Anxious Scrutiny in Hong Kong Administrative Law: Origins, Adaptations, and Techniques

Abstract

‘Anxious scrutiny’ has become one of the most used terms within the lexicon of judicial review throughout the common law world, including Hong Kong, yet surprisingly remains understudied in the scholarly literature. In contrast to the considerable body of literature on substantive review of discretion in relation to proportionality and Wednesbury unreasonableness as rival standards of review, there is still much to explore in relation to the foundation, purpose, and techniques of anxious scrutiny review, including how the concept may have come to depart from its English roots in other common law jurisdictions. Using Hong Kong as a case study, this article examines how anxious scrutiny has been received in an Asian common law jurisdiction, considering both the scope of application and the techniques used by judges under this standard. Through a detailed examination of the case law, it traces the origins and evolution of the standard and its future role within the sliding scale of substantive review within Hong Kong's system of public law.

Introduction

‘Anxious scrutiny’ has become one of the most used terms within the lexicon of judicial review throughout the common law world, including Hong Kong, yet surprisingly remains understudied in the scholarly literature.¹ In contrast to the considerable body of literature on substantive review of discretion in relation to proportionality and Wednesbury unreasonableness as rival standards of review, there is still much to explore in relation to the foundation, purpose, and techniques of anxious scrutiny review, including how the concept may have come to depart from its English roots in other common law jurisdictions.

In this respect, Hong Kong provides an interesting case study in the use of anxious scrutiny review. In 2004, Hong Kong's Court of Final Appeal (CFA) first recognised the standard as applicable in refugee screening decisions.² Since then, both the number of judicial reviews and those that apply anxious scrutiny have grown exponentially.³ Like its English counterpart, anxious scrutiny in Hong Kong has roamed beyond the sphere of refugee status determination review into other areas, including the evaluation of restrictions on unincorporated human rights treaties, immigration decisions affecting families, and the protection of cultural assets such as Victoria Harbour.⁴ The volume of cases before the Hong Kong courts, coupled with their diversity, thus commands attention for a more detailed study of the application of anxious scrutiny in this jurisdiction.

This article is structured in four parts, excluding this introduction. The first part begins with an overview of the origins of the concept in the English courts and its different spheres of application, followed by an analysis of its conceptual structure and the variety of techniques employed by reviewing judges. In turn, outlining the contours of English judicial practice on anxious scrutiny allows for comparative insights into the approaches adopted by the Hong Kong courts. Attention then shifts in the subsequent part to the applicability of anxious scrutiny review in the Hong Kong courts, including the various areas in which it has been applied. The present analysis shows the scope of application of anxious scrutiny review and the techniques used by Hong Kong judges. The following part then provides a general analysis of these developments and identifies areas where greater conceptual clarity is needed to improve the understanding and application of the concept in future cases. The final part concludes.

English Origins and Typology

The origins of ‘anxious scrutiny’ in the lexicon of the common law judge is traced to Lord Bridge's dictum in Bugdaycay.⁵ One of the appellants (Musisi) had come to the United Kingdom (UK) from Kenya and sought asylum on the basis that he was a refugee from Uganda.⁶ Musisi argued that if he were returned to Kenya, the Kenyan authorities would not allow him to re-enter the country and would instead return him to Uganda, where Musisi believed he would be killed.⁷ While the Secretary of State proceeded on the assumption that Musisi was a refugee, they did not seek to ascertain the likelihood of Kenya returning Musisi to Uganda. Lord Bridge took the opportunity to articulate the standard of review when approaching a decision of such gravity:

The limitations on the scope of that power are well known and need not be restated here. Within those limitations the court must, I think, be entitled to subject an administrative decision to the more rigorous examination, to ensure that it is in no way flawed, according to the gravity of the issue which the decision determines. The most fundamental of all human rights is the individual's right to life and when an administrative decision under challenge is said to be one which may put the applicant's life at risk, the basis of the decision must surely call for the most anxious scrutiny.⁸

The House of Lords, for its part, was not convinced that the decision-maker had devoted sufficient attention to the issues that pertained to the decision. Rather, the Secretary of State's decision was ‘taken on the basis of a confidence in Kenya's performance of its obligations under the [Refugee Convention] which is now shown to have been, at least to some extent, misplaced.’⁹ As this decision was taken without consideration of evidence to the contrary, the Secretary's decision was quashed.¹⁰

As the outcome in Bugdaycay alludes to, and as later case law confirms, both the primary decision-maker and the court are under a duty to conduct an anxious scrutiny in applicable cases.¹¹ As Buxton LJ noted in WM (DRC), the ‘consideration of all the decision-makers, the Secretary of State, the adjudicator and the court, must be informed by the anxious scrutiny of the material that is axiomatic in decisions that if made incorrectly may lead to the applicant's exposure to persecution’.¹² Anxious scrutiny is therefore not only a standard of review applied by judges (discussed below), but an orientation towards more conscientious and resourced decision-making given the gravity of the issues at stake. The implication is that the decision-maker must do more – practically speaking, by devoting more attention and resources – in order to have confidence in the soundness of the decision made. As Moses LJ noted in ML (Nigeria), the irreversible nature of the harm arising (in that case, from an adverse refugee status decision) means that the decision-maker must show that they are as ‘rigorous as possible’ and ‘carry out a careful investigation’ which must ‘remove all doubt, legitimate as it may be, as to the invalidity of a request for protection’.¹³

Lord Bridge's statement above was somewhat unclear about the scope of application of anxious scrutiny review. It referred to both the ‘gravity of the issue’ and the ‘most fundamental of all human rights’ as warranting anxious scrutiny review.¹⁴ These two concepts are distinct – a decision might gravely impact on a person even though no human rights are formally applicable – although they also overlap.¹⁵ Nonetheless, it is clear that the English case law has developed to apply anxious scrutiny in two main areas.¹⁶ The first is in relation to immigration and asylum decisions, where the relevant question is whether the decision-maker has made a decision consistent with the statutory and legal framework.¹⁷ Such decisions – exemplified in Bugdacay no less – are in substance concerned with redressing public law errors rather than with human rights as such.¹⁸ They involve questions such as whether the decision-maker had acted with a proper purpose, took into account relevant considerations, and acted rationally in evaluating the evidence.¹⁹ The second category, by contrast, has indeed been more directly concerned with the evaluation of interferences with human rights, the crucial context for most such cases at that time being that the European Convention on Human Rights had yet to be incorporated into English law.²⁰ Anxious scrutiny in this respect was a means for the court to signal some sort of departure from the classic Wednesbury unreasonableness – which set the bar for judicial intervention at the extremely high threshold of manifest unreasonableness – in evaluating restrictions on human rights.²¹

