Previous Next

Indigenous Data Sovereignty


Data politics and Indigenous representation in Australian statistics

Maggie Walter


Accepting the philosophical premise that numbers exist, as per Quine (1948), is ontologically different to accepting that numbers have a fixed reality. This differential is the essence of the reality of numbers as they are applied to indigenous populations. In First World colonised nations such as Australia, Aotearoa/New Zealand, Canada and the United States, the question is not just ‘are these numbers real’, but also ‘how are these numbers deployed and whom do they serve’. The reality query is not of the numbers themselves but of what they purport to portray.

Numbers, configured as population or population sample data, are not neutral entities. Rather, social and population statistics are better understood as human artefacts, imbued with meaning. And, in their current configurations, the meanings reflected in statistics are primarily drawn from the dominant social norms, values and racial hierarchy of the society in which they are created. As such, in colonising nation-states, statistics applied to indigenous peoples have a raced reality that is perpetuated and normalised through their creation and re-creation (Walter 2010; Walter & Andersen 2013). The numerical format of these statistics and their seemingly neutral presentation, however, elide their social, cultural and racial dimensions. In a seemingly unbroken circle, dominant social norms, values and racial understandings determine statistical construction and interpretations, which then shape perceptions of data needs and purpose, which then determine statistical construction and interpretation, and so on. Just as important is that the accepted persona of statistics on indigenous people operates to conceal what is excluded: the culture, interests, perspectives and alternative narratives of those they purport to represent—indigenous peoples.

This chapter investigates how Australia’s racial terrain permeates statistics on Indigenous Australians. I examine the shape and context of these statistics as currently ‘done’ in Australia (Walter & Andersen 2013) and also the absences—how they are ‘not done’. Within this, I interrogate the construction and dissemination of the contemporary Australian statistical Indigene and its wider social and cultural contexts and consequences. The chapter also challenges researchers to consider how reversing the analytical lens to generate data conceptualised through an Indigenous methodological framework might alter the narrative, concepts, discourse and, ultimately, policy directions of Indigenous Australia.

Five-D data and the statistical Indigene

If you Google the term ‘Indigenous statistics’, the list that comes back in a millisecond is a depressingly predictable one. The first 10 entries are associated with eight different entities presented from 10 slightly different perspectives. But all focus in one way or another on statistical representations of the dire, and longstanding, socioeconomic and health inequities between Aboriginal and Torres Strait Islander peoples and non-Indigenous Australian people. I summarise these as the five ‘Ds’ of data on Indigenous people (5D data): disparity, deprivation, disadvantage, dysfunction and difference. For example, the Australian Human Rights Commission ( uses statistical data to highlight overall inequality between Indigenous Australians and the rest of the population; the Australian Bureau of Statistics ( entries look at homelessness and education disparities; the Australian Institute of Health and Welfare ( discusses the overrepresentation of Indigenous people in the numbers of deaths from preventable causes; takes a more original approach and uses the data to map out the depressing average Aboriginal Australian’s life; while looks at the direction of the ‘Closing the Gap’ policy and determines there is a long way to go. And so it goes. There is, seemingly, no shortage of data on, or data usage to compile portrayals of, Aboriginal and Torres Strait Islander inequality.

If you are interested in data on contemporary Aboriginal social phenomena that are not directly related to the five Ds, your search will likely be less productive. For example, Ting et al. (2015), in their examination of the division of household labour, found that not only did Aboriginal and Torres Strait Islander women do less housework per week than a non-Indigenous Australian-born sample, but also the division of labour was more egalitarian in Indigenous Australian households. The problem for the researchers was that the Aboriginal and Torres Strait Islander sample in their dataset—the Household Income and Labour Dynamics in Australia (HILDA) survey—is small and groups Aboriginal and Torres Strait Islander households into one base category. The authors concede, based on these limitations, that despite their tantalising findings, their results cannot reliably tell us anything about how Aboriginal and Torres Strait Islander households do domestic labour.

