Campbell v Northern Territory – The Lingering Uncertainty over Comparators and Comparisons in the Racial Discrimination Act
Alan Zheng
23.03.2022
Note: Aboriginal and Torres Strait Islander readers are advised this article contains the names of people who have passed away.
Over the past decade, the Racial Discrimination Act 1975 (Cth) (RDA) has found its place as a remedial avenue for First Nations persons and communities seeking justice against unlawful racial discrimination. This has become particularly prominent in cases framed under s 9(1)’s broad prohibition against:
[A]ct[s] involving a distinction, exclusion, restriction or preference based on race, colour, descent or national or ethnic origin which has the purpose or effect of nullifying or impairing the recognition, enjoyment or exercise, on an equal footing, of any human right or fundamental freedom in the political, economic, social, cultural or any other field of public life.
The finding of unlawful discrimination in Wotton v Queensland (No 5) [2016] FCA 1457 (Palm Island Class Action) resulted in an award of $30 million to class members. Wotton has been cited by Aboriginal and Torres Strait Islander people as a rare court decision that delivered ‘justice’ (Allison 2019, pp 241-242). The Federal Court’s decision in Wotton joined the ranks of Baird v Queensland [2006] FCAFC 162, a stolen wages case where complainants succeeded on appeal and were awarded damages of between $17,000 and $85,000 as well as an apology.
More recently, there has been a surge in RDA class actions and associated settlements (see eg, Davison v Commissioner of Police, NSW Police Force [2021] FCA 1324; Dawson v Commonwealth [2021] FCA 1354; Cumaiyi v Northern Territory [2020] FCA 1299; Pearson v State of Queensland [2020] FCA 619; Sister Marie Brigid Arthur as Litigation Representative v Northern Territory [2020] FCA 215; Mununggurr v Channel Seven [2019] FCA 2188). There are also ongoing proceedings seeking to argue that the pension eligibility age for Aboriginal and Torres Strait Islander people is discriminatory, in effect, to Indigenous Australians who have lower average life expectancies.
At the same time, the RDA is also being used in novel ways. For the first time, there has been an RDA complaint in respect of police conduct preceding a death in custody in Constable Zachary Rolfe’s fatal shooting of Warlpiri man Kumanjayi Walker in Yuendumu in November 2019.
Following Rolfe’s recent acquittal, it is necessary to consider alternative remedial avenues to ensure police accountability. Significantly, an RDA claim enables a claimant to frame the precise ‘acts’ which are said to constitute discrimination. These ‘acts’, in turn, are not viewed in isolation but inclusive of all consequences and surrounding circumstances ‘involved’ in the act on an objective assessment (Wotton at [559]-[560]). This makes s 9(1) a particularly flexible vehicle for exploring a broader array of conduct including matters otherwise unsuitable for criminal trial such as alternative policing approaches in the lead up to the shooting. Moreover, the RDA’s well-established interface with the denial of human rights would allow the court to properly consider key issues in the case including the fact that Walker was attending sorry business before the fatal shooting. It would also enable scrutiny of whether there was an arbitrary deprivation of life which is inconsistent with the right to life under article 6 of the International Covenant on Civil and Political Rights. Although an RDA complaint may not provide the expression of societal opprobrium which a criminal conviction secures, it may be that it can do something to vindicate the family and community’s experience of harm as that of a racist institution.
Putting to one side the remedial effectiveness of the RDA compared to other remedial avenues in tort and breaches of statutory duty, it is unmistakable that the RDA is increasingly being considered and deployed. Against this backdrop of renewed attention on the RDA—and s 9(1)—in particular, it has become clear that complainants have been thwarted by lingering uncertainty over the role of a comparator in the RDA, as the recent decision in Campbell v Northern Territory (No 3) [2021] FCA 1089 demonstrates.
The role of comparators in discrimination law
The comparator is the predominant means of determining causation in Australian discrimination law. The comparator is an actual or hypothetical person who does not have a particular protected attribute (sex, race, disability). Courts construct the comparator and are required to place them in similar factual circumstances to the complainant. Then, it is necessary to ask whether the complainant was treated less favourably than the comparator in those circumstances. If the answer is yes, that generally establishes the relevant discrimination (subject to defences). The comparative formula is found in the formulation of ‘less favourably’ tests in the Sex Discrimination Act 1984 (Cth) (s 5A), the Disability Discrimination Act 1992 (Cth) ( s 5) and the Age Discrimination Act 2004 (Cth) (s 14).
