Geographic disadvantage: the demographics of social exclusion...
35
Geographic disadvantage: the demographics of social exclusion in Australia Paper for the University of Melbourne Social Inclusion Forum 25 th to 26 th of June 2009 Scott Baum Urban Research Program Griffith University William Mitchell Centre of Full employment and Equity University of Newcastle Abstract The question of disadvantage, deprivation and social exclusion cuts across many layers of society. It clearly impacts on individuals and this is where policy is often directed. However, space and place are also important. Many suburbs bare the negative outcomes of transitional periods of social and economic change. These transitions, that impact on individuals and families, are also reflected in the geographic patterns of disadvantage because of where people live and their interactions within the wider metropolitan systems. Using output an index of relative disadvantage this paper considers how these patterns of disadvantage are a feature of Australian cities and how the spatial patterns of disadvantage might be shifting as a result of the current global economic crisis. About the authors Associate Professor Scott Baum is a Research Fellow in the Urban Research Program, Griffith University. Trained in economics and sociology, he has had a long standing interest in understanding the social conditions that shape local communities and the lives of their residents. His work has appeared in international journals including Urban Studies (Sage Publishers), Papers in Regional Science (Blackwell Publishing) and Regional Studies (Taylor and Francis). He is an editor of Urban Policy and Research (Taylor and Francis) and is the general secretary of the Asia-Pacific Sociological association. Professor William Mitchell is Professor and Director of the Centre of Full Employment and Equity, University of Newcastle. An expert in labour economics he is a regular commentator on labour market issues in the national and international media and has research interests ranging from large scale macro-economic modelling to regional models of unemployment and spatial econometrics. His work has appeared in international journals including Regional Studies, Applied Economics, Rutgers Journal of Law and Urban Policy, Urban Studies and the International Journal of Environment, Workplace and Employment. His most current book ‘Full Employment Abandoned: Shifting Sands and Policy Failures’ (with J. Muysken) is published by Edward Elgar.
Transcript of Geographic disadvantage: the demographics of social exclusion...
Geographic disadvantage: the demographics of social exclusion in Australia
Paper for the University of Melbourne Social Inclusion Forum 25th to 26th of June 2009
Scott Baum
Urban Research Program Griffith University
William Mitchell
Centre of Full employment and Equity University of Newcastle
Abstract The question of disadvantage, deprivation and social exclusion cuts across many layers of society. It clearly impacts on individuals and this is where policy is often directed. However, space and place are also important. Many suburbs bare the negative outcomes of transitional periods of social and economic change. These transitions, that impact on individuals and families, are also reflected in the geographic patterns of disadvantage because of where people live and their interactions within the wider metropolitan systems. Using output an index of relative disadvantage this paper considers how these patterns of disadvantage are a feature of Australian cities and how the spatial patterns of disadvantage might be shifting as a result of the current global economic crisis. About the authors Associate Professor Scott Baum is a Research Fellow in the Urban Research Program, Griffith University. Trained in economics and sociology, he has had a long standing interest in understanding the social conditions that shape local communities and the lives of their residents. His work has appeared in international journals including Urban Studies (Sage Publishers), Papers in Regional Science (Blackwell Publishing) and Regional Studies (Taylor and Francis). He is an editor of Urban Policy and Research (Taylor and Francis) and is the general secretary of the Asia-Pacific Sociological association. Professor William Mitchell is Professor and Director of the Centre of Full Employment and Equity, University of Newcastle. An expert in labour economics he is a regular commentator on labour market issues in the national and international media and has research interests ranging from large scale macro-economic modelling to regional models of unemployment and spatial econometrics. His work has appeared in international journals including Regional Studies, Applied Economics, Rutgers Journal of Law and Urban Policy, Urban Studies and the International Journal of Environment, Workplace and Employment. His most current book ‘Full Employment Abandoned: Shifting Sands and Policy Failures’ (with J. Muysken) is published by Edward Elgar.
2
Introduction In the face of continued deterioration of the global economic outlook and despite the fact Australia has enjoyed a sustained period of aggregate national prosperity since the early 1990s social disparities have got worse with certain communities appearing to be particularly vulnerable (Baum et al. 2005). Social and economic transitions that have characterised the past three or four decades have left the scars of change across society. For many social scientists and politicians an inclusive nation is one that not only sets in place mechanisms for dynamic economic and social change, but also ensures that all its citizens, living in its disparate cities, towns and regions, can participate in and benefit from this economic development (Stimson 2001, Stillwell 2002). A nation that grows at the expense of inclusiveness risks significant social dysfunction that can in the extreme, challenge all levels of government and bring into question the shape and function of the nation’s social fabric. Stillwell (2002 p. 1) in particular argues
inequality has significant consequences for economic efficiency, social justice and environmental sustainability. Its sources are properly a central concern for political economic analysis. Its reduction is properly a concern for public policy.
In short, inequality and disadvantage matters. The establishment of a ministerial appointment in charge of social inclusion at the federal level is a sign that the types of issues raised by academics such as Stillwell and others are important in the national policy agenda. Reflecting this renewed policy interest are the comments by Prime Minister Kevin Rudd (The Australian 2006) who while in opposition argued
There is [a] view… which says that this nation and this people are at their best when we are a people and a nation committed to building a prosperous nation while at the same time not jettisoning our vision for a fair Australia and a fair society.
The question of disadvantage and deprivation cuts across many layers of society. It clearly impacts on individuals and this is where policy is often directed. However, space and place are also important. Many suburbs bare the negative outcomes of transitional periods of social and economic change. These transitions that impact on individuals and families are reflected in spatially concentrated disadvantage because of where people live and their interactions within the wider metropolitan systems. Disadvantage is reflected in the distribution of and access to social networks, role models and a range of essential services often necessary for complete inclusion and participation in society. The extent to which this disadvantage is in evidence across cities is not even. Recent research has repeatedly made this point (Baum et al. 2005, Baum et al. 2007). Swan (2005) argues that a consequence of growing spatial inequality in Australian society is a greater separation of rich and poor. In particular he points out
As the wealthy take over real estate close to the good jobs, the best schools and hospitals they lift the cost of entry to those areas. On the other hand, as the splintering middle and poorer people move further away they are paying more to get to work, school and see the doctor. If they lose their job, distance compounds their disadvantage (Swan 2005 p.172).
Internationally the OECD recognises that ‘[d]eprived areas, which have grown in number in recent years, limit the opportunities and prospects of people who live in them. Without a vision of their potential, a nation [not] only bears the costs but fails to realise the possibilities inherent in these places and their populations’ (OECD 1998, 11). The spatial distribution of disadvantage has, traditionally, been important to an understanding of the way cities were structured, how they operated and how these structures and processes reflected broader social, economic and demographic trends. The identification
3
and interpretation of patterns of deprivation are as important today as they have ever been. This paper focuses on Australia’s metropolitan suburbs and uses the most recent census data. It offers both a critical analysis of contemporary debates on the socio-spatial structure of cities and also develops an index of relative deprivation (the General Deprivation Index). It proposes, in short, both an analytical framework and an empirical base for policy debate regarding a socially inclusive nation.
The geography of disadvantage A focus on the geography of disadvantage necessarily rests initially on understanding the transformations that have been part of the post-industrial metropolitan landscape. New national and international socio-economic forces have reshaped national geographies in general and the characteristics of cities in particular, resulting in a range of diverse social and spatial outcomes (O’Connor, Stimson and Daly 2001). Where once socio-spatial outcomes may have been clearly defined in research focusing on cities of the industrial era—here we might refer to the early and subsequent work by social area analysts (Timms 1970; Theodorson 1982)— a new or different set of divisions have been seen to emerge in post-industrial or post-Fordist cities. These new divisions do not necessarily exist in complete isolation from divisions that have appeared in earlier periods, but rather have developed from these existing patterns. Contemporary patterns are therefore reflective of the socio-spatial histories of cities. Nevertheless, what we are now seeing, and have been seeing over the past two or three decades, is a complex and wide ranging set of interlinked factors impacting in new ways on the social and economic fabric of the contemporary city (see Benassi et al 1997; Kesteloot 1998; Mingione 1996). This wider frame of reference takes into account three inter-connected areas of social life that are crucial factors in generating unequal socio-economic outcomes across a range of levels of abstraction–from individuals to aggregate regions and nation states. Internationally, sociologists talk about changes in the economic system and the nature of work, transitions in primary support systems, and a crisis in the welfare state as all being important to understand division occurring in society. With respect to these processes Kesteloot (1998, p.126) points out:
… there is a growing consensus about three distinct spheres in which the sources of polarisation originate, namely transformations in the division of labour,…; the restructuring of nation-states and particularly the slow dismantling of the welfare state; and finally the second demographic transition, which results in the appearance of new household forms and the parallel increase of single people and social isolation.
