In the US during the 1980s, income inequality increased sharply and many low-skilled workers, disadvantaged by the high-interest-rate policy of the Reagan administration and the economic restructuring which ensued, found themselves on wages below the official US poverty line. There was an accompanying increase in the spatial concentration of low-income people in ‘underclass areas’, areas which were largely cut off from the mainstream economy and characterised by multiple forms of disadvantage,¹ including high levels of welfare dependency, female-headed households, and poorly-educated young persons neither in school nor in employment.²

Since the late 1980s, there has been growing awareness of the emergence of a low-income ‘underclass’ in Australia. Yet, historically, many Australians have perceived their society to be egalitarian, a perception which dates to the Federation era. Although the extent to which this egalitarianism was achieved in practice has been questioned, there were sound reasons for the perception. By the early post-Federation period, the implementation of economic protection provided the principal basis for nation-building and for a tacit class compromise between nascent business interests and organised labour. This was because protection promised to provide both profits and ‘fair wages’.

Despite the social and economic disruption of the 1930s depression, by the mid-1970s, the economic progress of the post-war period revived the perception of Australian egalitarianism. Writing in 1995, Gregory and Hunter argued that in 1976 the gap in household incomes between the lowest one per cent of Census Collection Districts (CDs) and the median CD could be, on average, bridged by an additional part-time job of nine hours per week at $12 per hour.³

But the mid-1970s appear to have been a turning point. From this time, the historic compromise between capital and labour began to break down. Gregory and Hunter’s analysis of the growth in household income inequality between 1976 and 1991 led them to speak of ‘two Australias’. Their research indicated that there was one Australia of the work and income rich, and another Australia which was becoming economically marginalised and spatially isolated. They argued that, between 1976 and 1991, the gap in household income between the lowest and highest five per cent of CDs or neighbourhoods (ranked by socio-economic-status) almost doubled.⁴

The authors further argued that between 1976 and 1991 the basis of the growing differentiation in household income across Australian neighbourhoods shifted. From 1976 to 1981, falling incomes in low socio-economic-status neighbourhoods accounted for the growing gap, whereas, between 1981 and 1991, this trend was accompanied by income increases in high socio-economic-status neighbourhoods.⁵ Similarly in the US, during the 1980s, widening income differentials reflected both a decline in the incomes of low-income earners and a marked increase in those of many high-income earners.⁶

Gregory and Hunter’s study could not identify the sources of household income and thus could not assess the impact of the redistributive effect of government non-cash-transfer social-wage expenditure, such as in the areas of health and education. Work by Raskall and Urquhart suggests, however, that although social-wage expenditure had a mitigating influence upon widening income differentials during the 1980s, particularly in the mid-1980s, it did not stem the general trend.⁷ Nor did Gregory and Hunter’s study examine the ethnic make up of CDs characterised by low incomes.

The 1980s saw the onset of industry and labour-market restructuring and bipartisan political support for economic internationalisation. Globalist ideologies were used by elements within political, business and even labour elites, to promote the idea that standards inherited from the protectionist past were no longer valid, and that continuous change was now the ‘norm’. Unemployment levels which would have been politically unthinkable a decade earlier became acceptable.

A central factor behind the growth in income inequality has been the significant decline in the proportion of males employed. In the 1966 to 1996 period, male employment declined across all age groups. In 1966, approximately 96 per cent of males in the 35-44 years age group were in paid employment. By 1996, this figure had fallen to 86 per cent. The situation was worse for younger and older men. For males between 60 and 64 years of age, the proportion in employment fell from 80 to 41 per cent.⁸ A recent Evatt Foundation study expressed the gravity of this development:

...the traditional Australian foundation for household economic security, has been a male full-time job, and any fall in the relative incidence of full-time employment provides an apparent warning of more general social upheaval.⁹

Between 1966 and 1993, industrial restructuring was accompanied by a decline in the proportion of the work force in manufacturing employment, but increases in the proportion in wholesale and retail, community services, and finance sectors.¹⁰ In 1966 manufacturing accounted for 28 per cent of Australia’s total employment. It currently accounts for only about 14 per cent.¹¹ Given the tendency for manufacturing industry to be concentrated in more or less well-defined geographic areas, the dramatic decline of manufacturing since the late 1960s as a source of semi-skilled and low-skilled employment probably affects the spatial concentration of low-income persons.

These developments also raise the question of the impact of immigration policy during this period of economic adjustment. The family-reunion and humanitarian components of the program, both of which expanded during the 1980s, brought into the Australian labour market large numbers of low-skilled persons with poor English. Where do these people fit into the spatial pattern of growing low-income concentration?

