July 2005

Draft. Comments welcome

Are occupational choices affecting housing choices?

Judith Yates*

School of Economics and Political Science

University of Sydney

Paper prepared for the Australian Social Policy Conference 2005, Looking back, Looking Forward, A Quarter of a Century of Social Change, University of New South Wales, 20-22 July 2005

*The work presented here forms part of a larger study funded by AHURI and being undertaken in collaboration with Bill Randolph and Darren Holloway. Whilst they are not responsible for the use to which the results of this larger study have been put in this paper, they have contributed significantly to the process of conceptualising the issues and analysing the data.

This material was produced with funding from the Australian Government and the Australian States and Territories. AHURI Ltd gratefully acknowledges the financial and other support it has received from the Australian, State and Territory governments, without which this work would not have been possible.

ABSTRACT

Are occupational choices affecting housing choices? In both the UK and US, many of the jobs for lower paid workers in occupations considered important to the proper functioning of the city are in areas of high housing costs. This has led to concerns that employers are experiencing recruitment and retention problems for such workers.

This paper will provide some early results from an AHURI research project on the interaction between housing and labour markets in Sydney. The results will focus specifically on affordability outcomes in 2001 and changes between 1996 and 2001 in the residential and employment locations of low and moderate income households, with a focus on those in specific occupations (nursing, computing, cleaning and hospitality workers) selected as "indicator occupations" to reflect the emergence of what in the UK and US has been called the "key worker" or "essential worker" problem.

They are occupations on the general skills shortages lists in Australia at present. They are also occupations for which there is expected to be a growing demand in the future as a result of the continuing economic restructuring from old to new economy activities and from production of goods to service provision. At the same time, they are occupations which include both low and moderately paid workers as well as better paid workers and for which there is likely to be a quite different geographical spread of employment opportunities and, as a result, of housing opportunities.

key words: housing affordability, spatial mismatch, key worker, residential location

1

Are occupational choices affecting housing choices?

“Central Sydney is being starved of basic-wage workers - those who clean offices and hotel rooms, wait on tables, work in kitchens - because they cannot afford to live within reasonable travelling distance from the city.” (O'Malley, Sydney Morning Herald, 2nd June, 2005).

Introduction The research reported in this paper was motivated by a concern such as that expressed in the above quote from a recent press report: viz. by a concern that high housing costs in central city regions are excluding many lower paid workers from jobs in those regions. Such concerns suggest labour market shortages in some occupations can be attributed to shortages of affordable housing in locations that provide ready access to the central city. Because many of the jobs for lower paid workers are in occupations considered important to the proper functioning of the city, the recruitment and retention problems that employers experience potentially can have broader ramifications through their negative impact on the competitiveness of the city. An associated policy issue, therefore, is whether affordable housing and affordable housing policies should be targeted to workers in the specific occupations where problems have been identified.

This paper has four specific aims. The first is to place concerns about the impact of housing costs on labour market outcomes into a broader context relating to structural change over the past twenty-five years; the second is to review the US, UK and Australian literature that focuses on what more generally is described as the key worker or essential worker problem in order to determine whether there is agreement on which occupations are central to the broader debate on the issues raised in the press report quoted. The third is to raise some of the conceptual issues that arise in attempting to relate occupations and housing outcomes. The final aim, and the key contribution of this paper, is to provide new evidence for Sydney on the housing affordability problems faced by workers in specific occupational categories.1 Its focus is on the affordability outcomes of employed persons or, more specifically, of households in which at least one person is in employment.

Background Concerns about the impact of high housing costs on labour market outcomes in the centre of Sydney have emerged with the increasing pressures associated with the economic restructuring and growth of new jobs associated with Sydney's emergence as a global city. In Newman’s words, “Increasingly these jobs are occurring within city centres, and inner city areas, and they are promoting a re-urbanisation process, a resurgence of the inner city and central city living, that is associated with the need to be part of this global city phenomenon.” (Newman, ABC Background Briefing, September 7th 1997)

1 The larger study of which this paper is a part provides data for Australia as a whole and for Melbourne and Brisbane similar to those presented here for Sydney (Yates et al, 2005). The results presented here represent the first two stages of this study. A third stage will focus specifically on relating results for individual workers to their household circumstances.

2

Economic change associated with labour market reform has resulted in an increased casualisation of employment, a growth of part-time rather than full-time jobs and increased earnings disparities.2 At the same time, there has been strong growth both in the number of people employed in the highly skilled professional group and in the numbers employed as relatively unskilled clerical, sales and services workers (Watson et al., 2003). This is consistent with the Reichian notion of the symbiotic relationship between the growth of well-paid symbolic analysts or knowledge workers in new economy jobs and the lower paid in-person service workers such as retail sales workers, hospitality workers, cleaners, health care workers and the like whose services tend to be person to person, cannot be sold world wide and have not been eliminated by the restructuring that has taken place (Reich, 1991). These changes have contributed to a significant polarisation of household incomes (Saunders, 2003, 2005). Social change, which has resulted in increases in the incidence of single person and of dual income households, has compounded the effects of these economic changes.

Income and occupational polarisation is one aspect of the global city-social polarisation thesis in Reich's work. Excellent overviews of the broader literature on this are provided in Dodson (2004) and, in the context of Sydney, Baum (1997). Baum specifically highlights the shift in occupational structure which has resulted in the growth of high paid professionals associated with new economy jobs along with a cluster of lower paid service workers and suggests this shift "results in an increasing polarisation of the occupational structure as jobs become concentrated in the high-skill/high-status and low-skill/low-status sectors" (Baum, 1997, p1885). Stimson (2001) highlights the extent to which the growth in Reich's symbolic analysts between 1986 and 1996 has been concentrated in Sydney and has been closely associated with re-urbanisation in the inner suburbs of Australia's big cities.

Economic and social change also has contributed to significant spatial polarisation in housing markets. As Hamnett (1994, p401) has argued, occupational polarisation “is linked to changes in housing demand leading to a gentrification of parts of the inner city and to a concentration of the less skilled in the less desirable parts of the housing market. Thus, occupational polarisation is accompanied by growing social, tenurial and ethnic segregation." Sassen (1991:185) likewise has suggested that the rapid growth and concentration of high income workers has contributed to the rapid growth of high-priced real estate market with a premium for central locations. Randolph (1991) provides an early analysis of why such outcomes might arise. Winter and Stone (1998, 1999) and Yates (2002) provide more recent evidence for Australia. Within Sydney, the gap between dwelling prices in central and outer locations has increased. At the same time, the supply of low rent housing has decreased and the bulk of that which remains is located well away from the city centre.3

The impact of widening house price differentials on employment opportunities has been a key concern of the spatial mismatch literature for well over a quarter of a century (reviews of which can be found in Ihlanfeldt and Sjoquist, 1998 and Kain, 1992).4 This (largely US 2 The proposals to continue with industrial relations reform and to change workplace relations that are currently in place in Australia suggest these trends will continue. 3 Data provided in the Appendix illustrate these claims for Sydney. 4 A number of relatively recent papers in Urban Studies have uses spatial-skill-mismatch (Ong and Blumenberg, 1998 or Immergluck, 1998) or skills mismatch (Stoll, 2005) rather than spatial mismatch or have distinguished between spatial mismatch and skills mismatch (Houston, 2005). Whilst these terms may have different implications in terms of residential and occupational mobility, these difference are not germane to this paper.

3

based) literature focuses on the adverse employment outcomes for low-paid (Black) workers and is predicated on the assumption of a steady dispersal of jobs (and particularly low skilled jobs) from central cities to the suburbs and a concentration of minority workers in inner city ghettoes. In broader terms, however, it applies to cases where there is a spatial disconnection between residential and work locations for particular workers. In other words, it applies equally well to the case where to the case where jobs are concentrated and where low-paid workers are dispersed.

Gobillon et al (2003, p21) provide an overview of a number of underlying mechanisms that potentially explain how distance to job opportunities could be harmful and develop theoretical arguments to support the main intuitions embodied in these. Other than those that directly relate to discrimination, the potential mechanisms are as follows:

i. The efficiency of job search decreases with distance to jobs because workers obtain less information about distant job opportunities or firms resort to local recruiting methods.

ii. Incentives may be too low for workers residing far away from jobs to search intensively. If dwelling rents are sufficiently low or search costs sufficiently high, workers may be discouraged from searching.

iii. Workers may refuse jobs that involve too long commutes because commuting is too costly relative to the wages paid. As a result, they may restrict their job searches to their residential neighbourhood.

iv. Transport is inadequate, thus reinforcing the search and commuting costs at (ii) and (iii).

v. Employers may refuse to hire, or may pay lower wages, to distant workers because commuting long distances makes them tired, more likely to be absent and hence less productive.

The first mechanism listed, focuses on both the demand for and supply of labour; the second and third mechanisms, reinforced by the fourth, focus on labour supply; the fifth focuses on labour demand.

Smith and Zenou (2003) develop a similar framework. They suggest spatial mismatch can be the result of optimizing behavior on the part of workers who may choose low amounts of search and long-term unemployment if they reside far away from jobs. They choose not to relocate closer to jobs because the short-run gains (represented by low land rent and large housing consumption) are big enough compared to the long-run gains of residing near jobs (represented by higher probability of finding a job). In other words, distance to jobs is seen as the main culprit.

In broad terms, the concerns raised by the international literature both on global cities and on spatial mismatch, despite starting from different geographies about the changing location of jobs and workers, form the basis of a rethinking of the focus of affordable housing policies in the UK and US as well as in Australia. Recent affordable housing debates have broadened from focussing on targeting the most needy (often those excluded from the labour market) to including workers seen as essential to the efficient operation of the local economy but at risk of being squeezed out of local housing markets by high housing costs.

In the UK, affordable housing initiatives have been targeted on a group of what have been called key workers following a Housing Green Paper commitment (DETR, 2000) to help

4

them buy homes in high demand, high price areas so that they can live within or near the communities they serve (Renewal, 2002). In the US, the Urban Land Institute (ULI) Workforce Housing Forum was convened because “the lack of affordable housing in urban areas is leading many households to locate far from their jobs, creating all the problems associated with sprawl, including traffic congestion, air pollution, environmental degradation, and requests for public funds to be used for the construction of new roads, schools, libraries, etc. In some areas, the lack of workforce housing has become an economic development issue as corporations decide not to locate in areas where their employees cannot acquire decent, safe, and affordable housing." (Haughey, 2001, p2). In other words, concerns with affordability problems for those in the workforce extend beyond the difficulties faced by employers in recruiting and retaining staff. They extend to issues of urban form, the environment, public infrastructure and local economic development.

This broad overview suggests there are a number of structural reasons why high cost locations may be starved of essential workers and a number of reasons why this might be seen as a problem. It also provides some insights into the processes that might lead to the need for, but exclusion of, particular groups of workers within global cities. The following section more specifically on who these essential or key workers might be.

The key worker/essential worker literature One of the first questions to be addressed is which low-paid jobs are seen as essential for the proper functioning of the city. In some sense it could be argued that, almost by definition, all jobs satisfy this criterion. Much of the UK and US and the emerging Australian literature, however, has started from a pre-determined definition of who are key workers or essential workers.5

UK literature

Early key worker affordable housing policy initiatives in the UK, for example, were targeted at teachers, health workers and the police (DETR, 2000) and have since been extended to a broader range of public sector workers, a full list of which is available in the relevant documents from the Office of the Deputy Prime Minister (ODPM) (for example, ODPM, 2004a). The fact that the list was extended to include planners in London and fireman in Hertfordshire underlines a political aspect to the selection.6 In general terms, a key worker in the UK is someone who is employed by the public sector; in a frontline role delivering an essential public service; or in a sector where there are serious recruitment and retention problems (ODPM, 2004b). In specific terms, affordable housing policies are directed at key workers in London and the South East, where housing has become increasingly unaffordable; where problems are being encountered recruiting and retaining key workers; where the social rented sector is unable to provide low cost housing for rent or shared ownership; and where concerns have been raised about the staffing of key services (Renewal, 2002).

