Federal Court of Australia

RDYQ v Minister for Immigration and Citizenship [2026] FCA 645

ORDERS

THE COURT ORDERS THAT:

1.    The application for an extension of time within which to appeal be allowed.

2.    The appeal be allowed.

3.    The first respondent is to pay the appellant’s costs of the application for an extension of time within which to appeal and the appeal as agreed or assessed.

4.    The costs in order 3 are to be paid directly to Mr Wood SC pursuant to rule 4.19(3) of the Federal Court Rules 2011 (Cth).

Note:    Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011.

REASONS FOR JUDGMENT

LENEHAN J

1.    INTRODUCTION

1    This proceeding concerns an appeal from a decision made by a judge of the Federal Circuit and Family Court of Australia (Division 2) (the primary judge), dismissing an application for judicial review of a decision made on 27 November 2024 by the second respondent, the Administrative Review Tribunal (the Tribunal). By that decision, the Tribunal affirmed a decision of a delegate of the first respondent, the Minister for Immigration and Citizenship (the Minister), under s 65 of the Migration Act 1958 (Cth) (the Act), to refuse the grant of a protection visa to the appellant.

2    For the reasons set out below, the application for an extension of time within which to appeal and the appeal should be allowed.

2.    BACKGROUND

3    The Tribunal accepted that the appellant is a citizen of Canada and also of the United States of America (the USA). He arrived in Australia in 2013 – although the record is unclear as to what visa he held at that time. It was variously described as a “Work and Holiday (Class US) (Subclass 462) visa” (in an earlier decision of the former Administrative Appeals Tribunal dated 22 August 2022) and as a “Partner (Temporary) (Class UK) (Subclass 820) visa” (by the Tribunal making the decision in issue in this case). Nothing turns on that for present purposes. He was subsequently granted a Partner (Class BS) (Subclass 801) visa in 2017.

4    On 5 July 2019, the appellant was convicted of two offences: possessing child exploitation material, for which he was sentenced to two years’ imprisonment; and using a carriage service to transmit, make available, publish, distribute, advertise or promote child pornography material, for which he was sentenced to twelve months’ imprisonment. In respect of the first offence, the appellant’s sentence was suspended after four months’ imprisonment on condition that he be of good behaviour for two years. In respect of the second offence, the appellant was released upon his recognisance on 5 November 2019 on condition that he be of good behaviour for two years.

5    On 10 September 2019, the appellant’s visa was cancelled under s 501(3A) of the Act (the visa cancellation decision). On 26 May 2022, a delegate of the Minister decided, under s 501CA(4) of the Act, not to revoke the visa cancellation decision (the non-revocation decision). On 22 August 2022, the Tribunal affirmed the non-revocation decision.

6    On 18 September 2024, RDYQ applied for a protection visa. He claimed to fear harm in both Canada and the USA. According to the appellant, as a consequence of his criminal convictions in Australia, he would be required to register as a sex offender and could be identified as such by members of the public by reference to a (potentially public) sex offender registry. As a result, the appellant claimed he would be ostracised from society and exposed to a risk of violence or death, especially from vigilantes.

7    On 17 October 2024, a delegate of the Minister (the Delegate) refused to grant RDYQ a protection visa (the Delegate’s decision). Relevantly, the Delegate found that effective protection measures were available to the appellant in Canada and the USA (see s 5LA of the Act), so as to engage the operation of s 5J(2) and deny the appellant the status of having a well-founded fear of persecution (s 5J(1)), and therefore of being a refugee (s 5H(1)). As to the so called “complementary protection” provisions, the Delegate was satisfied as to the matters in s 36(2B)(b): that is, the appellant could obtain, from an authority of either country, protection such that there would not be a real risk that he will suffer significant harm as provided for in s 36(2B)(b). For that reason, the Delegate was not satisfied that there was a real risk that the appellant will suffer significant harm as defined in s 36(2A).

8    RDYQ applied for review of the Delegate’s decision with the Tribunal.

9    The Tribunal affirmed the Delegate’s decision in a decision dated 27 November 2024 (the Tribunal’s reasons or Tribunal’s decision or TD). But it took a distinctly different path of reasoning. The Tribunal concluded that, because s 36(3) of the Act applies to the appellant in respect of Canada, Australian was taken not to have protection obligations in respect of the appellant, and that (accordingly) the appellant did not satisfy the refugee criterion in s 36(2)(a) or the complementary protection criterion in s 36(2)(aa) of the Act: see TD [105]. The critical premise for that conclusion was the Tribunal’s earlier finding at TD [101] that the appellant “does not have a well-founded fear of persecution in Canada” and that “there are not substantial grounds for believing that, as a necessary and foreseeable consequence of the [appellant] being removed from Australia to Canada, there is a real risk that the [appellant] will suffer significant harm”: see also TD [103]. Because the appellant had not taken “all possible steps to enter and reside in Canada”, and because the exceptions in s 36(4) did not apply in light of those earlier findings, the Tribunal was satisfied that s 36(3) applied to the appellant: see TD [102]-[107].

10    On 6 January 2025, RDYQ applied for an extension of time within which to seek judicial review of the Tribunal’s decision. On 26 August 2025, the primary judge granted RDYQ an extension of time (of five days) to file his application for judicial review: RDYQ v Minister for Immigration and Citizenship [2025] FedCFamC2G 1407.

11    On 8 October 2025, the primary judge ultimately dismissed the application for judicial review: RDYQ v Minister for Immigration and Citizenship (No 2) [2025] FedCFamC2G 1643 (the primary judgment or PJ). In that decision, her Honour observed that “[a]lthough the judicial review application purported to set out two grounds, neither party in their submissions referred to ground 1 or ground 2”: PJ [9]. Her Honour therefore proceeded to address the issues raised in RDYQ’s application, written submissions and oral submissions “thematically, consistently with the approach taken by the parties”: PJ [9]. Amongst those “themes” was the issue of whether the Tribunal applied the wrong standard in assessing whether the appellant “faced a ‘real chance’ of serious harm”: PJ [11(d)]. On that issue, her Honour concluded that the Tribunal’s finding – “the chance of harm to the [appellant] was remote and therefore the [appellant] did not face a real chance of serious harm or a real risk of significant harm” – was “a proper application of the real chance test … [and] the real risk test”: PJ [58]-[59]. Having regard to that and her Honour’s conclusions on the other issues raised by RDYQ, her Honour held that the Tribunal’s decision was not affected by jurisdictional error.

12    On 3 December 2025, RDYQ applied for an extension of time within which he could file a notice of appeal. My reasons for granting that extension are set out at [16]-[18].

13    The appeal therefore proceeds on the basis of the appellant’s (draft) notice of appeal which advances the following grounds of review.

(1)    The Tribunal erred by misapplying the “real chance” test or, in the alternative, by failing to consider information bearing on risk to the appellant (ground 1).

(2)    The Tribunal failed to give adequate reasons (ground 2).

14    I should add that while ground 2 was not advanced below, the Minister did not oppose the appellant being granted leave to rely upon it. Leave is granted. Ultimately, however, as I have found that the appellant has established ground 1, it is not necessary to determine ground 2.

15    Finally, the appellant was represented pro bono by Mr Wood SC (who did not appear for the appellant below) and the Minister by Mr Barrington. The standard of their advocacy was very high and of considerable assistance to the Court.

