THE HONG KONG HOUSING AUTHORITY v. INFINGER, NICK

Press Summary (English)

Press Summary (Chinese)

FACV Nos 2 and 3 of 2024

[2024] HKCFA 29

FACV No 2 of 2024

IN THE COURT OF FINAL APPEAL OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

FINAL APPEAL NO 2 OF 2024 (CIVIL)

(ON APPEAL FROM CACV NO 81 OF 2020)

__________________________

__________________________

FACV No 3 of 2024

IN THE COURT OF FINAL APPEAL OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

FINAL APPEAL NO 3 OF 2024 (CIVIL)

(ON APPEAL FROM CACV NO 362 OF 2021)

__________________________

__________________________

(HEARD TOGETHER)

_______________________

J U D G M E N T

_______________________

Chief Justice Cheung:

1.  These two appeals by the Hong Kong Housing Authority (“HA”), which have been heard together, concern the HA’s family provision policies for public rental housing (“PRH”) and the Home Ownership Scheme (“HOS”) as they apply to same-sex couples lawfully married abroad.

2.  Following two separate applications for judicial review, the Court of First Instance struck down the respective policies for discrimination.[1]  Both decisions were affirmed by the Court of Appeal, which heard the appeals at the same time.[2]

HA and its PRH & HOS policies

3.  The HA, established under section 3 of the Housing Ordinance (Cap 283), is tasked with carrying out significant governmental functions related to public housing. Its statutory mandate, as set out in section 4(1) of the Ordinance, is to provide housing and associated amenities for those classes of persons as it may determine, subject to the approval of the Chief Executive. In practice, the HA’s main responsibility is to ensure the availability of affordable housing for low-income families.[3] To a lesser extent, it is also in charge of facilitating home ownership for low- to middle-income households.[4]  Its overarching objective is to address the pressing housing needs of the most vulnerable members of society.[5]

4.  One of the primary functions of the HA is the administration of PRH.[6]  PRH units are heavily subsidised and offered to those whose financial circumstances preclude them from accessing the private rental market.[7]  Given the perennial challenge of demand far outstripping supply, the HA maintains an application system with stringent criteria. These include, but are not limited to, age, income, and asset limits, so as to ensure the fair and rational distribution of limited resources amongst those in need of housing.[8]  Within this system, there are two main categories of PRH unit applications: (1) “General Applications”, namely applications by family applicants with a household size of two or more persons and those by elderly one-person applicants;[9] and (2) “Applications by Non-elderly One-person Applicants”.[10]

5.  Under the first main category (“General Applications”), there are four sub-categories: (i) the “Ordinary Families”; (ii) the “Single Elderly Persons Priority Scheme”; (iii) the “Elderly Persons Priority Scheme”; and (iv) the “Harmonious Families Priority Scheme”.[11]

6.  For the “Ordinary Families” sub-category, only certain familial relationships are recognised for application purposes – the relationship between the applicant and their family members, as well as that between family members themselves, must be “either husband and wife, parent and child, or grandparent and grandchild”.[12]  An applicant may also apply with their single sibling(s).[13]  Notably, the marital relationship of a same-sex couple that is lawfully married overseas, unlike its opposite-sex counterpart, is not recognised. Consequently, same-sex couples are ineligible to apply for a shared PRH unit as spouses under this policy framework, but can only apply as separate individuals.[14]

7.  In terms of processing time, the “General Applications” category enjoys priority over the category for “Applications by Non-elderly One-person Applicants”.[15] Within the “General Applications” category, the three “Priority” sub-categories[16] are given precedence over the “Ordinary Families” sub-category.[17]  However, amongst the relationships recognised under the “Ordinary Families” sub-category, no distinction is made; all applicants join the same queue.[18]  As for the “Applications by Non-elderly One-person Applicants” category, the HA operates the “Quota and Points System” to determine priority for PRH unit allocation.[19]   

8.  According to the evidence filed, the average waiting time (“AWT”) for general applicants (i.e. family and elderly one-person applicants) as of 30 September 2018 was 5.5 years,[20] with the AWT for elderly one-person applicants being 2.9 years.[21]  The AWT for non-elderly one-person applicants, though not available,[22] must have been even longer than 5.5 years, given the low priority accorded to them as a class.[23]  In the first six months of the 2018/2019 financial year, approximately 12,000 PRH units were allocated to PRH applicants of all types.[24]

9.  Further to its primary role in PRH, the HA also administers the HOS, introduced in 1978.[25]  The HOS serves two purposes: first, it encourages financially better-off PRH tenants to vacate their units, thereby allowing these units to be re-allocated to more needy applicants;[26] second, it provides an opportunity for low- to middle-income families to purchase HOS flats at concessional prices.[27]  Applicants for both new and second-hand HOS flats are divided into two categories: (1) “Green Form applicants”, who are PRH tenants at the time of application and must relinquish their PRH units upon purchase;[28] and (2) “White Form applicants”, who are mainly households in private accommodation or family members of PRH tenants.[29]

10.  HOS flats are subject to restrictions, both in terms of use and alienation.[30]  They must be used solely for residential purposes by the purchaser and their registered family members.[31]  Applications to purchase HOS flats are restricted to households (not individuals) consisting of persons in a “blood or legal relationship”.[32]  In addition to opposite-sex married couples, an eligible household may comprise the applicant and their children, parents, grandparents, parents-in-law, grandparents-in-law, unmarried siblings, unmarried brothers-in-law or sisters-in-law, or other relatives dependent on the applicant.[33]  Again, same-sex couples lawfully married overseas cannot apply to purchase HOS flats as couples.

