Among the CFOs, opinion was divided on the question of how to involve local organisations in the census process. Some felt that they should be involved much more directly in the organisation of the count, so that best use could be made of their local knowledge and of their resources. They saw a need to develop a skill set out in the communities that could be tapped into at census time. Others thought that organisations should be kept at arm’s length because they had a conflict of interest: it was felt to be in their interests to maximise the census count in their area because this had resource implications.^[1] Given this division of opinion, it is helpful that the evidence from our observations is unequivocal: the single most important factor that can enhance the quality of delivery and outcome from the IES is greater and more sustained engagement with local organisations and their personnel. In the Northern Territory in the future, this sector will include the new regional authorities as well as local community organisations.

The IES is designed in part to take advantage of the local knowledge of CCs and CIs, but it is clear that in the current social dynamic that operates in most communities it is unrealistic to expect the CFOs to be able to find and recruit sufficient numbers of adequately qualified and strongly motivated local Indigenous field staff. A variety of factors are involved, including low levels of literacy and numeracy in remote communities, which reduces the size of the potential pool of workers. Nearly all other major problems with the count follow from this. With inadequate numbers of collectors, the process becomes unduly prolonged and the problem of ensuring an accurate count is compounded by the mobility of the population. Moreover, those collectors who are recruited are faced with very burdensome workloads and, not surprisingly, many of them lose motivation, particularly when the payment system is slow and unsatisfactory. Most recruits are people with many other social obligations and it is easy for those obligations to take precedence over working for the ABS, particularly if people have to work long hours for a prolonged period to complete their workload.

In 2006, it was also possible that the general political climate had alienated many of those people who might have acted as CCs and CIs—that is, those who were more literate and more aware of events in the world beyond the local. Such people are no longer persuaded that taking part in the census delivers benefits to them and their communities. Moreover, there is no sense among the general remote Indigenous population—at least in the communities where we observed the count—that the census enumeration is anything other than yet another ‘government’ intervention in their lives, which serves no direct purpose as far as they can see. It is viewed as irrelevant to their concerns, so more immediate local events always take precedence over taking part in the census. The inability to recruit sufficient numbers of field staff was a widespread phenomenon, and blame cannot be laid at the door of individual CFOs. The problem lies with the general nature of the engagement between the ABS and local populations.

In 2006, the lack of CCs and CIs, combined with the size of the regions they had to cover, made the job of the CFOs and their assistants well nigh impossible. One of the solutions advocated at the CMU debriefing was, in effect, to throw more bodies into the fray by increasing the number of CFOs next time, and maybe even the number of non-Indigenous CCs and CIs. It was suggested that greater use might also be made in 2011 of the team of Indigenous Assistants from Darwin. Such a solution begins to negate the rationale behind the use of local Indigenous people in the IES, and denies the value of their local knowledge.

There is another way to approach the problem, and that is to increase the sense of ownership of the census process at the local level, not so much by attempting a mass education campaign, but by strategic long-term engagement with local Indigenous organisations and their Indigenous and non-Indigenous staff, and with the staff of the regional authorities in the Northern Territory—once established. Such a relationship could be of mutual benefit.

At the moment, there is not enough capacity to deal with the inevitable contingencies that arise from the unpredictability and the scale of short-term mobility of remote Indigenous populations. It cannot be otherwise when the ABS’s primary engagement with remote Indigenous populations is on a short-term basis once every five years.

The detailed description of the course of the census count in part of the Yolngu area of Arnhem Land (Chapter 4) highlighted the major causes of mobility in that particular region: funerals, first and foremost, and other contributory factors such as the need to visit service centres and visit kin in other communities. Although contingent and therefore unpredictable events such as the death of a particular person initiate episodes of mobility, once that mobility is in train its patterns are to some extent predictable to people with detailed knowledge of the networks of kinship and ceremonial connections in a particular region—in other words, to locals with local knowledge. The IES could make much better use of this store of knowledge than it does at the moment, by giving greater responsibility for the planning of the count to locals. It could not do this at the moment, because locals are not qualified to take on such a responsibility.

The CC position therefore needs substantial rethinking, in the context of continuing engagement with local organisations and regional authorities. The emphasis should be on training selected employees—Indigenous and non-Indigenous—in basic demographic methodology, and on their employment on micro-demographic projects and other surveys between censuses, as has happened at Wadeye and Tangentyere, so that, come census time, there is for each Collection District (CD) a set of people already trained for the enhanced CC role. This will decrease the burden on the CFO, who will be able to take a genuinely regional approach to their task and make better and more systematic use of the local knowledge of the CCs. For example, CCs could be alerted to watch out for patterns in the PTA data as they emerge in the field, so that crosschecking with other communities can take place at that point, rather than after the fact. They could also be instrumental in preparing access to local administrative data sets that could assist in providing vital demographic data such as dates of birth.

The enhanced CC role should be extended to responsibility for recruitment and training of the CI workforce. This should enable the training of CIs to take place in a less hurried manner—because it will not be dependent on the presence of the CFO—and will most likely improve recruitment and retention rates because the CCs will be local, will not have to leave the area during the count and will have received some basic training in the management of their CI workforce.