5 A complete list of the occupations covered by the UK, US and Australian literature can be found in Yates et al (2005a). 6 The former were added in response to political pressure after a series of strikes in 2003 by the politically aggressive Fire Brigades Union which resulted in the army being called in to provide firefighting services (http://news.bbc.co.uk/1/hi/uk/2704501.stm). In London, planners were added to the key worker list because Keith Hill, the Minister for Planning at the time, was also the Minister for London.

5

Monk et al (2002) take a broader view to defining the target group for affordable housing policies. They suggest that the most relevant definition of key workers relates to their role in the local economy, as either employment in essential services or in the growth industries required to sustain the local economy into the future. There are also other signs of acceptance of the need for a broader definition as a result of a blurring of the distinction between public and private sector jobs arising from outsourcing and privatization (eg ATIS Real Weatheralls 2002) and because of concerns that local economic development and the health of the local economy could be threatened by recruitment and retention difficulties within the private sector (for example, Fordham Research, 2005). This latter rationale suggests generalization of the definition at a national level may be problematic.

US literature

A broader definition of key workers is closer to that employed in one of the several strands of the equivalent US literature. The first strand comes from a national level perspective and is reflected in the work done by the Essential Worker Immigration Coalition (EWIC). EWIC is a coalition of businesses, trade associations and other organizations across the industry spectrum concerned with the shortage of both skilled and lesser skilled ('essential worker') labour. The second comes from a local perspective and is reflected in the work done by the National Housing Conference (NHC) and by the Urban Land Institute (ULI).

The focus on shortages at a national level arises from a focus on immigration policies rather than housing policies. It is relevant, however, because the essential workers nominated by EWIC include restaurant workers, retail clerks, construction trades people, manufacturing line workers, hotel service workers, food production workers, landscape workers, and health care aides.7 These overlap considerably with shortages identified at a local level. EWIC have argued that these are the jobs that many Americans do not choose, but which are 'essential' to keep the American economy growing. To support their argument they claim companies are reporting difficulties in retaining permanent staff and hiring replacements and are curtailing expansion plans, and many small businesses are struggling to survive without enough employees.

It is the second strand of the US literature from a local perspective, however, which reflects a concern with the impact of housing affordability on labour shortages. This literature is as much concerned about the implications for urban sprawl as it is with attracting and retaining workers to support local business or to provide essential community services. It focuses specifically on workers in service industries whose location is determined by the service they provide. "Where affordable housing does exist…it usually is located far from where most people work. ...This...brings with it all the undesirable aspects of sprawl: grinding traffic congestion, school overcrowding, air pollution, and a loss of open space. Yet most major institutions - governments, hospitals, and the like - are located in or near the central city and cannot move out to follow the workforce. This dynamic makes it hard to recruit and retain moderate-income employees such as teachers, fire fighters, nurses, and so forth. Private businesses, on the other hand, are more mobile. Many are moving to the outer fringes to be closer to their workforce. While this might appear to solve the jobs/housing imbalance, it actually further compounds the cycle of sprawl by driving up land costs and forcing affordable housing even farther out." (Haughey, 2002, p2)

7 http://www.immigration.com/newsletter/news22ess.html

6

This view of essential workers has focused on those whose wages are tied to the old economy (Stegman et al, 2000). These include janitors, retail sales workers, teachers, nurses and police. Wages in these occupations are seen as being typical of wages in other essential occupations. Jobs have not been eliminated by economic restructuring and there are a reasonably large number of jobs that are attracting relatively low skilled first time entrants to the workforce (Fiore and Lipman, 2003). They might also be described as occupations in which the scope for productivity increases is relatively limited.

Australian literature

One of the earliest and most comprehensive studies in Australia that specifically is concerned with affordability, occupation and location is that undertaken by O'Connor and Healy (2002) whose work (on Melbourne) identifies only high and low status workers broadly defined according to their occupations. Their work suggests central Melbourne is becomingly increasingly separate from the rest of the metropolitan region and is consistent with Sassen's social and spatial polarisation or Fainstein's divided cities hypotheses (Fainstein, 1992; Sassen, 1991). However, it does not specifically focus on the potential difficulties faced by the low status workers employed in central Melbourne.

Sydney based studies have followed the pre-determined approach to defining occupations of interest that is the hallmark of the related international literature. Randolph et al (2004) focused on computer professionals, registered nurses, primary and secondary teachers, truck drivers and sales assistants. These five occupational groups were a pragmatic choice: they were chosen to provide examples from both the public and private sectors and because they contained relatively large numbers of the middle income workers who were the focus of their study. A significant result of their study was that outcomes differed considerably for different occupations, reflecting in part the extent to which jobs were dispersed.

In their study of the Northern Beaches area in Sydney, Epic DotGov (2004) employed a list of sixteen occupations seen as providing key services to the community in the area of health, education, transport, child care and property protection. These cover low or modestly paid workers such as teachers, nurses, police, sales assistants and cleaners defined as key workers in the UK and US work. They also include paramedics, motor mechanics, automotive electricians, firefighters, child-care workers, bus drivers and train drivers. The results of their study suggested there were more key workers than jobs in the region and that there were few problems in filling jobs as a result. Housing costs were seen as unimportant because many key workers fully owned their own homes.

Cottrell (2004) based her (Eastern Suburbs) study on the same list as Epic DotGov. Blunden et al (2004), who also covered the Eastern Suburbs, to date have examined only bus drivers.8 Cottrell's results suggest that the number of key worker (as defined) jobs in the Eastern Suburbs exceeded the number of key workers living in Eastern Suburbs and that the shortfall of workers was aggravated by an out-migration of younger key workers who can't afford to buy and remain in the region. The majority of workers covered by Blunden et al's survey traveled less than 30 minutes to work but almost 20 percent traveled more 60 minutes or more to get to work. Their interview based results suggested that the main issues arising for those who worked in the region were the long commute (generally undertaken by car as this was more predictable); poor work schedules (not

8 An update of this work is due to be presented at the 2005 Social Policy Conference.

7

enough time to go home with broken shifts); and not seeing enough of their family because they have to live further away than desired because of affordability problems.

In NSW, a limited demonstration policy initiative defined key workers were as those who provide a service that contributes to the well being of the community and are unable to afford appropriate accommodation on the open market. The definition includes, but is not limited to, hospital workers, teachers, childcare workers, police, transport workers or fire fighters (NSW DoH, 2004).

A Queensland Department of Housing report, whilst accepting the broad intent of identifying occupations necessary to the efficient functioning of a community and targeting policies to workers who provide an essential service in areas where they cannot afford to live and who may face long commutes as a result, suggests a number of criteria might be applied to identifying such workers (Queensland DoH, 2003). In their view, for policy purposes, key workers are likely to have low incomes; are likely to rely on penalty rates or overtime to boost basic wages; are likely to have a predominance of casual or part-time work, or lack long-term job security; may work night shifts, split shifts or irregular hours, including times when public transport is irregular or unavailable; may work in areas where employee car parking is not provided and alternative public transport is not practical; and generally lack higher tertiary qualifications (although some groups such as community service workers, teachers and nurses are seen as exceptions to this). These criteria move the concept of what constitutes a key worker away from a focus on occupational categories to a more constructive focus on the labour market conditions which govern work. They also raise the possibility that income earned and the conditions under which workers are employed, rather than occupation per se, are the key drivers of the concerns raised in the quotation at the start of this paper.

Conceptual issues and methodological approach Defining occupations of interest

The above suggests the issue of defining relevant occupations or workers is one that has economic, regional, industrial relations, and political overtones. In economic terms, key workers have been defined as those needed to support local business or provide essential services, whose shortage is likely to affect the ability of a region to grow or function efficiently. The definition of exactly what occupations are required is open to debate and, in regional terms, is likely to vary according to whether the region is growing or declining. In other words, it varies because local economies vary. In industrial relations terms, relevant workers are likely to be those in casualised or part-time jobs for whom the costs of travel or housing are not compensated for in their inadequate wages. In political terms, key workers are likely to be seen as those who can hold governments to ransom at politically inopportune times. Illustrations of how this might be done are strikes by transport workers or teachers. Each of these broad interpretations has been embodied in the literature reviewed above.

For the empirical part of this paper, the competing demands made by these different interpretations on defining key workers have been addressed by limiting the focus of the analysis to the issues that arise in a specific local economy (viz. Sydney and Inner Sydney) and by use of what can be called indicator occupations.

The indicator occupations chosen are, at the least skilled end of the spectrum, cleaners (ASCO code 911) and hospitality workers (ASCO code 632) and, at a higher skill level,

8

nursing professionals (ASCO code 232) and computer professionals (ASCO code 223). According to Reich's taxonomy workers in the first three occupations would all be described as 'in-service workers'. They are occupations where employers are constrained in their location choice by the nature of the service provided but the jobs are likely to be relatively dispersed. Workers in the fourth occupation, include those who would be described as 'symbolic analysts'. They include (but are not solely) those who are part of the new economy and who are likely to be better paid as a result. Their jobs are those that are likely to be concentrated in the centres of the cities.

This selection builds on the lessons learned from the studies reviewed above. It includes occupations with large numbers of workers who receive low or moderate wages and which meet a number of other criteria. It includes occupations likely to be regarded as essential to the proper functioning of the city and likely to be in short supply in high cost locations as a result of housing affordability issues (rather than as a result of general labour market shortages). It includes occupations that are representative of those used in other studies but are sufficiently different from each other to enable differences between them to be identified. It includes occupations that provide representation across private and public sector employers and occupations which have both a dominance of females and a dominance of males. It includes occupations whose workers have been identified as facing above average affordability problems and who therefore are most at risk of being excluded from jobs in high cost areas because of housing costs.

Defining regions of interest

This paper focuses specifically on workers who live in Sydney and who work in Inner Sydney. This decision was based on the significantly higher housing costs that are observed in Sydney compared with other metropolitan regions. Within Sydney, two key choices arise in relation to the choice of region when examining the impact of housing affordability on labour market outcomes. The first is the choice of the appropriate level of spatial disaggregation within the metropolitan region; the second, simpler choice is determination of which of the spatially disaggregated within metropolitan regions are of interest. The issues that underpin the first of these choices are the same as arise in attempts to define housing or labour markets. The Bureau of Transport and Regional Economics (BTRE), for example, describes a labour market region as "the area within which people are willing to commute from their place of residence to their place of employment." BTRE (2003, p17) They use residential containment, based on commuting patterns revealed by the 2001 census, to define these regions with a labour market defined as one in which the majority (typically at least 70%) of employed residents work. On this definition, they define each greater metropolitan region as one labour market.

However, this definition defeats one of the key aims of this paper. It does not allow for an examination of the extent to which affordability within a particular metropolitan region affects the residential location choices of workers in that region. The results from periodic transport surveys suggest that the journey to work time within Sydney and distance traveled has remained fairly constant since 1999 with an average work trip duration of around 30 minutes and an average commuting trip of around 16 kms per day (DIPNR, 2004). This suggests that an appropriate choice was to base the analysis on the statistical sub-division (SSD) level of which there are 14 in the Sydney statistical district (these are listed in Table A. 1 in the Appendix). These SSDs broadly define areas within which it is

9

possible to meet these average travel times and distances.9 To these regions are added Newcastle in the north and Wollongong and the balance of the Illawarra region in the south to capture those prepared to undertake commutes to work that are significantly longer than 30 minutes. The extent of residential and employment self-containment at the chosen level of spatial disaggregation is examined below.

The second choice, that of determining which regions are most likely to be those where labour recruitment and retention issues arise, was based on balancing the housing costs in each region with the extent to which there was a labour deficit (or jobs surplus). Almost by definition, these constraints lead to the choice of the region containing the CBD, which is that used in this study. Data to support these claims are provided below.

