1504.0 - Methodological News, Dec 2013  
Previous ISSUE Released at 11:30 AM (CANBERRA TIME) 11/12/2013   
   Page tools: Print Print Page Print all pages in this productPrint All RSS Feed RSS Bookmark and Share Search this Product

Constructing an experimental household level socio-economic index

The ABS produces Socio-Economic Indexes for Areas (SEIFA) which seek to summarise the socio-economic conditions of an area using relevant data collected from the Census. The indexes provide information about the area in which a person lives, but within any area there are likely to be smaller groups with characteristics different to the overall population of that area.

Constructing socio-economic summary measures for finer units such as households would enable researchers and policy makers in Australia to better differentiate between areas with varying concentrations of advantage and disadvantage.

Research was conducted into the construction and dissemination of an experimental household level index. The results were presented to the Methodology Advisory Committee (MAC) in June 2013. Mesh Block, household, family and individual levels were all considered as units of output for the experimental index.

A household level index was chosen as it complements the area level rankings by adding more depth to the information given by SEIFA, as well as providing its own valuable insights by allowing users to make more accurate inferences about smaller units. Households are a clearly definable fine-level unit in which individuals tends to share socio-economic conditions. Furthermore, a household level index maximises the population included in the index given the Census data available.

The summary measure was derived from binary indicator variables constructed from Census data which measured household advantage and disadvantage. Principal Component Analysis (PCA) was used to specify weights for the variables. Occupied private dwellings were the base unit for analysis. Households which had more than 10 non-responses for the relevant Census input data items were excluded, which accounted for 2% of the population. Remaining records with missing responses were coded to the neutral response (0) rather than the advantaging or disadvantaging characteristic (1). These decisions were deemed appropriate based on conceptual validity, a literature review and user familiarity with previous SEIFA research and products.

Avenues identified for disseminating an experimental household level index include using Census TableBuilder to enable cross-classifications with other Census variables and including counts of households with current SEIFA output to pinpoint diversity within areas.

There was interest from the MAC members to see the ABS continue pursuing the release of an experimental household level index product. There is great demand externally, especially from service providers who currently use SEIFA to plan and monitor their services and would benefit from a measure at a finer level. Currently the SEIFA team is investigating producing a finer-level index at the household level in response to MAC feedback and user demands.

The paper presented to MAC is available here Research Paper: Building on SEIFA: Finer Levels of Socio-Economic Summary Measures (Methodology Advisory Committee) (ABS cat. no. 1352.0.55.135).

Further Information
For more information, please contact Phillip Wise (02 6252 7221, phillip.wise@abs.gov.au) or Courtney Williamson (07 3222 6031, courtney.williamson@abs.gov.au)