A perspective from a jurisdiction without a doctrine of deference: Australia
4.10.23
This post is from a blog symposium on “Chevron Deference through the Lens of Comparative Law" edited by Susan Rose-Ackerman and Oren Tamir. The symposium brings together scholars from around the world discussing deference in their jurisdictions. The Introduction to the Symposium can be found in the following link which includes details about the various contributors and the jurisdictions covered: https://balkin.blogspot.com/2023/09/the-chevron-doctrine-through-lens-of.html
Among English-speaking common law jurisdictions Australia has been the most resistent to doctrines of deference in the administrative law context. It is often said that Australia’s High Court has rejected deference. In fact, a majority of the High Court has rejected Chevron deference, but only in obiter. Nevertheless, it is true that Australia has no ‘doctrine’ of deference, that Chevron deference specifically is generally thought to be inconsistent with the Australian conception of the separation of powers, and that the very mention of the word ‘deference’ seems to provoke the ire of some judges.
The High Court of Australia’s reasons for rejecting Chevron deference were its ‘undesirable consequences’ and that it involves an ‘abdication of judicial responsibility’. Australian public law rests on a different conception of judicial power and its separation from executive and legislative power than does US law. In the US, influenced by the theories of Thayer and others, judicial power is broad in scope but not exclusive in the realm of interpretation: courts need not supply all statutory meaning. By contrast Dicey’s conception of judicial power as narrower in scope, but exclusive in the domain of statutory interpretation has dominated and remains orthodox in Australia. Indeed, Australian administrative law may be described as ‘more diceyan than Dicey’ in its conception of judicial review of administrative action.
Under the Australian conception of the separation of powers interpreting legislation is an exclusively judicial function, as is finally and conclusively determining other questions of law. The separation of judicial power is constitutionally entrenched (albeit to differing degrees at the federal and state levels). The orthodox view is that allowing the executive to finally and conclusively interpret a statute provided that their interpretation is reasonable, as both the US’s Chevron doctrine and the Canadian approach to deference do, would permit the executive to exercise a judicial function and thus breach of the constitutionally entrenched separation of judicial power. One judge has even gone so far as to describe Chevron deference in the Australian context as antithetical to the rule of law.
The apparent incompatibility of Chevron deference with Australia’s constitutional framework seems to have caused judges and commentators to eschew the language of deference altogether. This is despite the fact that not all types of deference, in all contexts, are incompatible with Australia’s constitutional framework. For example, epistemic, or Skidmore deference (where weight is given to the executive’s interpretation, as opposed to doctrinal deference which gives the executive latitude to determine the meaning of legislation within the bounds of reasonableness) would be acceptable under Australia’s approach to the separation of powers and statutory interpretation. Furthermore, as I and the next Chief Justice of the High Court have (separately) pointed out, Australian courts do, in fact, defer to the executive on a range of questions of law, some of which arguably involve statutory interpretation. Australian courts use both epistemic and doctrinal forms of deference, but they prefer terms such as ‘judicial restraint’ or the giving of ‘weight’ rather than deference, due to the connotations of the term. While these may be examples of judicial deference to the executive, they do not form a coherent doctrine. It is often not clear why Australian courts chose to give weight or latitude to the executive’s view of the law in one case and not in another.
The absence of a doctrine of deference in Australia does not appear to have rendered the Australian administrative state incapable of performing its functions. Australian legislatures are not frequently having to re-draft legislation in response to court decisions which interpret administrative powers differently from the way they are understood by the executive (though there are some examples). It is not clear that Australian regulators’ inability to conclusively interpret their legislation within the boundaries of reasonableness means that their actions are more likely to be challenged, or are any less certain, compared with their US counterparts. Nor is there a general sentiment in Australia that courts become embroiled in policy-making as a result of their lack of deference to the executive on questions of statutory interpretation.
It could be that the absence of a doctrine of deference and the role of agencies in drafting legislation in Australia has led to differences in the drafting of statutes. For instance, Australian statutes might generally be more prescriptive in the powers they confer on regulators, and contain fewer ‘gaps’. I am not sufficiently familiar with US legislation to offer a considered analysis of whether that is true. But a cursory comparison of the equivalent piece of Australian federal legislation to the legislation at issue in Loper Bright Enterprises v Raimondo suggests it is at least not true in all cases. Just as the Magnuson-Stevens Fishery Conservation and Management Act PL 94-265 (US) confers broad powers to make a fisheries management plan which balances a range of objectives, the Fisheries Management Act 1991 (Cth) confers similarly broad powers on the Australian Fisheries Management Agency (AFMA) to make management plans which balance its objectives. The Australian Act confers similar powers on AFMA to require e-monitoring and observers as well as general powers to set conditions, without specifying who bears the costs of those requirements.
A more likely reason why the lack of deference is not a significant problem for the executive in Australia is the fact that it is a Westminster parliamentary system—with no separation between the executive and legislative branches. As the executive, by definition, has majority support in the lower house, the executive in Australia also has far greater control over the legislative agenda compared with the US. The vast majority of Bills originate within the executive, so that departments and agencies are almost always closely involved in the drafting and amendment of the laws they administer. While US agencies propose and draft some Bills and provide ‘technical drafting assistance’ for most others that they administer, Australian agencies have a far greater degree of input into their empowering legislation. In addition, Australian parliaments do not usually make the kinds of significant amendments that are common in the US. This would logically result in Australian laws better reflecting the administering agency’s views on how regulatory schemes ought to operate, and may therefore be less need for creative interpretations to achieve effective administration.