11.  Moreover, the HA’s policy on addition of authorised occupants does not permit the same-sex spouse of an HOS flat owner to be so added.[34] Nor does its policy on transfer of ownership without a premium cater for a transfer of ownership of the flat in the same-sex spouse’s favour as joint owner.[35]  By contrast, the opposite-sex spouse of an HOS flat owner, whether married locally or overseas (where the marriage is recognised in Hong Kong), is eligible for both addition and transfer under the HOS policies.[36]

12.  The HA’s PRH policy, and its HOS policies on addition and transfer, as applied to same-sex couples married overseas, were challenged in the two sets of judicial review proceedings below (the “PRH proceedings” and “HOS proceedings” respectively).[37] They continue to be the focal points of these appeals.

Facts and procedural history

13.  In the PRH proceedings, the applicant for judicial review, Mr Nick Infinger, married his husband in Canada in January 2018. With his same-sex spouse as his only family member, he applied for a shared PRH unit under the “Ordinary Families” sub-category. Pursuant to its policy, the HA rejected the application.[38] In November 2018, Mr Infinger initiated judicial review proceedings to challenge the policy on the grounds of discrimination against same-sex married couples, along with other grounds.[39]

14.  In the HOS proceedings, the original applicant for judicial review, Mr Edgar Ng, and the substituted applicant, Mr Li Yik Ho, were a gay couple lawfully married in the United Kingdom in 2017. Mr Ng, a former PRH tenant, was the registered owner of an HOS flat purchased on the HOS secondary market in April 2018 in his own name. Although, according to the evidence, Mr Li contributed more than 90% of the financial costs, Mr Ng could only purchase the flat as sole owner, as the HA’s purchase policy did not recognise his same-sex relationship with Mr Li.[40]

15.  As explained above, the HA’s policies also rendered Mr Li ineligible to be added as an occupant of the flat as Mr Ng’s same-sex spouse, or considered for a transfer of ownership of the flat without payment of a premium.[41]  In September 2019, Mr Ng commenced judicial review proceedings to challenge the exclusionary HOS policies on grounds including discrimination.[42]  Unfortunately, Mr Ng passed away on 7 December 2020, and Mr Li, as his surviving spouse under their foreign same-sex marriage, was substituted as the applicant in the HOS proceedings.[43]

16.  The two sets of proceedings came before Chow JA.[44] After separate hearings, the learned judge held that the respective policies of the HA discriminated against same-sex couples lawfully married overseas and declared them unlawful and unconstitutional.[45]  On appeal, the Court of Appeal held the same view and dismissed both appeals accordingly.[46]

17.  With leave to appeal granted by the Court of Appeal in each of the two appeals, on the basis that they each raised questions of great general or public importance,[47] these appeals now come before us for final adjudication.[48]

Approach to constitutional challenges based on discrimination

18.  The general approach to constitutional challenges is well-established. As this court has recently explained,[49] this approach involves five stages: (1) the court identifies the constitutional right relied on and the impugned measure; (2) it asks whether and on what grounds the impugned measure is said to encroach upon and thus to engage the right; (3) if there is such encroachment and the ground is that the encroachment excessively and unjustifiably infringes that right, the court undertakes the Hysan four-step proportionality test[50] to determine if this is indeed the case; (4) if the impugned measure fails to satisfy the Hysan test, the court proceeds to consider whether any remedial order should be made to preserve the validity of the impugned measure in whole or in part; and (5) if no such remedial possibilities exist, the court declares the impugned measure unconstitutional and invalid.

19.  Essentially the same approach applies to challenges based on the constitutional rights to equality and non-discrimination.[51]  That is, the court must first identify the rule or policy subject to challenge and determine whether the rights to equality and non-discrimination are engaged.[52]  Those rights are engaged if the court is satisfied that differential treatment exists in relation to a person in a comparable position.[53]  If the court concludes in the affirmative, it will proceed to examine the justification advanced by the rule/policy-maker to sustain the differential treatment.[54]  This justification is scrutinised under the Hysan four-stage proportionality analysis involving:[55] (1) identifying a legitimate aim; (2) establishing a rational connection between the challenged measure and its aim; (3) reviewing whether the measure is proportionate;[56] and (4) verifying that a reasonable balance is struck between the societal benefits of the measure and the inroads made into the rights of the affected individuals.[57]  Where the rule or policy cannot be so justified, the court will consider if it can nevertheless be preserved by a remedial interpretation.[58]  If not, the court will strike it down as being unconstitutional and invalid.[59]

20.  Furthermore, in a series of cases involving same-sex unions,[60] this court has explained and demonstrated how the legal principles on equality and non-discrimination should be applied in the context of the unique situations and challenges faced by same-sex couples in Hong Kong. The present appeals concern similar issues in the respective contexts of public rental housing and subsidised home ownership flats. I shall return to these principles when addressing the various issues raised in these appeals.

Impugned policies

21.  The HA policies challenged in these appeals have been outlined above.[61]

HA’s central argument on non-engagement of right to equality: BL36

22.  Under the common law, equality before the law is fundamental to the rule of law.[62] At the constitutional level, equality protection and non-discrimination are guaranteed under Article 25 of the Basic Law (“BL25”) [63] and Article 22 of the Hong Kong Bill of Rights (“BOR22”),[64] which is constitutionally entrenched under Article 39(1) of the Basic Law. [65] The first issue raised in these appeals, and indeed the main thrust of the HA’s argument,[66] is that these equality provisions are not engaged insofar as the challenged HA policies are concerned, because the matter is specifically governed by Article 36 of the Basic Law (“BL36”).[67]

23.  In essence, the HA contends that BL36, as interpreted by this court in Kong Yunming v Director of Social Welfare,[68] entrenches all pre-1997 social welfare benefits and entitlements as rights protected under the Basic Law.[69]  The government may develop and improve the pre-existing social welfare system in the light of prevailing economic conditions and social needs, as provided by Article 145 of the Basic Law (“BL145”).[70]  Any such modification is subject to constitutional review by the court based on a proportionality analysis.[71]

