This claim for judicial review challenges the revocation on 15th October 2024 (‘the Decision’) by the Defendant of the Claimant’s ‘sponsorship licence’ under the Defendant’s ‘Workers and Temporary Workers: Guidance for Sponsors’ guidance.

This ‘Sponsor Guidance’ (as I shall call it) is an adjunct to the Home Office’s role in regulating immigration under the Immigration Act 1971 and enables the Home Office to grant (or suspend or revoke) a ‘sponsor’ licence to businesses, authorising them to issue ‘Certificates of Sponsorship’ (‘CoS’) to migrant workers. This assists those workers to get visas and addresses shortages in the labour market. The Sponsor Guidance sets out a list of grounds for revocation by the Home Office in three Annexes. Annex C1 lists grounds under which it will either revoke or suspend a licence without warning. Annex C2 lists grounds under which it will normally revoke unless there are exceptional circumstances. Annex C3 lists grounds under which it may revoke, although it will not generally do so. I shall refer to these as ‘Annex C1 grounds’, ‘Annex C2 grounds’ and Annex C3 grounds’.

In R(Prestwick Care) v SSHD [2025] EWCA Civ 184, the Court of Appeal recently decided that when revoking a sponsorship licence, the Home Office was not under a duty to assess the wider impact of revocation on the community, service users, employees and the sponsor. The present case is one of a run of first instance cases where that ground of challenge has fallen away as a result. However, in one of the linked cases in R(Prestwick), the Court of Appeal also decided the Home Office had acted unlawfully in determining that conduct fell within ‘Annex C1 grounds’ as in effect ‘dishonest’ and had followed an unfair process in reaching that decision.

Whilst the Claimant here pursues three grounds of challenge, its overarching argument is the relevant Annex C1 ground for revocation - that the Claimant had assigned CoS to ‘non-genuine roles’ - requires (the Claimant argues R(Prestwick) decided it required) either dishonesty or other reprehensible conduct. Since it is accepted that was not found here, the Claimant argues the Defendant was not entitled to find ‘non-genuine roles’ from a combination of a high turnover of staff (Ground 1); one worker failing an English test in his visa application (Ground 2); and another unable to drive being allocated a role which required driving (Ground 3). The Defendant argues R(Prestwick) actually decided there is no requirement for dishonesty or reprehensible conduct for a ‘non-genuine role’ under Annex C1. It contends it was entitled to conclude there were ‘non-genuine roles’ by combination of those three factors; or even if it was not, they showed the Claimant had breached its Sponsor duties and so the Defendant was entitled to revoke under Annex C2 which it explicitly did and if necessary, it relies on s.31(2A) Senior Courts Act 1981.

So, this case squarely raises the meaning of ‘genuine role’ and ‘genuine vacancy’ under Annex C1 and Part 3 of the Sponsor Guidance. Different views have been taken at High Court level in England prior to R(Prestwick) whether ‘non-genuine role’ requires dishonesty or other ‘reprehensible conduct’. This case requires consideration of (i) whether R(Prestwick) has now settled that debate; (ii) if not, which view is correct; and accordingly (iii) whether the Defendant was entitled to find there were ‘non-genuine roles’ and revoke the licence either for that, or under Annex C2.

The Claimant company was incorporated in 2012 and provides specialist and technical care to high-needs patients, such as end-of-life care. It has been registered with the Care Quality Commission (‘CQC’) since 2013 and receives referrals primarily through local authorities and NHS Integrated Care Boards in the East Midlands. I stress I do not doubt it provides high-quality care for many vulnerable people in that area. The witness statement for this claim by its Company Secretary Mrs Chipatiko shows how seriously it takes its responsibilities to vulnerable people.

Mrs Chipatko explains the Claimant (like much of the care sector) relies heavily on migrant labour and the Sponsor scheme, as it is difficult to recruit British nationals to care roles which involve such long, strenuous and anti-social hours for relatively low pay. This reflects the wider picture: according to a January 2024 House of Commons report cited by the Claimant, it was chronic under-staffing in the care sector (a 9.9% vacancy rate in 2022/23) which led in February 2022 to the care sector being brought within the Shortage Occupation List on the Defendant’s Sponsor Scheme. Since July 2025, no new social care visas are being granted due to concerns about abuse of the Sponsor Scheme, rather than easing of shortages: Major immigration reforms delivered to restore order and control - GOV.UK. There is also considerable staff turnover in the care sector: according to a House of Commons report, the average was 28.3% in 2022/23 (24% for those had worked in care for over 10 years rising to 47.5% turnover for care workers with less than a year’s experience).

However, the Claimant’s staff turnover rate from 2021 to 2024 was significantly higher than the sector average: at least 40%. It assigned 111 CoS to different workers, but it later told the Defendant it longer sponsored 66 of them. The Claimant says 10 resigned, 14 did not take up posts (including one who had his visa refused I shall call ‘Mr F’); 7 workers were dismissed, 3 workers left the UK and at least 14 went on to work for other sponsors (including one I shall call ‘Ms K’). The balance of 16 workers is not clear. In any event, as the Defendant points out, a 40% turnover rate is itself 40% higher than the sector average and close to average for carers with less than a year’s experience, even though some of the Claimant’s staff had more.

On 26th March 2024, the Defendant’s UK Visas and Immigration (‘UKVI’) department suspended the Claimant’s sponsorship licence highlighting its concerns that since August 2021 it had assigned a volume of CoS (111) disproportionate to the profile of the company. It raised concerns for UKVI that some of the vacancies were not genuine and/or that the Claimant was supplying workers to third parties or running an employment agency in breach of the Sponsor Guidance. UKVI required the Claimant within 20 working days to provide evidence it required current levels of recruitment, a current staff list, payroll data, employment and client contracts and ‘evidence of the recruitment process’ for two workers: one being Ms K.

On 24th April 2024, the Claimant’s previous solicitors denied the allegations in the suspension letter and specifically suggested that before April 2024, there were only 58 members of staff: the rest had been dismissed, or had switched to other employers, with another 22 set to join in April 2024 until affected by the suspension. The letter (which was re-sent on 10th May 2024 after it was not received) also referred to financial data and a positive CQC inspection which it contended justified the Claimant’s recruitment rates.

On 18th June 2024, UKVI responded (‘the Minded to Find’ letter), remaining concerned the Claimant was in breach of the Sponsor Guidance and referring to para C7.26 of it (quoted below), empowering UKVI to suspend or revoke sponsor licences if there was a more serious breach such as ‘a significant or systematic failing’ or if the sponsor ‘poses a serious threat to immigration control’. But notably, the Defendant did not maintain its allegations of non-genuine vacancies, or running an employment agency (both Annex C1 grounds). Instead, it said it believed that various failures ‘constituted a failure to comply with sponsor duties’ (an Annex C2 ground). There were three key points made in the Minded to Find letter:

UKVI’s Minded to Find letter made clear the combination of: high staff turnover; the failed visa workers (at that stage it thought four, including Mr F); and Ms K being recruited into a position for which she was unsuitable, constituted an overall failure in recruitment practices which it believed was a failure by the Claimant to comply with its sponsorship duties (an Annex C2 ground for revocation of a sponsor licence, as I discuss below). Indeed, the Defendant could not have been clearer that if the Claimant ‘failed to alleviate’ those concerns, it would revoke the licence. The Claimant plainly had a ‘last chance’ to avoid that, including the possibility of putting forward ‘exceptional circumstances’ which can avoid revocation of a licence on an Annex C2 ground. There is no such proviso with an Annex C1 ground, but I emphasise no such ground was alleged in that letter.

