Table of Contents
Almost a decade ago, when looking at the 1994 NATSIS, I was able to be very complimentary about what that survey could add to our understanding of housing outcomes for Indigenous Australians. That was largely because, in the housing area, the 1994 NATSIS had improved on previous census collections in two important ways: it had developed a better tenure breakdown, distinguishing between community and private rental dwellings, and it had developed an accessible and useful geographic breakdown into capital cities, other urban, and rural or remote areas. Using the 1994 NATSIS, I was able to show just how different the community and private rental tenures were for Indigenous households and, perhaps more importantly, just how different tenure profiles for these households were in rural and remote areas compared to the cities and other urban areas (Sanders 1996).
Today, it is a little more difficult to be glowingly complimentary about what the 2002 NATSISS can add to our understanding of housing outcomes for Indigenous Australians. This is not because the Indigenous survey processes have, in any way, gone backwards, but because other data collection processes, most notably the five-yearly censuses, have moved a long way forwards. Recently, I was able to undertake an analysis of housing outcomes for Indigenous households in five geographic categories of Australia ranging from major cities to very remote areas based entirely on publicly available data from the 2001 Census (Sanders 2005). I could show, even more clearly than I had been able to with the 1994 NATSISS, how different were the tenure profiles of remote areas compared to the cities and regional areas, and again how different were community and private rental. This makes it harder to argue that the 2002 NATSISS now adds so significantly to our understanding of Indigenous housing outcomes. However, it does still add some degree of understanding, and it can be used to confirm and reinforce census-based analysis.
In this paper I will reflect on the housing data from the 2002 NATSISS, drawing figures primarily from the RADL, but occasionally also quoting figures directly from ABS (2004c). The RADL contains two files, one giving characteristics of an estimated 165 674 dwellings containing Indigenous households, and the other giving characteristics of an estimated 282 205 Indigenous adults. When analysing housing characteristics, the predominant practice is to report figures relating to dwellings containing Indigenous households. However, it is also possible, and can be informative, to report by ‘persons’. In what follows, I often give both sets of figures, noting in the process that the figures by ‘persons’ tend to accentuate the differences between tenures and geographic areas. In the first section of the paper, I look at housing tenure profiles and household size by remoteness. In the second section of the paper, I look at a range of affordability and adequacy issues analysed primarily by tenure, but also by remoteness. In the third section of the paper, I look briefly at comparisons with non-Indigenous Australians and changes over time. I conclude by reiterating that almost all of this analysis can also nowadays be done from the five-yearly census data.
Many of the tables that follow divide Indigenous households or people into remote and non-remote areas. This two-fold geography is a sub-set of the ABS’s five-fold remoteness classification ranging from cities to very remote areas. The two-fold categorisation combines remote and very remote areas, on the one hand, and cities and inner and outer regional areas on the other. While not as refined as the five-fold categorisation available in the census, this two-fold categorisation is still robust enough to show major geographic differences in housing characteristics among Indigenous households and people.
Table 7.1 presents tenure profiles of dwellings containing Indigenous households in remote and non-remote areas in 2002. In the non-remote areas, three tenures—buying, government and private rental—each account for approximately a quarter of households, with ownership and other less significant tenures making up the last quarter of households. In the remote areas, by contrast, buying, private rental and ownership all fall away to single figure percentages, while government rental also falls away, but not quite so dramatically, to 17.1 per cent. What rises in remote areas in their place is one of the very minor tenures in non-remote areas—community rental. This tenure accounts for 49.7 per cent of dwellings containing Indigenous households in remote areas, but only 6.8 per cent in non-remote areas. Community rental is hence the dominant tenure in remote areas, while being fairly insignificant in non-remote areas. This contrast suggests just how different the housing tenure system is for Indigenous people in remote areas compared to elsewhere.
Table 7.1. Tenure of dwellings containing Indigenous households by remoteness, 2002
|
Remote |
Non-remote |
Total |
|
|
Number 000s |
29.2 |
136.5 |
165.7 |
|
Owned % |
5.9 |
11.8 |
10.8 |
|
Buying % |
7.7 |
23.2 |
20.5 |
|
Government rental % |
17.1 |
24.0 |
22.7 |
|
Community rental % |
49.7 |
6.8 |
14.4 |
|
Private rental % |
4.9 |
26.3 |
22.5 |
|
Employer and other rental % |
7.7 |
5.9 |
6.2 |
|
Other % |
7.0 |
2.0 |
2.9 |
Source: Customised cross-tabulations from the 2002 NATSISS RADL
Some explanation of community rental may be of assistance at this point for those not familiar with Indigenous housing policy. Community rental is the product of a thirty-year government effort to improve Indigenous housing conditions at government capital expense by providing between 500 and 1000 dwellings per year through grant funding to Indigenous community organisations. While these dwellings have been government-funded at the point of construction or capital expenditure, the ongoing ownership and management of them has been vested in Indigenous community organisations. These organisations have then been encouraged to charge income-related rents to cover recurrent housing costs, such as repairs and maintenance, and this has led to the label ‘community rental’. The closest cousin of community rental housing within the housing tenure system is government rental, which is also provided at government capital expense, and which also charges income-related rents in the hope of covering recurrent housing costs. However, government rental in most instances is restricted to cities and smaller urban areas. It does not often extend into remote, non-urban places, where significant numbers of Indigenous people live. In these areas, community rental generally takes over the role from government rental of meeting the housing needs of those who do not seem able to do so through market processes. Sometimes government rental and community rental are together referred to as social housing.
