DM, R (on the application of) The Secretary of State for The Home Department

This case is about child refugees – that is, children under the age of 18 who have been granted asylum or humanitarian protection (Footnote: ¹) – who came to this country on their own and whose parents and/or siblings wish to join them here. Applications by family members for leave to enter the UK to join a refugee are described as applications for “family reunion”. Although the application is made by the family members, I will generally refer to family reunion as being “for” the child, who is their “sponsor”.

It is the Claimant’s case that the Secretary of State’s policy governing family reunion for child refugees is unlawful. I will have to analyse that policy in some detail later; but the essence is that a child refugee can sponsor their parents or siblings only where a refusal to admit them would lead to “unjustifiably harsh consequences”, which typically would constitute a breach of the right to respect for family life under article 8 of the European Convention on Human Rights (“the ECHR”). The Claimant says that that policy is too restrictive: he points out that the “unjustifiably harsh consequences” test is not applied to applications by the partner or minor children of an adult refugee, who can (broadly speaking) join them in the UK automatically.

For reasons which will appear, we are not concerned with the facts of the Claimant’s own case, save to the extent that they afford an illustration of the situation with which we are concerned, and I need not set them out in detail. He is a national of Eritrea, born on 10 October 2002. As a result of his fear of enforced conscription (which in Eritrea is notoriously abusive), he left home at the age of 11, and after spending time in Ethiopia, Sudan, Libya, Italy and France he arrived in the UK, concealed in a lorry, on 7 May 2017, aged 14. He claimed asylum, and in November 2018, when he was 16, he was granted refugee status. His parents and his five younger siblings (“the family”) initially remained in Eritrea, but some time in 2018 they moved to Ethiopia because of a fear that the other children would be conscripted. On 8 September 2020 they applied to the Entry Clearance Officer (“the ECO”) in Addis Ababa for leave to enter the UK on the basis of family reunion.

On 30 September 2020 the Claimant commenced the present proceedings challenging the lawfulness of the Secretary of State’s policy as regards family reunion for child refugees. That was only three weeks after the family’s application, and before any decision had been taken on it; but in due course the application was refused, and the claim form was amended so as to include a challenge to that refusal.

The claim came on for hearing before Lavender J on 15 and 16 June 2022. Some three months previously the First-tier Tribunal had in fact allowed an appeal by the family, and although they had not yet been granted leave to enter it was clear that that would occur shortly (as it in fact did on 7 July). The challenge to the refusal of family reunion in the Claimant’s own case accordingly became academic. However, the originally pleaded challenge to the Secretary of State’s policy was allowed to proceed: Lavender J rejected a submission that the Claimant had no standing to pursue such a challenge. Shortly before the hearing the United Nations High Commissioner for Refugees (“UNHCR” (Footnote: ²)) was given permission to intervene.

Before Lavender J the Claimant advanced three grounds of challenge, each of which remains live before us. I give the details later, but in headline terms they are:

As will appear, the parties’ cases on the section 55 ground were substantially clarified in the course of the hearing before Lavender J, but that clarification led to a series of post-hearing written submissions (and some further evidence) which were not concluded until the end of November 2022.

Lavender J handed down judgment on 31 March 2023 ([2023] EWHC 740 (Admin), [2023] 1 WLR 4109). He dismissed the section 55 and discrimination grounds, but he felt obliged to adjourn the rationality ground in order to give the Claimant the opportunity to consider issues emerging from the post-hearing submissions. During that adjournment still further evidence and submissions were filed.

The adjourned hearing took place on 17 January 2024. By a judgment handed down on 26 April ([2024] EWHC 967 (Admin), [2024] 1 WLR 5309) Lavender J also dismissed the rationality ground.

It is obviously highly unsatisfactory that the Judge was only enabled to reach a decision after the complex process outlined above; and, as will appear, there have been further developments in the case as presented to this Court. I will say more about this at the end of my judgment.

This is an appeal, with permission granted by Lewis LJ, against Lavender J’s decisions in both judgments. Mr Raza Husain KC, Mr Jason Pobjoy KC and Ms Eleanor Mitchell have appeared for the Claimant, and Ms Lisa Giovannetti KC and Mr Jack Anderson have appeared for the Secretary of State. UNHCR has again been given permission to intervene and has been represented by Ms Sonali Naik KC, Mr Ali Bandegani and Ms Rebecca Chapman. All counsel also appeared below. For convenience I will refer to the skeleton arguments and written submissions as if they were solely the work of the leaders, although I am sure that that is far from being the case.

The recent suspension of family reunion. On 4 September 2025 the Secretary of State announced that the Government aims next spring to introduce changes to “the current rules for family reunion for refugees” and that in the meantime she was suspending their operation for new applications (although she drew attention to the availability of other routes for family members which might be available under the Immigration Rules). That is not directly relevant to the substantive issues raised by this appeal, which must be concerned with the position at the time of the judgments appealed from (though it is potentially relevant to the question of relief); and to avoid complication in the drafting I will refer to the policy as if it were still in force.

Anonymity. When he granted permission to appeal, Lewis LJ extended the anonymity order which had been granted in the High Court, but he suggested that the Court which heard the appeal might wish to consider whether it should continue thereafter, given that not only the Claimant but also his family were no longer at any risk of persecution in Eritrea. He rightly emphasised that the principle of open justice normally requires that the parties to proceedings should be identified and that a powerful justification was required for any departure from that principle. We asked for further submissions, which were provided in writing together with a psychiatric report on the Claimant’s mental health, following the hearing. Not without hesitation, I am persuaded that in the light of those submissions it would be wrong to withdraw the anonymity which has been accorded to the Claimant thus far. One factor in this conclusion, though it would not be decisive on its own, is that the issues which we have to decide are not in truth specific to his circumstances. As explained above, following the admission of his family to the UK his claim has become the vehicle of a general challenge to the Secretary of State’s policy as regards family reunion for child refugees, which in other circumstances might have been expected to be brought by a charity or other interested body and to have involved no reference to the details of his, or any other individual’s, case.

In this judgment I will start by identifying the components of the impugned policy. I will then consider the three grounds of appeal, which correspond to the three bases of challenge identified above.

Neither the 1951 Refugee Convention nor the 1967 Protocol provides for refugees, whether adults or children, to have a right to family reunion. The Final Act of the Conference which produced the 1951 Convention contained only a recommendation that governments

“take the necessary measures for the protection of the refugee’s family, especially with a view to:

(1)

ensuring that the unity of the refugee's family is maintained particularly in cases where the head of the family has fulfilled the necessary conditions for admission to a particular country;

(2)

the protection of refugees who are minors, in particular unaccompanied children and girls, with special reference to guardianship and adoption.”

In July 1983 UNHCR published Guidelines on Reunification of Refugee Families (“the UNHCR Guidelines”), stating its support for family reunion and in particular for the reuniting of unaccompanied minor children with their parents and siblings. Para. 5 (a) (iii) reads (so far as material):

“An unaccompanied minor child should be reunited as promptly as possible with his or her parents or guardians as well as with siblings. If the minor has arrived first in a country of asylum, the principle of family unity requires that the minor’s next-of-kin be allowed to join the minor in that country unless it is reasonable under the circumstances for the minor to join them in another country. Because of the special needs of children for a stable family environment, the reunification of unaccompanied minors with their families, whenever this is possible, should be treated as a matter of urgency.”

But UNHCR’s position does not reflect any treaty obligation. As Sales LJ put it in Mosira v Secretary of State for the Home Department [2017] EWCA Civ 407 (see para. 13)

“The Refugee Convention does not impose an obligation on Contracting States to grant leave to enter or leave to remain in order to achieve family reunion with a sponsor who has been granted refugee status in the host state, but the UN Human Rights Committee exhorts Contracting States to do this.”

In the absence of any treaty obligation the UK has not provided for family reunion for refugees, whether adults or children, in any primary legislation. Nevertheless, the Secretary of State has since at least 2000 had a policy, expressed partly in the Immigration Rules and partly in published guidance, about the circumstances in which leave may be granted for the purposes of family reunion. Describing the elements in that policy, and where they are to be found, is not entirely straightforward, and I need to go through the relevant Rules and guidance in some detail.

At the time of the Judge’s decision paragraphs 352A-352F of the Immigration Rules provided for family reunion in the case of the partners and minor children of refugees: paragraphs 352A-352C cover partners, and paragraphs 352D-352F cover children. Those paragraphs (to which I will refer as “the old Rules”) were first introduced with effect from 2 November 2000 and had not been materially altered since then. Family reunion of this kind is of course of marginal significance for child refugees, who will typically wish to be joined in the UK not by a partner or child but by their parents and/or siblings; but it is nevertheless relevant to the issues, and I need to go through the relevant provisions.

In fact, in the interval between Lavender J’s two judgments the provisions relating to family reunion were re-framed (“the new Rules”): see paras. 29-37 below. It is common ground that that did not involve any change in the relevant policy, and since the real significance of this appeal is for the current and future position it might at first sight seem sensible to proceed solely by reference to the new Rules (and the accompanying guidance). However, the old Rules and guidance are a necessary part of the history, as well as being in the relevant respects rather more clearly formulated, and I need to explain them.

As regards both partners and minor children (Footnote: ⁴), I need only set out the paragraph in each case which sets out the substantive requirements for leave to enter or remain. These are as follows:

“Family Reunion Requirements for leave to enter or remain as the partner of a refugee

352A. The requirements to be met by a person seeking leave to enter or remain in the United Kingdom as the partner of a person granted refugee status are that:

(i)

the applicant is the partner of a person who currently has refugee status granted under the Immigration Rules in the United Kingdom; and

(ii)

the marriage or civil partnership did not take place after the person granted refugee status left the country of their former habitual residence in order to seek asylum or the parties have been living together in a relationship akin to marriage or a civil partnership which has subsisted for two years or more before the person granted refugee status left the country of their former habitual residence in order to seek asylum; and

(iii)

the relationship existed before the person granted refugee status left the country of their former habitual residence in order to seek asylum; and

(iv)

the applicant would not be excluded from protection by virtue of paragraph 334(iii) or (iv) of these Rules or Article 1F of the Refugee Convention if they were to seek asylum in their own right; and

(v)

each of the parties intends to live permanently with the other as their partner and the relationship is genuine and subsisting

(vi)

the applicant and their partner must not be within the prohibited degree of relationship; and

(vii)

if seeking leave to enter, the applicant holds a valid United Kingdom entry clearance for entry in this capacity.

…

Requirements for leave to enter or remain as the child of a refugee

352D. The requirements to be met by a person seeking leave to enter or remain in the United Kingdom in order to join or remain with the parent who currently has refugee status are that the applicant:

(i)

is the child of a parent who currently has refugee status granted under the Immigration Rules in the United Kingdom; and

(ii)

is under the age of 18; and

(iii)

is not leading an independent life, is unmarried and is not a civil partner, and has not formed an independent family unit; and

(iv)

was part of the family unit of the person granted asylum at the time that the person granted asylum left the country of their habitual residence in order to seek asylum; and

(v)

the applicant would not be excluded from protection by virtue of paragraph 334 (iii) or (iv) of these Rules or Article 1F of the Refugee Convention if they were to seek asylum in their own right; and

(vi)

if seeking leave to enter, holds a valid United Kingdom entry clearance for entry in this capacity.”

In broad terms, the effect of those rules is that the eligibility for family reunion of a partner or minor child of a refugee depends simply on the fact of their relationship with the sponsor (sub-paragraph (i) in both rules), provided that the relationship pre-dated the sponsor’s departure (paragraph 352A (ii) and (iii) and paragraph 352D (iv)) and is continuing (paragraph 352A (v) and paragraph 352D (iii) and (iv) (Footnote: ⁵)). It is important to appreciate that as regards a minor child eligibility is evidently based on a presumption that a child under 18 is dependent on their parents, in the sense of being entitled to the support – emotional and personal, just as much as material – that typically comes from a child having a home with their parents: that is why paragraph 352D (iii) provides that eligibility is lost where the child has become “independent” in one of the specified ways. (The bright-line cut-off at 18 is artificial, but it is of course applied across many areas of the law.)

That prima facie eligibility may be lost as a result of the public policy considerations identified at paragraph 352A (iv) and (vi) and paragraph 352D (v), but we are not concerned with those here.

The effect of those Rules was described before us as conferring “automatic” eligibility on a refugee’s partner and minor children. I am content to adopt that label, but it should not obscure the fact that, even apart from the public policy conditions, eligibility is not totally automatic: as noted above, minor children will lose eligibility if they have ceased to be dependent on their families.

Those provisions do not themselves distinguish between the positions of adult and child refugees. However, paragraph 277 of the old Rules, which was introduced in order to discourage forced marriages, provided:

“Nothing in these Rules shall be construed as permitting a person to be granted entry clearance, leave to enter, leave to remain or variation of leave as a spouse or civil partner of another if either the applicant or the sponsor will be aged under 18 on the date of arrival in the United Kingdom or (as the case may be) on the date on which the leave to remain or variation of leave would be granted. …”

Thus, while it seems that a child refugee could in principle sponsor a minor child of his or her own (though such cases will surely be very rare), they were unable to sponsor a partner.

As I have said, family reunion with a partner or minor child will rarely be of significance for a child refugee. Typically the family members with whom a child will wish to be reunited will be their parents and/or siblings. As to that, although the old Rules made no provision, the Secretary of State did have a policy governing family reunion outside the Rules which covered, though it was not limited to, the case of child refugees. This was set out in Home Office Guidance entitled Family reunion: for refugees and those with humanitarian protection (“the Family Reunion Guidance”, or simply “the Guidance”). Page 19 of the version of the Guidance which was before the Judge (version 4) contains a section headed Parents and siblings of a child recognised as a refugee. This reads:

“The parents and siblings of a child who have been recognised as refugees are not entitled to family reunion under the Immigration Rules. Where an application does not meet the requirements of the Immigration Rules, the caseworker must consider the ‘Family life (as a partner or parent), private life and exceptional circumstances’ guidance or consider whether there are any compassionate factors which may warrant a grant of leave outside the rules. Each case must be considered on its individual merits and include consideration of the best interests of the child in the UK [emphasis supplied]. As the Immigration Rules are specifically designed to meet our obligations under the European Convention on Human Rights (ECHR) in respect of family or private life, it is not expected there will be significant numbers granted outside the rules. However, it is important that evidence relating to exceptional circumstances is carefully considered on its individual merits.”

The reference to “the best interests of the child” which I have italicised is a summary reference to the requirements of section 55 of the 2009 Act: as to this, see para. 46 below.

