Bridging the gap between indigenous Australia and the broader community - a methodological note

I still do not want to get drawn back into the policy debate about the condition of Australia's indigenous people. I remain of the view that much of the discussion is misdirected. However, I do want to make two simple methodological points.

Earlier this year, Prime Minister Rudd pledged to "close the gap" between Aboriginal and non-Aboriginal Australians on a range of socio-economic indicators. Now an ANU team has concluded that this could take more than 100 years to achieve based on progress to date as indicated by census data. I have not been able to find the ANU report, so am relying on newspaper reports.

The first point is one apparently made by the ANU team, the danger of articulating sweeping aspirations. Here the researchers:

warned the Government that "a degree of policy realism and caution is required" and that its use of statistical benchmarks might lead to perceptions of policy failure in coming years.

"Under such circumstances we would counsel that commitments to 'reducing disparities' might be a more realistic policy goal for the Rudd Government than 'closing the gaps'."

I think that this is wise advice as former PM Hawke found when he promised to abolish child-hood poverty.

The second, linked point, lies in the dangers of using statistical averages. Take health as an example.

At national level, all the health indicators are national averages, concealing quite wide and arguably growing variations between groups within our community.

Our indigenous population is presently especially concentrated in regional, rural and and remote areas, along with some of the poorer suburbs in our capital cities.

As I have discussed before, general health services in parts of regional Australia have been in decline for some time. There are indications of a growing gap between health indicators for the whole population in these areas and those living in metro areas.

Within metro areas, the poorer suburbs have also been experiencing problems in health service delivery. Again, and for a variety of reasons, health indicators for the general population in these areas are below the national average.

So what does it actually mean to say that we will close the gap between indigenous health outcomes and the Australian average?

Are we implying that we can achieve better Aboriginal health outcomes than those applying to equivalent groups in the broader community? Alternatively, do we mean that we plan to reduce broader health disparities? Then again, are we proposing that the combination of broader social and economic advancement for our indigenous peoples in combination with physical re-location will get the required results? Or some combination of all the above?

My point is that we need to exercise great care in using statistical averages in setting aspirations and as performance benchmarks.

Postscript

David Nash kindly pointed me to the source document for the ANU paper on which the story is based - CAEPR Discussion Paper 287/2008 How realistic are the prospects for 'closing the gaps' in socioeconomic outcomes for Indigenous Australians? by J.C. Altman, N. Biddle and B.H. Hunter.

This is not the first time that David, an Honorary Visiting Fellow, Linguistics, Arts atANU and Visiting Fellow, AIATSIS, has pointed me to material.

The paper is really worth reading for a number of reasons.

It points to the difficulties involved in interpreting statistics in general, indigenous statistics in particular over multiple census periods. It suggests that, contrary to much popular reporting and public perception, the indicators show improvement in the indigenous condition relative to to the broader population since 1971.

Importantly, I think, in discussing the statistics the paper points to some of the variables that affect the indigenous position.

The paper does not affect the methodological points I was making. However, it did leave me thinking that in responding I was caught in a trap that I have complained about before, the way our attitudes and interpretation are affected by current reporting.

This is a very real issue just at the moment, and not just with the Aborigines, because of the way our collective thinking is affected by the constant flood of negative stories.

I am not saying that there are not bad things around, nor am I saying that the news must be good news week. I am saying that the perceptions that are created and that then drive policy and politics may be at considerable variance to the actual facts.

I was thinking about this because of discussions between Ramana and myself, most recently on Agriculture, the environment and Australia's future.

I think it fair to say that I am far less negative than Ramana about the future, although I am also very critical about some current approaches. Part of the reasons for this lies in the way I find a constant variance between short term reporting and commentary including sometimes my own and the reality when I come to check the facts.

Postscript 2

The Acadamy of Social Sciences in Australia has released a paper that bears upon the discussion in this post - Maggie Walter, Lives of Diversity: Indigenous Australia, Occasional Paper 4/2008 census series #2, Academy of Social Sciences in Australia, Canberra, 2008.

Using census data, this paper compares the indigenous position in three different communities - Maningrida (Northern Territory remote), Dubbo (NSW regional) and Perth (WA metro) - with each and with the non-indigenous populations in those communities. In this sense, it addresses one of the statistical concerns that I raised above.

I think that the results can be fairly summarised this way: while there are differences between the three, the indigenous populations are in some ways equally disadvantaged compared to the non-indigenous populations in the same locality. This is a relative, not an absolute measure.

This does not affect the broader point because the paper also draws out differences between the general populations especially in Perth and Dubbo.

There are a number of interesting issues raised by the paper that I will address at some point. For the moment, I just wanted to alert you to its existence.