However, on 16th July 2024 (though dated 18th June) in a very detailed letter (‘the Response Letter’) the Claimant’s - I emphasise, previous - solicitors did not attempt to ‘alleviate concerns’, nor put forward ‘exceptional circumstances’, but instead adopted a surprisingly adversarial tone, for example saying:

Mr Gajjar ruefully accepted in argument before me that this was an ‘own goal’ by the Claimant’s previous solicitors. Moreover, the Response Letter was also a missed opportunity. No mitigation was put forward, such as the effect on the Claimant (or indeed its clients or employees) of revocation, which argument then had been yet to be rejected by the Court of Appeal in R(Prestwick). That argument was pursued initially in this claim: Mrs Chipatiko’s witness statement detailed the impact of the revocation of the Claimant’s sponsor licence on vulnerable clients, (including children who are deaf, blind or have autism) having disrupted and compromised care, this in turn was due to resignation by sponsored workers, with the resultant reputation and financial impact on the Claimant. This was touched on in a chasing email to UKVI on 14th October from the Claimant’s solicitors complaining (with some justification) a delay of 3 months was unacceptable and having a disruptive effect, including resignation of workers and reputational/financial harm. However, it was framed as a complaint about delay, not as ‘exceptional circumstances’ under Annex C2 to avoid any revocation for breach of duties as warned in the Minded to Find letter.

This apparently prompted UKVI’s same decision-maker as had written the Minded to Find letter to send the Decision the next day on 15th October. To the allegations of breach of sponsor duties were now restored allegations (indeed a main heading) ‘Non-Genuine Vacancies’. He focussed on three main areas:

Then, perhaps the most critical paragraph in the Decision, para.12, stated:

The Decision went on to cite paras.C1.38, C1.44, C1.46, Annex C1z and Annex C2 (a) and (b) of the Sponsor Guidance which I will quote and analyse later. Then under the second heading ‘Decision’, it continued:

On 3rd December 2024, the Claimant’s previous solicitors sent the Defendant a pre-action protocol letter, making various complaints, including the speed the Decision was produced – namely the day after the chasing email; and providing more information on the second worker (not Mr F) who had a visa refused: namely, that she had an English degree and valid English test. In the Defendant’s response of 17th December 2024, this explanation was accepted as not illustrating a failure of recruitment with that worker, though the Defendant’s position otherwise remained, despite the point that (now) Mr F and Ms K represented only 2 failures out of 139 staff recruited. The Claimant then changed to its current solicitors who sent another pre-action protocol letter on 24th December 2024, but there was no response to that, doubtless as the Defendant had already responded to pre-action correspondence. Therefore, on 13th January 2025, the Claimant’s solicitors (as I shall now call its current solicitors) issued this claim for Judicial Review, drafted by Mr Gajjar. It raised four grounds of review: Ground 1 is that it was irrational for the Defendant to conclude that the Claimant’s high turnover of staff posed a threat to immigration control. Ground 2 is that it was irrational for the Defendant to conclude the refusal of Mr F’s visa showed a failure in the Claimant’s recruitment practices. Ground 3 is it was irrational for the Defendant to treat Ms K’s recruitment as demonstrating a non-genuine vacancy. Ground 4 was that it was wrong not to undertake a ‘global assessment, although that was pleaded before the Court of Appeal’s decision in R(Prestwick) rejecting that argument. On 11th February 2025, the Defendant filed its Acknowledgement of Service and Summary Grounds of Defence drafted by Mr Yarrow, contesting all for Grounds. It contested Grounds 1-3 on the basis that all three conclusions were rationally open to the Defendant and cumulatively justified the conclusions either that there were non-genuine vacancies or breach of sponsor duties; or even if there were some flaws in the Decision, if they had not occurred, it was highly likely the outcome would have been substantially the same under s.31(2A) Senior Courts Act 1981 (‘SCA’).

On 11th April 2025, by which time the Court of Appeal had decided R(Prestwick), HHJ Wall refused permission on Ground 4: the ‘global assessment’ point which it had settled, but granted permission on Grounds 1 to 3 as arguable, observing that:

Building on that observation, pursuant to HHJ Wall’s directions, on 5th June 2025, the Claimant applied out of time to Reply to the Defendant’s Detailed Grounds of Defence. I granted that on 13th August, especially as R(Prestwick) had been decided since the Claim had been issued. In the Reply, Mr Gajjar argued while R(Prestwick) had removed the ‘global assessment’ point, it had also decided that a conclusion of ‘non-genuine vacancies’ required dishonesty, deception, malice or dishonourable conduct, contrary to the view of Lord Malcolm in the Outer House of the Court of Session in R(Experience India) v SSHD [2016] ScotCS CSOH 161 (quoted later). In his Skeleton Argument drafted shortly afterwards, Mr Gajjar developed this point as part of an overarching submission that the Defendant’s decision taken at its highest did not allow the conclusion any vacancies were non-genuine under paras C1.44 and C1.46 of the Sponsor Guidance. Mr Gajjar put it this way:

In his Skeleton Argument, Mr Yarrow queried whether this overarching submission fitted the existing Grounds, but commendably addressed it on its merits anyway. He argued C1.46 of the Sponsor Guidance made clear that ‘non-genuine vacancy’ included one that ‘did not exist’ and adopted the analysis in Experience India that a finding of ‘non-genuine vacancy’ by UKVI would often involve - but did not always require - dishonesty or reprehensible conduct by a Sponsor. He argued R(Prestwick) had not decided otherwise, indeed it had decided the opposite. He submitted the Decision was entitled to find that Ms K’s role was a ‘non-genuine vacancy’. In his Skeleton, Mr Yarrow concluded by arguing s.31(2A) SCA applied:

With the benefit of these excellent Skeleton Arguments (for which credit should also go to Mr Gajjar’s junior Mr Jamali), at the hearing on 23rd September 2025, I suggested that rather than looking at the three Grounds in isolation, it was preferable to pull together the overarching submissions on each side into three key issues:

In R(Prestwick), Baker LJ set out the key legal framework at [4]-[9] (which had not materially changed since the revocations in issue in those cases in 2023):

However, as in R(Prestwick), this case principally turns on particular paragraphs of Part 3 of the Sponsor Guidance, which have changed since some of the cases on ‘non-genuine vacancies’. So, I set out the key paragraphs in the 2024 version of the Sponsor Guidance, starting with some of the overarching principles:

These overarching principles are then reflected in parts of the Sponsor Guidance specific to particular issues, for example immigration control in C1.38:

Those provisions are relevant ‘sponsor duties’ for Annex C2 grounds for revocation. The only relevant Annex C1 ground is ‘non-genuine vacancies’:

There is also detailed provision on the consequences of breach of the Guidance:

Annex C1 sets out a tabulated list of ‘mandatory grounds’, including these:

Annex C2, rows (a) and (b) read as follows:

I consider the authorities on ‘non-genuine vacancies’ below. But in R(Prestwick), Baker LJ set out the general principles on revocation generally at [19]- [22]:

The phrase defined by paras. C1.44-1.47 is ‘genuine vacancy’, though paras. C1.46-47 use ‘vacancy’ and ‘role’ interchangeably and Annex C1 Ground (z) only refers to ‘role’, not ‘vacancy’. But Counsel treated ‘vacancy’ and ‘role’ as synonymous and - subject to a short point I discuss later - so do I. Yet Counsel proffer opposing interpretations of ‘non-genuine vacancies’: for the Claimant Mr Gajjar submits it always entails a finding of dishonesty or other reprehensible conduct; but for the Defendant Mr Yarrow contends that it does not. Indeed, both said their position was supported by R(Prestwick) and/or by strongly persuasive High Court authority. So, I must first review that previous authority in detail, before reaching my conclusion.

