SYDNEY’S ETHNIC UNDERCLASS

Bob Birrell and Byung-Soo Seol

Immigration is back on the political agenda. Leading representatives of corporate Australia have worked assiduously during 1998 to put the issue on the policy table.¹ The leader of the Federal Labor Party, Mr Beazley, has declared his Party’s in-principle support for an expanded immigration program in the context of a renewed ‘nation building’ strategy for Australia’s future. As a result, further debate on the role of immigration in Australia’s long-term economic and strategic outlook is inevitable.

According to Mr Beazley, ‘Labor believes that increased immigration levels can be a solution to many of the problems Australia faces in the long term’.² He had in mind issues of population ageing and defence when making this statement. But like a number of recent advocates for a regenerated immigration program, Mr Beazley seems to assume that the slate is clean on the impact of past immigration flows.

As is well known, Australia has experienced the largest intake of overseas migrants relative to population of all Western nations. Currently, overseas-born persons make up 23 per cent of Australia’s population compared to around 9 per cent of the USA population and 15 per cent of Canada’s. The achievement of accommodating such a massive influx has been momentous. While the notion of embarking on a renewed nation-building immigration program clearly excites some business and political leaders, we should pause for a moment and ask whether the slate really is clean.

The central problem in justifying any renewed immigration program is how it fits with the current restructuring of Australia’s economic landscape. Most Australian businesses now confront far more intense competitive pressures. This is a direct consequence of the globalisation process and the decisions of successive Australian Governments to open the domestic markets to competition. Businesses operating in low-skilled, labour-intensive areas are particularly vulnerable if the goods or services in question can be imported.

Where would a renewed migration program fit in given this context? For the past couple of decades most of the growth in the migration program has come from family reunion, humanitarian and on-shore change of status flows, none of which are selected on the basis of skills. The majority of these people have come from Non-English-Speaking-Background (NESB) countries. In the 1950s and 1960s there were plenty of entry-level jobs for such persons, particularly in manufacturing firms. But in the decade 1986 to 1996 the proportion of Australia’s employed workforce holding manufacturing jobs fell from 16.2 per cent to 13.5 per cent.³ Meanwhile, there has been a rapid growth in employment in the service industries, particularly in services to property and business (much of it at the professional-level in computing, accounting and legal and business and marketing services), in health and educational services (again knowledge intensive) and retail services. Knowledge of English is usually essential in these areas. This is true even of routine retail work, given the customer orientation. There has been some growth in jobs accessible to less skilled NESB persons particularly in the accommodation and food service areas and in cleaning services. But these labour markets tend to be highly competitive and thus relatively low paid.

To understand what might happen with a renewed migration program we need to look at the experience of lower-skilled migrants since the onset of the economic changes summarised above. The analysis has to be regionally focussed because of the highly selective locational preferences of recently arrived migrants. The main destination of the family, humanitarian and change-of-status migration flows has been Sydney. By 1996, 21 per cent of Australia’s population lived in Sydney, but the proportion of Australia’s overseas-born residents living there was 31.7 per cent. The overseas-born were drawn heavily from recent arrivals. Over the decade from 1986 to 1996, Sydney has been the favoured location of most of the newer migrant streams, particularly those from Asia. By 1996, some 37.4 per cent of all migrants who arrived in Australia between 1991-96, and 36.5 per cent of those arriving 1986-91, were living in Sydney (compared with 26.1 per cent of those arriving before 1986).

Table 1 shows the proportion of migrants living in Sydney by major country of birth groups by time of arrival. The NESB streams are highly concentrated in Sydney. For example, 55.7 per cent of the Chinese arriving in Australia over the 1991-96 period were living in Sydney in 1996, as were 63 per cent of those born in Korea and 73.6 per cent of those born in the Lebanon. On the other hand, 25.7 per cent of 1991-96 arrivals born in the UK and Ireland and 26.4 per cent of those born in New Zealand were located in Sydney.

Sydney therefore constitutes a crucial test case of the consequences of structural change for the economic well-being of recent migrants to Australia. Should any expansion in the migrant intake occur it is likely to be drawn primarily from Asian sources because of the ‘push’ pressures from that continent. If they were primarily family and humanitarian migrants it is likely that they would follow the recent settlement pattern because of the location of family, friends and ethnic communities. If the emphasis was on skilled migrants the outcomes described below would be less likely be replicated in future. But whatever the future migration pattern, Australia, and Sydney in particular, has still to come to grips with the economic and social legacy of recent migration movements.

