Rainbow in the Dark
Judicial Strategies to Defend Same-Sex Partnership Recognition in Authoritarian Hong Kong
On 16 July, the Hong Kong government introduced the Registration of Same-sex Partnerships Bill (“Bill”) in the Hong Kong Legislative Council (“LegCo”). This move was mandated by two decisions of Hong Kong’s apex court – Court of Final Appeal (“the Court”) – in a 2023 case (collectively, “STK Decisions”). The Bill only grants same-sex couples who have already registered overseas the right to have their relationships legally recognised through government registration and to deal with each other’s medical and after-death matters. Albeit socially conservative, it was opposed by nearly half of LegCo. Some LegCo Members even suggested responding to the Court by curbing its constitutional review jurisdiction over “important social value issues” and by amending the Hong Kong Bill of Rights (“Bill of Rights”), the basis of the STK Decisions, to effectively declare them void for misinterpreting the Bill of Rights. On 10 September, LegCo scheduled to vote on the Bill.
The STK Decisions came as a beam of light at the grim time of Hong Kong’s authoritarian turn. Supporting their enforcement, I argue that, first, doctrinally, both curbing courts’ constitutional review jurisdiction and narrowing the Bill of Rights are unconstitutional propositions. Second, strategically, if the Bill is rejected, despite political pressure, the Court should grant “declaration of partnership at common law” and lay down statutory interpretive principles favouring same-sex partners. My analysis demonstrates judicial strategies for navigating a liberal enclave within the authoritarian regime1) and the correlation between “gender backlash and constitutional degradation”.
The STK decisions
In the STK Decisions, the Court held that the right to private life – protected under art. 14 Bill of Rights – requires the government to introduce a form of legal recognition of same-sex couples. It may or may not be same-sex marriage (para. 115). As to what rights and obligations this recognition entails, the Court largely left it to the government and legislature. Nevertheless, borrowing from ECtHR jurisprudence, it specified that “core rights” are necessary, while “supplementary rights” fall within a wider margin of discretion (paras. 179-188). It ordered the government to legislate accordingly before 27 October 2025, with possible extension given compelling reasons.
Basic Law-Bill of Rights supremacy
The Bill of Rights has constitutional effect based on the Hong Kong Basic Law (“Basic Law”). Conferring Hong Kong high autonomy within China, no restriction to freedom and rights in Hong Kong may contravene the Basic Law (arts. 11, 39(2)), including its art. 39(1) which provides that Hong Kong-applicable ICCPR provisions “shall be implemented” through Hong Kong laws. That implementation is the Hong Kong Bill of Rights Ordinance, containing the Bill of Rights. It is thus entrenched (para. 15) beyond other laws.
The association of the three instruments means that limiting either courts’ constitutional review jurisdiction or the Bill of Rights is unconstitutional.
First, courts’ constitutional review jurisdiction is constitutionally obligatory (para. 61), because only they could effectuate the Basic Law-Bill of Rights supremacy. Under the Basic Law, where citizens allege Basic Law contraventions, only the judiciary has the jurisdiction2) (arts. 19, 80) and therefore the duty ([106(c)]) to adjudicate and the power to interpret the Basic Law (art. 158) to verify the allegations. Thus, restricting or transferring courts’ constitutional review jurisdiction violates art. 19 Basic Law for usurping judicial power (paras. 49-60)3) and art. 11 for undermining its necessary component – timely detection and disapplication of Basic Law infringements.4)
Second, “shall be” in art. 39(1) means local implementation of applicable ICCPR provisions is “necessary” (para. 32). Restricting its “embodiment” (para. 15) in the Bill of Rights is therefore impermissible.5) The same follows from the Chinese text: the mandatory-toned adverbial “通過香港特別行政區的法律予以實施” (literally: “implemented through Hong Kong laws”) applies to “[ICCPR provisions] as applied to Hong Kong” without qualification.
Moreover, the Bill of Rights is not just an ordinary law, but part of the Basic Law which LegCo cannot amend (art. 159). Indeed, the Bill of Rights is “expressly incorporated by [art 39]” (para. 62) and rights in both the Bill of Rights and the Basic Law itself are “fundamental rights […] constitutionally guaranteed” (para. 14), suggesting equal footing.
