Constitutional equality and executive action – a comparative perspective to the comparator problem

Abstract

A general right to equality is a common feature of written constitutions around the world. Interesting questions arise when one seeks to apply such rights to discrete executive acts. The subject of such acts has necessarily been singled out from a multitude of possibilities for the purposes of the act. To determine whether a differentiation has occurred such that like cases have not been treated alike, to what or whom should this subject be compared? The question of how one selects the proper comparator becomes especially significant when one notes that whether the equal protection guarantee is triggered at all depends on the answer to this question. This paper will study how courts in Hong Kong and Singapore have addressed these difficulties. It argues that three categories of approaches can be discerned in these jurisdictions: class-focused, policy-focused, and justification-focused approaches. It critically evaluates each approach, argues in favour of a justification-focused approach to constitutional equal protection in the context of discrete executive acts, and explores the implications of such an approach for the proper relationship between constitutional equality and administrative law.

Introduction

The right to equal protection is a commonly-found constitutional guarantee in written constitutions around the world. Such rights come in two main forms: general equality rights enshrining a guarantee of equality before the law; and specified rights to equality articulating certain categories of proscribed differentiations.¹

The focus of this paper will be the general equality right. As a constitutional right, jurisdictions adhering to the doctrine of constitutional supremacy have generally recognised that it can be invoked as a means of challenging both legislative provisions and executive action. Interesting questions arise, however, when one turns to apply the constitutional guarantee of equal protection to discrete executive acts; for example, a land authority's compulsory acquisition of a plot of land. Indeed, the subject of a discrete executive act has necessarily been singled out from a multitude of possibilities. To determine whether a differentiation has occurred such that like cases have not been treated alike, to what or whom should this subject be compared? A wide range of comparators is possible, given the specific nature of discrete executive acts. How then should one select the proper comparator?

This issue can be described as the comparator problem. The question of how one selects the proper comparator becomes especially significant when one notes that whether the equal protection guarantee is triggered at all depends on how one answers this question. Indeed, if the chosen comparator results in the compared entity being considered sufficiently different from the subject of the executive act, then there is no question of equal protection being violated at all. Any differentiation would not contravene the principle that like cases should be treated alike, since both subjects were simply not alike to begin with.

This paper will seek to offer a principled response to the comparator problem. It will first set the background by discussing the comparator problem that arises when one seeks to apply general equality rights to discrete executive acts. It will then study how courts in Hong Kong and Singapore have sought to resolve the comparator problem. Both jurisdictions enshrine equal protection as a constitutional right and possess a shared common law heritage.

The paper will suggest that three types of approach to the comparator problem can be discerned: class-focused, policy-focused, and justification-focused approaches. A class-focused approach seeks to determine first whether there has been differential treatment for the purposes of equal protection analysis. This step would involve a consideration of whether the subjects sought to be compared were equally situated. This will be predominantly shaped by whether the compared subjects fall within the same class or belong to different classes, with the relevant classes being determined by the court. If the court finds that there was indeed differential treatment, the constitutionality of such treatment will fall to be determined by rationality or proportionality analysis.

A policy-focused approach adopts a similar two-step framework but, in contrast, is focused less on classes at the first stage of analysis and instead more on the substantive moral requirements of equality that are called for by the nature of each decision – compulsory land acquisitions, the grant of dependant visas, and so on. Whether there is differential treatment that needs to be justified is determined by whether the relevant treatment amounts to a deviation from the moral requirements of equality in the particular administrative context.

A justification-focused approach concentrates on the justifiability of any alleged factual difference in treatment. In other words, a justification-focused approach emphasises the test of justification as the principal heuristic by which one determines whether there has been unconstitutional differential treatment. This paper will argue that a justification-focused approach represents the best approach to constitutional equal protection in the context of discrete executive acts, and will explore the implications for the acceptance of such an approach for the proper relationship between constitutional equality and administrative law.

1. Equality and executive action – the comparator problem

The meaning of equality has been the subject of erudite philosophical discussion – indeed, the notion of equality is a common feature in works of philosophy.² The concept of equality has also found its way into the legal realm. It is a widely-accepted component of the rule of law, often expressed in the maxim ‘like should be treated alike’.³ Further, equality has found its way into law in a more concrete sense, given that a constitutional guarantee of equal protection is a common feature across written constitutions around the world.⁴

Concrete incorporation of the philosophical concept of equality into law necessitates its translation into legal doctrine amenable to practical application. There are two major categories of methods by which the requirements of equality have been effectuated in legal doctrine, each of which can be broadly described as equality as rationality and equality as impermissible differentiation. These categories broadly track Niels Petersen's classification of legal approaches to equality as specifying a ‘consistency requirement’ or ‘non-discrimination guarantee’ respectively, as well as Owen Fiss's categorisation of equality approaches into ‘anti-classification’ and ‘anti-subordination’ approaches respectively.⁵

As a means of effectuating the concept of equality in concrete legal doctrine, equality as rationality would require judges to consider whether a differentiation is rationally justifiable by reference to whether it is relevant to a goal sought to be accomplished.⁶ As long as a differentiation is rationally justifiable, the legal requirement of equality would be fulfilled. Scholars have indeed argued that the concept of justification is central to legal equality. For example, Jeffrey Jowell argued that equality ‘requires government not to treat people unequally without justification’, as a matter of basic democratic principle.⁷ In a similar vein, Trevor Allan proposed a connection between equality, which requires distinctions to be ‘capable of a reasoned justification’, and the ideal of the rule of law.⁸ Guy Lurie further underscored the closeness of the connection between rational justification and the concept of equality in his argument that the Aristotelian formula that like cases ought to be treated alike unless there is a justification for differentiation is ultimately a ‘relational means-ends test or measure, much like proportionality’.⁹

In addition to equality as rationality, a distinct means by which equality has been effectuated in legal doctrine can be described as equality as impermissible differentiation. Instead of focusing on whether differentiations are rationally justifiable, equality as impermissible differentiation would go further to highlight certain types of differentiation as problematic per se and therefore subject to justification or simply impermissible.¹⁰ Such a vision of equality would be directed at protecting vulnerable social groups – indeed, it has been described as an anti-subordination vision of equality seeking to ameliorate the systemic disadvantages of certain social groups which lack political power.¹¹ The types of suspect classifications that would attract attention can either be articulated by judges¹² or identified in the constitutional text.¹³ The question of how suspect classifications ought to be articulated has attracted intense academic discussion.¹⁴ A noteworthy contribution in this regard is Tarunabh Khaitan's proposal that suspect classifications can be identified on the basis of two cumulative requirements: first, the ground must classify ‘persons into groups with a significant advantage gap between them’; secondly, the ground must either be ‘immutable’ or constitute ‘a fundamental choice’.¹⁵

In view of the focus of this paper on the general right to equality, our discussion will be centred around the idea of equality as rationality.¹⁶ Indeed, where the relevant constitutional equality provisions are framed as general equality rights – Article 12(1) of Singapore's Constitution and Article 25 of Hong Kong's Basic Law are pertinent examples in this regard – courts have often effectuated such rights through the idea of equality as rationality.