This seemingly minor issue exposes the positionality of Aboriginal and Torres Strait Islander people within the statistical terrain of our contemporary nation-state. The problem is that there is a plethora of easily accessible 5D data. Attempting to move outside this trope of the statistical Indigene is to find yourself in a data desert. There are no existing datasets available for researchers wanting to further investigate their findings on household division of labour. The Longitudinal Study of Indigenous Children has some questions on household functioning in relation to children but none about how family life is lived. The National Aboriginal and Torres Strait Islander Social Survey, as it is currently constructed, is focused almost completely on indicators of socioeconomic, lifestyle, health and neighbourhood non-wellbeing, and the Census of Population and Housing, the other major source of data on Indigenous Australians, contains data on homeownership and occupation but not household functioning.

The crucial point is that the nation-state’s data collection topic priorities for its Indigenous and non-Indigenous peoples differ dramatically. This critique is not to undermine the necessity of the continued collection of data on socioeconomic and demographic disparities; the deep-seated and whole-of-colonisation period presence of inequality in the life outcomes and chances for Aboriginal and Torres Strait Islander peoples marks the obvious importance of these. Rather, the critique is of the non-existence of other data for Indigenous people—the kinds of data that are regarded as critical to collect on the majority population.

These data absences raise critical questions. Why, for example, did the federal government initiators and funders of the (very expensive) HILDA survey project, and the research consortium that conducts the project, not feel it necessary to generate an Indigenous sample that was large enough to yield robust statistics regarding their separate circumstances? In the early 2000s, the very wide range of household, income and labour fields, including data on household division of labour, collected in the HILDA survey were considered so important by policymakers that a large-scale national longitudinal study was established to collect and collate data on them. Yet, it seems there is no similar urgency, or perhaps even interest, in gathering such data about Indigenous Australians. This question leads to a second. Why is understanding Aboriginal peoples through anything but the lens of a social problem seemingly un-thought of and perhaps unthinkable within our major statistical institutions? It is through the unravelling of these conundrums that the racialised politics of contemporary data collection in Australia can be understood.

5D data and the deficit data/problematic people correlation

Current Australian practices in regard to the collection of data on Indigenous people are the cloned descendants of the data imperatives of colonisation. In what I refer to as the deficit data/problematic people (DD/PP) correlation, processes of enumeration have long been used to correlate the highly observable societal Aboriginal and Torres Strait Islander inequality with the concept of racial unfitness. This situation is not unique to Australia. As Tuhiwai Smith (1999) argues, in an argument that resonates around the colonised indigenous world, numbers rationalise our dispossession, marginalisation and even our right to be indigenous. The heritage and ubiquity of these statistics, everywhere, allow the reality of the indigenous peoples they depict to go largely unchallenged in public and political discourse.

The DD/PP correlation’s basic premise is that racial inequality and racially aligned social and cultural differences are directly connected. Many Indigenous and other researchers would agree with some aspects of that premise. What they strongly disagree with is the direction of the relationship. In the DD/PP correlation, the problematic people are the ones who, through their behaviour and their choices, are ultimately responsible for their own inequality. The power of the DD/PP correlation is such that it still works in contemporary times as a mechanism for disenfranchising and dispossessing. Echoes of this discourse are clearly evident in the rationales of both the federal and Western Australian Governments for why Aboriginal communities should be closed rather than supported. The Premier of Western Australia, Colin Barnett, is cited as repeatedly drawing a direct link between the necessity of community closures in the Kimberley region and the problems of violence and suicide experienced in some (but certainly not all) of these communities (ABC 2014). Ditto for the Northern Territory Emergency Response, and the list could go on.

