AKRAM MUHAMMAD v. DIRECTOR OF IMMIGRATION AND ANOTHER

HCAL 1329/2022

[2023] HKCFI 595

IN THE HIGH COURT OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

COURT OF FIRST INSTANCE

CONSTITUTIONAL AND ADMINISTRATIVE LAW LIST NO 1329 OF 2022

________________________

________________________

__________________________________

REASONS FOR DECISION

__________________________________

A.  Introduction

1.  This is my second ‘Reasons for Decision’ following from a single hearing on 17 January 2023. It arises as follows.

2.  The Applicant is a Pakistani national. He entered Hong Kong illegally in 2008 and has been overstaying in Hong Kong ever since. In 2015, the Director issued a Removal Order against him under section 19(1)(b) of the Immigration Ordinance Cap 115 (“IO”). In 2022, the Secretary for Security issued a Deportation Order against him under section 20(1) of the IO.

3.  Since 19 December 2020, the Applicant has been held in immigration detention pending his removal or deportation from Hong Kong pursuant various powers granted to the Director or Secretary under the IO.

4.  The Applicant was originally detained in the Castle Peak Bay Immigration Centre (“CIC”), but was transferred to the Tai Tam Gap Correctional Institution (“TGCI”) on 17 June 2021. Detainees in the two institutions are subject to different detention conditions. The Applicant viewed the detention treatment in CIC as more favourable than that at TGCI. He repeatedly requested to be transferred back to CIC, but the requests were not acceded to.

5.  By a ‘hybrid’ application form filed in this action on 24 November 2022, the Applicant specifically canvassed both (1) an application for leave to apply for judicial review arising from the difference in detention conditions between TGCI and CIC, and (2) a usual habeas corpus application seeking the Applicant’s immediate release from immigration detention (albeit pending his removal from Hong Kong).

6.  Both applications were heard by me on 17 January 2023. At the hearing, the Applicant was represented by Mr Jeffrey Tam and Mr Albert Wan of Counsel. The Director and the Secretary were represented by Mr Michael Lok of Counsel.

7.  At the end of the hearing, I dismissed both applications with reasons to be handed down later. It would obviously have been convenient to give reasons for the dismissal of both the judicial review leave application and the habeas corpus application in one go, and that was what I intended at the conclusion of the hearing. But I eventually had to bifurcate my reasons for the two applications, and first to deliver reasons for dismissing the habeas corpus application, necessitated by the subsequent events narrated in those reasons. I did so in my first ‘Reasons for Decision’ dated 31 January 2023, [2023] HKCFI 174.

8.  These are my reasons for dismissing the application for leave to apply for judicial review.

B.  The Amendment

9.  In the form filed on 24 November 2022, the Applicant identified two targets as the subject of challenge and two proposed grounds, with each target supported by one ground:

(1)  The Applicant identified Paragraph 3 of the Immigration (Places of Detention) Order Cap 115B (“115B Order”) – which governs detention conditions in TGCI – as the subject of challenge. This challenge was supported by Ground 1, which says that the 115B Order is Wednesbury unreasonable in that it violates the principle of equality. This was the judicial review component in this action (“Unequal Treatment Ground”).

(2)  The second target identified was the decision by the Director/Secretary to continue to detain the Applicant in TGCI (“Detention Decision”). This was supported by Ground 2, which says that the Applicant’s detention had become unlawful by reference to Hardial Singh principles. This is the habeas corpus component.

10.  As can be seen, the judicial review component and the habeas corpus component in this action were originally neatly divided. The Unequal Treatment Ground was only intended to support the proposed judicial review against Paragraph 3 of the 115B Order, and would not concern the Detention Decision.

11.  However, by a summons filed on 9 January 2023, the Applicant applied to amend his application form. The textual amendments sought were minor but the effect was to add the Unequal Treatment Ground to challenge not only the 115B Order but also the Detention Decision.

12.  I allowed the amendment at the hearing. I agreed with Mr Tam that the Director/Secretary would not be prejudiced:

(1)  Adding the Unequal Treatment Ground to the Detention Decision is consistent with, if not already forming part of, the original application. The Detention Decision speaks of not only continued detention, but detention in TGCI specifically. The relief sought included, as an alternative to immediate release from detention, an order of mandamus to transfer the Applicant from TGCI to CIC or a declaration that the Applicant is not subject to the Prison Rules (which are applicable in TGCI).

(2)  Mr Tam said the amendment has caused little change to his arguments for the Unequal Treatment Ground. This was reflected in the very minor textual change to the application form. Indeed, under the heading of Ground 1: Irrationality: violation of principle of equality, there are only three minor additional parts of text which do nothing more than indicating that this ground is now also applicable to the Detention Decision.

(3)  The Secretary/Director had filed 6 affirmations covering a wide range of topics, including evidence relevant to TGCI and CIC generally as well as evidence specific to the Applicant. Mr Tam said there would be no need for the Director/Secretary to file further evidence to cater for the amendment. Indeed, Mr Lok did not argue otherwise or indicate what any further evidence (if any) might cover.

(4)  The amendment application was taken out about 1½ months after the application form was filed. Whilst the oral hearing took place only a week after the amendment application, that was because I had fixed a rather early hearing in view of the habeas corpus component which would usually call for a hearing as early as possible.

13.  After the amendment, the judicial review component became an application for leave to challenge, on the Unequal Treatment Ground, both Paragraph 3 of the 115B Order as well as the Detention Decision.

C.  Legislative Scheme Governing Places of Detention and Detention Treatment

C.1  Introduction

14.  The places for detaining immigration detainees, pending their removal, and the relevant detention treatment are provided for in subsidiary legislation enacted under the IO.

15.  Section 35(1) of the IO empowers the Secretary, by order, to authorize immigration detainees to be detained in such places; and the CE, by order, to provide for treatment of such detainees during detention.

16.  The 115B Order was enacted under section 35(1). Paragraph 2 provides that places specified in Schedules 1, 2, and 3 shall be places for immigration detention.

(1)  TGCI can be found in Schedule 1 of the Order. Schedule 1 includes sites and buildings from time to time set apart for the purposes under the Prisons Ordinance Cap 234 (“PO”) and which are specified in Prisons Order Cap 234B. TGCI is specified in Cap 234B and thus is authorized by law as a place for immigration detention.

(2)  CIC is expressly listed in Schedule 3 to the 115B Order.

C.2  CIC: Schedule 1 of 115E Order

17.  The treatment of detainees in CIC is provided for in Schedule 1 of the Immigration (Treatment of Detainees) Order Cap 115E (“115E Order”) which is also enacted under the powers of section 35(1) of the IO. Paragraph 2 of the 115E Order provides that the persons detained in places specified in Schedule 2 should be accorded with the treatment specified in Schedule 1. Currently, CIC is the only place listed in Schedule 2.

18.  Schedule 1 of the 115E Order is short, containing less than 30 rules. The Secretary/Director published a “Castle Peak Bay Immigration Centre Operation Manual” (“CIC Operation Manual”) to provide guidance on the management of detainees in CIC with regard to their daily routine and welfare, as well as the security of CIC. The CIC Operation Manual is not subsidiary legislation.

C.3  TCGI: Paragraph 3 of 115B Order and Prison Rules

19.  Mr Tam understood Paragraph 3 of the 115B Order as the sole or main provision governing the treatment of detainees in TGCI, and the source of differential treatment between the two detention places. This is why the Applicant sought as part of the relief a declaration that Paragraph 3 is irrational and an order of certiorari to bring up and quash that paragraph. But as will be explained below, I think this reflects a misunderstanding as to the legal effect of Paragraph 3 in the overall legislative scheme governing detention treatment.

20.  Paragraph 3 of the 115B Order provides as follows:

3. Application of Prison Rules

A person detained in any place set out in Schedule 1 [including TGCI] shall receive the same treatment as that which is accorded to a person committed to prison for safe custody in any of the circumstances specified in rule 188(1) of the Prison Rules (Cap. 234 sub. leg. A) and, mutatus mutandis, rules 189 to 207 (inclusive) of the Prison Rules (Cap. 234 sub. leg. A) shall apply to him.

21.  Rules 188 to 207 of the Prison Rules Cap 234A (“Prison Rules”) are put under Division 1 of Part II of the Prison Rules. Part II is headed “Special Rules for Particular Class of Prisoners” and Division 1 is headed “Prisoners Awaiting Trial”. Part II is also comprised of divisions for other classes of prisoners, such as convicted persons who have already lodged an appeal and prisoners committed for contempt. The effect of Paragraph 3 of the 115B Order is that prison rules applicable to prisoners awaiting trial become applicable to TGCI detainees.

