A range of counts and estimates are available for the Indigenous and non-Indigenous populations of the East Kimberley. For example, the ABS provides a de facto count of people who were deemed to be present in the region on census night (7 August 2001 at the most recent census). Then, there is a de jure count of people across Australia who indicated that one of the two East Kimberley SLAs was their usual place of residence on census night. These two counts are also available for Indigenous Areas found within the Northern East Kimberley. Finally, in recognition of the fact that the census fails to count some people, the ABS develops post-censual estimates of the ‘true’ population by augmenting their SLA ‘usual residence’ counts according to an estimate of those missed (net undercount), as well as other demographic adjustments. This produces an Estimated Resident Population (ERP), which in effect becomes the official population of each SLA for the purposes of electoral representation and financial distributions.

It should be emphasised that official ERPs are only available at the SLA-level. Estimates of the population within the study region, and its constituent parts, therefore have to be derived by ratio allocation of the overall ERP for the two East Kimberley SLAs combined. In terms of the present exercise, and in regard to the quite separate structural circumstances of most Indigenous people in the region, it is helpful that separate calculations of Indigenous and non-Indigenous ERPs have also been made by the ABS (since 1996) using differential undercount rates, and by distributing (pro rata) those usual residents who did not answer the ethnicity question on the census form.

One difficulty with the measurement of demographic change in the region since operations at Argyle commenced more than 20 years ago is the fact that the ABS has altered its statistical geography in this area several times during this period. Although the two Local Government Areas (LGAs, more recently described as Statistical Local Areas—SLAs) of Wyndham-East Kimberley and Halls Creek have remained intact, many collection district (CD) boundaries have changed substantially and the number of CDs has also varied with far more in existence now than 20 years ago. Additional geography was introduced in 1996 with the creation of IAs, and their sub-category Indigenous Locations (ABS 1998a), for the purposes of disseminating Indigenous community profile data (including data for the non-Indigenous population). As shown in Figure 1.1, this provides for reasonably fine-grained analysis down to the level of most communities of interest, with the main exception of Mandangala which forms part of the Lake Argyle IA.

Thus, in some cases, Indigenous (and non-Indigenous) profile data are available for specific communities and towns (viz Kalumburu, Oombulgurri, Wyndham, Kununurra, Woolah, Warmun and Halls Creek), whereas other smaller communities (such as Mandangala and all outstation communities) are subsumed as part of larger geographic units (Wyndham-East Kimberley (S) west, Lake Argyle, and Halls Creek (S) North) (ABS 2002a). However, since 1992, the ABS has endeavoured to gather information on estimated population numbers for all discrete Indigenous communities^[1], no matter how small, via the Community Housing and Infrastructure Needs Survey (CHINS). A prototype CHINS was developed on a state-by-state basis in 1992, followed by more nationally integrated surveys in 1999 and 2001.

It should also be noted that alongside these ABS geographic units, various regional service providers construct population lists of clients drawn from often quite different catchment areas. In Halls Creek, for example, the Yuri Yangi Medical Service caters for the town of Halls Creek as well as many outlying settlements, most of which lie to the south and beyond the region of interest. In Kununurra, the Ord Valley Aboriginal Medical Service is utilised by individuals from as far as Kalumburu, as well as across the Northern Territory border. Schools tend to have highly localised catchments for primary level students, but data for the high schools at Wyndham, Kununurra and Halls Creek are necessarily a composite of students from much wider areas. Participants in individual Community Development Employment Project (CDEP) schemes may also be spread among a number of communities.

Population size

At the 2001 Census, a total of 13,854 persons were counted by the ABS as present on census night (7 August) in the two East-Kimberley SLAs of Wyndham-East Kimberley and Halls Creek (Table 2.1). Of these, 5,249 indicated an Indigenous status in response to the census question on ethnicity, and 7,115 indicated non-Indigenous status. With as many as 1,490 individuals (11%) providing no response to this question, their Indigenous status remains indeterminate. Of the entire population counted within Australia on census night, a smaller total (10,757) nominated one of these East Kimberley SLAs as their usual place of residence.

