The 2006 Census enumeration in the Thamarrurr region provides an example of the logistical and cross-cultural issues associated with implementation of the Indigenous Enumeration Strategy (IES) in a large Aboriginal town (Wadeye) and surrounding outstations (Figure 5.1). In the past 30 years, many former mission and government settlements across northern Australia that were established for the purposes of administering Aboriginal welfare polices have grown steadily in size and complexity, with several now achieving the status of ‘urban centre’ (more than 1000 people) within the Australian Standard Geographical Classification (ASGC). Among those with populations that now exceed this number are: Wadeye, Maningrida, Nguiu, Galiwinku, Milingimbi and Ngukurr (in the Northern Territory) and Aurukun, Palm Island, Yarrabah, Doomadgee, Mornington Island, Woorabinda and Cherbourg (in Queensland). The population trajectory for these towns is for continued growth while many more such ‘urban’ places are expected to emerge in time. Consequently, the observations made in respect of census operations at Wadeye are representative of a category of Aboriginal settlement that will be of increasing relevance for the IES and to Indigenous affairs policy in the future.
There are other ways in which Wadeye is representative of an emergent type. As a polyglot, overcrowded, under-resourced and growing settlement that has drawn disparate social groups from surrounding country (Taylor and Stanley 2005), Wadeye presents a set of social, economic and governance difficulties that are increasingly evident across remote Australia (Ah Kit 2002; Dillon 2007; Westbury and Dillon 2006) and that provide an essential backdrop to any assessment of the conduct of census operations and their effectiveness. Among these, at Wadeye, is a scale of anomie, especially among youth, that has led at times to outbreaks of civil disorder. For example, in April and May 2006, just two months before the planned census date, many of the town’s residents—anecdotal estimates of up to 500 were cited locally—had fled to set up temporary residence across the immediate region and beyond because of relentless inter-family feuding and destruction of property (a similar situation had occurred at the time of the previous census in 2001). Part of the fallout was a delay in conducting a ballot for the Thamarrurr Regional Council (TRC), with the elected body in abeyance for a period until new elections could be arranged in mid July. This not only complicated communication with community representatives regarding census preparations, it also meant that the council and the people were preoccupied with electoral matters in the crucial weeks leading up to census day.
Needless to say, these were not ideal conditions in which to plan and conduct a census enumeration, and they underline the fact that the IES does not operate on a blank canvas. In order to be successful as a strategy, it is—or at least it should be—as much about having the capacity to anticipate and successfully negotiate these sorts of contingencies as anything else. One of the key questions posed by the research team was whether the strategy adopted for the enumeration of remote Indigenous communities was suitably adapted to meet the cultural and contingent situations encountered. By force of circumstances such as those above, the focus of census observation at Wadeye was concerned substantially with this question.
Along the way, two weeks spent at Wadeye and its outstations observing the census training and most of the census enumeration between 24 July and 6 August, together with subsequent follow-up, also provided a unique opportunity to witness at close hand the process of translation between two cultures: the local Indigenous community and the nation-state. This was manifest in several arenas: in the handling of interactions with community leadership, in the strategies adopted to engage, train and supervise an Indigenous census workforce, in the logistics deployed to connect with a mobile and scattered population and in the categorisations and interpretations associated with the questions on the census questionnaire. Other, more specific, issues concerning innovations in the structure of the Interviewer Household Form (IHF) were considered secondary to these more generic concerns, not because they were less important, but because they turned out to be of less consequence. They are discussed in Appendix B.
As noted, the IES does not unfold in a vacuum. In each Indigenous settlement there will be particular cultural and governance dynamics that can substantially influence the conduct of field operations. This is because—in stark contrast with the mainstream census—the interview format of the IES in remote settlements creates an encounter not just with individuals and their households, but with a social collective and related representative and administrative structures. This creates an institutional arrangement whereby councils and other local organisations can act as the interface between populations to be counted and the state (instantiated in this case by the Australian Bureau of Statistics [ABS] and its census).
In terms of the collective, the population of the Thamarrurr region is embedded in dense social networks across some 40 extended paternal family groups with affiliations to 20 locally defined and recognised clans. While members of all of these social units are present in the town of Wadeye, country and family ties beyond the town produce a population that is widely and variously scattered at any one time at local outstations and in neighbouring communities and towns across an area from as far south as Kununurra and Timber Creek through the Daly region and north to the Cox Peninsula and Darwin. This is especially so in the dry season around census time and within this network there is considerable short-term mobility. This configuration immediately raises a set of questions regarding what properly constitutes the population of the Thamarrurr region and how best this might be counted. In 2006, the ABS adopted a de facto enumeration to address this question—that is, assigning people to the place where they were found at census time rather than to the place where they usually lived—but issues still arose regarding the impact of frequent mobility and the consequent adequacy of de facto and usual resident counts. These are dealt with later.