22.  It appears that Mr Tam sees Paragraph 3 as the source of the more stringent detention treatment imposed on TGCI detainees, and his intended consequence of quashing Paragraph 3 would be to relieve TGCI detainees from such stringent treatment. This gives rise to a misunderstanding because, under the current legislative scheme, quashing Paragraph 3 of the 115B Order would result in TGCI detainees being made subject to prison rules applicable to convicted prisoners – the very opposite of what the Applicant seeks to achieve with this application.

23.  This is because Rule 2 of the Prison Rules provides:

The rules in this Part [i.e. Part I] shall apply to all classes of prisoners except in so far as they may be inconsistent with the rules made to govern any particular class or classes of prisoners.

24.  “Prisoners” is defined in section 2 of the PO as to include “a person who is sentenced in a place outside Hong Kong and is brought to Hong Kong in order to serve the sentence imposed upon him (or any part thereof) in that place”. This definition is not helpful in that it is neither a prescriptive nor an exhaustive definition. But it is clear that “prisoners” includes not only convicted criminals serving their sentences but also widely covers those lawfully confined in places which are set apart for the purposes of prisons in Cap 234B. For example:

(1)  Section 10(1) of the PO reads “a person shall be deemed to be a prisoner and in legal custody whenever he is being taken to or from, or is confined in, any prison in which he may be lawfully confined, whether under criminal or civil process, or whenever he is working outside or is otherwise outside any such prison in the custody or under the control of an officer of the Correctional Services Department.”

(2)  Section 7 of the PO provides that “any prisoner sentenced to imprisonment or committed to prison on remand pending trial or otherwise may be lawfully confined in any prison to which this Ordinance applies.”

(3)  As already alluded to, Part II of the Prison Rules makes reference to different classes of “prisoners”, including those awaiting trial, convicted persons who have lodged an appeal and those who have been committed to prison for contempt of court.

25.  Thus, by setting apart TGCI for the purposes of prisons in Cap 234B, those detained in TGCI become “prisoners”. By virtue of Rule 2, all rules provided in Part I – headed “General Rules for the Government of Prisons” are applicable to TGCI detainees except when they are inconsistent with rules for particular classes of prisoners, which are provided in Part II.

26.  Part I currently contains 176 rules generally applicable to all prisoners. The special rules for particular classes of prisoners in Part II modify the application of the general rules in Part I when the two are inconsistent.

27.  Taking Division 1 – the governing rules for prisoners awaiting trial – as an example, most of the rules are meant to provide prisoners awaiting trial with privileges which are not enjoyed by convicted criminals serving their sentences:

(1)  Rule 188 defines what kind of prisoners should be subject to rules for prisoners awaiting trial, which specifically includes persons who are committed to prison for safe custody for detention by virtue of the IO, i.e. immigration detainees.

(2)  Rules 190 requires that they be segregated from convicted prisoners.

(3)  Rule 192 allows them to procure food or to receive food and malt liquor from the outside, whilst these are generally prohibited under Rule 31 for other prisoners.

(4)  Rule 196 allows them to wear private clothes, whilst other prisoners would wear clothing provided under Rule 26.

(5)  Rule 201 makes employment an option for them whilst other prisoners are generally required to work under Rule 38.

(6)  Rules 202 allows them to retain books and documents on the person’s person at the time of arrest and the supply of newspapers at thier own expense.

(7)  Rule 203 allows them to be visited by one or two visitors for a quarter of an hour on any week day. Rule 205 allows them to see friends and relatives on any week day for the purpose of providing bail. Other prisoners are subject to the more stringent visiting rules under Rule 48.

(8)  Rule 204 allows them to receive a private doctor selected by themselves for the purpose of their defence.

(9)  Rule 206 allows them to send and receive letters at all reasonable times and be provided with a reasonable amount of paper and writing materials for communicating purposes. Other prisoners are only provided with materials sufficient to write one letter per week: see Rule 47.

28.  Indeed, Paragraph 3A of the 115B Order refers to Rules 203 to 206 as “privileges” which might be withdrawn by the Commissioner of Correctional Services:

3A. Power to modify privileges under Prison Rules

Notwithstanding paragraph 3, the privileges specified in rules 203 and 206 of the Prison Rules (Cap. 234 sub. leg. A) may be modified or withdrawn—

(a) by the Commissioner of Correctional Services in his discretion in respect of a person detained in a place specified in Schedule 1 of which he has the control;

(b) by the Commissioner of Police in his discretion in respect of a person detained in a place specified in Schedule 1 of which he has the control.

29.  Therefore, contrary to what Mr Tam might have in mind, the bulk of the rules governing detaining treatments in TGCI actually comes from the 176 general rules as found in Part I of the Prison Rules, subject to modification by rules applicable to prisoners awaiting trial provided for in Part II. Quashing Paragraph 3 of the 115B Order would remove TGCI detainees’ privileges and subject them to the general rules applicable to convicted criminals. This does not appear to be what the Applicant intended to achieve with the proposed judicial review.

30.  To make the Applicant’s case intelligible, his true complaint must be that the legislative scheme as a whole (including the 115B Order, the 115E Order, Cap 234B and the Prison Rules) operates to the overall effect of subjecting TGCI detainees to less favourable detention treatment than applied to those detained in CIC. It appears to me that, thus, the proper relief sought to achieve what the Applicant truly intends is to remove TGCI from the list of places set apart for prison purposes in Cap 234B, and expressly to list it under Schedule 1, 2 or 3 of the 115B Order and to list it in Schedule 2 of the 115E Order.

31.  I do not see the mismatch between the relief sought and the Applicant’s true complaint as fatal to Mr Tam’s argument. If I agree with Mr Tam that the relevant subsidiary legislation should be amended such that treatment of TGCI detainees is brought on par with CIC detainees, I see no reason why the intended effect could not be achieved by amending the relevant subsidiary legislation as suggested above or left to be worked out.

32.  I would also note that I do not attach particular weight to parties’ references to whether detainees are treated as prisoners and are subject to Prison Rules. Mr Tam said TGCI detainees are treated like prisoners and are subject to Prison Rules. Mr Lok disagrees with both propositions. Without clarifying the meaning of “prisoners” as intended by them, a discussion on whether TGCI detainees are prisoners is not very meaningful. In ordinary language, the term prisoners might be an exclusive reference to convicted criminal serving their sentences. But as I have explained above, under the PO and the Prison Rules, prisoners are those lawfully confined in prisons for safe custody. Under the current legislative scheme, TGCI detainees are prisoners and are subject to the Prison Rules. But that is neither here nor there for the resolution of this application.

33.  However, as will be seen below, the failure to identify the real target of attack in the Applicant’s systemic challenge has some ‘ripple’ effect on both sides’ arguments. But, insofar as practicable or necessary, I will take any reference to Paragraph 3 of the 115B Order as reference to the relevant legislative scheme (including the 115B Order, the 115E Order, Cap 234B and the Prison Rules).

D.  Further Background on CIC

34.  Further background on CIC may be helpful.

35.  Before the setting up of CIC, there were no dedicated facilities for housing immigration detainees pending their removal. Such detainees were initially housed at the ‘Victoria Immigration Centre’ inside the Victoria Prison. Apparently immigration detainees housed in the Victoria Prison were also subject to the Prison Rules applicable to prisoners awaiting trial: see Paragraph 3B of the 115B Order which specifically makes reference to “a person detained in that part of Victoria Prison Known as Victoria Immigration Centre”. Victoria Immigration Centre’s capacity was gradually saturated and detainees were then diverted to other prisons.

36.  CIC commenced operation in 2005. It was the first institution set up exclusively to house immigration detainees. It had a capacity to house 400 detainees at inception, but the capacity has since been increased to 500. At its inception, it was specified as a prison under Cap 234B and was managed by the Correctional Service Department (“CSD”), and the detainees were subject to the Prison Rules applicable to prisoners awaiting trial. Therefore, the legislative scheme applicable to TGCI now is identical to what was applicable to CIC before 2010.

37.  It was explained in the Government’s evidence that CIC was initially arranged to be managed by CSD rather than Immigration Department (“ImmD”) due to manpower considerations between the two departments.

38.  In 2010, the management of CIC was transferred from CSD to ImmD. The main corresponding legislative amendments made to enable this to happen are: (a) CIC was removed from the Schedule of Cap 234B such that it was no longer designated as a prison; (b) CIC was added to Schedule 3 of the 115B Order such that it would remain as an authorized place for immigration detention despite its removal from Cap 234B; and (c) to list CIC in Schedule 2 of the 115E Order such that the rules in Schedule 1 of the same order become applicable to it. Therefore, to achieve what the Applicant truly seeks in this application, TGCI would have to go through the legislative amendments applied to CIC in 2010.