1996 CENSUS DATA

The release of 1996 Census data provides an opportunity to explore these issues. The data analysed below are based on male individual income by major birthplace by Statistical Local Area (SLA) for Melbourne and Sydney.

The focus is on males of 25-44 years of age. This serves the dual purpose of controlling for age and sex and allowing a more meaningful comparison between SLAs which vary in age structure and in the labour-force participation rates of men and women. The 25-44 age bracket is also significant in that it represents the prime of most males’ working lives. We chose the weekly income category of less than $300 to check those males in this low-income category because few would dispute that such men belong to a low-income ‘underclass’. Most men on less than $300 per week are on the margins of the mainstream labour market. They either receive labour-market benefits, or are working on a part-time, casual, or irregular basis outside the award full-time male rate. For Australia as a whole, a disheartening 20 per cent of men aged 25-44 fit this description.

Notwithstanding evidence of rural and non-metropolitan areas of low-income concentration, the focus here will be on Melbourne and Sydney.

Table 1 shows that in Melbourne in 1996 17.9 per cent of males (25-44 years of age) were receiving less than $300 per week, or less than $15,600 per year gross. Subsequently this category will be referred to as ‘low-income men’. In the remainder of Victoria, the proportion was significantly higher at 23.5 per cent. This implies that in Melbourne nearly one in five males were subject to labour-market marginalisation. We now turn to exploring the degree of spatial concentration of these men.

Within Melbourne, there is considerable variation across SLAs in the proportion of 25-44 year-old males with an income of less than $300 per week. There are seventy-four SLAs within Melbourne. Table 2 shows those with the highest proportions of males receiving less than $300 per week. In order to indicate the diversity of income patterns for men, the SLAs with the lowest concentrations are also given.

Many of the SLAs with a high proportion of low-income men are located within the Western and Northern working-class areas of Melbourne, areas which have been particularly hard hit by job losses due to manufacturing decline. The SLA, Greater Dandenong (C) - Balance, in the South-East of Melbourne, has been similarly affected. In Maribyrnong and Sunshine in Melbourne’s West, 29.2 and 27.7 per cent respectively of all males aged 25-44 were in the low-income category. This is well above the Melbourne average. In Brunswick, Preston and Broadmeadows in Melbourne’s North the proportion was 27.4, 26.4, and 25.2 per cent respectively. Such localised concentrations of labour-market disadvantage are consistent with Gregory and Hunter’s argument that economic restructuring is leading to a greater spatial distinction between the affluent and the poor.

Table 1 shows that the proportion of men in the low-income category was 16.7 per cent in Sydney. This is marginally lower than Melbourne’s 17.9 per cent. As with Melbourne, there is a considerable variation between Sydney SLAs above and below the city’s average. Table 3 shows the SLAs at the top and bottom of the range in Sydney. In the SLAs of Fairfield and Auburn, the low-income group make up more than 27 per cent.

Table 4 shows that, if men in the low-income category in Melbourne and Sydney are disaggregated according to country of birth, there are marked differences in the proportions on low incomes.

In both Melbourne and Sydney, there is a considerably smaller proportion of 25-44 year old Australian-born men in the low-income category than men from the combined category of all overseas birthplaces. In Melbourne the figure was 15.6 per cent for Australian-born males and 23.5 per cent for overseas-born males. For Sydney, the corresponding figures are 14.0 and 21.7 per cent respectively. But overseas-born males born in English-speaking countries are less likely to be in the low-income category than are Australian-born males. This means that the broad comparison between the overseas- and Australian-born masks the extent to which males born in some non-English-speaking countries are over-represented in the low-income group. The extreme cases in both Melbourne and Sydney tend to be the same: males born in Lebanon, Vietnam, the Republic of Korea, Turkey, Cambodia and Laos, China, and the former USSR and Baltic States.

A pattern of marked residential concentration for the Indochinese in Australia (particularly in Melbourne and Sydney) had become evident by the 1980s. Since then, their propensity for urban residential concentration has fuelled debate about the formation of ethnic enclaves or ‘ghettos’. Two contrasting interpretations have been put. The first is that these concentrations represent sites of relatively permanent underclass formation based on linguistic, cultural, and labour-market marginalisation. The second is that such residential concentrations are transitory phenomena which will dissipate with time and upward social mobility as these communities become more established. Amongst others, Bob Birrell¹² and Ernest Healy (the present author)¹³ have argued for the first position, and James Jupp¹⁴ has been an advocate of the second. Nancy Viviani¹⁵ has taken something of an intermediate position. She argues that areas of Vietnamese residential concentration in Sydney and Melbourne remain for the most disadvantaged Vietnamese, and that Vietnamese unemployment rates are likely to be high for a considerable period. But she also maintains that there is significant outward mobility from these enclaves for persons who have improved their position since arriving in Australia.¹⁶ Healy and Viviani argued their positions in People and Place, vol. 5, no. 3 (September 1997).

