Poverty rates by electorate
The Parliamentary Library has just released a research note on poverty rates by electorate.
When I first saw the link I thought to myself, let’s do an ecological analysis of the 2004 election results. Alas, the data turned out to be based on the 2001 electorates.
I skipped the text and went straight to the map. As I looked at the map I thought to myself, I do not believe it. The remote electorates with large Indigenous populations had fewer people living below the poverty line than many regional electorates.
It was time to read the text.
Based on a definition of poverty that is described below, 9.3 per cent of all Australians are living in poverty. The poverty rate varies widely between electorates, from 2.1 per cent in Bradfield (NSW) to 15.1 per cent in Braddon (Tas).
The poverty rate estimates in this Research Note have been derived from experimental estimates of poverty by postal area, compiled by the National Centre for Social and Economic Modelling (NATSEM). While the estimates are the most accurate available at this stage, they should be regarded as indicative only.
Methodology and Definitions
NATSEM has produced estimates of poverty at a detailed regional level by combining postcode information from the 1996 Census with income and demographic information from the 1998-99 Household Expenditure Survey (HES). Income data from the HES were then adjusted from 1998-99 to 2001 values, allowing NATSEM to estimate poverty rates for all postal areas in Australia for the year 2001. These estimates at the postal area level were then aggregated to electoral divisions by the Parliamentary Library for the purpose of this Research Note.
As there is no universally accepted measure of poverty, NATSEM has adopted a relative poverty definition in which persons are defined to be in poverty if they are living below some community standard. Here the standard is set at half the average disposable income for a household type, consisting of a couple with two children, with incomes adjusted using an equivalence scale to determine comparable poverty lines for other types of households. If a household’s disposable income is below the poverty line then all the members of that household are deemed to be in poverty.
Mmmm. I think I see the problem.
It looks like the researcher relied on income and demographic information from the 1998-99 Household Expenditure Survey (HES). But the scope and coverage of the HES does not extend to all Australians.
SCOPE
24. Only residents of private dwellings in Australia were in scope. Private dwellings were houses, flats, home units, caravans, garages, tents and other structures that were used as places of residence at the time of interview. These were distinct from special dwellings which included hotels, boarding houses and institutions. Residents of special dwellings were excluded because of differences in their lifestyle and accommodation. Also excluded were households containing foreign defence force staff, foreign diplomats or diplomatic staff.
COVERAGE
25. Information was collected from usual residents of private dwellings in all areas of Australia except remote and sparsely settled areas, where:
- usual residents were residents who regarded the dwelling as their own or main home. Others present were considered to be visitors and were not asked to participate in the survey.
- remote and sparsely settled areas were areas in which there were less than 0.06 dwellings per square kilometre.
The HES excluded remote locations and it excluded the really poor who typically do not live in private dwellings. At a guess, a large slice would have been excised from the electorates of Kalgoorlie, Lingiari, Leichhardt, and Kennedy. All of which performed better according to the research note than I would have predicted.
I also wondered whether the treatment of multi-family households affected the analysis. There are a significant proportion of multi-family households in remote locations.
My conclusion: while the analysis in this latest research note may be robust for the urban and regional electorates, I have my doubts about the remote electorates.
Also, I think I will pass on the ecological analysis. (After being picky about someone else’s work I don’t want to be accused of falling for the ecological fallacy).