Comparators have attracted stinging criticism. It was described as akin to conceptual ‘shackles’ in the joint dissent of Kirby and McHugh JJ in Purvis v New South Wales (2003) 217 CLR 92. Indeed, dicta from the House of Lords has suggested that comparators should be relegated to the status of analytical tools for reaching a conclusion of discrimination rather than necessary preconditions (Lord Scott and Lord Nicholls’ speeches in Shamoon v Chief Constable). Numerous bodies have suggested that the comparator test should be simplified at the federal and state levels. The NSW Law Reform Commission in its 1999 review of the Anti-Discrimination Act 1977 (NSW) identified ‘widespread dissatisfaction … conceptual difficulties … artificiality and resulting complexity’. More recently, the Australian Human Rights Commission noted, ‘the application of the comparator test … has presented significant difficulties, including complexity in interpretation and uncertainty of outcome’ (p 279).
On the comparator issue, however, the RDA is nonpareil. Other than the comparator tests in federal discrimination law and the ‘detriment’ model found in the ACT’s Discrimination Act 1991 (s 8) and followed in Victoria’s Equal Opportunity Act 2010 (s 8), s 9(1) of the RDA has been home to a third model which has not received significant scholarly attention.
Although some have pointed to s 9(1)’s unique feature of assessing discrimination according to its effects on the enjoyment of human rights — that is the ‘effect’ of s 9(1) discrimination—the approach to establishing discrimination is far less clear. When the RDA was introduced in 1975, Commonwealth Solicitor-General Maurice Byers even noted s 9(1)’s generality, and recommended its removal from the final bill (p 246).
To prove that an act involving a distinction, exclusion, restriction or preference was ‘based on race’, some courts have magically read the ‘less favourably’ comparator formula into the text of s 9(1). In Vata-Meyer v Commonwealth [2014] FCCA 463 (not modified on appeal) and Phillip v State of New South Wales [2011] FMCA 308, judges have found the need to prove, at least, a hypothetical comparator. Other cases, like Hagan v Trustees of the Toowomba Sports Ground Trust [2000] FCA 1615, include judicial acknowledgments of the need to establish that the claimant was treated ‘less favourably from other members of the community’.
In Wotton, Mortimer J identified that ‘[a]lthough … s 9 does not require a comparator in the way that some federal and state anti-discrimination statutes do, that is not to gainsay the need for a comparison’ (at [539], emphasis added). Accordingly, s 9(1) ‘requires some comparative analysis of the circumstances, albeit one that is not constrained by the complex comparator structure found in other federal anti-discrimination statutes’ (at [540]).
In finding unlawful discrimination, Mortimer J appeared to use more flexible forms of comparative reasoning such as comparing what was required by police policies in responding to local protests against what was actually done in order to isolate and identify a racial basis for particular conduct. For instance, Mortimer J notes (at [1061]):
From the evidence before me, while control of unlawful behaviour was an important function for [Queensland Police Service] officers on Palm Island, as in any other community, the particular circumstances on Palm Island at this time required a more sophisticated and multi-faceted response. The [Operational Procedures Manual] contemplated such a response. It did not occur.’
This flexible and unstructured comparison can be found in a range of RDA cases.
In Baird, Allsop J compared what was paid and what should have been paid to Aboriginal workers. Similarly, in Qantas v Gama [2008] FCAFC 69, the plurality examined Mr Gama’s allegation that he had not been nominated for training courses, comparing this to Qantas’ policies for development and training. In Vata-Meyer v Commonwealth, the Full Court compared Mr Lee’s words, ‘black babies’, with workplace standards, including his cultural competency training about First Nations issues.
