to attract gentrifiers and other affluent newcomers. He notes that these suburbs “have not experienced the dramatic surges in property price inflation that have occurred in the gentrifying suburbs in the inner city areas” and hence the ability of local governments in these areas “to effectively manage the maintenance of high residential amenity has become increasingly problematic” (Stilwell, 2005, p. 6).
Evidence for a suburbanisation of disadvantage, characterised by lower socioeconomic status, overrepresentation of public and private tenants rather than homeowners, migrant communities, above average unemployment, lower housing values, emerging problems of obsolescence of the housing stock and poorer standards of local services has been assembled by several writers including Randolph and Holloway (2005) and Murphy and Watson (1994). In characterising ageing middle suburbs as Australia’s ‘third city suburbs’, the contention is that they have fallen behind both the gentrified inner city and newer outer-city fringe developments to become exemplars of social polarisation (Randolph, 2002; Randolph and Holloway, 2004). These outcomes have been driven by a confluence of impacts of macroeconomic restructuring, generational change, new migration policies and other social changes. Older Anglo-Australian residents have been replaced by a mixed new population, some relatively poorer households, often young and new migrants, as well as second-generation immigrants. In some low amenity zones, the ageing post-war housing stock has become ‘residualised’, in the sense of being characterised by increasing concentrations of disadvantaged persons such as the unemployed or low-income earners (Healey and Birrell, 2006). This denting of perceptions of suburban affluence was eloquently pinpointed early by Stretton when he characterised cities acting as both physical and psychological devices for quietly shifting resources from poorer to richer, and for excusing or concealing—with a baffled but complacent air—the increasing deprivation of the poor (Stretton, 1970, p. 310).
However, the picture is not all gloom and doom. The classic filtering processes of the housing chain suggest deterioration to the ultimate result of abandonment and demolition (Bier, 2001). Yet in Australian cities there has not been the extensive abandonment of inner or even middle-ring areas that has so damaged American cities. Indeed, the housing market in these third-city suburbs has responded to a range of demand and supply stimuli in recent years, partly assisted by a series of state-government-driven strategic planning polices aimed to encourage higher-density housing. From the evidence of on-going studies (for example, Randolph, 2002; Pinnegar et al., 2010; Ruming et al., 2010), at least six processes of physical renewal can be identified since the 1980s
— Additions and alterations (‘alts and adds’) to existing property, frequently as first-storey additions and extending the rear to accommodate larger living spaces.
— Spot redevelopment in an ad hoc and unplanned way, often as two-storey and bigger houses replacing older houses—this is the ‘knockdown rebuild’ market.
— Dual-occupancy redevelopments, where an additional dwelling is constructed on a single-house plot.
— Redevelopment of single houses as multi-unit villa and townhouse developments, where residential zoning or spot rezoning permits.
— Redevelopment of older houses into higher-density flats, often on larger plots rezoned specifically to permit such development—the so-called gun-barrel form stretching back from the street frontage on single blocks.
— More comprehensive high density (above four storeys) and mixed use renewal near town centres and train stations encouraged by recent urban consolidation planning policies.
These interventions result in a ‘salt-and-pepper’ renewal approach with poor admixture of older fibro-asbestos or weatherboard houses and newer outsized ‘McMansions’ and intensive medium-density redevelopments on single-house lots. The resulting urban form is often a disorganised mix of redevelopment that is poorly integrated in terms of street design, with low-quality visual amenity and open space provision. Usually there has been little consideration of the longer-term needs of the local community.