Table 7.2, which reports housing tenure by Indigenous people aged 15 years or over (hereafter referred to as adults), accentuates the tenure difference between remote and non-remote areas shown in Table 7.1. While the figures for non-remote areas differ by only one or two per cent for any tenure category from Table 7.1, for community rental in remote areas the figure increases by almost 14 per cent so that 63.6 per cent of Indigenous adults in these areas live in community rental dwellings. The dominance of community rental among Indigenous people in remote areas thus becomes all the more stark when reported by numbers of people, rather than households. The key to this shift is different household sizes in remote and non-remote areas.
Table 7.2. Housing tenure of Indigenous persons aged 15 years or over by remoteness, 2002
|
Remote |
Non-remote |
Total |
|
|
Number 000s |
77.1 |
205.1 |
282.2 |
|
Owned % |
3.8 |
12.0 |
9.8 |
|
Buying % |
4.9 |
22.4 |
17.6 |
|
Government rental % |
13.2 |
25.4 |
22.1 |
|
Community rental % |
63.6 |
8.5 |
23.5 |
|
Private rental % |
3.4 |
24.3 |
18.6 |
|
Employer and other rental % |
5.4 |
5.6 |
5.6 |
|
Other % |
5.8 |
1.9 |
3.0 |
Source: Customised cross-tabulations from the 2002 NATSISS RADL
Table 7.3 gives percentage distributions of household size by dwellings containing Indigenous households in remote and non-remote areas. It shows that whereas 21.2 per cent of these dwellings in remote areas have seven or more members of the household, only 5.4 per cent do in non-remote areas. Table 7.4 accentuates this difference by showing that 39.3 per cent of Indigenous adults in remote areas live in households of seven or more people, compared to 8.7 per cent in non-remote areas. Conversely, the proportions of Indigenous households containing four or fewer people, and of Indigenous people living in such households, are much higher in non-remote areas. This means tenure differences between remote and non-remote areas are accentuated when reported by Indigenous people, rather than households, due to larger Indigenous households in remote areas.
Table 7.3. Household size of dwelling containing Indigenous households by remoteness, 2002
|
Remote |
Non-remote |
Total |
|
|
Number 000s |
29.2 |
136.5 |
165.7 |
|
1–2 persons % |
29.2 |
39.6 |
37.8 |
|
3–4 persons % |
29.2 |
38.6 |
37.0 |
|
5–6 persons % |
20.3 |
16.3 |
17.0 |
|
7 or more persons % |
21.2 |
5.4 |
8.2 |
Source: Customised cross-tabulations from the 2002 NATSISS RADL
Table 7.4. Household size of Indigenous persons aged 15 years or over by remoteness, 2002
|
Remote |
Non-remote |
Total |
|
|
Number 000s |
77.1 |
205.1 |
282.2 |
|
1–2 persons % |
14.9 |
30.8 |
26.5 |
|
3–4 persons % |
24.6 |
40.7 |
36.3 |
|
5–6 persons % |
21.2 |
19.8 |
20.2 |
|
7 or more persons % |
39.3 |
8.7 |
17.1 |
Source: Customised cross-tabulations from the 2002 NATSISS RADL
Table 7.5 presents one final set of figures relating to housing tenure profiles and remoteness. These figures are for Torres Strait Islanders living in Torres Strait, the rest of Queensland and Australia (excluding Torres Strait). The figures in this table are drawn from ABS (2004c) rather than the RADL, and are based on the number of people responding to the 2002 NATSISS (not the number of households). To ensure that the estimates are reasonably reliable, the table is restricted to just two tenure categories: owned (including buying) and rental. Islanders in the Strait show the usual heavy reliance on rental of Indigenous people in remote areas and the concomitant low levels of home buying or ownership, at just 6.2 per cent, though ABS considers this statistically unreliable. [1] Islanders in the rest of Queensland show higher levels of buying and ownership and those in the rest of Australia, even higher still. We should note in passing here that were this third geographic category to be defined exclusively, as Australia excluding Queensland, (rather than overlapping with the second category and hence including Islanders in the rest of Queensland again), then clearly the level of home buying and ownership among this group of Islanders would be significantly higher again.
Table 7.5. Housing tenure of Torres Strait Islander persons aged 15 years and over by area, 2002
|
Torres Strait |
Rest of Queensland |
Australia (excluding Torres Strait) |
|
|
Number 000s |
3.6 |
13.0 |
26.2 |
|
Owned % |
6.2** |
20.9* |
34.9 |
|
Rental % |
84.6* |
79.1 |
63.8 |
Source: ABS (2004c: Table 23)