Sometimes, dwellings that CIs and CCs thought were deserted turned out to be otherwise, but only because someone at the TRC chanced to point this out. Likewise, late in the count, no one in the census team seemed aware that a whole (and large) family group had been overlooked because the dwelling they were usually associated with was in disrepair. It transpired—when the team was informed again by the council—that these people were camping in nearby bush and hurried arrangements were made to track them down. While not reflective of mobility, a similar oversight occurred towards the end of the enumeration with regard to aged pensioners at the respite centre who were also overlooked until a casual remark by the housing manager brought this to light.

The extended timing of the census compounds the complications caused by this residential mosaic. Because people travel into and out of Wadeye to outstations and elsewhere on a regular, sometimes daily, basis, the question of whether individuals were either overlooked or picked up twice—or even more times—as the enumeration rolled out, was very real. Of course, ABS checking procedures should have dealt with any of the latter cases, but no procedures were deployed to address the former. One classic example of this dilemma was played out around the public holiday weekend of the Royal Darwin Show, beginning on Friday 27 July. Because of anticipated lack of activity in town, the CFO decided to swing through some of the region’s northern outstations during the holiday period and conduct the enumeration there. On the previous afternoon, after a large funeral for a young girl, numerous vehicles had turned up in Wadeye from outlying areas to stock up at the store for the long weekend. Among those moving in were many members of a large family group from one of the northern outstations that had a population in 2005 of about 80 people. By this time, the census at Wadeye had more or less wound down and was resumed again (slowly) only on the Monday. On that same day, the CFO was back in town to report that not many people were found at the outstations and that some dwellings were vacated for a funeral ceremony. Almost simultaneously, most of the same aforementioned family group members were beginning their circuitous route back home. This raised the prospect that these people were not counted, although it should be noted that the CFO’s trip also uncovered people at outstations who had already been counted in town.

Likewise, a group of older people who had been at Tchindi Outstation—which wasn’t visited by the census team on the assumption that no one was there—eventually turned up in town and were enumerated at a previously listed vacant dwelling, but only because one of the CCs was informed of their arrival by a council employee. Elsewhere, two dwellings at Wudapuli Outstation—which was not visited until the middle of August—were found empty because the occupants had moved temporarily to Warmun in the east Kimberley for ceremonial business. Interestingly, an attempt was made by the CMU in Darwin to extract information on this group from ABS colleagues in Western Australia, but to no avail. It is also likely that the lack of people at Kuduntiga Outstation as reported by one of the CCs was due to their absence at ceremonial business in Elliot. These types of oversight and failures to track people seem inevitable when following a dwelling-focused census methodology that is extended over time among a mobile population. While at one level the extended nature of the census count can assist by allowing time for people to move (literally) into scope, the main lesson from Wadeye is that this is likely to assist their enumeration only if constant input of local intelligence on the whereabouts of people forms part of the process.

Confusion of this sort was not aided by an at times somewhat culturally inflected interpretation of the wording to Questions 10 and 11 on the part of interviewers and respondents. Question 10 asks, ‘Are there any persons who live here most of the time but are away?’ Given that high levels of temporary mobility are known to exist among the regional population, it was curious to the observer that affirmative answers to this question were very few, yet in a number of households where individuals were known to the observer to be away in other locations—in the sense that the ABS meant—their absence was not recorded. On quizzing interviewers about this issue, the fact that they were ‘away’—and in cases could have been so for some days—did not register as an absence. This was not because respondents made a judgment that they were likely to be counted elsewhere—as per the instruction for Question 11, the ‘Persons Temporarily Absent’ (PTA) table (see Appendix A)—rather it was because they were still considered to be part of the present household and therefore not away. In any case, the idea that respondents, or CIs, could assess whether absent people might be counted elsewhere was highly presumptuous and generally avoided, at least in those instances observed.

This ‘not absent’ response was all well and good if such people subsequently appeared in answer to Question 12. A sometimes literal interpretation of ‘people who are living or staying here now’ (my emphasis, meaning ‘at this moment’), however, meant that in some cases (admittedly only three observed) people who were absent at the shops, or were out bush for the day, were omitted. In an interesting variation on this, an instance was observed of a new-born child who was omitted from a household count because it had just left with its mother for the day to Palumpa and no one knew its name. Some confusion also arose over whom to indicate as Person 1 in Question 12 if the head of the household was listed as a PTA in Question 11. One final aberration surrounded the category ‘visitor’. Though few people seemed to nominate this, where it was used the individuals referred to were fairly long-term residents and the term ‘visitor’ was clearly being used as a social category as much as a residential one.

As mentioned, a common characteristic of the Wadeye Aboriginal population—and of Aboriginal populations across the Northern Territory for that matter—is their temporary movement into regional centres such as Darwin for a wide variety of reasons ranging from service access and recreation to funerals. This movement is heightened in the dry season, and was augmented in 2006 by civil disorder and by the fact that the official start of counting at Wadeye (26 July) coincided with public holiday weekends, first for the Royal Darwin Show (27–29 July). The Darwin Cup (5–7 August) presented an additional attraction at that time for Wadeye residents. According to local informants, other places where usual residents of the Thamarrurr region were to be found during August 2006—aside from those already mentioned—included Palumpa, Peppimenarti, Daly River, Belyuen, the Tiwi Islands, Milingimbi, Wyndham, Kununurra and the greater Darwin area.

When visiting Darwin, Wadeye people locate themselves in various residential settings including in conventional housing with relatives in town or in motels, at town camps such as Railway Dam, Knuckey’s Lagoon and Palmerston, at the Bagot Community and at various camping spots near the Catholic Mission headquarters in Stuart Park, by the Murin air terminal, the Airport Hotel, Lim’s Hotel, Nightcliff Oval, the Catholic Mission hostel at Berrimah and along Rapid Creek. Given the likely numbers involved—estimated anecdotally in August 2006 at about 200 people—a key issue for the usual residence count of the Wadeye population was whether these people were captured by the census in Darwin and, if so, how the various forms used were filled out.

For example, those in conventional housing or in motels should have been captured as visitors on the standard mainstream form in answer to Question 2, but to be part of the Wadeye usual residence count they would need to have indicated ‘Wadeye’ in answer to Question 8 on usual residence. Those in town camps faced a similar combination of questions and issues, in Questions 12 and 15, on the IHF. For those camping out in urban areas, the general strategy employed by the ABS in the Northern Territory was to use the standard Special Short Form designed for homeless people. According to the CMU, this was deployed for two days from 9 August at known regular camping sites around the Darwin urban area. The particular feature of this form and its application that has importance for census counts is that it contains no question on usual residence. Information was primarily obtained through interview or self enumeration as the first preference, or by observation when circumstances prevented this. For Wadeye residents who were camping out in Darwin—and in any other urban area in the Northern Territory—the ABS, however, adopted a different approach in recognition of the large-scale dispersal of Wadeye residents that occurred due to community tensions just before the census. The procedure for this group was to use abridged mainstream forms (household and personal) rather than the Special Short Form, therefore activating the usual place of residence question. While this was well intentioned, it turned out that only a small number of ‘homeless’ Wadeye people were enumerated successfully using this approach, so any individuals enumerated in other camping locations using the standard Special Short Form method would have been lost to the Wadeye usual resident count.