At the same time, this activity suggests that, rather than widespread disinvestment and social decline, parts of these middle suburbs have been able to retain and attract a level of reinvestment from a range of directions. The high proportion of first- and, increasingly, second-generation immigrant households also points to the role of these suburbs as areas of developing cultural significance. They are places where new communities can gain a foothold and flourish. Without doubt, some of those who prosper eventually join the move to the new fringe suburbs. Yet increasingly, there is an attraction to the areas they or their parents first moved to. This implies that, rather than completely falling apart, middle-suburb Australian communities have held together through cultural, religious and ethnic networks, forming a range of cohesive communities, rather than retrogressing into socially excluded and disadvantaged spaces. While there is some anecdotal evidence that this is the case, a gap still exists in our knowledge of the precise drivers of this process (Burnley, 1999). The analysis reported in the next two sections begins to unpack our understanding of the complex processes at work with special reference to metropolitan Sydney.
Third-city Suburbs in Western Sydney
Sydney with a population of 4.3 million is Australia’s largest city—a beta global city with complex metropolitan governance. The state government of New South Wales is essentially the metropolitan government, being responsible for all major policies crucial to urban development. The state also provides the majority of education, health, police, transport and recreational services through an array of government departments, statutory authorities, trading enterprises and public–private partnerships. Like most Australian metropolitan areas, Sydney is further fragmented into a patchwork of local government areas, over 40 of them, which provide services such as road and parks maintenance, garbage collection, cultural facilities, building and town planning controls through revenue raised via property taxation and assistance from the state (Forster, 2004; Kübler and Randolph, 2008).
The second objective of the paper is to try and discern evidence of a geography of disadvantage correlated with Sydney’s middle suburbs. The Sydney metropolitan area has certainly become increasingly differentiated socioeconomically as a result of the impacts of broader labour market restructuring, welfare reforms, demographic change and migration played out spatially within an essentially market-driven housing system. In general social terms, the northern, eastern and inner suburbs, together with waterside locations, have pulled away from the middle-ring western suburbs. Within these general trends, the suburbs themselves have begun to diverge significantly. Figure 1 conveys the spatial expression of a ‘third city’ with the middle suburban areas of western Sydney having markedly lower household incomes than the northern and eastern suburbs and also lower incomes than many of the newer suburbs on the fringe. Sydney is a city of social extremes; in between are the middle suburbs subject to what Baum (2008, p. 15) describes as “a range of fortunes”.
Median household income by suburb, Sydney 2006.
To construct a more detailed profile of the housing and household characteristics of Sydney’s middle suburbs, special tabulations from the 2006 national census of the Australian Bureau of Statistics (ABS) were commissioned for the 11 local government areas defining western Sydney (Figure 2). The spatial scale of analysis followed the suburb definitions which the ABS classifies as being in ‘urban Sydney’. Each of these suburbs was characterised by its age of development with assistance from senior town planners from the constituent councils (Figure 3). Five distinctive sub-regions were identified: the suburbs developed prior to 1945, between 1945 and 1970, between 1970 and 1980, between 1980 and 1996, and after 1996. The main focus of discussion is on the first two sub-regions.
Local government areas comprising western Sydney.
Western Sydney suburbs, by age of development.
An immediate impression from the mapping exercise is that, while there is a degree of contiguousness in the distribution of suburbs by age, the unique geography of the city and the periodicity of development has resulted in a patchwork pattern of development, with ‘middle’ suburban suburbs located both in the central western Sydney area as well as in the two urban ‘arms’ to the west and south-west. The particular spatial configuration of Sydney and its direction of suburban expansion mean that there is no clear concentric pattern to the location of suburbs by age of development.
At the time of the 2006 census, Western Sydney was home to some 40 per cent of Sydney’s metropolitan population, a total of about 505 000 households. Tables 1 and 2 summarise the main housing, household and person characteristics of the sub-regions developed before 1945 and from 1945 to 1970 relative to the broader regional and metropolitan context. Together, these two sub-groups of suburbs account for around 60 per cent of the region’s households and population. The detached bungalow is the dominant housing typology, significantly above the metropolitan norm for the 1945–70 suburbs. There is also a considerable stock of public housing, often developed on broadacre estates (Table 1). The socioeconomic indicators in Table 2 reveal the areas developed before 1970 as generally characterised by higher unemployment, lower educational qualifications, larger number of single-parent families and lower incomes than western Sydney and the metropolis as a whole. Some of the key characteristics are elaborated next.