Prior to R(Prestwick), the High Court (and Scottish Outer House) interpreted ‘non-genuine vacancies’ differently, then defined in Para 15.13 and Annex 5 Ground (ae) of the 2014 Sponsor Guidance (different from para C1.44 and Annex C1 Ground (z) 2024 Guidance applied here quoted above). Annex 5 Ground (ae) used to say:

So, there was no reference to ‘reasonable grounds’ and ‘the role does not exist’ was not a free-standing example as now. Also, para 15.13 in the old Guidance stated:

There was no reference, as there is in C1.44 of the 2024 guidance, to a third element: ‘is appropriate to the business in light of its business model, business plan and scale’

The meaning of these provisions was not considered by the Supreme Court in R(New London College), Court of Appeal in R(St Andrews College), or High Court in R(Westech College), since as their names suggest, they all concerned sponsored students, not workers. Whilst R(Raj and Knoll) did concern workers, it did not consider ‘genuine vacancies’, but rather failure to comply with sponsor duties (to which I return later on Annex C2). But the High Court touched on the ‘non-genuine vacancies’ issue, as in R(Liral Veget) v SSHD [2018] EWHC 2941, where DHCJ Andrew Thomas QC noted at [53] the example of an ‘exaggerated or incorrect job description’ in Ground (ae) involved deliberate misstatement, but that was obiter as he said other mandatory grounds and discretionary grounds were made out. (I return to a comment he made on job descriptions generally as relevant to the new C1.44).

Mr Yarrow relies on the first case where the issue was actually considered directly: R(Experience India) in 2016 in the Scottish Outer House of Session, although Lord Malcolm upheld revocation after the sponsor sacked workers to prevent interviews by SSHD, so his observations on ‘genuine vacancy’ were strictly obiter:

Conversely, Mr Gajjar says those comments in R(Experience India) were contradicted by observations of Foster J (as she now is) in the English High Court in 2019 in R(Goldcare). However, she was not referred to R(Experience India) and indeed she held revocation was justifiable because of misdescription of duties of a sponsored member of staff, even though it did not justify finding dishonesty, so again strictly the following analysis of the guidance was again expressly obiter:

However, the difficulty with these cases now is that relevant paragraphs of the Sponsor Guidance have changed in the ways I described above, as I understand it in December 2020 when relevant parts of the Immigration Rules changed from ‘Tier 2 Workers’ and ‘Tier 5 Temporary Workers’ to ‘Worker and Temporary Worker Routes’. Indeed, some changes appear to respond to R(Goldcare) decided the year before, such as the clarification that Ground (z) (as it now is) applies if there are ‘reasonable grounds to believe’ a role was ‘not genuine’, presumably to make clear what Foster J had said in R(Goldcare) at [69]: the issue is for judgment of SSHD, not one of ‘precedent fact’ for the Court. Another change is that despite Foster J’s view in R(Goldcare) the first two bullet points in the old para 15.13 were ‘not definitive but descriptive’, paras C1.44-C1.47 are now entitled ‘Genuine Vacancy: Definition’. Indeed, C1.44 now has a third definitional bullet-point para 15.13 did not. Moreover, in R(Experience India), Lord Malcolm said that whilst some examples of a ‘non-genuine role’ involved deliberate deception and ‘often there will be at least an inference of dishonourable conduct’, he suggested that was not always required as those examples were not exhaustive. Perhaps the most important change for this case relates to the former example of ‘a job or role that does not exist in order to enable a migrant to come to or stay in the UK’. As noted, in R(Goldcare), Foster J found this was ‘the language of deception and dishonesty’, as were the other then-examples (now carried through to the new C1.46 and Ground (z) in slightly different language). However, that example has been split into two: ‘a role is not genuine for example because it does not exist’ or ‘has been created mainly so the worker can apply for entry clearance or permission to stay’. Whilst Mr Yarrow accepts the latter entails reprehensible conduct, he disputes the former does, which I consider below. However, what is clear is that key provisions of the Guidance have changed and so neither R(Goldcare) nor R(Experience India) remain binding.

For that reason, it is preferable to focus on cases about the 2024 Sponsor Guidance in issue in this case (the 2025 Sponsor Guidance provisions on ‘Genuine Vacancies’ and Annex C1 Ground (z) all remain the same for all sponsored workers). So, I referred Counsel to R(Hartford Care) v SSHD [2025] ACD 44, a decision on the 2024 Guidance from December 2024 by DHCJ Andrew Burns KC, who said at [52]:

Mr Yarrow submitted that I should follow R(Hartford) unless clearly wrong. But that principle only applies if it considered and was right to disagree with earlier conflicting High Court authority (Changtel v G4S [2022] EWHC 694 (Ch) [118]). But as Mr Gajjar pointed out, R(Hartford) did not cite R(Goldcare). This may be because R(Hartford) was not a case about revocation at all, but about grant of a sponsor licence, so those observations (and others I refer to below) relate to a different context which DHCJ Burns KC accepted at [52] only had ‘some parallel’ with revocation. So, R(Hartford) did not settle the meaning of Ground (z) either.

However, both Counsel submitted R(Prestwick) - or more particularly the linked case in the April 2025 judgment of R(Supporting Care Ltd) (‘R(SCL)’) - has settled whether a ‘non-genuine role’ always required dishonesty or reprehensible conduct, albeit each say to opposite effect. As discussed, the main point decided in the case was that revocation did not require UKVI to make a ‘global assessment’ of the impact of revocation on the care market, clients, employees and the sponsor. So, the Court of Appeal upheld the judgment of HHJ Kramer in R(Prestwick), but disagreed with HHJ Siddique in R(SCL). But it dismissed the appeal in R(SCL) on the basis that SSHD had breached procedural fairness. Baker LJ applied the Court of Appeal decision in Balajigari v SSHD [2019] 1 WLR 4647 at [55], which held where UKVI was minded to find ‘dishonesty or other reprehensible conduct’ by an immigration applicant, procedural fairness required it to put the allegation to them for a response and to consider that. In R(SCL), Baker LJ held this was breached, as he explained at [130]-[139] which given the debate as to their meaning I set out in nearly in full:

As I discuss later, in R(SCL) at [138], Baker LJ added UKVI’s decision was also unlawful because it failed to take relevant considerations into account, such as the fact that Ms Rima was the only worker under suspicion out of SCL’s 162 staff and although she was not performing two tasks distinguishing a ‘skilled care worker’ from a ‘care worker’, both categories are covered by sponsorship and she was performing the majority of tasks under her CoS job description. UKVI had drawn an inference of dishonesty, but not called for an explanation with sufficient clarity. He concluded at [139] the decision that both Grounds (z) and (s) applied was wrong.

Mr Gajjar submitted that Baker LJ’s analysis, especially at [132] and [137] (which accepted the submissions of SCL’s Counsel Mr Malik KC recorded at [120]), was predicated on the basis that an allegation that a role is not ‘genuine’ under Ground (z) is an allegation of dishonesty – or at least other ‘reprehensible conduct’ to which the same principles apply (c.f. Balajigari at [55]). Mr Yarrow disagreed and argued that Baker LJ’s analysis in R(SCL) at [137] assumed that the allegation of ‘non-genuine vacancy’ in the suspension letter was not an allegation of dishonesty or reprehensible conduct; that was only found in deciding the role was ‘deliberately exaggerated’. In my judgement, both of these skilful submissions effectively cancel each other out. Baker LJ simply did not say that Annex C1 Ground (z) either did or did not entail a finding of dishonesty or reprehensible conduct (nor did Mr Malik KC submit that it did in terms at [120]). What Baker LJ said was that the revocation decision was procedurally unfair because the particular basis on which the decision-maker found Ground (z) was made out – i.e. that Ms R’s role had been ‘deliberately exaggerated in order to facilitate her stay in this country’ - was a finding of dishonesty (which the judge had found and neither party disputed), which had not been put in sufficient detail in the suspension letter, which spoke only of ‘it being not a genuine vacancy’. That does not mean that Baker LJ said that ‘non-genuine roles’ generally in Ground (z) either did or did not entail dishonesty or reprehensible conduct, only that the particular finding of dishonesty leading to revocation had not been fairly put. The only conclusion on the interpretation of ‘genuine vacancy’ one can draw from Baker LJ’s analysis was that he was content to proceed on the basis that the particular example of deliberate exaggeration of a job description in para C1.46 was an allegation of dishonesty. That had already been said by Foster J (as she now is) in R(Goldcare) at [83]-[84], yet it is telling that Baker LJ in R(Prestwick)/R(SCL) at [24] endorsed her analysis in another part of that judgment, but did not mention in on this particular point. Nor was he referred to R(Experience India), R(Liral Veget) or R(Hartford) (though the last was decided after argument in R(Prestwick)). Therefore, in my judgment, the meaning of ‘genuine vacancy’ was not settled either in R(Prestwick)/R(SCL). But it seems clear the specific examples in Ground (z) (re-wording two of the examples in para.C1.46) of a role being a ‘sham (including but not limited to where the CoS contains an exaggerated or incorrect job description to deliberately make it appear to meet the requirements of the route you assigned it under when it does not)’ or where the role ‘has been created mainly so worker can apply for entry clearance or permission to stay’ should be seen as allegations of dishonesty or reprehensible conduct to which [132]-[136] of R(SCL) apply.