This contemporary social structure has resulted in unequal access to resources and life chances for individuals and households. This inequality is expressed in a polarised socio-spatial urban structure of as ‘rich and poor concentrate respectively in rich and poor environments in terms of the resources of collective consumption, housing, mobility and access to jobs’ (Kesteloot 1998, p.127). Within the international literature reference to these types of concerns can be seen in research by Marcuse (1997; 1989), Marcuse and van Kempen (2000a, 2000b), Soja (2000), Dear (2000), Mikelbank (2004) and Walks (2001) in North America and by Hamnett (2003), Wessel (2001), Rhein (1998) and Musterd and Ostendorf (1998) in the United Kingdom and Europe. Marcuse (1997, p.228-229) identifies changes in space and race as new processes of exclusion that are part of ‘a broad pattern that makes up the post-Fordist city’, while for Soja (1997, p. 193) the force of change has seen the form of the city explode in ‘unprecedented scale, scope and complexity’. Walks (2001, p. 440), in an investigation of the changing socio-spatial structure of Toronto, supports the earlier work suggesting that ‘the social ecology of
4
the post-Fordist/global city may be characterized by increasing social complexity and differentiation among, between and within neighbourhoods’. Outside of North America Hamnett (2003) has identified new dynamics of overlapping complexity and polarisation in London. He points to, among other things, the differentiation of the inner-city suggesting that it is ‘now one of the most prosperous areas in western Europe’ but it also has ‘one of the highest concentrations of deprivation’ (Hamnett 2003 p. 189). He argues this is a result of London’s global role and its specific industrial, occupational and social structure. Parallel research into the socio-economic change and socio-spatial structure has been undertaken for other cities and regions in Europe and presented in the work of Musterd and Ostendorf (1998) Rhein (1998) and Wessel (2001). The primate urban region of each Australian state (Sydney, Melbourne, Brisbane, Perth, Adelaide, and Hobart), together with the capitals of the two territories (Canberra, Darwin) have been similarly transformed. In all these urban regions new social realities are reflected in differentiated socio-spatial and socioeconomic processes. Where once large relatively homogeneous working class communities were the dominant feature of most Australian cities (Baum et al. 2002), the economic and social processes that are characteristic of post-Fordism have resulted in a more complex socio-spatial patterning. Major dimensions of this complexity include an increasing suburbanization of poverty into Australia’s middle ring and old outer suburban areas, the movement of an inspirational class of households to opportunities in new outer suburbs, the dividing up of the old working class communities into several groups of new disadvantaged communities with each being affected by the new economic processes in different ways and, last but by no means least, the development of new advantaged communities closely tied to developments in the world economy (Stimson et al. 2001; Baum et al. 1999, 2002, Randolph 2004, O’Connor, Daly and Stimson 2001). Recent work by Randolph (2004, p. 492) argues the cities of Australia are subject to a range of intensifying pressures that are producing a series of different outcomes within and between cities. Although he provides little empirical support, he argues:
[t]he turnaround in the inner-city, the suburbanization of disadvantage, the new inspirational suburbs and the increasingly multi-regional city structure with increasingly multi-scaling of processes and outcomes, all point to new forms of city structure that make a change from the prevailing patterns of the period between 1945 and 1980.
Similar arguments have also been made by O’Connor, Stimson and Daly (2001) and Gleeson (2004). The empirical work that has been conducted on Australian cities has been wide ranging and in a collective sense illustrates the changes postulated by Randolph (2004) and others. Early work by Stilwell and Hardwick (1973), Stilwell (1989) and Stretton (1970) point to the existence of significant divisions in cities during early periods of post-industrial development. More recent research continuing the investigation of the socio-spatial structure of the contemporary Australian city includes work by Baum and Hassan (1993), Gregory and Hunter (1995), Hunter and Gregory (1996) Raskall (1995, 2002), Beer and Forster (2002), Stimson et al. (2001) and Baum et al (1999; 2002). Reflecting the flavour of this empirical work Beer and Forster (2002, p. 13) argue:
In particular, job losses tended to be most severe in residential areas of manufacturing employment in lower income inner and middle suburbs. In contrast, job gains in service employment were strongest in the city centre and higher income inner and middle suburbs. Regions such as western Sydney, western and northern Melbourne and northern and western Adelaide, where manufacturing employment grew most rapidly during the 1950s and 1960s, were major losers in the process.
5
These changes have exacerbated ‘established patterns of residential differentiation and contrasts in well-being between high- and low-income suburbs’ (Beer and Forster 2002, p. 13). Similarly, empirical work by Stimson et al. (2001a) and Baum et al. (1999, 2002, 2006), support the conclusion that across Australian metropolitan regions some spaces are, increasingly, being identified as places of status and opportunity, while others are increasingly becoming known, as sinks of vulnerability and deprivation.
Building the General Deprivation Index for Australian cities Within the research literature there has been a range of empirically based indicators and indices designed to account for urban deprivation (for example see Bentham 1985; Carstairs and Morris 1989; Bradford et al. 1995; Chakravorty 1996; Baum et al. 1999, 2005; DETR 2000; Rahman et al. 2000; Midgley et al. 2003). While all of these approaches cover various aspects of socio-economic deprivation, they vary in terms of the method of indicator construction, the types of individual measures used and the spatial scale at which deprivation is measured. Common indicators of deprivation include income level (both of households and individuals), levels of unemployment and labour force participation, all of which are considered to be direct measures of relative deprivation. In addition to these variables some research (depending on data availability) uses indicators relating to housing condition or quality, while others make use of social problem indicators such as crime or health outcomes data (see for example Bentham 1985; Williams and Windebank 1995). From a design view point, some measures take standardised data and produce unweighted indices (Carstairs and Morris 1989), while others (Bentham 1985; Baum et al 1999, 2005) make use of multivariate methodology to derive typologies of localities based on a range of socio-economic variables. Finally, considering spatial units, the choice made depends on the level of aggregation at which data is available and includes suburbs, neighbourhoods, local boroughs and local government authorities and enumeration districts (Bradford et al 1995; Chakravorty 1996; Sloggett and Joshi 1998; Baum et al 1999). The index used in this paper was first developed by Langlois and Kitchen (2001) for Montreal, Canada and subsequently used by Baum (2004) in an analysis of Sydney. The index uses multivariate analysis to construct a dimensional measure of socio-economic deprivation and as such provides a compliment to the existing range of instruments.
Methodology Following the method outlined in Langlois and Kitchen (2001) this paper uses Principal Components Analysis and the resultant factor scores to develop a range of deprivation indicators and a General Deprivation Index (GDI). Principal Components Analysis is used to reduce a larger set of variables thought to be associated with urban deprivation to a smaller number of sub-sets or factors. The meaning of each sub- group or factor is determined by the variables most highly associated with that factor (as displayed in the rotated components matrix). Each observation is given a score on each factor selected for the particular indicator. The factor scores, taken in combination, are then used to represent aggregate dimensions of urban deprivation.
6
Figure 1: Types of Urban Deprivation, based on the Principal Components Operationally, deprivation can be defined in terms of a set of social, economic and housing conditions and problems that place some residents in a more disadvantaged position compared with the population as a whole (Kitchen 2001). Figure 1, above, aides the interpretation of the index. It shows possible combinations of factors associated with the various situations of urban deprivation. The factor considered to be the general indicator of deprivation (factor I in Figure 1) is considered to be a necessary condition for socio-economic deprivation. Once this condition is satisfied, the overlaps with other components in figure one (I/II, I/III, I/IV, I/V) define more specific situations of deprivation. Apart from situations where there is an overlap between Factor I and other factors ( II, III, IV, V) suburbs can be characterised has having low deprivation or deprivation only defined by membership to factor I. Hence the urban socio-spatial dimensions of deprivation are conditioned by a range of factors. These factors are likely to reflect the well known dimensions of social area analysis (Timms 1970, Theodorson 1982) including family type, demographic structure, attachment to the labour market and ethnic status. The overlap between these individual components in urban space can inform an understanding of a number of social conditions including deprivation. Langlois and Kitchen (2001) find, using this framework, that urban deprivation can be structured along a general component of socio-economic status (factor 1) plus components accounting for age, labour force participation and ethnicity. Considering these factors in turn they argue that the types of urban deprivation across the city include communities that have low socio-economic status and old aged households, low socio-economic status and a strong ethnic background and low-socio-economic status and low labour market participation.
Factor III
Factor IV
Factor V
Factor II Factor I I/ IV
I/ III
I/ II
I/ V
7
While suburbs can be analysed on the basis of the different types of deprivation (i.e. the intersection of low socio-economic status and low labour market participation), following Langlois and Kitchen (2001) it is recognised that places can suffer from several types of deprivation and that the intensity of urban deprivation might be important. This observation results in the development of a General Deprivation Index (GDI). The GDI is constructed using rescaled scores arising from the principal components analysis. Rescaling follows equation (1):
S*ij =)min(max
)min(
jj
jSij
−
− (1)
where ( 0 ≤ S*ij ≤1); and Sij is the factor score for locality i on principal component j; maxj and minj are the highest and lowest factor score on component j. This equation produced rescaled factor scores in the range of zero to one and allowed the following equation to be used to develop a General Deprivation Index.
GDIi=comptot
ijSSik
_
)*1(* ∑+ (2)
where ( 0 ≤ GDIi ≤1); and Sik is the rescaled factor score of locality i on component k which plays the primary role in deprivation; S*ij is the rescaled factor score of one of the secondary components; and tot_comp is the total number of components. This produces a simple weighted index that accounts for all the factors derived from the initial principal components analysis.
Data The data used to develop the deprivation index is sourced from the Australian Bureau of Statistics 2006 Census of Population and Housing. The basic community profile contains a range of census information for individuals and households including variables that may be used as indicators of deprivation. The variables were selected with reference to existing studies (most notably Langlois and Kitchen 2001) and with due regard to the data constraints imposed by the content of the Australian Bureau of Statistics basic community profile. Sixteen individual variables are used in the analysis. They are divided into demographic variables, income variables, engagement with work variables and housing variables (see Table 1). Demographic variables include the percentage of the population who are indigenous Australians; the percentage of persons aged 65 years and older; the percentage of persons requiring daily assistance with activities; the percentage of the overseas born population who moved to Australia between 1996 and 2001 (recent arrivals); and the percentage of the population who do not speak English well and the percentage of families headed by single parents. These variables were selected to include sections of the population identified across a number of studies as vulnerable to adverse economic and social change (Bentham 1985; Baum et al. 1999; Langlois and Kitchen, 2001). Four income measures were included in the analysis. Two of these account for the presence of low income. One is a measure of low family income while the other accounts for low individual income. In both cases ‘low’ income is defined as those in the bottom 10 percent of the income distribution. In addition to these variables, measures of median family income and median individual income are also included. Such income measures are often used in indicators of deprivation and disadvantage and can be considered as a direct indicator of deprivation (Chakravorty 1996; Baum et al. 1999; Langlois and Kitchen 2001).