Housing characteristics of selected suburbs in western Sydney, 2006
Source: ABS, 2006 census.
Household and person characteristics of selected suburbs in western Sydney, 2006
Source: ABS, 2006 census.
Suburbs Developed before 1945
These are the oldest suburbs in the region. With their origins in sub-divisions created before the Second World War, they comprise most of the older town centres in the region and their surrounding areas. They have been subject to subsequent processes of infill, renewal and redevelopment, including ‘gun-barrel’ redevelopment of larger single-family housing blocks into three-storey walk-up flats. In many ways, these areas are the most physically and socially diverse in western Sydney, but it is here that characteristics of urban stress and decline associated with the notion of the ‘third city’ are most pronounced. The population is the oldest; the proportions of childless couples, lone parents and divorced or separated people are above the average for the region; and the 23 per cent single-adult households is the highest proportion in western Sydney. These suburbs also contain the most ethnically diverse population in the region, with over half of all residents born overseas. All the immigrant groups are represented, with high proportions of Europeans (overwhelmingly from southern or eastern Europe), the Middle East and, in particular, Asians. In many ways, this group of older suburbs is the core of multicultural Sydney (Collins and Poynting, 2000).
In economic terms, these suburbs are among the poorest in the region, with 24 per cent of households with incomes under $500 (AUD) per week and only 7 per cent earning over $2000. These numbers reflect the high proportion of single-adult and older households, low labour force participation rate and chronically high unemployment rate. A significant proportion of households (17 per cent) do not have access to a motor vehicle and there are relatively few households with three or more vehicles. With 28 per cent living in flats and another 12 per cent in other forms of attached housing, the level of higher-density housing is one of the key features that differentiate these suburbs from the rest of the region. Much of this higher-density stock is found around the local town centres and along the main transport routes, which partly accounts for the higher proportions of households without cars, although this feature has much more to do with low income than lifestyle choices (Bunker et al., 2002). Associated with the high proportion of flats is a very high proportion of private rental—by far the highest proportion of all five suburb groups (25 per cent). One-third of households (33 per cent) are homebuyers, while a similar proportion own outright (34 per cent)—both figures below the metropolitan norm (Table 1). The correlation of substantial higher density with a low-value private rental market helps to explain these suburbs’ relatively high mobility and their heterogeneous population.
The highly diverse population, in contrast to the newer suburbs, is a reflection of a diverse housing market. The sizeable proportion of single-person households reflects the greater variety of low-cost housing in the area that is not available in the new suburbs or elsewhere in Sydney. It is in these suburbs that the bottom rungs of the Sydney housing market are found (Randolph and Holloway, 2005).
Suburbs Developed between 1945and 1970
Economically, the 1950s gradually restored material prosperity to a city buffeted by depression and war. This was the so-called post-war long boom (Bell and Quiggin, 2008). Spatially, the greatest outward push in Sydney was onto the good building land of the Cumberland plain. This period is widely seen as bequeathing the low-density suburban landscape “characterised by outward social homogeneity and a rough uniformity in the style of houses” (Gleeson, 2006, p. 17). The early post-war period was a time of high demand and housing shortages caused by labour and material constraints. ‘DIY’ or self-build became central to housing production and owner-building accounted for a significant proportion of new house building throughout the 1950s (Dingle, 2000; Holland, 1988). A cheap and popular building material was fibrous cement panels which created a distinctive ‘fibro belt’ across large swathes of the now-middle-suburban landscape from Bankstown out to Fairfield and Blacktown (Pickett, 1997; Pinnegar et al., 2010). Between 1947 and 1961, nearly 120 000 fibro homes were added to Sydney’s housing stock (Spearritt and DeMarco, 1988).