The STK Decisions and courts in authoritarian Hong Kong
The STK Decisions are not just a victory for LGBTQ+ rights. They also boosted judicial independence amid Hong Kong’s authoritarian turn. This context informs the judicial strategies that follow.
Despite the Basic Law’s constraints on central power, Beijing “self-aggrandised” (p. 55) to impose on Hong Kong a National Security Law in 2020 and a LegCo election overhaul in 2021 – and Hong Kong courts could invalidate neither (para. 37). The reforms seriously affected judicial independence and the political-constitutional constraints on penalising activist courts. First, they demonstrated the authoritarian sovereignty’s absolute power.6) Second, a Beijing-supervised, Hong-Kong-government-staffed National Security Committee now effectively (paras. 39-44) has higher jurisdiction than any Hong Kong court does. Third, post-overhaul election allowed only pro-Beijing candidates.7) Fourth, almost all major opposition parties, which traditionally supported LGBTQ+ rights8) and made court-curbing politically costly9), dissolved, with 45 opposition LegCo contenders jailed under the 2020 Law. Today, LegCo is dominated by pro-Beijing legislators traditionally opposing LGBTQ+ rights.
Against this background, the Court’s bold findings in the STK Decisions and the government’s willingness to comply, however minimally, were breakthroughs for judicial independence and rights protection. As the applicant in the STK Decisions recalled, hearing about them, he “yelled and teared”, grabbing bars of his prison cell – he, too, was among the jailed 45.
Lessons from the past and future judicial strategies
If LegCo rejects the Bill, what could the Court do to ensure the effectiveness of the STK Decisions – thereby maintaining the independence it reclaimed – when facing an omnipotent Beijing, an aggrandised executive, a hostile legislature, and the absence of an opposition movement? Its successful past may be illuminating.
The first lesson was discerning the varying political risks of different cases. LGBTQ+ rights were regarded as an issue in which Beijing had little core interest that warranted intervention, unlike national security. Indeed, it benefited from a legalistic-and-cosmopolitan-looking Hong Kong. Capitalising on this, the Court consistently expanded LGBTQ+ protection, blazing a trail among Asian jurisdictions.10)
Beijing’s position seems unchanged. To date, it has neither clearly opposed the STK Decisions nor the Bill, notwithstanding the former’s unprecedented and radical requirement for systematic change rather than piecemeal relief. This suggests a relatively low risk of Beijing’s retaliation if the Court acts assertively upon the Bill’s defeat.
Second, “focal point” judgments could channel popular support to pressure the authoritarian regime into compliance.
One must start by recognising that nothing compels authoritarians to follow unwelcome judgments. However, Hong Kong’s authoritarians (its executive head, legislature, and Beijing) – all undemocratically elected even pre-202011) – usually did follow. As Eric Ip suggested, this was because, despite having their preferences, if it cost them a large-scale clash with the public, they would concede (until recently). Thus, the effectiveness of an unwelcome judgment depended on whether authoritarians perceived it as enjoying strong enough public support to warrant concession.
Yet, Hong Kong had no genuine democratic institutions for the public to unite their power and put the authorities under pressure. It was the Court that partially stepped in. Its esteem, publicity and majoritarian reputation focused both the public’s and the authoritarians’ attention on its judgments, rendering them “focal points”. Therefore, when a judgment favoured the public over the authorities regarding an issue, it redirected popular expressions of discontent into support for the Court itself, thereby uniting popular voice and thus strengthening its perceived popularity. The same function enabled the authoritarians to efficiently evaluate the strength of popular will and decide, based on their perception of the judgment’s popularity, whether to concede to popular will by respecting the judgment.
While large-scale popular movement has disappeared, the Court’s judgments may remain focal points, increasing the likelihood of government’s compliance. The Bill’s 19.3% support rate during public consultation (para. 2) is evidence, as all supporters also expressed “respect for the court’s judgment” (para. 3).
While far-outnumbered and likely underestimated, 19.3% was actually impressive. Hong Kong people’s views on LGBTQ+ rights largely align with the pro-Beijing/opposition line.12) Yet, the total liquidation of the opposition camp, the nuisance and harassment inflicted on civil society (even Hong Kong’s largest LGBTQ+ event lost its venue during the consultation), and the written-submission-only requirement all suggested that mainstream political mobilisation for Bill supporters would be absent. Meanwhile, the pro-Beijing political machine remained intact. Accordingly, the support rate could have been much lower, since presumably only the most committed supporters – probably far less than 19.3% of the population – would have participated on their own volition. Against this backdrop, the 19.3% figure in fact demonstrated the Court’s focal point power in coordinating support on its own.