Indeed, in my view, whether Ground (z) always requires dishonest or reprehensible conduct as Mr Gajjar submits, has not been settled in the cases since R(SCL) either:

The only case on the present Guidance and since R(Prestwick)/R(SCL) where there is a clear conclusion that Ground (z) does not always entail dishonesty or reprehensible conduct is also not binding because it was refusing permission for Judicial Review: R(He Reigns Healthcare) v SSHD [2025] EWHC 1699 (Admin). UKVI suspended a sponsor’s licence on the basis that it was assigning CoS to more migrants (124) than it realistically had suitable work for. The sponsor suggested this was explained by a contract by a local authority, but this was rejected in the revocation decision, which held that on the evidence the sponsor supplied, not only did 50 of the sponsored migrants not appear on the payroll at all, the number of sponsored workers far outstripped the capacity for the work being done and that a number of roles ‘have either been exaggerated or do not exist, therefore we do not accept you are offering your sponsored workers genuine vacancies’. DHCJ Matthew Butt KC held it not arguable this was flawed and whilst R(SCL)/Balajigari procedural fairness standards had not been complied with, they did not have to be because the conclusion did not entail a finding of dishonesty. He observed:

Of course, even decisions on permission for appeal, let alone permission for Judicial Review like R(He Reigns), are not citeable authority as there is not full argument. Therefore, Mr Yarrow does not say that I should follow R(He Reigns) as such, but he does adopt DHCJ Butt’s reasoning on when a role ‘does not exist’ in Ground (z). However, as Mr Gajjar pointed out, DHCJ Butt KC was not referred to R(Goldcare) or the other cases on this issue besides R(SCL). Nevertheless, the revocation decision in R(He Reigns) was treated as a finding that at least some of the 124 sponsored roles ‘did not exist’, which was held not to be a finding of dishonesty at [39]/[41]. Since Mr Yarrow adopts that analysis, I will consider it in my own conclusions, to which I now turn, as this (doubtless overlong) review of authority confirms it has not been definitively settled at High Court level, let alone by the Court of Appeal, that Ground (z) does or does not always entail dishonesty or reprehensible conduct.

This issue relates to the interpretation of Annex C1 Ground (z) and paras C1.44–1.47 of the Sponsor Guidance as opposed to their application. The authorities on that guidance make clear its application to an individual case is the expertise of UKVI/SSHD which the Court should respect. But interpretation of guidance is an objective question of law for the Court which it is well-qualified to decide, as Lord Reed explained in Tesco Stores v Dundee CC [2012] PTSR 983 (SC) at [18]-[20]. The relevant approach was helpfully summarised by DHCJ Tinkler in Treal:

But I also bear in mind that in R(St Andrews College), Haddon-Cave LJ said at [39]:

Yet that approach to reading the Sponsor Guidance in my view applies as much to the SSHD/UKVI as it does to the Court. With this in mind, I approach interpretation of Annex C1 Ground (z) ‘objectively in accordance with the language used, read in its proper context’, starting with the language before moving to its context.

For ease, I repeat the key paragraphs in the 2024 (and 2025) of the Sponsor Guidance:

Mr Gajjar submitted that as it was conceded there was no finding of dishonesty or reprehensible conduct by the Claimant, Annex Ground (z) was an impermissible ground of revocation, as it always entails such conduct. He pointed to the examples in Ground (z) and C1.46, submitting they each entailed dishonesty or reprehensible conduct. He also relied on R(Goldcare) and R(SCL), which spoke at [132] of ‘deliberately’ or a ‘sham’ as synonymous with ‘dishonestly’ and that a role ‘created mainly so the worker can apply’ for leave to enter or remain was similarly interpreted in R(TJ) and R(Treal). The same applied to deliberately inappropriate job requirements in C1.46, or deliberately over-long duration of a job in C1.47.

Mr Yarrow did not dispute that interpretation of any of those examples and in my view, it is consistent with the analysis of Foster J (as she now is) in R(Goldcare) on the similarly-worded examples in the old Guidance. It is also consistent with the approach (even without any formal ‘interpretation’) in R(SCL), R(Liral Veget), R(TJ), R(J’s Supermarkets) and R(Treal). Even in R(Experience India), Lord Malcolm accepted at [14] that ‘examples are given in the guidance which clearly involve an element of deliberate deception’ and at [12] that such an inference could often be drawn (R(Hartford) did not address this issue). I do not believe DHCJ Butt KC in R(He Reigns) suggested a looser interpretation of ‘exaggeration’, but if so, he was not referred those cases except R(SCL) and I would respectfully disagree.

However, Mr Yarrow did submit that Ground (z) did not always require reasonable belief in a sponsor’s dishonesty or reprehensible conduct, for three reasons:

On submission (iii), of course, I accept that Ground (z) and para C1.46 make it clear the examples are just examples, but that does not mean the word ‘genuine’ can mean anything UKVI consider it means. As Lord Reed put it in Tesco Stores at [19] ‘public authorities do not live in the world of Humpty-Dumpty’, echoing the famous dissent of Lord Atkin in Liversidge v Anderson [1942] AC 206 (HL) of Humpty-Dumpty in ‘Alice through the Looking-Glass’ saying ‘When I use a word...it means just what I choose it to mean’. Indeed, the word ‘genuine’ cannot even mean whatever SSHD/UKVI rationally think it means, as that is the incorrect approach to the meaning of policies in cases like Gangadeen overruled in Mandalia. In Ground (z), the key phrase is ‘not genuine’ and the meaning of that phrase is an objective matter for the Court of the meaning of the language in its particular context. Sensibly, Mr Yarrow only touched on this point and did not really develop it. I only mention it as it underlines the issue is one of interpretation, not application.

On submission (ii) however, Mr Yarrow did enthusiastically adopt and develop DHCJ Butt’s view in R(He Reigns) at [39]/[41] that a role could ‘not exist’ and be ‘not genuine’ even due to negligence (e.g. if a vacancy had already been filled), ignorance (e.g. a CoS being incorrectly processed), or misplaced optimism (more CoS assigned than required): not considered in R(SCL). But speaking as a former fee-paid Employment Judge, it is odd use of language to say a role or job is ‘not genuine’, even in the sense of ‘does not exist’, simply because it has already been filled, or was incorrectly processed, or proves superfluous. It is better to say such roles are genuine but ‘filled’, ‘misprocessed’, or ‘redundant’. There is no need to strain the meaning of ‘genuine’ to avoid a defence of incompetence as suggested in R(He Reigns) at [41], as such conduct is caught by other Annex C1 grounds (e.g. Ground (s)) or by Annex C2, which were not discussed in R(He Reigns). Whilst interpreting a policy or guidance is not the same as a statute: R(St Andrew’s), both involve language in its context: Tesco. As Lord Sales said of statutory interpretation in R(Paccar) v CAT [2023] 1 WLR 2594 (SC) at [40]-[41], the purpose and wider scheme of a rule orientates its language, so a word or phrase takes meaning from its specific context or setting. Here, all the other examples in Ground (z) entail deliberate conduct by a sponsor; and as it can be expected to know whether its own role ‘exists’, reading the example ‘does not exist’ with the synonymous phrase in C1.46 in my view means that in Ground (z), the example ‘a role does not exist’ covers only a CoS deliberately assigned for a role that ‘does not actually exist’, in the sense of not being ‘genuine’ but the opposite: i.e. ‘fake’ or ‘bogus’. That is clearly dishonest or reprehensible. That was how Foster J in R(Goldcare) read ‘a role that does not exist to enable a migrant to [enter or stay] in the UK’ in the old Guidance. Whilst now been split into two examples, that does not mean it no longer need be reprehensible. It just means a role created to facilitate entry can be ‘not genuine’ even if it exists (as alleged in R(TJ Trading)) or if it ‘does not actually exist’ (i.e. is ‘fake’ or ‘bogus’) irrespective of its purpose. That is more egregious than a ‘sham’ (i.e. an actual role misrepresented to deceive third parties like UKVI or HMRC: Autoclenz v Belcher [2011] IRLR 820 (SC) at [28]), because the role ‘does not actually exist’ at all. So, I agree with Mr Gajjar: if a role is ‘not genuine’ as it ‘does not exist’ in Ground (z), it entails reprehensible conduct. That also means that before UKVI find reasonable grounds for any of the examples in Ground (z), C1.46 or C1.47, they must comply with the fairness safeguards in R(SCL) [132-6].