These, then, are the classic post-war ‘struggle street’ suburbs of Sydney’s ‘westie’ population (Powell, 1993). In demographic terms, the original residents of these areas are ageing: one-fifth of the population in these suburbs is aged 55 years or older, while the proportion of couples with children has fallen to just two in five (41 per cent) and lone-person households have increased to 20 per cent, reflecting an ageing population. At the same time, the proportion of single-parent families is the highest in western Sydney at 16 per cent (Table 2). Incomes are significantly lower than in the newer suburbs farther out, with only 9 per cent of households earning over $2000 per week and almost a quarter (23 per cent) on incomes below $500. Reflecting this, the proportion of households with no vehicle jumps to 13 per cent, compared with the metropolitan norm of 11.7 per cent (Table 2). Residents in this sub-region are the least likely to have a university qualification. These suburbs contain some of the largest public housing estates in Sydney, built in the 1950s and 1960s at a time of rapid suburban expansion (Spearritt, 2000). Together with a growing private rental market, the result is that 28 per cent of households are renters. In addition, over a fifth of dwellings are medium-density forms—semi-detached houses, blocks of flats and townhouses—evidence both of the prevalence of denser public housing as well as a private market process of higher-density infill and renewal. These suburbs clearly show signs of change, with the first wave of residents being replaced by an increasingly diverse population. Renewal through higher density is occurring. The population is both less affluent than newer and inner-city suburbs and less homogeneous.
Differentiating the Third City
This empirical analysis confirms the notion of a core of struggling suburbs in Sydney’s middle western suburbs. Horvath et al. (1989, p. 88) have noted that, apart from isolated pockets of sub-standard housing, “perhaps the most striking feature about relatively poor Sydney. . . is the absence of extensive residential slums anywhere in the metropolis”. However, there are tracts of extensive 1960s public housing estates to the west and south-west as well as cheaper private housing areas where new migrants and other less affluent sectors of the population have found more affordable homes. These areas stand in stark contrast to the more privileged communities in the gentrifying inner suburbs, the northern ‘north shore’ and beachside suburbs with an historically enduring affluence since the early 20th century, and the newer outer suburbs. The profile of the new post-1996 suburbs predominantly in the north-west of the region is one of relative wealth with almost one in three households earning in excess of $2000 per week, very high rates of home purchase, low unemployment and higher education levels. These indicators capture the ‘aspirational’ cohort for whom up-scale fringe suburbia represents the expression of social achievement (Latham, 2003; Gleeson, 2006).
The suburbs developed before 1970 clearly share less privileged population profiles (Table 2). They are characterised by high levels of social diversity, with the highest proportions of the most disadvantaged people in the region, as measured by household income and occupational characteristics. The defining characteristic is, in many ways, the income polarisation between the older and newer suburbs. It is the absence of higher-income households that largely defines these older suburbs, together with much more heterogeneous housing markets. The higher social diversity here is in great part linked to this much greater variety in housing market characteristics. It also reflects the fact that several waves of renewal, in-migration and change have moved through these older suburbs as the housing stock has been redeveloped or reused. The result is a mix of newer migrant communities among what remains of the older Australian community that moved in when these areas were originally developed.
They remain vulnerable communities. Dodson and Sipe’s (2008) correlation of household socioeconomic status with indicators of car dependence in Sydney shows that oil vulnerability is not only highly spatially differentiated but also a socially regressive phenomenon. While households in the inner and central city exhibit low levels of exposure to the impacts of declining global energy security, the situation deteriorates towards the western suburbs. It is the relatively less affluent households in Sydney’s ‘third city’ that are the most vulnerable to the adverse consequences of higher global petroleum prices. Since 2008, there has also been susceptibility in these older suburbs to the global financial crisis. While Australia’s housing finance system is more structurally robust than the US system (Stapledon, 2009), recent evidence suggests that ‘mortgage stress’ has had its major impact in Australian cities on areas of lower socioeconomic status, paralleling the incidence of rental unaffordability in these areas (Randolph and Holloway, 2002; Berry et al., 2011).