The phrases which I have underlined in that paragraph are hyperlinked. The relevant links for our purposes are to “exceptional circumstances” and “compassionate factors”. These appear to lead to the immediately following section in the Guidance (pp. 19-20) (Footnote: ⁶), which is headed Exceptional circumstances or compassionate factors and reads:

“Where a family reunion application does not meet the requirements of the Immigration Rules, caseworkers must consider whether there are any exceptional circumstances or compassionate factors which may justify a grant of leave outside the Immigration Rules. There may be exceptional circumstances raised in the application which make refusal of entry clearance a breach of ECHR Article 8 (the right to respect for family life) because refusal would result in unjustifiably harsh consequences for the applicant or their family. Compassionate factors are, broadly speaking, exceptional circumstances, which might mean that a refusal of leave to remain would result in unjustifiably harsh consequences for the applicant or their family, but not constitute a breach of Article 8 [emphasis supplied]. It is for the applicant to demonstrate as part of their application what the exceptional circumstances or compassionate factors are in their case. Each case must be decided on its individual merits. Entry clearance or a grant of leave outside the Immigration Rules is likely to be appropriate only rarely and consideration should be given to interviewing both the applicant and sponsor where further information is needed to make an informed decision.”

The italicised sentences state the essence of the policy. The passage goes on to give some particular examples but none are material to the present case.

It is the passages from the Family Reunion Guidance quoted at paras. 25 and 26 above which state the Secretary of State’s policy, as at the date of the Judge’s decision, as regards family reunion for child refugees: I will refer to them as “the key passages”. I analyse their effect at paras. 39-45 below.

It seems that the key passages first appeared in the Family Reunion Guidance in July 2016. It does not, however, necessarily follow that prior to that date the Secretary of State had no policy permitting family reunion for child refugees: I do not consider that question here, but it is potentially relevant to the section 55 ground.

The relevant parts of the Immigration Rules were re-drafted with effect from 12 April 2023. Paragraphs 352A-352F were replaced by a new “Appendix Family Reunion (Sponsors with Protection)” (“Appendix FRP”).

Paragraphs FRP 4.1-4.2 and paragraphs FRP 5.1 and 6.1-6.2 of Appendix FRP set out the eligibility requirements for applications by, respectively, partners and minor children of refugees. They involve complicated cross-references, and since they are acknowledged to be to substantially the same effect as paragraphs 352A and 352D of the old Rules I will not set them out here (Footnote: ⁷).

However, Appendix FRP differs from the old Rules inasmuch as it provides for at least some family reunion applications otherwise than by partners or minor children. Paragraphs FRP 7.1-7.2 are headed Eligibility under Article 8 of the Human Rights Convention. Paragraph FRP 7.1reads (so far as relevant):

“Where an applicant does not meet all the suitability or eligibility requirements …, the decision maker must consider, on the basis of the information provided by the applicant, whether there are exceptional circumstances which would make a refusal of the application a breach of Article 8 of the ECHR, because such refusal would result in unjustifiably harsh consequences for the applicant or their family member, whose Article 8 rights it is evident from the information provided would be affected by a decision to refuse the application.”

Read literally, that says only that the decision-maker must “consider whether” refusal of family reunion would be a breach of article 8. But the effect of section 6 of the Human Rights Act 1998 (“the HRA”) is that if that were the case leave to enter would have to be granted.

Like the key passages in the old Guidance, paragraph FRP 7.1 is not concerned specifically with applications for family reunion for a child refugee: it applies to any application by a family member of a refugee other than a partner or child. However, its language plainly covers such applications.

It will be appreciated that the change effected by the introduction of paragraph FRP 7.1 is formal rather than substantive: it brings within the framework of the Rules cases which had previously been dealt with by way of guidance. Although the language is not identical to that of the old Guidance, it continues to refer to whether there are “exceptional circumstances” such that refusal would lead to “unjustifiably harsh consequences” in breach of article 8. But it should be noted that it does not reproduce the full effect of the old Guidance because it does not refer to “compassionate factors”: these remain the subject of guidance rather than Rules (see para. 35 below).

Paragraph 277 was not affected by the changes in April 2023 and remains in force.

The introduction of Appendix FRP was accompanied by a new version of the Family Reunion Guidance (version 11) (Footnote: ⁸) (“the new Guidance”). Guidance on the application of paragraph FRP 7.1 appears at pp. 26-27 in a section headed “Eligibility under Article 8 of the Human Rights Convention”, which has a part within it headed “Unjustifiably harsh consequences”. This is a good deal more elaborate than the equivalent passages in the old Guidance, but no points were taken before us on the details of its drafting: I need only note that it includes specific reference to section 55 of the 2009 Act. That section is followed (at p. 28) by a separate section headed “Compelling compassionate factors”. This corresponds, though the wording is not identical, to the treatment of “compassionate factors” in the section of the old Guidance quoted at para. 26 above. Thus the grant of leave in such cases remains part of the Secretary of State’s policy, albeit not brought within the framework of the Rules.

The upshot is that, although the most relevant provisions now appear in the Rules rather than only in the Guidance, the Secretary of State’s substantive policy as regards family reunion for child refugees is unchanged by the introduction of Appendix FRP. As I have said, that was common ground before us.

It is an oddity that, unlike the old Guidance, the new Guidance contains no express reference to applications for family reunion by the parents and siblings of a child refugee. This does not of course qualify the effect of paragraph FRP 7.1 as explained at para. 31 above, but it does mean that the attention of caseworkers is not specifically drawn to the fact that it can apply in such cases. No point was taken by Mr Husain on this omission, and accordingly we heard no explanation of it from Ms Giovannetti. But it seems desirable that the effect of the Rules as regards child refugees should be explicitly stated in the accompanying Guidance.

Finally, I should record that the suspension referred to at para. 12 above is effected by a Statement of Changes (HC 1298)whichsubstitutes for the previous paragraph FRP 1 a paragraph which begins:

“This Appendix, also known as refugee family reunion, is now closed to new applications pending a review.”

Under both the old Guidance and the new Rules and Guidance, the eligibility requirements in family reunion cases where the applicant is a parent or sibling of the sponsor – including where the sponsor is a minor child – contain four elements: (a) “exceptional circumstances”; (b) the “unjustifiably harsh consequences” of a refusal; (c) breach of article 8 of the ECHR; and (d) “compassionate factors”. The relationship between those elements is not very clearly spelt out, but I believe the correct analysis is as follows.

I start with element (a), “exceptional circumstances”. It is reasonably clear from the language both of the old Guidance and of paragraph FRP 7.1 (and the associated Guidance) that “exceptionality” is not intended as a criterion in its own right. Wherever the phrase “exceptional circumstances” appears, it is accompanied by an explanation of what the circumstances in question are – in short, that refusal of leave would have unjustifiably harsh consequences (whether or not also constituting a breach of article 8) – and it is accordingly the presence of those circumstances which should be regarded as the criterion. The phrase appears to be intended to express the Secretary of State’s expectation that it will in practice be exceptional for the refusal of family reunion to give rise to unduly harsh consequences except in the case of partners and minor children. That expectation is said to be appropriate because the Rules are designed to reflect the requirements of article 8 in the generality of cases: see the penultimate sentence of passage in the old Guidance quoted at para. 25 above. Reference to “exceptional circumstances” in this sense has long been a feature of Home Office drafting in the immigration field, and not only in the context of family reunion: see, for example, paras. 38-44 of the judgment of this Court in MF (Nigeria) v Secretary of State for the Home Department [2013] EWCA Civ 1192, [2014] 1 WLR 544 (referred to by Lord Reed at para. 37-38 of his judgment in Hesham Ali v Secretary of State for the Home Department [2016] UKSC 60, [2016] 1 WLR 4799). (I would add that the use of exceptionality as a distinct criterion is not generally appropriate in this field: see the observations of Lord Hamblen at para. 38 of his judgment in HA (Iraq) v Secretary of State for the Home Department [2022] UKSC 22, [2022] 1 WLR 3784.)

Although I have no doubt that that is the explanation for the use of the phrase “exceptional circumstances” in this context, there may be a need for caution about its relevance to the specific case of family reunion for child refugees. It is important to appreciate that the rules and guidance in which the phrase is deployed are not concerned only with such cases but with all applications for family reunion otherwise than by partners and minor children. I have no difficulty with the proposition that it would be exceptional for the refusal of leave to parents or siblings of an adult refugee to have “unjustifiably harsh consequences” or to constitute a breach of article 8. But it is not at all obvious that that would be so in the case of a child refugee. Although of course each case would depend on its own facts, the support of a parent (or sometimes an adult sibling) is inherently more likely to be important for a child refugee than for an adult, and there is no a priori reason to suppose that permitting family reunion should be exceptional in such a case.

The focus is thus on elements (b) and (c) – that is, “unjustifiably harsh consequences” and breach of article 8. The two might at first appear to be intended to constitute a single criterion. The language of “unjustifiably harsh consequences” is very close to the kind of test applied in the article 8 case-law (and indeed legislation (Footnote: ⁹)) in broadly analogous situations involving family separation; and I find it hard to conceive of a case where the refusal of entry to the parents or siblings of a child refugee would be held to result in unjustifiably harsh consequences and yet not constitute a breach of article 8 rights. Nevertheless, as explained above, both the old and the new provisions are drafted in a way which contemplates that there will be cases where element (b) is present but element (c) is not. I am not sure that that was really intended but it is prudent to respect the distinction in the formal analysis.

That being so, the correct approach to applying the policy in the case of an application for family reunion by a parent or sibling of a child refugee is as follows:

It follows that, under the policy, family reunion decisions in the case of a child refugee have to be made on a case-by-case basis involving an evaluative assessment, based on the evidence supplied by the applicant, of whether refusal of reunion (i.e. continued separation) would create unjustifiably harsh consequences for the child and/or the family member(s) concerned. It will be apparent that that approach is structurally different from that followed in the case of the partners and minor children of adult refugees, who are eligible automatically. That difference is at the heart of the Claimant’s case.

Although the policy refers without differentiation to parents and siblings, any case based on unjustifiably harsh consequences is likely to focus on the child refugee’s relationship with his or her parents: the policy behind allowing minor children to be reunited with adult refugee parents is based on the child’s dependence on the parents – see para. 21 above – and that would apply equally in this context. There may be unusual cases where a relationship of dependency exists with an older sibling (for example, if both parents have died), but typically the case for granting leave to enter to the siblings will be parasitic on the grant to the parents (because if they are granted leave their other dependent children cannot be left behind).

Ms Giovannetti was anxious to draw to our attention the fact that both the old and the new Guidance refer extensively to section 55 of the 2009 Act. Both have sections near the start headed “Application[s] in respect of Children” which set out the need in every such case to treat the child’s best interests as a primary consideration and cross-refer to the Secretary of State’s general guidance on the application of section 55 in the immigration context, Every Child Matters (see para. 64 below). Both also refer to section 55 in the sections which set out the policy with which we are concerned in this case – see paras. 25 and 35 above. However, those references can only be relevant to the taking of decisions in individual cases: they are not concerned with the terms of the policy itself, which is the framework within which those decisions have to be taken. I will have to return to this point later.

Finally, I should mention one point of terminology. Mr Husain formulates his various challenges as being to “the Rules”, on the basis that it is the Claimant’s case that the Rules should have provided for child refugees to have an automatic right to sponsor their parents and siblings. I understand that, but I have preferred generally to refer to the Secretary of State’s “policy”, since, as we have seen, the position of child refugees overall is the result of her policy expressed in a combination of Rules and guidance.

It is important to appreciate that the Secretary of State’s policy does not involve an absolute prohibition on family reunion for child refugees. The Claimant acknowledges that, but he says that the fact that child sponsors have to satisfy the unjustifiably harsh consequences criterion on a case-by-case basis means that it is inherently inferior to a policy permitting automatic reunion. In particular:

It does not appear to be possible to put a figure on how many child refugees suffer those impacts. The starting-point must be the number of child refugees who are present in the UK at any one time: that figure is apparently not known, though it would seem to be in the low thousands. (Footnote: ¹⁰) However, family reunion will not be relevant in all, possibly even most, of those cases: the children may be orphans or have lost touch with their families, or their relationship with their parents may be such that they do not wish to be reunited with them, or the parents may not wish to come to the UK. All that can safely be said is that there will be a substantial cohort of child refugees for whom family reunion is desired and desirable.

Nor is it possible to evaluate the extent of any impact on that cohort except in the broadest terms. There appear to be no figures for how many potentially eligible applications for family reunion are made by the parents and/or siblings of child refugees. There are some indications that the numbers are low, but it is impossible to know how many applications are deterred by the difficulty of the process. (Footnote: ¹¹) Nor do we know what proportion of the applications which are made are successful (either at initial decision or on appeal); and even if we had those figures it could not be assumed that every application which failed would have succeeded if eligibility had been automatic (see para. 23 above). The most that can safely be said is that impacts of the kind identified are liable to occur in a number of cases.

There has for many years been a good deal of criticism of the Secretary of State’s policy as regards family reunion for child refugees. This is relevant to the discrimination and irrationality grounds, which are concerned with the substantive merits of the policy; but it is also, for reasons which will appear, relevant to the section 55 ground. I need to summarise the principal sources of that criticism, but I do not at this stage say anything about its validity.

Parliamentary criticism. There were before the Court excerpts from reports of three Parliamentary Committees addressing the issue of family reunion for child refugees. It is not necessary to summarise them in any detail, but in headline terms:

The Chief Inspector. The Independent Chief Inspector of Borders and Immigration (“the Chief Inspector”) considered the family reunion policy for child refugees in paras. 6.12-6.17 of a report sent to the Secretary of State on 7 January 2020 and subsequently published, with her response, on 8 October 2020. The report is not itself explicitly critical of the Secretary of the State’s policy, but it identifies criticisms by “stakeholders” and the Home Office’s response and recommends that the Government clarifies its policy.

“Without My Family”. In January 2020 Amnesty International, the Refugee Council and Save the Children published a report entitled Without My Family: the impact of family separation on child refugees in the UK. Recommendation 1 was that the Home Office should “permit the right to family reunion for unaccompanied children with international protection needs when this is in their best interests”.

UNHCR. In 2019 UNHCR published three reports – Families Together: Family reunification for refugees in the European Union (February 2019); Destination Anywhere: the profile and protection situation of unaccompanied and separated children and the circumstances which lead them to seek refuge in the UK (June 2019);and A Refugee and Then …: participatory assessment of the reception and early integration of unaccompanied refugee children in the UK (June 2019) – together “the UNHCR reports”. Each of these is critical of current UK Government policy as regards family reunion for child refugees. (We were also shown a report dated July 2024 entitled UNHCR’s Recommendations to the Government of the United Kingdom, but that contains nothing material for our purposes.)