Sydney is Australia’s richest city. The total employed population has increased from 1.548 million in 1986 to 1.808 million in 1996 — an increase of 260 thousand or 16.8 per cent. This compares favourably with the city’s population growth over the same period of 11.8 per cent. But this job growth has been selective. Sydney has been a prime beneficiary of the economic adjustment process by virtue of its role as the main gateway for international capital into Australia and as the preferred location of leading Australian and overseas corporations. As a consequence much of Sydney’s employment growth is in services to property and business. Out of the total net growth in employment of 260 thousand, some 99 thousand jobs, or 38.1 per cent, were in services to property and business, particularly in the highly skilled fields described above. The other major areas of job growth were in education and health and in the retail industry. These industries offer few opportunities to NESB migrants with low skills or poor English. The relatively buoyant accommodation and food service and construction sectors offered some employment opportunities for such migrants. But, on the other hand, over the decade to 1996, there was a major contraction in manufacturing employment in Sydney of 41.1 thousand, thus diminishing opportunities in an industry which was once a major provider of entry level jobs for NESB migrants.

The implication is that economic change in Sydney does not fit with the parallel migration movements. Some of the consequences have been widely discussed, particularly the cluster of low income Vietnamese-born residents in the Cabramatta area of Fairfield. There is a bigger picture, however, which has not been acknowledged by commentators. This is the growing concentration of diverse NESB birthplace groups within a relatively tightly bounded region of Sydney’s south-west. We first describe the pattern then explore why it is occurring and some of the consequences.

Given the employment market for low-skilled migrants, it is not surprising that many have had to accept low paid work, or struggled to find employment at all. We have used annual income (as reported in the 1996 Census) as an indicator of Table 1 these outcomes. In order to control for the effects of age and gender we have focused on males aged 25-44. Four income categories were chosen, with the lowest category of $300 or less per week ($15,600 per year) selected to identify the very poor. Men in this age group would normally be expected to be responsible for a household. However, if they are in this very poor category they would barely be able to provide for themselves, let alone a dependent household. Yet 19.2 per cent of Australia-born men aged 25-44 were in this parlous situation. In Sydney the figure was 16.7 per cent and in Melbourne it was 17.9 per cent.⁴ However, the Sydney figure hides a sharp differentiation between overseas-born and Australia-born men, with just 14 per cent of the latter in the very-poor category compared with 21.7 per cent of the former.

There is also a sharp division between the various birthplace groups. Some have done as well or better than the Australia-born, notably those from South Africa, the UK and Ireland, New Zealand, Malaysia and India. These migrants have in common an English-speaking background, including those from Malaysia and India, many of whom came as skill-selected migrants with their original training in universities where English was the language of instruction. They have been able to successfully compete for employment in the growth areas within Sydney’s economy.

However, migrants from NESB countries who mostly entered via the family and humanitarian programs or via the various amnesties granted since 1976, are struggling. In almost all the South-East and East-Asian born groupings of men aged 25-44 listed in Table 1 (with the partial exception of Hong Kong) some 30 per cent or more are located in the low income category. Another 35 to 50 per cent are located in the $300-$599 per week income range, which by Sydney standards puts one in the battler class. It is this group of migrants, collectively designated here as Sydney’s ethnic underclass, whose situation is at issue in this paper.

Many low-skilled, blue-collar Australia-born workers are also feeling the pinch from structural change, especially those located in non-metropolitan areas. But, as Table 2 shows, relatively few are located in Sydney. We explore some of the reasons for this below.

Tables 3 and 5 provide information on the residential distribution of migrants by birthplace in 1991 and 1996, in this case for men aged 25-64. The reason for using data on males aged 25-64 rather than for all age groups is that this measure gives a better indicator of the social impact of the migrant presence. If all overseas-born persons were used, the figure would understate the ethnic presence because of the many young persons born in Australia with migrant parents.

Table 3 shows that there are high concentration of overseas-born males in a belt of suburbs several miles to the south-west of the Central Business District, beginning with Ashfield and Marrickville, then stretching to the west through Burwood, Strathfield and Auburn and, over recent years, into Parramatta. Another contiguous set of suburbs to the south, beginning with Canterbury and extending into Bankstown, and further out to Fairfield shows the same pattern. A few other suburbs on the periphery of this group, including Liverpool, appear to entering the high concentration category to judge from the rapid increase in their overseas-born populations (see Table 3). With the exception of Liverpool, Bankstown and Parramatta, substantial majorities of the male residents of these LGAs aged 25-64 were overseas-born. However, in the latter three cases, the overseas-born share rose sharply between 1991 and 1996 to reach just below 50 per cent.

