Getting a handle on individual diversity within areas
Analytical Services Branch (ASB) has recently completed and published a study on individual diversity within areas, as part of its ongoing research on Socio-economic indexes for areas (SEIFA).
SEIFA seek to summarise the socio-economic conditions of an area using relevant information from the Census of Population and Housing. The SEIFA indexes are widely used measures of relative socio-economic advantage and disadvantage at the Census Collector District level. The indexes provide contextual information about the area in which a person lives, but within any area there are likely to be individuals with different characteristics to the overall population of that area. If inferences are made about these individuals based purely on the characteristics of the area in which they live, they could be misleading and there is potential for error in any conclusions – this is referred to as the ecological fallacy.
Using 2006 Australian Census of Population and Housing data, ASB explored individual level diversity within areas by creating and analysing two person-based socio-economic indexes: one of relative disadvantage and the other of relative advantage and disadvantage. The conceptual and methodological basis for these indexes was established by Baker and Adhikari (2007). The primary purpose of the study was to illustrate how individual level index scores can be used to illustrate and measure the diversity of socio-economic advantage and disadvantage within area level SEIFA. Secondary to this analysis of diversity, the research served to highlight the advantages of SEIFA when compared with individual level indexes of relative socio-economic advantage and disadvantage, including maximising the proportion of the population receiving an index score.
Results showed that care that needs to be taken when using SEIFA information to draw conclusions about individuals who reside in those areas. Analysing the two individual level indexes of socio-economic advantage and disadvantage constructed in the study has facilitated the appraisal of diversity within areas, something that has until now not been possible except in Baker and Adhikari (2007). The results were important illustrations of how diversity of advantage and disadvantage within an area can exist, and the extent to which individuals with differing levels of socio-economic advantage and disadvantage reside in the same area.
The study has made clear some of the shortcomings of individual level indexes, and why SEIFA remains an important, robust product. Firstly, concerns with substantial population exclusions limit the applicability of the analysis; approximately one-third of the population counted in the 2006 Census were excluded from the individual level index construction process for applicability, compared to 0.6% of the population excluded for 2006 SEIFA. This vast difference reflects the robustness of the SEIFA indexes, namely that it maximises the proportion of the population that receives an index score. SEIFA is also more theoretically and conceptually sound because it is based on variables chosen for applicability in an area-based index, it is externally validated, and the aggregate nature of the data and stringent exclusion rules both work to ensure that there is sufficient meaningful data in an area to support index construction.
Given the analysis presented in the research, users of SEIFA will understandably be wondering if a product can be released that enables them to appropriately tackle the issues involved. Before attempting this, the following critical issues would have to be resolved: consensus on the definition of individual level advantage and disadvantage, the best set of variables to measure it, a means for validating individual level indexes and a strategy for integrating an individual level product with the existing SEIFA indexes to ensure that it is a useful addition. ASB is recommending that the analysis in the research paper be repeated after the release of SEIFA 2011 because of the introduction of the Australian Statistical Geography Standard (ASGS). There is an expectation that the ASGS will better capture the socio-economic gradient within areas. In the meantime, ASB recommends using the SEIFA indexes for socio-economic analysis, bearing in mind the caveats relating to those measures not being attributable to individuals, but only to the average relative socio-economic advantage and disadvantage in an area.
The published paper can be found on the ABS website, catalogue number 1351.0.55.036.
For more information, contact Phillip Wise on (02) 6252 7221 or Rosalynn Mathews (02) 6252 5257.