Such cases with a human rights dimension prompted further refinements of the Bugdaycay formulation. In Brind, for example, Lord Bridge noted that, when considering interferences with human rights, the court is ‘entitled to exercise a secondary judgment by asking whether a reasonable Secretary of State, on the material before him, could reasonably make that primary judgment’.²² Then came the most famous formulation on the issue, adopted by Bingham MR in Smith, where the human rights context was deemed to be ‘important’ to the court's standard of review.²³ In such cases, ‘[t]he more substantial the interference with human rights, the more the court will require by way of justification before it is satisfied that the decision is reasonable’.²⁴

Yet despite these refinements, jurists have also queried what exactly ‘anxious scrutiny’ means at a level of specificity. Lord Carnwath noted that the phrase is ‘uninformative’ and an ‘emotion based phrase’; read literally, ‘the words are descriptive not of a legal principle but of a state of mind’.²⁵ Lord Sumption also identified the problem with anxious scrutiny – conjuring up images of ‘nail-biting anxiety’ – as its lack of clarity, thereby acting as a ‘substitute for analysis’ and a front for ‘what the court is really doing and why’.²⁶ To Lord Sumption, the court should come clean and admit that it is engaged in merits review, with the intensity of that review dependant on the width of the decision-maker's discretionary area of judgment.²⁷ With these criticisms in mind, it is beneficial to look beyond the phrase to consider more specifically the type of judicial technique that has been applied, where this can be deduced from the case law. This is done here according to four themes associated with Wednesbury unreasonableness as a ground of substantive review, and thus potential sites for modification using the concept of anxious scrutiny.

The first understanding – referred to as Type 1 in this article – is that anxious scrutiny is concerned with ensuring that the authority's decision is adequately reasoned.²⁸ In this sense, anxious scrutiny is, in effect, analogous to the common law duty to give adequate reasons. This idea of anxious scrutiny allows the court to demand more from the authority in terms of justification, meaning that bland or general justifications will not so readily establish that the authority's decision is rational.²⁹ In turn, the need for more detailed reasoning means that the evidence will have to be scrutinised more closely, so that there is generally an evidentiary foundation for the decision.³⁰ It also allows the court to focus on the adequacy of the reasons stated at the time of the decision, rather than ex post facto in the context of the judicial review proceedings.³¹ This is reflected to some extent in the Smith formula: ‘the more substantial the interference with human rights, the more the court will require by way of justification before it is satisfied that the decision is reasonable’ (this statement also lends support to the thicker versions of anxious scrutiny considered below).³²

This thinner understanding of anxious scrutiny (ie, the need for adequate reasons) would also fit within the conventional understanding of Wednesbury unreasonableness, specifically concerning the court's deference to questions of weight and balance in a decision.³³ The balance between competing considerations (such as human rights and public policy) is still largely a matter for the authority, unless such balance is so untenable as to be absurd.³⁴ This is no better illustrated than in Smith itself, which concerned a ban on homosexuals serving in the army in order to uphold the putative aims of group cohesion and fighting effectiveness.³⁵ The Court of Appeal did not overtly balance fundamental rights against these putative aims.³⁶ Rather, Bingham LJ upheld the policy as rational as it had a reasoned justification, given that it was underpinned by parliamentary support and professional advice, while changes towards acceptance of homosexuals in armed forces in other jurisdictions were still in their infancy.³⁷ On this understanding, therefore, anxious scrutiny is ultimately concerned with ensuring that a decision is adequately reasoned rather than appropriately balanced.

A second understanding of anxious scrutiny – labelled Type 2 in this article – is that the courts are entitled to define, to a greater extent than under conventional Wednesbury, what considerations are relevant to a decision. The common law broadly recognises two types of considerations: those which are expressly or impliedly required by the statutory scheme (mandatory considerations), and those to which the decision-maker ‘may have regard if in his judgment and discretion he thinks it right to do so’ (discretionary considerations).³⁸ In relation to the latter, there is ‘a margin of appreciation within which the decision-maker may decide just what considerations should play a part in his reasoning process’, subject to Wednesbury unreasonableness.³⁹ Anxious scrutiny might narrow this margin of appreciation, enabling the court to insist that a particular norm or fact be considered before a decision is made.⁴⁰ This seems to echo the oft-cited dictum of Lord Carnwath, who described anxious scrutiny as ‘the need for decisions to show by their reasoning that every factor which might tell in favour of an applicant has been properly taken into account’.⁴¹ Indeed, there have been numerous cases where the court has quashed a decision on the basis that the decision-maker had failed to give proper consideration to material evidence favourable to the persons concerned.⁴² Still, in the fundamental rights context, anxious scrutiny did not allow the Court of Appeal in Smith to pierce the dualist veil and oblige the decision-maker to take into account the then unincorporated ECHR.⁴³ By contrast, the House of Lords in Launder, under the guise of an anxious scrutiny review, was able to consider whether the decision-maker had ‘correctly [taken] into account the scope and content’ of the ECHR where it had chosen to take that instrument into account in reaching its decision.⁴⁴ The general point here is that this understanding of anxious scrutiny potentially, in the circumstances of a particular case, gives the court more power to determine what is relevant to a decision than under conventional Wednesbury review.

A third perspective on anxious scrutiny – Type 3 here – allows the court to also make an assessment of the relative importance of considerations in a decision. Courts generally avoid such assessments. As Lord Keith noted in Tesco Stores, the weighing of considerations is ‘entirely for the decision maker’, and ‘the court will not interfere unless he has acted unreasonably in the Wednesbury sense’.⁴⁵ By contrast, an enhanced approach to relevancy review might well entitle the court to specify the importance of considerations to a decision. For example, Lord Woolf MR in Saville, in applying the anxious scrutiny approach, observed that the relevant decision-maker had failed to attach ‘sufficient significance’ to the right to life when removing the anonymity of soldiers involved in the 1972 ‘Bloody Sunday’ massacre, given the possibility of reprisals against the soldiers.⁴⁶ Lord Woolf underscored that ‘when a fundamental right such as the right to life is engaged, the options available to the reasonable decision-maker are curtailed’,⁴⁷ as it is unreasonable to reach a decision that contravenes or could contravene human rights unless there are sufficiently ‘significant countervailing considerations’.⁴⁸ Compared to earlier iterations of anxious scrutiny (Smith), this approach is being more explicit about the importance of fundamental rights and indeed the need for the decision-maker to give them substantial weight. It advances an approach that moves away from the agnostic stance towards the relative importance of relevant considerations typical under the classic understanding of Wednesbury unreasonableness.