3.    APPLICATION FOR AN EXTENSION OF TIME

16    The orders of the primary judge were made on 8 October 2025. The period for commencing an appeal expired 28 days later on 5 November 2025: see r 36.03 of the Federal Court Rules 2011 (Cth) (the Rules). This proceeding was commenced 28 days out of time on 3 December 2025. RDYQ therefore required an extension of time within which to file a notice of appeal: see r 36.05 of the Rules.

17    The principles which govern an application for extension of time to appeal are well established and are conveniently summarised in BQQ15 v Minister for Home Affairs [2019] FCAFC 218 (Yates, Wheelahan and O'Bryan JJ) at [33]:

Under rule 36.05, the Court may grant an extension of the time within which an appeal is to be filed. The principles applicable to the exercise of the Court's discretion were set out in Hunter Valley Developments Pty Ltd v Cohen (1984) 3 FCR 344 at 348-9, which were adopted by the Full Federal Court in Parker v R [2002] FCAFC 133 at [6]:

a.    Applications for an extension of time are not to be granted unless it is proper to do so; the legislated time limits are not to be ignored.

b.    There must be some acceptable explanation for the delay.

c.    Any prejudice to the respondent in defending the proceedings that is caused by the delay is a material factor militating against the grant of an extension.

d.    The mere absence of prejudice to the respondent is not enough to justify the grant of an extension.

e.    The merits of the substantial application are to be taken into account in considering whether an extension is to be granted. Leave will not be granted where there are no reasonable prospects of success on the appeal: Kalanje v Minister for Immigration and Multicultural Affairs [2006] FCA 1618 at [5]. The applicant will have no real prospects of success where the case is devoid of merit or clearly fails; is hopeless; or is unarguable. In making an assessment the Court is not required to go into too great a detail, but is to ‘assess the merits in a fairly rough and ready way’: Jackamarra v Krakouer (1998) 195 CLR 516 at [7] - [9].

f.    The discretion to extend time is given for the purpose of enabling the Court to do justice between the parties. Special circumstances often arise in immigration cases. Where the delay is short and no injustice will be occasioned to the respondent, justice will usually be done if the extension of time is granted: WAAD v Minister for Immigration and Multicultural Affairs [2002] FCAFC 399 at [7]. This will be so particularly if the applicant is in immigration detention, requires the assistance of an interpreter and has limited knowledge of law and practice including the time for any appeal: Ariaee v Minister for Immigration and Multicultural Affairs [2001] FCA 1627 at [14]; N1202/01A v Minister for Immigration and Multicultural Affairs (2002) 68 ALD 21 at [12] (N1202/01A). Further, where an applicant is an asylum seeker, the gravity of the circumstances that may confront him or her may be a reason for not applying the time rule too strictly: Applicants M160/2003 v Minister for Immigration and Multicultural and Indigenous Affairs (2005) 223 FCR 441 at [5]. Nevertheless, the merits of the appeal will remain a relevant factor: N1202/01A at [13].

18    Mr Barrington indicated that he was instructed not to consent, but not to oppose the application for an extension of time. For the following reasons, the interests of justice plainly favour the grant of an extension of time.

(1)    The delay was relatively short.

(2)    The appellant explained the delay in affidavits (dated 1 December 2025, 4 December 2025, and 12 December 2025) filed with this Court, all three of which were prepared by him while self-represented and in immigration detention. Consistent with the position noted by Mr Barrington, the Minister did not seek to cross-examine the appellant on that evidence or suggest that it did not adequately explain the delay. I accept the appellant’s submission that an adequate explanation was given.

(3)    The Minister did not suggest that he will suffer any specific prejudice in this case by the grant of an extension. The absence of prejudice is not, in itself, a reason to grant the extension sought: BQQ15 [2019] FCAFC 218 at [33](d). But I accept the appellant’s submission that there is no prejudice caused to the Minister.

(4)    Both grounds of appeal sought to be agitated plainly have real merit (as to which, see further below).

19    I now turn to consider the grounds of appeal.

4.    GROUND 1

20    Ground 1 was put in writing as either:

(1)    that the Tribunal misapplied the “real chance” test; or

(2)    that the Tribunal erred by failing to consider information bearing on risk.

21    Although he did not resile from the first characterisation of the error, Mr Wood focussed on the second, which he (correctly) described as the “simpler” way of putting the putative error the subject of ground 1. In substance, he contended that the Tribunal “simply failed to engage with critical evidence or important evidence that bore on the predictive task that it was required to engage in”. That “non-engagement”, he submitted, supports “a conclusion that the Tribunal has either misapplied the test or failed to consider material bearing on the test”.

4.1    Relevant legal principles

22    The authorities dealing with this asserted failure, namely a failure to consider the material regarding risk of harm, are well known, and the principles did not appear to be seriously disputed between the parties. They fall to be considered against some bedrock propositions regarding the nature of the Tribunal’s statutory task.

(1)    The Tribunal was here under an obligation to conduct a “review” (s 348(1) of the Act). Implicit in the scheme providing for that “review” is an obligation to reconsider the merits of the decision under review in light of the information, evidence and arguments that are relevant to the application and are either provided to it or obtained by it: Applicant WAEE v Minister for Immigration and Multicultural and Indigenous Affairs (2003) 236 FCR 593, 604 [44] (French, Sackville and Hely JJ); Minister for Immigration and Border Protection v SZMTA (2019) 264 CLR 421, 435-436 [13] (Bell, Gageler and Keane JJ), 463 [104]-[105] (Nettle and Gordon JJ).

(2)    Whilst it is for the Tribunal to assess the relevance of, and the weight to be attributed to, any item of evidence, the Tribunal will fail to perform its duty of review if it fails to take account of cogent evidence providing substantial support to the appellant’s case: Minister for Immigration and Citizenship v SZRKT (2013) 212 FCR 99, 130-131 [111]-[112] (Robertson J); Minister for Immigration and Border Protection v MZYTS (2013) 230 FCR 431, 444-445 [38]-[39] (Kenny, Griffiths and Mortimer JJ); SZMTA, 435-436 [13] (Bell, Gageler and Keane JJ).

(3)    Statements of a formulaic kind or sweeping statements that all the information, evidence and arguments that are relevant to the application have been considered “will not shield from scrutiny whether in substance they have”: Hands v Minister for Immigration and Border Protection (2018) 267 FCR 628, 638 [38] (Allsop CJ with whom Markovic and Steward JJ agreed).

23    In approaching those obligations, the appellant (correctly) emphasised two points.

24    First, the lack of utility in seeking to erect some rigid distinction between so-called “claims” and so-called “evidence in support of claims”. In that regard, as Robertson J noted in SZRKT, a line was sometimes sought to be drawn between overlooking (or misunderstanding) “mere items of evidence on the one hand”, and “overlooking (or misunderstanding) whole categories of evidence on the other hand”: at [69]. That distinction was sometimes expressed as being between “failing to understand some evidence, even if it was important, and failing to understand the claim (or an integer of the claim) for which the evidence was adduced”: at [69].

25    For the reasons his Honour gave, that is a conceptual distinction that is unstable and which has largely had its day. As his Honour emphasised at [111]-[112], there is “no clear distinction in each case between claims and evidence” (referring to SHKB v Minister for Immigration and Multicultural and Indigenous Affairs [2004] FCA 545, [24] (Selway J)). Rather, as his Honour observed, the “fundamental question” must be the “importance of the material to the exercise of the Tribunal’s function and thus the seriousness of any error”: at [111]. It follows that the distinction between claims and evidence provides (at most) a “tool of analysis” (SZRKT at [111]), and one which may prove very difficult to apply (SHKB at [24]). It is certainly not the discrimen itself.

