In order to meet the new requirement for statistical data on Aboriginal Australians, the ABS introduced special arrangements for the 1971 Census in the Northern Territory and Western Australia. Over time, the system has been gradually modified and geographically extended and it now constitutes a multi-faceted Indigenous Enumeration Strategy (IES) that is an integral part of the general census operation. In urban areas, this strategy addresses issues arising from distrust of officials, population mobility and large households; in remote areas it is designed to take into account geographic isolation, high mobility, traditional culture, communication problems due to language, and unfamiliarity with form-filling. The focus of the present analysis is on the strategy as applied in remote areas.

Choi and Gray (1985) provide a history and evaluation of the initial steps taken by the ABS in its attempt to improve the enumeration of Indigenous people in remote areas at the 1971, 1976 and 1981 Censuses. Much of what they report is paraphrased here. They note that attempts to improve the coverage of Indigenous people in isolated situations (those 'out of contact') can be traced to the 1966 Census with the recruitment of Aboriginal welfare bodies, mission superintendents, station owners, patrol officers and police to assist in the count. However, they consider the outcome to have been less than satisfactory. Knowledge of the location of many Aboriginal people was lacking, too few people were allocated to undertake the count, and the personnel chosen were inappropriate because of general distrust of authority (Choi & Gray 1985: 5).

By 1971, the new requirement to provide comprehensive data on the Indigenous population led to the development of dedicated procedures for application in the Northern Territory and Western Australia. For example, collection districts (CDs) were created for the first time in remote areas to cover most missions and government settlements, as well as certain pastoral properties known to contain concentrations of Aboriginal people. In the latter case, letters were sent to station owners, through the offices of pastoralists' associations, informing the owners about the census and requesting cooperation. Where possible, collectors were recruited from station staff, while Northern Territory patrol- 6 - officers and Western Australian Native Welfare officers were also enlisted (Choi & Gray 1985: 6). Some provision was also made for the first time to deal with enumeration problems among itinerant Indigenous populations in urban areas, although only in Darwin and Alice Springs. Logistical difficulties in conducting a simultaneous count over vast areas were addressed by commencing a population list one month prior to the census and adjusting for additions and departures at census date. Not surprisingly, this was found to be more effective in former mission and government settlements than in large reserves with indeterminate populations.

In the lead-up to the 1976 Census, the Department of Aboriginal Affairs (DAA) had been created, and responsibility for the census count of Indigenous people was centralised at the ABS in Canberra. As part of its expanding need for reliable population data, the DAA collaborated with the ABS in the design of special enumeration procedures for Indigenous people in town camps and remote communities, and input from the National Aboriginal Consultative Council (NACC) and other Aboriginal organisations was sought. Among the initiatives adopted were the use of Aboriginal census collectors in town camps, and the development of a special simplified questionnaire for communities where communication and literacy problems were identified. Flexibility in the timing of the count was also allowed in some Northern Territory communities in an attempt to accommodate significant movements of population.

Following a round of consultations with Aboriginal organisations and visits to major Aboriginal communities, the review of 1976 procedures recommended two significant changes for the 1981 enumeration that still resonate. The first involved the development of a special census schedule for use in interviewing individuals for whom English was a second or third language. The other was the devolving of central control of field procedures to the States and Territories, to take account of local conditions and to raise local awareness of the census. Over time this has led to some disparities in the implementation of the census methodology across jurisdictions, with the count being conducted sometimes on a place of enumeration (de facto) and sometimes on a place of usual residence (de jure) basis, and with individuals being recorded in a manner sometimes dictated more by expediency than by rigid adherence to the official methodology (Taylor 1993).

Enumeration by interview was pioneered in the Northern Territory at the 1976 Census and was extended to South Australia and Western Australia in 1981 (Queensland did not follow suit until 1991). In each of these jurisdictions, special CFOs were appointed to recruit and train collectors in Aboriginal communities, to promote pre-census publicity, and to physically locate communities (including outstations) to ensure their inclusion. In the Northern Territory the old system of relying heavily on police and other authorities for census collection was abandoned and a decision was made to recruit local Aboriginal people as interviewer-collectors in all discrete communities, with assistance in the major communities from adult educators (Loveday & Wade-Marshall 1985). A modification of the standard household census form was also introduced. It was broken down into separate components: a household form listing household members along with information pertaining to the household, and a personal form for recording the socioeconomic characteristics of each individual within the household. Information on these forms was acquired by interviewing a responsible adult member of each household. A Dwelling Check List form was also developed for use at the community level.

- 7 -

The benefit of devolving the jurisdiction of field operations is described by Loveday and Wade-Marshall (1985: 253) as enabling administrative flexibility resulting (in their estimation) in minimal underenumeration. Such flexibility included the production and distribution of customised pre-census publicity materials, the use of administrative records to fill information gaps, and return visits to communities where necessary.