One point picked up in many of these criticisms is that the great majority of developed countries allow family reunion for the parents and/or siblings of child refugees on the straightforward basis of their relationship with the sponsor: the only exceptions are apparently the United States, Canada and Switzerland. (Footnote: ¹²)

There are references in the materials before us to other occasions where the policy on family reunion for child refugees has been the subject of criticism both in Parliament and from the Chief Inspector and others; but the documents to which we were referred are sufficient by way of background.

In addition to those materials, there were before the Court witness statements from the Claimant and his mother, together with a report from a clinical psychologist, addressing the impact that the difficulties of the process of seeking family reunion had had on him. As regards the policy more generally, there were witness statements from Judith Dennis, Policy Manager at the Refugee Council, Anna Jones of RefuAid, Sheona York of Kent Law Clinic and Lawrence Bottinick, a Senior Legal Officer at UNHCR. Part of the purpose of these is to exhibit the reports referred to above, but they also contain, to varying extents, the witnesses’ own evidence about the impact on child refugees and their families of aspects of the policy.

The Government has expressed its response to these criticisms in various forms. I need only refer here to the most recent statement which was before the Court, being the Home Office’s response to the Chief Inspector’s recommendation that the Government clarify its position on child sponsors. This reads:

“4.3

The Government has made clear in the past its concern that allowing children to sponsor parents would risk creating incentives for more children to be encouraged, or even forced, to leave their family and attempt hazardous journeys to the UK. This would play into the hands of criminal gangs, undermining our safeguarding responsibilities.

4.4

Government policy is not designed to keep child refugees away from their parents, but in considering any policy we must think carefully about the wider impact to avoid putting more people unnecessarily into harm’s way. There is a need to better understand why people choose to travel to the UK after reaching a safe country. It is important that those who need international protection should claim asylum in the first safe country they reach – that is the fastest route to safety.”

The phenomenon referred to in para. 4.3 of children being sent on hazardous journeys so that they can then be joined by their families is elsewhere described as the use of “anchor children”. I return to this later.

It is convenient if I also quote at this stage a passage from the Chief Inspector’s report which is particularly material to the section 55 ground. At para. 6.15 he says:

“Home Office policy staff told inspectors that most major decisions about family reunion policy were made by ministers and ‘sometimes decisions taken are inevitably political. That is out of our control ultimately.’ The Home Office did not share any advice that it had put to ministers regarding policy options for family reunion. Inspectors asked for the rationale for excluding children from sponsoring family reunion applications. The Home Office’s response echoed what ministers had previously told Parliament:

‘If children were allowed to sponsor parents, this would risk creating incentives for more children to be encouraged, or even forced, to leave their family and risk hazardous journeys to the UK. This plays into the hands of criminal gangs who exploit vulnerable people and goes against our safeguarding responsibilities. This position supports our commitment to protecting vulnerable individuals.’”

Para. 6.17 notes that Home Office staff had told his inspectors that “child sponsors was a ‘ministerial red line’”. The relevance of these references is that they show that the Secretary of State has consistently declined to reconsider her policy as regards family reunion for child refugees.

Section 55 came into force on 2 November 2009. So far as material, it reads as follows:

“Duty regarding the welfare of children

(1)

The Secretary of State must make arrangements for ensuring that—

(a)

the functions mentioned in subsection (2) are discharged having regard to the need to safeguard and promote the welfare of children who are in the United Kingdom, and

(b)

any services provided by another person pursuant to arrangements which are made by the Secretary of State and relate to the discharge of a function mentioned in subsection (2) are provided having regard to that need.

(2)

The functions referred to in subsection (1) are—

(a)

any function of the Secretary of State in relation to immigration, asylum or nationality;

(b)

any function conferred by or by virtue of the Immigration Acts on an immigration officer;

(c)

any general customs function of the Secretary of State;

(d)

any customs function conferred on a designated customs official.

(3)

A person exercising any of those functions must, in exercising the function, have regard to any guidance given to the person by the Secretary of State for the purpose of subsection (1).

(4)

-(5) …

(6)

In this section—

“children” means persons who are under the age of 18;

…

(7)-(8) …”

The genesis of those provisions is helpfully summarised at para. 70 (i) of the judgment of David Richards LJ in R (Project for the Registration of Children as British Citizens) v Secretary of State for the Home Department [2021] EWCA Civ 193, [2021] 1 WLR 3049, (“the PRCBC case”) as follows:

“Section 55 was enacted to give effect in domestic law, as regards immigration and nationality, to the UK’s international obligations under article 3 of the 1989 United Nations Convention on the Rights of the Child (‘UNCRC’). The UK is a party to the UNCRC and in 2008 withdrew its reservation in respect of nationality and immigration matters. Article 3 provides that:

‘In all actions concerning children, whether undertaken by public or private social welfare institutions, courts of law, administrative authorities or legislative bodies, the best interests of the child shall be a primary consideration’.

Although section 55 uses different language, it is conventional and convenient to refer to a duty under section 55 as being to have regard, as a primary consideration, to the best interests of the child.’”

Like David Richards LJ, I will in this judgment use the phrase “have regard to the best interests of the child” as a convenient paraphrase for the language of section 55 (1) (a).

I should note that subsections (2) and (4) of section 11 of the Children Act 2004 impose duties in substantially the same terms as section 55 (1) and (3) of the 2009 Act on a variety of public bodies whose functions may affect children.

Although section 55 does not as such impose an obligation on the Secretary of State to issue guidance about its application, it is evident from subsection (3) that the intention was that she would do so. In November 2009 the Minister of State for Borders and Immigration and the Parliamentary Under-Secretary of State for Children, Young People and Families issued guidance entitled Every Child Matters on the application of both section 55 of the 2009 Act and section 11 of the 2004 Act. This sets out a number of arrangements designed to ensure that the functions of the UK Borders Agency (Footnote: ¹³), including the making of individual immigration decisions, are discharged having regard to the need to safeguard and promote the welfare of children. But, as noted at para. 46 above, it is not concerned with the making of the Rules or other general policies governing decisions outside the Rules.

There is a good deal of case-law about the meaning and effect of section 55, but most of it is of no more than background relevance to the issue before us. I need only quote the remaining points under para. 70 of the judgment of David Richards LJ in the PRCBC case, which is a distillation of the case-law up to that date:

“(ii)

The duty is imposed on the Secretary of State. She is bound by it, save to the extent (if any) that primary legislation qualifies it; we were not referred to any qualifying legislation.

(iii)

The duty applies not only to the making of decisions in individual cases but also to the function of making subordinate legislation and rules (such as the Immigration Rules) and giving guidance. The fact that subordinate legislation or rules are subject to the affirmative vote of either or both Houses of Parliament does not qualify the Secretary of State’s statutory duty under section 55.

(iv)

The best interests of the child are a primary consideration, not the primary consideration, still less the paramount consideration or a trump card. This does, however, mean that no other consideration is inherently more significant than the best interests of the child. The question to be addressed, if the best interests point to one conclusion, is whether the force of other considerations outweigh it.

(v)

This in turn means that Secretary of State must identify and consider the best interests of the child or, in a case such as the present, of children more generally and must weigh those interests against countervailing considerations.”

I will refer to the process referred to at (v) as “the section 55 exercise”. At para. 108 of his judgment in R (SG) v Secretary of State for Work and Pensions [2015] UKSC 16, [2015] PTSR 471, Lord Carnwath defined the essence of what was required as follows:

“[T]he evaluation needs to consider, where relevant, the interests both of children in general and of those directly affected by the action. It also needs to indicate the criteria by which the ‘high priority’ given to children’s interests has been weighed against other considerations. In so far as that evaluation shows conflict with the best interests of the children affected, it needs either to demonstrate how that conflict will be addressed, or alternatively what other considerations of equal or greater priority justify overriding those interests.”

It will be seen that section 55 imposes two duties –

We are in this case concerned with the duty under subsection (1). Two things should be noted about its structure:

We asked counsel if they were aware of any precedent for the use of a “make arrangements” structure of this kind, or any discussion of it in the authorities, but they could find none. I think, however, that it must reflect a deliberate distinction between the Secretary of State’s functions in making individual decisions in relation to immigration, asylum or nationality, which will in practice be discharged by civil servants on her behalf in accordance with the Carltona principle, and her role in setting the policy which will be applied in the making of such decisions, which will be discharged by her (or Ministers in her Department – see para. 84 below) personally. This distinction is recognised, albeit in a different context, at para. 55 of the judgment of Lord Sales and Dame Siobhan Keegan in CAO v Secretary of State for the Home Department [2024] UKSC 32, [2024] 3 WLR 847.

It appears from subsection (3) that the principal means by which it is envisaged that the Secretary of State will perform her duty under subsection (1) is by giving guidance. But para. 70 (iii) of the judgment in the PRCBC case makes it clear that the duty applies also to the making of Immigration Rules. That reflects a statement by Lady Hale and Lord Carnwath at para. 92 of their judgment in R (MM (Lebanon)) v Secretary of State for the Home Department [2017] UKSC 10, [2017] 1 WLR 771) that:

“The duty imposed by section 55 of the 2009 Act … applies to the performance of any of Secretary of State’s functions including the making of the rules [my italics].”

That was, as I have said, common ground before us; and I am content to proceed on the basis that it is correct. I would, however, venture to suggest, in case the issue falls to be considered elsewhere, that a better analysis might be that the making of the Rules does not itself constitute the discharge of a function of the kind referred to in subsection (2), which is concerned only with the function of making decisions in individual cases: rather, the Rules are to be regarded as constituting “arrangements” within the meaning of subsection (1). That is in my view conceptually more satisfactory, and it seems to be how the drafter regarded guidance – see the reference in subsection (3) to “guidance … for the purpose of subsection (1)”: if guidance is treated as part of the “arrangements”, surely the same should be the case as regards the Rules. However, the two analyses have the same effect in substance, since in order that individual decisions are taken with regard to the best interests of children it is necessary that the Rules which provide the framework for those decisions should themselves have regard to that consideration.

It will be convenient to mention at this point section 149 of the Equality Act 2010, which establishes the public sector equality duty (“the PSED”). It is to some extent analogous with section 55 of the 2009, although its drafting is not identical, and we were referred to several authorities about it. The PSED is established by subsection (1), which reads:

“A public authority must, in the exercise of its functions, have due regard to the need to—

(a)

eliminate discrimination, harassment, victimisation and any other conduct that is prohibited by or under this Act;

(b)

advance equality of opportunity between persons who share a relevant protected characteristic and persons who do not share it;

(c)

foster good relations between persons who share a relevant protected characteristic and persons who do not share it.”

The similarities between section 149 and section 55 of the 2009 Act are obvious. (Footnote: ¹⁴) Like section 55, it imposes a process duty – that is, a duty in the exercise of public functions to have regard to a specified need: the differences between “regard” and “due regard” and between the “exercise” and the “discharge” of the functions are surely immaterial. But it is notable that section 149 (1) does not impose a duty to “make arrangements” but imposes the “have regard” duty directly on the authority.

As already noted, the issues in relation to the section 55 ground have developed substantially over the course of the case. It is necessary to summarise that development in a little detail.

Paras. 94-95 of the Claimant’s Statement of Facts and Grounds read as follows:

“94.

In summary, section 55, taken with Article 3(1), requires that – in formulating and maintaining [emphasis supplied] her position in relation to family reunion for refugee children – the Secretary of State must:

(1)

identify with reasonable accuracy and assess the best interests of refugee children in the United Kingdom (R (Project for the Registration of Children as British Citizens) v SSHD, §96);

(2)

treat these interests as a ‘primary consideration’, outweighed only by strong countervailing factors (MM, §§91-92; R (Project for the Registration of Children as British Citizens) v SSHD, §82); and

(3)

show how these interests have been weighed against other considerations (R (JS) v Work and Pensions Secretary [2015] 1 WLR 1449, §108) – or, otherwise put, show that she has evaluated the impact of her position on the group of children concerned (MM (cited above); R (DA) v Secretary of State for Work and Pensions [2019] 1 WLR 3298, §178 per Lord Kerr; R (Project for the Registration of Children as British Citizens) v SSHD, §112).

95.

On the basis of the evidence presently available, there is nothing to suggest – let alone show – that the Secretary of State has complied with any of these obligations in the context of family reunion for refugee children.”

(Paras. 96-97 amplify the statement in para. 95 but I need not set them out.)

The essence of that case is that the Secretary of State has never “in formulating and maintaining” her policy as regards family reunion in the case of child refugees conducted a section 55 exercise. That case is reflected in the terms of the relief sought in section 7 of the Claim Form.

The Secretary of State’s response to that case, pleaded at paras. 27-34 of the Detailed Grounds of Defence (“the DGD”), was twofold:

At paras. 101-116 of his first judgment Lavender J recorded how the arguments on ground 1 had developed in the course of the hearing and the subsequent written submissions.

His starting-point is a careful analysis of the Claimant’s references to the Secretary of State “formulating and maintaining” the impugned policy. At paras. 103-104 he says:

“103.

As to this, three points emerged during the course of the hearing:

(1)

The claimant’s complaint is really about the relevant provisions of the Immigration Rules, rather than the published guidance on the application of those Rules. As I have already noted, Mr Husain acknowledged that the Secretary of State could not have used an amendment to the Family Reunion Guidance, or anything else short of a change to the Immigration Rules, to introduce the change for which the claimant contends.

(2)

The ‘establishment’ of the relevant ‘position’ under the Immigration Rules took place in 2000, long before section 55 came into force. The claimant cannot complain about a failure to comply with section 55 in 2000, when section 55 was not in force.

(3)

Consequently, the claimant’s case is that the Secretary of State failed to comply with section 55 when ‘maintaining’ the relevant position after section 55 came into force.

104.

This last point led me to ask during the hearing what the claimant contended was the relevant ‘function’ which the Secretary of State had discharged on one or more occasions since 2 November 2009. I invited the parties to make written submissions on this point after the hearing. …”

Before us Mr Husain did not dispute this analysis of his case. I respectfully agree with the Judge that it is an essential part of the analysis to identify the function for the purpose of discharging which the section 55 exercise is said to have been required.