26    To the extent the Minister contended otherwise, that submission should be rejected.

27    The Minister suggested that Feutrill J’s reasons in DXI22 v Minister for Immigration and Citizenship [2026] FCA 43 may indicate the ongoing importance of the conceptual distinction between claims and evidence. However, I do not read his Honour’s observations as going any further than Robertson J’s acceptance that those concepts may serve as a tool of analysis: see particularly DXI22 at [37], where his Honour refers to that very passage from SZRKT.

28    For the reasons given below, I have not found that (difficult) “tool” to be of any particular assistance here in answering the “fundamental question” regarding the importance of the material to the exercise of the Tribunal’s function.

29    The second point of principle made by the appellant concerned the nature of the obligation to take account of such material (assuming there was an obligation to do so).

30    The appellant relied upon Hill J’s useful distillation of some of the relevant authorities in BQQ25 v Minister for Immigration and Citizenship [2025] FCA 1279. In the context of a Ministerial refusal decision under s 501(1), his Honour drew an analogy with authorities dealing with the power conferred by s 501CA(4) of the Act to decide whether to revoke the mandatory cancellation of a non-citizen’s visa. His Honour observed, in that regard, that the authorities dealing with the content of the obligation to consider representations under s 501CA(4) provided guidance on the scope of the corresponding obligation under s 501(1): see [24]. The decision-making context here is different again (involving s 65), but I accept that that body of authority provides useful guidance on the general principles to be applied here.

31    Accepting those contextual differences, the appellant emphasised the following matters arising from Hill J’s reasons in BQQ25 [2025] FCA 1279.

(1)    Referring to what was said by Kiefel CJ, Keane, Gordon and Steward JJ in the context of the obligation to consider representations under s 501CA(4) in Plaintiff M1/2021 v Minister for Home Affairs (2022) 275 CLR 582 at [24], Hill J noted at [25] that the content of that obligation is to “read, identify, understand and evaluate the representations”. As explained in Plaintiff M1 at [24], that entails an obligation on the decision-maker to “have regard to what is said in the representations, bring their mind to bear upon the facts stated in them and the arguments or opinions put forward, and appreciate who is making them”. Applied in the current context, as regards the material put forward by the appellant, the Tribunal was required to bring its mind to bear upon the facts stated in that material and the arguments or opinions contained in that material.

(2)    The appellant accepted that that proposition is subject to the limitations explained in Plaintiff M1 (2022) 275 CLR 582 at [24] (and see BQQ25 [2025] FCA 1279 at [25], first dot point), which reflect what I have said at [22(2)] above: that is, having read, identified, understood and evaluated the material, the weight to be given to that material is a matter for the decision-maker, and the decision-maker is not obliged “to make actual findings of fact as an adjudication of all material claims” made by the person. Their Honours also emphasised that expressions such as “active intellectual process” and “proper, genuine and realistic consideration” must be understood in their proper context, and carry the risk of becoming “a kind of general warrant, invoking language of indefinite and subjective application, in which the procedural and substantive merits of any [decision-maker's] decision can be scrutinised”: see Plaintiff M1 at [26]. That is not the correct approach: see Plaintiff M1 at [26]; BQQ25 [2025] FCA 1279 at [25], third dot point.

(3)    As the appellant submitted, those cautionary words did not mean that Plaintiff M1 affected any radical alteration to the law in this area. In particular, as observed in BQQ25 [2025] FCA 1279 at [26]:

Plaintiff M1 does not mean that mere lip service is permitted. It is still the case that “[w]hat is required is the reality of consideration by the decision-maker”: Minister for Immigration and Border Protection v Maioha [2018] FCAFC 216; (2018) 267 FCR 643 at [45] (Rares and Robertson JJ). Equally, it is still the case that a person affected by a decision should not be “left to guess” what role, if any, a matter has played in the exercise of discretion: Lafu v Minister for Immigration and Citizenship [2009] FCAFC 140; (2009) 112 ALD 1 at [49] (the Court).

(4)    The appellant sought to apply those statements to the Tribunal’s reasoning in this case (see similarly BQQ25 [2025] FCA 1279 at [46]-[47]).

32    Mr Barrington quite properly did not contest those propositions as a matter of principle. Rather, he submitted that, having regard to the Tribunal’s reasoning and the nature of the material, I should not infer that the Tribunal failed to consider the material advanced by the appellant. As Mr Barrington put it: the Tribunal “expressly had regard to it at [19] of its reasons, and against that explicit recognition of the material, the Court should not lightly infer it was not considered”. Even if he was wrong in that, he contended that none of the material in issue here was so central or important to the Tribunal’s task that a failure to consider it would amount to jurisdictional error. In this regard, the Minister appeared to take issue with the age and generality of the evidence adduced by the appellant: namely, that “[s]ome of the information was significantly dated, some [o]f it was general in nature, and none of it related to Canada specifically”.

33    To evaluate those submissions, it is necessary to first identify the material, its nature and its relevance to the appellant’s case. I will then turn to how the Tribunal dealt with that material (to the extent it did so).

4.2    Material before the Tribunal

34    Candidly accepting that he was relying on the “best, most recent significant evidence that was before the Tribunal”, Mr Wood put at the forefront of his argument two papers.

4.2.1    Article by Cubellisa et al

35    The first paper was entitled “Sex Offender Stigma: An Exploration of Vigilantism against Sex Offenders”, written by Assistant Professor Michelle A. Cubellis, Dr Douglas N. Evans, and Adam G. Fera. In the “Notes on contributors” at the end of the paper, it was noted that:

(1)    Michelle A. Cubellis was:

an Assistant Professor at the Central Connecticut State University and Research Associate at the Research and Evaluation Center, John Jay College of Criminal Justice. She received her Ph.D. in Criminal Justice from John Jay College. Her research focuses on institutional responses to deviance, child sexual abuse, macro-level responses to violence and victimization, and program evaluation.

(2)    Douglas N. Evans was:

a Senior Investigator and Project Director at the Research and Evaluation Center, John Jay College of Criminal Justice and also teaches in correctional facilities in New York and New Jersey through Hudson Link for Higher Education and NJ STEP. His research focuses on the stigmatization resulting from criminal justice system contact, evaluation of violence reduction and alternative to incarceration programs, the effects of education programs in prisons, and the public health consequences of mass incarceration. He received his Ph.D. in Criminal Justice from Indiana University.

(3)    Adam G. Fera was:

a doctoral candidate at the Graduate Center CUNY/John Jay College of Criminal Justice. He is a Research Analyst on the Misdemeanor Justice Project. His research interests include jury decision making, capital jury research, methods of execution research, plea bargaining and the trial penalty, escapes from correctional custody, testing criminological theory, and sex offender research. H[e] is currently writing his dissertation which is a longitudinal examination of the trial penalty in federal cases.

36    On its face, that article was published in a journal entitled “Deviant Behavior”, having been received by that journal on 26 June 2017 and accepted on 28 July 2017.