It is interesting to observe, however, that substantiating the requirements of equality as rationality in the specific context of executive action raises some difficult issues. Executive acts necessarily make distinctions.¹⁷ But in the context of such acts, identifying the relevant differentiation that ought to be scrutinised is much less straightforward. Indeed, what is the appropriate comparator to adopt in order to identify whether there has been a differentiation which needs to be justified? While equality requires like cases to be treated alike, this raises the preliminary question of how to select the proper comparator to determine whether two entities are alike in the first place.¹⁸

Consider, for example, the compulsory acquisition of a plot of land by a public authority for the purposes of land development. Suppose that this acquisition is then challenged on equal protection grounds, on the basis that other adjacent plots of land were not selected for compulsory acquisition. In such a situation, what should be the appropriate comparator to determine whether there had been a differentiation that needs to be justified? One possibility might be to adopt general geographical location as the basis for comparison – in other words, why has this plot of land been singled out for compulsory acquisition when there are other plots of land just adjacent to this one? Since all these plots of land are within the same general geographical location, the selection of this plot of land amounts to a differentiation among equivalents that therefore has to be justified.

The obvious problem that arises with such an argument, however, is that there is a whole myriad of other possible comparators that can be adopted. If one adopts specific geographic location as the basis of comparison, then the selected plot of land is clearly not equivalent to the adjacent plots of land, and any differentiation between the selected plot and the adjacent plots would not raise any equal protection concerns at all. Other possible bases of comparison include the proximity of plots of land to main road connections, or the valuation of the plots of land.

The point to be made is that there is a range of possible comparators that one can adopt when analysing discrete executive action for violations of constitutional equality. How then does one select the comparator to adopt?¹⁹ The trouble is that the idea of equality as rationality in itself does not answer this question. To the extent then that equality as rationality has been adopted on its own as a means by which the concept of equality has been instantiated in legal doctrine, a zone of ambiguity arises when applying equal protection doctrine to executive action.

This ambiguity is not merely a theoretical matter. Important consequences follow from it. Indeed, as the preceding paragraph has illustrated, the comparator that one selects as the basis of equal protection analysis has a significant impact on whether the right to equal protection is at all implicated – if the relevant parties are not alike, then no concern of equal protection will be triggered.

The Singapore High Court decision of Syed Suhail bin Syed Zin v Attorney-General ²⁰ (Syed Suhail (HC)) illustrates the practical importance of this comparator problem. The court was concerned with the merits of a judicial review application against the scheduling of a prisoner's execution. The applicant argued, inter alia, that the scheduling of his execution ahead of other prisoners who had been sentenced to death earlier amounted to an equal protection violation. By failing to schedule executions in the order that prisoners were sentenced to death, prisoners would not be entitled to a fair amount of time during which they could adduce new evidence to challenge their convictions.

In addressing this argument, the High Court considered first of all the question of whether the compared persons or entities were equally situated. The High Court held that the applicant failed at this first stage – he was not equally situated with the other prisoners he sought to compare himself to, since he had no realistic expectation of his case being reviewed and potentially reopened on the merits, while the other prisoners did.²¹ This was sufficient to dispose of the applicant's equal protection challenge. This illustrates the importance of the comparator problem. The selection of the proper comparator is quite capable of resolving equal protection challenges by determining to begin with whether an equal protection concern has arisen at all.

This comparator problem applies to a lesser degree where legislation is concerned. Indeed, when analysing legislation for a violation of constitutional equality, the relevant comparators can be more obvious. Many legislative provisions draw clear distinctions between persons and acts captured by a provision, and those which are not. This is not to say, however, that legislation is immune to such problems. Where the relevant legislation is pitched at a general level, identifying the relevant comparator to analyse can still raise challenges. But a further unique issue raised by equal protection challenges to executive action is that they raise the possibility of overlap with administrative law principles which would be broadly applicable to executive action. The import of such overlap will be revisited subsequently, and it suffices for present purposes to note that the unique issues raised by equal protection challenges to executive action merit specific attention and will be the focus of this paper.

Returning to the specific context of executive action, one might suggest at this juncture that the range of possible comparators can be narrowed if one enjoins judges to have regard to the comparator actually intended by the executive authority. For instance, if the public authority in the above hypothetical scenario intended valuation to be the comparator, the courts ought to adopt this as the relevant comparator. Insofar as the underlying instinct beneath this suggestion is that the identification of the relevant comparator to be analysed ought to be analytically linked to the purpose of the differentiating executive action, this suggestion foreshadows the argument that will be made subsequently in this paper.²² The crucial point, however, is that this unfortunately has not been the only way that courts have addressed the comparator problem, as the subsequent sections of this paper will illustrate.

2. Legal approaches to constitutional equality for executive acts

We turn then to identify and evaluate the solutions which courts have articulated in response to the comparator problem.

A few preliminary points should first be made about the methodology. First of all, the goal of this section is to identify any possible means by which judges have sought to address the comparator problem, even if they may not be the prevailing legal approaches in the jurisdictions selected for this study. Also, in selecting the jurisdictions to focus on, attention was paid to selecting jurisdictions which were reasonably comparable. Jurisdictions possessing generalised constitutional guarantees of equal protection as well as a common law administrative law heritage were shortlisted for this study. The former requirement would ensure that selected jurisdictions have a general equality right entrenched in legal doctrine at similar levels in the hierarchy of legal norms – thereby excluding jurisdictions such as the UK and Australian federal law from this study. The latter requirement was intended to ensure that the selected jurisdictions would possess broad commonalities in judicial culture, reasoning, and administrative law principles.