Relating individual to household characteristics

Some of the results from the literature reviewed above signal other conceptual issues that need to be addressed when examining the relationship between occupation and affordability. The most important of these is that occupation is defined at an individual level whereas affordability and choices regarding residential location are affected by household characteristics.

For the UK, Tym et al (2003) and Morrison (2003) distinguished problems of recruitment for people aged under 30 from problems of retention for those in the 30-34 age range. Workers in the 25-34 age range were seen to be moving from a housing solution where house sharing was not just acceptable but even popular because of the social benefits associated with it. Living with friends or family was a significant option for younger key workers under age 25 but its acceptability dropped sharply thereafter as workers partnered, had children and aspired to a garden and access to good schools. A report for prepared for the Greater London Authority showed "commuting was relatively low among certain groups, particularly ethnic minorities and women, among those with fewer qualifications and less experience and in certain employment sectors - notably health and education" (GLA, 2001, p16).10 It also showed a significant proportion of key workers were looking for jobs outside of London, with those most wanting to leave London being younger workers aged 18-34 rather than older workers (GLA, 2001, p18).

In Australia, one of the conclusions of Epic DotGov (2004) was that there were relatively few problems in filling key worker jobs in the Northern Beaches region despite a marked deterioration in affordability. In part this was attributed to the fact that many key workers lived in owner-occupied housing. Key workers can afford to live there because they have done so for a long time. Cottrell (2004) showed a significant reduction in the numbers under 40 who can't afford to live in the Eastern Suburbs and so look elsewhere when they are ready to purchase. The interviews undertaken by Blunden et al (2004) provide support for this. Analysis undertaken at an individual level shows many of the current young workers in the occupations of interest as living in owner-occupied housing. When the analysis is undertaken at an individual rather than household level, this outcome is as likely to reflect the tenure of their parents as of the individual of concern. This highlights a further difficulty in defining the target group of interest and indicates that results based on housing outcomes for individual workers must be interpreted with care.

9 Local Government Areas and Statistical Local Areas (of which there are more than 40 in Sydney) were seen as being too small to define a labour market. 10 This report also highlighted differences in commuting patterns between occupations with nurses and bus drivers wanting to live near where they work but police officers not wanting to.

10

These results from the literature also suggest that issues of retention may differ from issues of recruitment, primarily because of the difference in the ages and lifestyle preference of the workers involved. In this paper, the persistent conclusion that life-cycle factors are relevant is taken into account only indirectly. Housing affordability is assessed at a household level with the occupation of the reference person in the household being taken as the identifying occupational characteristic. One justification for this is that the reference person may reasonably be taken as the person who is most likely to be dominant in the decisions that affect housing and location choices. It is, however, problematic in that there are more than twice as many persons in employment as there are households where the reference person is in employment and a focus on the reference person is likely to mean many secondary and part-time workers, who form an increasingly important part of the labour supply, are likely to be excluded from consideration. This issue will be addressed below.

Determining the impact of housing affordability on labour market outcomes

The final and most difficult conceptual issue from the inherent difficulty in determining the impact of housing costs on labour markets shortages. This cannot be determined solely from affordability measures because lower income workers who work in high cost areas might simply undertake longer commutes from areas where housing is more affordable. Consequently, the analysis of their housing cost or affordability position per se may show relatively little to be concerned about. The downside, however, is increasing commuting costs, social stress and poor traffic and environmental outcomes, rather than labour market shortages.

This difficulty is being addressed in the larger study of which the paper is a part by taking a three staged approach. However, only the first two are reported in this paper. The first stage is to examine the extent to which different households are in housing stress. The analysis at this first stage is undertaken at a household level because individual circumstances do not reflect capacity to pay for housing costs and because affordability outcomes are inherently related to household income and household structure. This step also is used to add to refine the choice of indicator occupations for which the more detailed analysis is undertaken.

The second stage is to focus on a specific high cost region within Sydney and to determine the travel to work patterns for all those who work in this region. The inner city region of Sydney is chosen as a region which is a high housing cost region and which has the lowest degree of residential self-containment and, hence, the largest net in-movement of workers.

The third stage, for which data have yet to be supplied by the ABS, is to link these travel to work patterns which are defined at an individual level, to the characteristics of the household in which the individual lives (as either a reference person or as a secondary earner). This will provide additional insights into the extent to which younger workers in particular currently have housing solutions (such as sharing, renting, living at home) which may not meet their longer term aspirations problems as their household circumstances change.

11

Results Housing stress levels

The results presented in this sub-section focus initially on outcomes for the household; that is, from stage one described above. Table 1 provides an overview of the proportion of households in housing stress according the occupation of the reference person. The results are limited to those households were the reference person is employed as occupation is not defined for those who are unemployed or not in the workforce. There are 115 occupations defined at the 3 digit ASCO category but those with fewer than 1000 households in Sydney have been deleted to maintain a focus on the more significant categories. Households are defined as being in housing stress if they are paying 30 per cent or more of their gross household income in meeting their housing costs. Because the results in Table 1 have been generated from census data which contains information only on rent and mortgage repayments, outright owners have zero housing costs attributed to them and home purchasers have their costs understated.11 As a result, the reported measures are likely to understate the true extent of housing stress. The results have been ranked according to the proportion of all households who are in housing stress.

Table 1 shows there were just under 1 million households in Sydney in 2001 where the reference person was in the workforce. Of these, 14 per cent (or 141,000 households) were in housing stress. This incidence of housing stress amongst working households in Sydney is higher than the 10 per cent incidence that applies for working households in Australia as a whole (Yates et al. 2005b).

In broad terms, the results presented in Table 1suggest that working households with the highest proportion of in housing stress are predominantly those where jobs are casualised, part-time and offer few career prospects. At the top of the list are hospitality workers (including, inter alia, bar attendants, waiters and hospitality trainees – ASCO code 632).12 High on the list (with an incidence of housing stress that is roughly 50 per cent higher than the Sydney wide average) are food preparation workers (993), food tradespersons (451), receptionists (613), hospitality and accommodation managers (332), carers and aides (631), cleaners (911) and, with the greatest absolute numbers in housing stress, sales assistants (821).

Table 2, which provides a summary of the household income distributions for all households as well as those in housing stress, highlights the relationship between housing stress and household income. Whereas only 23 per cent of all working households in Sydney have a household income of less than $800 per week, 56 per cent of those in housing stress have incomes less than $800 per week. Alternatively, whereas 47 per cent of all working households in Sydney have a household income of less than $1200 per week, 79 per cent of those in housing stress have incomes less than $1200 per week.

The results also indicate that, whilst the high proportions of households in housing stress are found amongst the less skilled occupations, there are a number of occupations classified in the top occupational groupings (categories 1 and 2 at the single digit ASCO level) in which there is at least an average incidence of households in housing stress 11 Households where income or housing costs or occupation were not stated or partial also have been excluded from the data. 12 It is for this reason, plus their identification in the quotation given at the start of the paper, that these workers were included in the set of indicator occupations.

12

despite the generally higher incomes enjoyed by such households. Notable amongst these are occupation categories 253 (artists and related professionals) and 222 (sales, marketing and advertising professionals). It is likely that these are occupations which, although skilled, are subject to fluctuating fortunes and little long term job security.

One additional observation that can be drawn from these results is that, at an aggregate level, not all of the households with a reference person classified as a key worker or essential worker household in the UK or US face affordability problems any worse than the population as a whole. In the UK, nurses, teachers, police officers and prison staff are amongst those classified as key workers. In the US, the list of essential workers includes cleaners, retail sales workers, teachers, nurses and police. The results above suggest that, of these, only households with a reference person who is a cleaner (911), a sales worker (821) or a miscellaneous intermediate worker (639, which includes prison officers) have a disproportionate share of households in housing stress. Households with a reference person in occupation 241 (school teachers) or 391 (police), for whatever reason, have a very low proportion of households in housing stress and those with a reference person in occupation 232 (nurses) have roughly the same incidence of housing stress as the population as a whole. In absolute numbers, however, the sheer size of the teaching and nursing professions means there are significant numbers of households in these professions in housing stress.

As suggested, however, housing stress is only one way in which housing affordability can impinge upon labour supply. An alternative to spending a high proportion of income in meeting housing costs is to locate in lower cost regions and undertake long journeys to work. This question forms the basis of the stage 2 results covered in the following sub-section.

Workplace-home locations in Sydney

One indication of the extent to which workers rely on commuting as a solution to housing affordability problems can be seen in the extent to which they live in a less costly region than that in which they work.

Table 3 and Table 4 show the degree of residential and employment self-containment of workers within Sydney. Residential self-containment provides a measure of the extent to which workers who live in the region work in the region. A low level of residential self-containment means that there is a significant amount of out-migration as residents travel to work in a different region. Employment self-containment provides a measure of the extent to which those who work in a region live in that region. A low level of employment self-containment means that there is a significant amount of in-migration from workers who live outside of the region.

For Sydney as a whole, 50 per cent of employed persons in Sydney live and work within the same region. Table 3 shows residential self-containment is above average both in the centre (Inner Sydney) and in the outer regions (Illawarra, Gosford-Wyong, Newcastle and Wollongong) with containment ratios from 66 to 89 per cent. 13 Residential self containment is around 50 per cent in the remaining regions which are located furthest from the city centre (Northern Beaches, Outer Western Sydney) and below average elsewhere. In other words, compared to workers who live in more centrally located regions, a higher proportion 13 The residential self-containment proportions include destination data where there was no fixed address or where the work was outside the regions covered by those listed for the origin data but these latter data have not been reported.

13

of workers in the regions located further from the centre work in the same region in which they live. Residential self containment varies from a low 25 per cent in Inner Western Sydney to a high 89 per cent in Newcastle.

Table 4 indicates there is an even greater variation in employment self-containment. Only 21 per cent of those who work in Inner Sydney (basically the CBD) live there whereas 98 per cent of those who work in Newcastle, 96 per cent of those who work in Wollongong and 93 per cent of those who work in Gosford-Wyong live in those regions.

The absolute numbers that underpin these self-containment ratios and changes in them between 1996 and 2001 provide a measure both of which regions are job-surplus or job-deficit regions and of where job-growth has been most rapid. In broad terms, areas which are job-surplus areas (that is, in which the number of jobs exceeds the numbers of workers who live in that region) are more likely to be those which face labour shortage difficulties. Table 5 provides this information along with median rent data for each of the regions for 2001 as an indicator of the relative housing costs in each region.

As can be seen from the results in column five of Table 5, there were only 3 regions at a statistical sub-division level within the greater metropolitan region in Sydney where there were jobs surpluses. These are Inner Sydney (central Sydney or the CBD), with a surplus of 283,000 jobs, Lower Northern Sydney, with a 53,000 surplus and Central Western Sydney (which includes Parramatta) with a 41,000 surplus. For the first two of these, described by Randolph et al (2004) as being within Sydney's 'Global Arc', median rents were amongst the highest in Australia in 2001 (Yates et al, 2005a). For the third, associated with the manufacturing and warehousing belt, rents are still high by Australia wide standards but some 25 to 50 per cent lower than those in the more centrally located regions.14

Columns two and four in show the changes in the residential location of workers in Sydney between 1996 and 2001 and the change in the location of jobs over the same period. This shows that the greatest job growth (given by the destination of workers) was generally in the lower cost regions on the outskirts of the city. The greatest loss of workers (given by their origin or residential location) was in the higher cost regions.

These results are illustrated in Figure 1 which shows the change in the proportion of workers by place of residence (origin) and place of work (destination) ranked from highest to lowest cost regions. With the exception of jobs in Central Northern Sydney and residential location in Inner Sydney (reflecting the growth in apartments between 1996 and 2001), and declines in the old economy locations of Newcastle and Wollongong, there was a relative shift of jobs residential location of workers from the higher cost inner to the lower cost outer suburbs of Sydney. A formal measure of the extent to which these patterns have converged or diverged is presented below.

