Overcoming Graham: The s 75(v) constitutional guarantee and non-disclosure in migration and citizenship decisions

Isolde Daniell

30.07.2021

The Migration and Citizenship Legislation Amendment (Strengthening Information Provisions) Bill 2020 (Cth) (the Bill) was introduced to the Commonwealth House of Representatives on 10 December 2020. It has since been considered by the Senate Standing Committee for the Scrutiny of Bills and the Parliamentary Joint Committee on Human Rights, and referred to the Senate Legal and Constitutional Affairs Legislation Committee (which recommended on 10 March 2021 that it be passed) and the Parliamentary Joint Committee on Intelligence and Security (which is yet to report).

The Bill replaces, and substantially extends, the non-disclosure framework in s 503A of the Migration Act 1958 (Cth). That framework permits the Minister administering the Migration Act to make a decision to cancel or refuse a visa on character or conduct grounds in reliance on confidential information, suspicion or intelligence obtained from the Australian Federal Police (AFP), the Australian Security Intelligence Organisation (ASIO), or a foreign government, without providing that information to the affected individual. The Bill amends that framework to provide that the information must be disclosed to a Court exercising the constitutional review jurisdiction, but the Court must not order disclosure of the information to the individual unless it finds that a limited public interest test is met.

This post reflects on the specific question of the Bill’s constitutional validity. This is important given that the Bill’s express purpose is to overcome the decision in Graham v Minister for Immigration and Border Protection; Te Puia v Minister for Immigration and Border Protection [2017] HCA 33 (Graham). It is also of broader public importance given that the nature of the constitutionally entrenched minimum provision of judicial review (provided for in s 75(v) of the Constitution) is not yet fully articulated.

It is not yet clear whether the Bill will progress with or without amendment, although the Government has so far resisted making changes to the Bill and appears committed to the policy underlying it (see Scrutiny Digest 5/21).

Section 503A

The historical justification for the s 503A framework was that the Department of Immigration and Multicultural Affairs was in possession of information, suspicion and surveillance received from law enforcement and intelligence agencies which might not qualify for protection under legislation which prohibits the disclosure in proceedings of Commonwealth national security information (presently the National Security Information (Criminal and Civil Proceedings) Act 2004 (Cth)) or public interest immunity. The Department was reluctant to rely on this type of information in case it might come to be disclosed, in a case before a court, to the individual.

It would appear to follow that the necessity for the framework is to protect suspicion, or other intelligence, which has no connection to national security. For example, prior to 2017, the Minister administering the Migration Act had developed a practice of relying upon the framework to act upon suspicion of non-security domestic criminal offences, such as in the case of Alessio Vella, the president of the Rebels Motorcycle Club, whose permanent residency visa was cancelled on the basis of information, provided by law enforcement, which was partly protected under s 503A in 2015.

The Bill

The Bill prohibits (except in very limited circumstances) the disclosure of information to the person affected by a visa or citizenship decision where the information was communicated to the Department of Home Affairs by a gazetted agency. The definition of gazetted agency presently includes a comprehensive range of domestic law enforcement agencies, including the AFP and ASIO, as well as law enforcement bodies in over 200 ‘foreign law enforcement countries’ and ‘parts of foreign law enforcement countries’, and other Australian government departments (such as the Department of Social Services).

That information is protected if it was communicated on condition that it be treated as confidential and is relevant to an exercise of certain powers under the Australian Citizenship Act 2007 (Cth) and the Migration Act. The relevant powers include, for example, the power to determine that a person’s citizenship has ceased as a result of international terrorist conduct under s 36B of the Citizenship Act and the power to refuse or cancel a visa on character grounds under s 501 of the Migration Act.

The Bill seeks to replace a provision (s 503A(2)(c) of the Migration Act) which was held to be constitutionally invalid by the High Court in Graham. That provision was held to be invalid to the extent that it prohibited the disclosure of designated information to the High Court exercising review jurisdiction under s 75(v) of the Constitution (or the Federal Court in its equivalent jurisdiction). The Department of Home Affairs says that, since the High Court’s decision, it has been unable to rely upon the type of suspicion and intelligence which would be protected by s 503A in making visa decisions, due to uncertainty about whether the information would be disclosed to the individual before a Court (Legal and Constitutional Affairs Legislation Committee Hearing, 2 March 2021, 38-39).