This was reinforced by two features of the Bill’s consultation. First, it was unique, being the only public consultation held on the STK Decisions. Second, it was a low-risk venue for expression, owing to its nature and its “reversed focal point effect” – informing citizens of the government’s attitude. This was crucial given current speech restrictions. It was no coincidence that 20% of supporters (para. 3) echoed the government’s narrative – “government respects rule of law, hence the Bill” – without citing rights protection.
Learning from this, if LegCo rejects the Bill, follow-up reliefs should amplify their popularity by providing unique and low–risk venues and loci for coordinating liberal expression. This and other incentives (e.g., economy and pink-washing)13) may be just enough to dissuade the authoritarian government from potentially disregarding the reliefs and its continuing duty to legislate.
Third, as Dixon and Yap argue, “responsive judicial remedy”14), such as suspended declarations, has helped the Court avoid backlash and promote dialogue while maintaining normative demands and the “democratic minimum core”, since such remedies carry fewer “stings” (e.g. less immediacy, more discretion).
The reliefs could and should be granted by the court
Tying these threads together, if the Bill is defeated and the government applies for an extension, the Court may consider the following interim reliefs.
First, for legal recognition, the Court could direct lower courts to grant a “declaration of partnership at common law” for same-sex partners.
To buffer opposition, this remedy resorts to a pre-existing jurisdiction. In NF v R, only one party in a lesbian couple was registered as a “parent” to an infant born to the them via RIVF. The other party applied for recognition as a parent under statutory law. Constrained by statutory language, the judge declined. However, finding the result absurd and discriminatory, the judge declared her “a parent of [the infant] at common law” (para. 161).
The remedy demonstrates both firmness and efficacy, since judicial recognition – in lieu of statutory recognition – confers legitimacy and dispels a sense of inferiority (para. 142), as the Court promised same-sex couples.
Second, for related rights and obligations, the Court may lay down a statutory interpretive rule that, as far as possible (para. 42), all laws applying to married couples should be read as applying to holders of the above declaration.
This avoids confrontation with LegCo. It merely reflects the established principle of constitution-consistent statutory interpretation, whose specific effects remain indeterminate until stated by policy or judgment, and which may be overriden by unambiguous legislative intent. Yet, this interpretation is broad and immediately effective.
Moreover, its three implicit “strengths” could keep LegCo engaged. First, provisions clearly displacing the proposed interpretive rule would likely invite constitutional challenge. Second, abolishing this rule is unconstitutional, as it flows from the constitution’s supremacy. Third, LegCo may not want to abolish this rule or the underlying principle, since both help reduce declarations of unconstitutionality by enabling courts to adopt strained or remedial interpretations instead.
LegCo would then face two choices: either amend each individual statute to exclude same-sex couples – a prohibitively cumbersome task that would ultimately leave same-sex couples with wider protections than the Bill – or reconsider the narrower Bill. Either way, same-sex partners obtain statutory protection, and the Court’s authority is reinforced.
Finally, given Bejing’s nonchalance, both remedies may function as standing focal points, reinforcing their resilience. The government’s stalling of legislating for the STK Decisions suggests that interested parties will focus their attention uniquely on litigating the details of the two remedies, thereby forming a specific, structured and low-risk venue (courtroom) where government and public could continuously exchange views. This is a pivotal function which authoritarians may not lightly discard (p. 353). Moreover, as same-sex couples keep invoking the two remedies – that is, application or clarification – the cumulative effect signals their and their allies’ strong commitment, deterring government from repealing the remedies.
Since the passage of the 2020 National Security Law, Hong Kong has witnessed a general deterioration in fundamental rights protections. Its performance on constraining government power and fundamental rights now ranks near the bottom of its income group. Today, even peaceful criticism may face up to a 10-year imprisonment if it is deemed, inter alia, to sow hatred, contempt or disaffection against the government (Division 4 of Part 3).