An initial answer to the question of whether this general outward trend differentially affected workers in different occupations can be seen by disaggregating the data presented in Figure 1 by the indicator occupations selected for consideration. This is done in Table 6, Table 7 and Table 8. Figure 2 and Figure 3 do the same for the data illustrated in Figure 1. As above, the data illustrated are ranked with the highest housing cost regions on the LHS of the chart. 14 A more detailed description of the localities covered by the statistical sub-divisions used as the basis of analysis in this paper is provided in the Appendix, along with a rough indicator of whether the region is an inner, middle or fringe/outer region.

14

The final column in Table 6 provides a measure of the extent to which the growth in the number of workers has been in the outer regions of Sydney. With the previously noted exception of Inner Sydney, the only regions that have had rates of growth of the numbers employed that exceeds the Sydney average are the outer or fringe regions. Whilst this is likely to reflect the greater residential opportunities that are available in these locations, in part it is also likely to reflect the impact of a reduction in what were high rates of unemployment in these regions in the mid 1990s.

Even though there are only 4 indicator occupations selected for consideration, the aggregate results presented in the final row of Table 6 clearly illustrate the outcomes of the significant restructuring in the nature of work over the inter-censal period. The growth in the number of employed as computing professionals (that is, of one particular group of new-economy workers) has exceeded the overall growth in the total numbers employed by a factor of 10 and the growth in those who are hospitality workers is almost double the overall growth in total numbers employed. The numbers employed as nursing professionals has grown more slowly than the general growth in employment and the number employed as cleaners has actually declined.

Table 6 also indicates there is considerable variation in the residential location of these changing numbers of workers, varying from decreases of 10 to 20 per cent in the number of cleaners in the highest cost regions (Eastern Suburbs, Inner Sydney, Inner Western Sydney and Lower Northern Sydney) to increases of 50 to 80 per cent in the number of computing professionals in the same locations. The results suggest (but by no means prove) that the latter are displacing the former. Figure 2 adjusts for differential growth rates amongst the various occupational categories by showing changes in the proportion of workers in each indicator occupation by their residential location. In broad terms, the outward migration is least for the highest paid computer professionals and greatest for the lower paid cleaners.

Table 7 shows similar information for the place of work. Amongst other things, the growth rates in the final column, which apply to all workers, indicate that the highest rate of growth in the various places of work is 'No fixed address'. Whilst this applies to only 4 per cent of all destinations, its growth may be indicative of future trends where workplaces are varied rather than fixed in location. Figure 3 reinforces the presumption that the growth of jobs in the inner regions (notably Inner Sydney and Lower Northern Sydney) is a result of growth in new economy jobs (as indicated by the growth of jobs for computing professionals).

Table 8 shows that the pattern of job deficits and surpluses by regions at a disaggregate region is not quite as stark as at the aggregate level. The largest surplus of jobs requiring specific occupational skills over resident workers with those skills for all occupations reported arises in two of the three regions that show the greatest surplus in aggregate (viz. Inner Sydney and Lower Northern Sydney). It is greatest in these regions for all but nursing professionals where the greatest surplus is in the third region that has a surplus at an aggregate level (viz. Central Western Sydney). For nursing professionals, surpluses of jobs over resident workers also exist in the Eastern Suburbs, Fairfield Liverpool and Inner Western Sydney. These surpluses are more likely to reflect the location of large hospitals rather than high housing costs.

15

The disparity between places of work and residence, the concentration of each and the changes that have taken place over time can be summarised in two different measures.15 The extent to which there is a mismatch between jobs and workers can be provided by the Dissimiliarity Index which is a measure of segregation.16 In the context in which it is used here, it can be interpreted as the percentage of workers who live in each region who would need to re-locate so that the proportion of workers with specific occupational skills was the same as the proportion of jobs requiring those skills in each region. The extent to which jobs in each occupation are becoming more concentrated can be provided by the Herfindahl or Hirschman-Herfindahl index.17 Increases in the index suggest that a higher proportion of workers in the specified occupation is found in a smaller number of SSDs or, in other words, that the residential (or workplace) location of workers with that occupation is becoming more concentrated in a limited number of SSDs.

These summary indices are presented in Table 9. A number of key points stand out. At an aggregate level (as shown in the final column) the Dissimilarity index is relatively low which suggests there is a relative match rather than mismatch between places of work and residence and this has been virtually unchanged between 1996 and 2001. Similarly, at an aggregate level, there is a low level of concentration of both place of work and place of residence. This, too, has remained unchanged over time.

At an occupation specific disaggregate level, similar conclusions can be drawn for all occupations other than computing professionals. The Dissimilarity index is even lower than the aggregate measure for nursing professionals, hospitality workers and cleaners, which suggests an even greater match rather than mismatch between places of work and residence for workers in these occupations. This, too, has been unchanged between 1996 and 2001. Similarly, for each of these three indicator occupations, there is a low and unchanged level of concentration of both place of work and place of residence which also has remained unchanged over time.

It is only for computing professionals that these broad conclusions are different. The destination based Herfindahl index increased by more than 10 per cent in just 5 years from 0.22 from 1996 0.25 in 2001 highlighting the increased concentration of jobs for workers in this occupation in the high cost areas of Inner Sydney and Lower Northern Sydney. Despite the increase in the share of computer professionals who live in Inner Sydney, the origin index has remained unchanged with the results that the Dissimilarity index has increased significantly (also by 10 per cent from 0.40 in 1996 to 0.44 in 2001). Both the destination based concentration index and the Dissimilarity index are considerably greater than the aggregate indices for all workers which is a reflection of the fact that these new economy jobs are spatially concentrated. 15 Massey and Denton (1988) provide a systematic overview of a number of different measures that have evolved for measuring concentration, dispersion, segregation or dissimilarity. Theil (1972) provides a detailed evaluation of the strengths and weaknesses of the various measures. 16 This index, described in Duncan and Duncan (1955) is frequently used in the sociological literature to describe racial segregation. 17 This index is borrowed from the industrial organization literature where concentration indexes are used to determine market structure and the extent to which an industry is competitive. In its original use, the index was based on the sum of the squared percentages of market share of all firms in an industry. Hannah and Kay (1977) suggest that the Herfindahl Index is the only measure of concentration that meets their criteria for judging the suitability of different concentration measures. These criteria are as follows: the index increases as concentration increases; when large firms take customers from smaller firms; if a merger takes place; if a new firm below a significant size enters; and the index decreases if new firm, below a significant size, enters the market. Here, regions replace firms and the share of low rent dwellings replaces market share.

16

Overall the data presented in this section suggests that, with the exception of the concentration of jobs in Inner Sydney leading to a significant job surplus in the CBD, both places of residence and places of work are dispersed throughout the greater Sydney region and that, in broad terms, changes in the places of residence of workers are matched by changes in their places of work. For the occupations considered, the only exception to this is computing professionals. There is little evidence that hospitality workers, for example, are living further away from their jobs in 2001 than they were in 1996. The dispersal of in-service jobs across the city provides one explanation of why employers in high cost locations have difficulties in recruiting workers. Workers do have other choices. Workplace choices are likely to be most limited in those occupations where jobs are highly concentrated.

The final stage of this part of the analysis is to examine in more depth the travel patterns of those workers who work in the Inner Sydney region which has the largest potential claim to having problems of recruiting and retaining labour because of it has the lowest employment self-containment ratio (21 per cent) and the greatest surplus of jobs over workers (278,000).

Workers in high cost Inner Sydney

The places of residence (origin) of the workers who work in Inner Sydney can be seen from the relevant column in Table 4. This showed that 21 per cent of those who work in Inner Sydney lived there, and a further 28 per cent live in the broader inner zone (consisting of regions that are within 10 kilometres of the CBD). In other words, almost 50 per cent of Inner Sydney workers live within what must be called a reasonable commuting distance. A further 43 per cent live in the middle ring SSDs18 It is the remaining 8 per cent of workers who are those most likely to have traded off significant travel times for housing affordability. This represents approximately 34,000 workers or around 17,000 households across all occupations.

Figure 4 illustrates the places of residence of those in each occupational category who work in Inner Sydney. Figure 5 shows how their residential locations have changed between 1996 and 2001. For ease of comparison, Figure 2, which shows changes in place of residence for all workers in Sydney, is reproduced alongside Figure 5. A comparison of the results in these two charts clearly shows the impact that the increase in dwelling supply in the Inner Sydney region has had on the residential locations of those who work in this region. For each of the indicator occupations charted, there has been an increase in the proportion of Inner Sydney workers who live in the same region in which they work. There has been little change in the proportions of Inner Sydney workers who live in the outer regions of Sydney.

In the absence of information on the housing outcomes of those who work in Inner Sydney or on their socio-economic and demographic characteristics, few conclusions can be drawn about which workers trade-off lower commutes for the higher housing costs of the inner zones. The analysis undertaken by Randolph and Holloway (2004), however, provides an excellent basis on which to make an informed assessment. Their work, which focussed solely on middle income households, which examined outcomes only for the reference person in the household and which included some, but not all, of the indicator occupations examined here, suggests a broad generalisation is that key workers tend to rent in the inner zones and buy out west. This tendency is strongly related to life-cycle 18 The categorization of regions into inner, middle and fringe zones is provided in the Appendix

17

characteristics with younger key workers renting in inner city locations whereas older workers were more likely to own or be purchasing in middle or fringe regions (Randolph and Holloway, 2004, chapter 8). These outcomes are consistent with those generated by Yates (2001) from earlier census data. Her work showed that tenure, dwelling type and location outcomes were also strongly correlated to household income and household type. Both of these studies highlighted the limited housing choices faced by lower income households.

Table 10 and Table 11 show the affordability outcomes for those workers who do live in Inner Sydney. Table 3 showed two thirds of these also work in Inner Sydney and three quarters work either in Inner Sydney or in Lower Northern Sydney. A comparison of the results in Table 10 and Table 11 with those in Table 1 and Table 2, which show the equivalent data for Sydney as a whole, highlights the financial cost of this location choice. For Sydney as a whole, 14 per cent of all working households are in housing stress. For those living in Inner Sydney, the proportion is 22 per cent. For Sydney as a whole, 33 per cent of the lower paid hospitality workers were in housing stress. For those living in Inner Sydney, this proportion increases to 49 per cent. Similar comparisons can be made for each of the indicator occupations.

The final results to be presented in this paper are in Figure 6. These show the change between 1996 and 2001 in the proportion of workers who work in Inner Sydney but reside in each of the statistical sub-divisions in Sydney disaggregated by their household income.19 It clearly shows that the increase in residential opportunities in Inner Sydney have benefited those on higher incomes and that there has been a significant re-location of all but the least well paid from the even higher cost regions which are generally part of the inner zone of Sydney. The least well paid have moved further out.

Conclusions A number of conclusions can be drawn from the discussion and data analysis above that can be used to provide insights into the relationship between occupation and housing choice.

(1) Claims (such as that attributed to Newman in the introduction to this paper) that structural change has led to a re-urbanisation of the inner city regions of global cities and an increasing reliance on new economy jobs in the inner city regions is only partly correct when applied to Sydney.

There has been significant growth and concentration of computing professionals, the occupation selected in this study as an indicator of new economy jobs, and there has been a move towards inner city living of those employed in this occupation. Overall, however, there has been much greater growth in jobs in the locations where an increasing proportion of workers are living. In the case of Sydney, this is in the fringe regions of the city. The results for Sydney lend support to O’Connor and Healy’s claims that jobs do follow people although care does need to be taken in drawing the implications about the direction of causality from correlations between growth in workplace and residential locations of workers.