24.  The HA goes on to argue that BL36 is the lex specialis concerning social welfare rights and policies,[72] in the same way that Article 40 of the Basic Law (“BL40”)[73] specifically and exclusively governs the preservation and protection of the lawful traditional rights and interests of indigenous inhabitants of the New Territories.[74]  When the Basic Law and the Hong Kong Bill of Rights are construed as a coherent whole, it becomes apparent that, much like the rights and interests protected under BL40 are not subject to the general equality provisions, the entrenched social welfare rights under BL36 are likewise excluded from such review.[75]

25.  The HA argues that opposite-sex married couples have always enjoyed exclusive rights to apply for PRH units and HOS flats under the pre-1997 social welfare system.[76] Their exclusive entitlements are protected by BL36 and cannot be diluted or otherwise negatively impacted by extending the same privileges to same-sex couples married overseas.[77]  The relevant HA policies, which reflect and uphold the constitutionally entrenched social welfare entitlements of opposite-sex married couples, are therefore immune from the scrutiny of the general equality provisions.[78]  Accordingly, the equality provisions are not engaged, and the constitutional challenges do not even get to first base.

26.  The Court of Appeal refused to allow the HA to run its argument on BL36 for lateness and because it did not raise a pure question of law.[79] Nonetheless, it considered the merits of the argument in full and rejected it on the basis that under the pre-1997 social welfare system, opposite-sex married couples only had rights to apply for PRH units and HOS flats.[80]  Their rights did not extend to a guarantee on the AWT.[81]  The addition of otherwise eligible same-sex married couples to the same queue of opposite-sex married couples would at most lengthen the queue and thus the AWT, but would not affect the latter’s rights to apply for PRH units or HOS flats.[82]  BL36 was simply not engaged.[83]  It was therefore unnecessary to decide whether BL36 is the lex specialis concerning social welfare rights and excludes the operation of the general equality provisions.[84]

Is BL36 engaged (Question 1)?[85]

27.  I see good reasons to entertain the argument based on BL36. The Court of Appeal has dealt with it on the merits. Two of the questions on which leave to appeal was granted are based on this argument.[86] The parties have advanced full submissions on them in their written cases and at the hearing.[87]  Furthermore, a proper understanding of BL36 and its relationship to other relevant provisions of the Basic Law and the Hong Kong Bill of Rights is a matter of importance that this court should address.

28.  I would approach the BL36 argument on two levels. First, on the facts. Admittedly, allowing same-sex married couples to apply does not, as such, deprive opposite-sex couples of their rights to apply. However, the HA’s real point is that this would lengthen the queue and the AWT, and therefore dilute, in substance although not in form, opposite-sex couples’ pre-existing rights to apply.[88] This argument might carry weight if a separate queue for opposite-sex married couples had been reserved, such that the inclusion of a substantial number of same-sex married couples in the same queue would significantly lengthen the AWT for opposite-sex couples, thereby rendering the claim of a dilution of their pre-existing rights more tenable.

29.  However, as described, for PRH units, opposite-sex married couples apply under the “Ordinary Families” sub-category, which also includes parent and child, grandparent and grandchild, and single siblings.[89]  Together, they form one queue. There is no separate queue for opposite-sex married couples as such. The entitlement of opposite-sex married couples to apply under the “Ordinary Families” sub-category can hardly be said to be “exclusive”. As its name suggests, the sub-category under which opposite-sex married couples apply for PRH units is one for “Ordinary Families”. An opposite-sex married relationship is just one of the several familial relationships included within this sub-category.[90]

30.  Nothing in the evidence suggests that the “Ordinary Families” sub-category is a closed sub-category such that no other familial relationships can be recognised for inclusion into this sub-category. The mere fact that pre-1997, same-sex married couples were not included cannot be equated with an “exclusive” entitlement of opposite-sex married couples to be included.

31.  The truth is that different familial relationships have always been included in the same “Ordinary Families” sub-category and they all join the same queue. None of them can claim to have, or to have had as at 1 July 1997, any “exclusive” entitlement to being in the queue, to the exclusion of other familial relationships which the HA may from time to time decide to recognise and include. There is nothing in the documentary evidence produced by the HA concerning the PRH policy and its implementation to suggest that no new familial relationship can be added to the “Ordinary Families” sub-category. There is equally no suggestion that the HA cannot or will not revise and enlarge the “Ordinary Families” sub-category, or even create new sub-categories that enjoy priority over the “Ordinary Families” sub-category.

32.  The same may be said in relation to the HA’s overall HOS policy (of which the policies on addition of occupants and transfer of ownership form part).[91]  Households eligible to apply for purchase of HOS flats are not limited to those of opposite-sex married couples. Rather, an eligible household may include the applicant’s children, parents, grandparents, in-laws, siblings, as well as other relatives if dependent on the applicant. Again, there is nothing “exclusive” about opposite-sex married couples’ entitlement to apply as couples to purchase HOS flats.

33.  The HA responds that one should relevantly focus on married couples, and amongst them, only opposite-sex married couples are eligible.[92] However, this argument still cannot overcome the objection that there has never existed a distinct, exclusive spousal sub-category or queue catering for opposite-sex married couples only, and their rights to apply have always been non-exclusive. This being the case, there can be no complaint of a dilution of their entitlements by the admission of other relationships to the same queue – be it same-sex marital relationships or other familial relationships which the HA may choose to recognise.

34.  The Court of Appeal was therefore right to conclude that permitting same-sex married couples to apply would not affect opposite-sex couples’ protected right to apply. [93] There being no constitutional guarantee on the AWT, BL36 would not be engaged at all.