With the requirements that the selected jurisdictions had to meet articulated as such, Hong Kong and Singapore were chosen for this study as jurisdictions capable of serving as good case studies for the issue under interrogation here. Both jurisdictions possess a body of case law centring around a constitutional general equality right as applied to executive action, and both jurisdictions share a common law administrative law heritage. It is worth noting, however, that the significance of this paper is not strictly limited to the type of jurisdictions chosen for this study – indeed, the conceptual possibilities illustrated by the chosen jurisdictions will be of relevance to any jurisdiction applying constitutional or quasi-constitutional guarantees of equality.

The focus of the following discussion will be on identifying and evaluating how courts have grappled with the issue of how to determine whether there was a differentiation giving rise to an equal protection concern to begin with. It will be observed that this issue has attracted a variety of solutions, and that the approaches taken by the Hong Kong and Singapore courts to this issue fall into three broad categories: class-focused, policy-focused and justification-focused approaches. Each approach will be discussed in turn.

(a) Class-focused approach

A class-focused approach seeks to identify whether there has been differential treatment for the purposes of equal protection analysis by considering whether the compared subjects fall within the same class or belong to different classes. Should the compared subjects indeed fall within the same class, differential treatment of the compared subjects will implicate the constitutional equal protection guarantee and will have to be justified as lawful on the basis of rationality or proportionality analysis. An analogous type of approach is adopted where categories of impermissible differentiation are expressly articulated in the relevant constitutional text. Where such clauses are concerned, courts would have to determine whether the relevant differentiation falls within these express categories of impermissible differentiation. However, where general equality rights are concerned – the focus of this paper – courts would have to construct the relevant classes themselves.

A distinctive characteristic of this approach is its focus on classes in determining whether there are any equal protection concerns with the executive action in question. It is worth noting, however, that a class-focused approach is not hermetically isolated from moral assessment (as a policy-focused approach emphasises) or a requirement of rational justification (as a justification-focused approach emphasises). What makes a class-focused approach distinct from these approaches is the significant weight accorded to the formulation of classes in resolving equal protection challenges. Indeed, as the subsequent cases will illustrate, on such an approach the formulation of classes in itself is quite capable of being dispositive of equal protection challenges, even if how exactly the classes have been formulated receives little judicial discussion. Moral assessments or rational justification remain implicit or are secondary to the formulation of classes – the significance of which will be revisited in our subsequent evaluation of this approach.

(i) Hong Kong

The approach which various decisions of the Hong Kong courts have adopted to analyse equal protection in the context of executive action falls within this category of approaches. Guarantees of equal protection are enshrined in Hong Kong's Basic Law under Article 25 as well as the Hong Kong Bill of Rights Ordinance under Article 1²³ – documents which enjoy constitutional status in Hong Kong and which justify judicial review of both legislation and executive action.²⁴ The Hong Kong courts have generally proceeded on a two-step approach to analysing whether executive acts or legislation have infringed the Basic Law or Bill of Rights Ordinance.²⁵ First, the courts will consider whether a constitutional right has been infringed. If so, the courts will evaluate whether such infringement was justifiable by reference to the test of proportionality.²⁶ In applying the test of proportionality, the Hong Kong courts will consider whether a restriction on a constitutional right was rationally connected to a legitimate purpose, as well as whether it was a proportionate response to the legitimate objective in mind.²⁷

A similar two-step approach applies in Hong Kong to the context of equal protection challenges to government action. Articulated by the Hong Kong Court of Final Appeal in Secretary for Justice v Yau Yuk Lung (Yau Yuk Lung), this approach requires the court to first consider whether there had been a differentiation giving rise to a concern that like cases had not been treated alike, and if so, whether this differentiation satisfied the test of justification.²⁸ This test of justification is essentially a proportionality test.²⁹ The Court of Final Appeal has explained that while this test does require the court to evaluate the proportionality of the alleged restriction on the equality right, the court will generally not ‘put itself in a place of the executive or legislature or other authority to decide what is the best option’, and will ‘only interfere where the option chosen is clearly beyond the spectrum of reasonable options’.³⁰ However, where the case touches upon certain ‘fundamental concepts or core-values’, the court would be ‘particularly stringent or intense in the application of the justification test’.³¹

What is of particular interest is that in the context of executive action, the Hong Kong courts have several times placed significant weight on the first step of the approach – the issue of whether there has been a differentiation at all giving rise to a concern that like cases have not been treated alike³² – and have analysed this step by way of the formulation of classes and comparators. Such an approach is paradigmatic of a class-focused approach. As will be noted later, however, the Hong Kong Court of Final Appeal has not proceeded along the lines of a class-focused approach. This will be revisited later.

The Hong Kong Court of First Instance and Court of Appeal decisions in Fok Chun Wa v Hospital Authority ³³ provide useful illustrations of a class-focused approach. These cases concerned a challenge in judicial review brought against a decision of the hospital authority to reclassify non-residents of Hong Kong who were spouses of Hong Kong residents, specifically Mainland Chinese women who possessed two-way permits and who had applied for one-way permits to stay in Hong Kong, as persons not eligible to receive subsidised medical treatment at public hospitals. This challenge was brought under the equality provisions of the Basic Law and the Bill of Rights Ordinance. The applicants argued that Mainland Chinese spouses of Hong Kong residents living in Hong Kong who had applied for one-way permits were for all intents and purposes identical to Hong Kong resident women with Hong Kong resident spouses. Differentiating between these two categories of persons therefore amounted to discrimination, and such discrimination failed to satisfy the test of justification.

The Court of First Instance held that the applicants failed at the first stage. The court held that there was a fundamental difference between the groups of persons sought to be compared – the applicants were visitors to Hong Kong, while Hong Kong resident women were resident in Hong Kong. The positions of both groups of persons were therefore ‘materially different’.³⁴ Since both groups of persons belonged to different classes to begin with, there was no infringement of the principle that like cases ought to be treated alike. In any case, the court thought that the alleged differentiation would satisfy the test of justification.