Finally, the thickest understanding of anxious scrutiny – labelled Type 4 in this article – is to treat the concept as lowering the threshold of Wednesbury unreasonableness, primarily by introducing a simple unreasonableness or proportionality test.⁴⁹ A simple unreasonableness threshold is indeed alluded to by Lord Bridge in Brind (as above), in focusing on ‘whether a reasonable Secretary of State, on the material before him, could reasonably conclude that the interference with freedom of expression was justifiable’.⁵⁰ On this understanding, it is not necessary to establish a decision to be ‘extremely’ unreasonable; mere unreasonableness will suffice. Unfortunately, the case law has not always been consistent or clear as to whether a lower threshold has been applied, this being a question of interpretation of the case law.⁵¹ Another point of debate in this regard is the extent to which anxious scrutiny has come to resemble proportionality review in cases involving human rights, particularly given that the test formulated in Smith and Saville (above) seems to suggest a role for the court in evaluating and balancing countervailing considerations.⁵² Some judges have, in turn, suggested that the division between Wednesbury unreasonableness and proportionality should not be exaggerated and in certain respects has been narrowed.⁵³ Another – more extreme – construction of anxious scrutiny is that it allows, in certain cases, a review of whether the public authority has made the correct decision, rather than merely a decision within a spectrum of permissible decisions (so called ‘correctness’ review).⁵⁴ More recently, Lord Sumption in Pham noted that how broad the range of permissible rational decisions is ultimately depends on the circumstances of a given case, taking into account the significance of the right interfered with, the degree of interference, and the court's competence ‘to reassess the balance’.⁵⁵ On this basis, it may be that the ‘range of rational decisions is so narrow as to determine the outcome’ – as was indeed the case in Saville (above).⁵⁶

Mapping Anxious Scrutiny Practice in Hong Kong

Before setting out the practice of anxious scrutiny in the Hong Kong courts, it is instructive to situate it in the wider context of the major standards applied by judges in reviewing executive action. Apart from anxious scrutiny, there are two other major standards of review: Wednesbury unreasonableness and the commonly formulated four-stage proportionality test.⁵⁷ The Wednesbury unreasonableness test, mandating a judicial remedy to address a manifestly unreasonable decision, remains the default standard of review in Hong Kong common law, absent the applicability of a statutory or constitutional source to justify the use of a proportionality review.⁵⁸

The limitations of review under classic Wednesbury unreasonableness have two main implications that are relevant to the proceeding analysis. The first concerns the ‘daunting task’ for the claimant of establishing that an executive decision is Wednesbury unreasonable, with the burden of justification on that claimant.⁵⁹ The flip side is that the test is conventionally undemanding for decision-makers, who need only establish that their decision is not so unreasonable, irrational, outrageous, or perverse (all of which adjectives convey the same sentiment of an extremely wrong decision).⁶⁰ The second relevant point here concerns the approach under classic Wednesbury unreasonableness to the review of what considerations are relevant to a decision and the weight to be given to them. As generally noted in the ‘English Origins and Typology’ section, and as recognised in Hong Kong, the court will not generally dictate or assign weight to a (discretionary) relevant consideration; rather, it is for the primary decision-maker to decide what is relevant to its discretion or the weight to be given to a consideration.⁶¹ It follows that, under the classic Wednesbury unreasonableness test, the decision-maker is not obliged to consider everything that the claimant considers relevant to support a decision in their favour, unless it would be manifestly unreasonable to disregard such considerations.⁶²

With this context in mind, this section examines the scope of application of anxious scrutiny review as well as the type employed by the Hong Kong courts. It identifies four main areas in which the test has been applied, with some variation in the types.⁶³

Refugees Status Decisions

As with anxious scrutiny review in the English courts, the first substantive articulation of this standard in Hong Kong was in the context of the review of a refugee status decision.⁶⁴ Since then, there have been several thousand cases in which anxious scrutiny has been applied to the review of refugee status decisions.⁶⁵ In fact, given that judicial review applications have been dominated by asylum seekers challenging adverse decisions, anxious scrutiny has become (at least for the time being) the most commonly used standard of review in the Hong Kong courts – more so, that is, than classic Wednesbury unreasonableness and the proportionality test found in constitutional law.⁶⁶ Space precludes a detailed consideration of all anxious scrutiny cases, including the success rates of claimants in comparison with classic Wednesbury unreasonableness, so the focus here is instead on the principal contours of this standard as it has been substantively articulated and applied in the judicial review of refugee status decisions in Hong Kong.

The first substantive consideration of the anxious scrutiny standard came in the CFA's 2004 landmark decision of Prabakar.⁶⁷ Prior to this case, the standard was barely mentioned in the case law.⁶⁸ In 1994, the Court of Appeal rejected the use of anxious scrutiny in refugee screening cases, finding that in England the applicant had only a ‘one shot’ chance to persuade the authority that they feared persecution; it was therefore especially important for the court to engage in closer scrutiny, unlike in Hong Kong where applicants had the opportunity to appeal to an administrative tribunal.⁶⁹ In 2000, further doubt was cast on the applicability of anxious scrutiny in Bahadur.⁷⁰ In Hong Kong, both the Basic Law and the Hong Kong Bill of Rights Ordinance (Cap 383), the two constitutional instruments that entrench human rights in the territory, generally exclude those without the right to remain from exercising constitutional rights to challenge immigration decisions.⁷¹ Given that judges have associated anxious scrutiny to the protection of fundamental human rights (see Bugdacay and Smith above), the Court of Appeal accordingly expressed ‘serious doubts’ that this more intensive standard applied to judicial review of immigration decisions.⁷² But any such ambiguity as to its applicability in Hong Kong was dispelled by the CFA in Prabakar, where Chief Justice Andrew Li adopted a formulation redolent of Lord Bridge's in Bugdacay: ‘the courts will on judicial review subject the Secretary's determination to rigorous examination and anxious scrutiny to ensure that the required high standards of fairness have been met’.⁷³ The context here was the power to deport, which in the relation to a torture non-refoulement claimant was of ‘momentous importance’ because the claimant's ‘life and limb [were] in jeopardy’ and the ‘fundamental right not to be subjected to torture’ was at stake.⁷⁴

The decision-making process, which was itself subject to challenge in Prabakar, provides an insight into the particular meaning given to anxious scrutiny by the CFA. The Hong Kong Special Administrative Region (HKSAR) had been a party to the Convention Against Torture (CAT) since 1992 which, pursuant to its Article 3(1), prohibits the return of an individual to another State where there are substantial grounds for believing that they would be at risk of being subjected to torture.⁷⁵ However, the Convention had not been incorporated into Hong Kong law, and the executive had also avoided involvement in the screening of torture claimants. Instead, the HKSAR would allow individuals to enter the territory to make a claim to the Hong Kong sub-office of the United Nations High Commissioner for Refugees (UNHCR).⁷⁶ The UNHCR would then determine whether the applicant met the definition of a refugee; crucially, the Director of Immigration, having played no part in assessing the refugee claim, would exercise his discretion to deport the torture claimant solely on the basis of the UNHCR's rejection decision (which, when conveyed to the Director, did not state the reasons for such rejection).⁷⁷ The CFA in Prabakar set out four understandings of anxious scrutiny, developed further in subsequent jurisprudence, to establish that the process adopted by the Director of Immigration was flawed.