At paras. 105-108 Lavender J also analysed the Secretary of State’s response in the DGD. He concluded that neither of her answers addressed the Claimant’s case as pleaded. Specifically:

Against that background, the Judge says at para. 108:

“In the hearing, it did not appear to me that Miss Giovannetti was submitting that the Secretary of State had ever made arrangements which were compliant with section 55 in relation to ‘maintaining’ the position established when the Immigration Rules were changed in 2000. Miss Giovannetti confirmed that that was the case. In essence, therefore, the case which she advanced was that the Secretary of State had not discharged a relevant function since 2 November 2009. In putting the Secretary of State’s case in that way, Miss Giovannetti, as I understand it, quite properly accepted that, if the Secretary of State had discharged a relevant function during those years, she had not complied with her duty under section 55 when discharging that function.”

Ms Giovannetti did not suggest to us that that was a mis-statement of her position before the Judge.

The upshot of all that is that the issue as it crystallised at the hearing came down simply to whether the section 55 duty had ever arisen – or, more precisely, whether in “maintaining” her position under the Rules following 2 November 2009 the Secretary of State could be said to have been discharging any function for the purpose of section 55 (1). It is important to appreciate that it was never the Secretary of State’s case that she had conducted a section 55 exercise in formulating the policy as it stood as at 2 November 2009.

As noted above, there were extensive written submissions following the hearing. The sequence is as follows:

At paras. 110-111 Lavender J summarised the Claimant’s case appearing from those submissions as being that, taking a “broad and non-technical approach”, the relevant functions for the purposes of section 55 “include formulating, amending/updating and considering whether to alter or maintain [emphasis supplied] the Immigration Rules and associated policies”. Mr Husain’s primary position was that it was not necessary for him to identify precisely the point when the Secretary of State had discharged a relevant function or functions; but in case that was not accepted he identified at least ten occasions since 2 November 2009 when he said that she had done so. These are recorded at para. 110 of the judgment but I need not set them out because, as will appear, in his submissions before us Mr Husain took a more focused approach.

The Judge records the Secretary of State’s response to that case at para. 113 as follows:

“The Secretary of State submitted that she had not, since 2 November 2009, discharged any function (such as making a rule or subordinate legislation) which triggered section 55 and that:

‘… it would be a significant extension of the ambit of s. 55 to hold that it applies where the SSHD does not propose to make any changes to her current policy or practice. While the SSHD will, of course, consider representations or recommendations that she should do so, declining to accede to such representations or recommendations is not a ‘function’ engaging s. 55.’”

Her third round submissions, settled by Ms Giovannetti, included the following passage:

“9.

The Secretary of State is not aware of any occasion since s.55 came into force (2 November 2009), when the relevant decision makers (namely Home Office Ministers or the Secretary of State) decided to review the Immigration Rules in order to consider providing a route to family reunion for child refugees (i.e. introducing criteria within the Rules governing decisions whether or not to grant leave to enter to the parents and siblings of refugee children).

10.

Records since 2015 indicate that the consistent position of the relevant decision makers, as communicated to officials, has been that they are not prepared to change the existing and long-standing policy of considering applications for leave to enter by immediate family members of child refugees on a case-by-case basis outside the Immigration Rules. Thus, for example, Ministers were clear that changing that policy was not one of the options to be included in 2021 consultation on the New Plan for Immigration (which fulfilled the statutory obligation to review legal routes to the UK from the European Union (EU) for protection claimants, set out in the Immigration and Social Security Co-Ordination (EU Withdrawal) Act 2020).

11.

As to the position before 2015, a search has been conducted, but the Secretary of State been unable to find relevant communications from Ministers to officials dating back beyond that date. To the best of the Secretary of State’s knowledge, even prior to 2015, the relevant decision makers were consistent in their position that they intended to maintain the existing policy, as summarised above. This is supported by Family Reunion Guidance from 2007 to 2011 (see Jason Büültjens’ witness statement, para 7).”

The factual content of these paragraphs was expressly confirmed in Mr Büültjens’ witness statement.

I should note one point arising out of the reference in the submissions quoted by the Judge to “the relevant decision makers (namely Home Office Ministers or the Secretary of State)”. The reference to “Ministers or the Secretary of State” reflects the language of the contemporary documents, which tend to refer generically to “Ministers”. Although we were not referred to any authority on the formal constitutional position, I am sure Ms Giovannetti is right that the duty imposed by section 55 (1) (a) on the Secretary of State extends also to Ministers in her department taking decisions in their areas of responsibility. However, for economy I will in this judgment continue to refer simply to the Secretary of State.

On the basis of that analysis Lavender J in his first judgment addressed the issue before him by reference to two questions – (1) “What constitutes the discharge of a function?”; and (2) “Did the Secretary of State discharge a relevant function?”.

He considered the first of those questions at paras. 120-136 of the judgment. After reviewing the authorities, at paras. 133-134 he concludes:

“133.

It is not disputed that the Secretary of State discharges a function when she makes a change to the Immigration Rules. In order to discharge that function, she has to consider from time to time whether to make any and, if so, what changes to the Immigration Rules. It seems to me that she can properly be described as discharging a function when she actively engages in that consideration. As part of that consideration, she may have to choose between various options, one of which may be to make no change to the Immigration Rules. It seems to me that when she decides to choose one option rather than another, including the option of making no change to the Immigration Rules, she is discharging her function of reviewing the Immigration Rules and considering and deciding whether to change them in one or more ways.

134.

However, the decision in Adiatu [R (Adiatu) v Her Majesty’s Treasury [2020] EWHC 1554 (Admin), [2020] PTSR 2198] is a salutary reminder that both the public sector equality duty and the duty imposed by section 55 of the 2009 Act have to be kept within sensible bounds. In this context, there are a spectrum of possibilities.

(1)

At one end of the spectrum, Badmus[R (Badmus) v Secretary of State for the Home Department [2020] EWCA Civ 657, [2020] 1 WLR 4609] concerned a relatively formal process, resulting in a decision which was amenable to judicial review, and including a review of existing policy, a report containing a recommendation that consideration be given to changing the policy and ‘a clear and considered policy choice’ between identified options. It seems to me that if the Secretary of State were to take a decision in the immigration, asylum or nationality context similar to the 2018 Review Decision in Badmus, she would be discharging a function for the purposes of section 55 of the 2009 Act, even if the decision were to make no change to existing arrangements.

(2)

At the other end of the spectrum is the claimant’s submission that the Secretary of State was exercising a relevant function when she responded to the pre-action protocol letter. It cannot be the case that the section 55 duty is triggered whenever a claimant sends a letter contending that the Secretary of State should change her policy.

(3)

Indeed, once the Secretary of State has decided to adopt one policy rather than another, I do not consider that she is to be treated as, in effect, re-making that decision every time she applies, repeats, defends or declines to change the policy which she has adopted.”

As regards the second question, at paras. 137-140 he considered whether, in accordance with his answer to the first question, the Secretary of State had at any time after 2 November 2009 “taken a positive decision, following active consideration, not to change the Immigration Rules so as to create a route to family reunion for refugee children” (para 138). He identified at para. 139 various occasions appearing from the documents which suggested that ministers might have taken such a decision. However, he concluded at para. 140 that

“… the evidence served after the hearing [i.e. the evidence of Mr Büültjens confirming the Secretary of State’s post-judgment submissions (see para. 83 above)] makes clear thatthe relevant decisionmakers, i.e. the Secretary of State and Home Office ministers, have not given active consideration since 2 November 2009 to the policy option of changing the Immigration Rules so as to create a route to family reunion for refugee children. That evidence is consistent with what is recorded in paragraph 6.17 of the Chief Inspector’s 2020 [report], i.e. that, while Home Office staff were considering the issue of child sponsors, ‘child sponsors was a “ministerial red line”’.”

Ground 1 of the Claimant’s Amended Grounds of Appeal reads:

“The Administrative Court erred in holding that the Respondent had not exercised a relevant function for the purposes of s 55 of the Borders and Citizenship Act 2009 at any point since that provision came into force. On a proper interpretation and application of s 55, the Respondent exercised a relevant function (i) on the multiple occasions when, as the Administrative Court found, she made judicially reviewable decisions not to review the relevant paragraphs of the Immigration Rules to consider whether (inter alia) she should provide a more straightforward route to family reunion for refugee children and/or (ii) when she operated her overall system for refugee family reunion. Further or in the alternative, the Respondent is in breach of s 55 as that section does not permit (and it would frustrate Parliament’s intent in enacting it for) the Respondent to refuse to discharge a relevant function indefinitely, as has been the case here.”

The underlined words were added by amendment, with the consent of the Secretary of State, by order of Lewis LJ dated 16 January 2025. It is accepted that they raise an argument not advanced before Lavender J. I will refer to the two arguments identified at (i) and (ii) as “the original case” and “the new case” respectively.

By a Respondent’s Notice the Secretary of State seeks to uphold the Judge’s decision for the reasons that he gives but also relies on the following additional ground:

“In relation to ground 1, at §133 of his judgment, Lavender J held that when the Respondent ‘actively engages’ in considering whether to amend the Immigration Rules she is engaging a function. The Respondent submits that she is not exercising a function when she does not propose to make any changes to her current policy or practice (whether or not she is considering any representations or recommendations that she should do so).”

It is important to appreciate that the Judge’s analysis at paras. 105-108, not challenged by the Secretary of State, means that the defences originally advanced in the DGD are no longer relied on. In particular, the Secretary of State no longer contends that the fact that decision-makers in individual family reunion applications treat the best interests of the child as a primary consideration is an answer to the Claimant’s case that they were not so treated in the formulation of the overall policy.

I will consider the original and the new cases separately, addressing the Respondent’s Notice in the context of the former.

THE ORIGINAL CASE

The Claimant’s case is that the Secretary of State was discharging a relevant function within the meaning of section 55 on each occasion that she decided, as pleaded in ground 1 (i), “not to review the relevant paragraphs of the Immigration Rules to consider whether (inter alia) she should provide a more straightforward route to family reunion for refugee children”; and that Lavender J had himself recognised that she had made such decisions on “multiple occasions”. The relevant discharge of a function for the purpose of section 55 is thus said to consist in the making of a decision not to review the Rules. (Footnote: ¹⁵) Before us Mr Husain in the end relied on only two such alleged decisions; but I need to explain how that position was reached.

The starting-point is that as part of his order granting permission to appeal Lewis LJ gave a direction requiring the Claimant to “identify precisely each act, and the date of each act, said to amount to an exercise of a function within the meaning of section 55”. His concern was in fact with a potential time point; but for present purposes what matters is that, in order to be able to comply with the direction as to dates, the Claimant made a Request for Further Information (“RFI”) to the Secretary of State, and her response prompted a further request. In summary:

It does not appear that the Claimant proceeded, with the benefit of those responses, to file a separate document identifying the relevant decisions as directed by Lewis LJ; but that does not matter, since his case was sufficiently identified in paras. 54-58 of Mr Husain’s amended skeleton argument, which are headed “The decisions made by the Respondent”. The decisions relied on are the four identified in the RFI; but before us Mr Husain abandoned his reliance on two of them (being those identified in para. 57), and relied only on the two decisions identified in para. 58. These are said to have occurred in 2016 (para. 58.1) and 2018 (para. 58.2): I will refer to them as “the 2016 decision” and “the 2018 decision” respectively. There is no direct evidence of either decision: instead, the Court is asked to infer that they were taken from various references in the contemporary documents. However, the Secretary of State has adduced no detailed evidence to rebut those inferences and indeed has appeared to accept that decisions of some kind were taken: the real issue is about their nature and who took them. I review them in turn.

The starting-point as regards the 2016 decision is a passage in the decision of the Upper Tribunal (McCloskey P sitting alone) in AT v Entry Clearance Officer of Abu Dhabi [2016] UKUT 00227 (IAC), promulgated on 29 February 2016. The case concerned the refusal of leave to enter on the basis of family reunion to the mother and younger brother of a child refugee. The appeal proceeded on the (correct) basis that the appellants had no such right under the Rules, but McCloskey P held that the refusal was a breach of the sponsor’s rights under article 8 of the ECHR. At paras. 10-11 of his Decision he described the Secretary of State’s policy at the material time as follows:

“10.

The Secretary of State’s policy in the realm of family reunification, as expressed in the Immigration Rules, dates from the year 2000. Its most important feature, for the purposes of these appeals, is that no provision has ever been made for family reunification in the case of a child who has gained refugee status in the United Kingdom. This discrete regime is currently contained in Part 8 of Appendix FM to the Rules, at paragraphs 352A-352G and 819L-819U. In short, spouses and minor children of a ‘sponsor’ can, subject to satisfying the governing conditions, secure family reunification in the United Kingdom by the grant of leave to enter. However, this possibility does not exist where the sponsor is a child.

11.

Thus a blanket prohibition is in operation. Historically, there was a short lived exception to this prohibition relating to the parents of unaccompanied children who had fled Kosovo and secured asylum in the United Kingdom. This concession was confined to the short time frame of July to September 1999. With effect from 02 October 2000, the family reunification regime enshrined in the Immigration Rules contained the aforementioned blanket prohibition. From then to 2006 the Secretary of State operated a policy of permitting the parents or siblings of unaccompanied minor refugees to enter the United Kingdom for the purpose of reunification only where compelling and compassionate circumstances were demonstrated. Since 2006 the Secretary of State’s policy has extinguished this possibility. …”

McCloskey P is of course right that the Rules did not provide for family reunion in the case of a child refugee. That would not itself constitute a “blanket prohibition”, since it would be perfectly possible for the Secretary of State to have a policy of permitting family reunion in some circumstances outside the Rules. However, there is no mention of any such policy in the version of the Guidance then in force (being the July 2011 Guidance). The first version of the Guidance which contains the “key passages” quoted at para. 25 and 26 above appears to be what is described as version 2, dated 29 July 2016.

Mr Husain submits that the clear inference from the foregoing is that the key passages were inserted in the Guidance in response to the decision in AT and that that evidences a “decision” by the Secretary of State attracting the duty under section 55 (1) (a). On that basis the first RFI referred to “the decision … to remove the ‘blanket prohibition’ on children sponsoring their parents for family reunion and instead to allow them to do so in ‘exceptional circumstances’ outside the Immigration Rules”. Mr Husain points out that the Secretary of State in her response did not query that description but simply answered the request, giving the date “April-May 2016”.

I turn to the 2018 decision, which Mr Husain described as being his best case on this issue. He relied on two pieces of evidence.