37    The “Introduction” described the background to, and methodology of, that paper as follows:

Sex offenders (SOs) are a highly stigmatized group. Individuals convicted of a sexual offense are subject to harsh public disdain, in addition to the loss of privacy due to registration and community notification laws. The stigma attached to SOs is so pervasive that it extends to individuals suspected of having committed a sex offense. It even extends to others who are in close proximity to a suspected, accused, convicted, or registered SO. SOs experience stigma in a variety of ways, ranging from discrimination and exclusion from social participation to violent victimization and murder. Private citizens who act outside of the law to target a group or specific individual for the purpose of causing harm are known as vigilantes. Studies on the collateral consequences of SO registration and notification have acknowledged and explored awareness of vigilantism against SOs (Brannon et al. 2007; Tewksbury 2005), but research has yet to explore the nature of vigilantism against SOs (and non-SOs who are mistakenly connected to a sex offense), characteristics of vigilantes and their victims, and the ways in which the legal system responds to vigilantes who target SOs.

There are currently more than 850,000 registered SOs in the United States (Special Analysis Unit 2016) and many more individuals that will never be detected, reported, arrested, convicted, or registered for having committed a sexual offense. The current study used open-source searches to identify incidents of vigilantism in which the victim was a suspected, accused, convicted, or registered SO and produced data on a small sample of known incidents of vigilantism. Reports of these incidents were compiled and coded for characteristics of the offender, incident, and victim, and entered into what we refer to as the "Sex Offender-Vigilante Database.

(Emphasis in original.)

38    As will be immediately apparent, the nature of the study was such that it was likely to underrepresent the scale of the problem: it relied on reports of incidents that were accessible via “open-source searches”, with other incidents – be they unreported or not accessible via open-source searching – falling outside its scope.

39    Later, under the heading “Stigmatization of sex offenders” the authors said this:

Registered SOs may experience harassment and victimization as a result of their placement on registries (Tewksbury and Zgoba 2010; Zevitz and Farkas 2000) although this victimization is not isolated to only individuals listed on SO registries. One study suggests that between 5 and 16% of SOs have experienced physical assault (Levenson and Cotter 2005). For those offenders that are able to find a job, they are more likely than non-SOs to be denied promotions at work. They also face rude treatment in public spaces, being asked to leave businesses, loss of friends who find out about their SO status, harassing and threatening phone calls and mail, and high levels of shame (Tewksbury 2005; Tewksbury and Lees 2007).

40    Some of the real-world consequences of that “stigmatisation” were further described in the next section of the paper under the heading “Vigilantism”:

A serious consequence of the stigmatization of SOs is their risk of victimization at the hands of vigilantes. When a community becomes aware of a SO, citizens may harass, discriminate against, or victimize them out of anger or disgust (Tewksbury 2005). Vigilantes are private citizens who take the law into their own hands to harass, assault, or physically harm another individual. According to Johnston (1996), there are several components to the definition of vigilantism: it involves premeditation by those engaging in it; vigilantes are private citizens who engage voluntarily without government support; acts of vigilantism constitute a social movement; vigilantism is a reaction to acts of crime or deviance and a perceived failure of the system to respond adequately to them; and acts of vigilantism seek to control crime. An additional element of Johnston's (1996) conception of vigilantism is the inclusion of violence as a common element. However, violence is not always relevant when a SO is the target of vigilantism because exposing a SO in the community, a non-violent act that does not even necessitate contact between vigilantes and their targets, is a tactic in which citizens outside of law enforcement are not generally permitted to engage.

41    Later in that section (and reflecting the point that I have made above about the study’s under-representation of the problem), the authors said this:

The prevalence of vigilantism directed toward SOs is unknown, but research indicates that they experience more acts of vigilantism than the public is aware (Brannon et al. 2007).

42    A similar point is made later by the authors under the heading “Current study” and the sub-heading “Research questions”:

It would be extremely difficult to determine the extent of acts of vigilantism against SOs because the number of unreported and unrecorded incidents is unknown …

43    A further reflection of that point appears later under the heading “Discussion”:

It is impossible to know the prevalence of vigilante acts against SOs because many of these incidents are not reported in the media. The media usually focuses on interesting and unique events that attract public interest, which gives a limited indication of the nature of vigilante acts against SOs. The true extent of vigilantism against registered SOs is unknown, but the number of incidents identified in the current study is far less than the actual number of incidents that have occurred.

44    With those recognised limitations, the purpose of the study was described in this way:

[T]he purpose of this study is to explore the nature of vigilantism against SOs. Specifically, we are interested in determining trends in the relationships between vigilantes and their SO victims, reasons that they would target a suspected or convicted SO, their method of attack, and the legal outcome of the attack. We also collected information on frequencies and demographics of vigilantes and their victims, weapons used in the attack, and locations where incidents occurred. With a limited sample it is difficult to identify consistent trends among vigilante attacks, but this research will provide a better indication of the consequences of the intense stigma that convicted SOs, and even those that are suspected of having committed a sexual offense, experience.

45    Under the heading “Method”, the nature of the study (and its geographic focus) was described by the authors in this way:

To identify vigilante incidents, researchers used the following search engines: Google, Yahoo, Bing, and LexisNexis. The research team used a variety of keywords (e.g., vigilante, SO, attack, beat, shot, stabbed) to locate stories involving vigilante attacks on SOs. Stories of potential incidents were screened using specific inclusion criteria that the vigilante incident occurred in the United States and that the victim was a registered SO or they were targeted because they were suspected of having committed a sex offense. The authors were interested in the total number of incidents of vigilantism against SOs or those accused of committing sexual offenses, regardless of date, leading to a data sample with incidents ranging from 1983 until 2015.

(Emphasis added in bold.)

46    The study therefore involved two categories: cases where the victim was a registered sex offender; and cases where the victim was targeted because they were suspected of having committed a sex offense. Both categories were limited to incidents in the USA. And only the first category was directly relevant to the issues before the Tribunal.

47    The results identified 279 separate incidents that took place between 1983 and 2015 in which a vigilante (a private citizen acting without legal authority) targeted and/or attacked an individual who had previously been accused (credibly or otherwise), arrested, convicted, or registered for a sex offense.

48    Although the study did not directly report its results in terms of registered sex offenders, 178 of the 279 recorded incidents involved people who had been convicted of a sex offence and therefore, presumably registered on that registration system. That is reflected in the “Discussion” section where the authors noted:

The majority of victims of these attacks were registered SOs, leading to concern about the role of the registry in aiding these attackers. The personal information contained on SO registries about an offender's address, physical characteristics, and what they look like enables members of the public to identify these individuals in the community. While this identification is intended to empower community members to protect themselves and their family members from a stranger SO attack, the current study suggests that a few citizens have used this information to proactively locate and harm registered SOs.

(Emphasis added in bold.)

49    Also, in the “Discussion” section, the authors said this:

This study indicates that the stigmatization of SOs, and individuals suspected or accused of sexually offending, potentially results in harm to them in the form of vigilantism. The risk of vigilante victimization may be a consequence of having been labelled (accurately or inaccurately) as a SO. The SO label is so powerful that it becomes an individual's “master status" (Edwards and Hensley 2001). The stigmatization associated with the SO label is revealed in the derisive sentiments that citizens feel about SOs and discrimination and actions expressed toward this stigmatized group. Many of the incidents discussed in this study demonstrate the abhorrence, fear, and anger that citizens and incarcerated individuals have regarding SOs and the legal ramifications that some are willing to risk to administer their own form of street justice against SOs.

