The origins of judicial hostility to ouster clauses

The seminal case of Anisminic Ltd v Foreign Compensation Commission, decided back in 1969 by the UK House of Lords, heralds the beginning of contemporary jurisprudence on ouster clauses. The impugned ouster clause in Anisminic was section 4(4) of the Foreign Compensation Act 1950, which provides that ‘[t]he determination by the [Foreign Compensation Commission] of any application made to them under this Act shall not be called into question in any court of law.’ In ruling that the clause did not preclude the courts from inquiring into whether a decision made by the Foreign Compensation Commission amounted to a nullity,Footnote ¹⁶ Lord Reid famously held:

Similar observations emphasising the need for clear parliamentary language before the court's supervisory jurisdiction could be ousted were also expressed in the judgments of Lord MorrisFootnote ¹⁸ and Lord Pearce.Footnote ¹⁹ Anisminic thus establishes the principle that ‘there is (at least) a strong presumption against statutory exclusion of review by the High Court of any decision of an inferior court or tribunal treated as made without jurisdiction and so a “nullity”.’Footnote ²⁰ The effectiveness of an ouster clause, therefore, hinges to some extent on how it is drafted.Footnote ²¹

While no doubt controversial,Footnote ²² the Anisminic framework has been embraced throughout the common law world. It is said to predicate on a simple yet formidable conviction that ‘Parliament cannot entrust a statutory decision-making process to a particular body, but then leave it free to disregard the essential requirements laid down by the rule of law for such a process to be effective’.Footnote ²³ Indeed, it would be non-sensical for the legislature to circumscribe the authority of a decision-maker while denying courts the authority to determine what those limits are.Footnote ²⁴ This conviction is shared by Lord Irvine of Lairg LC in Boddington v British Transport Police, albeit in a different context, who opined that ‘it is well recognised for the maintenance of the rule of law … that individuals affected by legal measures promulgated by executive public bodies should have a fair opportunity to challenge these measures and to vindicate their rights in court proceedings.’Footnote ²⁵

These observations were recently echoed by the UK Supreme Court in Privacy International. Lord Carnwath, faced with an ouster clause in section 67(8) of the Regulation of Investigatory Powers Act 2000,Footnote ²⁶ held that an ouster clause applies ‘only to a legally valid decision relating to jurisdiction … [and that] a decision which is vitiated by error of law, whether “as to jurisdiction” or otherwise, is no decision at all.’Footnote ²⁷ Interpretation of an ouster clause cannot be treated as ‘one of ordinary statutory interpretation’, since to do so would ‘[downgrade] the critical importance of the common law presumption against ouster.’Footnote ²⁸

Reception and endurance of the Anisminic orthodoxy in Hong Kong jurisprudence

Anisminic soon found its way into Hong Kong's jurisprudence. As early as 1976, in Ng Chun-Kwan v Commissioner of Inland Revenue, the Court of Appeal affirmed ‘the correctness of the proposition’ that an ouster clause ‘does not prohibit the courts from inquiring whether a purported determination is in truth a nullity’.Footnote ²⁹ Anisminic was further endorsed without any question or resistance in Re Mr Tse Cho,Footnote ³⁰ a case decided in 1979.

The first major local decision that fleshed out the jurisprudential underpinnings of the Anisminic approach, beyond mere passing reference, is Chan Yik Tung v Hong Kong Housing Authority.Footnote ³¹ The ouster clause in that case, found in section 19(3) of the Housing Ordinance, reads: ‘[n]o court shall have jurisdiction to hear any application for relief by or on behalf of a person whose lease has been terminated under [section 19(1) of the Housing Ordinance] in connection with such termination.’ Upon a judicial review challenge brought by a cooked food hawker whose licence was revoked, Liu J, ruling against the Housing Authority, rehearsed the familiar lines: ‘[a]ll ouster clauses must be construed strictly’, for it is ‘a cardinal principle that the courts’ ordinary jurisdiction to review should not be whittled down except by clear and comprehensive language’.Footnote ³²

In articulating why ouster clauses had traditionally been met with judicial hostility, Liu J excerpted a heavily-cited passage by Sir William Wade and stated: ‘[t]here are sound reasons for the courts’ ordinary jurisdiction to supervise by judicial review to be jealously guarded’, for ‘[w]ithout which, the tribunal could be made a law unto itself.’ If an ouster clause is given unqualified effect, the tribunal ‘would virtually become a potential dictator with uncontrollable given jurisdiction, and the personalities presiding it would become sole judges of the validity of their own decisions. That, needless to say, is repugnant to a coherent legal system.’Footnote ³³

The decisions of Ng Chun-Kwan, Re Mr Tse Cho, and Chan Yik Tung were all decided before the handover in 1997, a watershed in the political and legal history of Hong Kong. On the stroke of midnight on 1 July 1997, the Basic Law, a constitutional charter spanning 160 Articles, officially finalised its solemn entrance into the legal system of Hong Kong. The Basic Law enjoys precedence over all other local legal norms, including the common law, which is preserved by virtue of Article 8 to the extent that it does not contravene the Basic Law. The common law, including all the substantive doctrines and procedural safeguards, became hierarchically subordinate to the Basic Law.

In the context of such constitutional change, which undoubtedly has and will continue to implicate and challenge the endurance of various parts of the common law, questions arise as to whether and to what extent the Anisminic principles can remain intact or will evolve under the new system. In a number of post-handover decisions, courts have maintained their grip on the Anisminic framework, comforting those who feared it would slide into irrelevance. In Gurung Bhakta Bahadur v Director of Immigration,Footnote ³⁴ the Court of First Instance affirmed the need to preserve the restrictive approach to interpreting ouster clauses, echoing ‘the jurisprudential rationale … that if a tribunal could become a law unto itself, it would move dangerously towards dictatorship.’Footnote ³⁵ The common law presumption that access to courts is not to be whittled down, except by the clearest of words, also counsels in favour of a restrictive interpretation. In Thai Muoi v Hong Kong Housing Authority,Footnote ³⁶ the same Court of First Instance noted that, in any jurisdiction ‘with separation of the powers’, ‘the right of the court to supervise the decisions of the executive is always zealously guarded against’.Footnote ³⁷ It would not be easy to override ‘the presumption of the legislative intent that decisions by the executive, the tribunal or other official are justiciable by way of judicial review’.Footnote ³⁸

In neither of these decisions, however, was the Basic Law invoked to bolster or add gloss to the Anisminic principles. That is notwithstanding the discernible alignments between key tenets of the jurisprudential bedrock of Anisminic and numerous provisions in and principles emanating from the Basic Law. For example, the right of access to courts is given the pride of place in Article 35; and the text is replete with references to and safeguards for the independent adjudication by judges of the lawfulness of executive acts and decisions,Footnote ³⁹ as well as the imperative for government organs to abide by the law.Footnote ⁴⁰ The reticence to put a constitutional gloss on the Anisminic principles represents a missed opportunity to develop and entrench them further.

On top of the Basic Law, readers may wonder whether it is pertinent to also consider the Hong Kong Bill of Rights, which incorporates the International Covenant on Civil and Political Rights, in the analysis. For all intents and purposes, the Hong Kong Bill of Rights contains a list of human rights safeguards, including the right of access to courts,Footnote ⁴¹ in a manner akin to that of the Basic Law. A comprehensive analysis would undoubtedly consider the relevance of the Hong Kong Bill of Rights. However, because the relevant human rights provisions largely overlap in language and scope, it is clearer and more concise to focus solely on the Basic Law.