The first is an “ad-hoc query” submitted to the European Migration Network (“the EMN”), being an arm of the European Commission, in February 2018 by a civil servant in the Home Office designated as the “national contact point” (“NCP”). The subject-matter of the query was described as “Evidence on the impact that policy changes on the right to refugee family reunion may have on asylum intake and the number of family reunion applications received”, and it comprised six questions about aspects of member states’ policy and procedures as regards family reunion for refugees. For our purposes it is only necessary to set out qu. 5, which reads:

“Do you allow children recognised as refugees or granted subsidiary protection to sponsor relatives for purposes of family reunion? If yes; When did you start allowing child sponsors and what impact did this have? If yes; How many family reunion visas have you granted, in the last five years, where the sponsor was a child?”

Those questions were preceded by a section, evidently drafted by the NCP (or someone else in the Home Office), headed “Background information”, which reads (so far as relevant) as follows:

“The UK are currently reviewing the policy on refugee Family Reunion and listening to the concerns from Non-Government Organisation’s [sic] that the current policy and the Immigration Rules on family reunion are too narrow. This work is part of our wider asylum and resettlement strategy. We are gathering evidence on whether changes to policy creates a ‘pull factor’ that may lead to more people risking dangerous journeys to Europe, and on the number of refugee family reunion applications that could be expected with associated analysis of the impact of the cost on public services.”

The second piece of evidence arises out of a report from the Chief Inspector in 2018 in which he had observed that family reunion policy development had ceased to be a priority within the Home Office. The Home Office’s response, published in September 2018, included a statement that it was “… reviewing the approach to Family Reunion as part of the wider asylum and resettlement strategy” and considering recent parliamentary debates on the issue and discussions with NGOs.

In the first RFI the Claimant asserted (a) that that evidence established that a substantive review of the Government’s child refugee policy was being undertaken and (b) that since it was the Secretary of State’s case that her consistent position had been that she was not prepared to change her policy it must necessarily be inferred that she “made a decision to this effect either before, during, or following the review”. The RFI asked for the date of that decision. Again, the Secretary of State in her response did not dispute that there had been such a decision but merely responded “after July 2018”.

Ms Giovannetti accepted that the making of Immigration Rules constituted the discharge by the Secretary of State of a function for the purpose of section 55: cf. para. 69 above. She also, unsurprisingly, accepted that the same was true of the making of a change to the Rules.

More significantly for our purposes, Ms Giovannetti also accepted that a decision not to change the Rules could in principle constitute the discharge of a function. However, she contended that that would only be so in the case of a decision taken by the Secretary of State as the result of a process under which she considers a particular option, or options, for change. Making a decision to maintain the status quo following such a process could properly be regarded as the discharge of a function; but a mere failure, or refusal, to undertake a review (which is what is relied on by the Claimant – see ground 1 (i)) could not. Ms Giovannetti said that typically such a process would take the form of a proposal by officials setting out the various options – which would include (explicitly or not) the option of doing nothing – on the basis of which the Secretary of State would make her decision. I will as a shorthand refer to that as a process of “substantive review”.

Ms Giovannetti submitted that on the basis of that approach Lavender J was right to conclude that Mr Büültjens’ evidence established that the Secretary of State had not discharged a relevant function – that is, a function as regard family reunion for child refugees – at any time since the coming into force of the 2009 Act. Her policy was established prior to that date and she has not been prepared to reconsider it at any point since then: as the Judge said, it was a “red line”. Neither of the “decisions” relied on by the Claimant contradicted that evidence: in so far as any decision was taken, it was not of the necessary character.

That is essentially the approach taken by Lavender J at para. 133 of his judgment, answering his question “What constitutes the discharge of a function?”. The Respondent’s Notice reflects a concern that his reference to the Secretary of State “actively engaging” in a consideration of whether to change the Rules may not fully reflect the kind of formal decision-making process which she says is necessary; but in so far as there is a real difference between his formulation and Ms Giovannetti’s it seems to me to be marginal.

I prefer Ms Giovannetti’s submissions. It is common ground that when the Secretary of State makes (or amends) a particular rule she discharges a function within the meaning of subsection (1) (subject to my caveat at para. 69 above). But that involves a substantive act on her part. It does not follow that she should be regarded as discharging a function on every occasion that she declines to consider whether to change her policy. That could in some sense be described as a decision, but it does not involve what would on any ordinary understanding be regarded as an exercise of the Secretary State’s statutory functions. Neither section 55 itself (subject to the Claimant’s new case) nor any other statutory provision requires the Secretary of State to review her policy from time to time, whether with regard to the best interests of children or otherwise. (Footnote: ¹⁷) I cannot see how the fact that she may from time to time confirm (typically in response to pressure from parliamentarians or campaign groups) that she does not intend to change her position should convert the maintenance of the status quo into the discharge of a function within the meaning of section 55.

It may be a little over-simple to regard the distinction in question as being between action and inaction, since Ms Giovannetti concedes – I think correctly – that in some circumstances a decision to maintain the status quo will constitute the discharge of a function. But in my view her contention that that will only be so where the decision is the result of a substantive review process makes sense: it is the process of substantive reconsideration, embodied in a formal process, that makes it appropriate to describe the Secretary of State as discharging a function. In short, she will only have discharged a relevant function within the meaning of section 55 if she made a decision not to change the Rules (or policy expressed in guidance) as a result of a substantive review process in the sense identified at para. 103 above: she will not have done so if her only decision was not to embark on such a process.

On that basis, I also believe, for the reasons given by Ms Giovannetti, that Lavender J was right to find that the Secretary of State had not at any material time discharged a relevant function as regards family reunion for child refugees. Although it is unsatisfactory that we do not have fuller evidence about the 2016 and 2018 “decisions”, it is in my view reasonably clear that neither constitutes a decision of the necessary character. I take them in turn.

As regards the 2016 episode, I agree that it is reasonable to infer that the content of the Guidance was changed in response to the decision in AT; and that no doubt involved a decision of some kind, possibly taken or approved by a Minister. But the question is whether the making of that change justifies the inference that there was a substantive review of the policy as regards family reunion for child refugees. The starting-point must be para. 10 of the written submissions quoted at para. 83 above, confirmed by Mr Büültjens, which says explicitly that no such review took place. It would be possible to reject that evidence if the only possible explanation of the changes in the 2016 Guidance were that there had been a substantive review by the Secretary of State of the necessary kind. But the introduction of the key passages in 2016 is equally explicable on the basis that AT had revealed a need, not for a substantive review of the Secretary of State’s policy, but only for that policy to be clearly stated in the Guidance. It is in fact prima facie most unlikely that the Secretary of State had not previously had a policy of permitting family reunion where that was necessary in order to avoid a breach of article 8 and/or in compelling and compassionate circumstances. I accept that at para. 11 of his decision McCloskey P says that a previous policy of broadly that kind was “extinguished” in 2006, but he gives no particulars and what he says not only is unlikely to be the case but is inconsistent with Mr Büültjens’ evidence. In short, this episode does not justify an inference that the Secretary of State discharged a relevant function in 2016.

As for the 2018 decision, the statements from the Home Office that it was “currently reviewing the policy on refugee Family Reunion” and “reviewing the approach to Family Reunion” do not by themselves get the Claimant home. “Reviews” can be of a variety of kinds and for a variety of purposes. It appears from the Chief Inspector’s 2018 report (see para. 5.36, quoted by the Judge in para. 57 of his judgment) that the relevant teams within the Home Office aimed to “review” family reunion policy every twelve months. It is also unsurprising that officials should “review” the policy in response to criticisms from parliamentarians and others. But it does not follow that such reviews would develop into proposals for a change of policy to be considered by the Secretary of State; and the evidence is that throughout the relevant period the Secretary of State made it clear that she was not prepared to consider any such change.

Although I am content to base that conclusion on the material that was before Lavender J when he delivered his first judgment, there is also relevant evidence in the witness statement of Dr Meirav Elimelech, Deputy Director of the Asylum and Protection Unit in the Home Office (filed following the first judgment, for the purpose of ground 3). In particular, at para. 8 she refers to a “compelling, compassionate circumstances” policy being in place in 2007, which is inconsistent with McCloskey P’s statement that such a policy was “extinguished” in 2006. And at para. 12 she says that

“Home Office ministers have been consistently clear with officials that they do not wish to amend the policy position with regards to children sponsoring parents or other family members under the family reunion policy”.

That reinforces the Chief Inspector’s reference to a “ministerial red line”.

I am not shaken in that conclusion by the terms of the Treasury Solicitor’s responses to the RFI. I accept that, if on the Secretary of State’s case there were indeed no decisions by her of the kind identified by the Claimant as the premise for his request, the response ought to have been prefaced by a challenge to the validity of that premise. But it is much more likely that the failure to include such a challenge is the result of the author failing to appreciate the potential significance of the introductory parts – the request was on the face of it simply concerned with dates – than that it represents a considered acceptance of a case which was contrary to the finding of the Judge and to the Secretary of State’s consistent case throughout.

For the process of the reasoning above, I have remained loyal to the parties’ common position that the crucial question is whether the Secretary of State’s alleged decisions not to review the relevant Rules constituted the discharge of a function. But I would add that if, as I suggest at para. 69 above, the making of Immigration Rules is treated by the statute as the making of “arrangements” rather than as the discharge of a relevant function, the answer is the same, if not indeed a fortiori: I do not see how a decision not to review the Rules can possibly be regarded as the making of an arrangement.

I have reached that conclusion without reference to authority. We were in fact referred to three authorities in connection with this issue. None of them seems to me of central importance, but I will address them briefly. I take them in date order.

In R (Badmus) v Secretary of State for the Home Department [2020] EWCA Civ 657, [2020] 1 WLR 4609, the claim concerned the lawfulness of the amounts paid to immigration detainees for work voluntarily undertaken by them. The impugned rates were first set by a detention service order (a “DSO”) made in 2013, but they were maintained by a decision of Home Office Ministers following a review in 2018. It was important for limitation reasons to establish whether the only judicially reviewable decision was made at the time of the original DSO. This Court (Sir Terence Etherton MR, Hickinbottom LJ and Simler LJ) held that the 2018 review gave rise to a distinct decision. I agree with Lavender J that the facts were fundamentally different from those of the present case; and the outcome is entirely consistent with Ms Giovannetti’s submissions. As appears from paras. 11-15 of the judgment of the Court, the 2018 decision was the product of a formal review process commissioned by Ministers which produced a report offering them a choice between four options, including (which was the option chosen) to leave the rates unchanged. The case was not concerned with section 55, but it is clear that a decision taken as a result of a process of that kind would indeed be the discharge of a function in the relevant sense.

InR (Adiatu) v Her Majesty’s Treasury [2020] EWHC 1554 (Admin), [2020] PTSR 2198, the claimant contended, among other things, that in deciding not to accord benefits to the self-employed under the Coronavirus Job Retention Scheme the Government failed to comply with the PSED established by section 149 of the Equality Act 2010. The claim was rejected by the Divisional Court (Bean LJ and Cavanagh J). Among other things, it held that the PSED only obliged an authority “to have regard to the equalities implications of the decision that is actually taken, not other decisions that might have been taken instead”. At para. 242 of its judgment the Court said:

“The ‘exercise of the [public authority’s] functions’ for the purposes of s 149 (1) consists of the implementation of the measures that the public authority decides upon. … A public authority must have regard to the equalities implications of the steps that it intends to take. It need not have regard to the equalities implications of other steps, which it is not taking, and is not even considering.”

I respectfully agree with that, and it is broadly in line with Ms Giovannetti’s submission; but the Court was not directly considering the question which arises in the present case.

In R (the 3million Ltd) v Minister for the Cabinet Office [2021] EWHC 245 (Admin) the claimant contended, among other things, that in omitting to make certain amendments to regulations governing the registration of EU nationals who wished to vote in the 2019 European Parliamentary elections the defendant minister failed to comply with the PSED. The Divisional Court (Lewis LJ and Supperstone J) did not in the end need to reach a decision on that contention, but at para. 130 of its decision dismissing the claim it expressed some doubt whether the duty arose: it acknowledged  that “if regulations are made, and quite possibly, when the issue of whether to amend regulations or not is being actively considered, a minister may be exercising functions”, but pointed out that on the facts the issue did not appear ever to have been actively considered. The passage in question may be the origin of Lavender J’s reference in the present case to “active consideration” but it does not otherwise advance the argument. The Divisional Court gives no detailed consideration as to the kind of process that would be necessary in order to give rise to the exercise of a function.

I should record, finally, that in her Respondent’s Notice the Secretary of State took the point that any claim based on decisions taken in 2016 or 2018 would be out of time and contends that permission to extend time should not be granted. That objection was not developed in either Ms Giovannetti’s skeleton argument or her oral submissions, but in any event on the basis of my conclusion above it does not arise.

THE NEW CASE

The Claimant’s new case, as pleaded in the Amended Grounds of Appeal, is that “the Respondent exercised a relevant function … when she operated her overall system for refugee family reunion”. On that basis, he says, the Secretary of State has unquestionably failed to comply with section 55 because it is common ground that she has never made arrangements for ensuring that the policy governing family reunion for child refugees, as opposed to decision-making in individual cases, has had regard to the best interests of children.

The submissions of Mr Pobjoy, who argued this part of the case for the Claimant lucidly and intelligently, can be summarised as follows. He candidly acknowledged that the new case might at first sight appear ambitious, because in the great majority of cases it was sufficient to analyse the relevant functions of the Secretary of State as comprising the initial establishment of the system, in the form of the relevant rules and guidance, and taking individual decisions within that framework, so that there was no need to rely on a distinct function of “operat[ing] her overall system”. However, he submitted that it was nevertheless necessary to recognise the existence of a broader function of that character because otherwise in at least some cases the evident statutory purpose would not be achieved. He gave two examples. The first was where there had been a material change of circumstances since the system was first established: as he put it, “you cannot just set up a system and forget about it”. The second, which was the case here, was where the system was designed prior to the coming into force of section 55 and had not been the subject of a section 55 exercise on its implementation. In cases of this kind it was necessary to treat the operation of the system as itself a function as a “guard-rail” to ensure that it always took into account the best interests of children.