Having said that, the situation in R(He Reigns) itself at [39] (a proportion of sponsored workers not being on the payroll and there being insufficient work to support all sponsored workers) certainly ‘calls for an explanation’ as Baker LJ put it in R(SCL) at [132], which if not forthcoming, could justify an inference some of the roles (even if unclear which) are ‘not genuine’ as they do ‘not actually exist’ in the sense of ‘fake’/’bogus’. That could also be a rational inference for a single ‘role’ e.g. if a sponsored worker never actually starts work and is not replaced. Indeed, that may be what the UKVI decision-maker in R(He Reigns) meant and on its facts I respectfully agree with DHCJ Butt KC. However, those are legitimate inferences that a role is not ‘genuine’ in the sense of ‘does not exist’, rather than dilution of its true interpretation in the context of all the other examples in Ground (z)/C1.46-47.

Moreover, I do accept the essence of Mr Yarrow’s submission (i), that Ground (z) must be read in its own context: namely what is explicitly called (in the guidance since R(Goldcare)) the start of ‘definition’ of ‘genuine vacancy’ in para. C1.44:

As Mr Yarrow said, these three characteristics are not just examples, or even a paradigm, of a ‘genuine vacancy’, but definitional of it: a ‘closed set’ as Mr Gajjar realistically accepted. Whilst neither the word ‘vacancy’, nor para C1.44, are specifically referred to in Ground (z), the effectively identical examples in it and para C1.46 make clear the meaning of ‘genuine’ is the same. This is also consistent with C1.45 entitling UKVI to request information or evidence ‘to establish this requirement (in C1.44) has been met’. Under the old Guidance, DHCJ Thomas QC in R(Liral Veget) at [55]-[58] confirmed such requests did not reverse the burden of proof, but reflected the requirement on sponsors to co-operate and inferences could be drawn from failure to do so. Whilst the ‘definition’ is the whole of C1.44-1.47, C1.44 is not cut down but illustrated by the ‘examples’ in C1.46 and C1.47. Whilst they all entail either dishonesty or reprehensible conduct as discussed, the absence of any of the three factors in para C1.44 does not necessarily do so, as Lord Malcolm said of the first two in the old Guidance in R(Experience India) at [14]. There is a difference between a role that is not ‘genuine’ as it ‘does not (actually) exist’ and a role which is not ‘genuine’ as it is ‘inappropriate to the business given its scale’ etc.

That said, like the rest of the Sponsor Guidance, Ground (z) and C1.44 must be read not rigidly or myopically, but sensibly, purposefully and holistically (St Andrew’s College) - by SSHD/UKVI as well as by the Court - in the context of the rest of the Guidance given that Ground (z) defines a ‘genuine vacancy’. Three points follow:

Drawing those together, I would put it this way: focussing on the job not the worker, a finding that a role in the sense of vacancy (as opposed only to the way it is later performed) lacks any of the C1.44 characteristics properly interpreted, even without dishonesty or reprehensible conduct, can justify but not does not inevitably mean, that the role is ‘not genuine’ under Ground (z). It is not enough for there to be some practically irrelevant technical non-compliance with C1.44, the non-compliance must give reasonable grounds to believe that the role is ‘not genuine’. For example, a trivial ‘dissimilarity’ in the duties in a job description for a role is insufficient on its own to give reasonable grounds to find the role ‘not genuine’. Also, ‘not genuine’ in Ground (z) must not be stretched so wide by C1.44 to render other Annex C1 grounds otiose, or to overlap confusingly with Annex C2 Grounds. So, outside the examples in Ground (z) and C1.46-1.47, it may be helpful for UKVI to consider: are there reasonable grounds to believe one or more of the requirements of a ‘genuine vacancy’ in C1.44 are not met so that the role is ‘not genuine’ ?

The first characteristic at C1.44 (‘C1.44(1)’) is that a genuine role ‘requires the jobholder to perform the specific duties and responsibilities for the job and meets all of the requirements of the relevant route’. If the role does not meet immigration requirements, or if it does not ‘require performance of duties’ (e.g. a job with written terms that do not reflect ‘the true agreement between the parties’: Autoclenz at [29]), the role can be ‘not genuine’, even without dishonesty or reprehensible conduct. But C1.44(1) / Ground (z) must be interpreted alongside other Annex C1 Grounds:

Unlike Ground (z), Grounds (q), (s) and (w) do not require reasonable grounds to believe a role is not ‘genuine’. Of course, more than Ground can apply to the facts. But if a role does not meet immigration skill requirements, if (z) and C1.44(1) inevitably deemed it ‘not genuine’, Ground (w) would not just overlap with (z) but be otiose. Ground (s) also does not require deliberate conduct (R(J’s Supermarket) at [112]) and plainly relates to the role ‘undertaken by a worker’ in practice. But Ground (z) relates to the ‘genuineness’ of the initial vacancy/role, although how the job is done may enable an inference the initial role/vacancy was not ‘genuine’, (c.f. R(SCL) at [132]/[138] and R(J’s Supermarket)). For C1.44(1), it may assist to ask whether there are reasonable grounds to believe the role/vacancy does not ‘require the jobholder to perform the specific duties and responsibilities for the job [or] meet all requirements of the relevant route’; and if so, that the role is ‘not genuine’ ?

The second characteristic at C1.44 (‘C1.44(2)’) is that a genuine vacancy/role ‘does not include dissimilar and/or predominantly lower-skilled duties’. This appears to focus on the internal consistency of duties in the vacancy / job description, rather than whether they are actually required to be performed. However, like C1.44(1), the ultimate focus is on the duties as they are described in the original vacancy, job description or CoS, as DHCJ Thomas QC explained in R(Liral Veget) at [46]:

However, again how the role is performed in practice may enable an inference that the vacancy/role was ‘not genuine’ in the way stated in C.1.44(2). However, a vacancy with similar but predominantly lower-skilled duties could just be a genuine lower-skilled job, e.g. office junior. Of course, a genuine predominantly lower-skilled job may not merit a CoS at all: R(J’s Supermarkets) at [65], or justify revocation on other grounds (e.g. Grounds (w) or (q)). But if that role is ‘genuine’, Ground (z) does not apply. So, for C1.44(2), it may assist to ask whether there are reasonable grounds to believe that the role/vacancy includes dissimilar and/or predominantly lower-skilled duties; and if so, that the role is ‘not genuine’ ?