First, as Chief Justice Li's statement above makes clear, anxious scrutiny was not limited to substantive review, but also applied to review of the fairness of the decision-making procedure, encompassing factors such as the absence of oral hearings and legal representation. Indeed, a key purpose of the Prabakar litigation was to ensure greater procedural fairness for applicants, given the UNHCR's inability to provide this effectively.⁷⁸ The use of anxious scrutiny in conjunction with ensuring a ‘high standard of fairness’ led to both structural changes and individual successes in relation to the procedure used to screen asylum claims, creating a strong presumption in favour of procedural safeguards such as, as mentioned above, the provision of an oral hearing.⁷⁹ At a structural level, the Court of First Instance (CFI) held in FB v Director of Immigration that asylum seekers should be provided with legal aid and legal representation before status determination officers; the denial of legal aid meant that the authority had effectively denied asylum claimants the right to legal representation where they could not afford such assistance.⁸⁰ In FB, the CFI also formed a view on the lack of training provided to the Secretary; it required ‘appropriate’ training for the procedure to be considered fair.⁸¹ As Swati Jhaveri has noted, FB was a remarkable judgment for its intensive consideration of the organisational components of a fair procedure, venturing into areas not usually considered in procedural fairness reviews, including the adequacy of training of decision-makers.⁸² The implication here is that anxious scrutiny provides a signal for more intensive, structural level fairness review.

Second, another key principle expounded by the CFA in Prabakar and developed in subsequent cases was the Director of Immigration's own duty to engage in anxious scrutiny (and not just that of the court).⁸³ It is not sufficient for the Director to rely solely on the UNHCR's unexplained rejections in exercising his discretion to remove the migrant; rather, the Director must independently assess the circumstances of the claim to satisfy himself that the rejection has been properly substantiated.⁸⁴ The Director also has a duty, albeit not an absolute one, to proactively conduct an investigation and obtain relevant information on country conditions in relation to the circumstances arising in a case, and to make reasonable enquiries into a claimant's answer or omission.⁸⁵ In doing so, ‘it would not be appropriate for the Secretary to adopt an attitude of sitting back and putting the person concerned to strict proof of his claim’.⁸⁶ Rather, it may be ‘appropriate for the Secretary to draw attention to matters that obviously require clarification or elaboration so that they can be addressed by the person concerned.’⁸⁷

Third, and relatedly, the CFA articulated a set of relevant considerations that the Director of Immigration is obliged to consider when deciding whether to remove a torture claimant, including, for example, the obligation to look for evidence of torture and to examine the country conditions (ie, Type 2 anxious scrutiny).⁸⁸ Under the guise of anxious scrutiny, the nature of the CFA's decision thus delved into ‘advisory opinion’ type reasoning, where the court set out matters to guide the future construction of a decision-making procedure that, strictly speaking, went beyond the issues to be decided in the appeal before it. Furthermore, in order to satisfy the court that all relevant matters had been considered, the decision-makers had to demonstrate this in their reasoning. On this basis, according to the Court of Appeal in a later case, the court will consider the substance of the reasoning rather than its format, so that where a particular case requires more, an assertion by the decision-makers that they have considered all the circumstances may not be enough (ie, Type 1 anxious scrutiny).⁸⁹ It would therefore not be sufficient to provide a bland, general justification for a decision. In the context of refugee determination, the authority will have to explain how it has assessed the ‘level of risk’ to an asylum seeker's life if they were returned.⁹⁰ Indeed, this enhanced duty to provide reasons has been significant in practice, leading to refugee status decisions being quashed as unreasonable for lack of explanation.⁹¹

Fourth, the CFA in Prabakar also indicated the appropriate weight to be given by the Director of Immigration to certain considerations (ie, Type 3 anxious scrutiny). The court held that a favourable decision by the UNHCR on a claimant's refugee status should be given ‘great weight’.⁹² By contrast, a negative decision of the UNHCR should only be given ‘appropriate weight’.⁹³ This is because the UNHCR's adverse determination may be based on certain assumptions (in relation to the Refugee Convention, which is the UNHCR's remit) that are not relevant to the specific issues arising under the CAT non-refoulement principle.⁹⁴ In any event, the UNHCR's adverse determination should not be dispositive, given the Director's duty ‘to come to an independent judgment’ on the refugee status case.⁹⁵ This aspect of the Prabakar judgment is now partly obsolete, following the establishment of a unified screening mechanism by the Hong Kong executive, thereby removing the UNHCR's role in status determination (although the Director will of course still have to give ‘great weight’ to the favourable decisions of his refugee status adjudication officers). Nevertheless, it illustrates that the court is prepared to dictate the weight of a consideration to a decision-maker, more so than it would be prepared to do under the classic Wednesbury unreasonableness test.

Since Prabakar, as already noted, there have been many thousands of judgments in which the court has been involved in an assessment of the decision-maker's exercise of substantive discretion. It is impossible to provide a comprehensive evaluation of all these cases here, although the general impression is that the overwhelming majority of these cases have led the reviewing judge to find no unreasonableness in the decision-maker's assessment and weighing of the evidence.⁹⁶ In this respect, it is perhaps ironic that, given that anxious scrutiny demands more justification, many such judgments often provide only highly formulaic and conclusory remarks as to how a decision has withstood anxious scrutiny.⁹⁷ The majority of cases thus provide no specific insight into the extent to which anxious scrutiny is seen as a departure from the high threshold of classic Wednesbury unreasonableness. In fact, there is support for the view that anxious scrutiny has not lowered the Wednesbury unreasonableness threshold in the context of refugee status determination. A frequently cited formulation comes from AM, where Justice McWalters noted that the ‘enhanced Wednesbury test is simply an application of the Wednesbury test by means of a rigorous examination and anxious scrutiny of the decision-making process and the reasons by which the decision-maker reached his decision’.⁹⁸ Justice Lam was perhaps more specific, stating that as long as the court is satisfied after anxious scrutiny that the process was fair and that the decision was ‘made after taking account of all relevant matters and supported by reasons (so that it is not Wednesbury unreasonable or irrational), the Court will not intervene.’⁹⁹

At the same time, there are occasional decisions which arguably did not confine themselves to redressing extremely wrong exercises of substantive discretion.¹⁰⁰ A notable example of this type is Singh v Permanent Secretary.¹⁰¹ The applicant was facing a deportation order for committing a serious crime (rape) in Hong Kong; he argued that his life would be at risk if he were returned to India, at the hands of family members of his rape victim.¹⁰² The CFI held that the authority had failed to weigh the relevant factors with sufficient rigour in the light of the serious threats to the applicant if deported.¹⁰³ Justice Reyes analysed each of the reasons given for deportation and found them to be ‘far from cogent or persuasive’.¹⁰⁴ The judge delved into the evidence to undermine the authority's evaluation, including questioning the assumption that the Indian authorities would provide the migrant with adequate security against vigilante attacks.¹⁰⁵ The judge then formed the opinion that the applicant would be safer remaining in Hong Kong rather than returning to India.¹⁰⁶ Most strikingly, the judge noted that if these factors were ‘rigorously weighed’, then the ‘only reasonable decision’ was to allow the migrant to remain in Hong Kong on compassionate grounds.¹⁰⁷ The fact that a possible outcome of anxious scrutiny review is that there is an ‘only reasonable decision’ is, of course, far removed from the classic Wednesbury unreasonableness orthodoxy, which erects a significant zone of immunity for the decision.¹⁰⁸ Nonetheless, it suggests that the review was approached on the basis of a lower threshold than the classic formulation of extreme unreasonableness, even though the judge concerned did not explicitly acknowledge that he was applying a lower threshold of unreasonableness.