Mr Pobjoy submitted that that approach was supported by a number of authorities. He referred us to six cases, concerned almost exclusively with the PSED, but he relied principally on the judgment of this Court (Sir Terence Etherton MR, Dame Victoria Sharp P and Singh LJ) in R (Bridges) v South Wales Police [2020] EWCA Civ 1058, [2020] 1 WLR 5037. That case concerned the deployment by the South Wales Police (“the SWP”), on what was said to be a trial basis, of a system of automatic facial recognition (“AFR Locate”, or simply “AFR”). The use of the system, both generally and on two particular occasions, was said to give rise to breaches of article 8 of the ECHR, the Data Protection Act 2018 and the PSED. The Court held that there had been a breach of the PSED. The relevant part of its judgment is at paras. 163-202. The SWP had in fact carried out an equality impact assessment (“the EIA”) before AFR Locate was deployed, but the claimant made clear that his essential challenge was to what was described (see para. 167) as a “continuing failure to discharge the PSED [italics in original]”: his case, as I understand it, was that the EIA was necessarily based on predictions of how AFR would operate but that it was inevitable that new information would be acquired in the course of actual operation. The Court accepted that submission. Most of its reasoning is in fact directed to other issues, but it quoted with evident approval item (4) in McCombe LJ’s well-known summary of the effect of section 149 inR (Bracking) v Secretary of State for Work and Pensions [2013] EWCA Civ 1345, namely that “the duty is a continuing one”.

I should refer in more detail to one passage in Bridges on which Mr Pobjoy relied. At para. 200 of its judgment the Court says:

“Finally, we would note that the Divisional Court placed emphasis on the fact that SWP continue to review events against the section 149(1) criteria. It said that this is the approach required by the PSED in the context of a trial process. With respect, we do not regard that proposition to be correct in law. The PSED does not differ according to whether something is a trial process or not. If anything, it could be said that, before or during the course of a trial, it is all the more important for a public authority to acquire relevant information in order to conform to the PSED and, in particular, to avoid indirect discrimination on racial or gender grounds.”

The observation by the Divisional Court to which that paragraph is directed is to be found at para. 158 of its judgment ([2019] EWHC 2341 (Admin), [2020] 1 WLR 672). It is not entirely clear what point the Court was making there or, therefore, what point this Court was seeking to address in para. 200. However, Mr Pobjoy relies on the statement that it did not matter whether the introduction of AFR was by way of trial or not: in either case the SWP was required to “continue to review events” – to put it another way, to monitor its effects – to ensure that due regard continued to be had to the section 149 criteria.

Mr Pobjoy submitted that there was in this respect no real difference between the PSED and the section 55 duty. The PSED is a process duty (a point explicitly made in para. 176 of the judgment in Bridges), but the Court clearly regarded it not just as applying to the decision to introduce AFR Locate but also as imposing obligations so long as it remained in operation: if the reference in section 149 (1) to “the exercise of [the] functions” of a local authority could be read as covering the monitoring of a system, so also could and should the reference in section 55 (1) (a) to the discharge of the Secretary of State’s functions.

The other cases relied on by Mr Pobjoy were R (DMA) v Secretary of State for the Home Department [2020] EWHC 3416 (Admin), [2021] 1 WLR 2374, (Robin Knowles J); R (Elkundi) v Birmingham City Council [2022] EWCA Civ 601, [2022] QB 604, (Court of Appeal); R (DXK) v Secretary of State for the Home Department [2024] EWHC 579 (Admin), [2024] 4 WLR 46, (Paul Bowen KC sitting as a deputy High Court Judge); R (Refugee and Migrant Forum of Essex and London (RAMFEL)) v Secretary of State for the Home Department [2024] EWHC 1374 (Admin), [2024] 1 WLR 4950 (Cavanagh J); and R (Awale) v Secretary of State for Justice [2024] EWHC 2322 (Admin) (Ellenbogen J).I do not propose to review them individually. Only one – Elkundi – is a decision of this Court, and it is far from clear whether the short observation in para. 114 of the judgment of Lewis LJ on which Mr Pobjoy relied was directed to the question before us. In response to questions from the Court Mr Pobjoy did not press his reliance on DXK (Footnote: ¹⁸); and the passage to which he referred us in Awale does no more than summarise the relevant passages in Bridges. DMA and RAMFEL are rather more substantial. In the former Robin Knowles J found a breach of the PSED on the basis of a failure by the authority in question to monitor the operation of the policy in question (see paras. 311-312 and 324-325 of his judgment); and in RAMFEL counsel for the Secretary of State conceded that the PSED was “engaged” in circumstances where the Secretary of State had simply maintained a long-standing policy (para. 225). But the circumstances in which the question arose were very different from those of the present case and it is not clear exactly what points were argued. It is fair to say that in broad terms all these cases show an understanding that a breach of the PSED may occur, at least in some circumstances, by reason of a failure to monitor the ongoing effect of a policy or practice which was PSED-compliant at the point of its introduction; but they contain no consideration of how that position is to be analysed by reference to the language of section 149 (1), and accordingly they do not advance the argument beyond what appears from Bridges.

Ms Giovannetti started by noting that the structures of section 55 (1) of the 2009 Act and section 149 (1) of the 2010 Act are not identical: as to this, see para. 71 above. She nevertheless accepted that the duty under section 55 could, like the PSED, be continuing in character. Her case, however, was that whether that was so in a particular case was context-specific: more particularly, it depended on whether the function to which it related was itself continuous in character or was a one-off. At paras. 53-56 of her skeleton argument she commented briefly on each of the individual authorities relied on by the Claimant, with a view to establishing that they were consistent with that analysis. The only one of those cases to which she specifically referred in her oral submissions was Bridges, which she said was a case where there was an ongoing duty to monitor because AFR Locate was in the nature of a trial. The present case was different. The Secretary of State was not discharging any continuous function. She had discharged the relevant function by putting in place a policy, prior to 2 November 2009, and had (assuming that the original case is rejected) exercised no distinct function since then.

Although the new case is framed in terms of the relevant function being “the operation of the system”, the substance of Mr Pobjoy’s submission was that the Secretary of State is under an obligation to review her policies in the field of immigration (whether expressed as Rules or guidance) from time to time or in any event in response to a material change of circumstances. I see difficulties with that submission. Mr Pobjoy made it clear, in response to questions from the Court, that it was his case that such an obligation was implicit in section 55 (1) itself, but the way that section 55 works is by reference to existing functions rather than by the creation of new ones. If it were sought to get round that difficulty by relying on a general public law duty to review existing policies, that would raise very wide issues on which we were not addressed.

It is true that Bridges appears to lend support to Mr Pobjoy’s argument, but I am not sure that it is decisive. Even assuming for these purposes that the differences in the drafting of section 149 (1) and section 55 (1) are immaterial, I see force in Ms Giovannetti’s argument that in the particular circumstances of that case the authority was under a duty to monitor the operation of AFR Locate because it was a new technology being introduced avowedly by way of trial. At first sight that argument might appear to be contradicted by para. 200 of the judgment. However, that passage is rather opaque, and I am inclined to think that the Court’s point was only that the potential equality impacts of AFR required ongoing monitoring because of its inherent nature and/or its novelty, irrespective of whether it was explicitly labelled as a trial.

However, I do not believe that I need to reach a concluded view on this aspect of the case, and I prefer not to do so. In my view the Claimant is entitled to succeed on the basis of Mr Pobjoy’s much more particular argument that the Secretary of State was obliged at the point that section 55 came into force – that is, on 2 November 2009 – to consider whether her policy as regards family reunion for child refugees had regard to the best interests of children. As at that date it became the Secretary of State’s duty to make arrangements to ensure that individual immigration decisions which provided the framework for those decisions were taken with regard to the best interests of children. In my view those arrangements necessarily involved ensuring that the Rules and policy guidance applying to such decisions were themselves formulated with regard to the best interests of children. That way of putting it, focusing on the language of making arrangements, seems to me the most natural fit with the statutory language; but if it is necessary to use the language of the exercise of a function the point of substance is the same. Mr Pobjoy’s overall label of “the operation of the system” may not be particularly apt to this aspect of his case, but what matters is that the Secretary of State was obliged as at 2 November 2009 to conduct, or have conducted, an exercise that ensured that her family reunion policy had regard to the best interests of children.

If the Secretary of State was, as I believe, under such an obligation it is clear that she did not discharge it. It has never been part of her case that she conducted a section 55 exercise as regards the policy at or around the time of the coming into force of the section, and, as we have seen, it is a positive part of her case that she has not conducted a substantive review since then. Since the obligation has never been discharged, it remained in force up to and including the date of Lavender J’s judgment.

The explanation of what may seem to be a surprising omission on the Secretary of State’s part seems to be that she took the view in 2009 that it was a sufficient discharge of her duty under section 55 (1) to issue the guidance in Every Child Matters; and indeed at para. 58 of her skeleton argument (though not in her oral submissions) Ms Giovannetti advanced that as a fallback answer to the Claimant’s case. But, as Lavender J pointed out, the fact that the decision-makers in individual cases are required to have regard to the best interests of the child does not address the Claimant’s case that the policy which they are obliged to implement (whether expressed in the Rules or in guidance) must also be formulated with regard to those interests: see paras. 78 (first bullet) and 79 above.

For those reasons, I would allow ground 1 of the appeal, though only on the basis of an argument not advanced before the Judge.

It follows, if My Lords agree with that conclusion, that the Secretary of State is obliged now to conduct a section 55 exercise as regards her family reunion policy as it relates to child refugees; and the normal form of relief would be to make a declaration having that effect. It is important to emphasise that it does not necessarily follow that she would be obliged to change that policy. The duty is a process duty; and if, having gone through the necessary process, she concluded that the policy is in the best interests of children in the UK or that, to the extent that it is not, those interests are outweighed by other primary considerations, that would discharge her duty under section 55.

The effect of the Secretary of State’s announcement of 4 September (see para. 12 above) is that she is now planning in any event to conduct a comprehensive review of her family reunion policy, which will presumably cover the position of child refugees. It may be that the terms of our order should take account of that circumstance. I would give the parties the opportunity following the hand-down of this judgment to consider precisely what form of order is appropriate and to file written submissions if they cannot agree.

Article 14 of the ECHR provides as follows:

“The enjoyment of the rights and freedoms set forth in this Convention shall be secured without discrimination on any ground such as sex, race, colour, language, religion, political or other opinion, national or social origin, association with a national minority, property, birth or other status.”

The effect of that provision is that discrimination is outlawed as regards matters falling within the ambit of one of the substantive Convention rights. As already noted, in the present case the Claimant relies on article 8, which confers the right to respect for private and family life.

At para. 136 of her judgment in R (DA) v Secretary of State for Work and Pensions [2019] UKSC 21, [2019] 1 WLR 3289, Lady Hale summarised the questions raised by a claim under article 14 as follows:

“In deciding complaints under article 14, four questions arise: (i) Does the subject matter of the complaint fall within the ambit of one of the substantive Convention rights? (ii) Does the ground upon which the complainants have been treated differently from others constitute a ‘status’? (iii) Have they been treated differently from other people not sharing that status who are similarly situated or, alternatively, have they been treated in the same way as other people not sharing that status whose situation is relevantly different from theirs? (iv) Does that difference or similarity in treatment have an objective and reasonable justification, in other words, does it pursue a legitimate aim and do the means employed bear ‘a reasonable relationship of proportionality’ to the aims sought to be realised (see Stec v United Kingdom (2006) 43 EHRR 47, para 51)?”

It was common ground before Lavender J that the first two questions were to be answered in the Claimant’s favour – specifically, (i) that the effect of the policy fell within the ambit of article 8; and (ii) that being a child refugee was an “other status”. The issues related to the third and fourth questions – “differential treatment” and “justification”.

As regards differential treatment, this can, as Lady Hale’s formulation recognises, be of two kinds – (1) treating persons in (relevantly) similar situations differently and (2) treating persons who are in (relevantly) different situations the same. I will refer to the first kind as “primary discrimination” and to the second by the usual label of Thlimmenos discrimination (the reference being to the decision of the ECHR in Thlimmenos v Greece 34369/97, [2000] ECHR 162) (Footnote: ¹⁹).At the risk of spelling out the obvious, the labels “primary discrimination” and “Thlimmenos discrimination” do not mean that the discrimination in question is necessarily unlawful: that will only be the case if it is not shown to have been justified.

Authoritative guidance on the correct approach to claims under article 14 is now to be found in the judgment of Lord Reed in R (SC) v Secretary of State for Work and Pensions [2021] UKSC 26, [2022] AC 223. For present purposes I need only note para. 37, which reads:

“The general approach adopted to article 14 by the European court has been stated in similar terms on many occasions, and was summarised by the Grand Chamber in the case of Carson v United Kingdom (2010) 51 EHRR 13, para 61 (‘Carson’). For the sake of clarity, it is worth breaking down that paragraph into four propositions:

(1)

‘The court has established in its case law that only differences in treatment based on an identifiable characteristic, or “status”, are capable of amounting to discrimination within the meaning of article 14.’

(2)

‘Moreover, in order for an issue to arise under article 14 there must be a difference in the treatment of persons in analogous, or relevantly similar, situations.’

(3)

‘Such a difference of treatment is discriminatory if it has no objective and reasonable justification; in other words, if it does not pursue a legitimate aim or if there is not a reasonable relationship of proportionality between the means employed and the aim sought to be realised.’

(4)

‘The contracting state enjoys a margin of appreciation in assessing whether and to what extent differences in otherwise similar situations justify a different treatment. The scope of this margin will vary according to the circumstances, the subject matter and the background.’”

Lavender J addressed the issue of differential treatment at paras. 148-156 of his first judgment. He concluded that neither primary nor Thlimmenos discrimination had been established.

As regards primary discrimination, at para. 154 he says:

“I do not accept … that the matters complained of in this case constitute a difference of treatment for the purposes of Article 14. The relevant Immigration Rules treat child and adult refugees the same: neither child nor adult refugees are permitted by the relevant Immigration Rules to sponsor applications for leave to enter by their parents or siblings.”

That statement needs a little amplification. Lavender J is right to say that under the Rules neither child nor adult refugees could sponsor their parents or siblings. That is not the whole picture because under the guidance they were entitled to do so in exceptional circumstances; and that has since become the position under the Rules themselves. However, that does not affect his basic point: the Secretary of State’s policy as regards reunion with parents and siblings was the same for both child and adult refugees.

At para. 155 he observed that Mr Husain had not explicitly relied on Thlimmenos discrimination, but he said that in substance his case appeared to be that child refugees were in a relevantly different position from adult refugees because their “nuclear family” consisted of their parents and/or siblings whereas an adult refugee’s nuclear family consisted of their partner and children and that that difference required different treatment. He rejected the case so formulated at para. 156, where he said:

“… I do not find this to be a helpful way of looking at the matter. The concept of ‘nuclear family’ is not as clear-cut as Mr Husain suggested. For instance, some adult refugees will be individuals who are 18 or more years old and who may have been living with their parents as dependent relatives before they left their country of habitual residence. For these adult refugees, their nuclear family may well consist of their parents and siblings, so that their situation is substantially similar to that of many child refugees.”