The concept of the DD/PP correlation fits within the theoretical frameworks aligned with the sociology of new racism. Predominantly emerging from the United States, theories of new racism attempt to explain how contemporary African American/white American relations have not changed substantially despite the fact that racism per se is now almost universally regarded as socially, culturally and politically unacceptable. Researchers such as Bobo (1997) and Kinder and Sears (1981) argue that the continuation of racism can be explained by the replacement of discredited ideas of racial biological inferiority with rationales of non-white cultural and moral inferiority. These moral and cultural racial differences, just like old-fashioned notions of biological inferiority, are then problematised as the cause of and explanation for socioeconomic disparity. Under this new reasoning, Bonilla-Silva (2010) argues that it is now possible for claims for non-white inferiority to be made simultaneously with claims of non-racism—or what he refers to as racism without racists. Bonilla-Silva takes these ideas further, arguing that the ‘new’ morally and culturally pejorative interpretation of racial differences is structurally, not only individually, situated, and that the embeddedness of these ideas in the institutions and functions of the state may be even more powerful than old-fashioned racism. Under the individual and systemic promulgation of this discourse, race-based inequality is undisturbed within an almost hegemonic argument that (individual) racism is an anachronism.

Race relations that emerge from colonising settler states, however, add a complexity to the black or brown/white binary of theories of new racism that are strong in the United States. In nations like Australia, the primary race relations locus is between the majority European population, especially the dominant Anglo-heritage group and its historical and contemporaneously dominant instruments of state, and the first peoples of Australia. The act and practice of colonisation, historically and through its current day realities, saturate this relationship. It is colonisation that pervasively frames Australian racial/social hierarchies. In turn, these hierarchies are supported and rationalised by racialised discourses that circulate through the dominant society, defining and positioning the Indigenous peoples they have dispossessed and from whose lands and resources the now-settler nations draw their wealth and identity (Walter 2014). As I have argued elsewhere (Walter 2010; Walter & Andersen 2013), these discourses draw on the projected 5D data depictions of Aboriginal and Torres Strait Islander people as their evidentiary base. Statistical portrayals of Aboriginal and Torres Strait Islander people sit at the centre of how they are understood by the dominant settler society. They also frame the lived realities and the socially, politically and culturally framed understandings of the Australian nation-state’s relationship with ‘its’ Indigenous population.

This racialised ‘politics of the data’, therefore, has powerful consequences in the determination, and practice, of the nation-state/Indigenous population relationship. In the absence of other portrayals, stereotype-enhancing data pictures of Aboriginal ‘deficits’ and ‘inadequacies’ are all the more glaringly visible. 5D data provide an infinitely variable circular rationale for Aboriginal and Torres Strait Islander inequality, to the convenient exclusion of other less palatable explanations. More insidiously, they provide a virtuous veil to draw over the use and misuse of the power of the nation-state in its ongoing interactions with Australian Indigenous peoples—being cruel to be kind as it attempts to ‘help’ those who, obviously as per the 5D data, are incapable of helping themselves. The silencing of Indigenous voices within this discourse can also be justified through the presentation of the state/Indigenous relationship as akin to that between a stern but caring parent and a wayward child.

Academic research is not immune from the lure of the DD/PP correlation, which in turn adds a scholarly legitimation to the picture of Indigenous people as unfit and blameworthy. Weatherburn’s (2014) analysis of arrest, incarceration, socioeconomic and other statistics relating to Australian Indigenous people, for example, concludes that the primary reason for the heavy overrepresentation of Indigenous people in incarceration is widespread criminality among Australian Indigenous peoples. The growth in this overrepresentation, he argues, can be explained by the change in the relative rates of Indigenous and non-Indigenous involvement in serious crime. Uncritically reiterating the correlation mantra that the cause and the remedy for inequality (in this case, the over-incarceration of Indigenous people) can be found within those people themselves, Weatherburn posits data on poor parenting, poor school performance, early school leaving, unemployment and drug and alcohol abuse as the social correlations of offending and, therefore, its causes. But these phenomena are not social facts in and of themselves; they do not just exist. They are the predictable outcomes of longstanding social, cultural and racial inequality that is the signature product of colonising settler states (see, for example, Aotearoa/New Zealand, Canada and Hawai’i for a near mirror image of these inequalities and social outcome phenomena). Moreover, the classic ‘correlation equals causation’ error is made by Weatherburn, as it is in much of the DD/PP correlation interpretations. It is not that these things—that is, poverty, low educational attainment, unemployment and so on—‘cause’ offending, but that offending, overrepresentation and these indicators are part of the same landscape of inequality.