8
Table 1: Variables included in the Analysis Demographic/household
• Indigenous population (%)
• Persons aged older than 64 years of age (%)
• Person requiring assistance with daily activities (%)
• Recent immigrants to Australia (arrived in the between 2001 and 2006) (%)
• Population who do not speak English well (%)
• Single parent families (%) Income
• Median family income ($)
• Families with low incomes (bottom 10% of the distribution) (%)
• Median individual income ($)
• Individuals with low incomes (bottom 10% of the distribution) (%) Housing
• Households in public housing (%) Engagement with work
• Youth unemployment rate (persons aged 15 to 24) (%)
• Male unemployment rate (%)
• Male labour force participation rate (%)
• Female unemployment rate (%)
• Female labour force participation rate (%)
Like income, labour force participation is a well established component of deprivation indicators (Slogett and Joshi 1998; Baum et al. 1999, 2006; Langlois and Kitchen 2001; Midgley et al. 2003). The selection of variables used here included measures of a lack of labour force engagement (youth and overall unemployment) as well as measures of participation in work. Youth unemployment is measured as the number of persons aged between 15 and 24 years (in a given area) who were unemployed as a percentage of the total labour force aged between 15 and 24 years. Overall unemployment is divided into both male and female unemployment and defined as the number of unemployed males and females in an area as a percentage of the total gender specific labour force in that area. Male and female labour force participation was measured by the number of males and females over in the labour force as a percentage of the total males and females in the age appropriate population. One housing variable is included in the analysis. It is a measure, essentially, of those unable to afford private shelter. The percentage of households living in public housing is, therefore, readily associated with social disadvantage, low income and a lack of labour force engagement (Williams and Windeband 1995; Slogget and Joshi 1998).
Spatial units In developing the index reported here several possible levels of spatial aggregation were considered. Australian Bureau of Statistics data are available at levels ranging from collector districts (CDs) comprising approximately 200-300 households, to state and national level
9
aggregations. The difficulty in identifying locations in a meaningful way using a collector district codes (rather than a place name) meant that data at this low level of aggregation was less useful. Given this and the goal of considering disadvantage across local suburban communities, the General Deprivation Index is developed for suburbs within what are largely considered to be the capital city regions.
The general deprivation index for Australian cities The methodology described above was applied to Australian Bureau of Statistics defined suburbs across the capital city regions of Sydney, Melbourne, Brisbane, Adelaide, Perth and Hobart. The Principal Components Analysis with varimax rotation resulted in four components accounting for 78 per cent of variance. The four components and the variables associated with each are outlined in Table 2. One of the components (component three) represents the overarching indicator of deprivation as set out by the methodology of Langlois and Kitchen (2001). This component was negatively associated with median individual income (-0.886) and median family income (-0.804) and positively associated with low income families (0.590) and low income individuals (0.507). The other components represent situations where deprivation might be thought to exist so that suburbs with high scores on component three and one or more of the remaining three components will represent the level of intensity or variation in deprivation across the cities. Considering the factors in detail, the figures in bold in Table 2 indicate high loadings between a given individual variable and a particular component. Factor one, an indicator of the presence of disadvantaged households and families and disadvantage associated with negative labour market outcomes was positively associated with the households in public housing (0.774), indigenous population (0.774), single parent families (0.760) male unemployment (0.584), female unemployment (0.571) and youth unemployment (0.503). This factor accounted for 21.41% of the overall variance. The second factor, accounting for suburban patterns of old age and low labour market participation was positively associated with persons aged over 64 years (0.932) and persons requiring daily assistance (0.763) and negatively with male labour force participation (-0.752) and female labour force participation (-0.742). This component accounted for 21.1 % of the variance. The final component accounted for approximately 15 % of the variance and indicated suburban concentrations of people who had recently moved to Australia (0.838) and members of the population who had poor English skills (0.722). Following Langlois and Kitchen (2001) the individual components of deprivation across Australian suburbs are related to a general socio-economic or income component and the intersection of this with components accounting for old age and low labour market participation, unemployment, pubic housing and the presence of indigenous population and ethnicity. It is the combination of these separate components that are used to derive the General Deprivation Index. Table 2: Rotated Components Matrix Component 1 Component 2 Component 3 Component 4 Households in public housing (%) 0.774 0.203 0.117 0.202 Indigenous population (%) 0.774 -0.063 0.106 -0.243 Single parent families (%) 0.760 0.247 0.412 0.084 Male unemployment rate (%) 0.584 0.299 0.446 0.441 Female unemployment rate (%) 0.571 0.141 0.529 0.458 Youth unemployment rate (%) 0.503 0.174 0.437 0.388 Persons aged older than 64 years of age (%) -0.151 0.932 0.021 -0.065
Person requiring assistance with daily activities 0.201 0.763 0.268 0.009
Male labour force participation rate (%) -0.303 -0.752 -0.222 -0.293
10
Female labour force participation rate (%) -0.243 -0.742 -0.422 -0.279
Median Individual income ($) -0.168 -0.320 -0.886 -0.081 Median family income ($) -0.340 -0.253 -0.804 0.036 Families with low incomes (%) 0.469 0.472 0.590 0.289 Individuals with low incomes (%) -0.388 -0.381 0.507 0.438 Recent immigrants to Australia (arrived in the between 2001 and 2006) (%)
-0.016 -0.044 -0.246 0.838
Population who consider they do not speak English well (%) 0.025 0.165 0.292 0.722
% variance explained 21.41 21.100 20.91 14.60 Following the methodology described above and taking the components outlined in table 2 a General Deprivation Index for Australian suburbs was estimated. The index ranges from 0 to 1 with suburbs being relatively more deprived as the index approaches 1. To facilitate discussion of the GDI the suburbs are divided into six groups according to the relationship of each suburbs score to the average. Band or group 1 comprises those suburbs with a General Deprivation Index that is significantly higher than the average. Band or group 6, conversely, comprises those suburbs with significantly lower than average index scores1. In addition, a regional concentration ratio is developed to illustrate the relative distribution of suburbs in each band across each city. The regional concentration ratio is a version of a location quotient (Hill et al. 1980). It determines the extent to which any metropolitan region has an over concentration of suburbs in any given band. The RCR is calculated by considering the percentage distribution of a particular band in each metropolitan region divided by the percentage distribution of that band across all metropolitan regions. Like a location quotient, a RCR greater than 1 indicates that the number of suburbs in a particular band in a particular city is overrepresented in that band. An RCR less than 1 indicates the opposite outcome. Table 3: Regional concentration Ratio, Australian metropolitan cities
GDI band
1 2 3 4 5 6
highest
relative
deprivation
high relative
deprivation
low relative
deprivation
lowest
relative
deprivation
Sydney 1.54 1.29 0.93 0.86 0.96 2.35
Melbourne 1.60 1.00 1.31 0.80 0.67 0.33
Brisbane 0.00 0.74 0.93 1.22 1.20 0.22
Adelaide 1.16 1.34 1.19 0.97 0.33 0.00
Perth 0.00 0.51 0.93 1.41 0.63 0.12
Hobart 1.25 1.45 1.38 0.84 0.00 0.00
1 The suburbs were grouped according to the mean and standard deviation of the distribution of scores. Band 1 suburbs have a score more than 2 standard deviations higher than the mean and band 2 suburbs have a score between 1 and 2 standard deviations higher than the mean. Conversely band 6 suburbs have scores more than 2 standard deviations below the mean and band 5 suburbs have a score between 1 and 2 standard deviations below the mean.
11
To supplement the RCR, and analyse how the bands of suburbs differ from one another this paper presents a visual interpretation of the data (see Figure 2). The method essentially pictures confidence intervals (CI) to allow the drawing of inferences about the outcomes in a given band that take account of both the mean and also the wider spread of the data (Masson and Loftus 2003). Comparing the mean for the distribution of each variable across the 6 bands allows consideration of the potential importance of each individual variable to the distribution of the GDI. The distribution of the individual variables used in the GDI across the 6 bands is largely as expected given the way the index has been developed. For example Band 1 suburbs had far and away the highest means for indicators such as the percentage of low income families (25.57%) and individuals (17.72%) and the various unemployment variables (male unemployment 14.08%; female unemployment 14.76%; youth unemployment 21.53%). In all cases these were significantly different from the average for all suburbs and considerable higher than the suburbs in band 6. Interestingly for the variable accounting for the percentage of the population aged 65 plus the pattern across the 6 bands was not clear. Neither was the variable accounting for the presence of recently arrived immigrants. In both cases this may suggest that these variables are less important to the overall GDI.