Does BL36 override the equality provisions (Question 2)?[94]

35.  Secondly, and indeed more fundamentally, I do not accept the premise of the argument based on BL36. The HA’s contention is founded on the undoubted principle of constitutional interpretation that the Basic Law should be read as a coherent whole.[95]  However, it does not follow that this can only be achieved, as the HA suggests as the main plank of its argument, by reading BL36 as overriding the equality provisions in BL25 and BOR22. In other words, there is no warrant to read a Basic Law guarantee on pre-existing social welfare rights as trumping the constitutional requirement of equality before the law under BL25 and of non-discrimination under BOR22. Rather, one should proceed on the footing that, as a core tenet of the rule of law, equality can only be displaced by compelling justifications and unequivocal language.[96]

36.  Social welfare entitlements and benefits, covered by BL36, are almost by definition concerned with the allocation of limited societal resources. Such allocation must involve differentiating amongst members of society by classes and demographics, whereby some are favoured or privileged over others in view of their needs and circumstances, connections to Hong Kong, past contributions to society and so forth. There is, however, no reason why such differentiation should be made on grounds other than ones that are fair, rational, and otherwise justifiable. Still less would one expect any distinctions to be drawn on inherently suspect grounds[97] or grounds that cannot be justified.

37.  In fact, precisely because social welfare benefits inevitably involve favouring some categories or strata of the populace over others, the need for drawing demarcation lines on a fair and defensible bases is all the stronger. Admittedly, those responsible for making social welfare policies often have to make hard choices, given limited resources and competing demands. However, this pertains only to the margin of discretion to which policy-makers are entitled when their decisions are subject to judicial scrutiny,[98] but it in no way lessens the necessity of allocating social welfare benefits on a rational and justifiable basis, free from discrimination. Thus analysed, there is no necessary conflict between the protection of equality under BL25 (and BOR22) and the preservation of pre-existing social welfare rights under BL36.

38.  Moreover, BL145 expressly envisages that the pre-existing social welfare system may be improved and developed in the light of the prevailing economic conditions and social needs of society. There is no basis for saying that a pre-1997 social welfare entitlement is constitutionally entrenched and free from equality scrutiny even if it is subsequently found to be wanting in justification. One can discern no such drafting intention of the Basic Law. If anything, BL145 tends to suggest that the pre-existing social welfare system may be modified to remove any differential treatment that cannot or can no longer be justified in the light of prevailing social values and conditions.

39.  BL40, concerning the preservation and protection of the traditional rights and interests of the indigenous inhabitants of the New Territories, is an entirely different story. As explained in this court’s judgment in Kwok Cheuk Kin, right from the beginning, the drafters of the Basic Law were fully aware of the “inherently discriminatory” nature of the “Ding rights”[99] of male indigenous inhabitants of the New Territories.[100]  After debates and discussions, BL40 was included in the final draft of the Basic Law despite those concerns. As a result, as held by this court, the “Ding rights” are preserved and protected post-1997 despite their known inherently discriminatory nature.[101]  The drafting material and background of the Basic Law clearly demonstrated that its drafters exceptionally intended to accord favourable differential treatment to the male indigenous inhabitants of the New Territories.[102]  Thus, this court concluded in Kwok Cheuk Kin that BL40 constitutes a lex specialis and overrides the general equality provisions under BL25 and BOR22.[103]

40.  BL36, in contrast, is a far cry from that situation. Whereas BL40 has always been known to be inherently discriminatory with respect to the “Ding rights” preserved thereunder, this is not the case with BL36. Nothing in the drafting history of BL36, so far as is made known to the court, suggests that there was any drafting intention to entrench any social welfare entitlements that were known or suspected to be discriminatory in nature. In particular, it should be noted that back when the Basic Law was drafted in the 1980s or promulgated in 1990, the question of same-sex individuals getting married overseas or their applying for PRH units or HOS flats as couples was simply non-existent. There was then no question of thinking that the now challenged policies might be discriminatory at all. It is true that with the advance of human rights law and the availability of same-sex marriages in many overseas jurisdictions in recent years, doubts have now arisen regarding whether the subject policies discriminate against same-sex married couples. However, this fact provides no support for suggesting that the drafting intention of BL36 was to preserve social welfare entitlements even if they were or were subsequently found to be discriminatory, despite the equality provisions in BL25 and BOR22.

41.  Rather, BL25, BL36, and BL145 (as well as BOR22) can and should be read consistently to mean that pre-1997 social welfare benefits and entitlements are entrenched rights that are subject to improvement and development pursuant to BL145, as well as judicial scrutiny under BL25 and BOR22. BL36 does not insulate pre-existing social welfare rights from the oversight of the equality provisions.

Nor does BL37

42.  To a lesser extent, BL37, guaranteeing the right to heterosexual marriage in Hong Kong, is similarly relied upon by the HA to contend that the impugned policies are immune from the equality provisions because they give effect to BL37.[104]

43.  This argument can be disposed of shortly. BL37, a lex specialis in relation to the right to marriage, is a constitutional guarantee to opposite-sex couples, and these couples only, to the institution of marriage, and thus the status of marriage.[105]  It is not concerned with legal rights and obligations that are usually based on or associated with the status of marriage as such.[106]  Entitlements to apply for PRH units and HOS flats are only some of those rights, but they do not go to the status of marriage itself. BL37 therefore does not remove the subject HA policies from the purview of the equality provisions.

Comparability of same-sex and opposite-sex married couples (Question 3)[107]

44.  Accordingly, I turn to the next issue that divides the parties, namely whether same-sex couples who are married overseas are comparable to heterosexual married couples.[108]  The equality provisions are only engaged if they are so comparable in the context of applications for PRH units and HOS flats.