19 The income classifications in the 1996 and 2001 Censuses remained unchanged in nominal terms, and there are changes in the proportions of households in each of the income categories between the two years. This limitation of the data, however, does not affect spatial analysis presented here.

18

(2) There is considerable evidence to support the claims of those who express concerns that employers in high cost areas such as the inner city cannot attract key workers because of housing affordability problems. However, there is also considerable evidence to suggest that this problem is a subset of a wider problem of general income and spatial polarization which is reducing housing options for lower income earners in general. The affordability problems faced by key workers are the same as those faced by any lower paid household.

For employers in high cost regions, the problem of retaining low paid workers is a key concern as there are high costs associated with labour market turnover. Such costs can arise from recruitment, from the need to retrain new employees and from failures to meet service standards or delays in meeting production deadlines (ANCER, 2004:16). Recruitment and retention difficulties increase the pressure on wages for employers who are likely to be faced with shortages of their lower skilled work-forces. Such wage increases are unlikely to be compensated for by productivity increases given the service orientation of a large number of these occupations.20 Such difficulties are likely to be greater, the greater is the reliance on younger and/or secondary workers in a household. The extent to which this is the case in Inner Sydney is a question for stage three of this research project.

(3) One of the key difficulties of relating the workplaces to the residential locations of workers is that labour is inherently heterogeneous in its skills (and hence in the income it can generate). Housing markets, however, are becoming spatially homogeneous as general economic and population pressures increase land prices in locations where jobs are concentrated with the result that housing in those regions is becoming increasingly unaffordable.

The implications of a lack of appropriate affordable housing are that many workers either pay a high proportion of their incomes in meeting their housing costs and/or travel long distances in order to work in their chosen location. Neither of these options is likely to be particularly sustainable. The pressure of high housing costs may be reduced by natural life-cycle events as young households move away from high cost well located rental housing to less well located owner-occupied housing, much of which is still likely to impose a high housing cost burden on them throughout the early home purchase years. If these costs are reinforced by high commuting costs, a natural outcome may well be to search for new employment closer to the place of residence. High commuting costs, particularly if they are not compensated for by high housing consumption, also are likely to lead to a search for new employment closer to the place of residence.

The consequence is that the inner high cost regions of the city will be populated by the young, the affluent and those without children. This suggests a bigger issue that needs to be addressed is whether such spatially polarised cities are sustainable in the long run. Back in the 1960's Jane Jacobs argued that living cities needed diversity (Jacobs, 1962). More recently, Richard Florida has suggested that it is creativity rather than knowledge that is at the base of the new economy and that the creative class wants to live in places

20 It is important to remember that earnings are not the only factor that affects labour market supply. A detailed analysis of the persistent state of shortage of nurses by Dockery (2004), for example, suggests that this is attributable more to job satisfaction and poor career prospects than it is to wages (which are estimated to be in line with those of other women when typical human capital variables are taken into account). He concludes shortages are due "not only to institutional and policy settings in Australia, but also to intrinsic characteristics peculiar of the occupation." (Dockery 2004, p74)

19

that reflect this diversity. Vibrant cities need hospitality workers; they need cleaners; they need workers who work at all times of the day or night. Such diversity does not come from the occupational and age-based segregation that currently defining our cities. (Florida, 2002, 2005).

The results presented here suggest that the inability of employers to recruit and retain key or essential workers is likely to be part of a much broader and potentially much more insidious process. Low paid workers are being displaced from the high cost regions. However, much of this displacement is because their jobs are also being displaced from high cost regions. It is part of a process whereby high-cost residential development is displacing many of the traditional sources of employment for the residents of those areas. With the exception of the CBD, the numbers of persons employed in jobs in the high cost regions of Sydney have declined. These jobs have been replaced by jobs in the middle and outer regions. It is part of a process by which the low paid increasingly are being excluded from the range of choices available to those who are benefiting from the economic growth that has occurred as a result of the economic restructuring that has taken place in the last quarter of a century. Employers will be able to address their problems by employing a continuing supply of young, mobile workers by providing better wages and better conditions. The continued costs associated with retraining will be passed on in higher prices.

The ultimate cost, however, will be a city which has lost the diversity that made it great in the first place. Jane Jacobs knew the importance of diversity as long ago as 1962. Richard Florida has reminded us of it today.

Both the UK and the US have implemented affordable housing policies. In London these require 50 per cent of all new housing developments to be affordable. In a number of States in the US, they require 15 to 25 per cent of all housing developments to be affordable. In both countries, these policies are underpinned by national and state based financial arrangements that ensure the policies are feasible. It is time to introduce such policies in Sydney.

20

Table 1: Household in stress by income and occupation, Sydney 2001* Proportion of households in stress by household income distribution (%)

Occupation (3 digit ASCO category)In stress

(%) $0-$399$400-$599

$600-$799

$800-$999

$1000-$1199

$1200-$1499 $1500+

Total no. of households

in stressTotal no. of households

632 Hospitality Workers 33 32 36 16 8 5 3 1 3,470 9,451993 Elementary Food Preparation and Related Workers 26 38 36 14 6 3 1 1 1,408 4,670829 Miscellaneous Elementary Sales Workers 24 29 34 17 8 5 3 3 2,125 7,975821 Sales Assistants 23 27 34 17 10 5 3 3 6,617 25,338639 Miscellaneous Intermediate Service Workers 22 19 29 20 13 8 5 7 2,238 9,268451 Food Tradespersons 21 21 35 21 12 4 3 3 1,694 6,956613 Receptionists 21 18 39 22 8 5 4 4 1,871 7,746332 Hospitality and Accommodation Managers 21 13 25 23 15 9 6 9 2,996 13,029631 Carers and Aides 20 30 40 14 6 4 3 2 3,044 13,061911 Cleaners 20 27 38 17 9 5 3 2 3,473 15,368253 Artists and Related Professionals 20 14 20 18 15 13 6 13 3,365 15,628831 Elementary Service Workers 18 24 37 19 8 5 3 3 2,311 11,390331 Shop Managers 18 14 22 22 14 9 9 11 3,508 17,169462 Horticultural Tradespersons 18 14 29 23 15 9 6 4 1,081 5,576612 Keyboard Operators 17 14 36 27 10 6 4 4 1,115 5,757442 Final Finishes Construction Tradespersons 16 10 26 28 15 9 7 5 1,203 6,486811 Elementary Clerks 16 21 34 20 11 5 4 5 1,482 8,252421 Automotive Tradespersons 16 7 23 28 17 10 9 6 2,239 12,611591 Advanced Numerical Clerks 16 11 23 21 17 9 8 11 1,232 7,125799 Miscellaneous Intermediate Production and Transport 15 15 37 22 11 6 6 3 2,849 16,692731 Road and Rail Transport Drivers 15 16 31 24 13 7 6 4 6,054 35,622511 Secretaries and Personal Assistants 15 7 22 26 20 9 6 8 2,421 14,545621 Intermediate Sales and Related Workers 14 7 20 22 17 12 8 13 3,365 21,115921 Process Workers 14 15 39 20 11 6 6 3 1,522 9,425222 Sales, Marketing and Advertising Professionals 14 7 12 17 18 15 5 25 1,739 11,354111 General Managers and Administrators 14 3 5 7 8 11 6 60 2,856 18,103498 Miscellaneous Tradespersons and Related Workers 14 9 23 21 19 12 7 10 1,361 8,779614 Intermediate Numerical Clerks 14 8 27 29 13 7 6 11 2,913 19,274611 General Clerks 14 12 32 26 11 6 5 7 2,013 13,005

…contd./

21

Table 1: Household in stress by income and occupation, Sydney 2001 …/contd.

Proportion of households in stress by household income distribution (%)

Occupation (3 digit ASCO category)In stress

(%) $0-$399$400-$599

$600-$799

$800-$999

$1000-$1199

$1200-$1499 $1500+

Total no. of households

in stressTotal no. of households

619 Miscellaneous Intermediate Clerical Workers 14 10 27 29 14 7 6 7 2,100 14,138232 Nursing Professionals 13 9 21 21 18 14 7 10 2,173 14,204999 Miscellaneous Labourers and Related Workers 13 15 32 22 11 6 7 6 1,154 7,819441 Structural Construction Tradespersons 13 8 22 25 20 11 8 6 2,798 18,677119 Miscellaneous Generalist Managers 13 7 16 18 17 12 10 19 1,878 12,538321 Finance Associate Professionals 13 4 8 13 11 12 5 47 1,840 12,996991 Mining, Construction and Related Labourers 13 13 32 24 14 8 6 3 1,189 8,100251 Social Welfare Professionals 13 12 24 21 14 12 8 9 1,287 8,991123 Sales and Marketing Managers 13 4 8 10 12 16 7 43 2,580 18,838339 Miscellaneous Managing Supervisors (Sales and Servi 12 7 15 20 17 15 7 20 1,651 12,064431 Electrical and Electronics Tradespersons 12 6 17 23 21 13 10 10 2,834 21,018615 Material Recording and Despatching Clerks 12 7 22 26 17 9 8 11 1,530 11,452329 Miscellaneous Business and Administration Associate 12 5 15 22 18 13 8 19 4,378 32,925711 Mobile Plant Operators 12 6 30 27 15 10 6 6 1,056 8,120238 Miscellaneous Health Professionals 11 8 13 15 17 12 8 27 1,073 8,500252 Miscellaneous Social Professionals 11 5 6 7 7 13 5 57 1,256 10,132229 Miscellaneous Business and Information Professionals 11 5 9 16 15 17 6 33 2,300 19,292121 Resource Managers 10 1 4 7 10 14 7 58 1,365 12,521223 Computing Professionals 10 5 8 12 14 20 8 33 2,574 25,233129 Miscellaneous Specialist Managers 9 4 7 10 14 21 8 37 1,431 13,913221 Accountants, Auditors and Corporate Treasurers 9 4 9 15 16 16 8 32 1,909 18,930312 Building and Engineering Associate Professionals 9 4 14 16 19 16 13 17 1,267 12,796411 Mechanical Engineering Tradespersons 8 5 16 25 19 12 11 12 1,390 14,999122 Engineering, Distribution and Process Managers 8 3 6 11 14 15 9 43 1,792 19,702241 School Teachers 8 7 12 17 19 19 9 19 2,236 25,195212 Building and Engineering Professionals 8 5 9 15 16 19 11 26 1,397 16,204391 Police Officers 8 0 1 13 21 20 14 31 443 5,255All households 14 13 24 19 14 10 6 14 141,014 902,576 *Occupations with less than 1,000 households in stress generally not reported

Source: Special request tabulation, 2001 Census

22

Table 2: Households in stress by income and selected occupations, Sydney 2001 Proportion of households by household income distribution (%)

In stress (%) $0-$399

$400-$599

$600-$799

$800-$999

$1000-$1199

$1200-$1499 $1500+

Total no. of households

in stressTotal no. of households

Summary distributionsHouseholds in stress 100 13 24 19 14 10 6 14 141,014 141,014All households 14 4 8 11 11 12 12 41 141,014 902,576Selected occupations223 Computing Professionals 10 1 2 3 6 13 8 68 2,574 25,233232 Nursing Professionals 13 3 7 11 15 16 12 35 2,173 14,204241 School Teachers 8 1 3 6 9 19 12 50 2,236 25,195391 Police Officers 8 0 0 3 7 14 16 60 443 5,255632 Hospitality Workers 33 18 23 16 12 9 10 12 3,470 9,451639 Miscellaneous Intermediate Service Workers 22 8 14 14 12 11 12 28 2,238 9,268821 Sales Assistants 23 13 19 15 13 10 11 19 6,617 25,338911 Cleaners 20 13 21 16 13 10 11 16 3,473 15,368All households 14 4 8 11 11 12 12 41 141,014 902,576