The Court of Appeal, on the other hand, came to a different conclusion on the first stage of analysis. The court affirmed a doctrinal approach that can be characterised as a class-focused approach – the proper approach in the first stage of analysis was to ‘ask whether the applicant and the people who are treated differently are in “analogous situations”’.³⁵ Returning to the facts, the Court of Appeal disagreed with the lower court's finding that the relevant persons belonged to different classes. The Court of Appeal thought that ‘the comparators were in clearly analogous situations’.³⁶ This was because women non-resident in Hong Kong belonging to the applicant's class ‘bore children of Hong Kong resident fathers, children who upon birth would become Hong Kong residents themselves, where the woman and child were part of a family unit whose centre of life was Hong Kong and where each woman had a substantial de facto residential connection with Hong Kong’.³⁷

With the relevant comparators framed as where the family unit had their centre of life and substantial residential connection, the applicants and Hong Kong resident women with Hong Kong resident spouses were indeed in the same class. Accordingly, any differentiation between these persons in the same class fell to be justified at the second stage of analysis. Nevertheless, while the Court of Appeal disagreed with the lower court's assessment of the first stage of analysis, it agreed that the applicants’ challenge ought to fail on the basis that the differentiation in question was indeed justifiable under the Yau Yuk Lung test.

These two decisions provide a helpful example of a class-focused approach to equal protection analysis.³⁸ Indeed, one can observe in the courts’ reasoning a clear focus on determining the class of persons into which the applicants fell and the class of persons to which they should be compared. The courts’ articulation of the relevant classes was quite capable of being dispositive of the issue – if the applicants and the compared group did not fall within the same class to begin with, then no equal protection concern would arise at all, as the Court of First Instance held. As described earlier, this is a distinctive feature of a class-focused approach – on such an approach, the articulation of classes by the courts acquires a significant amount of weight in equal protection analysis and is capable of resolving equal protection challenges on its own.

(ii) Singapore – Syed Suhail (CA)

There are also indications in Singapore jurisprudence of the usage of a class-focused approach. By way of background, Singapore is a constitutional democracy, and a right to equal protection of the law is enshrined in Article 12 of its constitution.³⁹ The power of the Singapore judiciary to perform judicial review of both legislation and executive action on the basis of unconstitutionality has also been widely accepted.⁴⁰

The Singapore Court of Appeal, Singapore's apex appellate court, has recently paid specific attention to the question of how executive action should be assessed for violations of constitutional equality. Indeed, in Syed Suhail bin Syed Zin v Attorney-General ⁴¹ (Syed Suhail (CA)), the Court of Appeal articulated a new legal approach to equal protection analysis, an approach that can be characterised as a class-focused approach. The factual background of the Syed Suhail (CA) decision is the same as that of the Syed Suhail (HC) decision described earlier. To clarify the procedural relationship between the two identically-titled decisions, Syed Suhail (CA) concerned an application for leave for the judicial review application to proceed, while Syed Suhail (HC) was the High Court's decision on the merits after the Court of Appeal had granted leave for judicial review in Syed Suhail (CA).

Faced with an equal protection challenge to the scheduling of executions, the Singapore Court of Appeal in Syed Suhail (CA) articulated a new legal approach to equal protection challenges against executive action. The court acknowledged that the hitherto-prevailing approach to such challenges was the test of deliberate and arbitrary discrimination. Outlining several problems with this test,⁴² the Court of Appeal held that a two-step approach ought to apply to analyse equal protection challenges to executive action generally. First, the applicant would bear the evidential burden of proving that he was treated differently from other ‘equally situated’ persons.⁴³ Secondly, should the applicant succeed in discharging this burden, the evidential burden would shift to the decision-maker in question to prove that this differential treatment was ‘reasonable in that it was based on legitimate reasons’.⁴⁴ The reasonableness of this differential treatment would be assessed by determining whether such treatment bore ‘a sufficient rational relation to the object for which the power was conferred’.⁴⁵ This new test bears some parallels to the requirements of the reasonable classification test, a test which remains the prevailing approach to equal protection challenges to legislation in Singapore.⁴⁶

Applying this new approach to the facts, the Court of Appeal found that there was some inconsistency in the tendered evidence as to whether the applicant was first to be sentenced to death among equally situated prisoners. Accordingly, the court considered that the threshold requirement for leave to be granted to commence judicial review was met⁴⁷ – which led then to the earlier-mentioned High Court decision in Syed Suhail (HC) on the merits of the judicial review application.

What is noteworthy for present purposes is that the Syed Suhail (CA) test amounts to a class-focused approach. Indeed, it requires the court to decide whether an applicant is ‘equally situated’ with an entity being compared to – in other words, whether the compared entities are in the same class. Judicial declarations of what are the appropriate classes and comparators would therefore be quite capable of being dispositive of equal protection challenges, even if how these classes and comparators have been formulated receives little discussion. One might observe that the High Court's decision on the merits in Syed Suhail (HC), described earlier, adopted precisely such an approach. The High Court essentially declared that the relevant comparator was the potential for one's case to be reviewed and reopened, with limited justification proffered for the selection of this comparator, and the selection of this comparator had fatal consequences for the applicant's equal protection challenge.⁴⁸

(iii) Evaluating the class-focused approach

It is therefore clear that a class-focused approach has gained some traction in the equal protection jurisprudence of Hong Kong and Singapore. A key reason for its appeal is undoubtedly that it is directly connected to a foundational principle behind equal protection: like cases should be treated alike.

A class-focused approach is, however, not devoid of difficulty. There are two major issues with such an approach to equal protection, especially in the context of executive action. First, such an approach is not well-equipped to provide a principled means by which one can determine what is the appropriate comparator to adopt for the purpose of deciding whether the compared entities are alike to begin with. The differing views between the Court of First Instance and the Court of Appeal in Fok Chun Wa v Hospital Authority as to whether the groups of persons being compared were in similar classes highlights the ambiguity in determining who is equally situated under a class-focused approach. Depending on whether the appropriate comparator was taken to be Hong Kong residency status or where the family unit had their centre of life, one could come to very different views on whether the groups of persons being compared belonged to similar classes – with consequently significant implications for equal protection challenges. Given this ambiguity as to the selection of comparators, an important problem with a class-focused approach is that it provides little guidance in itself as to which comparator to adopt. This opens the doorway for unprincipled and arbitrary reasoning.