39.  In the paper submitted by the Security Bureau to the LegCo in February 2010, the Secretary explained that, after the transfer of management from CSD to ImmD, CIC detainees’ treatment would be governed by the 115E Order as well the CIC Operation Manual. The Secretary’s position was that the treatment of CIC detainees after the transfer would remain the same if not more generous. The Secretary annexed to the paper a table setting out the comparison of detainee treatment under the Prison Rules and under the 115E Order together with the CIC Operation Manual. It can see that for aspects which are provided for in the Prison Rules but not in the 115E Order, the Secretary tried to filled the gap by providing for matching arrangements in the CIC Operation Manual.

40.  The Secretary noted in that table that there would be no comparable body cavity search in CIC after the transfer because, at that time, the CIC seldom housed convicted persons known to be involved in drug trafficking or addition and such power could thus be dispensed with. It was also noted that after the transfer, the CIC Operation Manual would only ask detainees to open a letter in the sight of an officer but the officer would not read the content of the letter.

41.  In the current application, the Applicant clearly sees the Prison Rules applicable to prisoners awaiting trial as less favourable than the combined effect of the 115E Order and the CIC Operation Manual. That view was not exactly shared by the LegCo members in 2010. In the report submitted in April 2010 to the House Committee of the LegCo (“2010 LegCo Report”), members questioned why the 115E Order was not amended to align with provisions in the Prison Rules. They saw the CIC Operation Manual as offering lesser protection against breaches of its rules than the Prison Rules or the 115E Order which were subsidiary legislation.

42.  Prison Rules do not only empower the CSD to manage prisoners, but also establish the prisoners’ legal entitlements when they are remanded in custody. For example, the rules operate to ensure that prisoners should be entitled to sufficient food and diet, clothing, and bedding; to make complaints or to appeal against CSD’s arrangements or orders, to be visited by Justices of the Peace, to be visited by family and friends, to communicate with the outside world by letters, etc.

43.  In the 2010 LegCo Report, it was also recorded that some members worried that the absence of a power to conduct urine examination and body cavity searches would mean that drug-taking detainees might not be easily detected. They requested the Secretary/Director to consider transferring CIC detainees with drug problems or long term CIC detainees to centres managed by CSD and be governed by Prison Rules. The Secretary again said that it did not see the need to have such powers because CIC did not have detainees to be found to be in possession of drugs in the previous 5 years, i.e. since its inception in 2005 until 2010.

44.  About a decade later, the Secretary eventually changed his view and decided greater powers would be needed for ImmD to uphold the discipline and order in CIC in view of what transpired in the 2010s.

45.  It is a matter of public record that there has been an influx of non-refoulement claimants into Hong Kong in the 2010s. It has been usual that years pass before a claimant has exhausted all the available avenues for the adjudication of his claim and meanwhile his illegal stay in Hong Kong would be tolerated by the Secretary/Director. As of October 2022, there was about 14,900 such non-refoulement claimants in Hong Kong.

46.  Since 2016, the Secretary/Director has geared up in removing claimants who were due to be removed. That resulted in more and more detainees being detained in CIC pending their removal. The length of detention has also increased substantially for reasons as explained in the series of habeas corpus decisions delivered by the Court last year.

47.  In a LegCo paper submitted by the Secretary in November 2022 for the discussion of the Panel on Security (“2022 LegCo Paper”), the Government has proposed to amend the 115E Order by reference to Rules 9(1A), 34A, 58, 61 and 63 of Prison Rules in order to confer similar urine examination and body cavity search powers on ImmD, as well as to impose heavier punishment by increasing the maximum period of separate confinement to 28 days.

48.  In the 2022 LegCo Paper, the Secretary pointed to the change in the demographic of CIC detainees over the years in support of the need for greater powers to uphold discipline and order in CIC:

(1)  The period of detention for CIC detainees has generally become longer over time, leading to higher propensity to uncooperative behaviour. From December 2012 to October 2022, the proportion of detainees at CIC having been detained for less than a month decreased from 52% to 44%, and the proportion of detainees detained for over 180 days increased from 2.6% to 10.6%. I would also add that (as was mentioned by the Secretary in an earlier LegCo paper rather than the 2022 LegCo Paper) that from 2007 to 2009, the average detention period in CIC was only 16 to 17 days.

(2)  Detainees with a criminal background have become more prevalent in CIC. From 2015 to 2022, the percentage of newly admitted detainees who were discharged prisoners increased from 28% to 62%. In 2022, around 25% of the discharged prisoners had committed serious crimes such as murder, trafficking in dangerous drugs, robbery, burglary, wounding, blackmailing and rape, etc. Some of them are repeated offenders.

(3)  It has become increasingly common for non-refoulement claimants to be arrested for drug offences. From 2013 to 2022, amongst 9,600 non-ethnic Chinese Form-8 holders arrested, over 13% of them were arrested for drug offences.

(4)  Indiscipline or uncooperative cases have become more prevalent in CIC. In 2019, 2020 and 2021, there were 67, 69 and 87 indiscipline cases in CIC, with 61, 53 and 109 detainees ordered to separate confinement ranging from one to seven days. Since ImmD took over CIC until October 2022, 21 immigration officers were injured in handling indiscipline cases.

E.  Recommissioning of TGCI and the Policy to Concentrate Detainees with Higher Security Risk there

49.  As of 2019, CIC was approaching its full capacity; the overall occupancy has reached 80% and that for the male dormitories had reached over 90%. It became necessary to increase the Government’s detention capacity.

50.  Eventually TGCI was chosen as the suitable institution to provide the extra capacity to house immigration detainees. TGCI was a decommissioned detention facility under the management of CSD. It was originally listed as a prison under Cap 234B and thus already authorized for detaining immigration detainees under the 115B Order. As said, being designated as a prison means that persons lawfully confined in TGCI would be subject to Prison Rules. TGCI commenced operation in May 2021 with a detention capacity of 160.

51.  In the affidavit evidence filed by the Secretary/Director, it was said that the recommissioning of TGCI was not only to increase the Government’s detention capacity but there was also a “policy intent” to concentrate immigration detainees who pose or likely to pose a higher level of threat or security risk to the community in TGCI. It was said that CIC, being a low level security facility, is unsuitable for housing detainees with criminal records for a long term.

52.  The Secretary/Director referred to a number of exhibits authored by the Secretary or the CSD either in submission to LegCo or in reply to media enquires in support of the existence of the policy. The earliest reference that detainees with criminal records or higher security risk would be concentrated in TGCI was first mentioned in Security Bureau’s reply to media enquiry on 24 September 2021. The same message was also repeated in public documents subsequently issued by the Government.

53.  The Government did not exhibit any policy document which sets out in more detail how the Secretary/Director would assess if a detainee should be housed in TGCI rather than in CIC, or how the policy works generally. More detail about this policy only comes from the affidavit evidence. It was said that it would not be possible to list out all the serious crimes which would be accorded priority in detention in TGCI as each case must be considered on a case-by-case basis. It is said that the factors to be taken into account in according priority in transferring detainees to TGCI include (a) the nature of offences committed by the detainee, and (b) the gravity of the offence as reflected in the length of sentence. It is also said that other factors such as the status of the detainee’s non-refoulement claim and removal progress would also be taken into account in deciding if a detainee should be transferred to TGCI. The evidence did not give a complete set of statistics comparing the security profiles of TGCI detainees and CIC detainees.

54.  For TGCI, what the Secretary/Director could say conclusively is that, as of 15 December 2022, all TGCI detainees are convicted criminals who were sentenced to imprisonment. Other than that, there are no statistics on the categories of crimes committed or the sentences served by TGCI detainees. But it was said that some detainees committed serious and violent crime such as rape, trafficking in dangerous drugs or wounding with intent, and were sentenced to more than 10 years. Some of them are repeated offenders, persons considered by the Police to be a security concern or with a history of absconding.

55.  For CIC, the Secretary/Director did not refer to any statistics concerning the criminal background of CIC detainees in the affirmation evidence, but Mr Tam pointed to what the Secretary has said in the 2022 LegCo Paper about CIC detainees’ criminal background.

56.  The Secretary/Director also pointed to the higher frequency of disciplinary cases in TGCI than in CIC as a reflection in the difference in detainees’ security risk profile. From May 2021 to 14 December 2022, the number of detainees in TGCI was only about one-third of that in CIC, yet the number of disciplinary cases in the former was double that in the latter. In TGCI, there were 257 disciplinary cases involving 203 detainees. In CIC, there were 136 disciplinary cases involving 265 detainees.