The imperative to reconceptualise and develop new approaches to ouster clauses in Hong Kong

This article seeks to defend the master proposition that a reconceptualisation of ouster clause analysis, unshackled from the shadows of Anisminic and anchored in Hong Kong's post-1997 constitutional context, is necessary. Indeed, as highlighted in the introduction, a similar story has unfolded against the backdrop of macro-constitutional change in the UK and Singapore. Take the UK. The waning appeal of parliamentary sovereignty as a meta-constitutional principle and the advent of other concepts, such as the rule of law and rights-based constitutionalism, have for some time been the talk of the town,Footnote ⁴² prior to their explicit manifestation in Privacy International. Consider Lord Steyn's ‘striking’Footnote ⁴³ remarks in R (Jackson) v Attorney General: ‘The classic account given by Dicey of the doctrine of the supremacy of Parliament, pure and absolute as it was, can now be seen to be out of place in the modern United Kingdom … it is not unthinkable that circumstances could arise where the courts may have to qualify a principle established on a different hypothesis of constitutionalism.’Footnote ⁴⁴ Of course, Hong Kong's peculiar encounter with parliamentary sovereignty – a point to which this article will return – implies that it may not trail the same trajectory. That said, the UK does provide a compelling comparative case demonstrating how approach to ouster clauses may evolve under constitutional tutelage.

With the above remarks in order, this article now turns to examining the novel approaches to ouster clauses available to Hong Kong courts, beyond the most orthodox and traditional Anisminic framework, that are grounded in Hong Kong's current constitutional context.

Localised and more aggressive Anisminic approach

This section commences by examining how Anisminic itself may be recalibrated in the age of post-handover Hong Kong. Anisminic in its original and orthodox form, boiled down to its essence, is an interpretive approach, under which most of the interpretive gymnastics takes place at the statutory level without inviting constitutional scrutiny. Nevertheless, this does not categorically preclude the Basic Law from asserting its relevance in the interpretive process. Quite the contrary, the Basic Law is capable of assuming a formidable presence within the macrocontextual backdrop and forging an organic unity with Anisminic to create a ‘localised’ version of the same.

In contextualising the discussion, it is helpful to first embark on a brief exploration of the modern ouster clause jurisprudence in the UK. This involves dissecting two seminal cases decided by the Supreme Court, Cart Footnote ⁴⁷ and Privacy International. Among the many themes explored in these decisions, this discussion zeroes in on how these cases have respectively tackled the intricate interplay between two meta-constitutional principles, the rule of law and parliamentary sovereignty, the tension and symbiosis between which have informed the trajectory of the judicial approach to ouster clauses.

The first principle is the rule of law. In Cart, Lady Hale observed that ‘the scope of judicial review is an artefact of the common law whose object is to maintain the rule of law – that is to ensure that, within the bounds of practical possibility, decisions are taken in accordance with the law … and not otherwise.’Footnote ⁴⁸ Similarly, for Lord Clarke, the real question was the level of independent scrutiny required by the rule of law. He wrote that judicial review did not require ‘the same degree of scrutiny in every case. All depends upon the circumstances.’Footnote ⁴⁹ The rule of law arguably took a more prominent role in the plurality's reasoning in Privacy International. Lord Carnwath remarked that, to be consistent with the rule of law, there was a ‘strong case’ for holding that ‘binding effect cannot be given to a clause which purports wholly to exclude the supervisory jurisdiction of the High Court to review a decision of an inferior court or tribunal’.Footnote ⁵⁰ The rule of law should be given effect in a ‘sufficient and proportionate’ manner,Footnote ⁵¹ such that the scope of judicial review would not fall below what was required by the rule of law.Footnote ⁵² This was also supported by the fact that Parliament has recognised the ‘rule of law’ as a constitutional principle in section 1 of the Constitutional Reform Act 2005.Footnote ⁵³

On the other hand, the lingering influence of parliamentary sovereignty remains conspicuous in both cases. In Cart, Lord Phillips remarked that hopefully ‘the proposition that Parliamentary sovereignty requires Parliament to respect the power of the High Court to subject decisions of public authorities to judicial review … will remain academic.’Footnote ⁵⁴ In Privacy International, Lord Carnwath wrote that, in the spirit of the rule of law, ‘Parliament cannot entrust a statutory decision-making process to a particular body, but then leave it free to disregard the essential requirements laid down by the rule of law for such a process to be effective.’Footnote ⁵⁵ This apparent limitation on what the Parliament can legislate on ‘[goes] beyond a merely conceptual form of limitation in that it treats rule of law-derived requirements upon statutory bodies as matter whose supervision Parliament cannot exclude.’Footnote ⁵⁶ Similarly, Lord Lloyd-Jones agreed that ‘it is a necessary corollary of the sovereignty of Parliament that there should exist an authoritative and independent body which can interpret and mediate legislation made by Parliament.’Footnote ⁵⁷

Voluminous commentaries have since been produced deliberating on these themes and cases.Footnote ⁵⁸ Especially noteworthy is Lord Carnwath's opinion in Privacy International, which sparked intense debates surrounding the Supreme Court's readiness to impose apparent limits, under the banner of the rule of law, on the Parliament's ability to oust the High Court's supervisory jurisdiction. This stems not least from an explicit call for a reorientation from ‘the Anisminic-derived categorical approach to the scope of review based on notions of jurisdiction and nullity … to a more contextual balancing approach’,Footnote ⁵⁹ but also from Lord Carnwath's bold statement that ‘it is ultimately for the courts, not the legislature, to determine the limits set by the rule of law to the power to exclude review’Footnote ⁶⁰ – a ‘striking’Footnote ⁶¹ blow to parliamentary sovereignty. Jettisoning Anisminic-esque reasoning paves the way for a judicial approach that ‘rests not on the formal legal criterion of an error of law, but requires judges to engage directly with the substantive reasons for and against upholding an ouster clause's effectiveness’.Footnote ⁶²

The ramifications of Privacy International have yet to be fully worked out by the English courts,Footnote ⁶³ but Parliament has reacted by enacting the Judicial Review and Courts Act 2022 to insert a new section 11A into the Tribunals, Courts and Enforcement Act 2007, with the publicly declared aim of abolishing Cart-style judicial review of Upper Tribunal decisions. Of significance is that the language in section 11A is clearly and distinctly tailored to neutralise the effect of Anisminic by incorporating Anisminic-esque phraseology. It provides that a decision of the Upper Tribunal – which includes a ‘purported decision’Footnote ⁶⁴ – is ‘final’Footnote ⁶⁵ and the Upper Tribunal is ‘not to be regarded as having exceeded its powers by reason of any error made in reaching the decision’.Footnote ⁶⁶ In the first judicial review challenge involving section 11A, R (Oceana) v Upper Tribunal (Immigration and Asylum Chamber),Footnote ⁶⁷ Saini J sitting in the High Court decided to give full effect to the provision:

Some commentators argued that Oceana has ‘poured cold water’Footnote ⁶⁹ on Lord Carnwath's statements in Privacy International, especially in light of parliamentary rebuke. Others view Oceana as illustrating the proposition that it is difficult to extract ‘general constitutional rules’ about ouster clauses from individual cases and reduce the relationship between parliamentary sovereignty and the supervisory jurisdiction of the High Court to a hierarchy.Footnote ⁷⁰ Whatever the fate of the Anisminic approach in the UK, which is far from certain at this point, the key message this article wishes to bring home is that the UK provides a compelling case demonstrating how ouster clause jurisprudence may evolve under constitutional tutelage, and Hong Kong courts stand to reap substantial jurisprudential dividends by closely scrutinising and extracting insights from their counterparts’ reasoning.

The starting point is to place the constitutional configurations of the UK and Hong Kong side-by-side, a crucial part of the macrocontextual backdrop. In the UK, despite the powerful remarks uttered by Lord Steyn in Jackson and Lord Carnwath in Privacy International, parliamentary sovereignty perseveres as the bedrock of the UK's constitutional fabric. As the Supreme Court declared in Miller (No 1), this principle continues to be ‘fundamental to the United Kingdom's constitutional arrangements’.Footnote ⁷¹ Elliot offered an equally realistic account: whilst parliamentary sovereignty remains ‘a fixed point onto which we can lock, even when the constitution is otherwise in a state of flux’, it is ‘being shaped by the changing nature of the constitutional landscape within which it sits’.Footnote ⁷² That explains why the principle or doctrine of parliamentary sovereignty continues to be influential in shepherding the UK's ouster clause jurisprudence.