A second and related issue with a class-focused approach is that without the provision of any principled means by which the relevant comparators should be framed, the courts may perform such framing by referring to implicit but unarticulated notions of what kinds of differential treatment are justifiable. In other words, a finding that the compared entities do indeed belong to similar classes, with differential treatment therefore giving rise to an equal protection concern, might actually be implicitly reliant on an assessment that the relevant executive decision has resulted in unjustifiable differential treatment. This would raise the concern of a lack of candour. Despite the ostensible separation of both stages of analysis, both steps would in fact be closely related to each other, since the first step of determining whether there is an equal protection concern to begin with would already incorporate an assessment of whether the differential treatment is justifiable.

Further, if it is indeed correct that the first step of analysis in a class-focused approach already allows for (or perhaps even requires) the incorporation of normative judgments about what amounts to equal treatment, then an additional danger arises: the application of such a step might risk obscuring the judicial incorporation of contentious value judgments about the requirements of equality at this stage of analysis. Indeed, if such contentious value judgments have to be incorporated in judicial reasoning, one might think that the preferred approach would be to incorporate them transparently – a point which segues nicely to our discussion of the next approach.

(b) Policy-focused approach

An alternative approach to equal protection analysis in the context of executive action can be characterised as a policy-focused approach. It is similar to a class-focused approach in the sense that it also adopts a two-step framework: has there been a differentiation giving rise to an equal protection concern at all, and if so, is such a differentiation justifiable? However, instead of answering the first question by reference to posited classes, a policy-focused approach would do so by shifting attention away from the formulation of classes and instead seek to directly discern the substantive moral requirements of equality in the specific context of each decision. Whether there is differential treatment giving rise to an equal protection concern that has to be justified would depend substantially on how the court articulates the requirements of substantive moral equality.

(i) Hong Kong

A useful illustration of such an approach is the Hong Kong Court of Appeal's decision in QT v Director of Immigration.⁴⁹ In this case, the applicant had entered into a same-sex civil partnership in England with her partner. Her partner was granted an employment visa in Hong Kong, but the applicant was denied a dependant visa on the basis that she was not in a monogamous marriage between a male and a female. She subsequently challenged the decision, arguing that she had been unlawfully discriminated against for her sexual orientation.

The Court of First Instance rejected the judicial review challenges. On appeal, however, the Court of Appeal held that the Director of Immigration's decision was indeed problematic on equal protection grounds. The judgment of Andrew Cheung CJHC (as he then was) is of particular significance for present purposes. Indeed, Cheung CJHC paid very close attention to the substantive moral requirements of equality in the specific context at hand. Reasoning that the issue at hand was inextricably tied to the nature of marriage, Cheung CJHC discussed in detail what the core rights and obligations unique to a marital relationship were. In Cheung CJHC's view, matters relating to divorce, adoption, or inheritance were examples of matters where the ‘status of marriage carries rights and obligations unique to married couples’.⁵⁰

This substantiation of the core rights and obligations connected with marriage was translated into the equal protection context in the following manner. Cheung CJHC considered that if a differentiation was based upon a matter falling within these unique core rights and obligations, then the differentiation would ipso facto be rendered constitutional – in such situations, ‘the status of marriage provides the obvious, relevant difference between a married couple and one that is not’.⁵¹ If a differentiation was not based on a matter falling within these core rights and obligations, then the differentiation would have to be subject to the test of justification.⁵² Cheung CJHC offered a variety of arguments in support of such an approach as opposed to an approach which would require all differentiations to be justifiable, several of which centred around the special significance of the marital relationship as a social and legal institution.⁵³

Applying this approach to the facts, Cheung CJHC held that the issue at hand related to immigration, which did not fall within the core group of rights and obligations unique to marriage.⁵⁴ This accordingly meant that the differentiation was not rendered ipso facto constitutional on the basis of the special status of marriage, and that the differentiation had to satisfy the test of justification. Turning then to evaluate the Director's proffered justifications for the differentiation, Cheung CJHC considered that the test of justification was not met – the justification based upon attracting talent to Hong Kong was plainly contradicted by the differentiation at hand, and the justification based upon maintaining strict and stringent immigration control had little connection with a differentiation based upon the sexual orientation of the spouse.⁵⁵ Since these were the only justifications offered by the Director, Cheung CJHC found that the differentiation was unconstitutional.⁵⁶

It will be observed subsequently that when this decision was appealed, the Hong Kong Court of Final Appeal took a rather different approach towards analysing the equal protection challenge in this case. At this juncture, however, the key point to emphasise is that Cheung CJHC's approach represents an example of what a policy-focused approach looks like. Indeed, the emphasis paid in his decision to the substantive moral requirements of equality in the context of the equal protection challenge being brought, and his application of such value judgements as a means to resolve equal protection issues, are all indications of a policy-focused approach.

(ii) Evaluating the policy-focused approach

A policy-focused approach offers an alternative approach to equal protection analysis where executive action is concerned. But it does present its own problems. Its advantage over a class-focused approach – requiring judges to reason about what equality requires in each context – is a double-edged sword. Indeed, insofar as such reasoning requires a moral assessment of the requirements of equality, one may be concerned about the prospect of judges bearing the responsibility of substantiating these requirements.

This concern is exacerbated by the fact that such moral assessments can be very contentious. Consider, for example, an executive decision to deny the public housing application of a couple in a same-sex marriage entered into in a foreign jurisdiction because the applicant's jurisdiction does not recognise such marriages as legal. What would substantive equality require in such a context? A couple of options are possible. One might think that substantive equality requires the equal treatment of housing applications between same-sex and opposite-sex marriages. Others might think that substantive equality simply requires the public authority to treat equally housing applications from couples married legally under the laws of its jurisdiction.

It is clear that what one thinks about the requirements of substantive equality in this context will have a major impact on whether there is an equal protection concern at all in such a situation. And what one thinks about the requirements of substantive equality in this context will obviously very much depend on one's moral assessment of the requirements of equality.

As mentioned earlier, such moral assessments may also occur at an implicit level even where a class-focused approach is adopted. One may, therefore, argue that an express incorporation of such moral assessments would at least possess the advantage of candour. This is indeed an advantage of a policy-focused approach over a class-focused approach. Nevertheless, notwithstanding the advantages of a policy-focused approach relative to a class-focused approach, it ought to be recognised that a policy-focused approach would independently raise concerns about the wisdom of opening a doorway for morally contentious considerations to directly influence legal reasoning – and would prompt one to wonder if any alternative approaches may exist. It is to such an alternative that we will next turn.