F.  Actual Difference in Treatment in CIC and TGCI

57.  As might be imagined, there could be myriad actual differences in treatment, major or minor, in the lives of detainees held in TGCI and CIC respectively.

58.  At the oral hearing, Mr Tam relied upon 8 areas of actual difference in treatment, and sensibly did not press on some other differences also mentioned in the application or the Applicant’s affidavit evidence. Mr Tam said the treatment in TGCI is more intrusive to detainees’ privacy and personal liberty. As said, he understood the difference as arising from Paragraph 3 of the 115B Order and I have explained why that is incorrect. Further, by looking at the 8 areas themselves, one can also see that not all of them are provided for in the subsidiary legislation. Some of the treatment (particularly as applied in CIC) is merely the result of practices or policies adopted by the ImmD in-charge of CIC, or the CSD in charge of TGCI).

59.  These are the 8 items of differential treatment relied upon by Mr Tam:

(1)  Body cavity search: TGCI detainees, prior to their admission to the institution, may be subject to a search in their rectum, nostrils, ears and other external orifice. The Director/Secretary’s affidavit evidence says that the CSD has adopted X-Ray body scan as an alternative to replace the manual rectal search. The power for such search is provided in Rule 9(1A) of the Prison Rules. Such power is not to be found in Schedule 1 of the 115E Order, nor is such practice adopted in CIC.

(2)  Urine examination: TGCI detainees may be required to submit a urine specimen for examination. Currently, detainees’ urine is tested to see if prohibited drugs could be detected. Such power is provided in Rule 34A of the Prison Rules. Such power is not to be found in Schedule 1 of the 115E Order, nor is such practice adopted in CIC.

(3)  Letters: In TGCI, both inward and outward letters may be opened, read, searched for articles which may pose a threat to security or order and discipline in the prison, or be stopped for specified reasons under Rule 47A of the Prison Rules. In CIC, letters will only be opened and searched for unauthorized articles, but not read or stopped.

(4)  Telephone calls: Telephone conversations of TGCI detainees are recorded in a system known as the “Integrated Intelligent Communication System” and retained for 31 days. The system can also detect sensitive keywords during phone calls and will immediately bring them to the CSD staff’s attention. Such practice is not grounded in the Prison Rules but falls within the Government’s initiative/policy to build TGCI as the first generation ‘smart prison’. Telephone conversations in CIC are monitored by staff in CIC but are not recorded.

(5)  Social Visits: In TGCI, conversation between visitors and detainees will be both monitored by CSD staff and also recorded. This practice is not grounded in any Prison Rules. In CIC, there will be a staff member at the end of the visit room to observe detainees and their conversation with visitors but the conversations will not be recorded.

(6)  CCTV: The CCTV coverage in TGCI is more extensive than in CIC. In TGCI, day and night accommodation including cells, dormitories and hospital wards are under full CCTV coverage. The toilets are also covered by CCTV but its system is programmed to block out the urinal, and thus the private part of a detainee. There is no CCTV coverage in the bathroom. Further the CCTV system in TGCI is equipped with multiple high-tech functions such as facial recognition and body temperature detection. These are also part of the Government’s initiative to build TGCI as the first generation ‘smart prison’. In CIC, there is no CCTV coverage in toilets, shower rooms, dormitories, special units on detention floors and all wards in the sickbay. But there is CCTV coverage in dayrooms, exercise yards, recreation rooms, sally ports, corridors and lobby on detention floors, isolation wards and protected rooms in sickbay, main gate, visit booths and interview rooms of visit room and waiting areas in reception office.

(7)  Disciplinary action: There are more categories of offences against institutional discipline and heavier punishment for breaches in TGCI than in CIC. In TGCI, there are currently 17 categories of offences against institutional discipline the breaching of which could attract 28 days of separate confinement as punishment: see Rules 61 and 63 of the Prison Rules. In CIC, there are 5 categories of offences the breaching of which attract 7 days of separate confinement: see Rule 13(1) of the 115E Order. But Rule 63 of the Prison Rules also provides for a mechanism for TGCI detainees to appeal against punishments imposed on them, whilst there is no equivalent appeal mechanism in CIC.

(8)  Removal from association: Both the Prison Rules and the 115E Order enable the authority to order a detainee to be separately confined if it is the authority’s opinion that such order is in the interests of the detainee or in the interests of good order and discipline in the institution: see Rule 68B of the Prison Rules and Rule 13(2) of the 115E Order. But the maximum period of separate confinement in CIC is 7 days, whereas in TGCI detainee may be so removed for initially not more than 72 hours and subsequently not more than 1 month and from month to month if CSD has reasonable grounds to do so.

60.  Mr Tam originally pressed for 9 areas of differential treatment in his written submissions, including that TGCI detainees would have to wear a ‘smart wristband’. By the time of the hearing, it had become clear that the wearing of a smart wrist band is optional and Mr Tam sensibly dropped this item. In the application form filed in this action, reference was also made to TGCI detainees having to make their own beds whilst CIC detainees do not, that there are some differences in the cleaning duties in the two institutions, and that CIC detainees are allowed to bring in “reference materials” – not clearly defined – but TGCI detainees are not. These points were not really relied upon by Mr Tam in the oral hearing, and I would not focus on them.

61.  It may be convenient to note a few points at this juncture:

(1)  As said, not all the treatment in TGCI or CIC, such as the CCTV arrangement or the recording of telephone calls, is grounded in the Prison Rules or the 115E Order. But there is no challenge to the legality of any of the treatment in TGCI per se.

(2)  As an extension of the previous point, whilst Mr Tam says that the treatment in TGCI amounts to greater intrusion into detainees’ privacy and personal liberty as compared to that in CIC, there is no challenge to the treatment in TGCI on its own, be it based on alleged human rights violation or otherwise.

(3)  Mr Tam made no distinction in advancing the 8 areas of actual differential treatment in regard to Paragraph 3 of the 115B Order (or the legislative scheme) vis-a-vis the Detention Decision. But, as said, not all treatment could be said to have arisen from the legislative scheme.

62.  It can also be usefully noted here that, whilst it might be thought generally more appropriate for immigration detainees to be detained in an immigration detention centre, rather than a prison (hence, operated by ImmD rather than the CSD), that would not make detention in a prison or a place subject to the Prison Rules generally in breach of any public law requirement, or in any way arbitrary: see, for example, R (Idira) v Home Secretary [2016] 1 WLR 1694 at §§61-62. Reference can also be made to §52 of the same case (which was concerned with whether detention in a prison rather than an IRC was arbitrary under Articl 5.1 of the Convention for the Protection of Human Rights and Fundamental Freedoms), which paragraph identified that the phrase “unduly harsh” captures the essential point that the place and conditions of immigration detention must be seriously inappropriate before a detention can properly be regarded as arbitrary and therefore unlawful.

G.  The Challenge to Paragraph 3 of the 115B Order on the Unequal Treatment Ground

63.  This is a systemic challenge. As explained, I shall take Mr Tam as referring to the entire legislative scheme as is said to produce the offending effect, rather than Paragraph 3 alone. Mr Tam said the principle of equality requires that immigration detainees should be treated alike even if they may be detained in different places of detention.

64.  He relied on QT v Director of Immigration (2018) 21 HKCFAR 324, at §§19-23, where the CFA ruled that the very wide powers granted to the Director in immigration controls are nonetheless still subject to the principle of equality in their exercise, and the principle of equality is an important aspect of rationality.

65.  Mr Lok’s replies could be summarised as follows:

(1)  The 115B Order is subsidiary legislation. The Court engaging its conventional administrative law reviewing powers cannot invalidate subsidiary legislation on the ground of unreasonableness. The only ground to do so is if the Court has found the subsidiary legislation to be ultra vires the enabling parent legislation which confers powers for the enactment of that subsidiary legislation.

(2)  Paragraph 3 of the 115B Order has built-in flexibility to allow the Government to adjust the Prison Rules or treatment as applied to TGCI. The legislative scheme does not lead to any inherent inequality of treatment between TGCI detainees and CIC detainees.

(3)  In any event, TGCI detainees and CIC detainees are not proper comparators such that any difference in treatment requires no justification, and/or is in any event justified.

66.  Each of the three arguments would be sufficient on its own to defeat the systemic challenge to the legislative scheme. The third argument was also advanced in respect of the Detention Decision.