It is challenging to measure expressions of stigmatization that the public exhibits toward SOs. Quantifying discrimination, prejudice, rejection, dirty looks, harsh language, and avoidance are difficult tasks. Vigilantism may be one of the easier expressions of stigmatization to measure, particularly when the act involves unambiguous violence. However, for the purposes of this study, measurements of vigilantism against SOs are limited to incidents that are reported to law enforcement and/or discussed in the media. The Sex Offender Vigilante database includes 279 cases of vigilantism against SOs in the U.S. While this is likely only a small fraction of the total amount of offenses vigilantes have perpetrated against suspected or convicted SOs, it provides a foundation for exploring the nature of vigilantism, the networks of those who are associated with or mistaken for a SO and tangentially affected by vigilantism, and gradually, the extent of vigilantism perpetrated against SOs.

4.2.2    Article by the Australian Institute of Criminology

50    The second article before the Tribunal on which the appellant focussed was prepared by the Australian Institute of Criminology (an entity of the Australian Government) in May 2018. That article was entitled “What impact do public sex offender registries have on community safety?”, and was published in a publication titled “Trends & issues in crime and criminal justice” (on its face, a publication of the Australian Institute of Criminology). The authors of that paper were Sarah Napier, Christopher Dowling, Anthony Morgan and Daniel Talbot.

51    At the conclusion of that article, the following information about the authors appeared:

Sarah Napier is a Principal Research Analyst at the Australian Institute of Criminology.

Christopher Dowling is a Senior Research Analyst at the Australian Institute of Criminology.

Anthony Morgan is a Research Manager at the Australian Institute of Criminology.

Daniel Talbot is a police officer with Western Australia Police Force and a PhD candidate at Edith Cowan University.

52    The aims of the paper were identified as being “to review the available empirical evidence to address” a number of “key research questions”, including “[w]hat additional issues [being issues in addition to reduction in sexual offending and re-offending, and perceptions of safety among the wider community] need to be considered when discussing the feasibility of a national public sex offender registry?”

53    Early on in that article, the authors drew attention to the experience of the USA:

In the United States, information on the name, appearance and location of high-risk sex offenders has been available to the public for 20 years.

54    Later in the article, reference was made to South Korea where, like the USA, “information is made accessible to the community via the internet”. The other comparative examples identified (including Australian examples) differed. As regards the United Kingdom, the authors noted:

In 1997, the United Kingdom passed the Sex Offenders Act, whereby convicted sex offenders were required to keep police notified of their address (Beard & Lipscombe 2016). In 2003 the Sexual Offences Act (UK) repealed and replaced the earlier legislation, mandating a more stringent register, although still without provision for public disclosure (Beard & Lipscombe 2016). In 2008 limited disclosure was piloted in four UK police districts, with the scheme being rolled out nationally in 2010. The amendment enabled members of the public to request information from police regarding whether an individual had a record for child sexual abuse (Beard & Lipscombe 2016) …

55    Likewise, in Australia, the schemes in the various states and territories provided for the collection of certain kinds of information (including addresses and other contact details, and information on motor vehicles, employment, club memberships and any children with whom the offender has contact), but “[t]his information is not made available to the public”. Some (limited) steps towards greater accessibility for the public were noted, not all of which had resulted in any change:

There have been some recent developments towards making registries more accessible to the public. In 2012, a restricted access public sex offender registry was established in Western Australia, in which local residents are required to enter their name and drivers licence to request information on missing registered sex offenders and sex offenders living in their area. Parents may submit a request as to whether an individual who has contact with their child is a registered offender (Taylor 2017). In the Northern Territory, a bill was proposed that would introduce a public sex offender registry similar to those in the United States, in which the names, whereabouts, physical descriptions and photographs of convicted serious sex offenders would be made available to any member of the public (not just local residents who submit a request) via a website. The bill was to be named 'Daniel's Law', in memory of 13-year-old Daniel Morcombe, who was sexually assaulted and murdered by a sex offender released on parole (Elferink 2014). However, following concerns raised by stakeholder groups the legislation was deferred in 2015 (Elferink 2015).

56    The material before the Tribunal thus indicated that it was primarily the USA, with its comparatively lengthy history of public disclosure laws, which provided the most obvious point of comparison when considering the possible consequences of such laws. Indeed, the only other country with any kind of comparable measure identified in that material was South Korea (which was not further discussed). Unsurprisingly, it was to the USA that the authors of the study looked when considering the “available empirical evidence” as regards one of the issues they identified as necessary to consider when discussing the feasibility of a national public sex offender registry, being “Vigilantism”:

Critics of public sex offender registries often highlight the potential for widespread public vigilantism, and concerns for the physical safety of registered sexual offenders. This is a plausible argument given the extremely negative public attitude towards sexual offenders in Australia. In the United States, Lasher & McGrath's (2012) review of multiple studies found that, on average, 44 percent of registered sexual offenders reported experiencing threats or harassment by neighbours, while around 20 percent experienced threats or harassment in general. Importantly, 16 percent of offenders reported that their family members or other cohabitants had been harassed, attacked or had property damaged as a result of their registration. Physical vigilantism (ie physical attack) targeting registered sexual offenders was less common, with (on average) eight percent experiencing physical attacks and 14 percent reporting some form of property damage. Policymakers and law enforcement should therefore be aware of the potential for a variety of forms of vigilantism to occur with a public sexual offender registry, along with the potential for vigilante activity targeting those related or otherwise close to registered sexual offenders.

Notably, a number of unofficial, community-run registries that attempt to compensate for the absence of a public registry in Australia have been established to keep members of the public informed of sexual offenders. These registries, while unreliable, essentially constitute online forms of vigilantism that have arguably resulted from the lack of a public sexual offender registry in Australia

(Emphasis added in bold.)

4.2.3    Relevance of the articles to the Tribunal’s task

57    I have set that material out at some length, together with biographical information about the authors. In doing so, I am, of course, not drawing any conclusions about the issues discussed therein or the weight that ought to have been given to that material. Those were all questions for the Tribunal. My point rather is to illustrate that what the Tribunal had before it was a detailed analysis, clearly relevant to the issues it needed to decide and (on its face) prepared by people who were highly qualified to express the views and opinions set out in that material (some of whom were employed by the Commonwealth or a Commonwealth entity). It was, on any view, cogent evidence providing substantial support to the appellant’s case.

58    Armed with that understanding of the content and nature of that material, I turn to the question of whether the Tribunal met its obligation to take account of that material in the sense identified in the authorities. But before I do, I should address submissions made by the Minister, seeking to downplay the significance of that material. None are convincing.

4.3    Minister’s submissions

59    The Minister made several submissions, seeking to diminish the significance of that material.

60    First, the Minister argued that the material was, in some respects, “significantly dated”, observing that “the academic article was published 7 years before the Tribunal’s decision, and the Australian Institute of Criminology paper was published 6 ½ years before the Tribunal’s decision”. That is so. It should also be noted that the studies referred to in those articles (and on which they are based) are even older.

61    But none of that is really to the point. There are two difficulties with the Minister’s submission.

(1)    Nowhere in its reasons did the Tribunal suggest that the age of the material was a basis upon which it could be put to one side. In those circumstances, the Minister is effectively inviting the Court to speculate as to how the decision might have been made: see LPDT v Minister for Immigration, Citizenship, Migrant Services and Multicultural Affairs (2024) 280 CLR 321, 329 [15] (Gageler CJ, Gordon, Edelman, Steward, Gleeson and Jagot JJ), 335 [36] (Beech-Jones J).