Contrariwise, Hong Kong, historically speaking, ‘never had a deep-rooted tradition of parliamentary sovereignty’.Footnote ⁷³ Before 1997, one of the main constitutional instruments of Hong Kong were the Letters Patents. Issued by the British Crown under royal prerogative, the Letters Patents established, among others, the Legislative Council and office of the Governor, who was authorised to, with the advice and consent of the former, make laws for the ‘peace, order, and good government of the Colony’. The colonial legislature never possessed, like the British Parliament did, unfettered power to make and unmake laws in the Diceyan sense. According to the Supreme Court of Hong Kong in 1912, the Legislative Council was ‘a strictly non-sovereign body, deriving its powers wholly from the Royal Letters Patent. Any enactment it may purport to pass which is not within the scope of the Letters Patent is made without jurisdiction, and the Courts would have no hesitation in pronouncing it bad.’Footnote ⁷⁴ A more accurate description of Hong Kong's constitutional order at the time would be what Ip called the ‘doctrine of the supremacy of the Letters Patent’.Footnote ⁷⁵

China's resumption of exercise of sovereignty in July 1997 finally and conclusively excised this doctrine from the legal fabric of Hong Kong. The Basic Law superseded the Letters Patent as the principal constitutional document. As the umpire over the entire legal order, the Basic Law explicitly prohibits the legislature from enacting any legislation that is inconsistent with its terms.Footnote ⁷⁶ This leaves little doubt that the post-handover Legislative Council ‘does not enjoy supremacy in the Diceyan sense’.Footnote ⁷⁷ As the Court of Appeal explained, ‘[a]s our written constitution, the Basic Law is supreme … The court is constitutionally tasked to protect the provisions of the Basic Law and to ensure that all governmental organs and branches, including [the Legislative Council], stay within its bounds.’Footnote ⁷⁸

Against this macrocontextual backdrop, Hong Kong courts are free from any obligation to grapple with the jurisprudential complexities posed by parliamentary sovereignty in fashioning their own ouster clause jurisprudence.Footnote ⁷⁹ They can comfortably appeal to the demands of the rule of law in order to justify a restrictive interpretive approach, and are spared the dreadful exercise of arbitrating between these two potentially conflicting fundamental precepts in UK constitutional law. Lord Carnwath's remarks in Privacy International will thus hold considerably more sway: ‘regardless of the words used, it should remain ultimately a matter for the court to determine the extent to which such a clause should be upheld, having regard to its purpose and statutory context, and the nature and importance of the legal issue in question; and to determine the level of scrutiny required by the rule of law’.Footnote ⁸⁰

Discarding the historical baggage of parliamentary sovereignty marks the first step towards reconfiguring the Anisminic framework with a local touch. Furthermore, the Basic Law itself can be called upon, not only to negate the perceived relevance of parliamentary sovereignty, but to actively reinforce and amplify Anisminic's efficacy and bite to an extent that goes beyond what had originally been contemplated by the House of Lords. The conduit through which this is achieved is the modern approach to statutory interpretation known as the ‘purposive approach’. Grounded in common law, the purposive approach requires that statutes be construed in light of their purpose and the existing state of the law,Footnote ⁸¹ which includes the Basic Law and Hong Kong Bill of Rights. The court should opt for a constructional choice that gives effect to important rule of law principles and constitutionally-protected fundamental rights and freedoms.Footnote ⁸² In essence, the common law and the purposive approach operate as a portal through which an organic synthesis between the Anisminic approach and the Basic Law becomes achievable.

Key tenets of the jurisprudential rationale of Anisminic, which include preventing the governmental body from becoming a law unto itself and safeguarding access to justice, as well as the rule of law, echo key provisions in the Basic Law, a document that is itself laden with rule of law discourse. Article 11(2) states that no locally enacted law can contravene the Basic Law, requiring the legislature to ‘bow to the supremacy of the Basic Law’.Footnote ⁸³ The right of access to courts, which mirrors the common law presumption of access to courts, is saliently inscribed in Article 35. Articles 2 and 19 vest independent judicial power in the courts, and Articles 16 and 64 enjoin the executive to ‘abide by the law’. The Basic Law, therefore, furnishes ready-made ammunition that can be invoked to bolster the restrictive interpretive approach in Anisminic. It tilts the balance further in favour of elevating ‘the level of scrutiny required by the rule of law’, and transforms Anisminic from being a means merely to protect the common law presumption of access to courts to a device that can be deployed to defend the fundamental right of access to courts. In sum, the Basic Law adds a fervent constitutional gloss to Anisminic, augments its aggressiveness unhampered by the shackles of parliamentary supremacy, and consequently amplifies its efficacy and bite.

A number of other jurisdictions have ventured down the same path. The Supreme Court of Namibia, endorsing Privacy International, stated that the Anisminic approach is ‘trenchantly reinforced by the supremacy of the Constitution and the rule of law entrenched in Art 1.’Footnote ⁸⁴ The High Court of Lesotho, citing Anisminic, declared that, notwithstanding the existence of a constitutional ouster clause, the jurisdiction to examine the constitutionality of parliamentary prorogation ‘arises ex lege because of the imperatives of the rule of law and the principle of legality, which as we have earlier said, are a part of the basic features of our Constitution’.Footnote ⁸⁵ In New Zealand, litigants have relied on the right of access to justice and judicial review in section 27 of the New Zealand Bill of Rights Act 1990 to neutralise the effect of ouster clauses. In Samuels v Employment Relations Authority, claimants argued that ‘regard should be had to the New Zealand Bill of Rights Act 1990’Footnote ⁸⁶ when interpreting the ouster clause in question. While Inglis CJ found it unnecessary to rule dispositively on this issue, preferring to base her decision on Anisminic and its local counterpart, Bulk Gas,Footnote ⁸⁷ she had no difficulty finding that the reasoning in Anisminic and Bulk Gas ‘sits comfortably with the Bill of Rights Act’.Footnote ⁸⁸ This may partly be rooted in section 6 of the New Zealand Bill of Rights Act, which obliges the courts to adopt ‘a meaning that is consistent with the rights and freedoms’ in statutory interpretation.

One microcontextual factor is earmarked for final consideration. The subject matter of the litigation in question bears significance as regards the reach of the localised Anisminic approach. If defending the supremacy of the constitution is marshalled as a justification for augmenting the Anisminic orthodoxy, as the Supreme Court of Namibia did, such interpretive synergy is likely more accentuated in cases of constitutional judicial review, perhaps more than judicial review of administrative acts and decisions with little or no constitutional dimensions. This is well articulated by the Kenyan High Court: ‘Breach of fundamental rights and freedoms enshrined in the Constitution including the right to protection of the law and respect for fundamental human rights will entitle a court to intervene, notwithstanding the existence of a finality or ouster clause.’Footnote ⁸⁹

In summary, a recontextualised Anisminic approach will be unshackled from the historical burden imposed by the doctrine of parliamentary sovereignty. This frees the courts from the necessity of arbitrating between the rule of law and parliamentary sovereignty when fashioning their own ouster clause jurisprudence. Furthermore, the Basic Law, a document which echoes key tenets of the jurisprudential rationale of Anisminic and embodies the rule of law, is capable of adding a constitutional gloss to Anisminic in the orthodox form, augmenting its aggressiveness, and amplifying its efficacy and bite. These macrocontextual factors, nevertheless, do not entirely eclipse microcontextual considerations at play. Importantly, the reach of the localised Anisminic approach will depend on the nature of the precise subject matter of the litigation at hand.

