BY LISA BURTON CRAWFORD

In the recent case of ABT17 v Minister for Immigration and Border Protection, the High Court unanimously held that a decision of the Immigration Assessment Authority (‘IAA’) to refuse a protection visa to the applicant was invalid. But within this unanimous result lies profound disagreement about the way in which the legal limits of executive power are ascertained.

The IAA was reviewing an earlier decision of a delegate of the Minister via the ‘fast track’ system of review established by Part 7AA of the Migration Act 1958 (Cth). The delegate had conducted a face-to-face interview with the applicant, in which the applicant claimed to have been tortured by the Sri Lankan Army as a suspected member of the Liberation Tigers of Tamil Eelam. The delegate concluded that the applicant’s claims were plausible, despite his hesitating delivery and some inconsistencies between the story he told in the interview and his previous statements. But the IAA disagreed: after listening to an audio recording of the interview, it concluded that the applicant’s story was incredible. The IAA also found that changed circumstances in Sri Lanka meant that the applicant no longer faced a risk of persecution there.

The plurality of Kiefel CJ, Bell, Gageler and Keane JJ concluded that this was unreasonable. If the IAA had these doubts, it should have used the powers conferred on it by the Migration Act to obtain ‘new information’, by inviting the applicant to a second face-to-face interview in which it could assess his credibility for itself. Justices Nettle, Edelman and Gordon, all writing separately, also found that it was unreasonable for the IAA to overturn the delegate’s finding of credibility but for a different reason. These Judges did not think the error lay in failing to convene a second interview because they did not think (or in the case of Nettle J, seriously doubted) that inviting the applicant to re-tell the same story would amount to obtaining ‘new information’ within the meaning of Part 7AA, especially given that the purpose of that Part was to provide a quick system of review. But the delegate had been in a much better position than the IAA to assess the applicant’s credibility, having both heard and seen the applicant tell his story, and as matters stood there was no logical basis for the IAA to overturn the delegate’s findings in this regard.  

This disagreement indicates that the legal standard of unreasonableness is difficult to apply. But on the whole, the case endorses the more searching, ‘process focused’ ([122]) standard of unreasonableness endorsed by the High Court in Minister for Immigration and Citizenship v Li (2013). The case also raises further questions about the operation of the fast track regime and the balance it strikes between procedural fairness and efficiency. In particular, the case highlights the ‘informational gap’ ([13]) created by the administrative practice adopted by the IAA of reviewing referred decisions by listening to audio recordings of face-to-face interviews. It also demonstrates the challenges that can arise when administrative decision-makers are required to assess the credibility of asylum seekers, an issue previously highlighted by Australian scholars. But of particular interest, and the subject of this post, is the contribution the case makes to the ongoing debate about whether the administrative law concept of jurisdictional error incorporates a requirement of materiality.

The challenge of the immaterial error

I know what you might be thinking: surely that is the least interesting aspect of the case! But while it might sound unbearably technical, the question of how courts respond to immaterial errors of law is one of great practical and principled concern. It goes to the heart of how courts balance the competing demands of legality and the practical realities of the administrative state, as well how the legal limits of executive power are drawn.

In R (Cart) v Upper Tribunal (2011), Lady Hale of the UK Supreme Court stated that ‘a certain level of error is acceptable in a legal system which has so many demands upon its limited resources’ ([42]). Or as Lord Brown put it, ‘[t]he rule of law is weakened, not strengthened, if a disproportionate part of the courts’ resources is devoted to finding a very occasional grain of wheat on a threshing floor full of chaff’ ([107]). But where to ‘draw the line between those legal errors that [courts] can (or should) remedy and those that they cannot (or should not)’?

A more specific problem which has plagued Australian courts in recent years is: how should they respond to errors of law that for some reason or another make no difference? What if, for example, a decision-maker departs from the legal limits of her power, but only in a trivial or inconsequential way? Or, what if a decision-maker acts on two bases, one of which is legally flawed but the other of which is sound? That appeared to be the case in ABT17, because the IAA had refused the applicant’s visa because of both his lack of credibility and general information about circumstances in Sri Lanka.