Table 2.1. Indigenous and non-Indigenous census counts and post-censual estimates: East Kimberley region,^a 2001

	Indigenous	Non-Indigenous	Not Stated	Total
Census count (de facto)	5,249	7,115	1,490	13,854
Usual residence count (de jure)	5,226	4,224	1,307	10,757
Estimated usual residents (ERP)b	6,000	5,176	Pro rated	11,176
Notes: a. Wyndham-East Kimberley, and Halls Creek SLAs combined b. Preliminary estimates only Source: ABS customised tables

Also indicated in Table 2.1 is the fact that the number of Indigenous usual residents of the East Kimberley was slightly lower than those counted as present there on census night. However, the number of non-Indigenous usual residents counted was substantially lower (by 41%) than the place of enumeration count. On this evidence, almost half of the non-Indigenous population present in the East Kimberley at any given time is visiting the region from elsewhere in Australia, at least during the dry season when the census is conducted. Given this scale of visitation, plus the fact that these visitors to the region may have somewhat different characteristics from those resident in the region, any comparative analysis of regional population characteristics that informs social impact assessment should be conducted using usual residence data only.

At the time of writing only preliminary ERP figures for the East Kimberley were available and these are shown in the final row of Table 2.1. As noted earlier, these purport to represent ‘true’ levels of the Indigenous and non-Indigenous populations of the region. However, when interpreting these, it is important to note that ABS ERPs have been observed to differ from other (unofficial) population estimates generated by alternate means (Taylor and Bell 2001, 2003). It is also necessary to take into account those methodological tendencies within the special procedures adopted by the ABS in remote communities and urban town camps in Northern Australia that are likely to produce an undercount of Indigenous people (Martin and Taylor 1996; Sanders 2002). It is debatable whether the standard ERP methodology adequately compensates for these shortcomings (Taylor and Bell 2003).

The study region

As noted, ERP figures are only available at SLA-level. One way to develop similar estimates of the population resident in the study region, and in each of its constituent parts, is to divide up the East Kimberley Indigenous and non-Indigenous ERPs, as reported above, according to the observed pro rata share of the relevant Indigenous Areas. Results of this ratio allocation are shown in Table 2.2. Thus, the Indigenous usual residence count in Kununurra (755) represented 16.8% of the Indigenous usual residence count for the whole of the East Kimberley. This same percentage of the East Kimberley Indigenous ERP produces an Indigenous population estimate for Kununurra of 1,009. In turn, the equivalent non-Indigenous proportion is 56.5%, which produces an estimate of 2,923 for the non-Indigenous population of Kununurra. All told, then, the 2001 total ERP of Kununurra is calibrated at almost 3,932, one quarter of whom are Indigenous.

Applying this methodology throughout, the resident population of the Northern East Kimberley is estimated at 9,259 in 2001, almost half of which (4,317) is Indigenous. These derived estimates for the Indigenous and non-Indigenous populations are not inconsequential as they are 22% and 27% higher than the reported census usual residence counts respectively.

Table 2.2. Ratio allocation of the 2001 East Kimberley ERP to Indigenous Areas in the Northern East Kimberley

Indigenous Area	Indigenous % of E-Kim UR count	Non-Indigenous % of E-Kim UR resident	Derived Indigenous ERP	Derived Non-Indigenous ERP	Derived total ERP	Indigenous % of total derived ERP
Kununurra	16.8	56.5	1,009	2,923	3,932	25.7
Wyndham	10.3	7.2	617	374	991	62.3
Oombulgurri	3.2	0.3	193	15	208	92.8
Kalumburu	5.3	0.9	319	46	365	87.5
Lake Argyle	4.5	20.2	270	1,046	1,316	20.5
W/E-Kim (W)	2.8	2.0	169	102	271	62.2
Woolah	0.9	0.1	54	3	57	94.4
Warmun	5.4	0.6	324	31	355	91.2
Halls Creek	18.9	5.7	1,131	294	1,425	79.4
Halls Creek (N)	3.8	2.1	231	108	339	68.2
Study region	71.9	95.5	4,317	4,942	9,259	46.5
Note:UR = usual residence

A striking feature of these estimates is the fact that most areas of the Northern East Kimberley have overwhelmingly Aboriginal populations. The main exceptions are the Kununurra and Lake Argyle Indigenous Areas where the demographic influence of urban development and the Argyle mine are in evidence. The overall effect of this is to reduce the Aboriginal share of population in the Northern East Kimberley to less than 50%—clearly a misleading statistic that does not account for the demographic influence of Kununurra and the Argyle mine site. Also worth noting is the fact that 72% of the Indigenous population resident in the East Kimberley as a whole is located within this Northern region. This compares to 95% of the non-Indigenous population.