Although my brief was to observe the taking of the Indigenous part of the census, as already mentioned, I was aware that there were significant numbers of Indigenous people who lived in what is known as the ‘town site’. The mixing up of people in certain areas should be relatively easy for the census to manage—it simply requires some form of continuing communication between the CFO and the non-Indigenous equivalent—the Area Supervisor—and some dedicated effort in the planning stage. As stated in the 2006 Census IES: ‘For 2006, the role of the Census Field Officer in areas covered by Area Supervisors [mainstream managers] will be adjusted so that lines of responsibility and communication between these staff are clearer.’ Regrettably, that did not occur, and not only did the CFO never even meet either of the Area Supervisors, the latter did not even meet each other! No ‘Indigenous Assistants’ were engaged to help with either the enumeration of Fitzroy Crossing ‘town site’, or with the stations and tourists enumerator who was attempting to count people across a similarly huge area—one quite distinct from, yet overlapping in parts with, the CFO’s area. The Area Supervisor in Fitzroy Crossing—a non-Indigenous woman but ‘married-in’ and resident for some 10 years—was aware that there were about 30 Indigenous households being enumerated in the mainstream collection. The Indigenous Assistant she thought she had recruited did not help out in the end, because she was already burnt out from working on the census at the nearby community of Bayulu. While the Perth ABS staff had encouraged the Area Supervisor to find an Indigenous Assistant to help out, in the end she was unable to.

Beyond the town of Fitzroy Crossing itself, there was much anecdotal evidence that similar issues were arising, such as that already mentioned relating to the pastoral stations. If census staff on the ground are not communicating with each other, and have been given the sense that they are not to collect from ‘the other’, significant gaps can easily emerge, as was the case. Overlaps can occur too, as when the CFO and the Area Supervisor on occasions visited the same stations, unaware of what the other was doing—this despite the fact that they were regularly camping at the same caravan park in Fitzroy Crossing. Arguably, the approach of enumerating Indigenous people separately has the potential to greatly improve the accuracy of the count, however, the process needs to be tempered with commonsense, so that, in some instances, the Indigenous CIs might also collect data from non-Indigenous people. The census will, of course, never be absolutely effective in these areas, but it seems that there are some quite simple matters to be remedied that would see it being a lot more effective next time around.