As for representative and administrative structures, a significant development at Wadeye was the formation in 2003 of the TRC as a local government body built around representation from the 20 clans. At the same time, the TRC entered into a Shared Responsibility Agreement (SRA) with the Commonwealth and Northern Territory governments as one of several Council of Australian Governments (COAG) trial sites for Indigenous Communities Coordination Pilot projects aimed at effecting whole-of-government cooperative approaches to service delivery. These were based on the idea of streamlining government processes and supporting some restoration to local Indigenous populations of responsibility for, and control over, decision-making regarding service delivery and general planning for social and economic development. Within this arrangement, the TRC gained a substantial local profile and presence and it is fair to say that the council assumed its new responsibilities with some vigour. One activity it conducted under the auspices of the COAG trial was its own census of usual residents in 2003, and again in 2005. These activities were undertaken for the explicit reason that previous census counts conducted by the ABS were felt by the TRC to have been deficient given their own estimation of numbers in the region. A number of salient points flow from all of this.
In the period leading up to the census, the TRC had established itself as a prominent institution heavily involved with government in a partnership approach to regional social planning with strong interests in census-taking. As part of these and related survey activities—including by the ABS, for example, in the 2002 National Aboriginal and Torres Strait Islander Survey—a small cadre of experienced local enumerators had been developed and the council was keen to nurture this expertise. By 2006, however, after almost three years of participation in the COAG trial, the experiment in whole-of-government approaches to service delivery had all but collapsed, with very few beneficial outcomes and a loss of confidence at Wadeye in the COAG process (Commonwealth of Australia 2006: 27–33). It was into this environment of initial local optimism and then despair at the state’s intentions that the ABS Census Management Unit (CMU) based in Darwin was about to step.
That is not to say that community–government relations were never propitious. In line with its pursuit of regional planning and cognisant of the approaching 2006 Census, the TRC and the ABS began in late 2005 to explore ways to ensure that people in the region were well prepared to participate optimally in the enumeration. Discussions with census officials were held at Wadeye and in Darwin, and an attempt was made to arrange for a workshop at Wadeye to explain the census questions and for this to be communicated in Murrinh-patha in line with council practice. The fact that this workshop failed to eventuate meant that by the time the enumeration finally began in late July, the council—and through it, the community—felt little ownership of or familiarity with the census content and process. By then, some council members expressed the view that the ABS alone had carriage and responsibility for the census. This was a far cry from the partnership approach to regional planning that had prevailed earlier and it meant that, for the most part, the census was going to be an encounter directly between the ABS and individual householders—a far more difficult task than working in effective collaboration with the council (see Sanders, Chapter 3 for the benefits of working in partnership with representative structures).
Consequently, communications regarding final census logistics between the ABS and the interface with the population to be counted became somewhat perfunctory. Almost inevitably, given the Census Field Officer’s (CFO) myriad responsibilities in preparing for the census across a vast area from Lajamanu to the Tiwi Islands, these communications were compressed in time and content. For example, before and during the conduct of the Community Housing and Infrastructure Needs Survey (CHINS) at Wadeye (at the end of April), a request was made to the TRC for assistance in mustering a census workforce of 20–30 local people. The limited opportunity for effective follow-up (just one visit in early July) meant, however, that by the time census training began in late July the task of mustering an interview team was still to be done. At an earlier visit, an approach was made to the Thamarrurr Regional School to obtain assistance with census workers, but to no avail, given staff commitments. As for advertising the census, the usual publicity packs were distributed via the council and the school, while a brief explanation of the purpose of the census was broadcast by a member of the TRC on Broadcasting for Remote Aboriginal Communities Scheme (BRACS) radio. Other attempts to pursue logistics were made by the CFO via phone and email in the lead up to census day, either from Darwin or on the road in other communities. Not surprisingly, such communication from a distance proved ineffective in terms of preparing a census workforce. According to the CFO, the TRC administration was keen to assist, but any prospective workers were already fully occupied in their normal jobs. In the meantime, the TRC provided a map of all dwellings in Wadeye and at outstations, but it was reluctant to supply a copy of the Thamarrurr population database, as requested, as it regarded this as confidential.
In practical terms, what eventuated from all of these interactions, on the very eve of planned commencement of the census, was the deployment of two TRC staff—one local Aboriginal person as a collector-interviewer (CI) and one recently arrived non-Aboriginal person as a Community Coordinator (CC)—a council vehicle for travel around town and to outstations and council assistance on the first day of interviewer and CC training in driving around the community to mobilise potential interviewers. The TRC also provided exclusive access to a training room, and later an office, to serve as a base for the duration of census operations, which turned out to be more than a calendar month.
Interestingly, at the same time that these arrangements had been established, another Commonwealth agency, Centrelink, had also assembled teams of six Commonwealth officers to conduct a survey in Wadeye and across the region asking householders many of the same questions as the ABS regarding resident population, family composition, income and employment. While no observation of the possible impact of this on the conduct of the census was possible, to the extent that any repetition involved might have confused or irritated residents, at the very least it suggests a lack of coordination between Commonwealth agencies.
By the time the census enumeration began in Wadeye on 26 July, the impression obtained from discussion with key informants was of a regional population that was largely ignorant of the imminent census activities, that was administratively detached from the process and that was otherwise diverted by issues of more pressing concern including elections and Centrelink processing. One important practical consequence, as we shall see, was a failure to engage the small cadre of experienced interviewers who had been used for a variety of previous local surveys and potential local CCs who had some standing in the community and knowledge of the population, because by census time they were all fully engaged in other work. It also meant that the time available for CC and CI training was highly restricted.