When we turn to income distribution, there is a clear pattern showing that in all the suburbs cited, a high proportion of the overseas-born men aged 25-44 were low income recipients, in every case much higher than their Australia-born counterparts (see Table 4).

The overseas-born persons living in these south-western suburbs are drawn primarily from the NESB countries identified above as the poorest birth-place groups living in Sydney. By contrast, Table 5 shows that 8.2 per cent of all men aged 25-64 living in Sydney by 1996 were born in the UK and Ireland. But no more than 2.8 per cent of men in this age-group living in Fairfield, Canterbury and Auburn were born in the UK and Ireland. A similar, if not quite so pronounced, pattern is evident for men born in New Zealand and South Africa. This contrast is important because the extreme end of the inter-linkage between low income and high concentrations of overseas-born people occurs in these three LGAs. In each, around 70 per cent of all males aged 25-64 were overseas-born. The result is that the low-income population living in these LGAs is overwhelmingly overseas-born and from NESB countries. Though not to the same striking degree, the other suburbs identified show a similar pattern. Thus, there is now a distinct and contiguous belt of suburbs through Sydney’s south-west, one of whose dominant characteristics is that a high proportion of their residents are part of Sydney’s ethnic underclass.

The regional ‘fit’ of low income and high NESB concentrations is not perfect. Botany and Rockdale which are located to the south of Marrickville and Canterbury and border Botany Bay also feature high overseas-born populations (see Table 3). But, because their income levels are higher than their neighbours (partly reflecting a high representation of European migrants who arrived pre 1980), we have not included them in the south-western region identified above.

The next question concerns the ethnic mix of these low-income overseas-born concentrations. Do they reflect large congregations of particular NESB-birthplace communities, as with the Koreans living in Koreatown in Los Angeles or Chinese in New York’s Chinatown?

Table 5 shows that, at the LGA level, no one birthplace group dominates in Fairfield, Canterbury or Auburn. Postcode level data show sharper aggregations. But nevertheless it is evident that we are dealing more with a phenomenon of accumulated clusters of low-income NESB migrants than of particular ethnic groups. There are some partial exceptions. The Vietnamese constitute 17.9 per cent of the entire Fairfield adult male population aged 25-64. They make up 41.7 per cent of all the Vietnamese men in this age bracket living in Sydney. In the case of those from Cambodia and Laos, though just 7.1 per cent of Fairfield’s 25-64 year-old male population were born in these two countries, these men represented 67.2 per cent of all the same age group of Cambodia- and Laos-born men living in Sydney in 1996. Also, the 10.2 per cent of Canterbury’s 25-64 year old male population born in the Lebanon made up 17.2 per cent of Lebanese-born same aged men living in Sydney, and the 7.3 per cent of Auburn’s male population which is Turkish-born constituted 23.5 per cent of same group of men living in Sydney.

If the residential pattern identified for 1996 was temporary, there might be less concern about the implications. However, when data for 1991 are compared with those for 1996, it is clear that the trend is towards higher overseas-born concentrations in Sydney’s south-western suburbs. Two, perhaps mutually reinforcing, demographic processes help explain what is happening. The first is that recent NESB arrivals are locating in the same areas as family, friends and community, thus adding to the numbers of overseas-born persons. The second is that there appears to be a net-out-movement of Australia-born residents. Between 1991 and 1996 the total male adult populations aged 25-64 resident in Fairfield, Canterbury, Auburn, Bankstown, Parramatta and Strathfield all increased. Nevertheless, as Table 3 shows, the numbers of Australia-born male residents declined.⁵ The result was that the Australia-born minority in Fairfield and Canterbury fell by three percentage points to 27 and 31 per cent respectively between 1991 and 1996, by six percentage points in Auburn from 34 per cent to just 27.4 per cent and three percentage points in Strathfield to 42.9 per cent. Though Australia-born males aged 25-64 were not yet a minority in Bankstown and Parramatta, the combination of declining Australia-born and increasing overseas-born resident numbers meant that the overseas-born share increased sharply to almost reach 50 per cent. Only in Liverpool, one of the more rapid growth areas of Sydney, did both the Australia-born and overseas-born male population increase over the 1991-1996 period. However, the rate of growth of the overseas-born was far greater than that of the Australia-born, thus the sharp increase in the former’s population share from 43.5 per cent to 49 per cent by 1996.