Source: Special request tabulation, 2001 Census

23

Table 3: Residential self-containment, Sydney 2001

Origin Blac

ktow

n

Can

terb

ury-

Bank

stow

n

Cen

tral N

orth

ern

Sydn

ey

Cen

tral W

este

rn

Sydn

ey

East

ern

Sub

urbs

Fairf

ield

-Li

verp

ool

Gos

ford

-Wyo

ng

Illaw

arra

SD

Bal

Inne

r Syd

ney

Inne

r Wes

tern

Sy

dney

Low

er N

orth

ern

Sydn

ey

New

cast

le

Nor

ther

n Be

ache

s

Out

er S

outh

W

este

rn S

ydne

y

Out

er W

este

rn

Sydn

ey

St G

eorg

e-Su

ther

land

Wol

long

ong

Tota

l

Blacktown 32 2 8 19 1 5 0 0 12 3 6 0 0 0 6 1 0 100Canterbury-Bankstown 1 32 1 7 2 5 0 0 27 6 5 0 0 1 0 7 0 100Central Northern Sydney 4 1 37 10 1 2 0 0 17 2 18 0 2 0 1 0 0 100Central Western Sydney 5 5 6 38 1 5 0 0 17 5 10 0 1 1 1 1 0 100Eastern Suburbs 0 1 1 2 34 0 0 0 46 1 8 0 1 0 0 2 0 100Fairfield-Liverpool 3 10 1 13 1 40 0 0 13 3 3 0 0 3 1 2 0 100Gosford-Wyong 0 0 6 2 0 0 68 0 6 1 6 3 1 0 0 0 0 100Illawarra SD Bal 0 0 0 1 0 1 0 83 2 0 0 0 0 2 0 0 3 100Inner Sydney 0 2 1 3 7 1 0 0 66 3 11 0 1 0 0 2 0 100Inner Western Sydney 1 5 2 8 2 2 0 0 36 25 11 0 1 0 1 2 0 100Lower Northern Sydney 1 1 4 4 2 1 0 0 31 2 47 0 3 0 0 1 0 100Newcastle 0 0 0 0 0 0 2 0 1 0 0 89 0 0 0 0 0 100Northern Beaches 0 0 3 2 1 0 0 0 18 1 17 0 52 0 0 0 0 100Outer S W Sydney 2 7 1 6 1 14 0 1 11 2 2 0 0 44 2 2 1 100Outer Western Sydney 11 2 3 11 0 4 0 0 7 2 3 0 0 1 50 0 0 100St George-Sutherland 0 6 1 3 3 2 0 0 30 2 4 0 0 1 0 42 0 100Wollongong 0 1 0 1 0 1 0 1 5 0 1 0 0 2 0 5 78 100Grand Total 4 4 6 8 3 5 4 2 21 3 10 8 3 3 4 5 4 100 Source: Special request tabulation, 2001 Census

24

Table 4: Employment self-containment, Sydney 2001

Origin Blac

ktow

n

Can

terb

ury-

Bank

stow

n

Cen

tral N

orth

ern

Sydn

ey

Cen

tral W

este

rn

Sydn

ey

East

ern

Sub

urbs

Fairf

ield

-Li

verp

ool

Gos

ford

-Wyo

ng

Illaw

arra

SD

Bal

Inne

r Syd

ney

Inne

r Wes

tern

Sy

dney

Low

er N

orth

ern

Sydn

ey

New

cast

le

Nor

ther

n Be

ache

s

Out

er S

outh

W

este

rn S

ydne

y

Out

er W

este

rn

Sydn

ey

St G

eorg

e-Su

ther

land

Wol

long

ong

Tota

l

Blacktown 46 3 7 13 1 5 0 0 3 5 3 0 1 1 7 1 0 5Canterbury-Bankstown 1 39 1 5 3 5 0 0 7 11 3 0 1 2 0 7 0 5Central Northern Sydney 10 3 61 13 2 3 1 0 8 7 17 0 5 1 3 1 0 9Central Western Sydney 8 6 6 27 1 6 0 0 4 9 6 0 1 1 2 1 0 5Eastern Suburbs 0 1 1 1 55 0 0 0 12 2 4 0 1 0 0 2 0 5Fairfield-Liverpool 5 13 2 10 1 50 0 0 4 6 2 0 0 8 2 3 0 6Gosford-Wyong 1 1 6 1 1 0 93 0 2 1 3 2 1 0 0 0 0 5Illawarra SD Bal 0 0 0 0 0 0 0 93 0 0 0 0 0 2 0 0 2 2Inner Sydney 1 3 1 3 15 1 0 0 21 6 7 0 1 0 0 3 0 7Inner Western Sydney 1 4 1 3 2 2 0 0 6 30 4 0 1 1 0 1 0 3Lower Northern Sydney 1 2 5 4 4 1 0 0 10 6 33 0 5 0 1 1 0 7Newcastle 0 0 0 0 0 0 5 0 0 0 0 98 0 0 0 0 0 9Northern Beaches 1 1 3 1 2 0 0 0 4 1 9 0 81 0 0 0 0 5Outer S W Sydney 2 8 1 4 1 14 0 2 3 3 1 0 0 77 2 2 1 5Outer Western Sydney 21 2 4 10 1 6 0 0 2 4 2 0 0 2 81 0 0 7St George-Sutherland 1 14 1 4 9 4 0 0 13 7 4 0 1 2 0 73 1 9Wollongong 0 1 0 0 1 1 0 4 1 1 0 0 0 3 0 4 96 5Grand Total 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 Source: Special request tabulation, 2001 Census

25

Table 5: Places of residence and work, 2001 and change from 1996

2001 growth

1996-2001(%)

2001 growth

1996-2001(%)

Job deficit/surplus

2001

Median Rent2001($pw)

Blacktown 107,911 10.5 69,966 15.3 -37,945 170Canterbury-Bankstown 112,777 -0.3 85,832 -3.3 -26,945 175Central Northern Sydney 190,912 6.3 110,176 11.7 -80,736 288Central Western Sydney 115,077 2.4 150,030 -1.0 34,953 195Eastern Suburbs 112,496 -1.3 66,183 4.0 -46,313 300Fairfield-Liverpool 127,366 12.6 92,709 6.0 -34,657 155Gosford-Wyong 108,851 11.9 75,409 11.9 -33,442 170Illawarra SD balance 42,562 10.5 36,345 13.7 -6,217 ~100Inner Sydney 142,576 13.7 420,679 4.9 278,103 250Inner Western Sydney 72,248 6.1 57,517 -3.1 -14,731 245Lower Northern Sydney 142,983 1.6 191,537 3.4 48,554 295Newcastle 180,892 3.9 157,148 3.0 -23,744 105Northern Beaches 111,215 0.3 68,762 3.6 -42,453 290Outer S W Sydney 99,236 11.3 53,660 16.8 -45,576 160Outer Western Sydney 141,913 7.3 83,473 11.2 -58,440 175St George-Sutherland 196,711 4.7 107,238 6.0 -89,473 225Wollongong 102,137 5.8 79,532 1.1 -22,605 110Total* 2,107,863 5.9 2,107,863 5.9 -201,667

Residential location Workplace location

* total for destination includes not stated, worked outside area, no fixed address

Source: Special request tabulation, 2001 Census

26

Figure 1: Change in proportions of workers living and working in each region: 1996-2001

Source: Special request tabulation, 2001 Census

-0.6%

-0.4%

-0.2%

0.0%

0.2%

0.4%

0.6%

Eas

tern

Sub

urbs

Low

er N

orth

ern

Syd

ney

Nor

ther

n B

each

es

Cen

tral N

orth

ern

Syd

ney

Inne

r Syd

ney

Inne

r Wes

tern

Syd

ney

St G

eorg

e-S

uthe

rland

Cen

tral W

este

rn S

ydne

y

Can

terb

ury-

Ban

ksto

wn

Out

er W

este

rn S

ydne

y

Bla

ckto

wn

Gos

ford

-Wyo

ng

Out

er S

outh

Wes

tern

Syd

ney

Fairf

ield

-Liv

erpo

ol

Wol

long

ong

New

cast

le

change in origin change in destination

27

Table 6: Residential location by occupation of workers, 2001 and change from 1996

Residential location

2001growth 96-01 2001

growth 96-01 2001

growth 96-01 2001

growth 96-01 2001

growth 96-01 2001

growth 96-01

(no.) (%) (no.) (%) (no.) (%) (no.) (%) (no.) (%) (no.) (%)Blacktown 1,936 64.8 1,962 16.4 1,218 13.4 2,281 6.5 100,514 9.8 107,911 10.5Canterbury-Bankstown 2,104 52.6 1,418 -2.3 1,605 4.0 2,922 -3.0 104,728 -0.9 112,777 -0.3Central Northern Sydney 6,902 56.8 4,196 1.6 2,457 18.2 1,685 -1.9 175,672 5.1 190,912 6.3Central Western Sydney 3,282 61.5 2,097 -7.8 1,574 5.2 2,530 -2.4 105,594 1.6 115,077 2.4Eastern Suburbs 3,118 44.7 2,024 -13.4 2,674 -9.1 1,253 -20.2 103,427 -1.5 112,496 -1.3Fairfield-Liverpool 1,785 74.0 1,612 17.5 1,547 18.8 3,119 8.6 119,303 11.9 127,366 12.6Gosford-Wyong 1,161 61.7 2,634 12.5 1,901 19.3 2,372 11.2 100,783 11.4 108,851 11.9Illawarra SD Bal 200 66.7 914 12.8 961 13.9 1,056 4.6 39,431 10.3 42,562 10.5Inner Sydney 4,787 90.4 2,353 -3.0 3,967 14.5 2,367 -17.5 129,102 13.1 142,576 13.7Inner Western Sydney 2,519 80.6 1,389 -6.0 1,259 15.1 1,216 -10.0 65,865 5.0 72,248 6.1Lower Northern Sydney 6,668 50.6 2,431 -9.6 2,410 18.3 1,403 -15.8 130,071 0.1 142,983 1.6Newcastle 1,242 37.7 4,948 4.0 3,743 12.2 4,912 4.0 166,047 3.5 180,892 3.9Northern Beaches 2,694 45.4 1,957 -10.4 1,739 7.7 1,461 -6.4 103,364 -0.3 111,215 0.3Outer S W Sydney 966 53.3 1,679 12.8 1,342 15.5 1,811 7.4 93,438 11.0 99,236 11.3Outer Western Sydney 1,567 32.7 2,933 3.2 2,130 9.5 2,402 1.5 132,881 7.2 141,913 7.3St George-Sutherland 4,067 53.5 3,428 2.5 3,112 13.7 3,290 -0.3 182,814 4.0 196,711 4.7Wollongong 1,228 43.1 2,126 12.6 1,983 10.8 2,504 3.4 94,296 5.3 102,137 5.8Total 46,226 57.1 40,101 1.5 35,622 11.1 38,584 -1.0 1,947,330 5.2 2,107,863 5.9