45.  Comparability is context-dependent.[109] Here, one is concerned with the allocation of finite social resources in the forms of public housing and subsidised home ownership. The HA argues that opposite-sex married couples are not true comparators of same-sex ones in the present context because only the former as a category have “reproductive capacities and procreative potential”, and this “general biological disparity is a clear objective difference which is highly relevant if not decisive”. [110]  The HA submits that its policies seek to support broad governmental objectives, and in particular that of promoting population growth through natural procreation.[111]  For this reason, procreative potential constitutes a material distinction. The HA also argues that as a policy-maker, it enjoys a margin of discretion in deciding who is and who is not a true comparator.[112]

46.  The Court of Appeal rejected these arguments.[113] So do I. The case law emanating from this court has established that in various contexts, same-sex couples who are lawfully married overseas are in an analogous and comparable position to opposite-sex married couples.[114]  None of the cases involves recognising the legal status of a foreign same-sex marriage as such.[115]  Rather, the emphasis on a foreign marriage that is lawful and valid according to the law of the place where the marriage was celebrated addresses the need to demonstrate that the relationship between the same-sex couple is characterised by publicity and exclusivity – essential characteristics that distinguish a heterosexual marriage.[116]  The HA argues that unlike the previous cases, public housing and subsidised home ownership constitute a different context in which opposite-sex married couples and same-sex ones are not truly comparable.

47.  I confess to having some reservation with the HA’s assertion that its policies are aimed to support the government’s population growth policy by encouraging procreation, given that the HA’s primary, if not overriding objective is to meet the pressing housing needs of the lower income strata of society.[117]

48.  In any event, the mere fact that the impugned policies also seek to support the government’s strategic objective to stimulate population growth through increasing housing availability to opposite-sex married couples does not impact on the question of comparability. Rather, it is only a matter pertaining to the issue of justification/proportionality. This is plain from the fact that the HA’s own policies on eligibility to apply do not differentiate amongst heterosexual married couples in terms of whether they have children[118] or are planning to have children, whether they are past child-bearing age, whether they are unable to have children for medical or other reasons, and so forth. The evidence does not reveal how many PRH tenants or HOS households have children or are in fact childless. At the very least, same-sex married couples are comparable to heterosexual married couples who do not have or plan to have children.

49.  Nor does the HA’s argument take into account the fact that same-sex couples may be able to adopt children or have children by artificial means.

50.  The HA’s suggested distinction based on procreative potential becomes more contrived when it is recalled that for PRH applications, opposite-sex married couples are included in the same sub-category of “Ordinary Families” alongside parents and children, grandparents and children, and single siblings. Procreative potential is relevant to none of these other eligible familial relationships. Yet all of them join the same queue as opposite-sex married couples. The only common thread to these different relationships is a close familial relationship that is in need of housing assistance. A same-sex married relationship with housing needs fits easily into that description.

51.  A similar observation may be made in relation to the HOS.

52.  In any event, as this court has observed, the notion whether the comparators are analogous or relevantly similar is elastic both linguistically and conceptually.[119] This is reflected in the court’s approach. As Lord Nicholls explained in R (Carson) v Secretary of State for Work and Pensions:[120]

“… the essential question for the court is whether the alleged discrimination, that is, the difference in treatment of which complaint is made, can withstand scrutiny. Sometimes the answer to this question will be plain. There may be such an obvious, relevant difference between the claimant and those with whom he seeks to compare himself that their situations cannot be regarded as analogous. Sometimes, where the position is not so clear, a different approach is called for. Then the court’s scrutiny may best be directed at considering whether the differentiation has a legitimate aim and whether the means chosen to achieve the aim is appropriate and not disproportionate in its adverse impact.”

53.  I do not accept the HA’s argument that when it comes to identifying the true comparator, the HA as the policy-maker is entitled to a margin of discretion. I have no difficulty in accepting that in appropriate circumstances, a margin of discretion should be accorded to the rule/policy-maker. This is particularly so when it comes to Stage 3 of the proportionality analysis.[121]  However, where, as here, the question is to identify the true comparator in a discrimination case, there is no room for according any margin of discretion to the rule/policy-maker. Comparability is partly a matter of values and partly a question of rationality.[122] It is a legal question for the court to decide. The HA has not been able to cite any relevant authority in support of its contention.[123] 

Justification under the proportionality analysis

54.  Once the issue of the true comparator is resolved in favour of same-sex couples who are lawfully married overseas, the differences in treatment that they experience under the impugned HA policies cannot be disputed, and are not disputed by the HA. The question thus becomes one of justification, or whether the policies can satisfy the proportionality analysis.

Stage 1: Family Aim as legitimate aim

55.  At Stage 1 of the proportionality analysis, the HA relies on what it says to be the “chief aim … to support traditional families founded on opposite-sex marriages” (“Family Aim”) as the legitimate aim underpinning the challenged PRH and HOS policies.[124]

56.  The Family Aim is said to have three connected aspects: (1) to support existing traditional families constituted by opposite-sex married couples (in and of themselves); (2) to support existing traditional families constituted by opposite-sex married couples along with their existing children; and (3) to support the institution of traditional family for the benefit of (a) opposite-sex unmarried couples whose marriage plans may be influenced by housing availability, and (b) opposite-sex married couples whose procreative plans may similarly be influenced. [125] 

57.  There is no dispute that the Family Aim is a legitimate aim as such,[126] although I repeat my earlier reservation regarding the HA’s primary objective to meet the housing needs of the underprivileged class.[127]

Stage 2: Whether the PRH and HOS policies are rationally connected to the Family Aim (Question 4)[128]

58.  As for the rational connection between the Family Aim and the challenged policies (i.e. Stage 2 of the proportionality analysis), the courts below have accepted that the PRH policy is rationally connected to the Family Aim,[129] whereas the impugned HOS policies are not.[130] Both conclusions are respectively disputed before us.[131]