H.  Can subsidiary legislation be invalidated on ground of irrationality?

67.  This issue arises from the fact that the judicial review component of this case was firmly framed as a conventional administrative law review, rather than a constitutional review on the right to equality or non-discrimination. This is, perhaps, designed to circumvent the difficulties of the Applicant’s being a non-Hong Kong resident and the immigration reservation under section 11 of the Bill or Rights (“BOR”).

68.  In QT, the applicant was also not a Hong Kong resident. At §24, the CFA observed that as the challenge was made to an administrative policy and not to primary legislation, there was no need to rely on the Court’s powers of constitutional review.

69.  Insofar as Paragraph 3 of the 115B Order (or the legislative scheme) is concerned, unlike QT, the Applicant was not challenging a policy adopted by the Secretary/Director in managing TGCI. But nor was the Applicant challenging primary legislation which, as hinted in QT, would require the Court’s constitutional review powers.

70.  The question thus arises as to whether the Court, invoking only its judicial review powers, can invalidate subsidiary legislation on the ground of irrationality. Mr Tam said the answer is ‘yes’, relying on Kruse v Johnson [1889] 2 QB 91 and The Attorney General v Tsang Kwok-Kuen [1971] HKLR 266.

71.  In the Kruse case, the local Council made a by-law under the Local Government Act 1888 prohibiting singing and dancing under certain conditions. Whilst there are references to the fact that a by-law could be invalidated by the Court on the ground of unreasonableness, the case is not clear authority for this proposition in the sense intended by Mr Tam. Closer examination of the majority view reveals that unreasonableness per se was not intended as sufficient to invalidate the by-law. Rather, the unreasonableness of the by-law was seen as evidencing that the by-law was outside the vires of the enabling parent legislation: see, for example, Lord Russell CJ (with whom the majority agreed) at 99-100:

Notwithstanding what Cockburn CJ said in Bailey v Williamson, an analogous case, I do not mean to say that there may not be cases in which it would be the duty of the Court to condemn by-laws, made under such authority as these are made, as invalid because unreasonable. But unreasonable in what sense? If, for instance, they were found to be partial and unequal in their operation as between different classes; if they were manifestly unjust; if they disclosed bad faith; if they involved such oppressive or gratuitous interference with the rights of those subject to them as could find no justification in the minds of reasonable men, the Court might well say. “Parliament never intended to give authority to make such rules; they are unreasonable and ultra vires”.

72.  In the Tsang Kwok Kuen case, the validity of a regulation – subsidiary legislation enacted under the Road Traffic Ordinance – was in question. Contrary to the submission of Mr Tam, this case lays down the principle that unreasonableness of the subsidiary legislation on its own is an insufficient ground for invalidation. Rather it is only a factor in assessing the ultimate question of whether the subsidiary legislation is ultra vires the enabling legislation: see 272-276, where Rigby CJ said:

For my part, I accept the argument of the Solicitor General that subsidiary legislation cannot be held to be ultra vires simply because a court may regard it as unreasonable. Different considerations arise where such subsidiary legislation is, in itself, not only unreasonable but contrary to natural justice and no express power is contained in the parent legislation enabling this to be done.

As to the question postulated under 4(b) it is, I think, clear that unreasonableness is not in itself a ground for holding that subsidiary legislation is ultra vires; it is only a factor to be taken into consideration when determining whether that subsidiary legislation falls strictly within the powers conferred by the enabling act or Ordinance from which such subsidiary legislation derives its existence.

73.  Mr Tam also drew my attention to the UK Supreme Court decision in AXA General Insurance Ltd v HM Advocate [2011] UKSC 46. There, the appellants challenged the lawfulness of an Act made by the Scottish Parliament on the basis that (a) the Act is incompatible with the European Convention on Human Rights and (b) the Act is open to judicial review as an unreasonable, irrational and arbitrary exercise of the legislative authority of the Scottish Parliament. In respect of the second ground, the UK Supreme Court took the view that Acts of the Scottish Parliament – which takes its place as a self-standing legislature in the constitutional arrangements in the UK – are subject to judicial review but not on the grounds of irrationality, unreasonableness or arbitrariness: see §§42-52 and 148. That case is plainly set in a wholly different context and does not lend assistance to Mr Tam. Nor do I see how §152 (which was specifically referred to by Mr Tam) assists the Applicant. The tenor of that paragraph is simply that Parliament must be presumed not to legislate contrary to the rule of law.

74.  The position in the UK is perhaps different from, or at least not as clear as, that in Hong Kong. In Nottinghamshire County Council v Secretary of State for Environment [1986] 1 AC 240, the House of Lords held that subsidiary legislation made by a local authority (and vetted by Parliament) which is within the scope of the enabling statute is not open to judicial review on the ground of unreasonableness unless the authority making the subsidiary legislation could be shown to have acted with bad faith, or for an improper motive, or that the consequences of the subsidiary legislation were so absurd that the authority must have taken leave of his senses: see 247G-H. This formulation was latter affirmed by the House of Lords in R v Secretary of State for the Environment Ex parte Hammersmith and Fulham London Borough Council [1991] 1 AC 521, replacing the phrase “taken leave of one’s senses” with “manifest absurdity”: see 597E-H.

75.  In a more recent case in O’ Connor v Chief Adjudication Officer [1999] 1 FLR 1200, the English Court of Appeal said that it is too narrow a view to say that courts could only find a subsidiary legislation invalid by reason of irrationality if the government has misled or deceived Parliament or has otherwise acted in bad faith. The Court affirmed the position that irrationality is a separate ground for challenging subsidiary legislation, and that means irrationality in the conventional Wednesbury sense and there is no need to refer to any notion of “extreme irrationality”: see at 1210. This view was endorsed by another bench of the English Court of Appeal in R (Asif Javed) v Secretary of State for Home Department [2002] QB 129, at 152E to 153C.

76.  But I am satisfied that the Hong Kong position is clear.

77.  Mr Lok referred me to the CFA’s decision in Noise Control Authority v Step In Ltd (2005) 8 HKCFAR 113. There, the validity of a Noise Abatement Notice (“NAN”) issued by the authority pursuant to power granted by section 13 of the Noise Control Ordinance Cap 400 was challenged on the ground, among others, that the NAN was ultra vires section 13 as the requirement stated on the NAN was too uncertain or unreasonable. The NAN does not have the status of subsidiary legislation but the CFA was apparently of the view that the NAN – an order issued by the authority pursuant to powers granted by primary legislation – was analogous to subsidiary legislation in the context. The CFA reviewed the law on the issue of whether a by-law could be invalidated by uncertainty or unreasonableness.

78.  The leading judgment was given by Sir Gerard Brennan NPJ. He noted that in earlier times the prevailing position was that a by-law must be certain and reasonable for it to be valid. But the law has gradually developed such that now reasonableness of the by-law is subsidiary to be question of whether the by-law was made within the powers of the enabling statute as intended by the legislature: see §§29-40. At §§35 and 40, he said:

35. Diplock LJ in Mixnam’s Properaties Ltd v Chertsey Urban District Council [1964] 1 QB 214 was surely correct in referring the question of validity of an impugned by-law, order or condition to the scope of statutory power to make or impose it. If the impugned instrument falls within the scope of the power to make or impose it, it is valid; if it falls outside the scope of the power, it is invalid…

40. The cases show that challenges to statutory instruments on the grounds of uncertainty and irrationality depend upon the scope of the power to make the instruments, not on a judicial policy or free-standing common law rule. Provided that the impugned instrument is not so uncertain in its terms as to be meaningless, uncertainty or unreasonableness in application or inconsistency with the general law are relevant to validity only insofar as they assist in answering the question whether the instrument falls within or without the power to make it.

79.  Therefore, it seems the position in Hong Kong is clear: subsidiary legislation enacted within the scope of power of the enabling statute is not subject to judicial review on the ground of irrationality.

80.  In the course of the oral hearing, Mr Tam did not seriously dispute the legal position put forward by Mr Lok. But, Mr Tam submitted that unreasonableness is ‘part and parcel’ of the ultra vires doctrine. In reply, Mr Lok said that if Mr Tam had intended unreasonableness as a subset of ultra vires, he should have clearly pleaded that in the application but he had failed to do so. For myself, I do not think Mr Tam really made clear what he meant by the submission that unreasonableness is part and parcel of the ultra vires doctrine, either in the pleadings or in oral submissions. If he meant to say that the 115B Order is outside the scope of power of its enabling statute (namely section 35(1) of the IO) on the basis that no legislature could have intended to enable the making of subsidiary legislation which would operate to produce unequal treatment between TGCI and CIC detainees, I do not think that argument was sufficiently developed, if at all.

81.  Construing the scope of powers of the enabling statute would be key to an ultra vires argument. No attempt was made to construe the scope of section 35 of the IO or, for obvious reason, section 4 of the PO, being the enabling provision of Cap 234B which sets aside TGCI for the purpose of prisons.

