What's behind Australia's exploding indigenous population?

While the mentioned factors are important, there are two large elephants in the room. 

The first (hugely taboo subject in Indigenous affairs) is that the majority of Aboriginal and Torres Strait Islander adults now have a non-Indigenous partner.  Non-Indigenous as well as Indigenous mothers therefore contribute to Indigenous population growth, with over 90 per cent of the children of mixed Indigenous/non-Indigenous unions being classified as Indigenous.  This alone (if it continues) can result in continuing big Indigenous population rises (a matter that occasionally receives grudging acknowledgement), though one must raise the question of whether generations with increasingly diluted Indigenous heritage can continue to legitimately claim Indigenous identity. 

The second, (affecting the Census count) is that there is a question mark surrounding the appropriateness of Australia's standard Census Indigenous Question, particularly in the light of the official definition of Aboriginality used in this country. 

While an Indigenous New Zealander is defined as "a person of the Maori race of New Zealand; and includes any descendant of such a person”, Australia uses a much narrower definition.  We require Indigenous persons to not only have an Indigenous ancestor.  We also require them to identify as Indigenous.  (A third requirement of community recognition is not easily ascertainable in a household questionnaire.)  The problem, however, is that our  standard Census Indigenous Question asks only about Indigenous origin.

The Standard Question for Indigenous status used in the 1996 and later Censuses is as follows:

 There are several things wrong with this question.  Firstly the question asks about Indigenous origin in isolation from other possible origins (e.g. European, Asian).  Secondly, while the question permits respondents to state both Aboriginal and Torres Strait Islander origins, there is no scope, for example, for a person to indicate that they are of both European and Aboriginal origin.  The biggest problem of all, however, is that the question does not address the issue of identification.  How therefore can the Indigenous Census question filter out those of Indigenous descent who either don't identify as Indigenous or record those may identify both as Indigenous and European (for example)? 

I would argue that the current question facilitates a count of those who have Indigenous origins rather than (a narrower) count of those of Indigenous descent who also identify as Indigenous.  Consequently, our Indigenous count probably is considerably overstated.

Additionally, because substantial numbers of people over time have changed from not declaring Indigenous ancestry to later acknowledging their Aboriginal or Torres Strait origins, time series data for Indigenous people have been corrupted.  Apparent progress in the economic and social circumstances of Indigenous people over time will reflect both actual progress and the effects of recruitment of (generally more advantaged) persons previously counted as non-Indigenous.  Consequently, using time series data from the various Censuses to see if Australia is "closing the gap" will exaggerate the extent of progress.

Overall, it is obvious that official statistics present a far from accurate picture of our Indigenous population but nobody in authority is willing to concede this.