These individuals would be a permanent resource that the ABS could call on. They could be involved actively in promoting the census to their own communities, in engaging and training the CI workforce for the census and in planning the count itself. The role of the CFO would be transformed, and the current regional structure could probably be maintained. In effect, what we suggest is an enhancement of the role of the CC, who would now be a trained person with experience in other survey work for the community, and a corresponding change in the CFO role to that of a regional coordinator and facilitator, with specialist knowledge of census procedures. The CFOs would still provide the CCs with training to carry out the particular census tasks, but they would be training people who already had the skill set required for the job.

These recommendations, if implemented, should also result in less drawn-out counts. It seems incontrovertible that the more the period of the count is extended, the more complex the PTA problem becomes, and the more scope there is for double-counting and for missing people altogether.

All of these recommendations would also improve other aspects of the count. They would result in substantially more complete sets of data for the questions on the form that relate to factors other than just basic age, sex and usual residence. More highly trained CCs would also be able to deliver better training to the CIs on the purposes and meaning of the questions, resulting in improvements in the quality of the data collected.

Underpinning all these changes there should be a much better use of information technology. Databases that are updated in the field for their areas of responsibility by the CFOs and/or the CCs to show the current progress of the count—and that are accessible to all CFOs—would make the task of assessing PTA data much easier in the field, particularly where mobility is occurring across regional boundaries.

It is likely that the institutional landscape of remote Australia, particularly in the Northern Territory, will look very different in 2011. The Northern Territory government’s plans for the development of regional authorities and shires is proceeding apace, and many local community organisations will cease to exist, or will be amalgamated into larger organisations with altered functions. The need for good local population data will, however, still be there. Many of the well-established organisations will survive. They will continue to deliver services and infrastructure under contract to their local shires, and increasingly they will become agencies that support and deliver economic development to their members. Unless the ABS keeps abreast of these changes it will experience considerable logistical difficulties in the field in 2011.

At one level, this reorganisation of service delivery in remote parts of the Territory will simplify the task of the ABS: there will be fewer organisations on the ground. At another level, it could increase the difficulty of compiling and updating information at the level of the community or the CD, unless the new regional authorities have the capacity to assist the ABS in this task.

For their part, many community organisations—and regional authorities once they are established—would probably welcome a continuing engagement with the ABS. Good-quality demographic data at the local level would help them in their planning at many levels, yet they do not currently have the expertise, time or resources to gather these data effectively for themselves. As a result, local administrative data sets are often very inadequate, and although there is much informal knowledge about patterns of mobility and their effects, formal analyses of mobility are almost non-existent. The ABS—by involving itself creatively in training staff of regional authorities and local organisations to carry out such work—would be helping them help themselves, while simultaneously building capacity in individuals that could be utilised at census time, and also in the context of other ABS surveys.

The symbiotic relationship between the ABS and regional authorities and community organisations would have benefits for all. It would lead to a vast improvement in local administrative data sets and to a more cooperative attitude on the part of organisations in allowing these to be used to validate the census count. Locally based research on patterns of mobility could also contribute significantly to our understanding of population mobility in remote Australia.

Between now and 2011, there will be considerable changes in the Indigenous landscape in remote Australia, particularly in the Northern Territory, as a result of Commonwealth government policy settings. It is possible that many small outstation or homeland communities will cease to exist, owing to the withdrawal of funding for community housing and infrastructure. It has been proposed that the Community Housing and Infrastructure Program be abolished and that new initiatives to address the need for housing should ‘[c]ontinue the shift away from building housing on “on country” outstations and homelands and focus on building new housing where there is access to education, health, law and order and other basic services’ (PricewaterhouseCooper 2007: 23). Recent changes to—and indeed the abolition of—the CDEP program, and the projected removal of the remote-area exemption for Newstart participants could also have the effect of turning many outstation dwellers into ‘economic migrants’, compelled to move to larger population centres in pursuit of ‘real’ jobs. Once again, the ABS needs to monitor these processes in the period leading up to the 2011 census, and the most efficient way to do so is to enlist the local knowledge of the staff of regional authorities and local community organisations.

These suggestions involve a new way of thinking about the engagement between the ABS and the Indigenous public, as mediated by the census. They would involve a radical change in the IES. Instead of being an intermittent strategy that manifests itself temporarily every five years, it would become a continual process of engagement. This would obviously have implications for the internal structure of the ABS, at least in the Northern Territory. It would involve the development of an Indigenous engagement unit to replace the single State Indigenous Manager. The unit’s initial role would be to make an audit of regional authorities and community organisations to assess their potential as, and interest in being, sites for the training of local people as, in effect, population specialists. The unit could also be responsible for keeping the DICD for its State or Territory up to date. As noted above, given the likelihood of sweeping changes in the institutional landscape of remote Australia in the next few years—and the possibility of substantial changes in settlement patterns—this will be an important task if the ABS is to have good local knowledge in 2011. The development of training materials and local projects would necessitate creative engagement with the TAFE system and/or Charles Darwin University and other tertiary institutions around the country. Certain projects could be university-based research projects with the local organisations as contributing partners.

What is being advocated here is not just incremental improvements to the IES and therefore to the quality of census data on Indigenous Australians. It is a proposal for a new kind of engagement with Indigenous organisations and local government agencies that will ultimately yield high-quality micro-demographic data that will be of substantial benefit to local communities, while simultaneously contributing to the success of future censuses. It is a proposal to harness the agency of local Indigenous people more effectively, in an organisational setting, in pursuit of a strategy that better addresses the complex contingencies that the census process confronts in the field in remote Indigenous Australia.