How 5D data construct the dominant discourse on indigeneity

The numerical form of statistics is a primary contributor to normalisation of the DD/PP correlation. Statistical analytical processes rely on the conversion of social and cultural phenomena, or measurements of social and cultural phenomena, to assigned numerical values. This transference allows examinations of relationships between objects to be represented in numerical form. It is here that the mental shift occurs. Indigenous statistics—these representations of phenomena such as relationship to the labour market, experience of high mortality and morbidity and housing positioning—in numerical form acquire within this conversion process a mantle of impartiality, if not full objectivity. Indigenous socio-structural realities are transmuted into neutral data points. Once social phenomena are perceived as ‘data’, it is an easy step to regard these data points as social facts—a dispassionate representation of Aboriginal and Torres Strait Islander reality.

Positioned as objective descriptors, these particular numbers operate now, as they have always done, as mechanisms of unequal power relations. They define who and what Indigenous people are. They also define what we cannot be. The Indigene remains the object, caught in a numbered bind, forever viewed through the straitjacketing lens of deficit (Walter & Andersen 2013). As such, relentless measurement, re-measurement and comparison of our invidious positioning within Australian society, to the exclusion of other investigations, reify and cement these 5D portrayals. The advent of big data, with its tendency to further distance lived social and cultural realities from their database embodiment, has only exacerbated the pejorative power of numbers to further marginalise and dispossess.

When the only Aborigine you know is the 5D statistical Aborigine

The DD/PP correlation’s grip on how the settler majority population, policymakers and statistical agencies ‘know’ Indigenous people is exacerbated by the intense disjuncture between black and white lives. Regardless of the fact that a predominantly urban Aboriginal population lives alongside the predominantly urban non-Aboriginal population, Aboriginal lives remain out of sight and mind—spatially, politically, socially and culturally absent from non-Indigenous Australia. The limited data available indicate very clearly that Indigenous and non-Indigenous Australians occupy different social and spatial realms; we live in different places even when living next to each other (Atkinson et al. 2010). Aboriginal people are largely invisible, as people and as peoples, in conceptions of everyday Australian life except as pejorative (statistically informed) stereotypes. This invisibility extends to the nation-state’s concept of itself and the business of state, except, reluctantly, as a seemingly unresolvable ‘equity issue’. Political and spatial marginalisation also insidiously support the perception of Indigenous peoples as remote outsiders, just another minority group, rather than Australia’s first nations. It is therefore unimportant, from within a majority Australian identity perspective, to know much about Aboriginal Australia.

This lack of knowledge fosters the building of non-Indigenous to Indigenous relations around pejorative stereotypes and this can be heard through the patter of almost thoughtless denigration and casual disrespect of Aboriginal and Torres Strait Islander people, culture and society that pervades our society’s conversations. As an Aboriginal person with pale skin, I hear this conversation everywhere—on public transport, at social gatherings and also in the university classroom. The widely held notions of Aboriginal responsibility for their own disparate socioeconomic position and a simultaneous but contradictory belief system about Aboriginal over-entitlement, which are doggedly resistant to the overwhelming evidence to the contrary, are repeated ad nauseum every day, between conversationalists who would (and do) take umbrage at any suggestion that such talk is founded on racism. Academic claims such as that Aboriginal culture is violent (see Weatherburn 2014) or that the deprived living conditions of many families in remote communities are culturally related (see Sutton 2005) just reflect and support this normalised terrain of disdain.

The almost complete absence of Aboriginal and Torres Strait Islander people within the life orbits of non-Indigenous Australia supports rather than restricts discourses of disregard. 5D data allow the non-Indigenous majority population to be assured in their knowledge of Aboriginal and Torres Strait Islander people regardless of the fact that they are unlikely to know any Aboriginal or Torres Strait Islander people. This is borne out by results from a battery of questions on attitudes to Aboriginal issues I asked in the nationally representative 2007 Australian Survey of Social Attitudes (AuSSA) (n = 2,699). Responses from the 34 survey participants who identified as Aboriginal or Torres Strait Islander were removed from the sample for analysis.