There are some exceptions to this movement. In the case of Ashfield and Marrickville, which are the two suburbs within the south-western region nearest to the inner city and thus most affected by Sydney’s escalating inner-city housing price boom, the proportion of Australia-born residents is increasing. There were substantial falls in the number of overseas-born males aged 25-64 living in these two suburbs between 1991 and 1996. The implication is that there is a movement of higher income residents into these LGAs, which is prompting an out-movement of poorer overseas-born persons to more remote south-western suburbs, probably including Bankstown, Parramatta and Liverpool.

The Australia-born movers from the south-western suburbs appear to relocating in outer suburbia, notably in Penrith to the south-west and Gosford and Wyong to the north. As Table 3 shows, these growth areas are predominantly composed of Australia-born residents, with no increase in the overseas-born share over the 1991-96 period. The data base used for this study does not allow us to identify the Australia-born who are moving out of Sydney, but it is also likely that some former south-western suburban residents have joined the exodus to northern NSW and Queensland coastal areas.

We can only speculate about the factors shaping the Australia-born exodus from Sydney’s south-west. It may be that the concentration of low-income NESB families occurring in the south-western suburbs presents problems of accumulated disadvantage, straining the provision of government services in the area. The changing social make-up of the schools is a case in point. When local schools have to cope with large majorities of children from NESB families, as is the case in much of the south-west, the stress of contending with language and social diversity and the change induced in the ethos of the schools may prompt parents to move to areas where the schools do not confront this problem. The accumulation of disadvantaged young people of NESB origin, particularly where manifested in media reports about ethnic gangs and crime, also adds to pressures on the Australia-born to consider moving.

Needless to say, the question of ethnic community concentrations is politically charged. The concern is that any publicity about concentrations will fuel anti-immigration propaganda. Advocates of immigration and multiculturalism tend to deny that such concentrations are occurring at all or, if they acknowledge their existence, suggest that they are a temporary phenomenon, likely to dissipate in the manner of the Southern European inner-city concentrations of the 1950s and 1960s. One recent account of residential patterns concludes, after some critical observations of earlier work by Birrell and Healy, that ‘the situation is one of concentration of disadvantaged groups, rather than ethnic concentration’.⁶

However, the Census data are unequivocal. There is a growing concentration of low-income families in Sydney’s south-west, and the families concerned are primarily derived from NESB countries. By contrast, the low income Australia-born population is widely distributed throughout Sydney.⁷ Sydney may be Australia’s economic dynamo, but those with limited skills and/or English are not well placed to benefit. Nevertheless, because of the dynamics of the migration process discussed earlier, much of Sydney’s population growth derives from migrants in the disadvantaged category. These recent arrivers are mainly locating in existing areas of high ethnic concentration. Some analysts acknowledge the point, but argue that there is a parallel pattern of dispersal amongst residents who arrived earlier. Viviani, for example, accepts that recent arrivals are focussing in areas of existing NESB concentrations, but argues that ‘there is now strong qualitative evidence that Vietnamese have moved in the decade 1981 to 1991 in rapidly increasing proportions into middle- and upper-class areas’.⁸ In the long-term she may well be correct, given the good progress of many young Vietnam-born students through Australia’s education system. But currently the level of ‘dispersal’ falls well short of counteracting the pressures for continued clustering. Between 1986 and 1996 the proportion of all Vietnamese-born persons in Sydney who lived in Fairfield and Bankstown increased from 54 per cent to 56.2 per cent. Also, during the decade there was no net movement of earlier residents (arriving in Australia pre-1986) out of Fairfield.⁹

However, the Census data does not support the view that these residential patterns reflect a drive for ethnic community solidarity or exclusiveness. The high community diversity, especially in Canterbury and Auburn belies this idea. What then is driving the concentration patterns in question? Why are low income migrants concentrating and not their Australia-born counterparts? The is little doubt, that for recently arrived migrants, especially those lacking English and coming from culturally different settings, it is a great advantage to have access to language and culturally specific services. The Australia-born, by contrast, can settle anywhere in Australia and feel reasonably comfortable. They can be sure that a Protestant church, a bowling club, a pensioner association and so on will all be handy. But the dispersed residential pattern of better-off NESB migrants living in Sydney indicates that the attractions of culturally specific services is not the only factor shaping residential concentrations. Indeed, anecdotal evidence suggests that where ethnic concentrations are associated with social problems, particularly amongst youth, ethnic parents, just like their Australia-born neighbours, are often keen to move out. The problem is that many do not have the resources to take up the option.

