As noted earlier, the 2006 Census took place at a time of considerable change in Indigenous affairs. In particular, the ‘sustainability’ of small homelands was being questioned and they were being portrayed in some circles as ‘cultural museums’ where the CDEP program had created a culture of dependency and fostered a ‘recreational’ lifestyle. Some commentators—and politicians—were advocating forms of social engineering that would ‘encourage’ people to leave the homelands and remote communities for ‘real’ jobs in the towns. The CDEP program was in the process of being refashioned, with less of an emphasis on ‘community development’ and an increasing emphasis on ‘training for real jobs’. Yolngu were aware of these developments and many were very worried about their implications. Many of their community organisations were under severe pressure as they attempted to adjust to the new demands that were being placed on them, particularly by the changes to the CDEP program.

Yolngu had been among the first people to embrace the homelands movement of the early 1970s—indeed, they were active instigators of it. They began the move back to the homelands from the missions before the time when, under the Whitlam government, the movement began to receive government support under the rubric of ‘self-determination’. Most homelands Yolngu want to continue living on their homelands. Many of them have been talking for some time about their desire to build local economies so that their young people will have jobs. Recent developments had shown them that they could no longer rely on government to ‘look after’ them, and had at the same time highlighted their vulnerability to externally imposed change.

For better or worse, most Yolngu perceived the census as an instrument of government, so the count took place in a politically charged atmosphere. I observed much more questioning of its purpose than in 2001, and more cynicism about the uses to which the data might be put. There was occasional resistance or politically motivated responses to some of the questions, particularly those concerning residence, and, interestingly, the ability to speak English. In 2001, Yolngu were not seeing fluency in English as a political issue—they tended to estimate their own and other people’s abilities using objective criteria and, from an English speaker’s point of view, they tended to overestimate people’s ability. In 2006, I heard the leader of one homeland, whose spoken English is reasonably fluent, exhorting the CIs—loudly, so that many people could hear him, in English:

The politicisation of the ‘residence’ issue was manifested in two ways. Some people questioned the notion of a single place of residence. One man commented, for example: ‘I live in both places [A and A12]. Sometimes at one, sometimes at the other, because of my job here [at A]. I can’t say which is “most of the time”. I should put both.’

More significantly, the political climate led to several homelands being enumerated ‘as if’ their residents were at the homeland rather than where they really were when they were filling out the form. People reasoned that they were only temporarily absent, for funerals or other reasons, but that if they allowed themselves to be enumerated as ‘visitors’ at A, where they really were at the time, the ‘government will say that we’re not really living at our home, they will say it’s empty’. Some of those who were included as residents of these homelands were indeed people who spent most of their time there, but others were people who went backwards and forwards between A and the homeland, much like the family from C3 discussed earlier. There was therefore potential for double-counting.

In most cases, the ‘as if’ practice did not lead to double-counting because these were small communities, all of the members of which were at A at the time of the count. They had made a joint and conscious decision to be counted as if they were at home, so they were not also counted as visitors at A. In one case (A11), however, there is a household whose ‘usual residents’ commute between A11 and the community of B because one of them has a job in the store at B. They were at B when the other usual residents of A11 decided to count themselves ‘as if’ they were at A11. The people at A also filled in a form for the B household. A crosscheck of the forms from B later at the CMU in Darwin revealed that this family had been double-counted: at A as residents of A11, and at B as residents of B.

Such an ‘as if’ count is not, however, the same as counting people at a place. Instead, it represents people’s idea of who would have been there had the count really happened there. In such circumstances, certain categories of people tend to be forgotten—most significantly children. I was able to satisfy myself that the children of at least one couple failed to gain a mention in one of these ‘as if’ counts—I do not know whether they were counted elsewhere.