12
Figure 2: 95% confidence intervals, selected variables
654321
GDI band
25.000000000000
20.000000000000
15.000000000000
10.000000000000
5.000000000000
0.000000000000
95%
CI p
erce
nt s
tate
hou
sing
Figure 2 (a): 95% confidence intervals, percent state housing
654321
GDI band
3.500000000000
3.000000000000
2.500000000000
2.000000000000
1.500000000000
1.000000000000
0.500000000000
0.000000000000
95%
CI p
erce
nt in
dige
nous
pop
ulat
ion
Figure 2 (b): 95% confidence intervals, percent indigenous population
654321
GDI band
30.000000000000
25.000000000000
20.000000000000
15.000000000000
10.000000000000
95%
CI p
erce
nt s
ingl
e pa
rent
fam
ilies
Figure 2 (c): 95% confidence intervals, percent single parent families
654321
GDI band
14.00
12.00
10.00
8.00
6.00
4.00
2.00
95%
CI m
ale
unem
ploy
men
t rat
e
Figure 2 (d): 95% confidence intervals, male unemployment rate
654321
GDI band
15.00
12.50
10.00
7.50
5.00
2.50
95%
CI f
emal
e un
empl
oym
ent r
ate
Figure 2 (e): 95% confidence intervals, female unemployment rate
654321
GDI band
25.000000000000
20.000000000000
15.000000000000
10.000000000000
5.000000000000
95%
CI y
outh
une
mpl
oym
ent r
ate
Figure 2 (f): 95% confidence intervals, youth unemployment rate
13
654321
GDI band
16.000000000000
14.000000000000
12.000000000000
10.000000000000
95%
CI p
erce
nt p
opul
atio
n 65
and
old
er
Figure 2 (g): 95% confidence intervals, percent populations aged 65 and older
654321
GDI band
7.000000
6.000000
5.000000
4.000000
3.000000
2.000000
1.000000
95%
CI p
eopl
e w
ho n
eed
daily
ass
ist
Figure 2 (h): 95% confidence intervals, percent populations needing daily assistance
654321
GDI band
75.00
70.00
65.00
60.00
55.00
50.00
95%
CI m
ale
labo
ur fo
rce
part
icip
atio
n ra
te
Figure 2 (i): 95% confidence intervals, male labour force participation rate
654321
GDI band
60.00
40.00
95%
CI f
emal
e la
bour
forc
e pa
rtic
ipat
ion
rate
Figure 2 (j): 95% confidence intervals, female labour force participation rate
654321
GDI band
1000
800
600
400
200
95%
CI m
edia
n in
divd
ual i
ncom
e
Figure 2 (k): 95% confidence intervals, median individual income
654321
GDI band
2400.0
2100.0
1800.0
1500.0
1200.0
900.0
600.0
95%
CI m
edia
n fa
mily
inco
me
Figure 2 (l): 95% confidence intervals, median family income
14
654321
GDI band
25.000000000000
20.000000000000
15.000000000000
10.000000000000
5.000000000000
95%
CI p
erce
nt lo
w in
com
e fa
mili
es
Figure 2 (m): 95% confidence intervals, percent low income families
654321
GDI band
20.000000000000
18.000000000000
16.000000000000
14.000000000000
12.000000000000
10.000000000000
8.000000000000
95%
CI p
erce
nt lo
w in
com
e in
divd
uals
Figure 2 (n): 95% confidence intervals, percent low income individuals
654321
GDI band
9.000000000000
8.000000000000
7.000000000000
6.000000000000
5.000000000000
4.000000000000
3.000000000000
95%
CI p
erce
nt re
cent
arr
ival
s
Figure 2 (o): 95% confidence intervals, percent of recent arrivals
654321
GDI band
12.000000000
10.000000000
8.000000000
6.000000000
4.000000000
2.000000000
0.000000000
95%
CI p
erce
nt o
f pop
ulat
ion
with
poo
r Eng
lish
Figure 2 (p): 95% confidence intervals, percent of population with poor English
The patterns identified within this paper refer to the relative distribution of deprivation across the Australian metropolitan cities and, of course, reflect the specific type of analysis undertaken including the variables used and the spatial units adopted. Using different variables for example may result in some difference, but the change at the margins is likely to be small. Likewise, if the analysis was undertaken on a city-by-city basis then the position of different suburbs at the extremes would also be different. Additionally, if this was an analysis of relative advantage or affluence a different set of variables would be used and a different picture presented. In this sense it is extreme relative deprivation that is the key focus as all other suburbs are all better off in relative terms. It should also be noted that while the suburbs as depicted here may appear to be homogeneous there may still be internal heterogeneity and thus analysis undertaken at a smaller spatial scale may well produce different outcomes. These issues need to be kept in mind when considering the analytical outcomes.
The geography of relative disadvantage: an overview Close to 13 percent of all suburbs included in the analysis were in band 1 (highest relative deprivation) or band 2 (high relative deprivation). The band 1 suburbs have been often associated with the negative externalities of economic restructuring and other social and demographic transitions. Both band 1 suburbs and to some extent band 2 suburbs have been
15
widely discussed both in the popular media and in academic research and policy. These suburbs are home to the real ‘battlers’ of the metropolitan regions. Some of these locales have become well known because of the range of social problems that seem to be concentrated there, but others just struggle along. Many of them tend to have concentrations of multiple disadvantages whereby one social problem is compounded by other problems. The geography of these deprived places also impacts on their social and economic scars as if the disadvantaged
…live in a community or locality where distance, diminished labour market opportunities and reduced services (particularly public transport, child care, education and training) make both social and economic participation very difficult if not impossible (Cass 1990, p. 9)
Recent years have seen some improvements, usually mobilised by local boosterism, but by in large, these battling localities remain relatively unchanged. The rising tide has lifted all the boats; some are just more buoyant than others. Many of these communities were developed with a strong emphasis on public housing estates associated with the rapid growth of manufacturing industries in the 1950s, 60s and 70s. They have become some of the most unfortunate victims of structural economic change. These localities dramatically illustrate the changing nature of the suburban industrial communities that were established during the post-World War II Fordist phase of expansion. For these localities, the post-Fordist economy has caused the old established economic and social geography to become obsolete as they are faced with new forces of production (Searle 1993). This is well illustrated by Peel’s (1995) case study of Elizabeth in Adelaide
The combination of economic growth and adventurous public planning that underpinned the workers’ city did not last. Crisis first arrived in the form of economic downturn of the mid 1970s. That was followed by restructurings which severed subsequent recovery from job creation…In this new environment places like Elizabeth faced an uncertain future…Their role in a reorganised and restructured economy would depend upon their ability to adjust, to attract and hold on to new investment and new kinds of jobs (p. 156).
In this sense these deprived suburbs have had a long history, but equally band 1 and 2 suburbs also contain the new suburban disadvantaged that have been discussed by Randolph (2004) and earlier by Badcock (1997). These band 1 and 2 suburbs also contain localities that have become known as welfare or retirement migration locations (Baum et al. 2005) often on the fringes of the city in the newly expanding peri-urban regions (tree change localities) or in coastal communities (sea-change localities). They also contain suburbs that are home to large proportions of students, for whom relative deprivation is (we hope) a short term concern or have become the new ethnic enclaves which are the face of some of the new urban deprivation (Birrell 1993) and are where ‘ethnic migrants from non-English speaking backgrounds make up the major disadvantaged component of a significant and growing socio-economic divide…(Forrest and Poulsen 2003: 9).
The 12 percent of suburbs located in band 5 and 6 are, in contrast to band 1 and 2 locations, the least deprived suburbs in metropolitan Australia. These locations too have been the subject of academic discussion and are generally well known. They have their equivalents in large metropolitan regions around the world, and in many ways these localities, especially those in band 6 have more in common with their international counterparts than they do with many urban localities in Australia. While Australia may not yet have the high security gated-community form that exist elsewhere (although there are some (Gleeson 2004a)), the exclusive content of such suburbia is readily found. Some of these suburbs are among Australia’s most established residential localities others reflect the new advantage associated
16
with a changing urban social reality. Most often considered in terms of gentrification and inner-city population turn around, the changes that have been underway in inner-city and near inner-city localities signal a transitional phase in the life cycle of many communities (Ley 1986, Forster 1995, Gleeson 2004b). Like the wealth-belt, these transitional ‘latte’ localities are advantaged, but at the same time they do exhibit glimpses of a less advantaged past. The inner-city change is stimulated by widespread redevelopment and by ‘show piece’ projects (think Sydney’s Darling Harbour, Melbourne’s Docklands or Brisbane’s South Bank). McHugh (1999 p. 135) made this comment about the changes:
These aren’t the only development’s currently underway in Australia but they are places to watch. In some, entire cities are about to spring up where once there were industrial eyesores. In others, new life will be breathed into old…the benefit is that our major capitals are regaining a heart. They are coming to life 24 hours a day, seven days a week.
These less deprived suburbs are the antithesis of Mark Peel’s (1995) places. They are a world away from the poverty traps and public housing estates that have become policy concerns in recent years (Swan 2005). In between the two extremes are the suburbs that might be considered to be part of middle Australia. These suburbs, the bulk of those included in the analysis, have been identified elsewhere as comprising the more traditional middle class suburbia and communities of aspirant working class battlers (Baum et al. 2005). In short a range of fortunes are found in the middling suburbs. Band 4 suburbs are likely to include examples of Pusey’s (2003) middle Australia project as well as tracts of McMansions where journalists argue people ‘love cocooning inside…[and] which are like castles, fun factories and mini resorts in one’ (Hawley 2003 p. 25). Others such as Randolph (2004 p. 489) refer to these as the ‘new urban rings of affluence, aspiration and security’, a sentiment reflected in the work of Gleeson (2004b). While these places are not the most advantaged localities in metropolitan Australia, they have managed to perform relatively well in socio-economic terms. In many cases the suburbs of band 4 include those which have been long established, now aged households that are asset rich but income poor. The changing occupational structure of Australia has impacted the traditional working-class localities in the metropolitan regions. Many of these areas fall within the band 3 designation. They are suburbs long identified with the working-class of old, but are now characterised as areas of aspirant working-class battlers (Gwyhter 2002). While the economic pain is not quite the same in these suburbs as in those that are most deprived the band 3 suburbs are also likely to feel the pinch of rising interest rates as many are located in Australia’s mortgage belt.