Refugee Work Authorisation Decisions

The use of anxious scrutiny review has been extended to decisions concerning mandated refugees seeking permission to work in Hong Kong pending their resettlement to a safe third country. The Director of Immigration adopted a restrictive approach to work authorisation requests, requiring the mandated refugee to demonstrate ‘exceptional circumstances’.¹⁰⁹ A group of mandated refugees sought to challenge this restrictive approach on the grounds that it does not adequately take into account their human rights and that it sets the threshold so high for permission that the policy is Wednesbury unreasonable.¹¹⁰ Yet, as they could not directly rely on domestic sources of human rights (as such instruments are generally excluded from immigration decisions), the refugees instead attempted to use unincorporated human rights treaties as a benchmark to test the rationality of the Director's work authorisation policy and its implementation.¹¹¹

In MA v Director of Immigration, the CFI surveyed English case law on anxious scrutiny (see above) and endorsed Lord Bingham MR's formulation in Smith that ‘the more substantial the interference with human rights, the more the court will require by justification before it is satisfied that the decision is reasonable’.¹¹² Justice Cheung (now the Chief Justice) then noted some salient developments in the substantive review of administrative discretion. According to the judge, under anxious scrutiny the burden of argument shifts to the authority, with the court being less inclined to accept ex post facto justifications.¹¹³ Justice Cheung thus appeared to define anxious scrutiny in ostensibly minimal terms, as a test concerned with the justificatory burden, rather than one that empowers the court to play a more prescriptive and evaluative role in relation to the decision at issue as against an external standard (ie, unincorporated human rights).¹¹⁴ This understanding was further confirmed in Justice Cheung's reasoning as to the context in which anxious scrutiny arose in England (as an attempt to give some effect to unincorporated human rights), meaning that the decision-maker is not required by law to act in accordance with the unincorporated right as such; nor can the court ‘require him to do so’.¹¹⁵ Accordingly, the anxious scrutiny approach did not entitle ‘the court to tell the Director that he must take into account humanitarian or similar considerations under any or any particular circumstances when exercising his wide discretions’.¹¹⁶ Indeed, even under anxious scrutiny, ‘a court does not substitute its own decision for that of the decision-maker’.¹¹⁷ Anxious scrutiny was then considered in relation to two different aspects.

The first was the Director's policy of limiting work authorisation to ‘exceptional circumstances’. The applicants in MA argued that setting the threshold so high infringed upon their human rights as recognised under international law, for example by failing to give due regard to the right to work under the International Covenant on Economic, Social and Cultural Rights.¹¹⁸ However, Justice Cheung was not prepared to conclude that the policy was ‘unreasonable, even under the anxious scrutiny approach’.¹¹⁹ This was because (given Hong Kong's dualist tradition) the Director was not obliged to devise his policy in accordance with unincorporated human rights.¹²⁰ Unlike the emergence of anxious scrutiny in England (which arose, at least in part, to mitigate the absence of a statute incorporating the European Court of Human Rights), Hong Kong does have legislation incorporating human rights treaties; this legislation specifically excludes human rights protection in immigration decisions.¹²¹ It seems implicit in this reasoning that it would not be constitutionally appropriate for the court to use anxious scrutiny as a backdoor to subvert the legislature's clear choice to exclude human rights in the immigration context. Furthermore, anxious scrutiny did not entitle the court to ignore the public interest justification for a stringent policy.¹²² On the contrary, the concept requires that ‘full regard to the context’ be taken in assessing rationality, which in this case meant assessing the need for strict immigration control as ‘an important, if not overwhelming, justification’ for the stringent policy.¹²³ Anxious scrutiny, then, could not be used to give greater weight to unincorporated human rights than to the authority's need to maintain strict immigration control.

This leads to the second aspect, which concerns the application of the restrictive work authorisation policy in individual cases. Although anxious scrutiny did not empower the court to impugn the policy for its failure to comply with international human rights law, it did give it a greater role in ensuring that the authority implemented its policy correctly in individual cases (similar, in this respect, to the House of Lords’ finding in Launder above).¹²⁴ The policy of considering whether to allow mandated refugees to work in ‘exceptional circumstances’ thus required the Director to apply this standard conscientiously and with sufficient regard to the individual humanitarian circumstances of the individuals concerned.¹²⁵ An inference from any consideration of such exceptional circumstances, according to Justice Cheung, would require the Director to turn his attention to five relevant conditions: (1) that the mandated refugee is a vulnerable person in a foreign land; (2) that they are likely to have been stranded in Hong Kong for a substantial period of time; (3) that they may have little prospect of resettlement in the foreseeable future; (4) that they have no choice but to remain in Hong Kong until resettlement; (5) and that all this leads to a risk that enforced unemployment will be detrimental to mental health.¹²⁶ The judge was therefore adding texture to the meaning of the Director's ‘exceptional circumstances’ policy in the context of refugee work authorisation decisions. Thus, the judge was prepared to use anxious scrutiny to construct a closer supervision of such future decisions through the specific articulation of relevant considerations flowing from the Director's policy, although a much stricter supervision that would have come through the imposition of considerations from international human rights law was rejected, as noted above.¹²⁷

The dual nature of anxious scrutiny – as a standard of review but also a set of duties on the authority – is also apparent in MA. Justice Cheung rejected the government's contention that the above five considerations were only relevant where the refugee applicant specifically raises them when making a work authorisation request.¹²⁸ This would be true ‘in a normal case’, rather than one concerned with acutely vulnerable individuals: their vulnerability must be recognised so that ‘proactive care be taken to avoid missing anything in their favour’.¹²⁹ This proactive, conscientious approach requires the Director to look into the each applicant's personal file to ascertain how long they had been ‘stranded’ in Hong Kong in order to see whether they are an exceptional case.¹³⁰ A conscientious approach would also require the authority to furnish sufficiently detailed reasons (where an application is denied) as to why the individual circumstances of each applicant do not amount to such ‘exceptional circumstances’.¹³¹ On the court's part, anxious scrutiny review will entitle it to ensure that the authority has proactively addressed the individual circumstances of each refugee. Accordingly, the court will ‘examine the record and evidence carefully to see whether the Director has really done so conscientiously’.¹³² Similarly, where reasons are inadequate, the anxious scrutiny approach would entitle the court to disregard ex post facto justifications arising during the judicial review proceedings and thus more readily find unreasonableness from the date of the decision.¹³³