The answers revealed that a modest to bare majority disagreed that equal treatment is now a reality (58 per cent), that injustices are all in the past (51 per cent) and that Aboriginal identity goes beyond traditional lifestyles (57 per cent). A similar proportion agreed with Aboriginal cultural autonomy (53 per cent). The responses to the restorative justice items—agreement that extra government assistance because of ongoing disadvantage is warranted (45 per cent) and disagreement that Aboriginal land rights are unfair to other Australians (33 per cent)—did not find majority support.

Table 5.1 OLS regression variable description and coefficients predicting ‘attitudes to Aboriginal issues’ scores







18–34 years



35–49 years



50–64 years#


50–64 years










< Year 12



Year 12









Bachelor degree or above











Community/personal service worker









Machinery operator/driver







Capital city



Other urban





Respondent individual income










$78,000 +








Social proximity

Mix regularly with Aboriginal people on a day-to-day basis


Know Aboriginal people but do not mix regularly with them



Do not know any Aboriginal people personally



Adj. R2


* p < 0.05

** p < 0.01

*** p < 0.000

# Collinearity diagnostics do not indicate multicollinearity between ‘Education’ and ‘Occupation’ variables.

Source: Adapted from Walter (2012).

This first set of responses suggests there is awareness, albeit very incomplete, within broader non-Indigenous Australian society that racial inequality is a contemporary reality for Aboriginal and Torres Strait Islander people. Explaining why this awareness does not translate into majority support for remedial action can be explained by both the prevalence of 5D data and the associated DD/PP correlation. There is no need to redress inequality if you can rationalise the cause of that inequality not within wider society, but within the people who experience it. This supposition is supported by findings from the set of social proximity questions asked in the same AuSSA. The results find more than 90 per cent of respondents do not interact with Aboriginal people regularly and more than half do not know any Aboriginal people (Walter 2012). An ordinary least-squares (OLS) multiple regression with ‘Attitudes towards Aboriginal issues’ constructed from a single-scale variable from the six statements1 as the dependent variable finds that a number of sociodemographic factors influence non-Indigenous Australians’ attitudes. As displayed in Table 5.1, in line with the literature on the topic (see Goot & Watson 2001; Bean et al. 2001; Pedersen et al. 2004; Goot & Rowse 2007; Walter & Mooney 2007), gender, education level, residential location, ancestry and occupation are all independently associated with non-Indigenous Australians’ attitudes to Aboriginal issues (for a full discussion, see Walter 2012).

The association that is my focus here is the social proximity variable—more particularly, it is the lack of a statistically significant independent association between attitudes towards Aboriginal issues and the level of interaction by non-Indigenous respondents with Aboriginal people. My theoretical explanation for this result is that few non-Indigenous people know Aboriginal people and that knowing, or lack of knowing, is not associated with attitudes. Therefore, in light of this lack of social proximity, it must be that dominant public discourses about Aboriginal people are the major informer of non-Indigenous attitudes (Walter 2012). In terms of this chapter, 5D data are central to the construction of these discourses and the impact of 5D data on attitudes operates independently of Aboriginal and Torres Strait Islander people.

Disrupting the paradigm of Indigenous statistics

Let’s return to our original proposition that for Indigenous statistical data the question is not merely ‘are these numbers real’, but also ‘how are these numbers deployed’, ‘what do they purport to portray’ and ‘whom do they serve?’ Our earlier discussion has established that the numbers are deployed in very limited ways and, while they purport to portray Indigenous reality, what they actually portray is primarily a picture of Indigenous deficit, contrasted with the (normal) non-Indigenous majority. Such numbers reinforce dominant discourses about Indigenous peoples and, in so doing, they support the status quo of the subordinate Indigene position within the nation-state. Disrupting this limited and limiting paradigm therefore requires that the established tropes of data on Indigenous people be disturbed, ontologically and epistemologically.