While the effectiveness of the Australian administrative state does not appear to have suffered due to the lack of a doctrine of deference, Crawford and I have (separately) argued that there are situations in which a deference doctrine may prove helpful in addressing existing and emerging issues in Australia’s system of administrative law. Deference in these circumstances would not, however, assist the administrative state by giving it more latitude to administer ambiguous laws. Rather, the circumstances in which deference would be useful would protect those who rely on an agency’s interpretation of the law (as Chevron deference also does). There are a range of situations in which the public relies on government agencies’ interpretations of statutes in order to understand their own rights and obligations. One is where the law is complex, and agencies provide guidance on what it means, such as taxation law and competition law. Another is when laws are changed frequently, making it difficult for the public to keep up with the current law, and agencies provide official or unofficial guidance as to relevant changes. The latter was a problem during the COVID-19 pandemic, with rules changing frequently, and the orders themselves often being difficult to locate. The public relied on the government’s announcements and websites to understand what the rules were at any given time. Another increasingly common example is when governments rely on automated systems in administering the law. As my colleagues, Professors Bennett Moses and Crawford and I have argued, automated systems require laws to be interpreted and, while those interpretations should be publicly available and accessible, they often are not, or are not available in a comprehensible format. Where a person relies on an agency’s interpretation in these circumstances, and that interpretation turns out to be incorrect, currently Australian law does not protect the person. A doctrine of deference to expert agencies interpretations in these limited situations would solve this problem, and be justified by the normative rationale of fairness. Though deference may not be the only, or even be the best adapted, legal solution to these problems.
While, in my view, the orthodox Australian view that deference is inconsistent with our constitutional and legal framework is overstated, I am not convinced that adopting a formal doctrine of deference in Australia is necessary or desirable beyond those limited situations set out above. I have two main reasons.
First, the intimate role that departments and agencies play in drafting legislation provides a strong argument against the adoption of Chevron deference (or any other variety of doctrinal deference) in Australia. I agree with Scalia J, Manning and others’ warnings that Auer deference provides an incentive for agencies to draft ambiguous regulations in order to enlarge their own powers (though I also note that incentive alone does not necessarily mean that this occurs). In a system where agencies are so closely involved in drafting primary legislation, the same concerns arise. Walker has argued that the significant ‘shadow’ role that agencies play in providing technical drafting assistance to congressional staffers in the US warrants a reconsideration of Chevron deference. The argument is stronger still in parliamentary systems like Australia. The level of control the executive has over the legislative agenda in parliamentary systems also means that there is less need for deference—if courts interpret a statutory provision in a way that is unworkable for the executive or doesn’t reflection parliamentary intention, legislation can usually be amended relatively quickly. I note that this argument does not preclude the adoption or utility of epistemic deference, and that the involvement of departments and agencies in the drafting of legislation gives them particular insight into the purposes of statutes which may be useful and compelling for courts in performing their interpretive function.
My second reason is more pragmatic. Australian law on when and how ‘weight’ should be given, or judicial restraint should be shown, to the executive on questions of law would benefit from greater clarity. This might be assisted by a greater willingness by courts to engage with deference doctrines overseas and draw parallels with Australia. However, the formal doctrines of deference in the US and Canada have proven difficult to apply. In Canada, questions of whether deference is due, how much deference ought to be given, and what exactly reasonableness requires now consume courts in judicial review of administrative action. The Canadian Supreme Court has dramatically changed its approach to deference at least three times since 1979. The US doctrine/s are even more complex, with at least three different kinds of deference. Some parts of the US doctrines of deference have never been clear, such as whether an agency should still be entitled to deference if it interprets a legislative provision to permit a change in policy under the statute, and whether Chevron deference applies to ‘soft’ law or informal adjudications as well as legislative rules and formal adjudications. Others have evolved and become controversial or difficult in their application, such as the legitimacy and scope of the major questions doctrine.
The existence of a formal ‘doctrine’ suggests that there is greater certainty and less room for judicial discretion in determining the boundary between judicial and executive power. But it is not clear this is accurate. Judges disagree about whether a statutory provision is ambiguous and so whether Chevron deference applies, on seemingly partisan lines. It is not clear that developing a formal doctrine of deference makes the law or the scope of statutory executive power clearer or less manipulable in application than the Australian position in which courts may decide to give weight to an agency’s view on certain questions of law, but offer no explanation of why they do or do not in each case.
Janina Boughey is an Associate Professor in the University of New South Wales Faculty of Law + Justice. Janina thanks Lisa Burton Crawford, Ros Dixon, Susan Rose-Ackerman and Oren Tamir for helpful discussions and feedback.
Suggested citation: Janina Boughey, ‘A perspective from a jurisdiction without a doctrine of deference: Australia’ (4 October 2023) <https://www.auspublaw.org/blog/2023/10/a-perspective-from-a-jurisdiction-without-a-doctrine-of-deference-australia>