17
The extremes of disadvantage: a city by city discussion
Sydney The pattern of relative deprivation in Sydney reflects conventional wisdom. Spatially Sydney is a tale of two cities that starts in the deprived western suburbs and ends on the storied north shore. Both the highest score (Claymore) and the lowest score (Milsons Point) are found in the Sydney region. In relative terms the RCR illustrates that Sydney has more suburbs in band 1 (highest relative deprivation) and band 6 (lowest relative deprivation) reflecting the polarised nature often associated with global cities (Sassen 1991). The suburbs with the greatest socio-economic scars are well known, with some being the focus of media attention for all of the wrong reasons. They have been widely commented on by academics in terms of the suburbanisation of disadvantage in the Sydney region (Burchell 2003, Randolph and Holloway 2005), with the suburbs of Western Sydney conjuring up symbols of an undifferentiated urban bad land (Burchell 2003). The suburbs of Airds, Cabramatta, Auburn and Fairfield in Sydney’s western suburbs all score highly on the General Deprivation Index. So do some localities on the city’s far north coast including The Entrance and Canton Beach. Contrasting these places, the north shore is where Sydney’s wealth belt reside (Baum et al. 2005). The suburbs of Kirribilli, Milsons Point and Double Bay are included in the Sydney suburbs with lowest relative deprivation. Interestingly, while suburbs including Redfern and Dawes Point once were identified as among the city’s deprived locales, heavy gentrification activity has seen recorded deprivation in these suburbs decline. Table 4: Band 1 and Band 6 suburbs, Sydney
Band 1 Band 6 Airds Heckenberg Alexandria Manly Ashcroft Lakemba Annandale McMahons PointAuburn Lethbridge Park Balmain Milsons Point Bankstown Miller Balmain East Mosman Bidwill Old Guildford Bellevue Hill Naremburn Bonnyrigg Punchbowl Birchgrove Neutral Bay Busby Sadleir Bondi Beach North Sydney Cabramatta South Granville Breakfast Point/ Mortlake Northwood Cabramatta West St Johns Park Cammeray Paddington Campsie Tregear Centennial Park Point Piper Canley Heights Villawood Chiswick Potts Point Canley Vale Warwick Farm Coogee Pyrmont Carramar Whalan Cremorne Queenscliff Cartwright Wiley Park Cremorne Point Rozelle Claymore Willmot Crows Nest Rushcutters Bay Emerton Yennora Darling Point St Leonards Fairfield Canton Beach Darlinghurst Surry Hills Fairfield East The Entrance Dawes Point/ The Rocks/ Tamarama Fairfield Heights Double Bay Waverton Edgecliff Wollstonecraft Elizabeth Bay Woollahra Erskineville Woolwich Fairlight Homebush Bay Kirribilli Lavender Bay
18
Figure 3: Relative deprivation, Sydney
Melbourne Melbourne has for many years worn the tag of Australia’s most liveable city. It has also been labelled as an expansionist events city (Engels 2000) or an up-and-coming knowledge city (Yigitcanlar 2005). However, Melbourne’s prosperity does not translate easily spatially equal outcomes. Relatively and considering the distribution of Australian suburbs in general, Melbourne has not got the extreme polarised structure of Sydney. However it does have suburbs at both extremes of the General Deprivation Index reflecting the generally uneven distribution of deprivation across the city (Baum et al. 2005, O’Connor and Healy 2004). The suburbs that have been most scarred (highest relative deprivation) include those in Melbourne’s post-war industrial growth heartlands including Broadmeadows and Sunshine. These are suburbs, which like the industrial regions in the other capitals, have been at the wrong end of economic and social transitions. An interesting finding is that the concentration of students in Carlton has contributed to that suburb’s poor showing. Band 6 suburbs (lowest relative deprivation) include East Melbourne, Docklands and Burnley; suburbs associated with Melbourne’s new economy activities and the gentrification that has occurred in the inner city. The spatial pattern of deprivation in Melbourne reflects long established trends with extreme relative deprivation located further out and lower deprivation closer to the central business district.
19
Table 5: Band 1 and Band 6 suburbs, Melbourne
Band 1 Band 6 Albanvale Heidelberg West Burnley Ardeer Kings Park Docklands Bangholme Lalor East Melbourne Braybrook Maidstone St Kilda West Broadmeadows Meadow Heights Campbellfield Springvale Carlton Springvale South Coolaroo St Albans Dallas Sunshine North Dandenong South Sunshine West Fawkner Thomastown Frankston North
Figure 4: Relative deprivation, Melbourne
Brisbane and Perth Australia’s two sun-belt capitals (Brisbane and Perth) have substantively different levels of relative deprivation when compared to the other main capitals. Both cities are considered to be residing over states which are on the up- side of Australia’s two-speed economy. Brisbane has no band 1 (highest relative) deprivation. The floor of Brisbane deprivation is band 2 (high relative deprivation) The distribution of relative deprivation measured by the regional concentration ratio shows a skewed distribution—more low deprivation than extreme deprivation suburbs— relative to the distribution of the GDI. The suburbs of Inala and Logan Central have the highest levels of relative deprivation in the Queensland capital, together with localities on the northern extremes of the Brisbane region such as Caboolture south, and other places such as Nathan and Robertson, located adjacent the Griffith University campus. Two suburbs are included in the band 6 suburbs - Newstead located on the northern bank of the Brisbane river and adjacent the CBD and Brookwater in Brisbane’s west. The spatial distribution of relative deprivation in Brisbane shows concentrations of
20
higher deprivation in the southern suburbs and lower relative deprivation north of the river and closer to the city centre. Table 6: Band 2 and Band 6 suburbs, Brisbane
Band 2 * Band 6 Beachmere Loganlea Brookwater Caboolture South Macgregor Newstead Carole Park Nathan Churchill Redbank Dinmore Richlands Donnybrook Riverview Gailes Robertson Goodna Sandstone Point Inala Stretton Karawatha Sunnybank Kingston Wacol Leichhardt Woodridge Logan Central * note no band 1 suburbs The other sun-belt capital, Perth, has been at the heart of the mining boom that has driven the Western Australian economy in recent times. The distribution of relative deprivation across the city is similar to Brisbane with less extreme deprivation than would be statistically expected. Perth has no suburbs in band 1. It does have suburbs in band 2 including Karawara and Bently, together with Crawley (adjacent the University of Western Australia). The suburbs with lower deprivation include Subiaco and Dalkeith (band 6) and Cottesloe and Leederville (band 5). Spatially, higher relative deprivation tends to be located further from the central city. Table 7: Band 2 and Band 6 suburbs, Perth
Band 2* Band 6 Armadale Kwinana Beach Dalkeith Balga Medina Subiaco Bentley Midvale Calista Mirrabooka Crawley Murdoch Girrawheen Parmelia Karawara Two Rocks Koondoola
• note no band 1 suburbs
21
Figure 5: Relative deprivation, Brisbane
Figure 6: Relative deprivation, Perth
Adelaide and Hobart The two state capitals of Adelaide and Hobart have long had a reputation as Australia’s rust-belt capitals as declining populations and declining economic fortunes have taken a toll on the level of socio-economic opportunity. For some time Adelaide in particular has been considered to be facing significant and entrenched economic disadvantage across several areas. Perhaps unkindly, Carlyon (2003 p. 52) argues that ‘South Australia has spent the past decade hurtling towards economic irrelevance’ a fact that has been reflected in a range of negative socio-economic outcomes, while Peel (1995 p. 3) discussing the northern suburb of Elizabeth says that it ‘is a shorthand for difference and despair, a symbol of what lurks in the darker spaces of a city’s life’. The picture of relative deprivation across Adelaide’s suburbs represents the long standing outcomes of earlier periods of economic, social and demographic change. Adelaide has no suburbs in band 6 (lowest relative deprivation), but does have Eastwood and Toorak Gardens in band 5. Adelaide does, consequently, have a much higher relative proportion of suburbs in band 1 (highest relative deprivation). This is reflected in its regional concentration ratio of 1.16. Suburbs with highest relative derivation are located in the city’s north and include Athol Park, Mansfield Park and Elizabeth Park and are those places that others have referred to when discussing the results of socio-economic transitions within the city (Peel 1995, Baum and Hassan 1993, Baum et al. 2005). Table 8: Band 1 and Band 5 suburbs, Adelaide Band 1 Band 5* Regency Park College Park Angle Park Dulwich Athol Park Eastwood Davoren Park Gilberton Dudley Park Millswood Elizabeth Mount George
22
Elizabeth Downs Northgate Elizabeth North Springfield Elizabeth Park Toorak Gardens Elizabeth South Unley Park Kilburn Walkerville Mansfield Park Ottoway Smithfield Plains Woodville Gardens * note no band 6 suburbs Hobart, like Adelaide has for some time been home to a relatively large socio-economically disadvantaged community. Hobart has no suburbs in band 6 (lowest relative deprivation) or band 5. Band 4 suburbs include Bellerive and Geilston Bay. Hobart has more than its fair share of suburbs in band 1 with a regional concentration ratio of 1.25. Band 1 suburbs include Gagebrook, Clarendon Vale and Bridgewater. Spatially there is no distinct pattern of relative deprivation. The distribution of Hobart deprivation may be seen in figure 7. Table 9: Band 1 and Band 4 suburbs, Hobart Band 1 Band 4* Bridgewater Acton Park Lenah Valley Clarendon Vale Austins Ferry Leslie Vale Gagebrook Battery Point Lindisfarne Bellerive Mount Nelson Bonnet Hill Mount Stuart Cambridge Orielton Cremorne Otago Dynnyrne Ridgeway Fern Tree Sandford Geilston Bay Seven Mile Beach Glebe Taroona Granton Tinderbox Hobart Tolmans Hill Honeywood Tranmere Howden West Hobart * note no band 5 or 6 suburbs
23
Figure 7: Relative deprivation, Adelaide
Figure 8: Relative deprivation, Hobart
Discussion As was pointed out earlier there has been considerable analysis of deprivation and social, economic and spatial differentiation in Australia (Australian Urban and Regional Development Review 1995; Gregory and Hunter 1995; Harding et al. 2002, 2004; Fincher
24
and Wulff 1998; Mitchell and Carlson 2003; Mitchell and Bill 2004, Vinson 2007). The foci of these studies have ranged from questions of segregation, spatial inequality, disadvantage or social polarisation, to the now fashionably-labelled concept of social exclusion and inclusion. While the semantics may differ, such studies have been concerned with much the same general issues, including locating the impact of social and economic transformation within a given spatial and socioeconomic context. This paper has made a contribution to this body of literature by developing a General Deprivation Index for Australian cities. The index provides a snap shot of urban deprivation at the time of the 2006 census and is an indication of the intensity of deprivation across the suburbs of Australia’s capital cities. The discussion presented outlines a specific geography of deprivation and results in a number of questions regarding possible policy approaches. These issues are considered below. One of the key statements taken from the introduction to this paper is that there are real reasons to be concerned about the suburban socio-economic scars that characterise our cities. It is worth restating these concerns against a common view that sees inequality and increases in inequality as simply the dominant, and irresistible, trend in the post-industrial era. Stilwell (2002 p. 1), on the contrary, argues that ‘inequality has significant consequences for economic efficiency, social justice and environmental sustainability. Its sources are properly a central concern for political economic analysis. Its reduction is properly the concern for public policy’. Social discrepancies between places that are advantaged and places that are disadvantaged matter, if for no other reason than the fact that disadvantages are cumulative rather than static. The impact of multiple disadvantages which are further heightened by locational disadvantage works to further heighten the socio-economic damage that confronts some suburbs. This is the theme of a book by Labour MP Wayne Swan and now Federal Treasurer. His view is that the result spatial inequality in Australian society is a compounding problem:
income, employment and economic growth vary by postcodes. As the wealthy take over real estate close to good jobs, the best schools and hospitals they lift the cost of entry to those areas. On the other hand, as the splintering middle and poorer people move further away they are paying more to get to work, school and see the doctor. If they lose their job, distance compounds their disadvantage (Swan 2005 p. 172).