In this regard, it is apparent that anxious scrutiny was of some assistance to the applicants in a manner that classic Wednesbury unreasonableness would not be. Although anxious scrutiny did not entitle the court to give greater prominence to unincorporated human rights, it did place additional obligations on the authority. Whereas the onus would conventionally be on the applicant to establish their humanitarian grounds for being able to work, under anxious scrutiny the authority shares with the applicant the burden of establishing whether there are exceptional circumstances for granting a work licence. Similarly, the court's willingness to construct a set of relevant considerations as necessarily flowing from the Director's policy also speaks to the value of anxious scrutiny requiring justification against a benchmark of considerations, even where, as here, the policy articulated (‘exceptional circumstances’) was very general and highly discretionary.¹³⁴

Immigration Decisions Affecting Family Unity

Another area where there has been judicial consideration of the scope of anxious scrutiny (but with less success) has been in relation to decisions impacting upon family members. This has arisen specifically in the immigration context, where a decision has been made against an individual that will result in their separation from other family members resident in Hong Kong. Here, as above, constitutional rights do not apply in relation to immigration decisions, so litigants have used anxious scrutiny as a creative alternative to fill the constitutional void – with mixed results.

This issue arose in the Comilang and Pagtama cases.¹³⁵ The Director of Immigration refused to grant the applicants (former foreign domestic helpers) an extension of stay even though they had children by local men with the right of abode in Hong Kong. Under Hong Kong law, a child born to a parent who has the right of abode also enjoys that right. The applicants therefore sought to remain on the basis that their children had the right of abode in Hong Kong and that their removal would therefore deprive them of their family life (mother and child) in the territory.¹³⁶ The applicants argued that the Director's decision should be subject to anxious scrutiny, which they argued required the Director to consider and give weight to their family life.¹³⁷ However, two differently constituted CFIs gave slightly differing answers to the question of anxious scrutiny.

In Comilang, the CFI rejected this invitation, distinguishing this context from that arising in refugee decision-making (see above).¹³⁸ The key difference, according to Justice Lam, was that refugees were not seeking to remain lawfully in Hong Kong on a permanent basis (they were in the territory pending resettlement to a safe third country), unlike the applicant in Comilang. ¹³⁹ Thus, anxious scrutiny did not apply because the decision pertained to the Director of Immigration's discretion to allow the applicant to remain in Hong Kong, an area which was considered to fall within the Director's wide discretion to maintain immigration control.¹⁴⁰ As a result, the CFI refused to impose a duty on the Director to consider the impact on family life as a relevant factor.¹⁴¹

Somewhat in contrast (though not in outcome), Justice Au in Pagtama undertook a deeper analysis of the sliding scale of review that he said now prevails in Hong Kong common law.¹⁴² He identified two ends of the spectrum: conventional Wednesbury, where no rights are engaged, and proportionality, where they are.¹⁴³ ‘In between’ these poles, the court will ‘review with increasing vigilance a subject decision which has increasingly grave and adverse impact on the affected person's interests’.¹⁴⁴ In doing so, the court will look at the reasons and all the matters taken into account ‘to see if there is Wednesbury unreasonableness in that decision, including for example, whether certain matters or factors should or should not be taken into account as a matter of relevance’.¹⁴⁵ Although the applicant had not invoked a family right, Justice Au accorded a ‘more vigilant review’, which meant ensuring that the Director had factually considered the applicant's family circumstances and potential hardship (which, on balance of the evidence, he had).¹⁴⁶ Yet, while Justice Au appeared to apply elements of Types 1 and 2 anxious scrutiny (see above), the judge rejected the proposition that the Director had to proactively inquire into possible factors that might support granting the applicant permission to stay, such as the degree of hardship if she and the child had to leave Hong Kong.¹⁴⁷ Thus, Justice Au adopted a much weaker variant of anxious scrutiny than, for example, in refugee status determination and work authorisation cases (see above), where the authority had to proactively inquire into various circumstances. Moreover, Justice Au still considered the high threshold of classic Wednesbury unreasonableness to be applicable (rather than a modified form of simple unreasonableness), thereby reducing the potency of this review even further.¹⁴⁸

Comilang and Pagtama were subsequently consolidated into a single appeal, during which justices in the Court of Appeal briefly presented divergent interpretations of the nature of anxious scrutiny review. The appellants’ complaint was that Justice Au had not actually applied anxious scrutiny (or a strong enough type of it for their liking).¹⁴⁹ Justice Cheung (as he was at the time) took a narrow approach, stating that anxious scrutiny would only apply where ‘non-derogable and absolute rights’ were engaged.¹⁵⁰ In turn, Justice Cheung cautioned that anxious scrutiny cannot be used to allow human rights to ‘overwhelm other considerations of the Director’ in split-family decisions.¹⁵¹ Perhaps more broadly, Justice Poon (albeit with some ambiguity) endorsed Justice Au's approach, which he said ‘corresponded well with the degree or gravity on the impact on the individuals’.¹⁵² Accordingly, Justice Poon examined whether the Director had factually considered all family circumstances and conjured with Justice Au that he had.¹⁵³ The difference between these judicial approaches in the Court of Appeal – premising anxious scrutiny on absolute/non-derogable rights versus a broader conception of individual impact – is discussed further in the section ‘Courts’ Role under Anxious Scrutiny’ below.

The scope for a sliding scale of anxious scrutiny has also been tested in the area of dependent visa decisions, where a migrant has sought to enter or remain in Hong Kong as the dependent of a family member. Some of these cases have been concerned with an applicant's failure to fully satisfy an element of the dependent visa policy, in particular the ‘no known record of detriment’ requirement.¹⁵⁴ The claimants’ arguments in these cases have thus focused on the importance of their family unity on the one hand, and the correspondingly lesser importance of the criminal offence used to disqualify them from a dependent visa on the other. In turn, claimants have invoked anxious scrutiny to engage the court in a weighting of relevant considerations in a dependent visa decision. This approach also sought to reduce Wednesbury unreasonableness to simple unreasonableness.¹⁵⁵ There have been isolated successes with this approach, such as when Justice Zervos in BI v Director of Immigration quashed a dependent visa decision on the basis that the criminal offence of working illegally in Hong Kong had been given too much weight against the importance of family unity.¹⁵⁶ However, this position remains an outlier and was denounced on appeal by the Court of Appeal, finding that Justice Zervos had stepped into the decision-maker's shoes to conduct a ‘balancing exercise’ and assigning too much prominence to family life.¹⁵⁷ In general, courts have affirmed that it is for the authority ‘to decide how much weight should be given’ to matters such as criminal convictions.¹⁵⁸ Unlike Prabakar, where possible torture was at issue, an adverse dependent visa decision has not been held to be of comparable gravity to warrant anxious scrutiny review.¹⁵⁹

