Current Issues Brief 13 2000-01

Socioeconomic Indexes for Electoral Divisions
(2000 Electoral Boundaries)

Gerard Newman and Andrew Kopras
Statistics Group
3 April 2001

Contents

Introduction
Highlights
Description of the Indexes
Data Limitations
Endnotes
Index of Relative Socioeconomic Advantage by Electoral Division
Electoral Division Ranked by the Index of Relative Socioeconomic Advantage
Index of Relative Socioeconomic Disadvantage by Electoral Division
Electoral Divisions Ranked by the Index of Relative Socioeconomic Disadvantage
Index of Economic Resources by Electoral Division
Electoral Divisions Ranked by the Index of Economic Resources
Index of Education and Occupation by Electoral Division
Electoral Divisions Ranked by Index of Education and Occupation
Appendix

Introduction

This paper provides data on four summary indexes, derived from the 1996 Census of Population and Housing, that measure different aspects of socioeconomic conditions for Commonwealth Electoral Divisions. It updates a paper published in 1998⁽¹⁾ to take account of the effects of the 1999-2000 electoral redistributions.

The five-yearly censuses of population conducted by the Australian Bureau of Statistics (ABS) collect data on a broad range of social and economic aspects of the Australian population. Each household is required to answer nearly fifty questions, covering such diverse topics as birthplace, occupation, educational qualifications, method of travel to work and ownership of dwellings. While it is possible to compare Electoral Divisions on the basis of each census variable it is often more useful to compare divisions on the basis of a summary of related variables. The indexes shown in this paper have been derived by the ABS for this purpose. A paper comparing Electoral Divisions on the basis of a number of individual census variables has been published previously by the Parliamentary Library.⁽²⁾

This paper shows each Electoral Division ranked on the basis of each index together with a description of each index and some notable features of the rankings. It has been produced from data obtained from the Australian Bureau of Statistics for Census Collection Districts (CCD) and aggregated to Commonwealth Electoral Divisions using the methodology recommended by the ABS (for details of the aggregation methodology see the Appendix pages 17-19). A more detailed explanation of the indexes is available from an Information Paper published by the ABS.⁽³⁾

Data for each index is presented in two tables. The first table shows each Electoral Division in alphabetical order and the second shows each Electoral Division ranked by the index (divisions are ranked from lowest to highest). The political party notionally holding the Electoral Division after the 1999-2000 redistributions is shown on each table.

In the Highlights section of the paper reference is made to the geographic classification of individual Electoral Divisions. The geographic classification used has been devised by the Australian Electoral Commission, which classifies Electoral Divisions into one of four categories: inner-metropolitan; outer-metropolitan; provincial or rural.

Highlights

The following section provides comment on the main features of each index. Because of the similar variables used to construct each index there are certain similarities in the rankings of Electoral Divisions for the various indexes. For instance, the division of Bradfield is ranked highest on each index while the divisions of Fowler and Bonython are the two lowest ranked divisions for three out of the four indexes. A number of divisions in the north shore area of Sydney and in the eastern suburbs of Melbourne appear in the top 20 rankings for each index. Similarly a number of rural and outer-metropolitan divisions appear in the bottom 20 rankings for all indexes.

Index of Relative Socioeconomic Advantage

This index measures the relative social and economic well-being of the area. The Electoral Division with the highest index score is the Sydney north shore division of Bradfield, while the division with the lowest score is the outer-suburban Sydney division of Fowler. The median divisions (index score where there are an equal number of divisions above and below that score) are the Queensland Sunshine Coast division of Fairfax and the Newcastle based division of Charlton

Most of the 20 highest ranked divisions are located on the north shore of Sydney (Bradfield, North Sydney, Berowra, Warringah, Mackellar and Bennelong) and in the Eastern suburbs of Melbourne (Kooyong, Menzies, Higgins and Goldstein). Other divisions to feature in the top 20 are the traditionally well-off areas of Brisbane (Ryan) and Perth (Curtin). No South Australian or Tasmanian divisions feature in the top 20. All but two of the top 20 divisions are held by the Liberal Party, the exceptions being the two ACT divisions (Canberra and Fraser).