82.  In any event, I am satisfied that for reasons set out in section H below, the detention treatment in TGCI and in CIC do not violate the principle of equality. There is no unreasonableness in the legislative scheme which could be relied upon by Mr Tam to say that the scheme is ultra vires section 35(1) of the IO or section 4 of the PO, if he had indeed raised such an argument.

I.  Does the legislative scheme produce unlawful inequality?

83.  Strictly speaking, because of the forgoing, it is unnecessary for me to deal with Mr Lok’s second argument. As I understand it, Mr Lok’s submissions was that the system under attack – being Paragraph 3 of the 115B Order – does not necessarily produce differential treatment. This is because the legislative scheme has in-built statutory flexibility to suit the needs of treating immigration detainees differently from other persons in the custody of a prison. He submitted that the existence of any actual difference (which does not arise as a matter systemic design) does not mean that the system violates the equality principle.

84.  He pointed to Rule 191 of the Prison Rules as providing for such statutory flexibility. It provides:

The Superintendent may modify the routine of the prison in regard to any class of prisoners awaiting trial, so far as to dispense with any practice which, in the Superintendent’s opinion, is clearly unnecessary in the case of that particular class of prisoners.

85.  “Routine” (or “例行程序” in the corresponding Chinese text) is not defined in the PO or the Prison Rules. But it is at least doubtful that it would cover all aspects governed by the 176 general rules in Part I or the other rules in the Prison Rules relating specifically to prisoners awaiting trial. It would be unnatural to describe, for example, rules governing prisoners’ entitlements to food, clothing, bedding, medical assistance, the right to write and receive letters or to be visited by friends and family, or the right to appeal against a punishment imposed by CSD as “routine”. Rather, it appears to me that “routine” in Rule 191 probably entails what the ordinary meaning of the word would suggest – the daily schedule of prisoners such as when to wake up or to go to bed or some other regular procedures they have to go through while detained.

86.  Even if “routine” is widely construed to include all aspects governed by the Prison Rules, Mr Lok did not explain and I do not see how CSD can, invoking its suggested Rule 191 power, flexibly adjust treatment in TGCI so that there would be no violation of the principle of equality, assuming that what he meant is the treatment of the two institutions would be aligned. Some rules are meant to benefit prisoners. For example, currently, TGCI detainees are entitled to appeal against punishment imposed on them under Rule 63 of the Prison Rules, and there is no comparable appeal mechanism in CIC. I do not think it can be argued that the CSD is empowered by Rule 191 to remove TGCI detainees’ appeal entitlement in order to bring them in line with CIC detainees.

87.  For rules which could be properly viewed as conferring intrusive powers on CSD rather than benefitting or giving entitlements to prisoners (for example, urine examination or body cavity search), the CSD would be able to refrain from exercising such powers even in the absence of any express provision to do so. But the fact that the CSD has power to refrain from exercise of powers would not be a sufficient answer to a systemic challenge that the legislative scheme has produced or is capable of producing unequal treatment on the two institutions.

88.  It was also referred to in the affidavit evidence of the Secretary/Director that the phrase “mutatus mutandis” in Paragraph 3 of the 115B Order also provides flexibility to consider if any part of Rules 189 to 207 shall not be applied to immigration detainees in TGCI. This seems to be a stretched reading of the phrase “mutatus mutandis” in context. Anyway, the bulk of the rules applicable to TGCI detainees are to be found in the 176 general rules.

89.  I am not persuaded that the 115B Order (or together with the 115E Order and Cap 234B) has sufficient in-built statutory flexibility, by virtue of Rule 191 or otherwise, to pre-empt an argument that the legislative scheme as a system would produce differential treatment between TGCI detainees and CIC detainees. So, I am not persuaded by Mr Lok’s second argument.

J.  Is there unjustified differential treatment?

90.  As to his third argument in respect of the 115B Order – that in any event any difference in treatment, if it exists, could be properly justified, see sections L and M below. The argument is advanced with regard to both targets of challenge. As said, amongst the 8 areas of differential treatment relied upon by Mr Tam, only those arising from powers granted by the Prison Rules are relevant to the challenge to the legislative scheme. These include body cavity search, urine examination, the stopping and reading of letters, disciplinary action and removal from association. On the other hand, the Detention Decision – being an administrative decision – is potentially capable of covering all 8 areas of factual difference. But. whilst the actual differential treatment relevant to the two targets of challenge differ, the analysis is substantially the same. Therefore, though the next section is titled by reference to the Detention Decision, it is equally applicable to the 115B Order/the legislative scheme.

K.  The Challenge to the Detention Decision on the Unequal Treatment Ground

91.  Mr Lok said that Mr Tam only added the Detention Decision as a target of challenge as a last minute ‘patch’ (not his exact word) to salvage the situation that the 115B Order could not be invalided on ground of irrationality. This is because the Detention Decision, being an administrative decision, would present no jurisdiction hurdle if the Court were to quash it on the ground of irrationality. But Mr Tam’s submissions on the Unequal Treatment Ground, whether in writing or orally, made no distinction between the 115B Order and the Detention Decision. Whilst Mr Tam did not expressly characterise the challenge to the Detention Decision as a systemic challenge, it is clear that in his submission on both targets, he intended both challenges as systemic ones.

92.  As identified in QT, at §§34-35, there are two main issues in unlawful discrimination cases: (1) first, whether there is discriminatory treatment at all, and (2) if so, whether the discriminatory treatment can be justified.

L.  Whether there is discriminatory treatment at all? Are the two groups comparable?

93.  It is helpful to start by recognising that the notion of whether comparators are analogous or relevantly similar is elastic both linguistically and conceptually. Therefore, it is generally not profitable to debate in the abstract whether a given comparator is always not sufficiently analogous to require like treatment. The context of the question is crucial: see QT at §45.

94.  The first question turns on whether the target challenged falls within the three recognized categories of discrimination. In short, they are where (1) like is not being treated as like in that the complainant is receiving treatment which is unfavourable when compared with treatment given to persons in “relevantly similar situations”; (2) a complainant disadvantageously receives the same treatment as persons in significantly different situations; and (3) the application of an ostensibly neutral criterion operates to the significant prejudice of a particular group: see §32 of QT.

95.  Mr Tam clearly relied on the first category. He argued that TGCI detainees and CIC detainees are alike, but the former received less favourable treatment than the latter. Detainees in the two institutions plainly receive different treatment. The battleground is whether they are “like” such they are entitled to like treatment. In other words, are the two groups proper comparators.

96.  Mr Lok argued that TGCI detainees and CIC detainees are not in a comparable or analogous situation such that like treatment is not required. The two groups of detainees are different, in that TCGI detainees are assessed by the Director to be posing a higher security risk. He referred to the policy intent to concentrate detainees with higher security risks in TGCI, and the information concerning the criminal records of TGCI detainees as mentioned above. In gist, by towards the end of 2022, all TGCI detainees were convicted criminals discharged from prison, and some of the them have been sentenced for more than 10 years for committing serious crimes, such as murder and rape.

97.  Mr Lok said simply because both TGCI detainees and CIC detainees are both immigration detainees pending removal, it cannot simply be said that the two groups are relevantly similar. He relied on Re Morrison and another’s Application [1998] NI 68 for the proposition that prisoners or detainees with different security classification should not be considered as comparators, and the different classification would warrant a difference in their detention conditions. In the particular case, prisoners held at one prison complained that the prisoners held at another prison received a more favourable arrangement concerning the receipt of parcels. At 72h to 73b, Kerr J asked and answered the following question:

Are the groups similar?

For the applicants it was claimed that no material distinction could be drawn between the two groups of prisoners. The applicants had the same or lower security classification than members of the group with which they sought to compare. There was no basis on which the groups could be treated as other than similar that they must therefore be accorded similar treatment.

It was implicit in the applicants’ argument that difference in security classification could warrant a difference in the conditions in which the prisoners are held. It could not be otherwise. But difference in security classification cannot, in my opinion, be that only circumstances in which the two groups of prisoners may be treated differently. If a group within a prison elects to associate together by reason, for instance, of political belief, it would be unrealistic to suggest that it was entitled to be treated in exactly the same way as the rest of the prison population in all circumstances. I do not consider, therefore, that the prison service is obliged to regard each group of prisoners in all the prions in Northern Ireland as similar solely because they are all prison inmates.