He goes on to amplify that point, but I need not reproduce the whole passage.

As a result of that conclusion Lavender J did not go on to consider the question of justification.

Ground 2 of the Amended Grounds of Appeal reads, so far as material:

“The Administrative Court erred in holding that there was no relevant difference in treatment, for the purposes of Article 14 ECHR, between refugee children seeking to reunite with their parents and minor siblings on the one hand and adult refugees seeking to reunite with their partners and minor children on the other; and/or in holding that refugee children seeking to reunite with their parents and siblings were not in a relevantly different situation from adult refugees seeking to reunite with their parents and siblings; and therefore in failing to consider the issue of justification.”

Thus the Claimant continues to rely on both primary and Thlimmenos discrimination.

In her Respondent’s Notice the Secretary of State relied so far as necessary on her case that any differential treatment was justified.

I will take the differential treatment and justification issues in turn.

DIFFERENTIAL TREATMENT

I believe, with respect, that Lavender J was wrong to find against the Claimant on the differential treatment issue, whether the case is analysed as one of primary or Thlimmenos discrimination.

I start with primary discrimination. In my view a minor child who has been granted asylum in the UK and who seeks to be joined by a parent is in a relevantly similar, or analogous, situation to a parent who has likewise been granted asylum and who seeks to be joined by a minor child. In both cases the relationship of parent and minor child has been disrupted by their separation, and the aim of family reunion is to restore it. I cannot see that it makes a difference in principle whether it is the child who is in the UK and the parent who is still abroad or vice versa. Yet the two situations are treated differently: in the former reunion will be possible only if the unjustifiably harsh consequences criterion is satisfied, whereas in the latter it will be automatic.

Lavender J’s reasoning at para. 154 depends on the fact that parents are not eligible for automatic family reunion whether the sponsor is a child or an adult. But the two situations are not the same. As pointed out at para. 21 above, the concept underlying the treatment of minor children under the Rules is dependency: a child is presumed to be dependent on his or her parents up to the age of 18 but not thereafter. Thus where a child and an adult refugee are both seeking reunion with a parent, the relevant relationships are fundamentally different: one is a relationship of (presumed) dependency and the other is not.

The point made in the last paragraph also provides the basis for a Thlimmenos analysis. For the reason given, a child refugee and an adult refugee seeking reunion with a parent are in different situations; yet both are treated the same. Lavender J’s point that in some cases there will be adult refugees for whom their parents and siblings constitute a “nuclear family” may be an answer to a particular way in which Mr Husain expressed himself below; but it does not address the fact the Rules recognise and apply the well-recognised conventional distinction between children and adults as regards dependency on their parents.

In the foregoing paragraphs I have referred only to the parent-child relationship. That is appropriate because in the great majority of cases siblings will be granted leave because leave is granted to the parent: see para. 45 above.

JUSTIFICATION

The burden of proving justification is on the Secretary of State. Her case on justification was pleaded at para. 42 of the DGD as follows:

“Any difference in treatment between groups in an analogous position is justified in any event. Age is not a suspect reason (nor, indeed, is the ‘status’ on which the Claimant relies). Furthermore, this is an issue of social policy where a wide margin should be accorded to the state: SC, paras 115(2), 129(2), 143-161, 203. The SSHD has explained on numerous occasions, including in Parliament, why she has adopted the policy she has. Plainly, this is not a matter capable of empirical proof; rather, it is a matter of judgement. Against that background, any difference in treatment has objective and reasonable justification.”

No reference is given for the “numerous” explanations of her policy, but para. 11 of the DGD quotes two statements by a Minister in the House of Lords and the Government response to the first report of the HL EU Committee.

The Secretary of State’s skeleton argument below closely followed the DGD but it was rather fuller. Paras. 42 and 43 read:

“42.

The SSHD has legitimate concerns about the welfare of child refugees. The SSHD’s consistent policy judgment has been that allowing children in this position to become sponsors could have the unintended consequence of incentivizing some families to send their children alone on long and dangerous journeys to the United Kingdom to act as ‘anchors’ to facilitate later asylum applications by family members. The SSHD is especially concerned that such journeys put children at risk of being trafficked or otherwise abused. Contrary to the Claimant’s claim, there is evidence to support the SSHD’s position: see, for example, an ‘ad-hoc query’ sent by Belgium to other members of the European Migration Network in which it asked several questions in relation to the ‘anchor child phenomenon’ which it described as ‘increasing’ in Belgium … Germany, Cyprus, the Netherlands, Sweden, Finland and Norway all replied to the effect that they recognised the phenomenon in their own States.

43.

But in any case, this is not a matter which is readily capable of empirical proof; rather, it is a matter of judgment. The SSHD’s judgment on this issue is not manifestly without reasonable foundation: to the contrary, it is entirely reasonable. Against this background, any difference in treatment between adult refugees and child refugees is justified.”

The Secretary of State filed no witness evidence in support of that case in advance of the first hearing, and the later witness statements of Mr Büültjens and Dr Elimelech do not address it. Concerns about the anchor child phenomenon were, however, referred to in various contemporary documents which were before the Court, including the Secretary of State’s response to the report of the Chief Inspector referred to at para. 53 above, the various ministerial statements referred to in the DGD, and the ad hoc query addressed by the Belgian government to the EMN (“the Belgian EMN query”) to which the skeleton argument refers and the responses to it.

Of those references, the Belgian EMN query, which is dated 8 June 2017, is the most substantial source of evidence for the existence of the anchor child phenomenon. The preamble to the request (headed “background information”) reads:

“The Belgian Immigration Office notes an increasing number of unaccompanied minors who lodge an asylum application in Belgium and for which it, soon afterwards, is requested (mostly by the Greek authorities) to take over the parents or other family members in the framework of Dublin III. The Belgian authorities are concerned about these young children who are left alone at the hands of smugglers and they wish to know whether other Members States are confronted with this phenomenon (‘unaccompanied minors who lodge an asylum application in the Member State and for which the authorities, soon afterwards, are requested to take over the parents and other family members’, sometimes referred to as ‘anchor child phenomenon’).”

The questions read:

“1.

Are your authorities confronted with the so called ‘anchor child phenomenon’? If so, please describe the overall phenomenon that you experience

2.

If yes to the first question: In which procedure is this phenomenon visible: Dublin, other?

3.

If yes to the first question: Which age group do the children fall into?

4.

If yes to the first question: Which nationalities are concerned by this phenomenon?

5.

If yes to the first question: Has you recently noticed an increasing number of such cases? Please provide estimates

6.

Which measure do you have in place to prevent and tackle the so-called ‘anchor child phenomenon’? Please elaborate on when the measure was taken, possible impacts etc.

The Belgian Immigration Office answered those questions itself in some detail. The answer to question 1 echoed the preamble, but I should also note answer 3, which reads:

“The Belgian Immigration Office notices that the children concerned are increasingly younger of age. The youngest was 3 years of age. A large portion of them is under 12 years old. These children are extremely vulnerable during their illegal trip from Greece to Belgium. Most parents pay smugglers to get them to Belgium. Younger children cost less to smuggle.”

As the Secretary of State notes, Germany, Cyprus, the Netherlands, Sweden, Finland and Norway all answered that they recognised the phenomenon (as also in fact did Austria, although it did not use the label). I need not summarise the details of their answers to the other questions.

Before us Ms Giovannetti submitted that the Secretary of State’s policy of denying automatic family reunion to child refugees represented a proportionate means of achieving the legitimate objective of disincentivising the sending of unaccompanied children on a hazardous journey initially (in most cases) across the Mediterranean and then across Europe and finally in small boats across the Channel: that was an objective peculiar to the case of child refugees and justified the adoption of a different policy from that applied to adults. As regards proportionality, she emphasised that the difference in treatment was not that family reunion was denied to child refugees but only that it was considered on a case-by-case basis, applying the “unjustifiable hardship” criterion, rather than granted automatically. A case-by-case approach was not inherently inappropriate: the circumstances of unaccompanied child refugees varied greatly, and not all, particularly those nearing the age of 18, would be truly dependent on their parents. Those points were summarised in the introductory section of her skeleton argument (at para. 2) as follows:

“The Respondent recognises that it will usually be in the best interests of children not to remain separated from their parents and siblings. However, the strength of that consideration will vary from case to case, depending on the age, maturity, personal history and life experiences of the child, and all the individual circumstances of the case. Conversely, the Respondent considers that, in general terms, there is a legitimate concern not to encourage or incentivise families to send a child on a long and dangerous journey to claim asylum, with the expectation that the rest of the family will then be entitled to join them. The best interest considerations do not all point in the same direction, nor will they be of equal force in all cases. The current system is a legitimate response to that tension.”

Ms Giovannetti also reminded us of factors which she said required a less intensive standard of review. These were enumerated at para. 62 of her skeleton argument as follows:

“(i)

This is a systemic challenge and so the question is whether the scheme is incapable of being operated in a proportionate manner: R (JCWI) v SSHD [2021] EWCA Civ 542; [2021] 1 WLR 1151 at [117-119]; MM (Lebanon) v SSHD [2017] UKSC 10, [2017] 1 WLR 771 at [55]-[58].

(ii)

The Court should afford a wide margin of appreciation to the state: age is not a core status or suspect ground (whether or not in conjunction with refugee status) and the more peripheral a status is, the lower the burden of justification: Mathieson v SSWP [2015] UKSC 47, [2015] 1 WLR 3250 at [21]; R (Akbar) v SSJ [2019] EWHC 332 (Admin) at [92].

(iii)

Immigration control is a matter of socio-economic policy, and therefore a matter in respect of which a particularly wide margin is appropriate: see JCWI at [134(iv)].”

The Claimant’s overall case is most conveniently summarised in the introduction to a schedule put in by Mr Husain in the course of the hearing setting out various references relevant to the issue of justification (“the justification schedule”). It reads:

“The burden of justification is on R. R is required to show that, balancing the severity of the effects of her differential treatments of refugee children against the importance of her objective, to the extent that the differential treatment will contribute to its achievement, the former outweighs the latter. The sources below establish that there is no evidence that the differential treatment of refugee children furthers R’s sole objective: namely, deterring children who would otherwise (if a straightforward path to family reunion were made available) undertake hazardous journeys to the UK. The absence of such evidence over a 25-year period, despite R’s efforts to obtain it, founds a strong inference that the contribution to R’s objective is non-existent, or (at best) minimal. The question for the Court is whether, in light of this, R’s view of the deterrent effect – unquantified, unevidenced, and dubious as it is – is sufficient to justify the well-established and extensive harms the differential treatment causes to vulnerable refugee children. A’s submission is that, on any standard of review (and certainly on the searching standard appropriate here), it is not.”

Those points were developed in Mr Husain’s oral submissions. He also made the point that there was no evidence that the Secretary of State had herself ever carried out any evaluation of the balance between her objective and the identified harms.

The justification schedule identifies the evidence on which Mr Husain relied on support of his submissions. This consisted of the evidence of his own and UNHCR’s witnesses, and the reports exhibited by them, as enumerated at para. 55 above, together with the Parliamentary materials and the report of the Chief Inspector. That evidence covers the elements that go into both sides of the proportionality balance. On the one hand, it questions the reality of the phenomenon of anchor children relied on by the Secretary of State. On the other, it contains accounts of the distress and difficulty which it is said that the absence of a straightforward route to family reunion causes to refugee children.

For reasons which will appear, I do not propose to go into further detail about that evidence or to attempt to assess its weight. However, it is fair to record that at paras. 69-70 of his first judgment Lavender J said that it was not seriously contested by the Secretary of State that:

“(1)

in general, it is in the best interests of unaccompanied refugee children to be reunited with their families; and

(2)

in general, it is in the best interests of unaccompanied refugee children to have a straightforward path to that result.”

Thus the focus of the dispute has mostly been on the other side of the balance – that is the weight to be given to the claimed risk that allowing automatic family reunion for child refugees would encourage the use of anchor children.

Normally in circumstances where the judge did not address an issue which has become live as a result of the appeal we would determine the issue ourselves, unless it had to be remitted so that findings could be made on disputed issues of fact. But the present case is unusual because of the combination of a number of circumstances.

First, our decision on ground 1 means that the Secretary of State will be obliged in any event to reconsider her policy as regards family reunion for child refugees. And in fact her recent announcement means that her entire policy as regards family reunion is suspended pending the introduction of significant changes. There is, to put it no higher, a real risk that any conclusion that we reach now on the justifiability of the current policy will be academic.

Second (though this may in truth be another aspect of the same point), Mr Husain made it clear that the only relief that he was seeking on ground 2 was a declaration that the rules in their current form were discriminatory, on the basis that the Secretary of State would then conduct a reconsideration: he conceded that that reconsideration might in principle result in her maintaining them in their current form. Such a reconsideration will now be required in any event. It is possible that if we determined ground 2 our conclusion, or the underlying reasoning, might be of assistance to the Secretary of State for the purpose of her reconsideration; but that is not obviously the case, particularly in view of the passage of time since the first hearing before the Judge (see para. 165 below).

Third, the fact that Lavender J did not himself address the issue of justification means that we are deprived of the benefit of his own analysis and judgment. That might not normally be a sufficient reason by itself for our not deciding the issue, but it is a relevant consideration.

Fourth, the state of the evidence on the issue of justification is unsatisfactory. It is rather surprising that the Secretary of State filed no witness statement setting out her case on justification and the evidence relevant to it: indeed in her Summary Grounds of Defence she had indicated that she might adduce such evidence, but that did not happen. It would be unfortunate if this Court had to proceed to a decision on the basis of piecemeal material such as the Belgian EMN request and the responses to it. When we asked Ms Giovannetti why there had been no witness statement on this issue she referred to the confused way in which the issues had developed, and also to her submission that the degree to which making automatic family reunion available for child refugees would incentivise the use of anchor children was a matter of judgement and not readily capable of empirical proof. I see some force at least in the latter point, and I do not say that the Secretary of State’s case could not in principle be proved without witness evidence. But I remain troubled by its absence. At the very least, it would have provided a coherent presentation of the materials before us, identifying the conclusions which the Secretary of State drew from them; but it may also be that the exercise of preparing a witness statement would have produced material that enabled the Court to arrive at a more solidly-based decision. (It would also avoid a potential issue of Parliamentary privilege to the extent that she wishes to rely on Ministers’ responses in Parliament.) If a determination on this issue were necessary in order to establish the Claimant’s entitlement to some substantive relief, we might have to do our best on the basis of such evidence as there was; but that would be unsatisfactory, especially on an issue of this importance.