The third characteristic at C1.44 (‘C1.44(3)’) is that a genuine vacancy/role ‘is appropriate to the business in light of its business model, business plan and scale’. However, as Mr Yarrow accepted, it is one thing to say that a role is not genuine because it is clearly inappropriate to a business (an extreme example we discussed was a dressmaker for a bakery). It is quite another for UKVI to second-guess a sponsor’s commercial judgment as to how many vacancies may be needed for its business, as DHCJ Burns KC explained in R(Hartford) at [37] and [60]:

Whilst R(Hartford) was a case of prospective application for a sponsor licence, this analysis applies equally to its retrospective revocation. Indeed, as DHCJ Burns KC also said in R(Hartford) at [56], it would be irrational for UKVI to find a role is not ‘genuine’ because a sponsor in the care field has not included a term that is not standard in the field anyway, like ‘guaranteed hours’. On the other side of the line, on the facts in R(He Reigns), the rational finding that there were far more sponsored roles than there was work, coupled with large numbers of sponsored workers not being on the payroll for long periods of time, meant it was rational to conclude at least some of those roles were not ‘genuine’. But this is not because a conclusion ‘they did not exist’ need not entail deception, but either because an inference of such deception could have been drawn as discussed above, or even if not, if the roles ‘actually existed’, they were clearly ‘inappropriate’ to that sponsor’s business. So, for C1.44(3) it may assist to ask whether there are reasonable grounds to believe the role/vacancy is (clearly) ‘inappropriate to the business in light of its business model, business plan and scale’; and if so, that the role is ‘not genuine’ ?

These conclusions are in my view consistent with the wider context of the other grounds in Annex C1, those in Annexes C2 and C3 and the Guidance generally:

That trust in the sponsor is also a recurring theme in the authorities, from the Supreme Court in R(New London College), the Court of Appeal in R(Raj & Knoll), R(St Andrews’ College) and R(Prestwick)/R(SCL); and the High Court, including in the ‘highwater mark’ for Mr Gajjar’s ‘dishonesty or reprehensible conduct’ submission: R(Goldcare), approved in R(Prestwick) not on that issue but on this:

Whilst Foster J (as she is) in R(Goldcare) at [83]-[84] described the examples in (now) para C1.46 and Ground (z) as entailing dishonesty or reprehensible conduct (and I have found this also applies to ‘it does not (actually) exist’), her actual decision was revocation was justified in that case even without such conduct. To require it always for Ground (z) to be engaged would go against the thrust of the Sponsor Guidance she encapsulated in R(Goldcare) at [21]-[22]. Since those principles were previously summarised in R(Raj Knoll), albeit in a different context, It is also consistent with Lord Malcolm in R(Experience India) commenting that a conclusion that dishonesty is not always required is consistent with R(Raj Knoll).

I would also add (though it does not formally arise for decision) that a public authority should apply its own policy (unless there are good reasons not to do so: Tesco, Mandalia and R(Prestwick) at [100]-[101])); and SSHD has adopted a definition of a role not being ‘genuine’. Therefore, if UKVI accepts a role is a ‘genuine vacancy’ within that definition in C1.44, it is not rationally open to it to decide the role is nevertheless ‘not genuine’ in Annex C1 Ground (z). That does not prevent it departing from its own Guidance if there is a good reason, but this may well be unnecessary given the plethora of revocation grounds besides Ground (z).

Finally, I apologise for the length of my analysis on this issue. It reflects the skilful submissions of both Counsel that the issue had been decided - different ways – in R(SCL) by the Court of Appeal or by the High Court. I disagreed with both Counsel on both, so it was necessary to review those authorities at length, before tackling my own analysis. I have tried to incorporate the previous observations, even though they are not binding, in the (possibly folorn) hope that, subject to any appeal, the issue can now at last be settled at this level (of course unless another Judge in the High Court considers my analysis wrong c.f. Changtel). Indeed, standing back from all the detail, I believe my analysis is consistent with R(Liral Veget), R(Goldcare), R(SCL), R(TJ Trading), R(J’s Supermarket) and R(Treal) on the examples of ‘non-genuine roles’ in C1.46 and Ground (z) requiring dishonesty or reprehensible conduct; but also consistent with R(Experience India) and R(Hartford) on my conclusion that Ground (z) does not always require such conduct. That was also the conclusion in R(He Reigns) and whilst I disagree with part of the analysis there on the ‘did not exist’ example in C1.46/Ground (z), on my preferred approach an inference that roles ‘did not exist’ in the sense of ‘fake’ was plainly made out in it.

Nevertheless, after all the detail, I hope a summary of my conclusions may assist:

Having dealt with the fundamental legal issue about the interpretation of Annex C1 Ground (z), I can deal more briefly with my conclusions in this case both on Ground (z) and Annex C2 Grounds (a) and (b). On Ground (z), as I said, the issue is whether the Defendant was rationally entitled to conclude the Claimant provided ‘non-genuine roles’ to workers generally, or specifically Ms K. I will first set out the relevant principles to this sort of irrationality challenge, then whether the Decision reached a rational conclusion the Claimant had assigned ‘non-genuine roles’ (breaking this down into (i) whether the Defendant reached rational factual findings and then standing back to consider whether its conclusion of ‘non-genuine role’ was rational); and if it was not, whether it is highly likely that without its error, it would have reached that conclusion under s.31(2A) SCA.

The general principles of rationality challenges apply to sponsorship revocation decisions, as DHCJ Tinkler observed in R(Treal) at [19]–[20]:

That theme runs through the principles set out by Baker LJ in R(Prestwick) (quoted above at paragraph 25 of this judgment), which there is no need to repeat save for the encapsulation by Tomlinson LJ in the Court of Appeal in R(Raj Knoll) at [32]:

There is no reasons challenge, but it is relevant to rationality that decisions are written by UKVI decision-makers not lawyers, need not be lengthy or legalistic and should be read as a whole in their proper context and as directed to recipients familiar with the system and facts: R(Tendercare) at [65] (similar to observations in R(SCL) at [136]). As DHCJ Tinkler noted in R(Treal) at [16], Lord Brown in South Bucks DC v Porter (No 2) [2004] WLR 1953 (HL) said at [36]:

Having said that, as Baker LJ explained in R(Prestwick) at [26]-[27], part of the rationality exercise is examining whether the decision-maker has failed to take into account relevant considerations (or taken into account irrelevant ones). If those considerations are neither ones that must be, or must not be, taken into account, they fall into the third category of considerations discussed by Lords Hodge and Sales in R (Friends of the Earth Ltd) v SST [2020] PTSR 2021 190 (SC) at [120]-[121]:

Whilst the main relevance of this principle in R(Prestwick) was the ‘global assessment’ point, Baker LJ also touched on it at [138] in observing the decision in R(SCL) was also flawed due to a failure to consider matters relating to Ms Rima:

These principes are relevant in the present case because Mr Yarrow accepted that neither the findings of high staff turnover, nor recruitment of Mr F whose visa was refused, individually justified a finding of any ‘non-genuine role’. However, he submitted against the background of these findings, those about Ms K did justify that conclusion, as she was recruited to a role it was known she could not do. By contrast, Mr Gajjar submitted that, as with Ms Rima in R(SCL), the Decision failed to take into account the administrative error in relation to Ms K was only one worker out of 139 CoS issued (or two including Mr F); and so it was irrational to leap from those two administrative failings to a conclusion that either role was ‘not genuine’. Given the way Mr Yarrow focusses his case on Ground (z) on Ms K’s recruitment, with Mr F and high staff turnover as background context, it is helpful to work through those three issues using the grounds of challenge, to decide whether the Decision reached rational factual conclusions on each of those issues individually. I will then stand back to decide whether the Decision reached a rational conclusion about ‘non-genuine roles’ and so whether Ground (z) justified revocation. (I defer consideration of Annex C2 grounds until after consideration of Ground (z)).

As discussed above, on the staff turnover issue, the Decision stated at paras.7-8:

It is important to note, based on the Claimant’s own staff list provided in April/May, the Decision reached a fresh calculation: not 66 departed CoS workers or 40% of the total; but 78 departed CoS workers, comprising over 60% of the total of 126.