59.  As regards the PRH policy and its rational connection to the Family Aim, the main argument raised for Mr Infinger is one on the lack of evidence.[132]  Apart from its repeated claim of a zero-sum situation,[133] it is argued, the HA has not adduced any evidence to demonstrate how its exclusionary policies would improve the actual allocation of units in the real world and thus support the Family Aim.[134]

60.  The HA essentially counters that whether there is a rational connection is a question that can be decided by simple logic and common sense.[135]  Given that the supply of PRH units is exceedingly limited, it must follow that the Family Aim is furthered by only allowing opposite-sex couples to apply. Evidence is not required to prove this self-evident proposition.[136]

61.  The question of rational connection is context-dependent.[137] Whilst there has to be a causal relation between the legitimate aim and the impugned measure in the sense that the implementation of the measure “can reasonably be expected to contribute towards the achievement of the [aim]”,[138] the threshold for establishing a rational connection, as compared to the standard of review often employed under Stage 3, is relatively low. Depending on the facts, it may be satisfied merely based on “reason or logic”, or “common sense”, without further proof.[139]  Moreover, the court has to “allow room for the exercise of judgment” by the executive or the legislature where the challenged measure is based on an evaluation of complex facts, or considerations which are contestable or may be controversial.[140]  Considerations of economic or social policy, or national security, have been said to fall within this latter category.[141] 

62.  In the present case, while I do not accept the HA’s zero-sum submission as an accurate description of the true position,[142] I am prepared to accept that there is a rational connection between promoting the Family Aim and favouring opposite-sex couples in the application for PRH units. Whether this differential treatment can be justified under Stages 3 and 4 of the proportionality analysis is an entirely different matter.

63.  As for the rational connection between the Family Aim and the HOS policies on the addition of occupants and transfer of ownership, it should be noted that since the original applicant (Mr Ng) in the HOS proceedings had already purchased an HOS flat in his own name, his and Mr Li’s challenge does not directly involve the HOS policy to exclude same-sex married couples from purchasing HOS flats.[143]

64.  The HA argues that the challenged HOS policies “could deter same-sex couples from initially purchasing HOS units and impact upon the number of available HOS units in the market” for opposite-sex couples, thereby advancing the Family Aim.[144]  The Court of Appeal as well as Chow JA have both rejected this argument on the basis that, in the absence of empirical evidence, the rational connection between the challenged HOS policies and the promotion of the Family Aim is tenuous to the extreme.[145]  The suggested impact is simply “de minimis”.[146]

65.  The HA again argues before us that in constitutional review, establishing a rational connection “requires nothing more than showing that the legitimate goal … [is] logically furthered by the means government has chosen to adopt”, as where a measure can “reasonably be expected to contribute towards the achievement of that objective”, without “insisting on [evidential] proof of a relationship between the infringing measure and the legislative objective”.[147]  The HA contends that “as a matter of logic and common sense”, the HOS policies increase the supply of HOS flats to opposite-sex couples and thus support the Family Aim.[148]

66.  The impugned policies on addition of occupants and transfer of ownership are, in fact, integral components of a broader, overarching HOS policy under which same-sex couples are disentitled from purchasing/owning or occupying HOS flats as couples.[149] Accordingly, to evaluate the rational connection of the challenged policies to the promotion of the Family Aim in isolation would be to adopt a myopic and unduly constrained perspective.

67.  I am prepared to accept that, amongst other things, the HA’s overall HOS policy is aimed at supporting the Family Aim and is rationally connected to it. And once the challenged policies on addition of occupants and transfer of ownership are viewed, as they should be, as part and parcel of the overall policy, and the entirety of the framework is considered holistically, it becomes plain that there exists a rational connection between the impugned policies and the Family Aim. Whether they can withstand scrutiny under Stages 3 and 4 of the proportionality analysis is, again, quite a different matter.

Stages 3 and 4: Whether the PRH and HOS policies are disproportionate and/or fail to strike a reasonable balance (Question 6)[150]

68.  Stages 3 and 4 of the proportionality analysis can be taken together. On the question of the intensity of the review, since the differential treatment is based on a suspect ground (i.e. sexual orientation) prima facie this calls for an intensive scrutiny of the policies challenged, nearing the top end of the continuous spectrum of the standard of review.[151]  However, sufficient regard must be had to the fact that the impugned measures concern the allocation of limited societal resources in the form of housing and the pursuit of governmental social policies on supporting traditional families and population growth.[152] These are matters in relation to which the HA should, subject to my observations below, enjoy an appropriate margin of discretion.[153]  Bearing these considerations in mind, I approach Stages 3 and 4 of the proportionality analysis.

69.  The Court of Appeal, in affirming the decisions below, came to the firm conclusion that the disputed policies were disproportionate.[154]  This was especially so when there was no empirical evidence from the HA to quantify the supposed impact on opposite-sex individuals regarding their plans or inclinations to get married and/or on married couples to have children, if the HA’s policies were to be relaxed to admit the applications of same-sex married couples.[155]

70.  The HA strongly criticises the Court of Appeal for its insistence on receiving empirical evidence.[156]  It contends that the court has failed to give proper weight to the fact that whenever a PRH unit or an HOS flat becomes available and is allocated to a same-sex married couple, at least one eligible traditional family will be unable to enjoy that unit for the duration of its occupancy.[157]  The HA also relies on the impact of any relaxation of the challenged policies on opposite-sex couples’ rights under BL36.[158]  Finally, in relation to the impugned HOS policies, the Court of Appeal, it is contended, has failed to take into account considerations of overall coherence in that those policies are integral parts of the overall HOS policy under which same-sex couples are excluded from purchasing, owning or occupying HOS flats as couples.[159]

71.  I am not persuaded by these arguments. I have no difficulty in accepting that the proportionality test does not always require positive justificatory evidence in every case, and the court must bear in mind the “realities” of the case in question.[160]  Whether, and the extent to which, evidence is required in a given case to support a justification must depend on the facts and the context, and especially, on the nature of the justification itself.[161]