Reinvesting in the Middle Suburbs
While the aggregate indicators of socioeconomic, demographic and household status described in the previous section point to suburban enclaves ‘doing it tough’ compared with areas both closer in to the CBD and the new suburban frontier of up-market project homes and master planned communities, there is demonstrable physical evidence that these areas still retain a certain resilient market appeal. The third objective of this paper—to recover evidence of on-going physical changes to the housing stock—was pursued through a case study of an archetypal slice of middle western Sydney: the local government area of Bankstown. Here we draw on data retrieved from the local council pertaining to residential development applications since 2000.
Bankstown’s current population is approximately 185 000 people. While opened up originally for suburban settlement in the earlier part of the 20th century with the rail line to central Sydney, Bankstown was predominantly semi-rural until the post-war years (Neutze, 1974). Public investment was limited and consequently the area has suffered from social and physical infrastructure deficits compared with the inner and northern suburb areas of Sydney. Its core housing stock is low-density, asbestos fibro, single-family houses on moderate to large plots. A second wave of higher-density renewal became a dominant trend from the 1960s onwards when newly introduced strata title legislation facilitated the development of blocks of flats for sale to investors (Spearritt, 2000). These were largely developed in and around town centres or near rail lines, aided by zoning ordinances. Large numbers of these blocks, typically three storeys and usually each comprising 6 to 12 flats, were constructed in the 1960s and 1970s where block sizes permitted, and they continue to be a focus of local planning in the region today driven by state-government-instigated urban consolidation policies (Bunker et al., 2006). As already noted, the medium-density housing stock is associated with some of the most disadvantaged households and Bankstown has a significant immigrant community. Yet other indicators point to a resilient suburban housing market.
Evidence as to the level of renewal activity is shown in Figure 4 which plots the location of the 414 development applications (DAs) approved by Bankstown City Council for the 2000–05 period that resulted in new dwellings. This shows the DAs disaggregated both by dwelling type and number and the location in relation to the local land use zoning scheme. A dispersion of activity across the council area is apparent. Single-house and dual occupancy redevelopments are scattered across the residential area, as are, to a lesser extent, villa and townhouse developments. Flats and larger townhouse developments are confined to the localities zoned for higher density around Bankstown town centre near the railway station. Altogether, these DAs resulted in 174 single-house redevelopments (known as ‘knock-down rebuild’) and 6 new houses on infill sites, with an additional 210 dual occupancy developments, 23 applications for townhouse redevelopments and 15 for blocks of flats. In addition, a further 157 DAs were submitted for major additions and improvements to existing houses (not plotted on Figure 4). The contrast between the old and new streetscape of Bankstown is apparent from Figure 5. There is evidence here of what has been termed the ‘McMansionisation’ phenomenon, with larger, predominantly two-storey homes in a contagious spot redevelopment process replacing the original, smaller, single-storey houses (Wiesel et al., 2011). As in the US, this selective infill raises issues of overconsumption and design compatibility (Nasar et al., 2007).
New dwelling development approvals, Bankstown, 2000–05.
The old and the new in Bankstown.
Additional data have been collected specifically on the continuing volume of ‘knockdown rebuild’ activity in Bankstown. Between 2004 and 2008, a total of 627 DAs was submitted, representing some 1.4 per cent of the total housing stock. Collected as part of a wider survey of this phenomenon across Sydney, the number was the highest figure for 29 local government areas. At the same time, the median household weekly income at the 2006 census was, at $926, the fourth-lowest in the metropolitan area (Wiesel et al., 2011). These analyses thus uncover a significant amount of activity for a seemingly disadvantaged and low-value suburban area.
What accounts for this wave of reinvestment in what is a relatively devalued housing market, at least compared with values elsewhere in the city? Preliminary research points to demand being driven by upwardly mobile households choosing to buy or redevelop new houses within the suburbs where they have previously lived or choosing to stay in their properties and undertake major renovations or rebuilds. There is evidence of a similar phenomenon in American cities which has been dubbed ‘suburban gentrification’ (Charles, 2011). However, the Sydney evidence suggests that this form of suburban renewal is primarily a locally generated process rather than invasion and succession from other parts of the city or social groups as in a classic inner-city gentrifying market. This form of housing sub-market challenges traditional concepts of centrifugal migration to the fringe for upwardly mobile households and evidence of ‘White flight’ from the ‘middle arc’ (Gwyther, 2004).