But it is in this tightened environment that, more than ever, the people of Hong Kong count on the judiciary to step up. The follow-up to the STK Decisions is both a strategic opportunity and moral obligation for the courts. While in national security cases judges may feel powerless amid Beijing’s wrath15), the foregoing arguments show that they may still make a difference by leveraging the preferences and moves of other actors’ to amplify the efficacy of their decisions.16)
I am indebted to Professors Cora Chan and Stefano Osella for their substantive and instructive help. All errors are mine.
References
| ↑1 | Cora Chan, ‘Pluralizing Constitutionalism’ in Madhav Khosla and Vicki C Jackson (eds), Redefining Comparative Constitutional Law: Essays for Mark Tushnet (Oxford University Press 2025). |
|---|---|
| ↑2 | Johannes Chan and C.L. Lim, ‘Interpreting Constitutional Rights and Permissible Restrictions’ in Johannes Chan and C.L. Lim (eds), Law of the Hong Kong Constitution (3rd edn, Sweet & Maxwell 2022), para 17.016. |
| ↑3 | Rediffusion (Hong Kong) Ltd v A-G of Hong Kong [1970] HKLR 231. |
| ↑4 | Eric C Ip, ‘Logical Foundations of Judicial Review of Legislation in the Hong Kong Special Administrative Region’ (2020) 50 Hong Kong Law Journal 19, 22-26; Johannes Chan and C.L. Lim (n 2), para 17.017. |
| ↑5 | Dinusha Panditaratne and Rehan Abeyratne, ‘Basic Law, Hong Kong Bill of Rights and the ICCPR’ in Johannes Chan and C.L. Lim (eds), Law of the Hong Kong Constitution (3rd edn, Sweet & Maxwell 2022), paras 16.059-16.060, 16.121. |
| ↑6 | Cora Chan, ‘From Legal Pluralism to Dual State: Evolution of the Relationship between the Chinese and Hong Kong Legal Orders’ (2022) 16 Law & Ethics of Human Rights 99. |
| ↑7 | Eric Chan, ‘The Legislative Council of Hong Kong’ in Po Jen Yap and Rehan Abeyratne (eds), Routledge Handbook of Asian Parliaments (Routledge 2023) 159-163. |
| ↑8 | See e.g. Amy Barrow, ‘Sexual Orientation, Gender Identity, and Equality in Hong Kong: Rights, Resistance, and Possibilities for Reform’ (2020) 15 Asian Journal of Comparative Law 126, 133. |
| ↑9 | Eric C Ip, Hybrid Constitutionalism: The Politics of Constitutional Review in Chinese Special Administrative Regions (Cambridge University Press 2019) 106, 190-191. |
| ↑10 | Julius Yam, ‘Approaching the Legitimacy Paradox in Hong Kong: Lessons for Hybrid Regime Courts’ (2021) 46 Law & Social Inquiry 153, 167-168; Rehan Abeyratne, Courts and LGBTQ+ Rights in an Age of Judicial Retrenchment (Oxford University Press 2025). |
| ↑11 | Albert H Y Chen and Po Jen Yap, The Constitutional System of the Hong Kong SAR: A Contextual Analysis (Hart Publishing 2023), 96-100. |
| ↑12 | Tien Ee Dominic Yeo and Tsz Hang Chu, ‘Beyond Homonegativity: Understanding Hong Kong People’s Attitudes About Social Acceptance of Gay/Lesbian People, Sexual Orientation Discrimination Protection, and Same-Sex Marriage’ (2018) 65 Journal of Homosexuality 1372. |
| ↑13 | Rehan Abeyratne (n 10) 228-230. |
| ↑14 | Rosalind Dixon, Responsive Judicial Review: Democracy and Dysfunction in the Modern Age (Oxford University Press 2023). |
| ↑15 | ‘Hong Kong’s Courtroom Dramas’ The Economist (London, 23 August 2025) 51. |
| ↑16 | Lee Epstein and Keren Weinshall, The Strategic Analysis of Judicial Behavior: A Comparative Perspective (Cambridge University Press 2021) 8-9. |
Updates and afterthoughts:
First, on 10 September, LegCo rejected the Bill by 71 to 14. The government stated that it would NOT seek an extension of time from the Court. This likely indicates that it considered its obligation fulfilled.
Second, the couple in NF v R had gone on to challenge the constitutionality of the statute in question. They succeeded in K v Secretary for Justice [2025] HKCFI 1974, handed down on 9 September. However, the judge there (“K-Judge”, a different one from that in NF v R) questioned (without rejecting) the validity of the declaration of parent at common law.