(2)    As the appellant submitted, this is not a case involving a particular dynamic environment where specific country conditions may be sensitive to changes in events: see eg Minister for Immigration and Border Protection v MZYTS (2013) 230 FCR 431 at [73]-[74] (Kenny, Griffiths and Mortimer JJ); GLD18 v Minister for Home Affairs [2020] FCAFC 2 at [77] (Allsop CJ and Mortimer J), referring to MZYTS. Rather, the material comprises empirically based conclusions, drawn from studies conducted over an extended period of time, concerning a discrete sociological and criminological phenomenon (namely, how members of the public respond, generally, to the availability of information about sex offenders). It is a particular form of “country information”, the potential relevance of which may be understood to have an enduring quality. As the Full Court in MZYTS at [74] observed, decision-makers may choose, “lawfully”, to rely on older information for a variety of reasons, material of this character being an obvious example. Indeed, had the Tribunal grappled with it, the persistent reporting of the phenomenon of vigilantism over time in response to a publicly available register would have been a matter that might have given further weight to the appellant’s argument. Again, those are issues for the Tribunal, not for the Court conducting judicial review; my point is that no consideration of those matters appears in the Tribunal’s reasons.

62    Second, the Minister submitted that the material was, in some respects, “general in nature”, noting that the Australian Institute of Criminology “simply stated that the potential for widespread public vigilantism was a ‘plausible argument’”.

63    The Minister’s submission must be rejected for the following reasons.

(1)    It needs to be recalled that the issue before the Tribunal was whether the appellant satisfied the refugee criterion in s 36(2)(a) of the Act (read with ss 5H and 5J) – amongst other things, that required the Tribunal to consider whether he had a well-founded fear of persecution within the meaning of s 5J (see also ss 5H(1)(a) and 36(4)(a)).

(2)    A person will have a well-founded fear of persecution if, amongst other things, there “is a real chance that, if the person returned to the receiving country, the person would be persecuted for one or more of [those] reasons”: s 5J(1)(b). A “real chance” is one that is not “remote” or “far-fetched”: Chan Yee Kin v Minister for Immigration and Ethnic Affairs (1989) 169 CLR 379, 389 (Mason CJ), 398 (Dawson J), 407 (Toohey J) and 429 (McHugh J); CGA15 v Minister for Home Affairs (2019) 268 FCR 362, 369-380 [22] (Murphy, Mortimer and O’Callaghan JJ).

(3)    “Plausible argument[s]” as to the potential for such harms are firmly within the realm of those concepts. That such material is expressed in general or tentative terms does not mean that it cannot rationally bear upon that question, or that it can simply be disregarded (although, again, the weight to be attributed to such material is a matter for the Tribunal).

(4)    Similar observations apply in respect of the Tribunal’s consideration of “real risk” in the context of ss 36(2)(aa) and 36(4)(b): TD [101]; see Minister for Immigration and Citizenship v SZQRB (2013) 210 FCR 505 at [245]-[246] (Lander and Gordon JJ).

64    Third, the Minister observed that none of that material “related to Canada specifically”. That is a truism, but not one that takes the Minister’s submission very far. The material was not specific to Canada, because Canada was yet to enact anything equivalent to the USA (or, it would seem, South Korea). The experience in the USA, with its extended history of implementation, provided an obvious comparator. The USA was, as the appellant described it, a “jurisdictional laboratory” that had “the very system of law that the Tribunal was contemplating might be introduced in Canada” – with 20 years of legislative experimentation and observation available to inform the questions confronting the Tribunal.

65    Fourth, relying on the point he sought to draw from DXI22 [2026] FCA 43 (see above at [27]), the Minister submitted (as regards the Australian Institute of Criminology paper) that the appellant had “disavow[ed]” a contention that one would compare the position in Australia with the position in the USA. The Minister relied, in particular, on the following passage from a letter (or submission) from the appellant to the Tribunal dated 25 February 2025:

Vigilantism violence is already alive and strong in Canada. As can be seen in the case of the Nova Scotia resident. The only thing that likely prevented him from committing the murders in Canada was the public vs private registration not a desire to only murder American sex offenders. In many ways Canada is sort of like the United States' little brother. They are always comparing themselves to the US and want to catch up. I believe that the geographic distance between the US and Australia or the UK as well as general societal differences allows for the latter countries to not feel such a desire to "catch up".

66    That submission was quoted by the Tribunal at TD [23]. Properly understood the appellant cannot be understood as disavowing any reliance on the Australian Institute of Criminology paper, or the views it expressed about what might happen if a similar law was introduced in Australia.

67    Although somewhat ambiguous, the reference to “a desire to ‘catch up’” appears to be connected to the appellant’s earlier observation in that letter that he was enclosing “some supplementary evidence”, providing:

additional information on the state of the Canadian registry. Including the legislation that was passed but not funded to create a public viewable registry. However, as you can also see the registry be on [sic] available to police has not stopped violence against sex offenders. In fact when you compare it to the US, it'll be much easier for Canada to have a national registry because in the US it required an act of Congress where they had threaten[ed] each states funding due to the 10th amendment.

68    In that context, the appellant’s reference to Canada having a desire to “catch up” is best understood as directed to legislative or policy developments, rather than any suggestion that Canadian citizens were seeking to “catch up” to the USA in terms of the prevalence of vigilantism. The appellant did not suggest that those jurisdictions were not relevantly comparable with Australia in the event similar laws were enacted here; nor did he suggest that the views expressed in the Australian Institute of Criminology paper (which he had earlier provided to the Tribunal) could be disregarded for some other reason. Indeed, if the Tribunal took his submission as conveying some broader point about the comparatively diminished risk of vigilantism in Australia, it would potentially give those expressed views more weight, in that even in such a society there was a view (expressed by officers of a Commonwealth body) that there could be an increased risk of such conduct in response to a public register.

69    In any event, none of that suggests that the claims versus evidence “tool of analysis” sheds some particular light upon the issues that arise here. The “fundamental question” remains the “importance of the material to the exercise of the Tribunal’s function and thus the seriousness of any error”: SZRKT (2013) 212 FCR 99, 130 [111] (Robertson J).

70    For the reasons developed below, the material was central to the question confronting the Tribunal on the path of reasoning it took.

4.4    Reasoning of the Tribunal

71    How then did the Tribunal approach that material?

72    It is true that the Tribunal referred specifically to each of the papers relied upon by the appellant at paragraph [19] of its reasons:

The [appellant] also relied on, and I have considered, the following country information:

a.    principally concerning the USA:

i.    …

ii.    What impact do public sex offender registries have on community safety?', 2018, Australian Institute of Criminology

iii.    'Sex Offender Stigma', 2017, Deviant Behaviour …

73    The description of the Australian Institute of Criminology paper as “principally concerning the USA” is itself revealing. As will be apparent from what I have said above, the point of that paper was to take the experience of the USA and use it to understand the possible consequences of the enactment of similar measures in Australia. It was a very similar exercise to that which the Tribunal was engaged in relation to Canada.

74    But, putting that to one side, those passing references and self-serving statements could rise no higher than “lip service”. Closer attention is required to the Tribunal’s reasons to discern whether it failed to take into account that material in the sense I have identified above by reference to the authorities: see again eg BQQ25 [2025] FCA 1279, [26]; Hands (2018) 267 FCR 628, 638 [38].