In the 2018 decision of Hossain v Minister for Immigration and Border Protection, a majority of the High Court purported to solve the problem by redefining the concept of jurisdictional error. The distinction between jurisdictional and non-jurisdictional error remains pivotal in Australian law. Certain remedies (mandamus, prohibition, and in many instances, certiorari) are only available to remedy jurisdictional errors. Judicial review for jurisdictional error is constitutionally entrenched to an extent that non-jurisdictional error review is not. And a decision vitiated by jurisdictional error will generally be invalid for that reason, whereas a decision vitiated by non-jurisdictional error remains legally valid unless quashed by a court. In Hossain, it was held that jurisdictional error means a ‘material breach of an express or implied condition of the valid exercise of a decision-making power conferred by that Act’ ([31] Kiefel CJ, Gageler and Keane JJ, emphasis added). In other words, the ‘magnitude’ of the error or the ‘extent of non-compliance’ with the requirements of the Act will inform whether or not that error is jurisdictional in nature ([27]). As Edelman J explained, this means the limits of executive power are not ‘binary’:

A legislative condition need not be construed as (i) always depriving a decision maker of power, or (ii) never doing so, no matter how it is breached. The question is always one of construction of the legislation: which breaches of a provision does the legislation, either expressly or, more commonly, impliedly, treat as depriving the decision maker of power? ([67])

I have previously described this as ‘redefining’ the concept of jurisdictional error, and I think that is a fair description. There is some limited support for this approach in prior case law. But there are other cases which state that this approach is wrong. As Mortimer J observed, ‘[t]he reality is that it is difficult to discern a consistent approach [to immaterial errors] throughout the authorities’ ([69]).

In Hossain itself, Nettle J and Edelman J queried how the materiality criterion would apply in certain cases. In the subsequent case of Minister for Immigration and Border Protection v SZMTA (2019), the materiality criterion was endorsed by a majority of the High Court, but Nettle J and Gordon J delivered strong dissents. Their Honours dissented again in ABT17, with Gordon J providing a powerful critique of the practical implications and normative foundations of the Hossain approach. The debate seems set to continue, with the applicant in another case currently before the High Court – MZAPC v Minister for Immigration and Border Protection – urging the Court to reassess the materiality criterion.

In light of this ongoing controversy, this post outlines two of the key questions raised by the materiality criterion. First, how do courts apply it? Secondly, what is its basis or justification? The post draws throughout upon my lengthier analysis in the Public Law Review.

How do the courts apply the materiality criterion?

As Mortimer J (politely) said in the Federal Court’s decision in MZAPC, the application of the materiality criterion ‘presents some challenges’ ([34]). First, it is not clear whether it has universal application. In Hossain, the plurality appeared to redefine the concept of jurisdictional error as a material departure from the limits of executive power, which would mean that in every application for judicial review courts must assess the materiality of an error before they can determine its legal consequences. While Nettle J and Edelman J joined the majority in Hossain, they disagreed on this particular point, suggesting that the materiality criterion could (or should) not apply to all legal errors that an administrative decision-maker might make.

Soon after, I queried whether Hossain might be an example of an easy case making bad law. That case concerned a decision of the Administrative Appeals Tribunal to affirm a decision of the Minister’s delegate to deny Mr Hossain a partner visa. There were two relevant statutory criteria for this type of visa, and the Tribunal concluded that Hossain did not satisfy either — but it misinterpreted what one of these criteria required. It was relatively easy to say that this error made no difference because even if the Tribunal had correctly interpreted the first criterion, it would still have had to deny the visa because Hossain did not satisfy the second. But in many other cases, the dividing line between material and immaterial errors of law will be far harder to draw.

Consider, for example, the rules of procedural fairness. These require that a person who is to be affected by an exercise of power be given notice of the basis on which a decision will be made, and an opportunity to put their side of their case. What happens if a decision-maker fails to disclose some piece of adverse information to the applicant — how can we know what the decision-maker would have thought or done, if the applicant had been given an opportunity to respond to it? This is why courts have previously said that a deficiency in a notice or hearing cannot be gainsaid by claiming that it made no difference; rather, the content of those rules must be determined in advance.

