New patterns of ethnic diversity: exploring the residential geographies of mixed-ethnicity individuals in Sydney, Australia

ABSTRACT

In multi-ethnic societies, the rise in mixed-ethnicity partnerships has contributed to strong growth in populations of mixed-ethnicity individuals. Yet scholarship on ethnic residential geographies has predominantly focused on individuals with singular ethnic identities. Using 2011 Australian census data, this paper explores the residential patterns of mixed-ethnicity populations in Sydney, Australia’s most populous city. I deploy a mapping analysis to show that mixed-ethnicity populations’ residential geographies are unique and do not match those of their constituent ethnic groups. In many cases, mixed-ethnicity individuals concentrate in inner-city areas, in contrast to the suburban hubs of their respective ethnic minority groups. They are also more likely to reside outside neighbourhoods with high proportions of their constituent ethnic groups, and instead gravitate towards moderately diverse neighbourhoods. The paper demonstrates the in-between nature of the geographies of mixed-ethnicity individuals, echoing established findings for mixed-race/ethnicity couples. Further, these geographies are powerfully differentiated according to birthplace and educational attainment. Australia-born mixed-ethnicity individuals and those with Bachelor degrees exhibit particularly extensive deviation from Sydney’s established ethnic landscape. The growing number of mixed-ethnicity individuals has implications for ethnic residential geographies both in this city, and in other diverse contexts.

KEYWORDS:

• Mixed race
• mixed ethnicity
• census
• diversity
• segregation

Acknowledgements

Many thanks go to Dr Natascha Klocker and the two anonymous reviewers for their helpful feedback on earlier versions of this paper.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 Other options include: American Indian or Alaska Native, Native Hawaiian, Guamanian or Chamorro, Samoan, Other Pacific Islander, Asian Indian, Chinese, Filipino, Japanese, Korean, Vietnamese, Other Asian and some other race.

2 From the early twentieth century until the 1960s, children of mixed Indigenous/European parentage were forcibly removed from their Indigenous families, with the intent of absorbing them into mainstream ‘white’ society (Katz 2012).

3 ‘Borderism’ refers to negative reactions towards those who traverse racial boundaries (Dalmage 2000).

4 This categorisation was constructed while cognisant that ‘Australian’ ancestry is conceptualised variously by census respondents. In the 2001 census, Khoo and Lucas (2004) found that second generation Australians were most likely to list ‘Australian’ in a dual-ancestry response. These are likely the children of mixed-ethnicity partnerships between migrants and persons of Australian ancestry. However, the authors conceded that some children of ethnic minority migrant parents, having been born in Australia, may list Australian as an indication of their nationality (i.e. an Australia-born individual with two Vietnamese migrant parents may list Vietnamese and Australian ancestries as part of a ‘transitional stage’ in their cultural affiliations, rather than an indication of ethnically mixed parentage). Nonetheless, Australian was included in the Anglo-European ancestry group, given the history of Anglo-Celtic predominance in the Australian population, and the instructions in the census guide to consider the origins of parents/grandparents in responding to the ancestry question.

5 This terminology is used for convenience and is not intended to suggest that an individual with multiple ancestries within the same regional ethnic grouping is not ethnically ‘mixed’ in the more general sense.

6 The scaled entropy index is calculated for each SA2 as$E=\text{s}\ast -\sum _1^n\left(K_i\right)\log \left(K_i\right).$ K is the proportional share of the SA2 population for each ethnic group (1 through n). The scaling constant (s) ensures that potential values range from zero (no diversity) to 1 (all groups present in equal proportions). For these calculations, the total population of each SA2 was divided into 10 ancestry groups largely based on the broadest level of the ASCCEG: Anglo-European, Pacific Islander, Southern and Eastern European, North African and Middle Eastern, South-East Asian, North-East Asian, Southern and Central Asian, Sub-Saharan African, Mixed and Other. Dual-ancestry persons with both responses in a single category are counted in that group, and those with ancestries in different groups are counted as Mixed. ‘Other’ includes all ancestries that do not fit within the other nine groups.

7 Following Wright, Ellis, and Holloway (2011), LQs are calculated for each set of mixed-ethnicity individuals as$\text{L}{\text{Q}}_j=\frac{\left(P_{ij}/P_j\right)}{\left(P_{im}/P_m\right)}.$ LQ_j is the location quotient for mixed-ethnicity individuals in SA2j. In the numerator, P_ij is the count of mixed-ethnicity individuals in area j and P_j the total population of area j. In the denominator, P_im is the count of mixed-ethnicity individuals in Greater Sydney, and P_m is the total population of Greater Sydney. All counts used in these calculations are restricted to individuals classified as likely residential decision-makers, defined in the methods section.

8 An SA2 was only included as a hotspot if its count for the group of interest was greater than 20. This excluded small counts, which are unreliable due to the effects of random cell adjustment in customised tables.

9 The variances of the dependent variables greatly exceeded their respective means.

Additional information

Funding

This research has been conducted with the support of the Australian Government Research Training Program Scholarship. It was also supported financially by a URC Small Grant (University of Wollongong) awarded to Natascha Klocker.