Disturbing and disrupting the dominant paradigm of these data is more difficult than might be imagined. Their unquestioned default position is founded on embedded ways of seeing the world, and these world views are what shape their discursive reality. The primary problematic is that the Indigenous ways of seeing the world are not doing the shaping. Let me explain. As elaborated in my co-authored book Indigenous statistics (Walter & Andersen 2013), the theoretical frame of social positioning within Indigenous statistics draws on the concepts of social space and habitus from the work of Pierre Bourdieu (1984). Bourdieu used the concept of an individual’s or group’s position in three-dimensional social space (consisting of social, cultural and economic capital positions) to explain how people from similar positions tend to share a similar world view.

This shared view, especially among groups with the highest levels of social, cultural and economic capital, leads to a ‘synthetic unity’ (Bourdieu 1984)—a presumption that their world view is the world view. For 5D data, the key change is to add race capital to Bourdieu’s group—a four-dimensional not three-dimensional social space. We argue that a similar positioning along the continuum of race, social, cultural and economic capital is a shared constitutive element of the world view of those who control the commissioning, analysis and interpretation of Indigenous data—a predominantly Euro-Australian and middle-class group. It is the world view of this group that shapes how Indigenous statistics are understood and ‘done’.

As cultural theorists Hofstede and Hofstede (2005) argue, similar groups of people are mentally programmed with ‘software of the mind’ to produce similar constructs, which they form into logical, affective and behavioural models. Thus, this shared habitus of the primary creators of data on Indigenous Australians and their lifelong positioning as Euro-Australian middle-class people shape (subconsciously mostly) the production of data on Indigenous Australians and their subsequent portrayal, thereby confining and/or prescribing how these data are ‘done’. As evidenced in the previous section, the majority of these controllers of such data, like the majority of the non-Indigenous population, are unlikely to personally know any Aboriginal people. Rather, the only Indigene they are likely to be familiar with is the portrayal drawn from 5D data. The contrasting (and distant) four-dimensional social space position of the object of the data (Indigenous peoples) reinforces the uncontested ‘synthetic unity’ (Bourdieu 1984) of dominant perspectives.

The central point is that dominant discourses of a society, not statistical methods, determine social data meanings. As Zuberi and Bonilla-Silva (2008) argue, claims of objective methodology allow dominant settler society questions to be perceived—largely unchallenged within the institutions and entities that pose them—as the only questions. The reality of these numerical data points, however, emerges not from mathematically supported computational techniques but from the social, racial and cultural standpoints of their creators. The power and the politics of the data are embedded in the ‘who’ of who has the power to make the assumptive determinations—to determine: what is the problematic, what it is that requires investigation, which objects to interrogate and which variables and variable relationships to test (see also Morphy, this volume). In the terrain of Indigenous Australian statistical data initiation, this ‘who’ among which these powers remain is most definitively not Aboriginal and Torres Strait Islander people.

Research constructed from statistics and data imagined from Indigenous ways of seeing the world will, by definition, change the terrain of Indigenous statistics. Yet it is important to stress that arguments around Indigenous methodological conceptual and practical distinctiveness are not to say: 1) that such statistics are in opposition to those emerging from the Western habitus, or 2) that differences from Western-framed statistics are what make Indigenous-framed statistics Indigenous. Neither are statistical techniques nor the ways of measuring per se what delineate 5D data from Indigenous-framed data. Rather, as argued in Walter and Andersen (2013), the Indigenous position in four-dimensional social space in a particular society makes apparent the gaps in current frameworks and in existing categories, concepts and conceptualisations of Indigenous data. In changing the ‘who’ of who has the power to make the assumptive determinations that shape data practices, the terrain of what is the problematic, what needs to be measured, how it is measured and very often who is doing the measuring is also changed (see also Lovett, this volume). Altering the paradigm of statistics on Indigenous people is critical if the statistical ‘recognition gap’ is to be addressed. As per Taylor (2008) and Kukutai and Walter (2015), the ‘recognition gap’ is the ongoing propensity for our official statistics agencies to misrecognise the social and cultural phenomena that are important to the wellbeing of Indigenous peoples. Expanding the ‘recognition space’ between Indigenous and non-Indigenous understandings allows us to speak back to the state in the language of statistical evidence that they both understand and culturally respect, reframing the narratives about us.