Ouster clauses and the constitutional right of access to courts

The localised Anisminic approach modifies the original and orthodox iteration partly by adding to it a constitutional gloss. Yet, it refrains from taking a step further and questioning the underlying constitutional foundation of the ouster clauses themselves. Ergo, that approach remains essentially an interpretive approach. This section shifts the gaze from interpretive to constitutional approaches, which involve techniques to pierce the statutory veil and interrogate the very constitutional soundness of the ouster clauses. If an ouster clause is adjudged constitutionally offensive, it will be void ab initio to the extent of its unconstitutionality. Unless it could be rescued by remedial interpretation, such as an Anisminic-esque interpretation deemed compatible with the constitutional mandates, the clause will be excised from the statute books.

This section ventures into dissecting the first constitutional approach, rooted in the alleged incompatibility between an ouster clause and the constitutionally-protected right of access to courts. Like the recontextualiesd and updated Anisminic approach, both macrocontextual and microcontextual considerations loom large, interacting both dialogically and dialectically.

One key building block that undergirds the Anisminic orthodoxy is the imperative to safeguard the common law presumption in favour of access to courts. This presumption or ‘right’, nevertheless, remains a creature of the common law and is susceptible to override by a stroke of legislative fiat. Following the enactment of the Hong Kong Bill of Rights in 1991, as well as the coming into force of the Basic Law in 1997, this common law right was elevated and given constitutional expression.Footnote ⁹⁰ Under Article 35(1) of the Basic Law, all Hong Kong residents have the right to ‘access to the courts’ and, more specifically in Article 35(2), ‘the right to institute legal proceedings in the courts against the acts of the executive authorities and their personnel’. Bokhary PJ characterised the right as an indispensable ‘arterial right’:

Consequently, the right is ‘no longer merely a creature of the common law’.Footnote ⁹² This begets two important consequences. Firstly, it is no longer necessary for the right to be content with performing the role of a mere, albeit integral, jigsaw block in an internal interpretive scheme, geared towards lobbying for a more restrictive construction. Secondly, the right has assumed independent significance as an external check against the impact of ouster clauses. Because the right has been constitutionalised, it can no longer be dismantled by ordinary legislative fiat. Instead, the right may now be marshalled as an external normative standard to evaluate whether an ouster clause is legally and constitutionally sustainable. Derogation from the right caused by an ouster clause will only stand if it is found to be rational and proportionate.

An ouster clause stands to subvert the right of access to courts and rob Article 35 of all practical significance. Indeed, that appears to be the straightforward effect of an ouster clause, removing recourse to the courts to challenge the immunised acts and decisions. That is why an ouster clause is said to ‘serve as a signal to decision-makers that they may operate without fear and intervention by the courts at a later stage’,Footnote ⁹³ and deprive citizens of legal redress against maladministration by ‘preventing those affected by such decisions from appealing or bringing judicial review proceedings against them.’Footnote ⁹⁴ Wade and Forsyth put matters simply in the English context: ouster clauses ‘may come into conflict with this right to judicial determination’ under Article 6 of the European Convention on Human Rights, for ‘they have the effect of cutting off judicial remedies, at least in so far as the courts allow them to operate.’Footnote ⁹⁵

A prima facie restriction of the right of access to courts, however, does not automatically render the ouster clause in question void. That is because Article 35 is a qualified, rather than absolute, right.Footnote ⁹⁶ An impugned ouster clause would only fail constitutional muster if it is found to be incompatible with the principles of legitimacy, rationality, and necessity, all of which are crystallised in the Hysan four-step proportionality test. To survive unscorched, the clause must be shown to pursue and be rationally connected to one or more legitimate aims, be no more than necessary in satisfying these legitimate aims, and strike a fair balance between the burden caused by a restricted right of access to courts and the societal benefits derived therefrom.Footnote ⁹⁷ While Hong Kong courts have yet to subject an ouster clause to constitutional test under this rubric, a few actionable principles can readily be gathered from the local jurisprudence on Article 35 and comparative law. For example, both New ZealandFootnote ⁹⁸ and IrishFootnote ⁹⁹ courts have relied on structured proportionality analysis to assess the constitutional validity of ouster clauses, sparing them the ‘intellectual gymnastics characteristic of the English courts’ approach’.Footnote ¹⁰⁰

Ouster clauses are commonly justified on the basis of preventing abuse of the judicial review process. The Kenyan Supreme Court summarised the rationale as follows: ‘protecting the integrity of the relevant body, by separating it from the formal legal process; and ensuring finality, preventing unnecessary litigation, or interventionist judicial proceedings’.Footnote ¹⁰¹ In Hong Kong, the requirement to obtain leave or permission to commence judicial review proceedings was upheld as constitutional as it amounted to a ‘proportionate measure[] to prevent the abuse of its process’.Footnote ¹⁰² Refusal to grant leave, similar to an ouster clause, would operate to prevent the litigant from mounting a challenge against the impugned act or decision.

Proportionality is touted as a ‘highly intrusive’ standard of review, leaving ‘no policy consideration or interest … screened from judicial scrutiny’.Footnote ¹⁰³ In assessing whether the ouster clause constitutes a proportionate response to the professed legitimate aims, the Hysan framework amplifies, most prominently at the necessity and balancing stages, clashes between diverse public and private interests that arise in the specific statutory context, the ‘microcontextual considerations’. These microcontextual considerations, which may be trans-substantive and apply across multiple statutory contexts, encompass, among others, the nature of the subject matter of the decision being precluded from review, the interests and rights implicated by the protected decision, and the nature of the decision-making body benefitting from the ouster clause. Thus, many microcontextual considerations are firmly captured within the proportionality framework, at the behest of the macrocontextual backdrop. For instance, if the protected act or decision pertains to matters typically considered to come within the purview and expertise of executive bodies, such as national security or social welfare, there may be a stronger argument for endorsing the constitutionality of the clause and affording it a laxer interpretation.Footnote ¹⁰⁴ Likewise, social benefits yielded from the speedy resolution of a matter may outweigh individual harm in terms of procedural justice on a marginal analysis.

Once the impugned ouster clause is adjudged to be constitutionally suspect, the court is faced with two principal avenues of redress. Firstly, the court may ‘rescue’ the ouster clause by way of ‘remedial interpretation’, which involves reading down, reading in, or severing part of the clause in order to bring it in conformity with Article 35.Footnote ¹⁰⁵ This remedy enables the court to keep the ouster clause ‘formally’ constitutional while ‘substantively’ altering its contour under the guise of interpretation.Footnote ¹⁰⁶ One way of ‘reading down’ the clause summons back the Anisminic approach. This was mooted in HKSAR v Chow Hang Tung.Footnote ¹⁰⁷ At issue was section 44A(7) of the Public Order Ordinance, which purports to make ‘the determination of an appeal’ by the Appeal Board on Public Meetings and Processions, responsible for hearing appeals against police decisions to ban protests, ‘final’. The Court of Final Appeal (CFA) held that, even if the contrary intent was unequivocally demonstrated, ‘given the constitutional guarantee to access to the courts in article 35 of the Basic Law, the finality provision in section 44A(7) would still be read down to mean subject to judicial review, unless the restriction on access to the courts could be justified’.Footnote ¹⁰⁸ Ramsden concurred, suggesting that instead of invoking Anisminic directly, ‘[t]he same conclusion could have been arrived at by reading down the ouster clause by invoking Article 35 of the Basic Law (which mandates access to the court)’.Footnote ¹⁰⁹ Joseph likewise wrote that, in the New Zealand context, ‘the courts must interpret privative provisions consistently with the s 27(2) guarantee, and preserve judicial review for manifest error of law.’Footnote ¹¹⁰ In sum, remedial interpretation potentially yields an outcome that is broadly similar to Anisminic, whether in its orthodox or modified form, the key distinction being that the particular outcome is reached instead indirectly via a constitutional route.