More broadly, who bears the onus of proving that the error made no difference? In SZMTA, the plurality of Bell, Gageler and Keane JJ appeared to state that the applicant bears this onus ([4]). But in ABT17, Gordon J suggested that this subverts the protective purpose of judicial review. As her Honour explained it, this would mean that the law of judicial review now

[p]laces the onus on an individual to show why public power should be re-exercised, rather than protecting that individual from exercises of public power which are contrary to the law ([109], Nettle J agreeing at [72]).

This question of who bears the onus of proof appears set to be a major issue in the pending case of MZAPC.

In other circumstances, it is difficult to see how the materiality criterion would apply at all. Consider, for example, an error of jurisdictional fact. A jurisdictional fact is a factual pre-condition for the exercise of power. For example, if a statute says that ‘a Minister may only grant a permit to a person who is 18 years old or over’ then it would seem the Minister has no power to grant a permit to a person who is in fact 12. A jurisdictional fact either exists or does not exist and if does not, the decision-maker has no power. Except in Hossain-like cases where there are several, independent bases for exercising a power, it is not at all clear how one could say that the absence of a jurisdictional fact made no difference.

The legal error in question in ABT17 was unreasonableness. As this case demonstrates, most statutory conferrals of executive power will be read on the presumption that they must be exercised reasonably. But unreasonableness review does not permit a court to substitute its own judgment for that of the executive decision-maker. Much ink has been spilt articulating the particular point at which some deficiency in logic or justification tips the decision-maker over the bounds of reasonableness so as to make the decision unlawful. How then is a court to apply an additional criterion of materiality to this error of law? As Nettle J explained in ABT17, the concept of materiality is in a sense ‘bound up in the concept of legal unreasonableness’ ([72]).

The plurality judgment did little to clarify matters, because their Honours did not consider or apply the materiality criterion at all. This appears to be because the premise of the applicant’s argument in this regard was flawed. As explained above, it seemed the IAA articulated two bases for rejecting the visa: the applicant’s (lack of) credibility and the general country information. That would have raised a real question of materiality, because only the first of these was vitiated by unreasonableness. But on closer inspection, the two bases were in fact interwoven: the IAA concluded that a person of the applicant’s profile no longer faced a risk of persecution in Sri Lanka (see the plurality at [32]). For now, the point of law remains uncertain.

What is the justification for the materiality criterion?

The orthodox position in Australian law is that judicial review is no more and no less than the enforcement of the limits of executive power found in the statute. But the courts play a role in shaping those limits, because they read empowering statutes in light of a set of interpretive presumptions which they themselves created. Hossain effectively created a new presumption which (subject to the questions raised above) will now shape the limits of executive power. But what is its justification?

Justice Edelman tied it to parliamentary intent:

Just as it is unlikely to be concluded that Parliament intended to authorise an unreasonable exercise of power, so too it is unlikely to be an intention that the legislature is taken to have that a decision be rendered invalid by an immaterial error ([67]).

The plurality’s explanation is more novel. Their Honours stated:

The common law principles which inform the construction of statutes conferring decision-making authority reflect longstanding qualitative judgments about the appropriate limits of an exercise of administrative power to which a legislature can be taken to adhere in defining the bounds of such authority as it chooses to confer on a repository in the absence of affirmative indication of a legislative intention to the contrary …. Those common law principles are not derived by logic alone …. They are not so delicate or refined in their operation that sight is lost of the fact that ‘[d]ecision-making is a function of the real world’ ([28]).

In the Australian context, this statement is quite extraordinary. It admits that the presumptions of statutory interpretation deployed in the administrative law context reflect where the courts think the limits of executive power ought to lie.

While it may be preferable for courts to frankly admit when they are making qualitative judgments, the nature of the judgment being made in Hossain is far from clear. Why do their Honours think that Parliament would not intend immaterial errors to lead to invalidity? One could well argue to the contrary. As Nettle J said in SZMTA:

Certainly, courts make assumptions in relation to statutory power. … [But] Parliament cannot be taken to intend that a decision-maker need only comply with laws to the extent that failure to comply would not bring about a different result. Any such conception would be contrary to the notion, central to the conceptual foundations of judicial review, that everyone (including a decision-maker) is bound by the law ([90]).