Given the escalation of events, it is necessary to add some afterthoughts:
1. The K-Judge’s questioning does not affect the argument in the article.
The K-Judge opined that (i) the case-laws did not provide for a definition of common law parent. They only described, from a child’s perspective, the diverse scenarios of parentage – legally recognised or not – so that the child’s best interests could be assessed roundly; (ii) parental rights and obligations in common law have been conclusively codified. There is thus no common law rights and obligations to be conferred by the declaration; (iii) accordingly, the declaration served no useful purpose, should not be termed “at common law”, and was just wrong.
I agree that the declaration of common law parent does not carry with it any right and obligation. That’s why in this article, the declaration of partnership at common law per se also does not confer rights or obligations – those are to be determined through statutory interpretation.
I disagree, however, that the declaration – be it of parent or partnership at common law – is wrong.
First, so long as a LEGAL state of affair – parentage to a child’s best interests, or partnership to the right to private life – is described in case-laws, there exists a “definition” “at common law” which an applicant could bring themselves under. Whether to judicially declare so is a different issue. Indeed, courts have even granted declarations that sort out “academic” legal positions without existing factual claim (Zamir & Woolf, The Declaratory Judgment (4th edn, Sweet & Maxwell 2011) para 4-99).
Second, while such declaration does not always come with rights and obligations, it still serves a purpose. In NF v R, even a nominal declaration of parent serves the child’s best interests – psychologically and socially. And in STK Decisions, the Court said unequivocally that legal recognition was a must so same-sex partners would not feel inferior and illegitimate.
2. What should be the eligibility requirement to a declaration of partnership at common law?
Given the escalation, the Court should invite parties to make submissions. Likely they will converge on the Bill’s term, and the Court should adopt it to maximise the chance of government compliance. At the same time, the Court should make clear that this is only interim, the requirement is only prima facie constitutional, and that the government is still obliged to provide a legislative framework which will be thoroughly scrutinised after its enactment. The Court could further stipulate default and permanent remedy if the government fails again.
3. What is the value of the proposed interpretive rule compared to simply striking down laws for infringing the right to private life?
The benefits are strategic. When granting follow-up remedies, the Court will be dialoguing with the legislature and executive while making its stance clear. Given the escalation and having conceded a margin of discretion, it is politically provocative for the Court to begin by saying that it will declare unconstitutionality. The interpretive rule, which is a milder remedy, could fill in instead. It preserves the look of division of functions by purportedly enforcing legislative intent – but only under the presumptions that legislative intent is “always speaking” and constitution-compliant.
Thus, where the legislative text, read naturally and ordinarily, admits of an updated and constitution-compliant meaning, the proposed interpretive rule is coyer than striking down the law but has the same effect in rights protection (Po Jen Yap, ‘Judging Hong Kong’s National Security Law’). Where the law’s language does not allow such construction, the Court could still strike it down.
4. Why the right to private life should mean equal treatment as-far-as-possible between same-sex common law partners and married (different-sex) couples? After all, right to private life is not the same as the rights to AND attendant on marriage; also, the Court in STK Decisions held that right to marriage is exclusive for different-sex partners, and same-sex partners could not assert it via the right to equality.
This is an oversight on my part, but it is not fatal to the argument. Indeed, the proposed rule is untenable if, by “as far as possible”, it means so long as the statutory language permits, courts should give same-sex common law partners equal treatments as married couples. However, if “as far as possible” means “so long as the statutory language permits AND within the extent of right to private life”, it is perfectly possible and desirable for courts to refer to married couples as the benchmark of same-sex common law partners’ entitlements.
The practical reality is that when same-sex partners claim a benefit under the right to private life, the benefit is usually also attached to the status of marriage. There is thus an intersection of the rights to equality and to private life. If same-sex partners are denied of part of the right to private life (especially the core rights), while such denial results in favourable treatment of different-sex partners, the cumulative detriment would likely lead to unconstitutionality. Thus, interpreting statutes consistent with right to private life WHILE referring to different-sex couples’ entitlement is both lawful and helpful for same-sex partners.
It is also strategically crucial for the Court to highlight equality between same-sex and different-sex partners within the extent of right to private life. It reiterates the Court’s commitment to protect minorities when much of LegCo’s opposition to the Bill was based on “mainstream value”.