The Bill also has other substantial purposes, including extending the regime for the first time to powers under the Citizenship Act, and introducing criminal offence provisions, capable of being breached by Commonwealth officers who disclose information which has been communicated by foreign and domestic law enforcement (or other government) agencies to the Department of Home Affairs.

The Bill has attracted strong criticism, which has been canvassed in submissions by the Law Council of Australia, Legal Aid New South Wales and other advocacy bodies. The Law Council (19 February 2021, 5) has recommended that the Bill ought not to progress until examination of whether it is ’likely to curtail the capacity of a court to exercise its judicial review jurisdiction under or deriving from s 75(v) of the Constitution to a substance or degree resulting in invalidity’.

Section 75(v) and the constitutional limitation

Section 75(v) of the Constitution provides that ‘[i]n all matters … in which a writ of Mandamus or prohibition or an injunction is sought against an officer of the Commonwealth … the High Court shall have original jurisdiction’. It is now established that the existence of s 75(v) gives rise to a constitutionally entrenched minimum provision of judicial review.

The modern High Court first adverted to that minimum provision in S157/2002 v Commonwealth [2003] HCA 2 (S157). The majority observed at [104] that the s 75(v) jurisdiction:

is a means of assuring to all people affected that officers of the Commonwealth obey the law and neither exceed nor neglect any jurisdiction which the law confers on them. The centrality, and protective purpose, of the jurisdiction of this Court in that regard places significant barriers in the way of legislative attempts (by privative clauses or otherwise) to impair judicial review of administrative action.

Bodruddaza v Minister for Immigration and Multicultural Affairs [2007] HCA 14 (Bodruddaza) confirmed that constitutional limitation. There, the High Court held invalid a time limit (28 days in the usual case) upon the lodging of an application for the exercise of the s 75(v) original jurisdiction. The Court held that the time limit was inconsistent with the place of s 75(v) in the constitutional structure because it curtailed or limited, whether directly or as a matter of practical effect, the right or ability of applicants to seek relief under s 75(v). The Court also confirmed that the purpose of s 75(v) is ‘to make it constitutionally certain that there would be a jurisdiction capable of restraining officers of the Commonwealth from exceeding Federal power’.

The decision in Bodruddaza indicated that the minimum constitutional provision is concerned with ‘practical, and not merely theoretical, access to the High Court’s s 75(v) jurisdiction’ (James Forsaith, 25-26).

Graham, the decision which has prompted the Bill, is the most recent in the line of jurisprudence.

Section 503A(2)(c) of the Migration Act provided that, where information had been communicated by a gazetted agency on condition that it be treated as confidential information, then ‘the Minister or officer must not be required to divulge or communicate the information to a court, a tribunal, a parliament or parliamentary committee or any other body or person’.

As adverted to above, the provision was designed to protect the information-gathering network of the Department of Home Affairs by permitting it to rely upon sensitive information confident in its non-disclosure (including to a Court in proceedings in its original jurisdiction for a writ against a Commonwealth officer). It was the latter which caused the constitutional problem, upon its intersection with s75(v) of the Constitution.

The High Court by majority (Edelman J dissenting) confirmed the existence of the constitutional limitation, and held that s 503A(2)(c) transgressed that limit because it ‘[o]perate[d] in practice to shield the purported exercise of power from judicial scrutiny’ (Graham at [53]). In circumstances where the Minister reached the relevant decision entirely on the basis of the confidential information, it would be impossible for the Court to analyse whether the Minister, in forming the suspicion and state of satisfaction and exercising the discretion, had done so reasonably on the material to which he had had regard (at [59]).

Taken together, S157, Bodruddaza and Graham indicate that a law which encroaches (in substance) on the Court’s practical capacity to conduct, or an applicant’s capacity to access, the original jurisdiction, will be invalid as a restriction of the s 75(v) jurisdiction.