Fifth, the unfortunate procedural history of this case means that over three years have passed since the first hearing before the Judge, and longer than that since the creation of the various pieces of evidence which were before him on this issue; and it is appropriate to take judicial notice of the fact that there have been significant developments relevant to immigration policy during that time. Quite apart from the particular points made above, a judgment, either way, on the justification issue based on that evidence is likely to be of limited value either generally or for the purpose of the Secretary of State’s reconsideration.

The effect of those factors in combination is that I do not believe that this Court ought itself to determine the issue of justification. The choice is between, on the one hand, dismissing the discrimination claim altogether on the basis that its determination is no longer of practical value and, on the other, remitting the issue to the Administrative Court, possibly coupled with a stay pending the outcome of the Secretary of State’s reconsideration. I would give the parties the opportunity to make written submissions following the hand-down of this judgment about which of those courses we should take.

I should say that this decision does not render redundant my summary of the parties’ cases on justification at paras. 151-159 above, since both relied on essentially the same material in connection with ground 3, to which I now turn.

It is fair to say that at the hearing before us the irrationality ground was rather the poor relation. It was only addressed by Mr Husain very briefly, and he largely relied on his skeleton argument. Ms Giovannetti’s response was correspondingly brief. That being so, in addressing the issue I propose to confine myself to the essentials.

I start with the pleading of the relevant ground of claim. During the adjournment between the first judgment and the second hearing the Claimant applied for permission to re-amend the parts of the claim form and the Statement of Facts and Grounds setting out ground 3. The application was supported by witness evidence. It was resisted by the Secretary of State, also with the support of witness evidence. In his second judgment Lavender J granted permission. The summary re-amended ground reads:

“[T]he Secretary of State’s failure to afford refugee children the opportunity to access reunion with their parents and siblings on the same basis as adult refugees are able to access reunion with their spouses and children is, and has since its inception been, irrational; further or in the alternative, her failure or refusal to give active consideration to amending the Immigration Rules to afford refugee children this opportunity is irrational”.

That ground was analysed by the Judge at para. 23 of his second judgment as comprising distinct challenges to the rationality of (i) the Secretary of State’s decision to make the original Rules in 2000; (ii) her subsequent refusals to give active consideration to amending them; and (iii) the Rules themselves. He dismissed all three challenges. The Claimant does not dispute the decision as regards limb (i), and we are thus only concerned with limbs (ii) and (iii).

Ground 3 of the grounds of appeal reads:

“The Administrative Court erred in holding that (i) the Respondent’s repeated refusals to review the relevant paragraphs of the Immigration Rules were rational, and/or (ii) the Rules themselves, the Respondent’s maintenance of them, and/or the Respondent’s refusal to amend them to provide a straightforward route to family reunion for refugee children were rational.”

Elements (i) and (ii) correspond in substance to limbs (ii) and (iii) respectively in the re-amended ground of challenge as analysed by the Judge. Element (ii) is elaborated so as to refer not only to the Rules themselves but to the Secretary of State’s maintenance of them and/or refusal to amend them, but I do not think that that refinement affects the substance of the distinction: whichever way one puts it, the ultimate challenge is to the policy as it stood throughout the period.

Mr Husain emphasised that limb (ii) represented a more modest challenge than limb (iii). If he succeeds on limb (iii) the result would be that the Secretary of State’s policy itself is unlawful, an outcome which he accepted was “ambitious”. If, however, he succeeds on limb (ii) the result is only that she is obliged to reconsider that policy: he expressly acknowledged that in principle the result of the reconsideration could be that she maintains the present policy.

I have considered whether it would be appropriate for us to take the same course as regards this ground as I propose in relation to the discrimination ground: since the Secretary of State will be obliged to consider her policy in any event as a result of our decision on ground 1, it might be thought that there was no value in our expressing a view on its rationality. But I do not believe that the position as regards the two grounds is the same. The Claimant’s case under limb (iii) is squarely that the Secretary of State’s current policy as regards family reunion for child refugees cannot be sustained on any rational basis. That is, at least arguably, a more fundamental challenge than that under article 14; and it would at least potentially reduce the options open to her in the reconsideration which she is obliged undertake. I believe that we should accordingly determine at least limb (iii), which I therefore take first.

I have found it easier as regards this ground to consider the issues directly rather than through the lens of Lavender J’s reasoning.

I start by identifying the primary elements involved in the Secretary of State’s assessment that it is right to maintain her current policy about family reunion for child refugees. The first is that there is in fact a phenomenon of “anchor children” – that is, that at least some families facing persecution in their own countries are willing to send a child unaccompanied on a hazardous journey to claim asylum in a European country in the expectation that they can then be joined by the rest of the family. The second is that the incidence of such children seeking to come to the UK would be significantly greater if family reunion for children were automatic. The third is that that benefit outweighs the harm to child refugees from having a less straightforward route to family reunion where that is both desired and desirable.

The first two of those elements are essentially matters of assessment rather than primary fact. They require taking a view on what motivates the conduct of the families and children involved, which in turn involves assessing what information is available to them, whether from reliable sources or from what they are told by fellow asylum-seekers or agents and smugglers. That inquiry can certainly be informed by what the families and children themselves say, and the opinions of those working with them; but those sources will not necessarily provide a clear or objective answer. The third element is obviously evaluative, since it involves balancing unquantifiable risks against harms which are not susceptible to absolute measurement and are in any event by their nature incommensurable with the risk. It is also material that the decision falls to be made in a sensitive area of public policy.

So understood, the Secretary of State’s decision to maintain the policy is quintessentially of a kind to which the Courts should apply the least intensive standard of review, on grounds both of constitutional principle and institutional competence, and that the threshold for proving a case of irrationality is correspondingly very high. I do not think it is necessary or helpful to consider the vexed question of the relationship between the standard of review in such a case and that which would apply if we had to consider the justification issue under ground 2.

I do not believe that the evidence before the Judge satisfied that threshold. I will review the evidence in question, but we were not addressed on it in detail and I will only do so briefly. I take in turn the three elements in the decision identified at para. 175 above.

Anchor children

The phenomenon of anchor children is not inherently implausible. A desperate family which could not afford to pay smugglers for more than one person, or which is unwilling to risk travelling with younger or sicker members, might well regard sending one child ahead, in the expectation that in due course they would be able to follow, as the least bad option, particularly if they had a poor understanding of the risks of the journey. (I would also add that a family might start the journey together but decide when they reach the English Channel to send a child ahead alone. (Footnote: ²⁰)) It is another matter how common such cases might be. What drives particular patterns of migration is of course not possible to judge definitively, but specialist immigration staff at the Home Office should have the experience to make a reasonable judgment.

So on the face of it the Secretary of State’s assessment that the anchor child phenomenon is real and significant is at least rational. I turn to consider the contrary evidence relied on by the Claimant, which is essentially the same as (some of) the evidence summarised in the justification schedule.

First, at para. 59 of the Summary of its 2016 report the HL EU Committee recorded its view that:

“We found no evidence to support the Government’s argument that the prospect of family reunification could encourage families to send children into Europe unaccompanied in order to act as an ‘anchor’ for other family members.”

The basis for that conclusion appears in paras. 56-61 of the report itself. The Home Office in its response made clear that it did not accept it. The relevant part begins:

“The Government believes that the reunion measures suggested in the recommendation will lead to more children setting out on unaccompanied journeys that will put their lives at risk. The Home Affairs Select Committee (HASC) acknowledged this in their recent report on the migration crisis published in August.”

(We were not shown the HASC report there referred to.) The response goes on to develop the point. If I were to undertake a detailed assessment of the cogency of the Committee’s reasons and the Home Office’s response, that would be liable to infringe article 9 of the Bill of Rights. It is sufficient to say that on no view was the Committee’s conclusion such that it could not rationally be disputed.

Second, the 2019 UNHCR report Destination Anywhere records research into three “common assumptions” about unaccompanied child asylum-seekers in the UK, including that “children arriving on their own are being sent by their parents or other adult family members in the hope that those family members can join them later”. The research took the form of interviews with 23 asylum-seekers who had arrived in the UK as unaccompanied children and with an unspecified number of frontline stakeholders. The results are said to “suggest a complex picture which casts doubt on the accuracy of these assumptions”. It is certainly true that the extracts from the interviews with the children show that they had varying degrees of knowledge about the UK as a destination and a variety of different reasons for choosing it (which may have been a late stage of their journey); but that is unsurprising and does not in itself exclude the possibility that they were sent ahead by their families with the intention of seeking family reunion when and if granted asylum in the UK. On that specific question, the report says that more than half “left in circumstances which strongly suggest that they were not sent ahead to the UK by their parents”. The circumstances are not identified, though it appears that the most significant is that a third left with their parents and only became separated from them later in the journey. I do not wish in any way to impugn the good faith of this carefully-expressed report, but such doubt as it might cast on the extent of the anchor child phenomenon is not even arguably of a degree that the only rational course for the Secretary of State was to change her previous belief in its reality: it is not of course necessary to her case that all or even most unaccompanied children are sent ahead in this way.

Third, section 4 of the Without My Family report (see para. 54 above) contends that research on the reason why unaccompanied children seek asylum in the UK “paints a more complex picture of decision taking”. However, as I have said above, that is not inconsistent with the belief that a substantial number of children are in fact sent ahead by their families in the hope of taking advantage of family reunion at a later date.

It is worth recalling that the Secretary of State was not alone in taking the view that the anchor child phenomenon was real: see the EMN query from the Belgian government and the responses to it. Mr Husain pointed out that the majority of the respondents to that query said that they did not recognise such a phenomenon; but where the question is whether the Secretary of State’s view was reasonably open to her it is sufficient that several did.

The incentivising effect of allowing automatic family reunion

I can deal with this much more shortly. If it was not irrational for the Secretary of State to believe that the phenomenon of anchor children was real and significant I do not believe that it can have been irrational to believe that making family reunion easier was liable to increase the number of children sent ahead in this way. No doubt the extent of that risk could only be assessed as a matter of judgement, but if anything that makes it more difficult to characterise the policy as irrational.

The balance of benefit and harm

This element too can be dealt with shortly. I cannot see how, if the Secretary of State could rationally conclude that there was a real prospect that making family reunion easier for child refugees could increase the number of unaccompanied children making hazardous journeys to the UK, it could be irrational for her to judge that it outweighed the difficulties which the absence of automatic family reunion causes to child refugees already here. It is important not to lose sight of the fact that, real as those difficulties are, they should not prevent family reunion in any case where refusal would be a breach of the article 8 rights of the child. As I have said, we have no information on how many family reunion applications for child refugees are refused under the current policy or, if refused, are allowed on appeal; but it may well be that the greater problem is the delay and difficulty of the process than a significant difference in outcomes if an application is pursued.

It is fair to say that, as Mr Husain emphasised, the Secretary of State did not present any evidence of the process by which she struck the balance between benefit and harm (or indeed how she assessed the other two considerations identified above). But we are not in this part of the case concerned with a process challenge but with the question of whether it was rationally open to her to adopt the policy that she did; and the burden is on the Claimant to show that it was not.

The submissions from Mr Husain which I have addressed above were in fact made in the context of other aspects of his case which overlap with limb (iii). The only submission in his skeleton argument which is peculiar to limb (iii) is directed to a particular aspect of Lavender J’s reasoning. He had in dismissing this limb of the claim relied on the reasons that he had given for dismissing limb (i) – that is, the claim (not pursued before us) that the introduction of the policy in 2000 was irrational. Mr Husain submitted that that was not a proper approach because the Secretary of State would acquire further evidence and experience about the relevant considerations as the years went by. That criticism has no bearing on my own reasons given above for dismissing limb (iii), which are not focused on the facts as they stood in 2000.

I believe that Lavender J was right to hold that the Claimant has not discharged the burden of showing that the Rules themselves – and thus also the Secretary of State’s maintenance of, and/or decision not to amend, them – were irrational.

I do not believe that it is necessary or appropriate to consider limb (ii). Even if the Claimant were to succeed on it the outcome would in substance be no different from what he has already achieved on ground 1 – that is, a reconsideration by the Secretary of State of her policy as regards family reunion for child refugees. Mr Husain said that there was a difference in as much as ground 1 established only that the Secretary of State would have to conduct a section 55 exercise, which she had avowedly failed to do, whereas his case on limb (ii) identified a number of particular matters – essentially the matters set out in the justification schedule – which it is said should have been sufficient to compel a reconsideration. But that is not a point of substance. In conducting the necessary section 55 exercise the Secretary of State will be obliged to take into account any evidence that she reasonably believes is relevant to the balance between the best interests of children in the UK who are affected by her current policy and the countervailing considerations which she identifies. That may include some or all of the evidence relied on by Mr Husain, though that may depend on the extent to which it remains relevant now. But nothing will be gained by the Court embarking on the artificial exercise of assessing whether that evidence would have required her to undertake a reconsideration at some earlier date.

I would dismiss ground 3.

For the reasons given above, I would allow this appeal on the basis of ground 1 and hold that the Secretary of State is obliged to reconsider her policy as regards family reunion for child refugees as it stood prior to 4 September 2025. I would not determine the issue of justification arising under ground 2 and would invite the parties’ submissions as to whether that issue should be remitted to the High Court or whether ground 2 should be dismissed. I would dismiss ground 3.

That outcome does not involve any decision by this Court about what the Secretary of State’s policy ought to be about family reunion for child refugees. It means only that she has not, in formulating her current policy, gone through the exercise prescribed by section 55 of the 2009 Act for decisions in the immigration field which affect the interests of children. She will now have to do so, in the context no doubt of the review of family reunion policy which she has recently announced. Whatever decision about child refugees she reaches as a result of that reconsideration, including a decision to maintain the current policy, will be a fresh decision that is capable of being challenged by judicial review if arguable grounds are shown: such grounds could in principle include a challenge under article 14.

I should finally say that I regret the length of this judgment and, which is related, the time that it has taken to produce it. However, it will be apparent that the case has been bedevilled, both before the Judge and on appeal, by difficulties in identifying the real focus of the issues. The procedural history summarised at paras. 8-9 above, and particularised at paras. 77-83, 93 and 168, speaks for itself. Those difficulties may also account for the absence of witness evidence from the Secretary of State on some points where it would have been useful. I will not try to attribute blame, and I acknowledge that, with the best will in the world, the process of arguing a case at first instance and on appeal will sometimes produce new insights. Nevertheless, the case illustrates the importance of rigorous analysis of the issues, by both parties, before the case reaches a hearing in order to minimise the risk of this kind of problem.

I agree.

I also agree.