Secondly, if the court determines that remedial interpretation is inept at redeeming the ouster clause's constitutionality, it may proceed directly to declare the clause invalid and strike it, to the extent of its unconstitutionality, from the statute books. All interpretive difficulties, whether stemming from Anisminic or otherwise, would then vanish.

This section has delved into how the constitutional right of access to courts furnishes a ‘powerful tool’Footnote ¹¹¹ to corner and reshape ouster clauses, through the well-established proportionality test. The appeal of this approach lies in the judges’ familiarity with the proportionality framework, transparency and clarity it provides in terms of the analysis, and its ability to systematically capture and incorporate a host of different microcontextual considerations. While this approach may not always precipitate a result that breaks free entirely from the Anisminic parameters, it nonetheless offers a valuable route to square ouster clause analysis with the post-handover constitutional order, a crucial macrocontextual backdrop.

Judicial power, separation of powers, and constitutional supremacy

The first reconceptualised constitutional approach discussed above is rooted in the tension between ouster clauses and the constitutional right of access to courts. This section probes into the next constitutional approach, which centres on the perceived incompatibility between an ouster clause and the interrelated principles of separation of powers, the supremacy of a written constitution or constitutional supremacy, and the allocation of independent judicial power, whether they are explicitly declared in or deduced from the constitutional text. Two points of distinction between these two approaches deserve emphasis upfront. Firstly, there are salient differences between the types of constitutional provisions that the ouster clause may be accused of subverting. The provision on the right of access to courts is an individual entitlement protected by the constitution, whereas the separation of powers, allocation of independent judicial power, and constitutional supremacy pertain to the structure of the polity and organisation of governmental organs. While the former readily lends itself to balancing against countervailing interests, it is less clear how the latter can similarly be shoehorned into a balancing exercise. Secondly, and as a result, while the first approach does not inevitably lead to the ouster clause being struck down, as the proportionality framework is ‘substantively neutral’Footnote ¹¹² and not outcome-dispositive, the second approach may more likely involve situations where the courts are bound to adhere to a priori categorical reasoning and invalidate the clause.

A helpful starting point is to consider an analogous scenario that has generated much jurisprudence, which involves the tension between a close relative of ouster clauses, ‘finality’ clauses, and the power of final adjudication provided for under Article 82 of the Basic Law. Finality clauses, as the name suggests, operate to bar appeal or final appeal to the CFA. Similar to ouster clauses, finality clauses are given a strict construction ‘to mean merely that there is no appeal’.Footnote ¹¹³ They should not be interpreted to exclude the supervisory jurisdiction of the High Court.Footnote ¹¹⁴ That being said, while finality clauses may appear to have limited reach, they do explicitly seek to strip away the power of final adjudication vested in the CFA under Article 82.Footnote ¹¹⁵ It is, therefore, instructive to explore how the CFA has traversed claims that explicitly rely on this argument and manage the tension between finality clauses and the power of final adjudication through a trilogy of cases. This will offer a more informed basis to consider whether the same approach can be applied in the context of ouster clauses.

In A Solicitor v Law Society of Hong Kong, section 13 of the Legal Practitioners Ordinance, which governs the legal profession, came under fire. The relevant parts of the section provide that, on appeal against any order of the Solicitors Disciplinary Tribunal, the decision of the Court of Appeal ‘shall be final’. The appellant complained that section 13 was in conflict with Article 82 of the Basic Law. The CFA, ruling in favour of the appellant, made the following points. Firstly, courts ‘do not have inherent appellate jurisdiction’ for appeals are ‘creatures of statutes’.Footnote ¹¹⁶ The appellate jurisdiction is distinct from the High Court's supervisory jurisdiction and the right to seek judicial review. Secondly, the power of final adjudication vested in the CFA may be limited by way of legislation. Thirdly, the justifiability of the limitation hinged on whether it passed the proportionality test. The CFA eventually held that section 13 of the Legal Practitioners Ordinance was unconstitutional and disproportionate. The reason is that, according to Chief Justice Li, section 22(1)(b) of the Court of Final Appeal Ordinance, which governs applications for leave to the apex court, has already set out an exhaustive and narrowly defined list of criteria that must be satisfied before the CFA can take up the case. Section 13 thus constituted a further limitation, one that was ‘absolute’, ‘total’, and hardly ‘reasonably proportionate to any legitimate purpose’.Footnote ¹¹⁷

Despite the shared recourse to proportionality, there is one fundamental difference between the analysis in A Solicitor and the analysis in the preceding section concerning the conflict between an ouster clause and the right of access to courts. Ribeiro PJ explained the difference in Sam Woo Marine Works Ltd v The Incorporated Owners of Po Hang Building:Footnote ¹¹⁸ ‘In most cases, constitutional challenges are founded on an applicant's claim that his or her constitutional rights have been violated. Thus, the analysis usually begins by identifying the constitutional rights engaged. However, this approach is inapplicable in the present case. This is because Article 82 of the Basic Law operates to vest the power of final adjudication in the Court of Final Appeal. It does not confer on parties to litigation any constitutional right of appeal to the final court.’Footnote ¹¹⁹ Put simply, the analysis under Article 82 cannot be equated with the analysis of constitutional rights, let alone proportionality balancing between rights and their countervailing interests.

The provisions under challenge in Sam Woo Marine Works Ltd are sections 63(1), 63A(2), and 63B of the District Court Ordinance, the combined effect of which is to restrict the ability of an appellant to take a relevant matter to the CFA. Section 63(1) provides that any appeal from the District Court in civil cases can only proceed with leave, and section 63A(2) limits the power to grant leave to situations where there is a reasonable prospect of success or some other reason in the interests of justice. Section 63B provides that no appeal lies from the Court of Appeal's decision to refuse or grant leave. The CFA did not, however, side with the party alleging unconstitutionality on this occasion and found that the jurisdictional limitation was proportionate. Ribeiro PJ contrasted the present case with A Solicitor: ‘the [District Court Ordinance] provisions limiting the right of appeal plainly do not erect a total ban on appeals. The Court of Appeal is entrusted with vetting the prospects of a potential appeal and enjoined to refuse leave unless the criteria specified in section 63A(2) are met.’Footnote ¹²⁰

One final case which merits analysis is Mok Charles Peter v Tam Wai Ho.Footnote ¹²¹ The impugned finality clause, section 67(3) of the Legislative Council Ordinance, purported to make decisions of the Court of First Instance ‘final as to the matters at issue concerning the election petition’. The CFA concluded that the restriction was disproportionate for it went beyond what was necessary to achieve the aim of determining election petitions speedily.Footnote ¹²² The CFA alluded to the absolute nature of the restriction and the bench's unease with the fact that another part of the same Ordinance curiously provided for an appellate procedure in respect of challenges against the qualifications of a Legislative Council member in a manner that is not parallel. Furthermore, the bench did not shy away from acknowledging that the electoral context, an important microcontextual consideration, played a key role in its analysis: ‘it is crucial at the outset to bear in mind that an election petition engages the public interest and not only the interests of the election protagonists themselves … an election petition involves substantive rights, and not merely procedural rights. These substantial rights are political in nature.’Footnote ¹²³

Drawing these threads together provokes the question of whether the same approach can be migrated to the ouster clause context. The argumentative matrix in the ouster clause context can be sketched as follows. Unlike finality clauses, which only prevent a matter that has already been processed by a lower court or tribunal from landing on the docket of the apex court, ouster clauses undertake to eliminate the supervisory jurisdiction of the first instance High Court and thus rob the power of any court to hear the case altogether. This exposes ouster clauses to the charge that they stand in contradiction to an array of provisions in the Basic Law concerning the allocation of judicial power. Under the Basic Law, the power to adjudicate cases, encompassing both original and appellate jurisdictions, is expressly vested in the judicial branch as a whole. Article 2 announces that the Hong Kong SAR is authorised to ‘enjoy executive, legislative and independent judicial power’. Article 19 repeats that the SAR ‘shall be vested with independent judicial power, including that of final adjudication’, and that its courts shall have ‘jurisdiction over all cases in the Region’. Article 80 stipulates that courts at all levels constitute the judiciary which collectively exercise ‘the judicial power of the Region’, and Article 85 reiterates that the courts ‘shall exercise judicial power independently, free from any interference’.