Planning Decisions

Outside of the context of asylum and immigration, the courts have also considered the scope of anxious scrutiny in relation to the protection of Victoria Harbour from reclamation. The executive had decided that further reclamation was required by the public need for more land, thereby prompting a series of judicial reviews. Such a decision would conventionally fall within the province of classic Wednesbury unreasonableness as a planning decision.¹⁶⁰ However, the legislation identified Victoria Harbour as a ‘special public asset’ and set a presumption against its reclamation.¹⁶¹ The ambiguity, however, was that the ordinance only stipulated that the relevant officers ‘shall have regard’ to the reclamation principle, leaving open for debate the extent to which Wednesbury unreasonableness should be modified in light of this enactment.¹⁶² The executive argued that the ordinance merely created a mandatory consideration for them to consider and weigh at their discretion; hence, as long as their decision was ‘rational and intelligible’, there was no grounds for judicial intervention.¹⁶³

In the first such case before the CFI, Justice Chu interpreted the presumption against reclamation as entailing three mandatory considerations: (i) there must be a compelling, overriding, and present public need which clearly outweighs the public need to protect the harbour; (ii) this need must be demonstrated on the basis of clear, cogent, and objective evidence; and (iii) any reclamation must be the minimum necessary (ie, proportionate).¹⁶⁴ In turn, a decision that was not ‘properly considered or attempted viable alternatives’ would be irrational.¹⁶⁵ Similarly, ‘a decision not based on an objectively demonstrated need but upon a subjective perception or a policy preference will be arbitrary and irrational’.¹⁶⁶ In short, the judge was outlining the basis for Types 1 and 2 anxious scrutiny review by requiring a greater evidential justification and prescribing considerations that the authority must proactively take into account. In applying this test, the judge found deficiencies in the decision-making process in that the need for reclamation was not objectively demonstrated and was based solely on an assertion of public need.¹⁶⁷ The CFA substantively agreed with this position (with some tweaks to the mandatory considerations) when considering the appeal.¹⁶⁸ In addition, in agreeing with the CFI's statutory interpretation of the presumption against reclamation, the CFA noted that the onus was on the authority to have ‘cogent and convincing’ material to rebut the presumption (ie, Type 1 anxious scrutiny).¹⁶⁹ Conversely, it would not be sufficient for the authority simply to assert, without evidence, that there was an overriding public need for harbour reclamation, as the burden to establish this would be on them.¹⁷⁰

The authority then commissioned an engineering report to show that reclamation was needed to satisfy the three factors identified by Justice Chu (above). The second judicial review, in turn, tested the degree to which the court would extend Wednesbury even further to assess the weight of this evidence and balance it against the importance of Victoria Harbour as a special public asset (ie, Types 3 and 4 anxious scrutiny).¹⁷¹ The CFI shied away from this. In setting the level of scrutiny, Justice Hartmann indicated that an intermediate approach between classic Wednesbury and anxious scrutiny was required in fundamental rights cases: ‘something more rigorous than the standard Wednesbury test is required although, in my judgment, the level of anxious scrutiny that must be applied when there is a substantial interference with a fundamental human right would be to set the test too high’.¹⁷² This was in the context of the statement in Mahmood that classic Wednesbury was inappropriate in fundamental rights cases because it did not allow the court to judge the ‘relative weight’ of the factors taken into account.¹⁷³ Justice Hartmann thus seemed to be implying that the court would not use Types 3 and 4 of anxious scrutiny review in assessing harbour reclamation decisions, as these tools are too intrusive into the executive's decision-making in planning matters. Indeed, the judge's reasoning tended to emphasise factors on the side of the authorities in favour of their autonomy and the public interest balancing they must undertake.¹⁷⁴ Justice Hartmann also did not delve into the engineering report and the degree to which it established the need for reclamation.¹⁷⁵ Ultimately, the result was that the mere presence of an engineering report that addressed the three tests set out by Justice Chu established that the reclamation decision was rational; it showed that the decision-maker had acted on expert recommendation on the need for reclamation.¹⁷⁶

Courts’ Role Under Anxious Scrutiny

The above judicial practice reveals a limited use of the anxious scrutiny concept within Hong Kong courts, both in relation to its scope of application and the types of techniques employed under this concept.

In relation to scope of application, two approaches have been referenced: one broader, the other narrow. A narrow approach – as applied in Prabakar and MA – was to apply anxious scrutiny to fundamental rights of ‘momentous importance’, which were subsequently defined in even narrower terms by Justice Cheung in Comilang to mean non-derogable and absolute rights.¹⁷⁷ A broader approach – evident in Justice Au's dictum in Pagtama – premises the applicability of anxious scrutiny on the existence of a decision of grave importance to an individual: ‘the court would review with increasing vigilance a subject decision which has increasingly grave and adverse impact on the affected person's interests (but short of referable human rights) to see if that decision should be quashed’.¹⁷⁸ Whether a broad or narrow approach should be adopted has yet to be answered authoritatively by the CFA or indeed the Court of Appeal; it remains an open question whether they would be prepared to extend the concept beyond the facts of Prabakar. The arguments in favour of this narrow approach would focus on the benefit of determinacy; confining anxious scrutiny to fundamental rights – or, better still, non-derogable/absolute rights – creates a bright-line and workable rule, avoiding the uncertainty and scope for misplaced judicial activism in situations where anxious scrutiny is based on a vaguer notion of the gravity of a decision on an individual. It also has the advantage of being tied to a body of human rights law that has become increasingly specific, both internationally and comparatively.

Nevertheless, there are good reasons to support the broader application of anxious scrutiny. First, it aligns descriptively with judicial practice. As already noted, anxious scrutiny has been considered in the Hong Kong courts (as in the English courts) in relation to two types of unlawful acts: public law wrongs and human rights violations. Concerning the former, which primarily involves reviewing the proper exercise of discretionary power, references to fundamental rights have been more of a rhetorical device to prioritise a claim where an administrative decision has a grave impact on an individual.¹⁷⁹ In light of this, it seems a more honest form of reasoning to justify anxious scrutiny of public law wrongs, as Justice Au did, according to the gravity of the impact of a decision on an individual, without the need to use the mantra (and inherent limitations) of fundamental rights discourse.¹⁸⁰ Second, tying anxious scrutiny to ‘fundamental rights’ in the Hong Kong context is confounding, given that this is the description given to constitutional rights in the Basic Law.¹⁸¹ Where such ‘fundamental rights’ are concerned, it is trite that the proportionality test is applied; a stricter standard of proportionality has in turn led the courts to apply ‘strict scrutiny’ and ‘reasonable necessity’ review, rather than using the phrase ‘anxious scrutiny’.¹⁸² It may be the case that the court's references to ‘fundamental rights’ in the anxious scrutiny context pertain to common law rights (distinct from constitutional rights) or human rights from unincorporated treaties, but even then the court has generally failed to explain rigorously the origin of such rights.¹⁸³ Third, it is also apparent that the use of ‘fundamental rights’ is under-inclusive in that it excludes those interests that cannot tenably be subsumed under human rights, including for example, the protection of Victoria Harbour as a community asset (see the section ‘Mapping Anxious Scrutiny Practice in Hong Kong’). To be sure, whether a standard of review is over- or under-inclusive is a value judgement and depends on underlying preferences about the scope of judicial review. The view taken here is that limiting anxious scrutiny to ‘fundamental rights’ – and, at its most extreme, to non-derogable/absolute rights – unduly curtails judicial power and the potential for the development of a sliding scale of review in response to patterns of abuse of power and societal recognition of important interests falling short of fundamental rights.