(c) Justification-focused approach

The final approach we will discuss can be characterised as a justification-focused approach. A justification-focused approach would not be centred on declaring or discerning comparators by which one can determine whether there has been differential treatment triggering the equal protection guarantee to begin with. It is instead, as its terminology suggests, focused on the justifiability of any alleged factual differences in treatment.

Put another way, both the class-focused and policy-focused approaches place significant weight on a first step requiring judges to determine whether there had been differential treatment giving rise to an equal protection concern, with judges shifting to the second step of evaluating the justifiability of such differential treatment only if the first step had been satisfied. The differences between the class-focused and policy-focused approaches lie in how they seek to address the first step of analysis. In comparison, a justification-focused approach would minimise the significance of the first step and place much greater weight on the second. The first step would be satisfied as long as there was a factual difference in treatment between alleged compared entities. The court's analysis would then be focused primarily on whether this factual difference was justifiable, with the court evaluating the relevant authority's reasons for the differentiation on the basis of rationality or proportionality. Indeed, a justification-focused approach is a direct expression of the idea, referenced earlier, that equality is closely connected to the concept of justification – justification can be described as the principal element in determining whether like cases are alike.

(i) Hong Kong – the Court of Final Appeal's approach

A justification-focused approach was in substance the approach taken by several decisions of the Hong Kong Court of Final Appeal to the analysis of equal protection challenges to executive action.

The Hong Kong Court of Final Appeal decision in Fok Chun Wa v Hospital Authority ⁵⁷ provides a useful illustration. The Court of First Instance and Court of Appeal decisions in this case have already been discussed, and it will be recalled that both courts applied a class-focused approach to the question of whether there was a differentiation giving rise to an equal protection concern.

The Court of Final Appeal, at the outset of its analysis of the equal protection challenge in this case, affirmed a doctrinal approach closely cohering with a class-focused approach.⁵⁸ Interestingly, however, the court went on to express concern about the problems that may ensue if such an approach were applied in an overly rigid fashion. Indeed, the Court of Final Appeal suggested that the single overall objective that courts should remain focused on in dealing with equal protection challenges is whether there was ‘enough of a relevant difference between X and Y [the comparators] to justify differential treatment’,⁵⁹ citing Lord Hoffmann's reasoning in R (Carson) v Secretary of State for Works and Pensions with approval.⁶⁰ Crucially, the court went further to suggest that in most cases, the two stages of the relevant approach would not be separable from each other, and that the first step of whether the compared persons were indeed in like situations would generally have to be answered by considering whether the differential treatment was justifiable. The Court of Final Appeal opined that ‘it may matter not at all whether the court's approach is seen as a two-stage one or not’,⁶¹ and that ‘in most questions involving the right to equality, there will be an overlap in the application of the two-stage test set out in Yau Yuk Lung’.⁶²

Turning to the facts, the Court of Final Appeal adopted a justification-focused approach to evaluate the equal protection challenge at hand. Instead of framing classes at the first step of the Yau Yuk Lung test, the court preferred to go straight to the question of whether the differentiation between the groups of persons being compared was justifiable.⁶³ The court paid attention to the importance of granting a margin of appreciation, ‘particularly in circumstances where the court is asked to examine issues involving socio-economic policy’⁶⁴ – which the case at hand involved. In such circumstances involving the allocation of public funds, the Court of Final Appeal held that decisions to demarcate who would be eligible for publicly subsidised healthcare services were ‘part of the Government's socio-economic responsibilities’, and that ‘it is no part of the court's role to second-guess the wisdom of these policies and measures’.⁶⁵ Accordingly, since the differentiation based on residence status was ‘within the spectrum of reasonableness’, the Court of Final Appeal ultimately held that the equal protection challenge was unsuccessful.⁶⁶

This justification-focused approach to analysing equal protection challenges to executive action was affirmed and applied in subsequent decisions of the Hong Kong Court of Final Appeal. Indeed, in the landmark equal protection decision of QT v Director of Immigration ⁶⁷ (QT), the Court of Final Appeal similarly applied an approach that can be characterised as a justification-focused one – a starkly different approach from that taken by Cheung CJHC at the Court of Appeal, as discussed earlier.

In response to one of the Director's arguments which had adopted a class-focused approach,⁶⁸ the Court of Final Appeal in QT expressed dissatisfaction with proceeding on a line of analysis focused on identifying comparators.⁶⁹ In the court's view, ‘the notion of whether the comparators are analogous or relevantly similar is elastic both linguistically and conceptually’.⁷⁰ Affirming the Court of Final Appeal's approach in Fok Chun Wa v Hospital Authority, the court in QT observed that the two stages of the Yau Yuk Lung test were intricately connected with each other – ‘where an issue of equality before the law arises, the question of whether a measure is discriminatory is necessarily bound up with whether the differential treatment which the measure entails can be justified’.⁷¹ Accordingly, the Court of Final Appeal thought that the proper approach to equal protection challenges would be to ‘examine every alleged case of discrimination to see if the difference in treatment can be justified’.⁷²

Applying this approach to the facts, the court held that a problem with the Director's decision was that it was not rationally connected with any legitimate objective. The objectives of bringing in talent to Hong Kong, ensuring strict immigration control, and setting out bright lines to facilitate the administration of immigration controls were examined in turn, with the court's ultimate conclusion being that the differentiation relied on by the Director was not rationally connected with any of these objectives. As such, the court upheld the equal protection challenge against the Director's decision.

This study of Hong Kong equal protection jurisprudence suggests that while the expressly articulated doctrinal framework to equal protection challenges in Hong Kong may be a class-focused approach on its face, the manner in which the Hong Kong Court of Final Appeal has actually applied this framework is more closely aligned with a justification-focused approach.⁷³ Indeed, the court's reasoning in QT provides a useful articulation of what a justification-focused approach would entail. Instead of framing classes and comparators, a justification-focused approach would be analytically focused on the justification for any alleged differential treatment. In contrast to Cheung CJHC's approach in the same case, the Court of Final Appeal's approach placed a central emphasis on the step of justification, negating the need to first determine whether there had been a departure from the substantive requirements of equality in the relevant context.

