Flawed official narrative on indigenous population growth

The 2011 Census counted 548,370 Aboriginal and Torres Strait Islander people, a rise of some 20 per cent since 2006. Even though such a large rise would normally be considered demographically impossible (given that immigration is negligible for this group), the reported 20 per cent increase raised hardly an eyebrow. This was because five yearly changes of over 30 per cent had been reported by the Australian Bureau of Statistics (ABS) on three previous occasions since the 1970s, including a huge 42 per cent rise in the 1986 Census.

A rationalisation of Australia's high recorded Indigenous population growth has developed across officialdom and has received largely unquestioned acceptance.

A typical version is given by the Human Rights Commission: "The increases in the Indigenous population cannot be accounted for by the birth rate alone. The ABS attributes the increase to a growing propensity of people to identify as Aboriginal and/ or Torres Strait Islander, and the greater efforts made to record Indigenous status in the censuses".

Important matters that receive little official recognition include the effects of inter-marriage between Indigenous and other Australians, shortcomings in the Census question used to determine Indigenous status, and whether Indigenous identification rates can keep rising.

Birth rates have been a relatively minor contributor to Indigenous population growth, since the ABS has found that the total fertility rate for Indigenous women has more than halved since the 1960s. Recent ABS statistics suggest that the total fertility rate for Indigenous women bottomed at around 2.2 children in 2001 (down from about 6 in the 1960s) and has since risen to about 2.7 in 2011. The other two attributed causes (increased rates of identification and (some) improvements in Census procedures) are valid, though exaggerated in terms of their actual implications. They also disguise other important but generally unacknowledged influences.

Indigenous population growth has been characterised by peculiar trends, which only become evident when you look at growth in individual age cohorts over time.

By way of example, the 1971 Census counted 20,200 Indigenous persons aged 0 to 4 years of age. Forty years later we might expect something like 80 per cent of this cohort (aged 40 to 44 years during the 2011 Census) to still be alive. Instead the 2011 Census counted 33,605 Indigenous 40 to 44 year olds. Clearly well over 15,000 40 to 44 year olds, previously counted as non-Indigenous, had their ethnicity (as measured in the Census) changed to Indigenous over this 40 year period. Similar effects occurred across other age groups. While this is consistent with the accepted "increased rates of identification" explanation, it could also be the result of fundamental changes made to the Census question that identifies people as Indigenous.

The number of Indigenous in the 0 to 4 age group itself grew even more rapidly, from 20,200 in 1971 to 67,414 (0 to 4 year olds) in 2011. Since Indigenous fertility has more than halved since the 1960s, there must be another explanation.

Academic research as far back as the1990s recognised inter-marriage as a major contributor to Indigenous population growth. Even at that time, only 44 per cent of couple families with one or more Indigenous children aged 0 to 14 years were families where both parents were Indigenous. Census figures also showed that 83 per cent of the children of couple families with one Aboriginal and one non-Aboriginal parent were recorded as Indigenous in the 1991 Census.

Research based on the 1996 and 2006 Censuses showed that, by 2006, 52 per cent of Indigenous males were partnered with non-Indigenous females, 55 per cent of Indigenous females were partnered with non-Indigenous males, and rates of Indigenous identification among the children of such mixed unions were close to 90 per cent.

The obvious conclusion is that high rates of marriage by Indigenous Australians to non-Indigenous partners, along with the vast majority of the resultant children being classified as Indigenous, has contributed in a major way to rapid growth in the Indigenous population. Put another way, the fertility of non-Indigenous women partnered to Indigenous males is an often unrecognised contributor to Indigenous classified births, and is a more important contributor to increases in the Indigenous population than high fertility amongst Indigenous women.

Acknowledgement of such influences is largely absent from key government publications. An ABS Occasional Paper does grudgingly state that: "The largest increase in Indigenous census counts was for child age cohorts. A possible reason for these increases.....is that they are children of couples where one partner is Indigenous and one is non-Indigenous."

While the effects of ethnic inter-marriage on Indigenous population growth is under-acknowledged, issues of shortcomings in the Census question itself are almost completely ignored in the official narrative.

The key requirement of a Census or survey question on Indigenous status is to identify respondents who meet the officially set requirements for Aboriginality. The Census question thus needs to ascertain whether respondents (a) are of Indigenous descent and (b) also identify as Indigenous. It is generally recognised that household questionnaires are unsuitable for verifying a third official requirement, which is that respondents' Indigenous status is accepted by the community in which they live.