The present author showed that, in the 1991-1996 intercensal period, there was no decline in the level of residential concentration for the Vietnam-born in Sydney. The largest concentration, in the SLA of Fairfield, maintained its share of Vietnam-born in Sydney, at around 41 per cent. There had been some movement in some other areas of Vietnamese concentration in Sydney, but the data indicated that this was due to movement between areas of high concentration rather than to outward movement to areas of low concentration.¹⁷ At the time, the data base for Melbourne was less clear due to changes in local-government boundaries between 1991 and 1996 which affected the boundaries of SLAs. Since September 1997, comparable Melbourne data for the 1986, 1991, and the 1996 Census have been made available by the Australian Bureau of Statistics. These data reveal that Melbourne’s Vietnam-born population was just as concentrated in 1996 as it had been in 1986. This is evident from index-of-dissimilarity calculations for the Vietnam-born in Melbourne. The index of dissimilarity measures the difference in the spatial distribution of two co-existent populations. The index indicates what proportion of the sub-population would have to move in order to have the same distribution as the total population. The indices of dissimilarity for the Vietnam-born in relation to Melbourne’s total population in 1986, 1991, and 1996 were 51.7, 51.5, and 53.0 respectively.

Table 5 shows the indices of dissimilarity in Melbourne and Sydney for some of the birthplace groups examined above. The index of 10.7 for low-income Australian-born males in Melbourne means that 10.7 per cent of them would have to change SLA if they were to have the same spatial distribution as all low-income males. The proportion of Australian-born males in Sydney who would need to move is higher than in Melbourne, no doubt reflecting the especially steep housing-price and rental gradient between favoured inner-city and coastal areas and the more remote outer-suburban locations.

The birthplace groups identified in Table 4 as being disproportionately over-represented in the low-income category are also characterised by high levels of spatial concentration within Melbourne and Sydney. Those born in Cambodia and Laos have very high indices of dissimilarity, at 57.7 and 70.0 for Melbourne and Sydney respectively. It is generally considered that an index of less than 20 represents little spatial difference between sub-group and total population. An index of over 30 can be read as a significant difference, and over 50 as a very significant difference.¹⁸ In these terms, low-income men born in Cambodia and Laos, and Vietnam are heavily concentrated, as are those born in Turkey (50.7 in Melbourne and 45 in Sydney), and in The Lebanon (47.4 and 49 in Melbourne and Sydney respectively).

Another aspect of this concentration pattern is that the birthplace groups in question often live in the same Local Government Areas (LGAs). Men (25-44) who were born in Vietnam, Lebanon, Turkey, China (excluding Taiwan), Cambodia and Laos, and the former USSR and Baltic States can be combined. When this is done, we find that forty per cent of this aggregated group resides within only five of Melbourne’s seventy-four SLAs. (See Table 6.)

Gregory and Hunter have argued that income inequality has increased in Australia since the late 1970s and that the spatial concentration of low-income households has become more pronounced. The findings of the 1996 census are consistent with this argument. In both Melbourne and Sydney, there is an underclass of men, 25-44 years, receiving less than $300 per week and this underclass is concentrated in specific geographic areas.

Higher concentrations become apparent when low-income men are disaggregated by country of birth. To the extent that those on low income are spatially concentrated as Gregory and Hunter claim, the concentrations are more associated with ethnic minorities than with the Australian-born. So far as Melbourne and Sydney are concerned, low-income Australian-born men are much more evenly distributed across SLAs than are low-income men from many overseas-born groups. Because the spatial concentration of disadvantage is a important aspect of underclass formation, this raises the issue of the contribution of immigration policy to the development of an Australian underclass.

The spatial concentration of low-income persons observed by Gregory and Hunter may largely be due to the settlement patterns of predominantly low-skilled migrant groups, particularly those from non-English-speaking backgrounds, who came to Australia when low-skilled workers were being increasingly marginalised within the labour market. Australian immigration policies of the 1980s and 1990s introduced large numbers of people with precisely those characteristics which were most likely to lead to marginalisation in the work force. Gregory and Hunter also argue that future employment growth is likely to be biased in favour of more affluent, low-unemployment urban locations and that this growth may not spill over into depressed neighbourhoods. If this argument turns out to be correct, then the ethnic dimension of Australia’s growing urban inequality may become entrenched.¹⁹