(ii) Evaluating the justification-focused approach

A justification-focused approach presents a credible alternative to a class-focused or policy-focused approach in analysing equal protection challenges to executive action. Indeed, the use of a justification-focused approach to effectuate a general equality right would track closely with developments in administrative law jurisprudence. Jowell and McCrudden have conducted detailed studies of case law to illustrate the point that judges have often specified legal equality as a requirement of rational justification.⁷⁴ The doctrinal connection between equality and rational justification is buttressed further by the development of the principle of consistency in UK administrative law. In pursuit of the normative ideal of consistency in public administration, such a doctrine would require a public authority seeking to depart from its publicly-stated policy to provide a rational justification for such a departure, failing which the departure would be considered unlawful.⁷⁵ Several decisions of the UK Supreme Court suggest that such a principle of consistency exists as a freestanding ground of review in administrative law,⁷⁶ and commentators such as Christopher Forsyth, Jason Varuhas and Mark Elliott have made powerful arguments in favour of such a doctrine.⁷⁷ Given the close connection between the normative ideals of equality and consistency, specifying a general equality right by way of a justification-focused approach would be very much aligned with such developments in administrative law jurisprudence. Further, to the extent that one accepts the body of scholarship suggesting that foundational common law principles of judicial review such as the requirement of rational justification in fact have constitutional foundations and are grounded in fundamental principles of the rule of law,⁷⁸ this would lend additional weight to the view that a general equality right in constitutional law can be aptly articulated by way of a justification-focused approach.

However, the adoption of such an approach faces a significant obstacle in the context of jurisdictions enshrining equal protection guarantees in their written constitutions – the jurisdictions which are the focus of this study. One may be concerned that a justification-focused approach to equal protection analysis would fail to give sufficient effect to the idea that equal protection is a constitutional right.

Indeed, Rowe J's concurring decision in Law Society of British Columbia v Trinity Western University ⁷⁹ contained an argument against a justification-focused approach. He suggested that the Supreme Court of Canada has sometimes favoured an approach to analysing challenges under the Canadian Charter of Rights and Freedoms which did not place much emphasis on determining first the scope of a right for the purposes of finding out whether it had been infringed in the first place, but instead focused the court's attention on the question of justification.⁸⁰ Rowe J expressed doubts about such an approach. He thought that doing so would blur ‘the distinction between infringement and justification’, going against ‘the architecture of the Charter’.⁸¹ Rowe J thought that analysis of fundamental liberties under the Charter required ‘a structured two-step process’: first, an analysis of whether a right had been infringed, which required a definition of the scope of the right; and secondly, a consideration of whether the infringement was justified.⁸² Conflating the two steps of analysis would, in Rowe J's view, result in ‘an unstructured, somewhat conclusory exercise that ignores the framing of the Charter and departs fundamentally from the foundational Charter jurisprudence of this Court’.⁸³ Accordingly, Rowe J would almost certainly not be in favour of a justification-focused approach to equal protection analysis.

The Singapore Court of Appeal decision in Syed Suhail (CA) raised a related concern with a justification-focused approach to equal protection analysis. In evaluating the previously prevailing deliberate and arbitrary discrimination test – which in substance is a justification-focused approach – one of the concerns that the court raised with this test was that it would ‘render Article 12(1) nugatory so far as it related to executive action’, since such a test was similar to the common law judicial review ground of irrationality.⁸⁴ Put another way, the court was of the view that since equal protection was a constitutional right, the legal approach to equal protection analysis should secure a greater degree of protection for the right than regular administrative law doctrines. Analysing the constitutional right of equal protection in a manner similar to that which administrative law doctrine would provide would not give effect to the special status of equal protection as a constitutional right.

These are important concerns that ought to be addressed. The first point that can be made in response is that it is questionable whether placing more emphasis on the first stage of analysis would indeed contribute anything meaningful to the analysis in the context of equal protection challenges to executive action. Rowe J's preferred mode of analysis is much easier to apply in the context of constitutional rights such as the right to life or the right to religious freedom. Such rights are amenable to substantiation of their scope – for example, a substantiation of what ‘life’ means would help demarcate the scope of the right. But the right to equal protection is by nature a comparative right. It is much less amenable to a priori substantiation of its scope in the absence of expressly enumerated categories of impermissible differentiation. It was an awareness of this issue that led the Hong Kong Court of Final Appeal to point out that the analysis at the first stage, in the context of equal protection, is necessarily connected to the test of justification. If so, then for the purposes of equal protection analysis, the first stage collapses into the second stage – the evaluation of whether any factual differentiations are justifiable.

To illustrate the point that placing more emphasis on the first stage of analysis in the context of equal protection challenges to executive action is unlikely to add anything meaningful to the court's reasoning, a simple example can be referred to. Consider a situation where a professor is grading student examinations. Grades are assigned to individual students. One student gets a B while another student gets an A. The first student alleges that he has been subject to impermissible discriminatory treatment and files a complaint with the school principal.

This equality-based challenge can be addressed in a variety of ways. The principal could adopt a class-focused approach, and frame the relevant comparators as quality of critical analysis or breadth of coverage of course material. Applying these comparators, the student receiving a B had turned in an examination that was not in a similar class as the student who received an A. There was therefore no contravention of the principle that like cases should be treated alike to begin with.

Alternatively, the principal could adopt a policy-focused approach. The principal could discern what equality requires in the specific context of grading. The principal might decide that what substantive equality requires in this context is that students with equally strong scripts will get equal grades. The strength of each script would depend on factors like the quality of critical analysis or breadth of coverage of the course material. Accordingly, the principal could decide that given that the strength of each script was not equal, there was no contravention of the requirements of equality here by giving the students different grades.

Finally, the principal could adopt a justification-focused approach. The principal could observe that there was a factual differentiation between the two students – one received an A and the other received a B. Given this factual difference, the principal could then move to consider whether this differentiation was justified. He could assess whether the differentiation was justified by reference to factors such as the quality of critical analysis and the breadth of coverage of course material. The principal could then decide that there was no impermissible discriminatory treatment, since the factual differentiation was well-justified by the differences between the two scripts on these factors.

The point of this illustration is to highlight that across all three approaches, the same considerations are taken into account. Adopting a class-focused or policy-focused approach to place more emphasis on the first stage of analysis does not serve to draw the principal's attention to anything that he may not have already considered under a justification-focused approach. It is therefore quite doubtful whether Rowe J's suggested two-step framework would add anything meaningful to the analysis in the context of equal protection challenges. Indeed, a justification-focused approach in fact serves to enhance the candour of legal reasoning, ensuring the closest correlation between the articulated legal doctrine and the reasons which are actually performing the heavy-lifting in helping judges reach their conclusions.