A case study

One way to disrupt established tropes surrounding Indigenous statistics is to reverse the presumed direction of the DD/PP correlation with place. In so doing, not only would Indigenous world views be incorporated into the assumptive determinations of what the problematic is and how it should be investigated, but also non-Indigenous Australians would become the sometimes difficult to comprehend ‘them’ and their social structural positioning would become the research object. To demonstrate this, I use a research example that takes the Indigenous perspective as its epistemic starting point. In so doing, it disrupts the trope of statistical production regarding Indigenous people and demonstrates an alternative Indigenous numbers paradigm. Yes, the numbers are real and their deployment bridges an ontological gap—providing a space for a discourse of Aboriginal perspectives on Indigenous and non-Indigenous social and cultural values, norms and life circumstances. It is this reality that these numbers purport to portray.

The research was conducted in 2014–16 by colleagues and myself. The project, ‘Telling it Like it Is’,2 was undertaken in partnership with Larrakia Nation, the organisation representing the traditional owners of the country where Darwin now stands and where Aboriginal people make up about 10 per cent of the total population. Our research rationale was that the unevenness of race relations has meant that Aboriginal people are rarely asked their views on Australian values, Australian society and their own place within it. This project’s aim was to redress this gap across multiple platforms. Initial results from the interviews of 40 respondents demonstrated a severe disconnect between Aboriginal and non-Aboriginal lives, lifestyles and values in Darwin. Respondents described how Indigenous and non-Indigenous people occupy different social worlds, with most social interaction being transactional rather than relational. Life disconnects were described in terms of the uncomfortableness of being Aboriginal in public spaces such as shopping centres, frequently feeling judged and feeling they did not belong. Value disconnects centred on what was perceived as the Western core value of material success versus central Aboriginal obligations of family and culture. Although respondents understood material benefits, they also saw the price many non-Indigenous people pay—stress, long working hours, with accumulation of material goods and career progression as the measures of personal success—as too high.

Results from the interview phase shaped the development of a stratified sample survey of over 400 Aboriginal people. Survey data were collected face to face by an Aboriginal survey team from the Darwin area. The results confirmed the qualitative findings as being representative of Aboriginal peoples’ views in the Darwin area. The survey data also revealed a deep lack of trust of the Euro-Australian–dominated institutions and governance bodies and a resentment of their refusal to recognise, in any meaningful way, Aboriginal, and particularly Larrakia, sovereignty of their own land. Disturbingly, for a majority of the respondents, regardless of their socioeconomic positioning, negative racialised encounters with non-Indigenous residents of Darwin remain an everyday, even normalised, experience (Walter 2016).


Alternative-paradigm Indigenous statistics cannot but disrupt the status quo of Indigenous data production: 5D data and the DD/PP correlation. But challenging long-established practices is likely to also disturb the ontological and epistemic security of those for whom the current way of creating such data is the norm. The alternative-paradigm results may also be hard to hear and potentially hard to understand for the wider non-Aboriginal audience. Nonetheless, such Indigenous-framed numbers are powerful and, by virtue of their framing of the ontological realities of Aboriginal life from an Aboriginal perspective, political. Most significantly—statistically significantly—the paradigm will reverse the hitherto one-way track of how Australia’s racial terrain permeates Indigenous statistics.


Atkinson R, Taylor E & Walter M (2010). Burying indigeneity: the spatial construction of reality and Aboriginal Australia. Social and Legal Studies 19(3):311–30.

Australian Broadcasting Corporation (ABC) (2014). Colin Barnett: closing WA’s remote Aboriginal communities will cause great distress but he has no choice. The World Today, ABC National Radio, 13 November 2014.

Bean C, Gow D & McAllister I (2001). Australian election study: user’s guide for the machine-readable data file: SSDA study no. 1048, Social Science Data Archive, The Australian National University, Canberra.

Bobo L (1997). Race, public opinion, and the social sphere. The Public Opinion Quarterly 61(1):1–15.