While there has no doubt been a process of relative devalorisation in the ‘third-city’ suburbs in Australian cities since the Second World War, suburban renewal suggests that a process of revalorisation may well be acting to achieve more cohesive social outcomes, sustain local economic performance through retaining the household income and tax-base, as well as generating local employment. While certainly not a panacea for the social issues facing these areas, it is nevertheless an indication that the socioeconomic processes at work compared with other areas of the city are much more complex and resilient than might first be appreciated. Yet larger urban policy challenges need to be confronted.
Policy Responses and Options
This penultimate section of the paper considers our fourth objective to canvass the policy issues arising from diffuse renewal processes in the less affluent suburbs. Newton (2010, p. 100) argues that the redevelopment of the middle suburbs is “one of the principal challenges for twenty first century urban planning”. Based on the evidence in the classic middle-ring suburb of Bankstown—the fourth-largest local government area by population in Sydney—housing markets are responding to the need for upgrading and renewal, albeit in a different form and scale from that which has occurred in the gentrification of inner-city areas (Salmon, 2006). However, renewal processes are largely occurring spontaneously, unco-ordinated and in a policy vacuum, thereby presenting a set of interrelated dilemmas. These challenges include
— replacement or upgrading of the ageing housing stock without displacing those in most need;
— integration of land use, social and transport planning;
— securing more effective urban design (streetscape) outcomes, improved local amenities and public realm solutions;
— reduction in private open space through oversized infill development;
— scatteration of housing reinvestment decisions;
— fragmented land holdings mitigating more comprehensive precinct renewal;
— high redevelopment costs producing dwellings for which there is little market demand; and
— lack of interest by major developers.
To date, areas of private-sector disadvantage have been left largely to the vagaries of the market to sort out. At the same time, as part of its compact city planning philosophy, the state government has set ambitious targets for new housing in these areas. The latest metropolitan strategy for Sydney aims to increase the housing stock in the ‘west central’ sub-region of Sydney (comprising Bankstown, Auburn, Parramatta, Holroyd and Fairfield in Figure 3) in the next quarter-century by 96 000, the highest total of any established ‘greyfield’ zone (New South Wales Department of Planning, 2010). Transit-oriented development and town-centre infill opportunities will play a key role, but prospects for the co-ordinated renewal of the in-between residential precincts are more problematic. There is a need for new initiatives that look beyond the largely fragmented, piecemeal processes currently in evidence (Newton, 2010; Newton et al., 2011).
A range of innovative community planning initiatives to stimulate more integrated solutions for these older suburbs that would use both public- and private-sector drivers might include
— affordable housing investment strategies through the community housing sector to replace poor-quality private rental housing and owner-occupied housing, address housing problems and rebuild community stability;
— development of renewal master plans to assemble land and implement the local renewal strategy;
— local renewal corporations or trusts based on public–private partnerships and local control to implement strategies over a 10–20-year period
— local renewal funds to help leverage private investment; and
— integrated place-focused and whole-of-government frameworks of targeted social and employment initiatives directed by local renewal corporations (Randolph and Holloway, 2004).
In the absence of effective governance structures, supportive planning and policy initiatives, and public funding, ‘third-city’ suburban renewal will continue to take place in an ad hoc way with little regard to the integration of broader social, economic and environmental goals including housing affordability (Ruming et al., 2010).
