In all the case-study areas, with the possible exception of the Alice Springs town camps, the case studies show that the role of the CFO, as presently conceived, is inherently impossible. Various factors contribute to this situation. The size of the areas for which the CFOs are responsible and the logistical difficulties arising from the remoteness of many discrete Indigenous communities from the main concentrations of population—and hence from administrative centres—is an intractable problem. The difficulties the CFOs experienced in three of the case-study areas in recruiting and retaining sufficient numbers of Community Coordinators (CCs) and collector-interviewers (CIs) threatened at times to undermine the whole exercise. In the way that the count is currently organised, the CFO cannot maintain control of the process across the entire region.

The training of the CFOs did not prepare them adequately for the realities of their task. Observation of this training (reported in Chapter 2) revealed two standout issues.

The combination of conducting the Community Housing and Infrastructure Needs Survey (CHINS), compiling the Discrete Indigenous Communities Database (DICD) and promoting upcoming census activities as a preliminary to census training proved a useful device for acquainting newly recruited CFOs with their field areas of responsibility, but the compression of these multiple tasks—especially the CHINS and the compilation of the DICD—into a single short field exercise also served to dilute the efficacy of efforts spent on census preparation and promotion.

Not only were CFOs new to the field operations they were about to encounter, so were those training them. The result was an inability to inject much corporate memory into the training process and this resulted in limited preparation for the many practical issues that individuals would face. To be fair, not all contingencies of this sort could be anticipated, but the truth is that many of the nuances involved in conducting the census were perforce learnt on the job. Essentially, CFO training requires more focus on practicalities and less on process. Greater attention should be given to the content of census forms and this could be aided by more explanation of the concepts, ideas and intentions that underpin form design. For example, the question of PTAs was rightly worrying the CFOs during their training, but they received only limited conceptual or substantive guidance on how to deal with them.

The support offered to the CFOs in the field was limited by structural and technological factors. Lack of access to computerised records that would have informed them of the progress of the count in nearby regions made it impossible for them to check, in the field, whether people ‘temporarily absent’ from where they were said to be residents were likely to have been counted in the places to which they were said to have gone. Lack of computers in the field also made keeping and updating their own records of progress much more arduous than necessary. They had no email access to the CMU, and were forced to use often-unreliable satellite phone links to maintain contact with the CMU. In the Northern Territory the managers at the CMU were responsible for many other tasks besides overseeing the IES, and often could not respond as quickly as was desirable to queries from the field.

It is significant that the most successful count observed in 2006 was in the Alice Springs town camps, where there was a vast improvement between 2001 and 2006. This can be attributed to two factors: the redesigned IHF and better engagement between the ABS and Tangentyere Council, the organisation that services and represents the town camps. Even in Alice Springs, however, there were factors that prevented the CFO and Tangentyere from cooperating fully with each other. For example, Tangentyere had developed a list of town camp residents through its own research work in 2005, but despite some early ideas about how it might be used, it was not in the end used to assist the census collection process.

At Wadeye and its outstations, there were signs of cooperation early in 2006 between the ABS and the Thamarrurr Regional Council (TRC), but when a suggested workshop at Wadeye failed to eventuate, the TRC began to feel little ownership of the coming census. Thereafter, the 2006 Census at Wadeye was destined to become an encounter between the ABS and individual households, with only limited involvement from the TRC. In the Arnhem Land case study, the local organisations felt no ownership of the census. They were not involved proactively in assisting the CFO, and little attempt was made to encourage proactive involvement. They were nevertheless willing to offer assistance of a limited kind when asked.

In Fitzroy Crossing, there were some boundary problems between the IES and the general census enumeration, which reflected Fitzroy Crossing’s status as an open, roadside town in a pastoral area of the Kimberley with a complex mix of Indigenous and settler populations. It appears that there were no formal procedures in place to ensure that the CFO and the Area Supervisor were aware of each other’s activities.

The recruitment and training processes for CCs and CIs were confounded by the contingencies of life in remote Indigenous communities. Acting as a CC or CI is a complex task. The individual, whose first language is not English in many cases, and who might not have highly developed literacy in English, must master the content of a long and complex form and the elaborate and unfamiliar administrative procedures (outlined in Chapter 1) involved in ensuring the consistency, accuracy and completeness of the count. The training was compressed because of the realistic perception that there was likely to be attrition in attendance if the training went for more than one day, but this compression, and the fact that the CFOs were not themselves trained as trainers, meant that the training delivered to the CCs and CIs was not very effective in many cases. We observed time pressures on the CFOs—the time it took to recruit people and ensure their attendance at training sessions, and the necessity because of this to deliver training to the CCs and CIs in a joint session rather than giving the CCs additional support and training. Time pressures too meant that very often the CFO could not provide sufficient backup and feedback to the CCs and CIs during the early days of the collection process because of the necessity to move on to another community to begin the process again.

As a result, administrative procedures often broke down in the face of the contingencies of everyday life. At Fitzroy Crossing, Master Dwelling Checklists (MDCs) and Interviewer Dwelling Checklists (IDCs) were used in an inverse administrative process for bringing together the paperwork after the enumeration, rather than as a planning tool before the enumeration. At Wadeye and in Arnhem Land, CCs and CIs did not follow the IDCs that they had been given, and arguably this was inevitable when the presence or absence of so many people was contingent on so many factors external to the census enumeration.

The census as it is currently constructed is essentially a dwelling-based count, which assumes some degree of sedentariness among the population being counted. Remote Indigenous populations, however, behave in ‘radically uncontained’ ways (Morphy 2007). In Chapter 3, it is shown how this led to major problems for census administration in Arnhem Land, as dwellings—and even whole settlements—were often found empty and people were to be found in all sorts of different places during the extended period of the count. Sometimes people were found away from their usual dwelling, but wanted to be counted and were in fact counted ‘as if’ they were back there. There was an interesting political aspect to this ‘as if’ counting, as well as some practical administrative issues. At other times, where someone at least was present at a dwelling, many people were counted as PTAs; however, the ability to judge whether these people would also be enumerated elsewhere was limited.

The account of the Wadeye and outstations enumeration also addresses the theme of the high mobility of the population being counted. In this case, the issue is not just constant mobility between Wadeye, the outstations and nearby Indigenous communities, but a high level of dry-season mobility into Darwin. Although many of the people who were in Darwin were enumerated in the town camps as visitors from Wadeye, many others were camping out. There was concern that these last people were not being counted in Wadeye as PTAs, and also might not be counted in Darwin; or that if they were counted in Darwin on the ‘Special Short Form’ used for people not associated with a dwelling that this did not identify their community of usual residence, and so they were lost to Wadeye’s final de jure count of usual residents.

The account of the 2006 IES in Fitzroy Crossing provides a useful widening of focus across a State/Territory border into Western Australia. The ABS administration in Western Australia clearly envisaged itself as doing a more ‘standard’ Indigenous enumeration than in the Northern Territory. The intention was that training of CCs and CIs would be done some considerable time before enumeration and that the time extension of enumeration would be kept to a minimum. At least in Fitzroy Crossing, however, and the southern Kimberley, there was not really that much difference from the time-extended rolling count that had been planned for the Northern Territory. Once training was done, there was encouragement to get on with the count straight away in case it took a while. Indeed, the enumeration dragged out in the southern Kimberley in just the same way as in the Northern Territory case studies.