To illustrate how this unstructured comparison could be applied in another case, let’s consider Rolfe. Rolfe’s text messages were recently published after suppression orders were lifted. In those messages, Rolfe described Alice Springs as ‘like the Wild West’ and his perception that his role in the Immediate Response Team was ‘a sweet gig, just get to do cowboy stuff with no rules’. Under unstructured comparison, such text messages might be compared to applicable police policies about culturally competent policing approaches in First Nations communities. If the text messages fell short of such policies, this derogation—along with the fact that it provides an insight into Rolfe’s mind and evinces stereotypes of First Nations people, ‘lifestyles, attitudes and tendencies’ as lawless and requiring a ‘law and order’ approach to policing—could support an inference of racial basis. They might also suggest Rolfe believed he could take whatever action he deemed appropriate in the course of policing in the community, and there was no ability or entitlement in the local community to protest or dispute that action.
Based on the circumstances of Wotton, Mortimer J drew a series of inferences such as this, which ultimately supported her finding of a racial distinction (at [1093]):
[Queensland Police] were not respectful, they were dismissive. They were not consultative, they dictated. They did not recognise grief and frustration, they over-policed. They behaved in this way because this was an Aboriginal community where the police had a sense of impunity and separateness, and an intention to exert control. The distinction was based on race.
See also Wotton (at [1029], [1441], [1475], [1711], [1748]).
It is important to caveat this analysis by noting that Wotton concerned police conduct following a death in custody and did not plead there was a racial basis in the lead-up to the death in custody at all. Any attitude of impunity brought to bear in the Northern Territory Police’s approach to Walker is a serious matter which would require a court to be mindful of the evidential standard in Briginshaw.
The decision in Campbell
Marley Campbell, a Kngwarraye and Aranda man, commenced proceedings concerning conduct which occurred during his period in custody at the Alice Springs Youth Detention Centre (ASYDC) and Don Dale Youth Detention Centre (Don Dale). The proceedings involved three causes of action, all of which were dismissed by White J in the Federal Court.
First, that particular incidents during custody constituted the torts of false imprisonment, battery, assault or negligence. Second, that the conduct of officers in the detention centres constituted breaches of the Youth Justice Act 2005 (NT) and the Youth Justice Regulations 2006 (NT). Third, and relevantly for this post, that certain conduct constituted discrimination in contravention of s 9(1) of the RDA.
The pleaded conduct included the Territory’s decision to transfer Campbell to Don Dale and the refusal to transfer him back to the ASYDC. In turn, it was argued these acts constituted differential treatment because they ‘failed to promote or protect his interests in maintaining connection with his family, community and country’ and accordingly denied his enjoyment of rights like the right to equal participation in cultural activities under art 5(e)(vi) of the Convention on the Elimination of All Forms of Racial Discrimination (CERD).
On this conduct, White J noted ‘it is not easy to see that the refusal to re-transfer the applicant back to ASYDC involved any form of differential treatment of him in comparison with other detainees’ (emphasis added) and ultimately ruled that the applicant had failed to establish that he was treated differently ‘based on’ his race (at [723]-[724]).
The applicant argued that ‘[n]o Non-Aboriginal detainees were transferred’. However, this was promptly rejected because there were no non-Aboriginal detainees at ASYDC at the time and the submission was not repeated by the applicant’s counsel in closing.
The difficulties of the comparator (whether real or hypothetical) are encapsulated by the applicant’s difficulty with establishing a comparator. One of the witnesses—the applicant’s case manager at Don Dale—said (at [733]):
I did not ever see a transfer decision made on the basis of a detainee’s race. Almost all of the detainees were Indigenous. I can only recall one or two non Indigenous detainees.
Two options arose at this point, though neither were taken further by counsel. First, even if we suppose that a comparator was doctrinally required by s 9(1), it could have instead been argued by the applicant that a hypothetical comparator, such as a non-Aboriginal white detainee, would not have been transferred to Don Dale. This could have drawn upon the literature surrounding systemic racism and historical patterns in the treatment of Aboriginal detainees in the Northern Territory compared to non-Aboriginal detainees in the Northern Territory and elsewhere.
However, it is doubtful that a hypothetical comparator is required by s 9(1).
