The IES, as currently practiced in northern and central Australia, relies on two major adaptations in how the count proceeds: an interview process, rather than household self-completion of a form, and a modified household plus personal form structure. While the interview process seems to us to be indispensable and working well as an adaptation, we do not believe that the two-form structure is working well at all. The two- or three-form structure—if we count the community-level Dwelling Check List as well—is administratively cumbersome and far too demanding on both Aboriginal interviewers and interviewees. The structure only worked in the Aurukun and outstation case studies which we observed because the SIHFs were largely filled out beforehand away from interviewees. But even in these cases, filling out just the SIPFs with interviewees was a time-consuming and onerous task which tested the endurance and interest of both the interviewers and the interviewees to the limit. In the Alice Springs case study, where an attempt was made to fill out both the SIHFs and SIPFs in the presence of interviewees, this proved far too demanding and cumbersome and had to be abandoned.

We believe that for Aboriginal communities in northern and central Australia it should be possible to design a single household form that can be administered by interview. This form should build on the existing SIHF, with its provision for large numbers of people in a household, and it should extend the numbers of questions asked of individuals on that form beyond the basic age, sex, usual resident/visitor demographics. Such a form would be a move towards the more standard census household form structure, and an argument could even be made that the appropriate adaptation is to do interviews using standard household forms in some Indigenous communities. But for the more traditionally-oriented communities in northern and central Australia with low levels of western economic activity and in which English is often not the first language of spoken daily interaction, we believe that there is a need for a specific, restricted Indigenous household form.

If we recall, from Chapter 1, the way in which the IES started as an adaptation of the census to the circumstances of more traditionally-oriented Aboriginal communities in the Northern Territory and Western Australia in 1976 and 1981, before spreading to South Australia, Queensland and, in 2001, also into New South Wales, we can perhaps see where it has gone somewhat astray. Its ability to focus specifically on the circumstances of more traditionally-oriented Aboriginal communities in northern and central Australia has probably been somewhat weakened over the years, while it has probably also been- 98 - somewhat unnecessarily applied in places like New South Wales where many people living in discrete Aboriginal communities could probably deal with standard census forms and with household self-completion of forms.

Our solution would be to distinguish the interview from the 'special form' aspect of census adaptation. While both could be used in the most traditionally-oriented Aboriginal communities, like those in our case studies, interviews with standard forms could be used in other Aboriginal communities. This would make it possible to design the new Indigenous household form specifically for the circumstances of the most traditionally-oriented Aboriginal communities in northern and central Australia. Some Indigenous communities might also have mixtures of households who used the standard and special household forms with and without interview. This leads us on the issue of what questions should be asked in this revised Indigenous household form.