This leads to the next issue, which concerns the types of anxious scrutiny employed in the Hong Kong courts. Anxious scrutiny, as noted above, can be disaggregated into several interrelated duties of the public authority, namely to give adequate reasons, to inquire and investigate whether an individual's claim should be approved, to consider and address those factors that are favourable to the individual, and to apply a high standard of fairness in procedural design. Much like in the English courts, anxious scrutiny has acquired a particular significance in the context of refugee decision-making, but has been seen as sufficiently adaptable to situations arising in Hong Kong, namely immigration decisions that split up families and the protection of Victoria Harbour. Anxious scrutiny has thus proved to be a versatile concept in Hong Kong public law, with the extent of the duties imposed on the authority being more or less intensive depending on the type of the decision in question. Thus, while anxious scrutiny required extensive procedural design in relation to refugee status determinations (see Prabakar and FB above), the duties under this concept were more modestly confined to obtaining a single expert report in the context of a harbour reclamation decision (see Harbour above). This variability in approach shows that anxious scrutiny is not defined by a singular feature, but involves judicial selection from a menu of possible duties to be imposed on a public authority. Greater openness about standard variability in the case law itself might have the advantage of providing a more specified and precise articulation of the different types of anxious scrutiny available and their appropriate triggers for application.

The development of anxious scrutiny also raises the question of whether there is now a sliding scale of reasonableness review in the Hong Kong common law, between the points of manifest unreasonableness and simple unreasonableness. In its 2004 judgment in Harbour, the CFA left this question open, noting that it was ‘an important question which has not been resolved in this jurisdiction’.¹⁸⁴ Unfortunately, the voluminous case law applying anxious scrutiny after the CFA's dictum has also not answered this question, with most cases rather blandly and superficially restating the Prabakar formulation (‘rigorous examination and anxious scrutiny’).¹⁸⁵ While most of these cases have been conceptually ambiguous, where the courts have defined anxious scrutiny, as already noted, they have tended to treat this concept as part of the inquiry into whether a decision was manifestly unreasonable, rather than modifying the standard as such.¹⁸⁶ On this basis, it could be said that any pressure for recognition of a sliding scale of reasonableness review will not come from the concept of anxious scrutiny as such.

This empirical reality, of course, is distinct from the question of whether the courts ought to use anxious scrutiny to embrace simple unreasonableness review. On the one hand, one might be concerned that a recognition of simple unreasonableness would lead to an exponential increase in judicial review applications: indeed, cases such as Prabakar and FB have led to a steep rise in legal challenges to refugee status decisions before the courts.¹⁸⁷ On the other hand, the central argument of this article has been that anxious scrutiny contains a menu of techniques open to a reviewing judge: it could conceivably support manifest unreasonableness review for certain types of decision and simple unreasonableness review for others (differing uses, as noted above, that can be seen in the English case law). Recognising that anxious scrutiny embraces different review techniques in turn allows for a more nuanced and tailored form of control beyond the bland and mundane slogans typically in use, such as ‘rigorous examination’.

Still, as ever, there are inevitably difficult questions about where decisions fall on the spectrum of review. It is beyond the scope of this article to articulate a theory of review here, although such an exercise is not unfamiliar to the Hong Kong courts: in fact, constitutional law review has become decidedly more structured over the past decade since the CFA's seminal Hysan decision in 2016.¹⁸⁸ In Hysan, Justice Ribeiro, writing for the court, introduced a new test for determining the appropriate standard of review in constitutional cases, on a spectrum ranging from ‘manifestly without reasonable foundation’ to ‘reasonable necessity’, with the placement on this spectrum depending on a host of prescribed factors, including the significance and degree of interference with the constitutional right in question, the identity of the decision-maker, and the nature and features of the encroaching measure.¹⁸⁹ Although arising in the different context of constitutional law review, these factors demonstrate the potential for a more nuanced consideration of the standards of unreasonableness review in the common law than has hereto been the case.

Conclusion

The purpose of this article was to provide an insight into the origins, evolution, and application of anxious scrutiny through the lens of Hong Kong judicial practice. It has shown that anxious scrutiny encompasses different techniques and spheres of application. At its core, anxious scrutiny denotes a requirement on the public authority to provide adequate reasoning and to ensure that it has considered all relevant factors that might militate in favour of the individual concerned (Types 1 and 2). It requires the authority to be conscientious in its decision-making and, in turn, to devote more resources to ensuring that it makes the right decisions. Yet, unlike the more expansive approaches taken by some English authorities, the Hong Kong courts do not appear to have generally used anxious scrutiny to dictate the weight of considerations or to lower the unreasonableness threshold (Types 3 and 4). These observations are subject to the caveat that most such judgments contain only superficial reasoning as to what anxious scrutiny entails, opening up the possibility of a covert use of Types 3 and 4 in some cases. Nonetheless, the general impression from reading the case law is that the courts have applied the thinner version of anxious scrutiny identified here.

Given this relatively conservative use of anxious scrutiny, the impact of this standard within Hong Kong administrative law should not be overstated. To be sure, the concept has stimulated wholesale change in the refugee screening mechanism, winning early structural victories for asylum seekers so that their claims are considered according to a high standard of fairness. The use of anxious scrutiny has also contributed to more accountable government through the articulation of a duty to give adequate reasons, regularly tested in judicial reviews. It has prompted the Hong Kong government to devote more resources to screening refugee claims and has placed a burden on the government to discharge its responsibilities more conscientiously and properly. The use of anxious scrutiny has also given more structure to the issues that are considered relevant to the assessment of work authorisation for mandated refugees. But it is also perhaps ironic that, despite the incantation of a ‘rigorous examination’, the largest category of cases rejected on judicial review in Hong Kong now involve the application of anxious scrutiny. In this respect, it could be said that anxious scrutiny has become a victim of its own (early) success; the sheer volume of cases since Prabakar has in turn led judges to seek to manage their jurisdiction more efficiently in relation to the review and appeal of refugee status decisions, including, for example, only quashing decisions of the reviewing judge on appeal where such decisions were ‘plainly wrong’.¹⁹⁰

Nonetheless, as this article has shown, anxious scrutiny is an adaptable concept and has a proper place in the review of decisions outside of the refugee context. It is unduly restrictive to confine anxious scrutiny to non-derogable/absolute rights (as some judges have sought to do); properly conceived, the standard is capable of being applied according to the gravity of the impact of a decision on an individual, with the various techniques explored in this article providing a means for the court to signal and adjust its supervision of public authority decisions more transparently.