Reflecting on the findings from this paper and taking Wayne Swan’s comments into account, we should be concerned, among other things, that there are whole groups of places that earn an ever larger proportion of their income from government transfers, and places where there are growing proportions of young people out of work. The result is a declining cycle of intergenerational poverty and disadvantage and a continued scarring of suburban Australia.
The geography of disadvantage: diagnosis and prognosis
The socio-economic disadvantage identified in this paper has in a lot of cases been an ongoing feature of Australia’s suburban landscape. Some of the most disadvantaged suburbs have witnessed long term and intergenerational deprivation. Others are part of the new suburbanisation of deprivation that has been alluded to by others (Randolph and Holloway 2005). Regardless of suburban histories individual level advantage and disadvantage gets reflected in local suburbs through the uneven spatial impact on local labour markets, and through the operation of housing markets. In short, broad changes in social and economic life are linked to the circumstances in local communities, because of where particular people live and their roles in society and the economy. For example, some groups are able to exercise a broader choice, across a wide range and diversity of living environments because their economic advantage provides them with the wealth and/or capacity to borrow, enabling them to choose to live in high-cost housing market areas. Others do not possess these economic means and have to make residential choices that are constrained within low-cost housing market locations.
25
But it is not only this differentiation between individuals and households, in the relative constraints within which their housing choices are exercised, that is a significant issue in the social differentiation that is readily discernable across our cities. Rather, it is in addition to the differences in the potential of people to engage in the labour market, influenced by among other things the supply of jobs and the ability of people to tap into new opportunities, that becomes crucial in that social and spatial differentiation. The patterns of variation in advantage and disadvantage across communities and suburban neighbourhoods will therefore reflect a complex set of both individual and societal-scale issues, and in addition will reflect the stages of communities in the transformation from the past to the contemporary economic era. As noted above a large part of the differentiation between suburbs in their performance is explained by economic outcomes. Differences in rates of engagement and disengagement from the labour market appear to be a key driver. Mitchell and Burgess (1999 p.1) argue
unemployment remains one of the most apparent forms of social and economic exclusion in our community. Without a job you are dependent on transfers and savings to survive. You are also stigmatised and you are also likely to be located in poverty and you are also more likely to be subject to health, family break-up and other social problems…The unemployed have few rights, indeed they are a group who can be easily stereotyped and condemned by anecdote.
The industrial and occupation structure as it relates to the transition from the Fordist industrial era to the post-Fordist services and information economy, levels of human capital, and the ability of individuals to tap into particular economic activities are all implicated in the uneven geography of deprivation. This is expressed in the level of monetary resources available to individuals, and at an aggregate level the suburbs they reside in, which in turn tend to structure wider social and demographic outcomes. Thus, single-parent families and non-earner families, lacking human capital and social networks, are more and more separated from meaningful labour market opportunity. Likewise the other demographic transition which involves the concentration of newly arrived immigrants has important implications. The analysis has also suggested that space and location are important. In situations where deprivation is spatially concentrated the impact of multiple disadvantages may act to further exacerbate negative outcomes through the influence of what are fashionably referred to as neighbourhood effects. A big picture view of disadvantage suggests that there exists a complex geography of deprivation across the Australian metropolitan regions. The divisions identified in the research presented here are most noticeable at the extremes— between the band 1 highest relative deprivation suburbs and the band 6 lowest relative deprivation suburbs. The group of suburbs identified within band 6 have for some time been identified as an emerging feature of Australian cities (See Baum et al 1999, 2002; Stimson et al. 2001). They are now however clearly differentiated from other localities and are establishing themselves as spaces of privilege in the global age (Baum et al. 2002). These places of privilege are not found in all metropolitan regions but are a distinct feature of Australia’s two most globalised cities Sydney and Melbourne. In Sydney, increasing global city functions associated with the presence of regional headquarters, national gateway functions and knowledge based industry have generated the wealth and high incomes necessary for the rise of ‘privileged communities’ close to the CBD in is now known as ‘global Sydney’. While there is little evidence of the high status citadels referred to in international literature (Marcuse and van Kempen 2000a) the communities of global Sydney are seen as being increasingly differentiated from other communities and localities often with symbolic walled suburbs as a
26
defence against crime and deprivation (McGregor 1997; Connell and Thom 2000; Gleeson 2004). At the other extreme, band 1 suburbs show the negative impacts of economic restructuring, and other social and demographic processes including changes in household structure and immigration, and are among the most disadvantaged places of the metropolitan region. Some are part of urban areas that prospered with the industrial development of earlier periods but which have, in the wake of deindustrialisation, fallen on hard times. For example, some areas of Adelaide are among the nation’s poorest suburbs having emerged as housing for prosperous workforces in the 1950s only to suffer profound demise through the 1970s and 1980s (Peel 1995; 2003) and other cities have similar examples. If the band 6 suburbs are in some sense distancing themselves from the rest of the country, then the band 1 suburbs are falling further and further behind new economic opportunity. This would seem to be especially the case with reference to those band 1 suburbs within the Sydney metropolitan region. While it is true that disadvantage in terms of unemployment and low incomes have been a feature of Australian cities for some time, these contemporary deprived localities may be thought of as being different with deprivation being more entrenched and less easy to escape. If these places are reflective in some senses of the ghettos of exclusion discussed by Marcuse and van Kempan (2000a p.19) then the following observation is pertinent
…the residents of the abandoned city, particularly in the new ghetto of the excluded, play a different role from those of the old ghetto in many respects….Older forms of the ghetto…remained an integral part of the mainstream economy, with residents of different classes and with a variety of prospects on the labour market. Their residents, when unemployed, were part of a reserve army of the unemployed, who had expectations of re-entering the mainstream labour force when conjunctural conditions changed. That holds less and less in the new ghettos.
The band 1 suburbs are Australia’s equivalent of the urban ghettos in the United States. The social and economic conditions might be at variance, but these are excluded places all the same and are likely to face the same issues of dependence, limited political participation, restricted socio-cultural integration, bad neighbourhood pathologies and, above all, increasing deprivation (Gregory and Hunter 1995; Musterd and Ostendorf 1998; Glesson 2002). In short, the prognosis of ongoing disadvantage deserves our full policy attention.
Healing suburban disadvantage This paper has already discussed the reasons why we need to be concerned about the uneven spatial outcomes that have been identified. It is now appropriate to consider something about policy outcomes. How should we begin to think about the inequalities we have identified and what input can we make in terms of policy questions and approaches? It is clear that to be successful any policy approaches must sit across a multiple of facets. Randolph (2004, p. 65) discussing spatially concentrated deprivation in particular notes that
Social exclusion which manifests itself in specific localities through the concentration of multiply disadvantaged households requires a full range of targeted and mainstream (general welfare and fiscal) policy interventions….