Bonilla-Silva E (2010). Racism without racists: colour-blind racism and the persistence of racial inequality in the United States, 3rd edn, Rowman & Littlefield, Lanham, Md.

Bourdieu P (1984). Distinction: a social critique of the judgment of taste, Routledge, London.

Goot M & Rowse T (2007). Divided nation: Indigenous affairs and the imagined public, Melbourne University Press, Melbourne.

Goot M & Watson I (2001). One Nation’s electoral support: where does it come from, what makes it different and how does it fit? Australian Journal of Politics and History 47(2):159–91.

Habibis D, Taylor P, Walter M & Elder C (2016). Repositioning the racial gaze: Aboriginal perspectives on race, race relations and governance. Social Inclusion 4(1):57–67, doi:10.17645/si.v4i1.492.

Hofstede G & Hofstede GJ (2005). Cultures and organizations: software of the mind, 2nd edn, McGraw-Hill, New York.

Kinder DR & Sears DO (1981). Prejudice and politics: symbolic racism versus racial threats to the good life. Journal of Personality and Social Psychology 40(3):414–31.

Kukutai T & Walter M (2015). Recognition and indigenising official statistics: reflections from Aotearoa New Zealand and Australia. Statistical Journal of the IAOS 31(2):317–26.

Pedersen A, Beven J, Walker I & Griffiths B (2004). Attitudes towards Indigenous Australians: the role of empathy and guilt. Journal of Community & Applied Social Psychology 14:233–59.

Quine WVO (1948). On what there is. Review of metaphysics, reprinted in his From a logical point of view (1953), Harvard University Press, Cambridge, Mass.

Sutton P (2005). The politicisation of disease and the disease of politicisation: causal theories and the Indigenous health differential, paper presented at 8th National Rural Health Conference, Central to health: sustaining wellbeing in remote and rural Australia, Alice Springs, NT, 10–13 March 2005.

Taylor J (2008). Indigenous peoples and indicators of wellbeing: Australian perspectives on United Nations global frameworks, Social Indicators Research 87(1):111–26.

Ting S, Perales F & Baxter J (2015). Gender, ethnicity and the division of household labour within heterosexual couples in Australia. Journal of Sociology (April):1–18.

Tuhiwai Smith L (1999). Decolonizing methodologies, research and indigenous peoples, Zed Books, London & New York.

Walter M (2006). Using the power of the data within Indigenous research practice. Australian Aboriginal Studies (2):27–34.

Walter M (2010). The politics of the data: how the statistical indigene is constructed. International Journal of Critical Indigenous Studies 3(2):45–56.

Walter M (2012). Keeping our distance: non-Indigenous/Aboriginal relations. In Pietsch J & Aarons H (eds), Australia: identity, fear and governance in the 21st century, ANU E Press, Canberra.

Walter M (2014). The race bind: denying Aboriginal rights in Australia. In Green J (ed.), Indivisible: indigenous human rights, Fernwood Publishing, Winnipeg.

Walter M (2016). Social exclusion/inclusion for urban Aboriginal and Torres Strait Islander people, Social Inclusion 4(1):68–76.

Walter M & Andersen C (2013). Indigenous statistics: a quantitative methodology, Left Coast Press, Walnut Creek, CA.

Walter M & Mooney G (2007). Employment and welfare. In Carson B, Dunbar T, Chenhall R & Bailie R (eds), Social determinants of Indigenous health, Allen & Unwin, Sydney.

Weatherburn D (2014). Arresting incarceration: pathways out of Indigenous imprisonment, Aboriginal Studies Press, Canberra.

Zuberi T & Bonilla-Silva E (2008). White logic, white methods: racism and methodology, Rowman & Littlefield, Lanham, Md.

1 Principle component analysis: Eigenvalue 2.70, 45 per cent of variance; Cronbach’s alpha 0.75.

2 ‘Telling it Like it Is’ is an Australian Research Council (ARC) funded research project conducted by Habibis, Walter and Elder. ARC Linkage Project 130100622. See also Habibis et al. (2016).

Previous Next