The judicial review jurisdiction of the High Court forms part and parcel of judicial power,Footnote ¹²⁴ clearly encapsulated by these provisions. An ouster clause, as the argument goes, undercuts the judicial power of the High Court. As such, following the case law of the CFA on finality clauses, the ouster clause ought to be subject to constitutional muster and processed under the proportionality framework. This includes scrutinising whether the encroachment on judicial power serves a legitimate aim, and whether the ouster clause can be defended as a rational, necessary, and proportionate means to achieve that aim. Similar to the analysis under Article 35 right of access to courts, a host of different microcontextual considerations, such as the subject matter of the shielded decision, will be captured, mainly at the necessity and balancing stages. As the case of Charles Peter Mok demonstrates, an ouster clause embedded in a statutory context involving high political affairs may tip the scales in favour of finding the ouster clause disproportionate. In terms of remedies, the court may similarly inquire into the possibility of remedial interpretation and, if impossible, proceed to invalidate the clause.

The neutrality of the proportionality framework meant the foregoing approach does not, again, ‘dictate the correct legal answers to legal questions’.Footnote ¹²⁵ The fate of an ouster clause may come down either way. Likewise, this approach rejects categorical reasoning, such that a violation of judicial power will not automatically render the ouster clause unconstitutional. It thus preserves flexibility and space for strategic judicial manoeuvring. Interestingly, Hong Kong cannot lay claim to exclusivity in the perceived tension between ouster clauses and allocation of judicial power. Courts in Singapore have also wrestled with this incredibly thorny and delicate issue. This article now pivots towards an examination of how Singaporean courts have navigated this issue, which, in turn, sheds light on how their methods may bear on the approaches that Hong Kong courts may pursue.

The idea that ouster clauses may be an unconstitutional curtailment of judicial power was first mooted by Chan Sek Keong, the former Chief Justice of Singapore, in 2012. In a lecture delivered to law students, Chan Sek Keong CJ posed an ‘academic argument’, suggesting to the audience that ouster clauses ‘may be inconsistent’ with Article 93 of the Constitution of Singapore, which vests judicial power in the courts.Footnote ¹²⁶ This line of argument, he believed, would ‘bypass’ the doctrinal difficulties presented by Anisminic.Footnote ¹²⁷ The Chief Justice, however, stopped short of offering a conclusive view on the matter, preferring instead to call it a ‘good examination question’.Footnote ¹²⁸

Chan Sek Keong CJ's ‘academic argument’ was later taken up by the High Court and Court of Appeal of Singapore, in the case of Nagaenthran a/l K Dharmalingam v Public Prosecutor. There, Nagaenthran, a Malaysian citizen who had been convicted and later executed for trafficking in drugs, challenged section 33B(4) of the Misuse of Drugs Act, which provides that ‘no action or proceeding shall lie against the Public Prosecutor in relation to any [determination as to whether an offender has substantively assisted the Central Narcotics Bureau].’ In the High Court of Singapore, Chan Seng Onn J upheld the validity and constitutionality of the clause.Footnote ¹²⁹ He devoted a large part of his judgment to the issue of the distinction between jurisdictional and non-jurisdictional errors of law, a dichotomy that Anisminic had rendered all but obsolete in the English context, and concluded in the final analysis that ouster clauses can only be used to defeat jurisdictional errors of law.Footnote ¹³⁰

In support of his conclusion that section 33B(4) was constitutional, Chan Seng Onn J put forward a number of reasons. To begin with, he ruled that prosecutorial discretion was a non-justiciable matter.Footnote ¹³¹ In his view, ‘courts are ill-equipped to consider whether an offender has rendered substantive assistance in disrupting drug trafficking activities, given that such a determination involves a holistic inquiry premised on a panoply of extra-legal factors, including in particular the operational considerations of the [Central Narcotics Bureau] in the disruption of drug trafficking activities’.Footnote ¹³² The Public Prosecutor, on the other hand, ‘possess[es] the unique qualities that render that office most suited to conduct the assessment under s 33B(2)(b)’.Footnote ¹³³ Rather than running afoul of Article 93 of the Singapore Constitution, the provision is ‘in fact an exemplar of the separation of powers principle in action’.Footnote ¹³⁴

Furthermore, Chan Seng Onn J observed that the impugned clause epitomised a ‘reasonable balance’Footnote ¹³⁵ struck by the Parliament, citing the case of Tey Tsun Hang v Attorney-General.Footnote ¹³⁶ In Tey Tsun Hang, the High Court was asked to review section 39A of the Immigration Act, the material parts of which read: ‘[t]here shall be no judicial review … except in regard to any question relating to compliance with any procedural requirement of this Act or the regulations governing that act or decision.’ The immunised decision in that case pertained to the cancellation of an application for renewal of re-entry permits. Adopting a deferential posture, Loh J was satisfied that section 39A of the Immigration Act was constitutionally valid. Firstly, ‘[i]n matters relating to national policy, good examples of which are matters relating to land planning … or immigration or defence, there are good and self-evident reasons why these matters are best left to the executive arm and not the courts.’Footnote ¹³⁷ Secondly, section 39A was not a complete or total ouster. It ‘does not oust matters involving compliance with any procedural requirements of the Act or the relevant regulations from the court's jurisdiction’.Footnote ¹³⁸ Taken together, Loh J was convinced that the Parliament has achieved a ‘reasonable balance’ in writing section 39A.

Chan Seng Onn J's conclusion in favour of the constitutionality of section 33B(4) of the Misuse of Drugs Act is very much informed by the fact that the provision did not bar judicial review in its entirety. Specifically, review on the more restricted grounds of bad faith and malice remains very much alive. As such, akin to section 39A of the Immigration Act, section 33B(4) was merely a partial ouster clause. Thio Li-ann characterised the High Court's approach as a ‘contextual’ approach, whereby the analysis would be pinned down to ‘factors like the field of regulation and the scope of the ouster clause … including constitutional norms.’Footnote ¹³⁹ That being said, it remains, to say the least, opaque from the judgment the mechanics of the balancing exercise, let alone its potential bite.

On appeal, the Court of Appeal charted a wholly different path. Instead of treating section 33B(4) as an ouster clause and analysing it as such, the Court classed section 33B(4) as an ‘immunity clause’. Thus, it was not strictly obligatory for the bench to proceed and assess the constitutionality of section 33B(4) as if it were an ouster clause. Nevertheless, Menon CJ, successor to Chan Sek Keong CJ, opted to seize the space and articulate staunch obiter remarks echoing his predecessor's ‘academic argument’. As the following paragraphs will demonstrate, Menon CJ's quest could prove consequential for the jurisprudential trajectory of ouster clauses, at least in Singapore.

