The case-study chapters (Chapters 3–6) describe what really happened on the ground during the count. In this section, we outline briefly the structures and processes of the count as envisaged and planned for by the ABS for the Northern Territory IES.^[2]

The IES was extended in time well beyond the process of enumeration itself—before and after—beginning in November 2005 with training for State Indigenous Managers (SIMs). SIMs were employed and CMUs established within ABS regional offices in each State and Territory. They were responsible for coordination of ensuing field operations associated with the Community Housing and Infrastructure Needs Survey (CHINS) and the census, including the training of the CFOs, who were responsible for organising the collection of CHINS and then census data in the regions to which they were assigned.

As in previous years, the IES involved firstly the designation of the communities and areas in which the remote strategy was to be applied. These then became the responsibility of the CFOs, whereas the ‘mainstream’ count was the responsibility of the local area supervisors. The IES involved the use of a form—the IHF—which differed in its structure and in the content of some of its questions from the form used in the mainstream count. In particular, the IHF was designed to attempt to capture data on household structure, education, employment and socio-cultural factors such as religion, language use and ethnicity, using questions and options for responses framed to allow for the ‘difference’ of remote Indigenous populations.

As in the past, the IES attempted also to mitigate the effects of remoteness and low levels of literacy in English in the remote Indigenous population. The census forms were not dropped off at households; rather, local CIs, ideally managed by local CCs, took the IHFs to the individual households in the communities and filled them in with the help of the household members. The responsibility for training these local temporary staff rested with the CFO for the region.

In the Northern Territory, it was decided—as in past censuses—that such a process made it logistically impossible for the count to take place on a single night. Accordingly—as in the past—the IES in the Northern Territory employed a rolling count over an extended period. The time frame initially allowed for the count was six weeks, and it was the responsibility of the CFO to organise the count within this time frame in their designated region. In 2006, for the first time, in acknowledgement that in reality this was a very exacting task, the CFOs were assigned an Assistant—also trained at the regional office—to go with them into the field. In the Northern Territory, a couple of ‘floating’ CFOs were trained, who were not assigned particular regions; their job was to provide backup wherever and whenever it was deemed necessary. In Western Australia, it was felt that the IES could be completed in a week, being in effect a slightly extended version of the more usual census method of counting on a single night.

In the best-case scenario, as detailed and emphasised in the training delivered to the CFOs, the IES should have proceeded as follows. During the initial visit to each community during the CHINS exercise, the CFO would also complete a form for each community, to be entered onto a new Discrete Indigenous Communities Database (DICD), which was being compiled for the first time for the National Centre for Aboriginal and Torres Strait Islander Statistics (NCATSIS). These data and the initial population figures revealed by the CHINS—together with the population figures from the 2001 Census for the community—would help the CFO to determine how many CCs and CIs would be needed to achieve a timely and complete count in the area. During this initial visit, the CFO would also carry out an awareness-raising exercise about the impending census by liaising with local community organisations and potentially helpful individuals, such as the census volunteers who were being recruited locally (for the first time) to provide logistical backup and assistance. If they had time, they might visit the local school and so on, and they had posters and other publicity materials to distribute in the communities. The CFO would also begin the process—with the help of local organisations—of identifying and meeting potential CCs for the census process.

After returning to the State or Territory CMU for a week’s training on the IES procedures, the CFO with their Assistant would then return to their designated region to begin work. Their task, in each of the communities in their region, would be to recruit and train the necessary CCs and CIs for each community, sign them up as temporary ABS employees and ensure that all the details necessary for their payment were relayed to the CMU.