Approaches need to be cognisant of both policy impacts on individuals and policy impacts on the places they live. A large body of literature has formed around the question of possible policy responses to spatially concentrated socio-economic deprivation. Many of these have had a particular place focus, recognising the extent of spatially concentrated deprivation. Past policies have included the post- World War II public housing estate construction and the more recent inner-city renewal projects such as Darling Harbour in Sydney or docklands in Melbourne
27
associated with city regeneration strategies. One particular approach has been to promote the concept of social mix, whereby the mixing, in a given space, of people of diverse social class, socioeconomic status, social category, stage in life cycle and so forth is facilitated to overcome, among other things, so-called concentration effects (see Arthurson 2002; Johnston 2002). In a similar vein is has been suggested that diversifying housing tenure to promote greater social mix, will mitigate problems associated with the concentration of disadvantage. While this policy formed part of the operational management plans of a number of state government housing authorities (Wood 2002), many have questioned the efficacy of such policies to create ‘inclusive, cohesive and sustainable communities’ (Arthurson 2002 p. 258). Another prominent approach is found on the premise that the root cause of the problem of unequal socio-economic outcomes and concentrated disadvantage across cities and regions is a lack of jobs and that job opportunities are, in certain circumstances, spatially limited. What are important are the interactions between where people live and the wider metropolitan system including the operation of local labour markets. The unemployed and other marginalised workers cannot get jobs if there are none available within reasonable distance of their homes. The consequent solution is located not in social mixing, but in the complex links between housing and job location in the wider metropolitan development process (Randolph 1991; Healy and O’Connor 2001; Moriarty 2002; O’Connor and Healy 2004). Local spatially sensitive employment safety nets are one suggested approach. A local job guarantee would see the government acting as an employer of last resort creating jobs when there was a net local shortage. This would amount to replacing work-for-the-dole schemes with on-going guarantees which provide income security to anyone unable to find work elsewhere. Those involved would be paid a safety net wage and work on local community or environmental projects. Such a policy would provide positive local demand injections that would go some way towards reducing the unemployment burden with only a small increase in cost. The sense of worth would go a long way towards addressing issues of social inclusion and local community strengthening. In marked contrast to policy arising from the conception of a social geography of deprivation are measures aimed at the individual. These policies recognise that a focus on space and place may not be enough or have more lately jettisoned the idea that space matters altogether. The last decades has seen the substantial institution of social policy requiring or empowering individuals to exercise market choices in areas of traditional state provision. Thus, for example, rather than providing capital funding for the construction of public housing in fixed locations, the Commonwealth and the states now prefer to provide low-income households with a housing rental subsidy. This subsidy can be used to rent a house where the recipient so chooses, within of course, the constraints of significant spatial and other variations in housing market costs. At the level of suburbs, engagement with the new economy and individual human capital assets are seem as key indicators of success. This suggests a need to enhance the ability of individuals to take advantage of growing opportunities. Individual job training schemes or skills training are a response to this. The people-focus on assistance programs is also seen in recent requirements by the Commonwealth Government for the payment of unemployment benefits to discourage relocation away from job opportunities or in the move to provide broad skills training so as to increase individual employability assets. Many of the existing arguments for what might or might not work in addressing unequal outcomes across the metropolitan cities take a very specific focus. In general however there is a tension between those which are place based and those which are people based. The place based versus people based debate goes back to the work of Winnick (1966) and has more recently been considered in the work of Bolton (1992), O’Connor et al (2005), Baum et
28
al. (2005) and Partridge and Rickman (2006). The focus on place involves dealing directly with places when policy is designed and implemented. In contrast, people based policy assumes a spatially indifferent world of individuals. However, different human beings, based on their individual capacities, can experience the same geographic context in profoundly different ways generating considerable variation in socio-economic outcomes. If individual capacities are the significant drivers of social outcome (i.e. if geography is less important) then the appropriate policy response may be to focus on improving individual capacities or assets. Alternatively individual capacities may be less important in comparison to the geographic context in determining socio-economic outcomes. Here, a different set of policies focusing on improving opportunities and outcomes based on where people live and their spatial interactions would be indicated. A third, more likely, alternative focuses on both individual capacities and the impact of the geographic context. The golden mean, in short, considers ways in which a combination of people and place based policies can aid in mitigating the negative impacts of social disadvantages. The question for policy makers pivots around the mix of people and place based policy. To this end there has been increased research into how a people and place focus allows for a better response to the questions surrounding deprivation and disadvantage. This international literature seeks to link a myriad of changes from the global to the local, or the macro to the micro, and understand the effects of those changes on human life (Sampson et al. 2002, Friedrichs et al. 2005). This agenda focuses on understanding the impacts of neighbourhood effects or area effects or more broadly the impacts of interactions between people and place. In particular, the research establishes that understanding the effects of people and place is crucial as individuals, households and local communities face continued economic restructuring and large-scale social and demographic change and as policy makers and researchers try to understand the impacts and outcomes of these changes. The recently published collection of research undertaken in Europe and the USA (Friedrichs et al. 2005) together with earlier work including Brooks-Gunn et al. (1997) clearly illustrates these concerns and points to the need for further research and policy debate. Within Australia, the work by Baum et al. (2007) illustrates the ways in which individual factors including human capital and family background are important for understanding labour market outcomes. Importantly however they also show that the strength of local labour markets matter in determining the distribution of negative labour market outcomes. They argue strongly for a policy focus that addresses both people and place.
Concluding comments The analysis and discussion presented in this paper illustrates the nature of socio-economic deprivation—the suburban scars—of our major urban centres. The interpretation of these patterns and the divides that are seen to exist should help to sharpen policy thinking about traditional and more recent policy approaches. Clear and fresh thinking is required regarding the later so as to achieve the most appropriate and effective policy actions. Given the entrenched nature of deprivation, there is a strong argument to be made that past policies have largely failed or at least been only partially successful. To drive new policy, however, a strong evidence base is also required. So is a multi-pronged approach. We clearly need both a political focus on Australia’s deprived suburbs as well as debate on the appropriate policy. More importantly we need to realise that Australia is far too prosperous to continue failing its most deprived citizens and that real and sustainable action is required to address the socio-economic scars that exist in our metropolitan cites.
29
30
References Arthurson, K. (2002) Creating inclusive communities through balancing social mix: a critical
relationship or tenuous link? Urban Policy and Research 20: 245-261
Australian Urban and Regional Development Review. (1995) Urban Australia: Trends and Prospects, Research Report No. 2. Canberra: Commonwealth Department of Housing and Regional Development
Badcock, B., (1997) Restructuring and spatial polarization in cities, Progress in Human Geography, 21, 2, 251-262
Baum, S. (2004) Measuring socio-economic outcomes in Sydney: an analysis of census data using a general deprivation index, Australasian Journal of Regional Studies, 10, pp.105 -133
Baum, S. and Hassan, R. (1993) Economic restructuring and spatial equity: A case study of Adelaide, The Australian and New Zealand Journal of Sociology, 29, 2, pp. 151-172.
Baum, S. Stimson, R. O’Connor, K. Mullins, P. and Davis, R. (1999) Community Opportunity and Vulnerability in Australia’s Cities and Towns: Characteristics, Patterns and Implications. Brisbane: University of Queensland Press.
Baum, S. Mullins, P. Stimson, R. and O’Connor, K. (2002) Communities of the post-industrial city, Urban Affairs Review, 37, 2, pp. 322-357.
Baum, S., O’Connor, K. and Stimson, R. (2005) Fault Lines Exposed: Advantage and Disadvantage across Australia’s Settlement System, Monash University ePress, Clayton.
Baum, S. Haynes, M. Han J. H. & van Gellecum Y. (2006) Advantage and disadvantage across Australia’s extended metropolitan regions: A typology of socio-economic outcomes, Urban Studies, 43, 9, 1549-1579
Baum, S. Mitchell, W. and Bill, A. (2007) Labour Underutilisation in Metropolitan Labour Markets: Individual and Contextual Factors, Centre of Full Employment and Equity, Working Paper07-17, University of Newcastle
Beer, A. and Foster, C. (2002) Global restructuring, the welfare state and urban programmes: federal policies and inequality within Australian cities, European Planning Studies, 10, 1, pp. 7-25.
Benassi, D. Kazepov, Y. and Mingione, E. (1997) Socio-economic restructuring and urban poverty under different welfare regimes, in: F. Moulaert and A. Scott (Eds.) Cities, Enterprises and Society on the Eve of the 21st Century, pp. 174-215. London: Pinter.
Bentham, C. (1985) Which areas have the worst urban problems? Urban Studies, 22, pp. 119-131.
Birrell, B. (1993) Ethnic concentrations: the Vietnamese experience. People and Place, 1, pp. 26-31.
Bolton, R. (1992) Place prosperity revisited: An old issue with a new angle. Urban Studies 29 (2): 185-203.
Bradford, M. Robson, B. and Tye, R. (1995) Constructing an urban deprivation index: a way of meeting the need for flexibility, Environment and Planning A, 27, pp. 519-533.
Brooks-Gunn, J. Duncan, G. & Aber, J. (1997) Neighbourhood Poverty: Vol 1 Context and Consequences for Children, Russel Sage foundation, New York.
Burchell, D. (2002) ‘The western Sydney factor’. Australian Policy Online. [cited 2004 Mar 11] Available from: http://www.apo.org.au.
31
Carlyon, P. (2003) ‘Vanishing state’, The Bulletin. Sydney, Dec 2, pp: 52–54.
Carstairs V. and Morris R. (1989) Deprivation and mortality: an alternative to social class? Community Medicine, 11, 3, pp. 210-219.
Cass, b. (1990) Expanding the concept of social justice: Implications for social policy reform, Social Issues in Town Planning, Proceedings of conference held at the University of New South Wales, February
Chakravorty, S. (1996) A measurement of spatial disparity: The case of income inequality, Urban Studies, 33, 9, pp. 1671-1686.
Dear, M. (2000) The Post-Modern Urban Condition. Lackwell, Oxford
DETR (Department of the Environment Transport and the Regions) (2000) Indices of Deprivation 2000. Regeneration Research Summary No. 31, DETR: London.
Engels, B. (2000) 'City make-overs: The place marketing of Melbourne during the Kennett years, 1992-99', Urban Policy and Research, 18:4, 469 - 494
Fincher, R and Wulff, M. (1998) ‘The locations of poverty and disadvantage.’ In Fincher R. and Nieuwenhuysen, J, editors. Australian Poverty: Then and Now. Melbourne University Press, Melbourne pp.144-164.
Forrest, J. and Poulsen, M (2003) Multiculturalism and the Spatial Assimilation of Migrant Groups: The Melbourne and Sydney Experience, paper presented at the State of Australian Cities Conference, Carlton Hotel, Parramatta, 3-5 December.
Friedrichs, J. Galster, G. & Musterd, S. (eds) (2005) Life in poverty neighbourhoods: European and American perspectives, Routledge, London
Gleeson, B. (2004a) The suburbs. Fear in the tank, hope on the horizon. What’s driving
suburban Australia Australian Policy On-line. [cited 2005 July 8] Available from:
www.apo.org.au.
Gleeson, B. (2004b) The Future of Australia’s Cities: Making Space for Hope. Professorial Lecture, Griffith University; February 19
Gleeson, B. (2006) Australian Heartlands: Making space for hope in the suburbs, Allen and Unwin, Sydney
Gregory, R. and Hunter, B. (1995) The Macroeconomy and the Growth of Ghettos and Urban Poverty in Australia. Discussion paper no. 325, Centre for Economic Policy Research, Canberra: Australian National University.