With the partial exception of the Marrickville and Ashfield LGAs noted above, the south-western suburbs under discussion are (by Sydney’s standards) areas of low cost housing. Fairfield, Auburn and Canterbury each feature relatively large numbers of cheap rental properties, often in the form of three-storey walk-up flats near railway stations. According to the 1996 Census data, over 80 per cent of all properties being rented in these LGAs were being let for less than $200 per week. In Fairfield, which is one of cheapest areas in Sydney, the proportion was 86 per cent. For Sydney as a whole the share of all rental properties in this bracket was 65 per cent. The only areas in Sydney offering similar cheap accommodation are those in even more remote outer suburban locations, including Campbelltown, Wyong and Blacktown. Though affordable, their Australia-born domination means they offer little in the way of ethnic specific services. The same point could be made about the option of moving to low-cost coastal areas. Such a move makes far more sense for Australia-born low income families than their NESB counterparts.

Another factor conducive to residential clustering is employment opportunities deriving from the provision of services to co-ethnics. Our hypothesis is that the attraction of this form of employment is largely limited to NESB migrants who do not have the skills or qualifications necessary to find employment with mainstream employers.

Sydney’s Korea-born community offers a significant test case for this hypothesis. In 1996 there were only 30,129 Korean-born people in Australia but some 69 per cent, or 20,734, were located in Sydney. They are well-known to be an industrious and enterprising people. Yet Table 2 shows that they are amongst the poorest of Sydney’s NESB communities, with 35.8 per cent of the men aged 25-44 earning less than $15,600 per year. Their residential pattern approximates that described above. They are located primarily in the poorer south-western suburbs identified, with the main clustering in Canterbury (in the Campsie area). However, as with most of the other NESB groups in question, they are not highly concentrated in particular LGAs. Some 13.5 per cent of all Sydney’s Korea-born population are located in Canterbury. Other areas within Sydney’s south-western suburbs where clusters of Koreans are found include Parramatta, Burwood, Bankstown and Strathfield. However, unlike the other low income NESB communities discussed, there are also some clusters in Sydney’s more affluent northern suburbs, at Hornsby, Ryde and Baulkham Hills. Nevertheless, it is notable that the incomes of residents in these LGAs are much the same as those located in south-western Sydney. In each case a third or more Korean-born males aged 25-44 reported income below $15,600 and another third between $15,600 and $31,200.¹⁰

A high proportion of Sydney’s ‘pioneering’ Korean residents entered Australia first as visitors or students. Some came direct from Korea, but many others came from Vietnam, Latin America, the Middle East and elsewhere. They subsequently changed their status to permanent residence via the successive amnesties offered by the Australian government in 1976 and 1980. These ‘pioneers’ later brought in their families through the family reunion program. More recently many entered as English-language students and some as business migrants. Very few entered as Independents, and thus only a tiny number have been subject to preliminary English language testing or to assessment of their qualifications. Nevertheless, a significant minority brought considerable capital with them, which helps explain the significant minority who live in relatively high priced areas of Sydney.

Though the proportion holding professional qualifications is similar to that of the Australia-born population,¹¹ many of these professionals have not been able to gain positions commensurate with these qualifications, at least in mainstream government or corporate organisations. Their poor English is a serious problem, as is their difficulty in getting recognition for their qualifications and experience. Thus most Koreans have been effectively excluded from participating in the upper levels of Sydney’s business sector. This is not for want of trying. The second author of this paper has conducted a number of interviews with former managers and professionals. Most stated that after repeated failed applications they had to give up any aspiration to work in their former field. Thus the low incomes reported by Korean-born men aged 25-44 living in the relatively affluent suburbs noted above.

Most of these people, along with many of their less qualified counterparts, turned to establishing their own businesses. Though there is a strong cultural tradition supporting this decision, for many there was little alternative if they were to ‘succeed’ in Australia. The only other option was to take low-skilled industrial or service work. There is a reluctance to take up this option. The aspiration to ‘succeed’ seems to have prompted most to at least try to set up their own businesses.