Yet the scope of this minimum provision has only so far been determined by reference to legislation which imposed fairly extreme restrictions upon the ordinary course of a constitutional review proceeding. The High Court would not allow a prohibitive and inflexible time limit; nor a complete prohibition on disclosure of relevant information to the Court. However, there is nothing to indicate that a lesser restriction would necessarily be valid: the majority in Graham did not find it necessary to analyse further ’matters of substance and degree which may or may not result in the invalidity of [a different] statutory provision’ (at [65]).

The Bill’s approach to the constitutional limitation

The Bill addresses the constitutional limitation by providing that a Court may order that the confidential information be disclosed to, or given in evidence before, it; and may give the information such weight in the substantive proceedings as it considers appropriate. The Minister for Immigration, Citizenship, Migrant Services and Multicultural Affairs has signalled an intention to rely on this element, stating (Scrutiny Digest 5/21, 62) ‘[t]he Bill will provide safeguards for the applicant by allowing the Courts to decide how much weight to give to the confidential information that has been submitted in evidence’.

However, the Court is prohibited from disclosing the information to any person (including the applicant or the applicant’s lawyer) if it determines that disclosure would create a real risk of damage to the public interest, and the Bill prescribes exclusive criteria which are relevant to that determination. Although framed as a weighing task, it is notably one-sided. The criteria do not include the applicant’s right to procedural fairness, or the proper administration of justice (see Scrutiny Digest 5/21 at [2.81]). The Government clearly considers the disclosure track to fall outside of the s 75(v) problem: this is clear from the pointed exclusivity of the range of considerations.

The question addressed here is how the measure taken by the Bill – that is, in the majority of circumstances to make the suspicion, understanding or intelligence available to the Court alone – will interact with the minimum provision.

There are a number of features of the Bill which are, at least, undesirable from the perspective of the High Court’s review jurisdiction. What follows is largely speculation as we have comparatively little information about the High Court’s view about the outer limits of that review jurisdiction. However, taken together, they may indicate an impairment of the review jurisdiction.

The information which may be designated lacks objective criteria

First, and in continuation of the current s 503A, the Bill does not prescribe criteria binding on a gazetted agency as to the designation of information as confidential. There is no objective criterion, either of type or underlying substance, which limits the information which may be designated confidential by the foreign and domestic law enforcement (or other government) agencies which share it with the Department of Home Affairs.

On its face, and even outside of s 75(v), this is undesirable from a law-making perspective, as it creates an unreviewable power on the part of the gazetted agency. The Bill does not stipulate the type of information which may be designated or the criteria which are relevant to that designation.  In this practical sense, the power is non-reviewable (because there is no standard in the legislation by which to measure the designation). There is no mechanism in the Bill to appeal from the gazetted agency’s decision to attach a condition of confidentiality to the information.

As a policy matter, this may flow from the fact that the regime established by the Bill is not really designed to protect confidential information. Rather, the federal government’s policy concern is to give a comprehensive range of foreign governments and domestic law enforcement agencies the strongest possible confidence in their own position, thereby encouraging information sharing. The Minister for Immigration, Citizenship, Migrant Services and Multicultural Affairs has stated ’[I]t is the agencies themselves who designate the information as confidential because of the intrinsically sensitive nature of its contents and scope’ (quoted in Scrutiny Digest 5/21, 62).

The broader, structural problem is that this non-reviewable power constitutes the point of entry into the non-disclosure regime. One of the consequences is that an applicant will have an extremely limited ability to contest the matters of fact on which the regime hinges. The Law Council has made this point persuasively (Submission dated 19 February 2021, 9). It also means that the Court has no ability to control which information falls within the regime, and which information falls outside of it.

In the separate realm of Chapter III jurisprudence on the conferral of limitations which are inconsistent with the judicial function, the High Court has considered it necessary that a confidentiality regime be underpinned by objective criteria, the existence of which can be reviewed (see, for example, Gypsy Jokers Motorcycle Club Incorporated v Commissioner of Police [2008] HCA 4 at [33]).

The Bill may prevent access to judicial review

Another real possibility is that the Bill limits access to the High Court’s review jurisdiction. Bodruddaza is the most direct authority on the question of access to the jurisdiction. There, the High Court stated (at [53]) that:

a law with respect to the commencement of proceedings under s 75(v) will be valid if, whether directly or as a matter of practical effect, it does not so curtail or limit the right or ability of applicants to seek relief under s 75(v) as to be inconsistent with the place of that provision in the constitutional structure.