Conclusion
This paper has documented the clear emergence of areas of relative disadvantage and low income in Sydney working- and lower-middle-class suburbs built approximately between 1930 and 1970. These are areas sandwiched between the different worlds of inner-city gentrification and the outer suburban ‘aspirational’ fringe, and distant from other more privileged élite, beachside and harbourside localities. These are the suburbs of the ‘third city’ with a very different dynamic from that powering global Sydney (Flood, 2000). An ecological small-area analysis employing census data reveals that the main feature differentiating these areas from other parts of the western suburbs, and the rest of the city, is arguably the lack of upper-end incomes, rather than a concentration of the very poor per se. Utilising recent development application data from a typical middle-ring suburb, the case is also made for evidence of a contemporary process of reinvestment and renewal. That is, counter to aggregate indicators of socioeconomic disadvantage is evidence of market-led processes leading to a selective and scattered upgrading of the housing stock that appears to be gaining momentum.
How globally distinctive is the phenomenon we have documented? The nature of Sydney’s struggling middle-ring suburbs built in the early post-war period of the 20th century do suggest parallels with the well-documented American ‘first-suburb’ phenomenon (Hanlon et al., 2006). ‘First suburbs’ are conceived as those areas falling between the inner and outer cities—‘neither fully urban nor completely suburban’ (Puentes and Warren, 2006)—and more forcefully as ‘grotesque’, ‘desolate’ and ‘gothic’ (Short et al., 2007). They are the ‘gloomburbs’ that contrast to the outer ‘boomburbs’ (Lang et al., 2008). These areas are basically synonymous with suburbs developed between the 1940s and the 1960s that now represent a declining band on the edge of older central-city jurisdictions. In many cases, disinvestment has led to a degraded physical, social and economic landscape, with these places becoming a ‘devalorised urban form’ (Short et al., 2007). Such neighbourhoods sometimes fulfil roles as gateway suburbs for migrant communities and have become increasingly shaped by the same type of pressures which previously afflicted the inner core, such as a rising concentration of socially disadvantaged groups and a deterioration of the physical fabric and local environmental quality (Hanlon, 2010). Neither desolate nor undeveloped enough to qualify for targeted state assistance, they have also been caught in a ‘policy blindspot’ (Leigh and Lee, 2005; Puentes and Orfield, 2002; Puentes and Warren, 2006).
Some clear convergence in trends and processes is observable with the kinds of places considered in this paper. However, based on the Sydney experience, the Australian ‘third city’ appears to have retained a more buoyant housing market and issues of both race and municipal resourcing are less relevant. While ageing middle-ring Australian suburbs have gone through a cycle of relative decline, the extent of decline has been moderated by continuing demand for housing in part due to continued immigration pressure on Australian cities, strong metropolitan economies and a more supportive public investment climate in which universalist state-government-level expenditures have ameliorated impacts on local government rate revenues from falling population and income bases. In terms of the housing sector, at least, the widespread disinvestment evident in the US (Short et al., 2007) is less of a problem than the twin issues recounted here—namely, policy challenges caused by inchoate and unco-ordinated revalorisation processes, together with local aspects of disadvantage associated with particular building forms and tenures.
The picture, then, is therefore far from bleak and such investment implies continued attraction of these areas. The positive prospects for the less affluent middle suburbs in Australian cities, judged by the Sydney experience, lie with investors and residents looking for new investment opportunities outside an overheated inner core and increasingly overvalued greenfield suburbs. Moreover, the role of the middle suburbs as locations for the consolidation and retention of newer immigrant communities which can act as a driver for reinvestment and community stability needs further consideration. These areas face an uncertain, but not necessarily unfavourable, future.
Footnotes
Acknowledgements
An early version of the paper was presented to the ‘A Suburban World’ Conference organised by Virginia Tech and held in Reston, Virginia, in April 2008. The authors acknowledge the commentaries of the referees and the journal Editors as well as Andrew Tice of the City Futures Research Centre at the University of New South Wales for preparing the figures.
Funding Statement
This paper links to a larger research project funded by the Australian Research Council on ‘The Drivers and Outcomes of Re-investment in Low Density Suburban Housing Markets’ (DP0986122).