Under Ground 1, Mr Gajjar criticises this factual conclusion in several ways:

However, I accept Mr Yarrow’s submissions that none of this undermines the rationality of the Decision’s conclusion on staff turnover. The starting-point is that the Defendant reached a fresh calculation based on the Claimant’s own staff list, so staff turnover is actually over 60%, not 40%. However, even sticking with the 40% in the Minded to Find letter, all that was said of relevance to that in the Response letter by the Claimant’s previous solicitors was to point to the House of Commons report and suggest: ‘Its estimate of nearly 30% is close to the 40% cited in the letter’. However, it was not: 40% is itself 40% higher than the 28.3% average. That is indisputably a higher-than-average staff turnover, which calls for an explanation, not about dishonesty etc, but simply why the Claimant’s staff turnover is significantly higher than average for the care sector. But no explanation at all was forthcoming in the Response letter. I would respond as follows to Mr Gajjar’s individual points:

Turning to Ground 2, the Decision reached this conclusion about Mr F:

As I have explained, whilst the Decision referred to two workers with failed visa applications, once the Claimant explained (after the Decision) the other worker had an English degree and test, the Defendant withdrew that criticism as it had in the Decision itself after the Response letter with the two other workers it had suspected. The Defendant cannot be criticised for accepting representations from the Claimant, especially given after the Decision. The Claimant’s Ground 2 focusses on Mr F, which is also my focus. Mr Gajjar points out that in the pre-action protocol letter, the Claimant’s previous solicitors pointed out that UKVI had initially awarded Mr F the English test, before rescinding that, showing how UKVI itself could make mistakes. Mr Gajjar argues the Claimant was only required to ‘believe’ that Mr F would pass the English test, which it did; and that a sponsor cannot be blamed if such a candidate later failed that test, for example due to nerves on the day.

Whilst Mr Gajjar and Mr Yarrow look at the same sponsor duties from different ends, they agree that a sponsor has both the following duties in para C1.38:

As Mr Gajjar observed, the duty to ensure a worker is ‘appropriately qualified’ does not extend to the English language requirement under Appendix SW Immigration Rules. Indeed, the Sponsor Guidance as a whole contains no specific obligation on sponsors to check likely compliance with the English Language test and UKVI is better placed than sponsors to do so in any event. However, the English Language test is an ‘immigration requirement of the route’ which a sponsor must ‘believe the worker will meet’ when they assign a CoS. Therefore, whilst Lord Malcolm in R(Experience India), cited by DHCJ Burns KC in R(Hartford), spoke of the sponsor giving UKVI a ‘guarantee’ in relation to sponsored workers, it is slightly more nuanced than that: the sponsor’s duties are set out in the Sponsor Guidance and carefully differentiated. Some duties are akin to guarantees (in a loose, not legal sense), such as the duty to ensure qualification for the role (C1.38 gives the example of a sponsor checking a doctor’s qualifications before sponsoring). But others are not. However, as Mr Yarrow observed, a sponsor must not assign a CoS unless they believe an application will meet the English language requirement. Moreover, because of the trust placed on sponsors, there must be some proper basis for that belief, even if it turns out to be misplaced. For example, with the other worker named in the Decision who had a visa refused, once it was eventually explained that she not only provided the Claimant a valid English test, but an apparently valid English Degree, the Defendant accepted there was a proper basis for the Claimant’s belief that she would meet immigration requirements and withdrew criticism of her recruitment.

However, in Mr F’s case, there was no such explanation of the Claimant’s ‘belief’ that he would pass the English test when assigning him a CoS. In the pre-action protocol letter, the Claimant’s previous solicitors simply pointed out the Defendant had initially awarded Mr F points but then rescinded them, showing it can make mistakes. Of course, UKVI can make mistakes, but that does not explain why the Claimant believed Mr F would pass the English test when recruiting him. In any event, that point was only made after the Decision. So too was the point about the other worker who failed to get a visa and without an explanation for her, the Defendant was rationally entitled to conclude in the Decision both workers were recruitment failures, even though it accepted the explanation for her when later given it. That was not an ‘error’ in the Decision given what it had been told. It cannot know what the Claimant ‘believed’ and why unless the Claimant says so before the Decision it as it had the chance to do in the Response letter. Yet in that letter, the Claimant’s previous solicitors offered no proper explanation of the basis of belief that either worker would pass the English test, other than asserting the robustness of their recruitment processes. So, the Defendant was rationally entitled to conclude in the Decision recruitment of both evidenced the lack of robustness in the Claimant’s recruitment processes, especially through the ‘lens’ of the high staff turnover. In any event, even just focussing on Mr F, the Defendant was entitled to find his recruitment alone was a failure of process, especially in the context of high staff turnover posing a threat to immigration control even if Mr F did not himself.

Turning to Ground 3, the Decision reached this conclusion about Ms K:

There is a difference between (i) whether this was a rational factual finding (which is what I am considering at the moment) and; (ii) whether it rationally justified a conclusion Ms K’s role was not ‘genuine’ against the background of Mr F’s recruitment and high turnover (which I consider next). Whilst Mr Gajjar submitted this was an isolated error which posed no threat to immigration control, that relates more to the latter Ground (z) conclusion than the former factual one. The Decision also does not say Ms K’s recruitment alone posed a threat to immigration control. Focussing initially on that factual finding, Mr Gajjar wisely recognised that Ms K’s recruitment was an ‘administrative error’ by the Claimant. However, he sought to mitigate it, in terms of the seriousness of the error and the time taken to address it. He took me carefully through the Claimant’s documents relating to Ms K, pointing out that she had a Degree in English and successfully passed an English test in July 2023. Whilst her ‘start date’ was recorded as June 2023, in fact her job application (where she clearly flagged she could not drive) was not until 7th August 2023 and her job interview and offer letter were on 9th August. The offer letter required her to ‘provide’ her Driving Licence and International Driving Permit, but Mr Gajjar suggested this should be read as she was required to ‘obtain’ them. She was then assigned a CoS on 11th August 2023. However, she never started work for the Claimant, who reported to UKVI on 31st October 2023 she was no longer sponsored.

However, whilst the Claimant corrected its error with Ms K’s ability within two months, it was still a clear and obvious error. Notably, despite Mr Gajjar’s ingenious re-interpretation of the offer letter, it did not say she had to ‘obtain a’ but ‘provide your’ Driving Licence. It was therefore a ‘requirement of the role’ for the purposes of para C1.44(1) of the Sponsor Guidance. However, as should have been clear to the Claimant from her job application two days earlier, Ms K did not have a Driving Licence. Nor is there any evidence in the Claimant’s own disclosed records of any correspondence chasing her for it, or querying why she had not ‘obtained’ one yet. This error was simple and entirely obvious (to anyone but apparently the Claimant’s previous solicitors). I accept that offer letter was not the ‘second dropping of the ball’ as suggested by Mr Yarrow. It was the first – and serious – ‘dropping of the ball’, compounded by the Claimant’s failure to spot the problem for two months, if indeed that was the problem, rather than ‘family reasons’. However, I am prepared to accept that lack of a Driving Licence was the reason for Ms K’s departure and certainly the Defendant was entitled to do so, as that was what the Claimant had said. Yet the only explanation of that clear and obvious failure in recruitment practice in the Response letter was that the Claimant had not been told by Ms K, when it was in fact clear from the documents attached with that letter that she had told the Claimant this from the start. In the circumstances, the Defendant’s conclusion that Ms K was recruited into a role with a requirement to drive she was unable to carry out was not only rational, it would have been extremely surprising had it reached any other conclusion. Whilst it was rationally entitled to find recruitment of Mr F was a failure of the Claimant’s recruitment process, its decision to hire Ms K was unquestionably a glaring failure of its recruitment process.