Divisions with a low score are typically either outer-suburban or rural. Of the 20 divisions with the lowest score, eight are outer-suburban (Fowler, Chifley and Werriwa in Sydney, Holt in Melbourne, Oxley and Rankin in Brisbane and Bonython and Kingston in Adelaide) and six are rural (Gwydir in New South Wales, Maranoa and Wide Bay in Queensland, Grey and Barker in South Australia and the new division of Lingiari in the NT). Fourteen of the 20 lowest ranked divisions are held by the Labor Party while the six rural divisions are held by the Coalition parties (Maranoa, Gwydir and Wide Bay by the National Party, Grey and Barker by the Liberal Party and Lingiari by the Country Liberal Party).

Index of Relative Socioeconomic Disadvantage

This index measures the relative social and economic hardship in an area and to a certain extent displays a similar pattern to the Index of Relative Socioeconomic Advantage described above. It should be noted that to maintain consistency with the other indexes this index has been inverted so that the higher the index score the better-off an area is while the lower the index score the less well-off an area is. The Electoral Division with the highest score is Bradfield while the division with the lowest score is Fowler. The median divisions are Perth and the southern NSW division of Riverina.

As with the Index of Socioeconomic Advantage most of the 20 highest ranked divisions are located on the north shore of Sydney (Bradfield, North Sydney, Berowra, Warringah, Mackellar and Bennelong) or in the eastern suburbs of Melbourne (Kooyong, Higgins, Menzies, Goldstein and Aston). Two southern Sydney divisions (Hughes and Cook) also feature in the top 20 divisions. All but two of the top 20 divisions are held by the Liberal Party, the exceptions once again being the two Labor Party held divisions in the ACT (Canberra and Fraser).

Rural areas are prominent in the lower ranked divisions, reflecting the lower family incomes and lack of educational opportunities in rural Australia. Of the 20 lowest ranked divisions, nine are rural divisions (Cowper, Lyne and Gwydir in New South Wales, Wide Bay, Hinkler and Longman in Queensland, Grey in South Australia, Braddon in Tasmania and Lingiari in the NT). Two provincial divisions, Throsby based on Wollongong and Brand in Western Australia, also feature in the bottom 20. Because of the presence of a number of rural divisions the political representation in the bottom 20 divisions is more evenly spread than in the Index of Advantage. Twelve divisions are held by the Labor Party and eight by the Coalition parties.

An indication of the lack of homogeneity within a particular Electoral Division can be obtained by comparing the relative positions on the rankings between the two indexes of advantage and disadvantage. For example, if an Electoral Division has a concentration of people who are relatively well-off in one part of the division and a concentration of people who are not so well-off in another part this will be reflected in a relatively high ranking on the Index of Advantage and a relatively low ranking on the Index of Disadvantage. By this measure the two least homogenous divisions are the new Darwin based division of Solomon and the outer-suburban Melbourne division of Isaacs. Solomon is ranked in the bottom third (49^th) on the Index of Advantage yet is ranked in the top third (111^th) on the Index of Disadvantage. The division of Isaacs has a similar pattern, being ranked 41^st on the Index of Advantage and 99^th on the Index of Disadvantage. Other divisions to display a wide discrepancy are: Melbourne and Lalor in Victoria and Makin in South Australia.

Index of Economic Resources

This index measures the economic resources or income and expenditure patterns of families. Not surprisingly the distribution of Electoral Divisions in this index are very similar to the distribution in the Indexes of Advantage and Disadvantage. Once again the Electoral Division with the highest index score is Bradfield in Sydney, while the division with the lowest score is the new Northern Territory rural division of Lingiari. The median divisions are Rankin in southern Brisbane and Melbourne Ports.

As with the Indexes of Advantage and Disadvantage the top 20 places in the rankings for this index are dominated by divisions located in the well-off suburbs in the capital cities, although neither Adelaide nor Hobart divisions are included in the top 20. All but two of the top 20 divisions are held by the Liberal Party, the exceptions being Canberra and the northern Perth division of Cowan.