What this means, therefore, is that if one were minded to safeguard the distinctiveness of constitutional equal protection analysis as opposed to administrative law by emphasising the importance of the first stage of analysis (as the class-focused and policy-focused approaches would do), it is unlikely that doing so would be an effective solution. One might in fact argue that a justification-focused approach, which places more weight upon the justificatory stage of equal protection analysis, would allow for a more candid and principled means of weighing the relevant considerations in analysing equal protection challenges. Indeed, if the resolution of equal protection challenges can often implicate contentious value judgments, it would be desirable to address such issues by way of the sophisticated tools which courts have already developed for use at the justificatory stage of analysis.

The concern raised by the Singapore Court of Appeal in Syed Suhail (CA) remains to be directly addressed, however – namely, should a justification-focused approach be adopted to effectuate a general equality right in legal doctrine, might there be a structural concern insofar as constitutional equality and administrative law would require substantially the same mode of analysis, even though they both exist at different levels in the legal order?

This question will benefit from a more detailed standalone analysis, unconstrained by space. But some thoughts will be proffered here. The similarity between administrative law principles and a justification-focused approach to effectuating a general equality right prompts three possible paths of legal development.

The first would be to accept that constitutional equality does nothing more than administrative law principles. The doctrinal way forward would then simply be to view constitutional equality as a transposition of existing grounds of judicial review in administrative law to the constitutional level. This pathway of development would be supported by academic literature from authors such as Westen and Bernard Williams highlighting the empty nature of the ideal of equality and the difficulties involved in substantiating the notion of equality ab initio.⁸⁵ Lurie's argument, referenced earlier, on the congruence between equality and proportionality analysis would lend further support to this pathway.⁸⁶ As much as this pathway may be supported by such reflections on the nature of equality, it is, however, quite difficult to see a future for it as a doctrinal way forward, given the weight of authority in favour of the elevation of equality as a constitutional principle driving the instinct of courts to seek to differentiate constitutional equality from administrative law – as the Singapore Court of Appeal in Syed Suhail (CA) sought to do.

The second pathway would similarly accept that general equality rights ought to be specified by way of equality as rationality, and would adopt the justification-focused approach as the best possible doctrinal means of doing so. This second pathway, however, in contrast to the first, would seek to draw a meaningful distinction between constitutional equality and administrative law principles. This could be done by applying different standards of review – in other words, proportionality review would apply where a differentiation implicating a general equality right has occurred, while ordinary reasonableness review would apply for challenges to executive action more broadly. In this regard, one may emphasise that the former is directed at a specific type of unjustifiable executive acts – where there are inadequately justified factual differentiations between entities – whereas reasonableness review is directed more broadly at all kinds of unreasonable executive acts.

The weight of authority would be in favour of this pathway of legal development. Indeed, this is ostensibly the approach that the Hong Kong courts have taken.⁸⁷ If this pathway of legal development were to be adopted in Singapore, however, this may prompt the need for a further doctrinal development: it would be necessary for the Singapore courts to accept proportionality review as the standard of review for constitutional equality in order to adequately distinguish constitutional equality from administrative law, a standard of review which the Singapore courts have thus far been hesitant to accept.⁸⁸

The preceding discussion also prompts a third, more radical, possible pathway of legal development that is worth raising. One might argue that if the effectuation of general equality rights through the idea of equality as rationality is the principal reason for the overlap between constitutional equality and administrative law, then a possible means of avoiding such overlap is to effectuate general equality rights through the alternative idea of equality as impermissible discrimination. One might be more inclined to recommend such a pathway to the extent that one prefers a more substantive vision of equality – for instance, where one is concerned that equality as rationality leads to an impoverished vision of constitutional equality, incapable of adequately addressing issues of substantive equality such as systemic discrimination against disadvantaged minorities.⁸⁹ Nevertheless, such an approach would raise concerns of its own – for example, the issue of how suspect classifications should be articulated – and would represent a radical departure from the weight of authority in both jurisdictions discussed in this paper. Accordingly, pending a more detailed interrogation of these issues, it is tentatively suggested that the second pathway of legal development is the preferred option.

Conclusion

It is therefore proposed that out of the three possible approaches to equal protection analysis based on a general right to equality in the context of executive action surveyed in this paper, a justification-focused approach is the most desirable. Such an approach, in contrast to a class-focused or policy-focused approach, would place an analytical emphasis on the question of justification.

This approach brings several important advantages. In the context of discrete executive action, discerning whether there has been a differentiation at all that can give rise to an equal protection concern is particularly challenging given the wide variety of comparators that can be selected. And as discussed earlier, the comparator that is selected can have a major impact on the equal protection challenge at hand – if one determines that the compared entities were not in analogous classes by reference to the chosen comparator, there would be no question of like cases not being treated alike to begin with and the equal protection challenge would fail. A justification-focused approach would avoid the potential for arbitrariness that can arise from such an exercise.

A justification-focused approach would also avoid the separation of powers concern that a policy-focused approach may entail. Instead of requiring judges to articulate the substantive requirements of equality in each context, and then determining whether there had been any contravention of these requirements on the facts, a justification-focused approach would instead place the burden of justification on the decision-maker and require the court to evaluate the reasons proffered by the decision-maker – an exercise well within the court's expertise, minimising the danger of the court entrenching contentious moral judgements about the requirements of equality in equal protection jurisprudence.

Finally, a justification-focused approach would promote clarity and candour of legal reasoning. It would provide for the most direct correlation between the expressly articulated legal doctrine and the reasons actually influential in directing judges to reach their conclusions, instead of cloaking such reasons behind the articulation of classes and comparators.

The discussion in this paper is envisioned to be most directly relevant to any common law jurisdiction containing generalised equal protection guarantees in its written constitution, although the ideas offered in this paper can also be of relevance to any jurisdiction seeking to apply constitutional or quasi-constitutional guarantees of equality to executive action. Other questions yet remain to be investigated further – for instance, how the relationship between constitutional equality and administrative law ought to be rationalised, a question touched upon here but which deserves separate analysis beyond the confines of this paper. Nevertheless, it is hoped indeed that this paper will contribute to the principled development of doctrinal frameworks governing equal protection challenges to executive action, and that the analysis offered here will provide a firm foundation for the resolution of these further challenges.