With approximately 18 per cent of the Australian continent under some form of Indigenous tenure (Altman, Buchanan and Larson 2007), and with this set to rise via native title determinations and land purchases, the demand for statistical information on Indigenous groups as proprietors of territory is also growing in the context of formal agreement-making. As with the broader government agendas of closing the gaps, applied demography has found a natural and successful disciplinary niche here by exploiting the rich seams of census, survey and administrative data that make up the burgeoning Indigenous statistical archive, even at local and regional levels. In particular, it is the predictive capacity of the demographic repertoire that has mostly caught the eye of regional stakeholders (Harvey and Brereton 2005) with its fiscal opportunity-cost message that business as usual is simply not an option in Indigenous affairs because of the weight of population momentum. However, there are two important constraints on the efficacy of this contribution that need attention if the interests of Indigenous stakeholders are to be truly represented and if the scope of applied demography is to be extended.

The first of these relates to geography and concerns the long-standing conundrum of determining which area/peoples are affected/implicated by mining. In regard to this, it remains the case that analysts are constrained by the configuration of official ABS and other administrative boundaries that are used for the collection and dissemination of official statistical data. While it is true that the Australian Indigenous Geographic Classification of the ABS attempts to best reflect the spatial distribution of the Indigenous population—to enable a ‘demography in situ’ as it were (to use Kreager’s (1982) term describing demography built from the ground upwards)—in some ways this misses the point entirely. The primary organising principles of Indigenous social formation are both spatial and socio-relational (Morphy 2007) and these invariably do not coincide to produce discretely bounded social groupings that neatly mesh with units of the Australian Indigenous Geographic Classification. In a nutshell, the statistical geography available to analysts is unlikely to provide a demography of Indigenous polities with rights and interests in particular places, or agreements.

In many ways this highlights an important distinction raised by Rowse (2008) between Indigenous populations and Indigenous peoples. Our current demography refers to Indigenous populations revealed by standard identification questions and is best suited to the provision of citizen rights. What it does not necessarily provide for are the interests of groups of Indigenous people in proprietory rights, in particular over areas of land. All across Australia we are witnessing a growing discrepancy between the best-intentioned of statistical output frameworks and the actual needs of Indigenous land-holding groups for an ethnographically-informed demography suited to their aspirations for managing the Indigenous estate via land use agreements.

The second constraint refers to the nature of available data themselves. While information on vital demographic events are constrained only by the degree to which this is gathered in respect of any population, the greater concern is to do with information on population characteristics as well as broader epistemological questions surrounding notions of well-being and what these imply about appropriate variables to be measuring. As things stand, the social, economic, and even some cultural, features of Indigenous populations are invariably established via mainstream categories. Whether these coincide with Indigenous categories in any given situation, and whether such categories can be identified, measured and are desired, remains a moot point.