Gwyther, G. (2002) Socio–spatial Differentiation and the Master Planned Community in a Global City. Research paper no. 8, Urban Frontiers Program. University of Western Sydney, Sydney
Harding, A. Greenwell, and H. Blake, M. (2002) AMP:NATSEM Income and Wealth Report, Issue 1. AMP Life Limited.
Harding, A. Yap, M. and Lloyd, R. (2004) AMP:NATSEM Income and Wealth Report, Issue 8. AMP Life Limited.
Hamnett, C. (2003) Unequal City: London in the Global Arena. Routledge, London
Hawley, J 2003 ‘Be it ever so humongous’. The Sydney Morning Herald, Good Weekend Supplement. Sydney: 23 August: pp. 24–31.
Healy, E. and O’Connor, K. (2001) Jobs and housing location in Melbourne 1986–1996. New insights on metropolitan development, Australian Planner, 38(1), pp. 9–15.
32
Hunter, B. and Gregory, R. (1996) An exploration of the relationship between changing inequality of individual, household and regional inequality in Australian cities, Urban Policy and Research, 14, 3, pp. 171-182.
Johnston, C. (2002) Housing Policy and Social Mix: an Exploratory Paper. Shelter NSW Sydney
Kesteloot, C. (1998) The geography of deprivation in Brussels and local development strategies, in: S. Musterd and W. Ostendorf, (EDs.) Urban Segregation and the Welfare State: Inequality, and Exclusion in Western Cities. Routledge, London
Kitchen, P. (2001) An approach for measuring urban deprivation change: the example of East Montreal and Montreal urban community, 1986-96, Environment and Planning A, 33, pp. 1901-1921
Langlois, A. and Kitchen, P. (2001) Identifying and measuring dimensions of urban deprivation in Montreal: An analysis of the 1996 census data, Urban Studies, 38(1), pp. 119-139.
Marcuse, P. (1997) The enclave, the citadel and the ghetto: What has changed in the post-Fordist U.S. city? Urban Affairs Review, 33, 2, 228-264.
Marcuse, P. (1989) Dual city: A muddy metaphor for a quartered city, International Journal of Urban and Regional Research, 13, 4, pp. 697-708.
Marcuse, P. and Van Kempen, R. (2000a) Introduction, in: P. Marcuse and R. Van Kempen (EDs.) Globalizing Cities: A New Spatial Order. pp. 1-21. Blackwell, Oxford.
Marcuse, P. and Van Kempen, R. (2000b) Conclusion: A changed spatial order, in: P. Marcuse and R. Van Kempen (EDs.) Globalizing Cities: A New Spatial Order. pp. 1-21. Blackwell, Oxford.
Midgley, J. Hodge, I. and Monk, S. (2003) Patterns and concentrations of disadvantage in England: A rural-urban perspective, Urban Studies, 40, 8, 1427-1455.
Mikelbank, B. (2004) A typology of U.S. suburban places, Housing Policy Debate, 15, 4, pp. 935-964.
Mingione, E. (1996) Urban poverty in the advanced industrial world: Concepts, analysis and debates, in: E. Mingione (ED.), Urban Poverty and the Underclass: A Reader. pp. 3-40. Blackwell Publishers, Oxford.
Mitchell, W. and Bill, A. (2004) ’Spatial dependence in regional unemployment in Australia.’ In Carlson E, editor. A Future that Works: Economics, Employment and the Environment. Proceedings of the 6th Path to Full Employment Conference and the 11th National Conference on Unemployment. University of Newcastle, Newcastle pp .312-326.
Mitchell, W. and Burgess, J. ‘Unemployment should be no longer tolerated: there is an easy solution’. On Line Opinion. [1999 Oct15]. Available from: http://www.onlineopinion.com.au /print.asp?article=1380.
Mitchell, W; Carlson, E. (2003) Common Trends and Cycles in Disaggregated Regional Employment Growth Rates. Working Paper 03-08, Centre of Full Employment and Equity. University of Newcastle, Newcastle.
Moriarty, P.(2002) ‘Environmental sustainability of large Australian cities’. Urban Policy and Research 20: pp. 233-244.
Musterd, S. and Ostendorf, W. (1998) (EDs.) Urban Segregation and the Welfare State: Inequality, and Exclusion in Western Cities. London: Routledge.
O’Connor, K and Healy, E. (2004) Rethinking suburban development in Australia: a Melbourne case study. European Planning Studies 12: pp. 27-40.
33
O’Connor, K. Stimson, R. and Daly, M. (2001) Australia’s Changing Economic Geography: A Society Dividing. Oxford University Press, South Melbourne.
Organisation for Economic Cooperation and Development. (1998) Integrating Distressed Urban Areas, Organisation for Economic Cooperation and Development, Paris
Partridge, M. and Rickman, D. (2005) The Geography of American Poverty: Is There a Need for Place-Based Policies? Upjohn Institute, Michigan
Peel, M. (1995) Good Times, Hard Times: The Past and the Future in Elizabeth. Melbourne University Press, Melbourne.
Pusey, M. (2003) The Experience of middle Australia: the dark side of economic reform. Cambridge University Press, Melbourne
Rahman, M. Palmer, G. Kenway, P. and Howarth, C. (2000) Monitoring Poverty and Social Exclusion 2000, Joseph Rowntree Foundation: York.
Randolph, B. (1991) Housing labour markets and discontinuity theory’. In Allen, J; Hamnett, C. (editors) Housing and Labour Markets. Building up Connections. Unwin Hyman, London pp. 16-51.
Randolph, B. (2004) The changing Australian city: New patterns, new policies and new research needs, Urban Policy and Research, 22, 4, pp. 481-493.
Randolph, B. and Holloway, D. (2005) The suburbanisation of disadvantage in Sydney: new problems, new policies, Opolis, 1, 1, pp. 49-65
Raskall, P. (1995) Who Gets What Where? Spatial Inequality Between and Within Australian Cities. Paper presented to the Commonwealth Department of Housing and Regional Development Seminar on Aspects of Spatial Inequality, Canberra.
Raskell, P. (2002) Equity issues in the new prosperity of the 1990s: the case of Sydney, Australian Geographer, 33, 3: 281-299.
Rhein, C. (1998) Globalisation, social change and minorities in metropolitan Paris: The emergence of new class patterns, Urban Studies, 35, 3, pp.429-447.
Sampson, R. Morenoff, J. & Gannon-Rowley, T. (2002) Assessing neighbourhood effects: social processes and new directions in research, Annual Review of Sociology, 28, 443-478
Sassen, S. (1991) The Global City: New York, London, Tokyo, Princeton University Press, Princeton, NJ
Searle, G. (1993) The Spatial Restructuring of Sydney’s Economy, 1981-1991: Case Studies of 12 Industries/Activities. Paper presented at Regional Science Association Conference (ANZ Section)
Sloggett A. and Joshi, H. (1998) Indicators of deprivation in people and places: longitudinal perspectives, Environment and Planning A, 30, pp. 1055-1076.
Soja, A. (1997) Margin/alia: Social justice and the new cultural politics, in: A. Merrifield and E. Swyngedouw (EDs.) The Urbanization Injustice. pp. 180-199. New York: New York University Press.
Soja, E. (2000) Postmetropolis. Oxford: Blackwell.
Stilwell, F. (1989) Structural change and spatial equity in Sydney. Urban Policy and Research, 7, 1, PP. 3-14.
Stilwell, F. (2002) Why bother about economic inequality? On Line Opinion, http://www.onlineopinion.com.au/view.asp?article=1866 [accessed July 2005].
34
Stilwell, F. and Hardwick, M. (1973) Social inequality in Australian cities, Australian Quarterly, 45, pp. 18-36.
Stimson, R. (2001) Dividing societies: The socio-political spatial implications of restructuring in Australia, Australian Geographical Studies, 39, 198-216
Stimson, R. Baum, S. Mullins, P. and O’Connor, K. (2001a) A typology of opportunity and vulnerability in metropolitan Australia, Papers in Regional Science, 80, pp. 45-66.
Stretton, H. (1970) Ideas for Australian Cities. Melbourne: Georgian House.
Swan, W. (2005) Postcode: the Splintering of a Nation. Pluto Press, North Melbourne.
The Australian (2006) Rudd’s first speech as Labor Leader, December 05, 2006, www.theaustralian.news.com.au/story/ 0,20876,20876230-601,00.html [accessed December 12 2006]
Theodorson, G. (1982) (ed) Urban Patterns: Studies in Human Ecology, Revised Edition. University Park, Pennsylvania: The Pennsylvania State University Press.
Timms, D. (1970), The Urban Mosaic: Towards a Theory of Residential Differentiation. Cambridge: Cambridge University Press.
Vinson, T. (2007) Dropping Off The Edge: The Distribution of Disadvantage in Australia, A report of Jesuit Social Services and Catholic Social Services Australia, Melbourne.
Walks, R. A. (2001) The social ecology of the post-Fordist/global city? Economic restructuring and socio-spatial polarisation in the Toronto urban region, Urban Studies, 38, 3, pp. 407-473.
Wessel, T. (2001), Losing control? Inequality and social divisions in Oslo, European Planning Studies, 9, 7, pp. 889-906.
Williams, C. and Windebank, J. (1995) Social polarisation of Households in contemporary Britain: A ‘whole economy’ perspective, Regional Studies, 29, 8, pp. 723-728.
Winnick, L. (1966) Place prosperity vs. people prosperity: Welfare considerations in the geographic redistributions of economic activity. UCLA Real Estate Research Program, pp. 273-83.
Wood, M. (2002) A Balancing Act? Tenure Diversification in Australia and Britain. Paper presented at the Housing Studies Association Conference, York.
Yigitcanlar, T., (2005), ‘The making of knowledge cities: lessons learned from Melbourne’, In the proceedings of the International Symposium on Knowledge Cities, 28-30 Nov 2005, The World Bank and Arab Urban Development Institute. Medina, Saudi Arabia, 1-18.
35