98.  Mr Lok also referred to R (on the application of Rangwani) v Secretary of State for Home Department [2011] EWHC 516, at §§54-58, to make good the point that the authority is to entitled to take account of a detainee’s security risk profile in determining the detention conditions imposed. There, the authority imposed a policy in which immigration detainees who have committed serious sexual offence would usually be held in prison rather than an Immigration Removal Centre (“IRC”). It was undisputed that the IRC regime offered a much more favourable or less stringent regime than the prison. The policy was challenged on the basis that there could be male only IRC where the particular risks – risks to women and children – could be controlled. The argument was rejected as the policy already has an inbuilt flexibility to allow assessment of each detainee on an individual basis.

99.  Mr Lok argued that the difference in security risk profile is a relevant distinction in the 8 areas of differential treatment highlighted by the Applicant. As set out in the Secretary/Director’s affidavit evidence:

(1)  Body cavity search and urine examination is necessitated by the increasing prevalence of drug offences amongst non-refoulement claimants. From 2013 to 2022, among the non-ethnically Chinese illegal immigrants arrested, over 13% were arrested for drug offences. Since the recommissioning of TGCI, body cavity search by way of X-Ray has once led to the discovery of a denture in the abdomen of a detainee. There was one successful detection of dangerous drugs in TGCI so far, and the Applicant was one of the detainees found to have dangerous drugs in his urine.

(2)  Recording of conversation over phone calls or during social visits assists the detection of illicit activities, preventing escape and also helping to detect any suicidal intention of the detainee. TGCI once detected a piece of crime-related intelligence which was passed to the Police leading to the arrest of a fugitive.

(3)  The reading and stopping of letters serves to enhance CSD’s management and to detect any lurking plot of unlawful activities.

(4)  The more categories of disciplinary offences, heavier punishment, and longer period of removal from social association are called for because the detainees in TGCI are of higher security risk with more serious criminal records. Tighter discipline is needed to maintain the order of TGCI.

100.  I agree with Mr Lok that TGCI detainees are not comparable with, or not in a relevantly similar situation to, CIC detainees in the context of the 8 areas of actual differential treatment now relied upon by Mr Tam.

101.  First, it is not very clear to me whether Mr Tam’s position is that all persons held in immigration detention pending their removal in Hong Kong must be held in the same institution or be subject to the same set of detention rules. In any event, if Mr Tam’s position is that immigration detainees must not be distinguished from each other or be subject to different detention rules by reference to their security risk profile, it must be wrong. What was said in Re Morrison (quoted above) accords with logic and common sense.

102.  Mr Tam said Re Morrison must be read together with the CFA decision in Leung Kwok Hung (Long Hair) v Commissioner of Correctional Services (2020) 23 HKCFAR 456, in which the applicant successfully challenged the differential hair cutting requirements between male and female prisoners. If Long Hair was intended by Mr Tam to rebut the proposition that prisoners need not necessarily be treated as if in similar position or that a difference in security risk profiles should not render two groups of detainees incomparable, I do not see Long Hair as a decision to such an effect. He also questioned the relevance of Rangwani as the present case does not concern risks to women or children in TGCI which might be posed by detainees. But, risks to women and children are obviously a possible facet of, or might come under the umbrella of, security risk.

103.  Mr Tam referred to In re Tran Quoc Cuong and Khuc The Loc [1991] 2 HKLR 312 for the proposition that “it is a requirement of the authorities to apply the detention centres rules to immigration detainees”. If what Mr Tam meant was that all immigration detainees, regardless of the institution in which they are held, must be subject to the same set of rules or that the Prisons Rules cannot be applied to them, In re Tran Quoc Cuong does not establish such a preposition.

104.  In that case, two Vietnamese refugees, who were suspected troublemakers, were transferred from one immigration detention centre to another immigration detention centre located inside the Stanley Prison and with a higher security setting. The refugees challenged the decision to designate Stanley Prison as a detention centre for immigration detainees under the then section 13D of the IO. At 335H-I, Jones J held that:

In my judgment, the Secretary for Security had power under s.13D to designate any place as a detention centre irrespective of the fact that the place was also used for another purpose such as a prison. Stanley Prison was both designated as a prison and as a detention centre and it was a matter for the authorities to determine which part should be used as prison and which part as a detention centre. It was not necessary to publish in the Gazette the discontinuance of a particular part of the prison which was to be used as a detention centre. The only requirement of the authorities was to ensure that arrangements were made, as they were in this case, to keep the Vietnamese boat people physically separate from the prisoners and to apply to them the Detention Centres Rules which would provide no difficulty for the Correctional Services Department.

105.  The last sentence was what was quoted and relied upon by Mr Tam. But Mr Lok identified that the phrase “Detention Centres Rules” is a term previously defined in the judgment, at 326, and it refers to the “Immigration (Vietnamese Boat People) (Detention Centres) Rules”. Those rules were amended to become the “Immigration (Vietnamese Migrants) (Detention Rules)” in 1992 and are currently to be found in Cap 115M. The point made by Jones J was only that the detention rules as set out in the subsidiary legislation under the IO as applicable to a particular type of detention centre should be applied, regardless of the physical location of the immigration detention centre. This cannot be elevated to a principle that, say, the rules applicable to CIC should be uniformly applied to all immigration detainees, even though it is not what is required under the currently applicable legislative scheme. It can also be noted that, at 338F, Jones J expressed satisfaction that a transfer between detention centres could be made for the purposes of good management and in the interests of security.

106.  Nor can the case of R (SM) v Lord Chancellor [2021] 1 WLR 3815 advance Mr Tam’s case that TCGI detainees and CIC detainees are comparable. There, immigration detainees held in an IRC were provided with free 30-minute legal advice and such service was not available for detainees held in prisons. It was decided that the two groups were in an analogous situation and the differential treatment was not justified. In that case, the comparability question was assessed in the context of the provision of free legal advice, which is different from the eight areas of differential treatment relied upon by the Applicant in the current case. Access to free legal advice is plainly different from security risk concerns.

107.  In any event, Mr Tam did not seriously dispute that detainees’ difference in security profiles could render them incomparable in the 8 areas of differential treatments. I believe implicit in his submissions is a recognition that if, as a matter of fact, the two groups of detainees truly differ in terms of security risk profile, they could be subject to different detention treatment. This is crucial. It means that Mr Tam’s remaining argument on whether TGCI detainees and CIC detainees are comparable could only be a factual one.

108.  Indeed, Mr Tam did raise a factual argument, namely that the criminal backgrounds of TGCI detainees and CIC detainees are not as different as the Secretary/Director would like to portray:

(1)  He referred to the what the Secretary said in the 2022 LegCo Paper about CIC detainees’ criminal background – in 2022, the percentage of newly admitted detainees who were discharged prisoners stands at 62%. Around 25% of them had committed serious crimes such as murder, trafficking in dangerous drugs, robbery, burglary, wounding, blackmailing and rape, etc. Some of them are repeated offenders.

(2)  Mr Tam said the Applicant committed robbery and handling of stolen goods, which could not be more serious crimes than murder.

(3)  Mr Tam said the statistics that over 13% of the arrested non-ethnically Chinese illegal immigrants were arrested for drug offences are equally relevant to both TGCI and CIC and could not be relied upon by Mr Lok to justify the differential treatments.

109.  As Mr Tam has accepted, the burden of proving that TGCI detainees and CIC detainees are comparable lies on the Applicant. I am not persuaded by Mr Tam that the two groups are comparable in terms of security risk profile:

(1)  Whilst the Court is not provided with a complete statistical comparison on the security risk profiles between the two institutions, at least it is known that as of the end of 2022, 100% of the TGCI detainees were discharged prisoners while the figure for CIC stood at 62%. Mr Tam cannot say that there is no difference. The Secretary/Director has also referred to the higher prevalence of indiscipline cases in TGCI.

(2)  In the Secretary/Director’s affidavit evidence, it is also said that the Government would take into account more than the nature of the offences committed and the length of imprisonment term in assessment of a detainee’s security risk.

(3)  Further, the distinguishing feature relied upon by the authority in the present case – security risk profile – is a matter traditionally recognized by the Court as being well within the Secretary/Director’s expertise: see Simona Mundai v Director of Immigration [2020] 2 HKLRD 1205 at §29. The Court has to give sufficient weight to the Secretary/Director’s assessment of detainees’ security risk. The position will be different if the distinguishing feature is, as in the QT case, marital status or sexual orientation which matters the authority will be in no better position than the applicants in ascertaining the factual position.

(4)  It is the Secretary/Director’s affidavit evidence that there was a policy intent to centralize to TGCI those detainees posing a higher security risk. The existence of that policy was unchallenged. The affidavit evidence also refers to how ImmD would decide who should be accorded priority to be transferred to TGCI under the policy. The formulation of the policy was not challenged, either on ground of reasonableness or certainty. There was also no challenge that the policy was not adhered to in its general execution or in relation to the Applicant’s specific case.