72.  In particular cases, reliance on common sense, logic or intuition may be appropriate,[162] and indeed cases exist where these are the only matters that may properly be relied on by the court for the evaluation of a justification. For there are “predictive and other judgmental assessments” of a kind “whose rationality can be assessed but whose correctness cannot in the nature of things be tested empirically”.[163]  Justifications grounded in moral or political considerations may not easily lend themselves to evidentiary support.[164]  Where firm factual conclusions are “elusive”, where evidence is “inconclusive or slight”, or where the truth is “inherently unknowable”, the court may be more prepared to accept an instinctive judgment by the executive or legislature about what the relevant facts are likely to be, without insisting on definite, concrete proof when examining a justification. [165]  By contrast, justifications of an economic or social nature would typically require evidential substantiation. As explained in R (Lumsdon):[166]

“The justification for the restriction tends to be examined in detail, although much may depend on the nature of the justification, and the extent to which it requires evidence to support it. For example, justifications based on moral or political considerations may not be capable of being established by evidence. The same may be true of justifications based on intuitive common sense. An economic or social justification, on the other hand, may well be expected to be supported by evidence.”

73.  The justification put forward for the impugned measures in these appeals is the Family Aim. The exclusionary policies are said to benefit opposite-sex married couples by increasing the supply of PRH units and HOS flats to them, thereby supporting existing traditional families (both with and without children) as well as the institution of traditional family by encouraging heterosexual individuals to get married and/or married couples to have children. Critical to this justification are therefore matters such as the estimated increase in supply of PRH units and HOS flats by pursuing the exclusionary polices and the likely impact on opposite-sex couples (married and unmarried) if the policies are relaxed in favour of same-sex married couples. These are matters on which empirical evidence is naturally expected.

74.  Yet the HA has not adduced any evidence at all concerning the likely effect on supply and its potential impact on opposite-sex couples if the challenged policies are relaxed. Here, the point taken against the HA is not one about the quality of empirical evidence adduced in support of the justification. Neither does it concern the appropriate weight the court should accord to the HA’s assessment if such evidence were forthcoming – especially since such material would likely involve statistics, estimates, projections and predictions, which inherently carry a certain degree of subjectivity and predictive judgment. The criticism of the HA’s case is a much more fundamental one – there is no evidence at all to back the HA’s assertions.

75.  This is not a case where it is suggested that it is difficult or impossible to carry out any meaningful study on the subject matter or to collect information and figures.[167]  The court has not been told why no such study was conducted or figures collected, or how unhelpful they would be. What has been repeatedly urged upon the courts, including this court, is that one PRH unit or HOS flat made available to a same-sex couple would mean one less unit or flat available to an opposite-sex couple. I do not think this is sufficient to support the justification claimed.

76.  When a measure as absolute as the total exclusion of same-sex married couples is invoked for the claimed objective of supporting the Family Aim, it is only natural to ask how that complete restriction is said to be actually contributing to the objective. As explained, the answer to this question must, to a significant extent, depend on the estimated number of same-sex couples that are potentially involved, and therefore how many PRH units or HOS flats may have to be made available to these couples, if the challenged policies are to be changed, at the expense of opposite-sex couples. The possible impact on opposite-sex couples must differ quite substantially, depending on whether one is potentially concerned with a few hundred same-sex married couples only or with, say, tens of thousands of same-sex married couples.

77.  But not only that. The position is further complicated by the fact that there are other relationships that also join the same queue as opposite-sex married couples. There is no evidence before the court on what the proportions of the different types of applicants under the “Ordinary Families” sub-category are, and therefore how the allocation of PRH units to same-sex married couples may affect these different types of applicants in terms of their waiting times. Presumably, the impact on the availability of units would be borne and shared amongst them, and thus the effect on opposite-sex couples’ AWT, as but one group amongst many, would be lessened. The same observation may be made in relation to HOS flats.

78.  The courts below did not have the benefit of any relevant empirical evidence or study.[168]  Neither do we. It is pertinent to point out that if the HA enjoys an institutional advantage over the courts on the issue under discussion, and for that reason should normally be accorded an appropriate margin of discretion, it has not made use of that advantage. One is left no wiser as to why the absolute policies under challenge are required to promote the Family Aim. Indeed, it is difficult to see why a substantial margin of discretion should be accorded to the HA’s assessment when all it relies on is “logic and common sense” to maintain that its policies are reasonably necessary to support the Family Aim. As has been said in Re Brewster:[169]

“... where the question of the impact of a particular measure on social and economic matters had not been addressed by the government department responsible for a particular policy choice, the imperative for reticence on the part of a court tasked with the duty of reviewing the decision is diminished.”

79.  I do not find the argument relating to where the burden of proof to adduce empirical evidence lies helpful. The issue here is not so much which party bears the burden of proof as whether the court can be satisfied that the impugned measures are not disproportionate. When the HA seeks to justify the challenged policies by asserting that they promote the Family Aim, the court is entitled to look at the real, rather than theoretical, impact that relaxing them may potentially have on the Family Aim, regardless of where that evidence comes from. Absent such evidence, the court is simply left bereft of any reliable basis to be satisfied that the impugned policies represent proportionate measures in support of the Family Aim.

80.  The HA’s case is further exacerbated by the total absence of any explanation as to why a less restrictive measure, such as prioritising opposite-sex married couples (or those with small children) while still allowing same-sex married couples to apply, could not be adopted to support the Family Aim.[170]

81.  Similar observations on the lack of empirical evidence can be made when it comes to Stage 4 of the proportionality analysis, that is, striking a reasonable balance between the societal benefits of the differential policies and the inroads made into the equality rights of the individuals affected.[171]

82.  Significantly, under the challenged policies, same-sex married couples cannot apply for PRH units together, but must apply separately as individuals where the AWT is longer. If one spouse is eventually successful in applying as an individual, the other cannot be added as an occupant as a same-sex spouse. A similar situation exists in relation to the purchase and occupation of HOS flats. In short, a same-sex married couple cannot share family life by residing in a PRH unit or an HOS flat.