In Baird, Allsop J (with whom Spender J and Edmonds J agreed) held ‘the need for a comparator or comparison for a distinction to exist are not necessary elements of s 9(1)’ and ‘one does not read into s 9(1) the need to demonstrate a direct comparator or comparison’ (at [63]). Some have argued (for example, in Vata-Meyer at first instance at [59]) that Allsop J’s reference to a ‘direct comparator or comparison’ should be understood as confined to an actual comparator such that a hypothetical comparator may still be required by s 9(1). However, this reading is flawed when considered alongside Allsop J’s other remarks. For instance, Allsop J also noted that ‘those suffering the disadvantage of discrimination may find themselves in circumstances quite unlike others more fortunate than they’ (at [63]). The very fact that circumstances would be ‘unlike’ militates against the viability of even hypothetical comparators.
The other option, which is the better view and consistent with the unique text of s 9(1), is that neither an actual nor hypothetical comparator is required by s 9(1). Rather, the comparator is one of a range of analytical devices which can be utilised to establish a racial basis for differential treatment. Another analytical device may be found in the unstructured comparison applied in Wotton. Justice White appears to leave scope for this by presupposing other analytical methods of identifying differential treatment after the applicant’s comparator arguments collapsed, noting (at [737]):
[T]he applicant’s counsel did not point to any other circumstance which might have indicated that the decision not to return the applicant to ASYDC could have been racially based.
Whether s 9(1) requires a comparator like other anti-discrimination statutes, a flexible comparison as in Wotton, or something else is still the source of some uncertainty for applicants.
Future directions for comparative reasoning in the RDA
Discrimination law is, as ever, in a state of flux. Even its name — ‘anti-discrimination law’ or ‘discrimination law’ is not easy to pin down in light of the evolution from negative prohibitions to positive duties in some jurisdictions such as the ACT. Moreover, discrimination law’s purposes, as Dr Alice Taylor writes, are ‘conflicting’.
As Taylor notes, some courts have treated discrimination law as having the purpose of formal equality, an approach which justifies the supremacy of the comparator device. On the other hand, s 9(1) is a paradigmatic example of the ‘liberty approach to anti-discrimination law’ (p 193) because it defines the scope of discrimination using human rights. Consequently, the wrong at the heart of s 9(1) is a denial of rights rather than necessarily using race as a criterion. This conclusion is consistent with the absence of any comparator device in s 9(1). As Drummond J noted in Ebber v Human Rights and Equal Opportunity Commission [1995] FCA 183 (at p 475):
[s 9(1)] … focuses on protecting from impairment by acts of racial discrimination certain fundamental rights which each individual has; it does not purport to aim at achieving equality of treatment in every respect of individuals of disparate racial and national backgrounds.
Uncertainty is compounded by the RDA’s status as a relatively recent addition to the Australian rights protection armoury (Taylor, p 190) and, more critically, the AHRC’s mandated conciliation process, which have contributed to a paucity of case law, particularly at appellate levels.
In December 2021, the Australian Human Rights Commission’s Report ‘Free & Equal: A Reform Agenda for Federal Discrimination Laws’ identified that the ‘development of [discrimination] case law and precedent is somewhat sporadic’, resulting in a ‘remarkably small body of discrimination case law’ (p 230). Other calls have been made to modernise discrimination law regimes at the state level. To top it off, the proposed Religious Discrimination Bill attracted significant public debate.
Lingering uncertainty over the analytical devices in s 9(1) of the RDA will further dissuade already vulnerable and resource-poor claimants.
Following Campbell, it is necessary to clarify the state of the comparator in s 9(1). If there is no comparator and there ought not be, what is in its place? As this post notes, one possible option is to unpack the flexible comparison in Wotton. As s 9(1) comes of age, the answer to this question is more important than ever.
—
Alan Zheng is a practical legal trainee at the Public Interest Advocacy Centre and completed his LLB thesis on racial discrimination law at the University of Sydney. The views in this post should be attributed to the author alone.
Suggested citation: Alan Zheng, ‘Campbell v Northern Territory - The Lingering Uncertainty over Comparators and Comparisons in the Racial Discrimination Act’ on AUSPUBLAW (23 March 2022 <https://www.auspublaw.org/blog/2022/03/campbell-v-northern-territory-the-lingering-uncertainty-over-comparators-and-comparisons-in-the-racial-discrimination-act>.