The first task after training would be for the CFO and the CC (or CCs) to go around the community—which had been assigned the unique identifier code for its Collection District (CD)—and compile a Master Dwelling Checklist (MDC) that would include all private dwellings, temporary dwellings and non-private dwellings in the community. Each separate dwelling would be assigned a unique Census Record Number (CRN). Non-Indigenous households and non-private dwellings would be identified and special arrangements made for the delivery and collection of the forms relating to them. Indigenous households, once identified, would be divided among the available CIs, using the local knowledge of the CCs to distribute the workloads most appropriately. Each CI would then be given an Interviewer Dwelling Checklist (IDC), on which the CD and the CRNs of each dwelling to be visited by them was to be listed, along with the surname of the ‘head’ of the household. They were to complete the details on this checklist as they went along. Once a household had been counted, the dwelling would be ticked off the list and the numbers of males and females in the household recorded on the IDC.

The task of the CCs—aided initially by the CFO until they were confident that the CCs had fully understood their responsibilities—would be to manage the CIs by ensuring that they visited all the dwellings on their list, and reassigning workloads if unplanned-for contingencies arose. It would be their task to collect and store securely the completed forms and to transfer the details of people counted from the IDCs to the MDC, having first checked that the forms had been completed fully and that the numbers of people on the IHFs tallied with the totals listed on the IDC. One copy of the IDC would be kept by the CI as a record of the pay owing to them, one would stay with the CC and one would be forwarded to the CMU by the CFO to initiate the payment process. It was envisaged—and it would be necessary if the count for the CFO’s region was to be completed in the time allowed—that a day or two after the beginning of the count, the CFO could leave the community to start the same process in another of the CDs in their region, leaving the Assistant CFO behind for a while if this was felt to be necessary.

After the count was completed in a CD, and all the IHFs were returned to the CCs, the CFO would return to the community and double-check with the help of the CCs that all dwellings had indeed been visited and accounted for, that the totals of males and females on the MDC and the IDCs tallied and that all forms had been completed correctly. The CFO would also use available administrative data and the figures from the CHINS exercise and the 2001 Census to determine whether or not the coverage was complete. They would be alert for large discrepancies in totals, and for under-counting of particular sectors of the population, such as young men, children and infants. If there were any such discrepancies, they would be documented, followed up in the field and accounted for.

The CFO would then return all the forms for a CD, boxed together, to the CMU. There, further checks would be made (see Chapter 7) before the forms were finally sent en masse to the DPC in Melbourne. The boxes containing the IHFs would contain—in addition to the forms themselves—copies of the MDC and the IDCs, the completed DICD forms and two checklists, one completed by the CFO before the forms left the CD and one completed at the CMU.

This very quick sketch of what is a very complicated administrative and logistical exercise is designed as a necessary background to the case-study chapters. It will become obvious that the reality fell far short of the ideal, in the face of the complex realities encountered by the CFOs and their teams on the ground.

One final important step in the process—which was new in 2006—was the extension of the Post Enumeration Survey (PES), designed to estimate census net under-count (ABS 2007) in remote areas and discrete Indigenous communities covered by the IES. This survey is conducted by the ABS a month after, but independent of, the census, and is used to estimate numbers of people missed or counted more than once in the census enumeration and then to adjust population estimates accordingly. In the past, the PES has not been carried out in remote Australia and discrete Indigenous communities and the rate of people missed or counted twice there has been estimated as roughly equivalent to elsewhere. This survey was not observed by the current research team because of its independence as a process from the census itself. John Taylor (2007b: 18) reports that, according to the ABS’s data, ‘In Western Australia around 24 per cent of the Indigenous population was estimated to have been overlooked by the census; in the Northern Territory the figure was 19 per cent.’ This recognition of a larger than normal under-count has significant implications for the use of census statistics in matters concerning Indigenous affairs funding. Our observations also show that the data collected for many of the people who were counted were incomplete (see Chapter 7). Together, these two facts point to some very serious problems with the IES.

We conclude that, despite the best intentions, the census is failing to capture adequately the characteristics of remote Indigenous populations. In Chapter 9, we call for a substantial rethink of the way in which the ABS engages with Indigenous communities and their organisations.