83.  In the case of the impugned PRH policy, when it is remembered that by definition, one is concerned with needy same-sex married couples who cannot afford private rental accommodation, the HA’s exclusionary policy could well mean depriving them of any realistic opportunity of sharing family life under the same roof at all.[172]      

84.  As to opposite-sex married couples’ rights enjoyed under BL36, I have already explained that their entitlements to apply under the PRH and HOS policies are not exclusive, and their BL36 rights are therefore not engaged.[173]

85.  Turning to the coherence considerations relating to the challenged HOS policies,[174] as explained, those policies simply form parts of the broader HOS policy framework, which excludes same-sex married couples from purchasing, owning or occupying HOS flats as couples. The HA argues that they cannot be disturbed without compromising the integrity and coherence of the overall policy framework – considerations that, it says, should attract significant if not decisive weight.[175] 

86.  The underlying premise of these coherence considerations is that the HOS policy on purchase can be justified. It is, of course, true that this other policy is not challenged in the HOS proceedings.[176] However, one should not assume that if challenged, it would necessarily withstand scrutiny.[177]

87.  In any event, if this coherence argument is taken to its logical conclusion, then unless and until an applicant has the standing and practical reason to challenge each and every discriminatory component of a policy framework, no individual component can be separately challenged, no matter how seriously the applicant may actually be affected by it. That cannot be right. On the facts of the present case, this administrative coherence argument carries little weight in the third and fourth stages of the proportionality analysis. A change to the challenged policies would merely leave an unchallenged policy on purchase in place. Indeed, without wishing to express any definite view on the matter, that policy would obviously be treated as vulnerable and of doubtful validity given the outcome of the HOS appeal. This simply shows the superficiality of the alleged “incoherence”.

88.  On the evidence before the court, the HA clearly fails in Stages 3 and 4 of the proportionality analysis. In other words, the challenged policies cannot be justified. There being no suggestion of any remedial orders to preserve them, the policies have been correctly struck down by the courts below.

Disposition

89.  For these reasons, I would dismiss the two appeals. I would make an order nisi that the HA pay Mr Infinger and Mr Li their respective costs of these appeals, with a certificate for three counsel. I would also direct that any application to vary the costs order nisi under each appeal be made within 14 days of the handing down of this judgment, and dealt with on paper.

Mr Justice Ribeiro PJ:

90.  I agree with the judgment of the Chief Justice.

Mr Justice Fok PJ:

91.  I agree with the judgment of the Chief Justice.

Mr Justice Lam PJ:

92.  I agree with the judgment of the Chief Justice.

Mr Justice Stock NPJ:

93.  I agree with the judgment of the Chief Justice.

Chief Justice Cheung:

94.  Accordingly, the court unanimously dismisses both appeals and makes the orders and directions indicated in paragraph 89 above.

APPENDIX

Questions of great, general or public importance

(1)  Does BL36 confer on opposite-sex married couples:

(a)   a constitutional right, as defined by the eligibility rules in existence as at 1 July 1997, to exclusively apply for PRH units as spouses under the “Ordinary Families” category;

(b)   a constitutional right, as defined by the eligibility rules in existence as at 1 July 1997, to exclusively apply to purchase HOS units under the spousal category.

(2)  Is the applicants’ reliance on BL25 and BOR22 precluded by a coherent and holistic interpretation of the Basic Law in line with BL36, BL37 and BOR19?

(3)  Are same-sex couples and opposite-sex couples proper comparators in relation to the differential treatment conferred by the PRH and HOS Spousal Policies under challenge, in circumstances where:

(a)   It is the HA’s policy decision to align with and support the broader and unchallenged governmental objective of promoting population growth by increasing hosing availability.

(b)   Opposite-sex couples as a general category (and notwithstanding exceptions in individual cases) possess inherent procreative capacities not possessed by same‑sex couples.

(4)   Where it has already been found that there is a logical connection between the Family Aim and the HOS Spousal Policy, is it open to the court to nonetheless find that the connection is “de minimis”, especially where there is no affirmative evidence to that effect?

(5)   Where the HOS Spousal Purchase Policy has not been challenged or found unlawful, is the administrative coherence between the HOS Spousal Policy and the HOS Spousal Purchase Policy a relevant factor in assessing the proportionality of the HOS Spousal Policy?

(6)   In assessing the proportionality of the PRH and HOS Spousal Policies:

(a)   Should the court take into account the BL36 right to social welfare enjoyed by opposite-sex couples as set out in paragraph (1)(a) above, as well as more generally opposite-sex couples’ interest in being able to hitherto apply exclusively for PRH and/or HOS units under the spousal category?

(b)   Is empirical or statistical evidence necessary when considering whether the PRH Spousal Policy would increase the number of PRH units available to opposite‑sex couples, in circumstances where (i) it is undeniable that whenever a PRH unit becomes available and is allocated to a same-sex couple, at least one eligible traditional family will be unable to enjoy that unit for the duration of its occupancy; and (ii) the PRH Spousal Policy serves as a long term housing policy.

(c)  Is empirical or statistical evidence necessary when considering whether the HOS Spousal Policy would increase the number of HOS units available to opposite‑sex couples, in circumstances where (i) some same-sex couples are likely to be deterred by the HOS Spousal Policy to apply to purchase a HOS flat, and (ii) whenever a HOS unit becomes available and is allocated to a same-sex couple, at least one eligible traditional family will be unable to enjoy that unit for the duration of its occupancy.