All others including not stated Total

223 Computing

Professionals

232 Nursing

Professionals

632 Hospitality Workers

911 Cleaners

Source: Special request tabulation, 2001 Census

28

Table 7: Workplace location by occupation of workers, 2001 and change from 1996

Workplace location

2001growth 96-01 2001

growth 96-01 2001

growth 96-01 2001

growth 96-01 2001

growth 96-01 2001

growth 96-01

(no.) (%) (no.) (%) (no.) (%) (no.) (%) (no.) (%) (no.) (%)Blacktown 417 37.6 984 -0.3 706 3.7 1,136 6.1 66,723 15.8 69,966 15.3Canterbury-Bankstown 549 35.9 1,272 9.0 987 -6.5 1,454 -7.3 81,570 -3.6 85,832 -3.3Central Northern Sydney 2,579 39.6 2,491 1.4 1,689 19.4 1,485 4.3 101,932 11.4 110,176 11.7Central Western Sydney 2,484 49.5 4,018 -0.5 1,497 3.0 2,070 -16.0 139,961 -1.3 150,030 -1.0Eastern Suburbs 876 26.4 2,225 0.1 1,799 -8.9 1,182 -8.7 60,101 4.6 66,183 4.0Fairfield-Liverpool 484 18.6 1,944 24.9 1,076 9.7 1,508 10.2 87,697 5.5 92,709 6.0Gosford-Wyong 341 68.8 2,289 12.4 1,745 21.4 1,672 16.2 69,362 11.4 75,409 11.9Illawarra SD Bal 131 37.9 815 15.3 942 14.2 875 8.3 33,582 13.7 36,345 13.7Inner Sydney 17,937 74.5 3,502 -13.8 8,271 9.6 4,279 -25.3 386,690 3.5 420,679 4.9Inner Western Sydney 1,356 1.0 1,613 -13.7 653 24.1 976 -6.0 52,919 -3.0 57,517 -3.1Lower Northern Sydney 12,614 65.7 3,431 -5.7 2,491 11.2 2,019 -11.4 170,982 0.9 191,537 3.4Newcastle 1,122 37.5 4,605 2.6 3,497 13.6 4,174 -0.9 143,750 2.7 157,148 3.0No fixed address 686 116.4 559 86.3 259 115.8 4,528 41.2 81,183 49.3 87,215 49.5Northern Beaches 1,144 39.5 1,199 -5.1 1,349 7.0 912 -18.1 64,158 3.6 68,762 3.6Not stated 842 160.7 1,082 66.7 1,594 79.1 3,087 44.6 86,610 34.2 93,215 36.0Outer S W Sydney 158 7.5 936 3.0 872 20.3 1,065 22.3 50,629 17.0 53,660 16.8Outer Western Sydney 405 52.8 1,955 6.7 1,809 10.3 1,685 7.8 77,619 11.3 83,473 11.2St George-Sutherland 945 39.2 3,109 4.9 2,185 14.9 1,855 -0.7 99,144 5.7 107,238 6.0Wollongong 857 21.4 1,792 8.5 1,789 11.4 1,943 -3.9 73,151 0.6 79,532 1.1Total 46,226 57.1 40,101 1.5 35,622 11.1 38,584 -1.0 1,947,330 5.2 2,107,863 5.9

223 Computing

Professionals

232 Nursing

Professionals

632 Hospitality Workers

911 Cleaners

All others including not stated Total

Source: Special request tabulation, 2001 Census

29

Table 8: Job deficit/surplus by occupation, 2001

223 Computing

Professionals

232 Nursing

Professionals

632 Hospitality

Workers

911 Cleaners

Others including Not

StatedTotal

Blacktown -1,519 -978 -512 -1,145 -33,791 -37,945Canterbury-Bankstown -1,555 -146 -618 -1,468 -23,158 -26,945Central Northern Sydney -4,323 -1,705 -768 -200 -73,740 -80,736Central Western Sydney -798 1,921 -77 -460 34,367 34,953Eastern Suburbs -2,242 201 -875 -71 -43,326 -46,313Fairfield-Liverpool -1,301 332 -471 -1,611 -31,606 -34,657Gosford-Wyong -820 -345 -156 -700 -31,421 -33,442Illawarra SD Bal -69 -99 -19 -181 -5,849 -6,217Inner Sydney 13,150 1,149 4,304 1,912 257,588 278,103Inner Western Sydney -1,163 224 -606 -240 -12,946 -14,731Lower Northern Sydney 5,946 1,000 81 616 40,911 48,554Newcastle -120 -343 -246 -738 -22,297 -23,744Northern Beaches -1,550 -758 -390 -549 -39,206 -42,453Outer South Western Sydney -808 -743 -470 -746 -42,809 -45,576Outer Western Sydney -1,162 -978 -321 -717 -55,262 -58,440St George-Sutherland -3,122 -319 -927 -1,435 -83,670 -89,473Wollongong -371 -334 -194 -561 -21,145 -22,605 Source: Special request tabulation, 2001 Census

30

Figure 2: Change in origin of all workers in Sydney by occupation, 1996-2001

Figure 3: Change in destination of all workers in Sydney by occupation, 1996-2001

Source: Special request tabulation, 2001 Census

-2.0%-1.5%-1.0%

-0.5%0.0%0.5%1.0%

1.5%2.0%

Eas

tern

Sub

urbs

Low

er N

orth

ern

Syd

ney

Nor

ther

n B

each

es

Cen

tral N

orth

ern

Syd

ney

Inne

r Syd

ney

Inne

r Wes

tern

Syd

ney

St G

eorg

e-S

uthe

rland

Cen

tral W

este

rn S

ydne

y

Can

terb

ury-

Ban

ksto

wn

Out

er W

este

rn S

ydne

y

Bla

ckto

wn

Gos

ford

-Wyo

ng

Out

er S

W S

ydne

y

Fairf

ield

-Liv

erpo

ol

Wol

long

ong

New

cast

le

223 Computing Professionals 232 Nursing Professionals 632 Hospitality Workers

911 Cleaners All others

-2.0%-1.5%-1.0%-0.5%0.0%0.5%1.0%1.5%2.0%

Eas

tern

Sub

urbs

Low

er N

orth

ern

Syd

ney

Nor

ther

n B

each

es

Cen

tral N

orth

ern

Syd

ney

Inne

r Syd

ney

Inne

r Wes

tern

Syd

ney

St G

eorg

e-S

uthe

rland

Cen

tral W

este

rn S

ydne

y

Can

terb

ury-

Ban

ksto

wn

Out

er W

este

rn S

ydne

y

Bla

ckto

wn

Gos

ford

-Wyo

ng

Out

er S

W S

ydne

y

Fairf

ield

-Liv

erpo

ol

Wol

long

ong

New

cast

le

223 Computing Professionals 232 Nursing Professionals 632 Hospitality Workers

911 Cleaners All others

31

Table 9: Measures of spatial concentration of workers by occupation Occupation

223 Computing

Professionals

232 Nursing

Professionals

632 Hospitality

Workers

911 Cleaners

Others including

Not StatedGrand Total

1996Dissimilarity index (O-D) 0.40 0.14 0.15 0.15 0.22 0.22Herfindahl index (origin) 0.09 0.07 0.07 0.07 0.06 0.06Herfindahl index (destination) 0.22 0.07 0.10 0.08 0.09 0.09

2001Dissimilarity index (O-D) 0.44 0.14 0.15 0.14 0.21 0.21Herfindahl index (origin) 0.09 0.07 0.07 0.07 0.06 0.06Herfindahl index (destination) 0.25 0.07 0.10 0.08 0.09 0.09 Source: Special request tabulation, 2001 Census

32

Figure 4: Origin of workers whose destination was Inner Sydney SSD, 2001

Source: Special request tabulation, 2001 Census

33

Figure 5: Change in origin of workers in inner Sydney, 1996-2001

Figure 2: Change in origin of all workers in Sydney by occupation, 1996-2001

Source: Special request tabulation, 2001 Census

-4.0%

-3.0%

-2.0%

-1.0%

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%E

aste

rn S

ubur

bs

Low

er N

orth

ern

Syd

ney

Nor

ther

n B

each

es

Cen

tral N

orth

ern

Syd

ney

Inne

r Syd

ney

Inne

r Wes

tern

Syd

ney

St G

eorg

e-S

uthe

rland

Cen

tral W

este

rn S

ydne

y

Can

terb

ury-

Ban

ksto

wn

Out

er W

este

rn S

ydne

y

Bla

ckto

wn

Gos

ford

-Wyo

ng

Out

er S

outh

Wes

tern

Syd

ney

Fairf

ield

-Liv

erpo

ol

Wol

long

ong

New

cast

le

223 Computing Professionals 232 Nursing Professionals 632 Hospitality Workers911 Cleaners All others

-2.0%-1.5%-1.0%

-0.5%0.0%0.5%1.0%

1.5%2.0%

Eas

tern

Sub

urbs

Low

er N

orth

ern

Syd

ney

Nor

ther

n B

each

es

Cen

tral N

orth

ern

Syd

ney

Inne

r Syd

ney

Inne

r Wes

tern

Syd

ney

St G

eorg

e-S

uthe

rland

Cen

tral W

este

rn S

ydne

y

Can

terb

ury-

Ban

ksto

wn

Out

er W

este

rn S

ydne

y

Bla

ckto

wn

Gos

ford

-Wyo

ng

Out

er S

W S

ydne

y

Fairf

ield

-Liv

erpo

ol

Wol

long

ong

New

cast

le

223 Computing Professionals 232 Nursing Professionals 632 Hospitality Workers

911 Cleaners All others

34

Table 10: Household in stress by income and occupation, living in Inner Sydney 2001* Proportion of households in stress by household income distribution (%)

Occupation (3 digit ASCO category)In stress

(%) $0-$399$400-$599

$600-$799

$800-$999

$1000-$1199

$1200-$1499 $1500+

Total no. of households

in stressTotal no. of households

632 Hospitality Workers 49 25 30 17 10 8 6 3 707 1,431493 Hairdressers 40 12 37 16 17 9 6 3 100 248821 Sales Assistants 40 24 31 18 12 7 4 4 810 2,024829 Miscellaneous Elementary Sales Workers 38 30 24 19 11 5 6 5 244 646993 Elementary Food Preparation and Related Workers 37 37 30 15 12 4 0 2 156 418613 Receptionists 36 18 31 23 12 7 4 6 243 682451 Food Tradespersons 35 20 36 19 14 4 4 4 170 481631 Carers and Aides 33 26 35 15 8 6 7 3 272 820811 Elementary Clerks 31 18 38 19 9 9 3 4 208 661831 Elementary Service Workers 31 25 40 19 5 6 2 1 263 845332 Hospitality and Accommodation Managers 30 11 22 24 17 12 7 8 488 1,633331 Shop Managers 30 14 19 23 15 11 8 10 310 1,044639 Miscellaneous Intermediate Service Workers 30 15 26 23 16 10 5 5 271 916799 Miscellaneous Intermediate Production and Transport 29 17 48 14 11 4 5 1 229 800612 Keyboard Operators 28 13 27 36 9 9 2 4 138 489911 Cleaners 28 29 39 20 8 1 1 2 288 1,022249 Miscellaneous Education Professionals 26 24 26 18 10 10 4 9 195 760731 Road and Rail Transport Drivers 25 22 38 20 10 3 4 3 370 1,470498 Miscellaneous Tradespersons and Related Workers 25 11 20 22 22 19 0 6 151 611621 Intermediate Sales and Related Workers 25 6 22 22 19 14 4 12 279 1,130111 General Managers and Administrators 25 3 6 6 10 14 2 59 371 1,507611 General Clerks 25 14 29 27 13 8 1 9 237 963253 Artists and Related Professionals 24 14 18 19 16 15 6 12 857 3,589591 Advanced Numerical Clerks 24 15 14 20 16 17 6 13 109 462619 Miscellaneous Intermediate Clerical Workers 23 9 29 29 16 6 5 5 301 1,306614 Intermediate Numerical Clerks 22 6 26 27 15 11 3 13 288 1,321All households 22 13 22 18 15 13 5 15 15,198 68,845

*Only occupations with above average proportions in stress reported; those with less than 100 households in stress not reported Source: Special request tabulation, 2001 Census

35

Table 11: Household in stress by income and selected occupations, living in Inner Sydney 2001 Proportion of households by household income distribution (%)