At the bottom end of the rankings rural divisions predominate, reflecting low family incomes and low rental and mortgage payments in rural Australia. The Index of Economic Resources is the only index where the Coalition parties hold more divisions in the bottom 20 than the Labor Party. The preponderance of rural divisions at the lower end of the rankings is further emphasised by the fact that the National Party holds no divisions with a ranking over 55.

Index of Education and Occupation

This index is designed to measure the educational attainment and occupational status of the population. Like the other indexes a familiar pattern emerges. The division with the highest ranking is Bradfield, while the division with the lowest ranking is Fowler. The median divisions are the northern Brisbane division of Petrie and Fremantle in Western Australia.

Once again the top 20 places in the rankings for this index are dominated by divisions located on the north shore of Sydney and in the eastern suburbs of Melbourne. However, there are a number of divisions appearing in the top 20 for this index that are not in the top 20 for the other indexes. These divisions are: Sydney, Melbourne, Brisbane and the inner- suburban Sydney division of Lowe. The high ranking for these divisions is probably a function of the location of tertiary education institutions in or near these divisions. Given the more diverse spread of divisions in the top 20 it is not surprising to find that the political complexion is also more evenly spread. Of the top 20 divisions, the Liberal Party holds thirteen, while the Labor Party holds seven.

Rural divisions make up the majority of the divisions at the bottom end of the rankings, reflecting the lack of tertiary education opportunities and professional occupational groups in rural Australia. Perhaps reflecting the gentrification of the inner suburbs of the capital cities, no inner-metropolitan divisions are in the bottom 20. Out of the 20 divisions at the bottom of the rankings, the Labor Party holds eleven divisions, while the National Party holds five and the Liberal Party four.

Description of the Indexes

There are four indexes shown in this paper. Each index summarises a different aspect of the socioeconomic conditions in the Electoral Divisions. The indexes have been obtained by summarising the information from a variety of social and economic variables. While there are similarities in the rankings of the Electoral Divisions, each index uses a different set of underlying variables.

The four indexes are:

• Index of Relative Socioeconomic Advantage
• Index of Relative Socioeconomic Disadvantage
• Index of Economic Resources
• Index of Education and Occupation.

All the indexes (including the Index of Relative Disadvantage) have been constructed so that relatively advantaged areas have high index values. For the Index of Relative Disadvantage this means that relatively disadvantaged areas have low index numbers. To enable easy recognition of high and low scores, the index scores have been standardised to have a mean of 1000 across all Collection Districts in Australia.

The Index of Relative Socioeconomic Advantage includes variables that measure relative social and economic well-being. Indicators included are: high income families; professional occupations; tertiary educational qualifications; dwellings owner occupied or being purchased; dwellings with a large number of bedrooms and a large number of motor vehicles. A higher score on this index means that the Electoral Division has a relatively large proportion of people with the above attributes (i.e. high incomes, professional occupations, tertiary qualifications, etc.). Conversely, a lower score on this index means that the Electoral Division has a relatively low proportion of people with these characteristics.

The Index of Relative Socioeconomic Advantage was constructed by combining two indexes-the Urban Index of Relative Socioeconomic Advantage and the Rural Index of Relative Socioeconomic Advantage. Because the index is a combination of two other indexes with slightly different components it should be treated with some caution. See the Appendix to this paper for a further explanation of the construction of the index.

As the name implies the Index of Relative Socioeconomic Disadvantage is derived from attributes that indicate relative social and economic hardship. Variables included are low income, low educational attainment, unskilled occupations, unemployment, one-parent families, renting households and Aboriginal and Torres Strait Islanders. To maintain consistency with the other indexes, the Index of Socioeconomic Disadvantage has been inverted so that advantaged Electoral Divisions have a high index score and disadvantaged divisions have a low score. Thus a high score on this index means that the Electoral Division has relatively fewer people with the above attributes, while a low score indicates relatively more people with these attributes.