75    The Tribunal’s evaluation of the prospect of the appellant being harmed by vigilantes in Canada seemingly turned on three contingencies:

(1)    Would the appellant be required to register on the NSOR?

(2)    If so, what is the prospect that the NSOR become public?

(3)    If so, what is the prospect that the appellant would be harmed by vigilantes?

76    As to the first contingency, the Tribunal found at [56] that the appellant would be required to register on the NSOR if he entered Canada:

I accept that the [appellant] would be required to register on the NSOR if he entered Canada.

77    As to the second contingency, the Tribunal noted at [57] (see also [81] and [91]) that the NSOR was not currently public, expressly drawing a comparison with the situation in the USA:

The NSOR is not public and can only be accessed by Canadian police agencies. In contrast, American states are required to post the names, photographs, addresses and licence plate numbers of registered sex offenders. Police agencies must use the personal information in the NSOR only for police purposes and as authorised by law. There are criminal penalties for the misuse of the data by persons authorised to access it or by third parties who may obtain the information.

…

I am also satisfied that the information that the [appellant] would provide for that registration is currently private and only accessible by law enforcement authorities in Canada. I also find that no public, open-source information-for example, public news reports-about the [appellant’s] offending in Australia is available.

…

The delegate referred to information about groups in Canada that collate information about sex offenders in that country. I have considered that information in making my assessment; however, as I have already found, there is no publicly available information about the [appellant’s] offending.

78    However, at [82]-[83], the Tribunal found that that position may change (or that there was a possibility or prospect that it would do so), following the then forthcoming Canadian election (which the Tribunal considered likely to be won by the Conservative party):

I am satisfied that, at present, it is more likely than not that the Conservatives will win the forthcoming elections in Canada and form either a majority or minority government. I also accept that the existing legislative amendments passed by the Harper-led Conservative government that would enable the NSOR to be transitioned to public accessibility remain law, even if they were not utilised or funded by the Harper-led Conservative government or the incumbent Liberal government. The country information available to me does not support a conclusion that such a course of action is likely. In that regard, I refer to country information that suggests that the Conservative party could have, but did not, allocate funding for a public register when they were last in government. Having said that, I do not know why that funding was not made available and whether that reason would continue.

I accept that there is a possibility, or prospect, that a Poilievre-led Conservative government would seek to fund changes to the NSOR to make it publicly accessible. However, whilst the [appellant] has been able to provide reports on what might be described as Poilievre's 'tough on crime' policies, including proposed amendments to sentencing and other aspects of the criminal law, no reports have been provided in which Poilievre comments on, or commits to, making the NSOR public.

(Emphasis added in bold.)

79    As such, the critical reasoning, for present purposes, concerns the third contingency, as to which the Tribunal said this at [96]-[98]:

Whilst there is a prospect that the [appellant] would experience harm, the prospect of the [appellant] being harmed, whether because of his registration on the NSOR or otherwise, is remote. I do not accept that there is a real chance vigilantes will harm him or that he will be unable to obtain employment or housing.

The country information available to me does not support a conclusion that the Canadian population is actively involved in vigilantism or would, if the register were made public, seek to engage in vigilantism. While there will always be groups of people in any society who seek to engage in vigilantism, I am satisfied that the risk to the [appellant] can be confidently described as remote.

I am satisfied that the prospect of the [appellant] facing harm is remote. Not only does there need to be a confluence of events to occur for the [appellant’s] information to become public, but there also needs to be a person sufficiently motivated to seek to harm the [appellant] based on his past offending. In relation to being unable to access employment, again on the assumption that the [appellant’s] information was to become public, I am not satisfied that employers in Canada would routinely check the NSOR against potential job applicants or that they would refuse job applicants generally, or the [appellant] specifically, because of their presence on the NSOR, particularly without having regard to the nature of the offending that the person is said to have engaged in.

80    The first sentence I have extracted from [96] is to be read with [88], where the Tribunal said:

I am prepared to accept that if the NSOR were made public, the information would likely be easier to access for the Canadian population, which could increase the prospect of harm to the [appellant] or others on that list. I have taken this into account in my overall assessment.

81    The questions raised by that tentatively expressed conclusion in [88] included these: What was that prospect and how much was it potentially “increase[d]” by making the NSOR public? Did that increased prospect of harm rise to the level of a real and not remote chance in a Chan sense? Central to the appellant’s case that it did was the evidence from the jurisdiction which had had the most extensive (and most extensively observed) direct experience with those measures and their consequences. That material pointed (squarely) to “potential” or “plausible” risks of harm resulting from vigilante activities.

82    The Tribunal was, of course, not bound to accept that evidence or give it any particular weight. There may well have been reasons for distinguishing that direct experience from the likely experience in Canada – although noting that the apparent assumption underlying the Australian Institute of Criminology paper was that the consequences of such schemes may be similar despite jurisdictional differences.

83    None of that presents any difficulty for the appellant’s case because no reasoning of that nature is to be found anywhere in the reasons of the Tribunal. The material is simply put to one side without any explanation whatsoever.

84    Perhaps recognising that difficulty, the Minister sought to emphasise other aspects of the Tribunal’s reasoning, relying in particular on [84]-[85] and [87], where the Tribunal said this:

I am satisfied that, at present, Canada manages public notifications of sex offenders considered to be high-risk on a case-by-case basis, which can include notifying the public in a particular area about an applicant's release or presence.

The [appellant], appropriately in my view, recognised that the country information that he has provided refers to the information of ‘high-risk’ sex offenders being made publicly available and there not being vigilantism associated with that. The [appellant] did submit, however, that the making of the entire register public would make the information more centralised than it presently is and submitted that insofar as different provinces do make some information available, if it were implemented at a Federal level, it would be easier for people to access and locate offenders, including him.

…

I also place weight on the fact that despite information about high-risk offenders being made available on a case-by-case basis, the [appellant] could not point to any country information indicating that the publication of their information resulted in vigilantism in Canada either on an isolated or systematic basis.

85    The Tribunal was not, of course, there dealing with a regime of the same kind as that which it found may come to be implemented in Canada in the future: it was not a generally publicly available sex offender registry of the kind that the evidence indicated existed in the USA. Indeed, that difference (and in particular the difference as to likely consequences) was seemingly accepted to some extent by the Tribunal in its tentative findings as to increased risk of harm at [88].

86    But, more fundamentally, the Minister’s reliance on that part of the Tribunal’s reasoning does not answer the appellant’s case. Giving some weight to an absence of information suggesting that vigilantism had taken place in Canada following the publication of information about high-risk offenders being made available on a case-by-case basis could not, as a matter of logic, render irrelevant the evidence put forward by the appellant. As Mr Wood SC observed:

[O]ne can still rationally give weight to [that absence of evidence as regards the very different Canadian high risk offender scheme] at the same time as consider[ing] the evidence that does exist of the jurisdiction that has done the very thing that was in the offing here, which was the making publicly available of the whole sex offender registry. So I don’t need to critique, as irrational, the Tribunal’s giving of some weight to the information referred to at [paragraph] 84. I simply need to point out that it gave no consideration to the significant information that I’ve focused on.