In stress (%) $0-$399

$400-$599

$600-$799

$800-$999

$1000-$1199

$1200-$1499 $1500+

Total no. of households

in stressTotal no. of households

Summary distributionsHouseholds in stress living in Inner Sydney 100 13 22 18 15 13 5 15 15,198 15,198Households in stress in Sydney 100 13 24 19 14 10 6 14 141,014 141,014All households living in Inner Sydney 22 5 8 10 10 13 8 45 15,198 68,845All households in Sydney 14 4 8 11 11 12 12 41 141,014 902,576Selected occupations (living in Inner Sydney)223 Computing Professionals 14 4 8 13 19 21 6 28 370 2,679232 Nursing Professionals 20 12 14 23 21 17 3 10 241 1,194241 School Teachers 14 4 13 19 18 25 3 19 214 1,534391 Police Officers 19 0 0 11 33 33 0 22 27 140632 Hospitality Workers 49 25 30 17 10 8 6 3 707 1,431639 Miscellaneous Intermediate Service Workers 30 15 26 23 16 10 5 5 271 916821 Sales Assistants 40 24 31 18 12 7 4 4 810 2,024911 Cleaners 28 29 39 20 8 1 1 2 288 1,022All households 22 5 8 10 10 13 8 45 15,198 68,845

Source: Special request tabulation, 2001 Census

36

Figure 6: Change in origin of workers in inner Sydney, 1996-2001

Source: Special request tabulation, 2001 Census

-3.0%

-2.0%

-1.0%

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%E

aste

rn S

ubur

bs

Low

er N

orth

ern

Syd

ney

Nor

ther

n B

each

es

Cen

tral N

orth

ern

Syd

ney

Inne

r Syd

ney

Inne

r Wes

tern

Syd

ney

St G

eorg

e-S

uthe

rland

Cen

tral W

este

rn S

ydne

y

Can

terb

ury-

Ban

ksto

wn

Out

er W

este

rn S

ydne

y

Bla

ckto

wn

Gos

ford

-Wyo

ng

Out

er S

outh

Wes

tern

Syd

ney

Fairf

ield

-Liv

erpo

ol

Wol

long

ong

New

cast

le

$0-$399 $0400-$599 $0600-$999 $1,000-$1,499 $1,500 or more

37

References ATIS Real Weatheralls (2002) Innovation and Good Practice: Affordable Rented Housing

for Key Workers, prepared for The Housing Corporation, UK. http://www.housingcorplibrary.org.uk/housingcorp.nsf/AllDocuments/32C43460E77D5E2380256C22005187F2

Blunden, H., Darcy, M. Perry, J. and Stubbs, J. (2004) "Housing Costs and Work Location in Sydney", paper presented at the ‘A future that works’ conference at Newcastle University 8-10 December, 2004

Department of Transport, Environment and the Regions (DETR) (2000) The Housing Green Paper - Quality and Choice: A Decent Home for All, DTLR: London

Department of Infrastructure, Planning and Natural Resources (DIPNR) (2004) 2002 Household Travel Survey Summary Report, http://www.planning.nsw.gov.au/tpdc/pdfs/htsreport_2004.pdf

Dockery, A. (2004) "Workforce experience and retention in nursing in Australia", Australian Bulletin of Labour, 30(2):74-102

Dodson, J. (2004) Affordable housing and employment opportunity: is there a spatial mismatch? Australian Housing and Urban Research Institute Positioning Paper, http://www.ahuri.edu.au/gobal/docs/doc723.pdf

Duncan, O. and Duncan, B. (1955) "A Methodological Analysis of Segregation Indexes", American Sociological Review, 20(2): 210-217

Epic DotGov (2004) Northern Beaches Key Workers Study, Final Report produced in association with Glazebrook and Associates for the Department of Infrastructure, Planning and Natural Resources, NSW Department of Housing, Manly Council, Warringah Council

Fiore, M. and Lipman, B. (2003) "Paycheck to Paycheck: Wages and the Cost of Housing in America", New Century Housing, 2(2), Center for Housing Policy/National Housing Conference http://www.nhc.org/nhcimages/paycheck03.pdf. Accessed 23rd November 2004

Florida, R. (2005) The Flight of the Creative Class: The New Global Competition for Talent, New York: HarperCollins

Florida, R. (2002) The rise of the creative class, New York: Basic Books Fordham Research (2005) Research into Key Worker Housing Needs, report prepared for

Chelmsford Borough Council, http://www.chelmsfordbc.gov.uk/planning/Policy/documents/KeyWorkerReport2004.pdf

Gobillon, l., Selod, H., Zenou, Y. (2003) “Spatial Mismatch: From the Hypothesis to the Theories”, Institute for the Study of Labor (IZA), IZA Discussion Paper No. 693, http://papers.ssrn.com/sol3/papers.cfm?abstract_id=382787

Hannah, L. and Kay, J. (1977) "Measures of Concentration", Journal of the American Statistical Association, 62:162-168

38

Houston, D. (2005) "Employability, skills mismatch and spatial mismatch in metropolitan labour markets" Urban Studies, 42(2):221-243

Immergluck, D. (1998) “Job Proximity and the Urban Employment Problem: Do Suitable Nearby Jobs Improve Neighbourhood Employment Rates?” Urban Studies, 35(1):7-23

Jacobs, J. (1962) The death and life of great American cities, London: Jonathon Cape.

Massey, D. and Denton, N. (1988) "The Dimensions of Residential Segregation" Social Forces, 67(2):281-315

Monk, S., Morrison N., Smith K., Stubley M. and Whitehead C. (2002), Research into Key Worker and Affordable Housing in the Cambridge Area, Research Report V, Department of Land Economy, University of Cambridge as quoted in Fordham Research (2005).

NSW Department of Housing (DoH) (2004) Thornleigh Key Worker Affordable Housing Program. a New South Wales Land and Housing Corporation Affordable Housing Demonstration Project, internal document

O'Connor, K. and Healy, E. (2002) The Links Between Housing Markets and Labour Markets in Melbourne, Australian Housing and Urban Research Institute Final Report. http://www.ahuri.edu.au

O'Malley, N. (2005) “Too far for clean run to city”, Sydney Morning Herald, June 2, 2005

Office of the Deputy Prime Minister (ODPM)(2003) Improving the delivery of affordable housing in London and the South East, http://www.odpm.gov.uk/stellent/groups/odpm_housing/documents/downloadable/odpm_house_023692.pdf

Office of the Deputy Prime Minister (ODPM) (2004a) Key Worker Housing, http://www.odpm.gov.uk/stellent/groups/odpm_housing/documents/page/odpm_house_027908.hcsp

Office of the Deputy Prime Minister (ODPM) (2004b) Creating Sustainable Communities, http://www.odpm.gov.uk/stellent/groups/odpm_housing/documents/page/odpm_house_027911.hcsp

Ong, P. and Blumenberg, E. (1998) “Job Access, Commute and Travel Burden among Welfare Recipients” Urban Studies, 35(1):77-93

Queensland Department of Housing (2003) Housing Analysis for Kelvin Grove Village, August 2003, internal report

Randolph, B. (1991) "Housing markets, labour markets and discontinuity theory", in Allen, J. and Hamnett, C. (eds) Housing and Labour Markets, London: Allen and Unwin

Saunders, P. (2003) Examining Recent Changes in Income Distribution in Australia, SPRC Discussion Paper 130, http://www.sprc.unsw.edu.au/dp/DP130.pdf

Saunders, P. (2005) “Wage inequalities among full-time workers”, paper prepared for the Australian Social Policy Conference 2005, Looking back, Looking Forward, A Quarter of a Century of Social Change, University of New South Wales, 20-22 July 2005, http://www.sprc1.sprc.unsw.edu.au/aspc2005/abstract.asp?PaperID=16

39

Smith, T. and Zenou, Y. (2003) "Spatial Mismatch, Search Effort and Urban Spatial Structure” Journal of Urban Economics, 54(1):129-56

Stegman, M., Quercia, R. and McCarthy, G. (2000) "Housing America’s Working Families", New Century Housing, 1(1), Center for Housing Policy, http://www.nhc.org/nhcimages/chprpt.pdf

Stimson, R. (2001) "Dividing Societies: The Socio-Political Spatial Implications of Restructuring in Australia" Australian Geographical Studies, 39(2):198-216

Stoll, M. (2005) “Geographical Skills, Mismatch and Race” Urban Studies, 42(4):695-717

Theil, H. (1972) Statistical Decomposition Analysis, Amsterdam: North-Holland

Tym, R. et al (2003) Cambridge Sub-Region Key Worker Housing Research, Final Report prepared for the Cambridge Sub-Regional Affordable Housing Working Group, http://www.cambridge.gov.uk/planning/reptdocs/key_worker_housing_research.pdf

Vipond, J., Castle, K. and Cardew, R. (1998) "Revival in Inner Areas", Australian Planner 35(4):215-222

Watson, I., Buchanan, J., Campbell, I. and Briggs, C. (2003) Fragmented Futures: New Challenges in Working Life, Leichhardt: The Federation Press

Wulff, M., Healy, E. and Reynolds, M. (2004) "Why don't small households live in small dwellings? disentangling a planning dilemma", People and Place, 12(1):58-71

Winter, I. and Stone, W. (1999) "Social Polarisation and Housing Careers: exploring the interrelationship of labour and housing markets in Australia" in O'Connor, K. (ed) Houses and Jobs in Cities and Regions: Research in Honour of Chris Maher, Brisbane: University of Queensland Press

Winter, I. and Stone, W. (1998) Social Polarisation and Housing Careers, Working Paper No.13, Australian Institute of Family Studies, Melbourne

Yates, J. (2001) "The Rhetoric and Reality of Housing Choice: The Role of Urban Consolidation ", Urban Policy and Research, 19(4):491-527

Yates, J., Randolph, B., Holloway, D. and Murray, D. (2005a) Housing Affordability, Occupation and Location in Australian Cities and Regions, Australian Housing and Urban Research Institute Positioning Paper, May 2005, http://www.ahuri.edu.au/global/docs/doc789.pdf

Yates, J., Randolph, B. and Holloway, D. (2005b) Housing Affordability, Occupation and Location in Australian Cities and Regions, Australian Housing and Urban Research Institute Final Report, forthcoming.

40

Appendix A: Supplementary data Figure A. 1: Intra-metropolitan median house prices, Sydney 1994-2003

Source: NSW DoH Sales Reports

Figure A. 2: Share of low rent dwellings in each LGA in Sydney, 2001

Source: Special matrix tabulation, ABS Census 2001

fringe

middle

inner

fringe

middle

inner

0

500

1,000

1,500

2,000

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

($ ' 000s)

Mosman Woollahra Willoughby Ku-ring-gai Warringah

Kogarah Parramatta Liverpool Wyong

41

Table A. 1: Detailed geography of Sydney Statistical Subdivisions

Zone* SSD SLA

middle Blacktown Blacktown middle Canterbury-Bankstown Canterbury, Bankstown middle Central Northern Sydney Baulkham Hills, Hornsby, Ku-ring-gai middle Central Western Sydney Auburn, Holroyd, Parramatta inner Eastern Suburbs Randwick, Waverley, Woollahra middle Fairfield-Liverpool Fairfield, Liverpool fringe Gosford-Wyong Gosford, Wyong Illawarra SD Bal inner Inner Sydney Botany, Leichhardt, Marrickville, Sydney (C) inner Inner Western Sydney Ashfield, Burwood, Canada Bay(Concord, Drummoyne), Strathfield inner Lower Northern Sydney Hunter's Hill, Lane Cove, Mosman, North Sydney, Ryde. Willoughby Newcastle middle Northern Beaches Manly, Pittwater, Warringah fringe Outer S W Sydney Camden, Campbelltown, Wollondilly fringe Outer Western Sydney Blue Mountains, Hawkesbury, Penrith middle St George-Sutherland Hurstville, Kogarah, Rockdale, Sutherland Wollongong

*Yates (2001) provides a rationale for this zoning classification which is based on a ring based system from the CBD. The inner zones contain statistical sub-divisions within ten kilometres of the city centre and have the highest population densities. The outer zones, contain statistical sub-divisions roughly 25 to 30 kilometres from the centre, have the lowest population densities and the greatest supply of land available for residential development.