According to Korean business directory sources, by 1997 there were some 1,600 Korea-run businesses in existence in Sydney. Given the low population base this is an extraordinary number. The largest number of businesses were in the food area, with 106 restaurants, 70 health food shops, 50 grocery stores and various other food speciality areas. Almost all provided Korean-style services. There were also a large number of education-related businesses, many serving as agencies for the thousands of Korean students (more than 7,000 in 1995-96) who, since the early 1990s, have flocked to Australia to study English. Another 99 acted as travel agents and a further 51 offered immigration counselling and related documentary services. Most of these businesses served a predominantly Korean co-ethnic market. One exception is cleaning services, where Koreans have found a niche outside the ethnic marketplace.

The co-ethnic service orientation means that Korean businesses have to be located close to areas of residential concentration. There are also social pressures favouring residential location close to other Korean residents. The Sydney Korean community has been very active in creating its own social organisations, the most striking being the establishment of churches with Korean-only congregations. There are currently some 106 Korean Protestant churches in Sydney, many on a small scale, as well as a few Buddhist temples.

A major outcome from this proliferation of small business within a relatively small ethnic community is a highly competitive business environment, with low profit margins. The reliance on Korean customers also means that the community is vulnerable to any downturn, such as is currently the case with the recent sharp drop in the number of English language students and tourists from Korea. This market situation helps explain why such an industrious community nevertheless features as one of the ethnic communities with the lowest incomes in Sydney.

The Korean situation differs from some of the other low-income NESB communities, in that only a small proportion of residents are engaged in manufacturing activities, such as in clothing based on outwork. Many Koreans were engaged in such work in the 1980s but these businesses collapsed with the import of cheaper Chinese clothing to Australia since the early 1980s. Much has been written about the Vietnamese involvement in manufacturing, notably in clothing and other labour-intensive goods. These activities tend to be located within areas of ethnic concentration too, because the work is often based in the home (as with out-work in the clothing industry) and is organised around clusters of small enterprises united by ethnic brokers or middlemen. However, as the Korean example shows, a high level of employment in intra-ethnic service provision can be maintained without the presence of out-work in manufacturing activities.

The Korean case is a representative of the larger ethnic community business scene. To the outside observer, the commercial areas where ethnic businesses concentrate appear to exhibit an air of bustle and enterprise, often favourably contrasted with the ‘run-down’ areas they replaced. In reference to Cabramatta, Jupp states that:

Our researches on the ground (in 1990) suggested that, far from being a deprived area, the commercial rents and purchase prices in Cabramatta were so high that many new arrivals cannot enter the booming commercial sector. Fairfield city council enjoys a strong rate base resulting from the development of John Street from a once sleazy and nondescript shopping strip to a major regional shopping centre.¹²

But this surface appearance shrouds a fiercely competitive environment. In the case of Cabramatta, drug trading, too, boomed during the 1990s. As a consequence some of the gloss has been taken off the legitimate commercial centre because of associated stereotyping of the area as dangerous and lawless. But apart from this problem, because it is relatively easy for competitors to initiate new enterprises where business is flourishing (there are, for example, some 100 hairdressers located in Cabramatta), returns tend to be low. Also the services can be provided elsewhere. Other regional Vietnamese shopping areas, as in nearby Bankstown appear to be cutting into Cabramatta’s status and profits.

Sydney is increasingly divided on ethnic, class and residential lines. Class divisions are an inevitable product of a competitive commercial world. Some overlap between class and residence is also inevitable, especially in a city like Sydney with such a sharply varied landscape and therefore intense competition for residential amenity. But the final overlap between residence, class and ethnicity is not inevitable. Rather it is a product, even if unintended, of Australia’s immigration program. According to our analysis, the increasing concentration of low-income NESB-origin residents in Sydney’s south-west is a consequence of a disjunction between the characteristics of the skills offered by the family, humanitarian and change-of-status streams of migrants locating in Sydney and the city’s changing economic base.

The concentrations identified are acute by any standard. In the case of Fairfield, Canterbury and Auburn, 70 per cent of all adult males residents aged 25-64 were overseas-born in 1996, in each case predominantly from low-income NESB backgrounds. Moreover, these concentrations are increasing. One response, common amongst scholars in the area, is to deny the situation. Alternatively, some romanticise the diversity and bustle of it all. This helps explain why the matter is not currently being given much attention. Thus the bland assurances from immigration advocates that the program can be expanded without risk.

There seems to be an analogy with the way the south-western zone of ethnic concentration is itself being pushed further to the periphery of Sydney. This process of social and geographical marginalisation may help to put the problem out of mind. However, it will certainly not solve it.