Were the plurality in Hossain animated by the need to balance the competing demands of legality with the autonomy and efficiency of the administrative state? Did they make a principled assessment that the rule of law does not demand the correction of trivial legal errors? These are discussions which occur openly in other jurisdictions (as in the UK case of R(Cart), discussed above). But there is no such explanation offered in the plurality judgment in Hossain. It occupies an uneasy half-way house between formalism and functionalism: admitting that the interpretation of statutes is informed by judicially constructed values without articulating what those values are or how they were applied in the case at hand.

Justice Gordon criticised this aspect of Hossain in her dissenting judgment in ABT17. As her Honour explained it, the materiality criterion qualifies the limits of executive power. Those limits are no longer absolute, but conditional. Her Honour uses the example of procedural fairness to demonstrate. According to Hossain, decision-makers are no longer required to afford procedural fairness, but to afford procedural fairness ‘if (and only if) to do so would make a difference to the ultimate decision’ ([104]). Her Honour concluded that

if [the limits of executive power] are to be limited or qualified, that limit or qualification must be expressed in clear words by the legislature. Limits or qualifications cannot be imposed by way of “qualitative judgments”made by courts. ([106] (emphasis added).

Indeed, the upshot of the materiality criterion is that the legal limits of executive power are ambulatory. Those limits cannot be determined by reading the empowering statute alone, or even through the lens of the various other presumptions known to Australian law. As Gordon J explains in ABT17:

the new obligation is inherently uncertain. The obligation on a decision-maker is not set in advance by the statute. Rather, its content now depends upon the particular circumstances of the decision at issue. It is difficult (if not impossible) for a decision-maker to know in advance what level of [compliance] might have made a difference. The decision-maker needs to know in advance how to act in accordance with the law, but a materiality analysis is necessarily backwards-looking. Nor is the criterion of materiality any easier for the judiciary to apply: the criterion is akin to, or feels like, a form of merits review ([105]).

This has profound implications. As I explained in my contribution to this edited collection, it has the potential to destabilise other doctrinal positions taken by Australian courts. For example, Australian courts insist that they cannot adopt some of the public law doctrines applied in other jurisdictions — like a doctrine of legitimate expectations, or some form of deference to executive interpretations of statute law — in large part because the limits of executive power must be treated as fixed in advance by the terms of the empowering statute. Yet, the materiality criterion means that that is no longer, or at least not always, the case.

Where to from here?

The materiality criterion is problematic for many reasons. But the question of how a reviewing court should respond to an immaterial error is not an easy one to answer, and I am mindful that it is far easier to critique a doctrinal solution to an administrative law problem than it is to propose one’s own.

The alternative approach endorsed by Nettle J and Gordon J in ABT17 is to treat the materiality of an error as a matter which informs the exercise of judicial discretion to refuse relief. That solution has clear practical appeal. Yet, it seems merely to relocate, rather than avoid, the constitutional objections to the Hossain approach. If a jurisdictional error is one which — according to the statute, properly interpreted — takes a decision maker beyond the realms of legal authority, then on what basis could a court allow a decision vitiated by jurisdictional error to stand? Arguably, this amounts to a court authorising that which Parliament has forbidden, or at least abdicating its responsibility to remedy invalid exercises of powers which come before the court. Janina Boughey and I explore this issue here.

There may be a way of circumventing these constitutional objections by reconfiguring the way in which the legal limits of executive power are understood. In brief, we might be forced to recognise that the limits of valid executive action depend both upon what the statute has to say, and the remedial powers of the courts. Given that the discretion to refuse relief is long and well established, it would not be implausible to assert that powers are conferred on that assumption. But this is not a way of thinking that has been articulated in the case law and undoubtedly, it raises questions of its own. For now, it seems that the question of how to respond to immaterial errors — and the broader challenge of balancing legality with more pragmatic concerns — remains.

Lisa Burton Crawford is a Senior Lecturer at the University of New South Wales.

Suggested citation: Lisa Burton Crawford, ‘How should a court respond to a material error of law?’ on AUSPUBLAW (28 October 2020) <https://auspublaw.org/how-should-a-court-respond-to-an-immaterial-error-of-law>