The primary import of an ouster clause, as underscored by Menon CJ, is ‘to oust the court's jurisdiction to exercise the power of judicial review.’Footnote ¹⁴⁰ This ‘power of judicial review’ is rooted in the principle of separation of powers as well as Article 93 of the Constitution of Singapore, which provides that ‘[t]he judicial power of Singapore shall be vested in a Supreme Court and in such subordinate courts as may be provided by any written law for the time being in force.’Footnote ¹⁴¹ Ouster clauses, therefore, besiege both Article 93 and the principle of separation of powers. This argument is further buttressed by the imperative to defend the rule of law. As Menon CJ said in no uncertain terms, ‘any society that prides itself in being governed by the rule of law, as [Singapore] does, must hold steadfastly to the principle that “[a]ll power has legal limits and the rule of law demands that the courts should be able to examine the exercise of discretionary power”.’Footnote ¹⁴² As noted by one commentator, these contentions parallel Lord Carnwath's insistence on placing limits on the Parliament's power to oust the High Court's jurisdiction in the name of the rule of law.Footnote ¹⁴³

Menon CJ's statements, despite being strictly obiter, are hailed as a marked departure from the Anisminic principles as well as the approach taken by the High Court centred on balancing and contextual reasoning, insinuating ‘some dissatisfaction with the prevailing Anisminic framework’ and signalling ‘a judicial intent to forge a different path.’Footnote ¹⁴⁴ Nevertheless, ‘the precise import of this new conceptual basis for the judicial approach to ouster clauses’Footnote ¹⁴⁵ and its clout remain uncertain. That engenders a situation where, but also stems from the fact that, the remarks can be interpreted in at least two plausible ways.

Firstly, the remarks can be interpreted as a categorical declaration that all statutory ouster clauses in Singapore are unconstitutional. Because all ouster clauses invariably undercut judicial power in one way or another, Menon CJ's remarks leave minimal, if any, constitutional breathing space for them regardless of the statutory context in which they are embedded. The Anisminic approach would also be consigned to the dustbin as a corollary. The basis for this view is that judicial power under Article 93 of the Constitution of Singapore is vested ‘exclusively’Footnote ¹⁴⁶ in the courts. The constitutional provision tolerates no notion of balancing: ‘There is no mention of any balancing, any truncation of judicial power, or any transfer of judicial power to other bodies; nor can there be, for the Constitution confers judicial power on the courts in absolute terms.’Footnote ¹⁴⁷ As a result, judicial power cannot be weighed against competing interests, regardless of the degree of curtailment or the force of the legitimate aim, within the proportionality framework. To rephrase Chan Seng Onn J, there can never be a ‘reasonable balance’ between an ouster clause and other considerations.

A similar analysis animates the decision of Desalination Company of Trinidad and Tobago v Registration Recognition and Certification Board,Footnote ¹⁴⁸ decided by Gobin J of the High Court of Trinidad and Tobago. The impugned ouster clauses, found in sections 23(6) and 23(7) of the Industrial Relations Act, purported to shield decisions of the respondent from review. The claimant contended that the ouster clauses violated the principle of separation of powers in the Constitution of Trinidad and Tobago, for they ‘render the [Registration Recognition and Certification Board], a non-judicial authority, the final arbiter as to the interpretation of the [Industrial Relations Act], even as to those provisions which define the limits of its own jurisdiction’.Footnote ¹⁴⁹ Persuaded by the argument, Gobin J ruled to nullify these ‘constitutionally aberrant’Footnote ¹⁵⁰ ouster clauses.

Consider, now, the second plausible interpretation.Footnote ¹⁵¹ Menon CJ's remarks may been construed as, rather than a categorical rejection of the constitutionality of ouster clauses, endorsing a contextual approach whereby the effect of an ouster clause is mediated by a panoply of ‘public law values’.Footnote ¹⁵² These public law values include the separation of powers, constitutional supremacy, judicial power under Article 93, the rule of law, the relative institutional competence of courts vis-à-vis other branches of government, or even an implied right to a judicial remedy.Footnote ¹⁵³ This second interpretation chimes with the approach taken by Chan Seng Onn J, which Menon CJ did not expressly overrule or disapprove. These public law values are components of a more nuanced analytical process premised on substantive rather than formalistic reasoning, softening the otherwise draconian ouster clause. Of particular note is that this interpretation bears a strong resemblance to the modified Anisminic approach, albeit with Anisminic itself not necessarily assuming similar focality and primacy.

Support for this interpretation can be gathered from the earlier case of Review Publishing. There, Menon JC, sitting then as a Judicial Commissioner of the High Court, cautioned, in the context of determining the range of matters that should be regarded as non-justiciable under Article 93, against adhering to ‘a highly rigid and categorical approach to deciding which cases are not justiciable’Footnote ¹⁵⁴ that is antithetical to a ‘sensible’ application of the principle of separation of powers. This remark should similarly counsel against a rigid and categorical approach towards ouster clauses. Likewise, the very existence of Article 149(3) of the Singapore Constitution, a constitutional ouster clause that shields decisions made under, inter alia, the Internal Security Act from review, notwithstanding Article 93, could be used to dispute the ‘exclusivity’ of judicial power.Footnote ¹⁵⁵

This interpretation of Menon CJ's remarks, therefore, again introduces a dose of contextuality and repels categorical reasoning premised on the supposed supremacy and inviolability of certain constitutional provisions. It offers the courts substantial leeway in fashioning an approach to ouster clauses that suits specific contexts. These may include invalidation, a restrictive or remedial interpretation, or giving the clause unqualified effect. Thio summarised elegantly:

To sum up, the obiter remarks penned by Menon CJ, and the multiple interpretations they are capable of sustaining, have placed ouster clause jurisprudence in Singapore back in a state of flux. Notably, the recent enactment of section 104 of the Foreign Interference (Countermeasures) Act 2021, which precludes decisions made under the Act from review, has intensified discussions in this direction.Footnote ¹⁵⁷

The jurisprudential trajectory of ouster clauses in Singapore furnishes robust comparative inspiration for and prompts consideration as to whether such approaches are transposable to Hong Kong. Like the Singapore Constitution, the Hong Kong Basic Law announces in numerous provisions that judicial power is vested in the courts. A straightforward application of the categorical interpretation of Menon CJ's obiter remarks yields the linear logic that all ouster clauses, irrespective of their statutory context, are unconstitutional as they violate these analogical provisions that admit no notion of balancing.

As for the contextual interpretation of Menon CJ's remarks, its linchpin of contextual balancing of various ‘public law values’, many of which are well recognised in Hong Kong, attunes well with both the modified Anisminic approach and the proportionality framework. In particular, the balancing stage ensures that these public law values are firmly captured and taken into account. On the other hand, the main difference between these two approaches lies in the fact that, as mentioned above, whilst the interpretive modified Anisminic approach still gravitates towards Anisminic as the focal point of analysis, the constitutional approach does not, treating an Anisminic-esque interpretation as no more than one potential outcome via remedial interpretation.

It is important to highlight, in the final analysis, that there are other peculiarities within Hong Kong's constitutional set-up that complicate the transposability of Menon CJ's remarks even further. Two of them are canvassed here.

Firstly, Hong Kong courts showcased unease with the proposition that they hold ‘exclusive’ judicial power, and the CFA has yet to render an unambiguous statement to that effect. There are, however, scattered dicta expressed by the lower courts. In Lee Yee Shing Jacky v Board of Review (Inland Revenue Ordinance), Lam J (now a Permanent Judge of the CFA), sitting in the Court of First Instance, observed that ‘[u]ltimately, one has to examine the nature, function and character of the decision in question in order to determine whether it involves the exercise of the judicial power of the state which by law reserves exclusively for the judiciary.’Footnote ¹⁵⁸ Such judicial ambivalence may be attributed to the difficulties involved in framing ‘an exhaustive definition of judicial power’, for there exists ‘a borderland in which judicial and administrative functions overlap’.Footnote ¹⁵⁹ Such ‘borderland’ is consistently traversed by administrative or quasi-judicial tribunals performing an ostensibly judicial function, such as the Market Misconduct Tribunal tasked with determining whether one is guilty of market misconduct.Footnote ¹⁶⁰ In the absence of high judicial pronouncements, it is plausible that the categorical interpretation of Menon CJ's remarks may face resistance.