(5)  In view of the forgoing, the information unearthed from the 2022 LegCo Paper by Mr Tam is not sufficient to justify a factual conclusion that the security profiles between TGCI detainees and CIC detainees are relevantly similar.

110.  For completeness, I do not attach much weight to Mr Tam’s submission that the Applicant himself was only convicted of robbery and handling stolen goods whilst some detainees in CIC were convicted of more serious crimes such as murder. Mr Tam did not object that detainees should be detained in groups and that uniform treatment could be applied across the whole group despite that there might be individual variation among members of the group. Indeed, Mr Tam has always advanced this judicial review on the systemic challenge footing by comparing TGCI detainees as a group against CIC detainees as another group. There is little value in singling out an individual in one group in making the comparison. Further, as said, the nature of the crimes committed is only but one factor in assessing a detainee’s security risk profile.

111.  Third, Mr Tam referred to the Secretary/Director’s affidavit evidence that they tried to align the employment option in both institutions. He submitted that such alignment efforts must reflect a recognition on the part of the authority that the two groups of detainees are comparable. I do not agree. The alignment of employment option is neither here nor there. This particular alignment may indicate that the Secretary/Director recognized detainees’ comparability in the context of employment. This is different from the comparability in terms of security risk or related treatment.

112.  I also note that, albeit not referred to by Mr Tam, besides employment, the Secretary/Director is also actively planning to align treatments in urine examination, body cavity search and maximum length of separate confinement by amending the 115E Order in the first half of 2023. The ground relied upon is the increasing prevalence of criminal records and worsening discipline in CIC over the years, which calls for greater power to uphold discipline an order in CIC. Unlike employment, these alignments plainly concern security risk.

113.  However, I think what the alignments indicate is that the security risk concerns in both institutions have reached the threshold where the Secretary/Director now considers that greater powers – similar to those conferred by the Prison Rules – would be needed for both institutions. But the current situation remains that those powers are only available in managing TGCI. In that light, it does not detract from the point that TGCI should be used to hold those with a perceived higher security risk, even though the Government’s view is that those held in CIC should also be subject to more stringent management as soon as possible.

114.  In view of the forgoing, I am of the view that TGCI detainees and CIC detainees are not comparable in the present context.

115.  As an aside, I would also note the following. Sometimes, the Court recognizes that in unlawful discrimination cases, the comparability question and the justification question cannot be neatly compartmentalized. The difficulty arises when, with regard to the comparability question, it is disputed whether the distinguishing feature relied upon by the authority is a relevant difference. An analysis of the relevance of the impugned distinction will inevitably have to be made by reference to whether such difference should justify differential treatment – thus running into the second question. In QT, it was much debated whether, in the context of dependent visa application, marital status is a relevant difference such that Government is entitled to treat homosexual couples in a civil partnership differently from married couples. The concern is one of circularity of reasoning.

116.  But in the present case, Mr Tam did not dispute the relevance of security risk profile in assessing the comparability of the two groups, and it follows that the two questions can be more easily compartmentalized. I do not think there is the same risk of circularity. If the Court is of the view that TGCI detainees and CIC detainees are not comparators, the second question on the justification of differential treatment (if it exists) does not arise. This would be the end of the matter.

117.  But for completeness, I shall consider the justification question.

M.  Justification

118.  In QT, at §§84-87, the CFA said that the proportionality concepts developed for scrutinising incursions made into constitutionally protected rights are equally applicable in deciding whether differential treatment could be justified in the judicial review context. The Court would have to consider whether the differential treatment (a) pursues a legitimate aim; (b) is rationally connected to the legitimate aim; (c) is no more than is necessary to accomplish the legitimate aim; and (d) strikes a reasonable balance between societal benefits of the encroachment on the one hand, and the inroads made into the constitutionally protected rights of the individual on the other.

119.  As for legitimate aim, Mr Lok said the more stringent measures to TGCI detainees were to safeguard the security of the institution, the safety of the officers and other detainees as well as the general public. Whilst he initially also relied on the institutional design of TGCI as a legitimate aim – relating to the initiative to develop it as a “smart prison” – he has not focused on the latter in the oral hearing. I shall only focus on the former.

120.  It will not be surprising that the first two steps substantially overlap with the analysis on the comparability question. Mr Tam did not dispute or could not dispute that treating detainees with a higher security risk adopting stricter rules or measures, and the contrary, is a legitimate aim. Nor did he offer any analysis on why the 8 areas of differential treatment are not rationally connected to that aim. Rather, his argument in the rational connection step is that the security risk profile between TGCI detainees and CIC detainees are “similar” as reflected in their criminal records. I have already dealt with this factual challenge above.

121.  As to the third and the fourth steps, I am of the view that in any event the less favourable treatment to TGCI detainees is not disproportionate and that it does strike a proper balance between the accompanying societal benefit and the incursion into detainees’ privacy or liberty:

(1)  Whilst Mr Tam said the detention treatment in TGCI amounts to intrusions into detainees’ privacy and liberty which are not trivial, there is no legal challenge to the treatment of TGCI detainees per se. If the treatment in TGCI is, on its own, not challenged as objectionable, the argument that any difference in treatment in comparison with CIC is objectionable will be harder to make.

(2)  The differences in treatment relating in telephone calls, letters and social visits are not substantial. They are subject to monitoring in both TGCI and CIC.

(3)  For CCTV arrangements, the difference turns on the coverage of CCTV rather than its existence. Where the CCTV system in TGCI also blurs the private parts of detainees, I do not think the difference is substantial.

(4)  Disciplinary action and punishments are undoubtedly potentially stricter in TGCI. The difference in the maximum period of separate confinement – 7 days as opposed to 28 days – may be substantial. But, this is to some extent alleviated by the existence of a statutory appeal mechanism in TGCI to safeguard against arbitrary punishment.

(5)  Body cavity search is conducted by X-Ray and thus the intrusion into bodily integrity has been minimized. For urine examination, it does not seem to me to be a substantial inroad into a detainee’s privacy or liberty in the detention setting.

(6)  Based on the Government’s evidence, the current inconsistency between treatment of TGCI detainees and of CIC detainees is temporary. The Government will soon amend the 115E Order to confer on ImmD the powers relating to body cavity search, urine examination and disciplinary actions and punishment in the first half of 2023.

122.  I have reached the above views bearing in mind that the proper standard of review should be the higher threshold of manifestly without reasonable foundation. The treatment of immigration detainees in TGCI falls within the Secretary/Director’s wide imperative of immigration controls and the designation of TGCI as a prison facility under the management of CSD obviously also involve man-power consideration as the historical background has reflected.

N.  Result

123.  For those reasons, I dismissed the application for leave to apply for judicial review.

124.  As to costs, in the exercise of my broad discretion on costs – where the Applicant was refused Legal Aid, and was represented by solicitors and counsel on a pro bono basis, and where this was the first occasion that the Court has closely examined any differential treatment between TGCI and CIC – it seems to me that the appropriate costs order to make is no order as to costs.

O.  Postscript

125.  In my first Reasons for Decision dated 31 January 2023 [2023] HKCFI 174, I explained why it was necessary to provide my reasoning for rejecting the Applicant’s habeas corpus application separately from the reasoning relating to the rejection of the judicial review application.

126.  In short, it was because the Applicant – this time acting in person – had made a further or renewed application dated 27 January 2023 seeking a writ of habeas corpus, which required being dealt with as a matter of urgency for reasons which included that he was due to be removed from Hong Kong on 1 February 2023 (a factor I had specifically noted and taken into account in my earlier decision just 10 days beforehand).

127.  The further habeas corpus application was made in HCAL 111/2023, and was heard by Campbell-Moffat J on 31 January 2023. The application was dismissed on the same date, with reasons provided on 1 February 2023 [2023] HKCFI 314. In summary (see §21), Campbell-Moffat J held that there was no reasonably arguable case for revisiting my very recent earlier decision at and following the hearing on 17 January 2023, and that the Applicant’s continued detention remained lawful. Indeed, she also stated that the interests of justice required the continued detention of the Applicant until such time as he could be repatriated to Pakistan (which was then considered imminent). She also made a summarily assessed costs order against the Applicant.

128.  However, for reasons which are unknown to the Court, it seems that the Applicant was not removed from Hong Kong when expected. This is because the Applicant wrote to the Court by letter dated 26 February 2023 asking for an extension of time for him to file a notice of appeal against my decision. I will deal, however, with that request separately.