Let's now consider whether the Census question addresses these requirements for official recognition as Indigenous.

The original (post-1967 Referendum) Census question (asked in the 1971 and 1976 censuses) was:

What is this person's racial origin?
(If of mixed origin indicate the one to which he considers himself to belong)

(Tick one box or give one origin only)

European Origin
Aboriginal Origin
Torres Strait Islander Origin
Other Origin

The inclusion in the 1971 and 1976 censuses of the direction "if of mixed origin indicate the one to which he considers himself to belong" clearly was an attempt to measure racial/ethnic identification, though respondents were only allowed to identify with one origin. By ticking either the Aboriginal or Torres Strait Islander origin box, respondents were indicating a clear preference for Indigenous identity over any alternative origin, with "European" or "Other" being explicit stated alternatives.

The Census question was changed in many subsequent censuses. The current Indigenous Origin question used from 1996 onwards has now become:

Is the person of Aboriginal or Torres Strait Islander origin?
For persons of both Aboriginal and Torres Strait Islander origin mark both 'Yes' boxes
No
Yes, Aboriginal
Yes, Torres Strait Islander

What is striking is that respondents are now only asked whether they are of Aboriginal or Torres Strait Islander origin. The question does not ask about any other origin or whether, if they are of more than one origin, which one do they regard as their main identity. Respondents may now acknowledge more than one ethnic identity but the only such multiple identity allowed for is joint Aboriginal and Torres Strait Islander origin. Dual Aboriginal and European identity (potentially far more common) is not a tick box option.

Not asking explicitly about identity seems a serious omission, given that our Indigenous population (in vast majority) is now of mixed ethnic origins and subject to increasing influence from out-marriage . A literal interpretation of the Census question asked since 1996 would suggest that Indigenous origin by itself could potentially get a tick from everyone who can trace any Indigenous origin regardless of their self-identity or community acceptance.

A fleeting acknowledgement of this issue did get raised in a 2001 ABS Working Paper, which noted that "this (standard) question measures the descent concept, although some respondents will interpret the question to mean both descent and identification. It does not take account of the third part of the definition, community acceptance."

So what are the practical implications of the Census question not explicitly addressing identification? I can think of quite a few.

On a personal level it puts those, who can trace some Indigenous ancestry but may not identify as Indigenous, in a quandary as to how they should answer the Census Indigenous question. If they interpret the question in its literal sense they ought to tick an Indigenous origin box. They may, however, be reluctant to do so knowing that this will result in them being included in the Indigenous count.

To the extent that such persons do tick an Indigenous origin box, official estimates of the Indigenous population will be an over-count, though it is difficult to quantify its exact extent. The over-count is probably substantial and likely to disproportionately affect counts for non-remote areas. It needs to be emphasised that the issue stems from respondents not being asked to nominate their self-assessed identity rather than any problem with Census responses per se.

If the Indigenous component of our population is not accurately identified, it follows that all the Indigenous data based on such identification will be contaminated by the data for persons incorrectly classified as Indigenous. In general, this will lead to measures of Indigenous economic and social well-being that are too high, thus misleading policy-makers.

There are also broader issues rarely broached in the official narrative.

The idea that persons of mixed ethnic origins have a single "main" identity is debatable. There may be a clear "main" identity for some persons of Indigenous descent, especially if they have experienced discrimination because of Aboriginal appearance or have stronger cultural or kinship ties to their Indigenous heritage. Others may have difficulty having to choose one ethnic identity in preference to another, and for such reasons censuses in the US, Canada and New Zealand allow respondents to nominate more than one racial/ethnic origin.

Countries also differ in the detail of their definition of "Indigenous". New Zealand regards anyone of Maori descent (however distant) as Indigenous and does not have a formal requirement for either Maori identification or community acceptance. The Canadian Census, on the other hand, uses four alternative questions to identify Aboriginal peoples (Ethnic origin (including Aboriginal ancestries); Aboriginal Identity; Registered or Treaty Indian; and Member of an Indian Band or First Nation) resulting in substantially different population counts. In the US the Census Reference Book directs that to be "American Indian or Alaska Native"aperson must (1) have origins in the original peoples of North and South America and (2) maintain tribal affiliation or community attachment.

The bottom line is that the measured size of a country's Indigenous population is very sensitive to both how "Indigenous" is defined and measured. The issue with Australia's approach to quantifying its Indigenous population is not so much that the methodology is not right. The main issue is that the Census question is not consistent with our officially set definition. Secondly, the official explanation for high measured Indigenous population growth is incomplete and misleading.