Second, granted that all formal textual features are in order, the issue of whether the principle of separation of powers is encapsulated within the Basic Law remains deeply contested.Footnote ¹⁶¹ Both the Basic Law and Constitution of Singapore create three branches of government and vest their judiciaries with independent judicial power accordingly. In Singapore, the principle of separation of powers was accepted as a core feature or ‘basic structure’Footnote ¹⁶² of its Constitution. In Hong Kong, on the other hand, while the CFA has on multiple occasions concluded that all formal features in the Basic Law add up to the same principle,Footnote ¹⁶³ such conclusion has not been particularly well-received by many others, including the political branches. The transposability of Menon CJ's remarks may hence also depend on the trajectory of the principle of separation of powers in Hong Kong.

A final offshoot of the growing global jurisprudence on ouster clauses was developed by the Federal Court of Malaysia, the highest court of the jurisdiction. In the 2022 case of Dhinesh Tanaphll v Lembaga Pencegahan Jenayah and Others,Footnote ¹⁶⁴ the Federal Court was urged, after many unsuccessful attempts, to nullify the ouster clause in section 15B of the Prevention of Crimes Act 1959. The relevant parts read:

After a meticulous review of the authorities, the Federal Court finally acceded to the invitation. The Federal Court held, with clear parallel to the remarks of Menon CJ, that the judicial power of constitutional review and the principle of separation of powers, housed in Articles 4(1) and 121(1) of the Federal Constitution, were eroded by section 15B of the Prevention of Crimes Act. More crucially, the bench took one step further, declaring that the provision has flouted the foundational principle of constitutional supremacy engraved in Article 4(1), which states: ‘This Constitution is the supreme law of the Federation and any law passed after Merdeka Day which is inconsistent with this Constitution shall, to the extent of the inconsistency, be void.’ The Federal Court wrote in emphatic terms:

In essence, the Federal Court found repugnant the fact that the ouster clause contained in section 15B erected an impregnable barrier to the exercise of constitutional review power, thereby unravelling the very principle of constitutional supremacy. As a result, the bench was compelled to declare the provision unconstitutional.

Consider the applicability of the Malaysian Federal Court's reasoning in Hong Kong. Some may question, as a preliminary matter, whether the Basic Law is supreme in the sense articulated by the Federal Court. This uncertainty stems from the fact that, despite all the formal and substantive appearance of a supreme law of the land, the Basic Law remains a piece of national law enacted by the Chinese national legislature, the National People's Congress, and is subject to the National People's Congress Standing Committee's ‘general and free-standing’ power of final interpretation.Footnote ¹⁶⁶ That is to say, the Standing Committee may, at any time and on its own initiative, issue a binding interpretation of any provision that has the effect of overriding any contrary decisions rendered by local courts. Despite such constitutional peculiarity, Hong Kong courts have for all practical purposes treated the Basic Law as a supreme written constitution, invalidating scores of constitutionally repugnant legislation under its command. As the Court of Appeal stated in no uncertain terms, ‘the supremacy of the Basic Law means that no one – the legislature included – is above the Basic Law … where a constitutional requirement under the Basic Law is in issue, even the legislature cannot act contrary to that requirement under the Basic Law.’Footnote ¹⁶⁷

Article 11(2), the single clearest manifestation of the principle of supremacy of the Basic Law, is a virtual analogue of Article 4(1) of the Malaysian Federal Constitution. The latter provides that: ‘This Constitution is the supreme law of the Federation and any law passed after Merdeka Day which is inconsistent with this Constitution shall, to the extent of the inconsistency, be void.’ Substituting the corresponding terms yields a provision remarkably similar to Article 11(2) of the Basic Law, which stipulates: ‘No law enacted by the legislature of the Hong Kong Special Administrative Region shall contravene this Law.’ A case could thus be made that the doctrine of Basic Law supremacy can be invoked to defeat ouster clauses. However, one important microcontextual caveat should be added. This approach to ouster clauses is more likely to hold sway where the clause is invoked to preclude constitutional judicial review, rather than judicial review of administrative actions with little or no constitutional dimensions.

This section has reconceptualised a set of approaches to ouster clauses available to Hong Kong courts, drawing on the intricate interaction between such notions as separation of powers, allocation of independent judicial power, and constitutional supremacy. It is important, however, to emphasise that the transposability of some of these approaches, particularly those drawing on comparative jurisprudence, should not be lightly assumed. Attention must be paid to the constitutional peculiarities of Hong Kong that belie simple generalisations.

Relevance of Article 19(2) of the Basic Law

One final issue that falls to be addressed is the implication, if any, of Article 19(2) of the Basic Law on the range of possible approaches to ouster clauses available to the courts. Article 19(2) reads:

The effect of Article 19(2) is to preserve the pre-existing jurisdictional constraints on the courts that were in place prior to the handover. At the highest level, the outer jurisdictional parameters, or ‘general jurisdiction’,Footnote ¹⁶⁸ of the courts are delineated by the Basic Law.Footnote ¹⁶⁹ As Cheung CJ, the current Chief Justice of Hong Kong, said: ‘In jurisdictions where there is a written constitution, the jurisdiction and positioning of the courts are entirely dependent on what is provided for in the Constitution, which governs what the courts can and cannot do.’Footnote ¹⁷⁰ General jurisdictional restrictions applicable to the judicial branch as a whole, prior and present, therefore, must either be found in the Basic Law itself, such as Articles 2, 19, 80 and 85, or be traceable back to it. An example is Article 19(3), which bars the courts from assuming jurisdiction over acts of state such as defence and foreign affairs. Some commentators have interpreted restrictions ‘imposed by the legal system and principles previously in force’ as meaning ‘common law limitations on the jurisdiction of courts, such as non-justiciability and the “one voice” doctrine’.Footnote ¹⁷¹ An example of the latter can be found in the case of Chong Bing Keung v Government of the United States,Footnote ¹⁷² where the Court of Appeal refused to inquire into the validity of a treaty and held that ‘the municipal courts of Hong Kong are not competent to adjudicate upon treaty obligations on the plane of international law’.Footnote ¹⁷³

Relying on Article 19(2), some may argue that the provision operates to shield ouster clauses enacted prior to the handover from judicial invalidation, a remedy available under the constitutional approaches, because they constitute legitimate and constitutionally recognised pre-existing jurisdictional restrictions. Arguments to that effect, nevertheless, falter upon closer scrutiny.

Firstly, these pre-handover ouster clauses are all, by definition, statutory creatures. They are, therefore, not ‘common law limitations’. In any event, the terms ‘system’ and ‘principles’ in Article 19(2) all point to broader and more systemic jurisdictional limitations, as opposed to individual limitations found in particular statutory contexts. Secondly, and even more fatally, even assuming that the ouster clauses do constitute pre-existing jurisdictional restrictions, it cannot be taken for granted that they will survive indefinitely. The CFA made this point extremely clear in Ng Ka Ling, the first constitutional case to reach Hong Kong's apex court:

To rephrase, pre-existing jurisdictional limitations survive and become part of Hong Kong law only if they are compatible with the constitutional order umpired by the Basic Law. The suggestion that Article 19(2) immunises all pre-existing ouster clauses from review creates a glaring inconsistency, whereby only ouster clauses enacted post-handover could be subject to constitutional scrutiny. In any event, Article 19(2), as part of the broader constitutional scheme, ought to be interpreted coherently with other provisions so as ‘to produce a harmonious and congruous instead of a discordant meaning’.Footnote ¹⁷⁵ If an ouster clause is found to be constitutionally offensive vis-à-vis other articles in the Basic Law, the decision to nevertheless uphold it would effectively elevate Article 19(2) over other articles. This is destructive of the coherence of the Basic Law and is wholly unwarranted.

Taken together, the contention that Article 19(2) immunises pre-existing ouster clauses from constitutional invalidation is unsustainable. As such, it in no way affects the viability of the constitutional approaches, as well as the interpretive approaches, proposed above.