87    I accept Mr Wood’s submission.

88    One is thus left with a factual position that is substantially similar to that which confronted Hill J in BQQ25 [2025] FCA 1279 (see [47]). The Tribunal did not give any explanation for putting aside (or perhaps rejecting) the evidence adduced by the appellant. That lack of explanation suggests a failure to engage with that material: FKP18 v Minister for Immigration and Border Protection [2018] FCA 1555, [46] (Kenny J). The appellant is left to guess as to the place (if any) those matters played in the decision. The fleeting citation of that material at [19] is mere “lip service” and not the reality of consideration. As his Honour found at [47] in BQQ25, the appellant’s complaint in this case is not “simply about the degree or quality of consideration … but an absence of consideration”.

89    For those reasons, ground 1 of the appeal should be allowed. I should note, in that regard, that while there are faint echoes of ground 1 in the submissions put below, the primary judge did not have the benefit of the argument that Mr Wood has put before me, which I have accepted.

5.    GROUND 2

90    My conclusions on ground 1 render it unnecessary to deal with the (large) questions that arise on ground 2 (which was put as a true alternative to ground 1).

91    Tempting though it is to venture into the interesting legal issues raised by that ground, the more appropriate course is to follow the prudent approach adopted by Kennett J (in BLD20 v Minister for Immigration, Citizenship and Multicultural Affairs [2023] FCA 294) and Anderson J (in AGD15 v Minister for Home Affairs [2019] FCA 896), and leave the resolution of those issues to a case in which it is necessary.

92    I should, however, say that, like Kennett J, I consider that there are some textual and structural features that point in favour of the appellant’s argument. For that reason, I have concluded, for the purposes of the extension of time application, that this ground has some merit: see [18(4)].

93    The appellant emphasised, in particular, s 368(6) of the Act which provides that, “[i]f the decision is not given to the applicant orally before the written statement is made” (which was the case here), “the decision is taken to have been made by the making of the written statement”. The appellant relied upon the addition of that provision to distinguish the current form of the Act from the provisions considered in Re Minister for Immigration and Multicultural Affairs; Ex parte Durairajasingham (2000) 74 ALJR 405 (and see also Re Minister for Immigration and Multicultural and Indigenous Affairs; Ex parte Palme (2003) 216 CLR 212).

94    Like the provisions considered in Durairajasingham, s 368(1) contains language that might be said to presuppose that a decision has been made, and then impose, subsequent to the making of that decision, an obligation to give a written explanation for the decision: Durairajasingham (2000) 74 ALJR 405, 417 [70] (McHugh J); Minister for Immigration and Citizenship v SZLSP (2010) 187 FCR 362, 387-388 [84]-[85] (Rares J).

95    As the Minister emphasised, s 368(1) provides:

If the ART makes a decision on a review, the ART must make a written statement …

96    The conditional language in the opening words (“if”) and the mandatory language following the comma (“must make”) naturally tends to connote an anterior “decision” and subsequent obligation to give reasons arising from the fact of making that “decision”.

97    However, the provisions considered in those earlier decisions did not contain an equivalent of s 368(6) of the Act. The appellant’s argument was that that provision had a substantive effect beyond its deeming language (“taken to have been made”), and, reading the statutory text as a whole, gave a different meaning to the words in subsection (1). According to the appellant, Parliament has, by the addition of s 368(6), clarified when and how a decision is made.

98    The appellant, with some force, illustrated the potential obscurities that otherwise arose by reference to the Full Court’s decision in Minister for Immigration and Citizenship v SZQOY (2012) 206 FCR 25. Building upon what had earlier been decided by Finn J in Semunigus v Minister for Immigration and Multicultural Affairs [1999] FCA 422 and the Full Court in the appeal from that decision (Semunigus v Minister for Immigration and Multicultural Affairs (2000) 96 FCR 533), the Full Court in SZQOY accepted that the making of a decision involves both the formulation of a conclusion through a mental process and translating that conclusion into a decision by an overt act “of such character as, in the circumstances, gives finality to the conclusion”: see Buchanan J at [25] (extracting Semunigus [1999] FCA 422 at [19]) and approving that statement of principle at [29].

99    That, it was said, was achieved only by communication of the decision to the parties:

[A] decision is no decision, in my opinion, until either it has been communicated to the applicant or irrevocable steps have been taken to have that done. I speak of communication to the applicant because, before the RRT, the applicant is the only party.

See SZQOY (2012) 206 FCR 25 at [26] and [29] (Buchanan J), approving the passage from Madgwick J’s reasons in the appeal in Semunigus (2000) 96 FCR 533 at [103]. See also SZQOY at [34] and [40] (Logan J), and at [50], [53] and [57] (Barker J).

100    The consequence in SZQOY was that the Tribunal was not functus officio prior to that time; and that it therefore erred when it refused to consider further material provided by the applicant after the decision record had been finalised in the Tribunal case management system but before the applicant had been notified of the decision.

101    That provided the contextual background to the insertion of s 368(6) of the Act (by the Administrative Review Tribunal (Consequential and Transitional Provisions No. 1) Act 2024 (Cth)). The insertion of s 368(6) seemingly addressed the perceived mischief that was thought to flow from that state of the law. But Parliament did not, in addressing that mischief, make the date of decision the date at which the Tribunal completed some uncommunicated internal mental process so as to enact the difficult and uncertain position rejected in Semunigus and SZQOY. Rather, it tied that date to a different objectively determinable time – being the day and time of the making of the written statement (required to be included by s 368(1)(d)), unless the Tribunal decided to give oral reasons. And all of that, the appellant argued, might be taken to suggest that it was by that overt act, and not some internal mental process, that Parliament intended a decision would be made.

102    Those submissions were to some extent supported by (and certainly not inconsistent with) the extrinsic materials: see Revised Explanatory Memorandum to the Administrative Review Tribunal (Consequential and Transitional Provisions No. 1) Bill 2024 (Cth) particularly at [438]-[439].

103    It followed, it was argued that the making of a written statement under s 368(1) is not “posterior” to making a decision. Rather, properly construed, the effect of s 368(6) is that a decision by the Tribunal is made “by” the making of a written statement.

104    The appellant also tied his argument to the fact that s 368(8) provides that the validity of a decision on review is not affected by a failure to comply with s 368(1)(d) (ie to record the day and time when the written statement was made). That was said to reinforce, by implication, that failure to comply with s 368(1)(b) does go to the validity of the decision. Kennett J noted the potential relevance of a similar provision in s 430(4): see BLD20 [2023] FCA 294 at [16].

105    The appellant submitted that the matters he advanced to distinguish Durairajasingham likewise provided a basis for distinguishing Palme (2003) 216 CLR 212 at [33], [37] and [43]-[47] (Gleeson CJ, Gummow and Heydon JJ) and [55] (McHugh J), on which the Minister also relied. In addition, the appellant observed that, unlike Palme, the provisions in issue here contain no equivalent to s 501G(4), which provided that a failure to comply with s 501G in relation to a decision did not affect the validity of the decision.

106    There is some force in the appellant’s arguments, albeit that there remains a degree of difficulty in the language of s 368(1). But, as I have said, I express no conclusion on those issues (beyond accepting that the appellant’s argument has some merit for the purposes of the extension of time application). Their resolution ought to wait for a matter in which it is necessary to decide them.

6.    CONCLUSION

107    The appeal should be allowed with costs.

108    I did not understand it to be disputed that Mr Wood SC had entered into a costs agreement with the appellant that meets the description in r 4.19(2) of the Rules. In his written submissions, the appellant submitted that, if the costs order he sought was made, I should order that the Minister pay those costs directly to Mr Wood SC pursuant r 4.19(3) of Rules. I make that order.

Associate:

Dated:    27 May 2026