In a case where a decision is made solely on the basis of information to which the Bill applies, and the Court does not make an order for disclosure of the information, the entire basis for the decision will be concealed from the applicant. The critical aspect of the Bill which achieves this effect is the nexus between the condition of confidentiality and relevance. It is a statutory secrecy regime which selects ‘relevance’ as the starting point for the secrecy regime.

In such a case, an applicant will have a very limited ability to assert judicial review grounds. This may, in the most practical sense, impair the ’protective’ nature of the s 75(v) jurisdiction (S157 at [104]). The Bill fails to remediate this difficulty, because the Court has no capacity to take into account the interests of the applicant as part of the ‘public interest’ test.

Combined with the point made above (that is, that the designation of the information lacks objective criteria), these factors may create a limitation, in the practical sense, of access to the s 75(v) jurisdiction. The problem might be partly addressed if the Bill included objective criteria to limit the circumstances in which information might be classified as confidential.

Interference with the Court’s judicial review task

Lastly – and again, as a matter of speculation – the modified adversarial process made available to the Court by the Bill may be inconsistent with the review jurisdiction itself.

The key deficiency identified in Graham was that the Court was unable to analyse whether the Minister had acted reasonably in the exercise of the power, because it was unable to view the information on which the Minister had acted. The Bill attempts to address that deficiency by permitting disclosure of the suspicion or intelligence to the Court itself.

Yet non-disclosure to the applicant may also create difficulties for the Court’s review function. The absence of submissions or evidence from the applicant limits the Court’s ability to engage in an analysis of whether the exercise of power was authorised by the material on which it was based. The Bill does not protect against this possibility because, in considering whether to disclose the information, the Court is unable to consider the impact of non-disclosure upon its own review function.

For example, a disproportionality argument (in support of an unreasonableness review ground) will likely rely on evidence which can only be sourced from the individual concerned, such as evidence that there was an alternative course available to the decision-maker or that the outcome is unduly harsh because of the applicant’s personal circumstances. It is improbable that such an argument will appear to the Court on the material itself, particularly given the high threshold usually required to establish a judicial review ground of this sort.

A similar difficulty arises where the material does not support the decision at all. This possibility cannot be excluded where a visa or citizenship decision has been made on the basis of untested suspicion, association information or intelligence. To take a hypothetical example, suppose that the Minister has obtained a law enforcement record from a foreign government relating to a tax debt, which the applicant has in fact paid.

Under the limited public interest test, the Court is unable to take into account its need for evidence from the applicant, which could be critical to establishing an abuse of power (whether as a failure to take into account a mandatory consideration or on the ground of legal unreasonableness) in either of these cases.

The Law Council has made a related point, commenting at [106] that ’in having the practical effect of preventing an applicant or their lawyer from knowing any essence of the case against them or making any submissions, the Bill impacts the Court’s ability to enforce the limits of an officer’s power and draw inferences adverse to the Minister to the degree required for a consideration of invalidity’.

Here, as elsewhere, it is relevant that there are no objective criteria attaching to the designation of information as confidential by a gazetted agency. It is more likely that there is a limitation, in the practical sense, of the constitutional review jurisdiction where a Court is unable to evaluate the type of information, or the reasons for confidentiality, which trigger this substantial narrowing of its adversarial process.

Conclusion

The minimum constitutional provision of judicial review is not yet fully defined. Whether, and to what degree, the legislature may impair judicial review of administrative action is likely to turn on fine distinctions in respect of the particular piece of legislation. However, there are real questions about whether the Bill, if passed, would prove consistent with the minimum provision.

—

Isolde Daniell works as a solicitor in private practice in Sydney. 

Suggested citation: Isolde Daniell, ‘Overcoming Graham: The s 75(v) constitutional guarantee and non-disclosure in migration and citizenship decisions’ on AUSPUBLAW (30 July 2021) <https://auspublaw.org/blog/2021/08/overcoming-graham-the-s-75v-constitutional-guarantee-and-non-disclosure-in-migration-and-citizenship-decisions/>.