In most cases Electoral Divisions that have a high score on the Index of Relative Socioeconomic Advantage will also have a high score on the Index of Relative Socioeconomic Disadvantage. However, it is possible for an Electoral Division to contain two distinct population groups, one relatively well-off and one relatively less well-off, thus giving a high score on the Index of Advantage and a low score on the Index of Disadvantage. The Division of Melbourne for example is ranked 108^th (i.e. relatively high) on the Index of Relative Socioeconomic Advantage, yet is ranked 53^rd (i.e. relatively low) on the Index of Relative Socioeconomic Disadvantage.

The Index of Economic Resources reflects the profile of the economic resources of families in Electoral Divisions. The indicators summarised in this index reflect the income and expenditure of families. Variables included are: home owner or purchaser households; high family incomes; high mortgage repayments and high rental payments. Variables covering non-income assets such as dwelling size and number of motor vehicles are also included. A high score on this index means that the Electoral Division has a higher proportion of families with these characteristics, while a low score indicates a lower proportion of families with these characteristics.

The Index of Education and Occupation is designed to reflect the educational and occupation of the population in Electoral Divisions. Education variables included in the index are the level of educational qualification attained or whether further study is being undertaken. Occupational variables include the major occupation groups and the unemployed. An Electoral Division with a high score on this index would have a high concentration of people with higher educational qualifications or undertaking further study and persons employed in higher skilled occupations. A low score indicates a concentration of people with low education attainment, low occupation skills or unemployed persons.

A list of the variables included in each index is shown in the Appendix to this paper.

Data Limitations

The indexes contained in this paper are subject to a number of limitations which should be borne in mind when the indexes are used to compare Electoral Divisions.

Firstly, the indexes include only some of the social and economic variables for which data was collected in the 1996 Census of Population and Housing. There are a number of social and economic indicators, such as wealth, savings, health, access to infrastructure, etc. that affect the well-being of the population but were not collected in the census and are therefore not included in the indexes. In addition, the indexes include only a limited number of the available census variables. Some notable exclusions from the indexes are age, country of birth, religion and hours of work.

Secondly, the indexes which have been produced depend upon the variables that have been included and the relative weights attached to those variables. The inclusion of a different range of variables or a different weighting pattern would result in a different index score. The indexes included in this paper are only four of the many indexes that could have been produced using census data.

In addition to the above conceptual limitations in the indexes there are a number of technical limitations. Non-response to individual census items may affect the accuracy of the indexes if there is a socioeconomic bias to non-response. Persons in non-private dwellings (e.g. boarding houses, etc.) are under-represented in the indexes as the variables pertaining to families and dwellings include only occupied private dwellings. The census results are based on place of enumeration rather than place of usual residence, thus holiday resort areas (e.g. Gold Coast) may be affected.

Other aspects of the indexes that should be borne in mind when comparing the indexes for different Electoral Divisions are that:

• The indexes are 'ordinal measures' and not 'interval measures', i.e. the indexes can be used to order Electoral Divisions into a ranked order but cannot be used to show that one Division is twice as well-off if its index score is twice that of an other Division.
• The indexes reflect the socioeconomic well-being of an Electoral Division rather than of individuals. Because all people in an Electoral Division are not identical the index scores do not apply to individuals but rather the way people are summed together for the area.
• The degree of heterogeneity within a Collection District influences the index score of that Collection District and hence the Electoral Division; the more homogeneous Collection Districts tend towards the extreme index scores.
• Partly because of the above, the interpretation of the index is more straightforward for Electoral Divisions that have extreme values. For example, it is usually easy to see why an Electoral Division with a high index score has that status, but it is more difficult to draw comparisons between Electoral Divisions with mid-ranked scores.

Endnotes

1. G. Newman and A. Kopras, 'Socio-Economic Indexes for Electoral Divisions', Current Issues Brief no. 4 1998-99, Department of the Parliamentary Library, 1998.

2. Kopras, 'Electorate Rankings: Census 1996 (2000 Electoral Boundaries)', Research Paper no. 11 2000-01, Department of the Parliamentary Library, 2000.

3. Australian Bureau of Statistics, 1996 Census of Population and Housing: Socio-Economic Indexes for Areas, Information Paper, Cat.No. 2039.0 ABS Canberra, 1998.