Therefore, I reject the rationality challenges to all three factual findings forming the ‘pillars’ of the Defendant’s decision to revoke. However, I must now examine whether those three factual findings even together could support the weight of its conclusion that at least Ms K’s job was not ‘genuine’ and Ground (z) applied as a result of those factual conclusions (as the Decision put it: ‘considering each of the issues detailed at paras 4-11 above’). Its reasoning for this was encapsulated at para 12 of the Decision, but developed at paras.19-21, which all bear repetition:

So, as is clear from para.21, the Decision had two alternative bases: Annex C1 Ground (z), for reasons at para.12, but relying on the three factual findings at paras. 4-11; and/or Annex C2 Grounds (a) and (b) which are based on those same factual findings, but leading to a different overall finding (which I will discuss later).

It is not entirely clear that para.12 of the Decision is talking about Ms K at all. It starts by saying ‘at least some of the vacancies you have filled are non-genuine’. When I first read para.12, I understood it to conclude the Claimant’s high staff turnover showed some of its CoS were non-genuine, including Mr F and Ms K. However, this is not a case like R(He Reigns), where there was evidence of a large number of workers not on the payroll, in combination with insufficient work for all the CoS granted, enabling an inference that ‘at least some of the roles did not exist’ (in the sense of ‘fake’ or ‘bogus’ as discussed above). Indeed, on the Defendant’s own turnover breakdown, the 13 rejected and 10 dismissed staff pointed away from ‘non-genuine’ roles and indeed paras.7-8 of the Decision do not suggest any of roles referred to there were ‘non-genuine’. The 27 workers who switched before employment might possibly have justified the Defendant’s initial suspicion in the suspension letter on 26th March that the Claimant was supplying workers to others or even running an informal recruitment agency. But those allegations were not pursued in the Minded to Find letter on 18th June that focussed purely on the three factual issues alleged as a failure to comply with sponsor duties. Indeed, the Decision itself does not renew those allegations, but found ‘non-genuine vacancies’, which is a different Annex C1 Ground of revocation: (z), not (x) or (y). Whatever else para.12 of the Decision says, it does not explain any basis for conclusion that ‘some vacancies were non-genuine’ other than Mr F and/or Ms K.

However, Mr Yarrow accepted the Decision did not find Mr F’s role was ‘non-genuine’ either. Whilst para.12 related to workers ‘unable or unwilling to perform the role’s specific duties and responsibilities’, which might have been said of Mr F failing his English test, para.6 instead said Mr F ‘failed to meet the eligibility criteria to apply for a Health Care visa or failed to meet language requirements’. That is not a conclusion about the ‘genuineness’ of his role, nor about his inability (still less unwillingness) to perform the role itself, as opposed to meet immigration requirements. Though para.6 is under the general heading of ‘Non-Genuine Vacancies’ above para.3 (before withdrawing reliance on the first two visa workers at paras.4-5), there is no reference to ‘non-genuine vacancies’ in para.6. Sensibly, Mr Yarrow did not argue either para.6 or para.12 found Mr F had a ‘non-genuine role’. Instead, he suggested against the context of the failure in recruitment practices involving Mr F; and the high staff turnover, the Decision at para.12 was referring to Ms K being ‘unable or unwilling’ to carry out the required duties of driving.

However, even assuming that is right, there is another problem with para.12 of the Decision which stated:

That first sentence correctly summarised C.1.44(1), so read with Ground (z), as driving was plainly a ‘duty’ of Ms K’s role (as made clear in her offer letter), the Decision should have examined whether there were reasonable grounds to believe Ms K was not ‘required’ to drive in her role; and if not, whether the role was ‘not genuine’. Instead, para.12 asked the wrong question and so got the wrong answer in finding Ms K’s inability to drive itself meant the job was not ‘genuine’. That was an irrational non-sequitur, as a worker recruited into a role apparently requiring driving when they cannot drive does not itself mean the driving role is not genuine: it could just be very poor recruitment for that role, just as with Mr F’s poor English. As I have said, on true interpretation, Ground (z) and C1.44(1) focus on what the role actually ‘requires’, not the worker’s ability to meet those requirements as such. Even against the context of poor recruitment practices shown by high staff turnover and Mr F’s selection, Ms K’s inability to drive in a role that apparently required it, did not itself ‘demonstrate’ reasonable grounds to believe her role was not genuine.

Of course, Ms K’s recruitment into a job supposedly ‘requiring’ driving when she clearly could not drive could have justified an inference that Ms K was not ‘required’ to perform the ‘driving duty’ contrary to C1.44(1) and so that the role was not ‘genuine’ under Ground (z), even without dishonesty or reprehensible conduct, as I have said. However, I do not accept there was any such inference in para.12 or elsewhere in the Decision. Even if there was, rather like Ms Rima’s discrepancy between job description and actual duties in R(SCL) at [138] (albeit of dishonesty, which as I have said is not the same as ‘non-genuineness’) any inference appears to have been drawn solely from the recruitment of Ms K into a driving role she could not perform. Certainly, as the Decision said, that demonstrated poor recruitment practice, supported by the conclusions about high staff turnover and Mr F. But the Decision went further to find that Ms K’s role was not ‘genuine’, without explaining why, or addressing the contrary evidence in her case. It would be one thing to infer Ms K’s role had been ‘non-genuine’ if she had retained it for a period despite being unable to drive. That might have showed she was not ‘required’ to perform her ‘driving duty’ under C1.44(1) so her role was not ‘genuine’ under Ground (z). Or it could have shown a mismatch between job description and duties performed under Ground (s), or an unapproved change in role under Ground (q). However, Ms K’s sponsorship was withdrawn within only two months. Assuming that was due to inability to drive as found, that suggests she was required to drive, so the role was genuine enough, but her recruitment for it was incompetent. In concluding Ms K’s role was ‘non-genuine’, the Decision irrationally failed to take into account this highly relevant consideration; or that like Ms Rima in R(SCL) at [138], Ms K was one of two specific and different examples of poor recruitment practice. Certainly, Ms K’s inability to drive ‘called for an explanation’. But that explanation was not ‘called for’ in the Minded to Find letter, which was not focussed on ‘non-genuine roles’, but on alleged breach of sponsor duties. I suspect the decision-maker in the Decision was led astray by the ‘own goal’ about Ms K in the previous solicitor’s Response letter into jumping to an irrational finding of ‘non-genuineness’.

Therefore, even against the background of the high staff turnover and the poor recruitment practice in Mr F’s and Ms K’s cases, I am driven to find the Decision’s conclusion that Ms K’s role was not ‘genuine’ under Ground (z) was irrational and I uphold Ground 3. But Mr Yarrow argued if I found such error, either that it was ‘immaterial’ at Common Law in the sense that the Decision would inevitably have been the same without it (see TJ Trading at [86]) or that s.31(2A) SCA applies:

Mr Yarrow takes two ‘bites at the s.31(2A) cherry’ and I shall have to consider it in more detail below, but there is a short answer to it salvaging Ground (z). It is difficult for s.31(2A) to apply (and even more difficult to find ‘immateriality’) without a fair opportunity to make representations: R(TJ Trading) at [91]. This is an objective issue for the Court: R(Prestwick) at [107] and at [28]-[29], Baker LJ cited Lord Neuberger in Bank Mellat v HM Treasury (No 2) [2014] AC 700 [179]:

This basic rule was developed in Balajigari at [55] and R(SCL) at [132]-[136] into stringent requirements for allegations of dishonesty and reprehensible conduct. So, in R(SCL), it was not even enough for UKVI to give the sponsor the chance to make representations before revocation, even where the suspension letter alleged a ‘non-genuine role’, as it did not clearly allege exaggeration to facilitate stay in the UK. But the more basic rule of procedural fairness in Bank Mellat applies generally. Whilst unfairness is not alleged by the Claimant, the burden lies on the Defendant to show procedural fairness for ‘immateriality’ and/or s.31(2A) SCA.

The ‘conduct complained of’ is the irrational finding that Ground (z) applied because Ms K was recruited for a role requiring driving she could not do and ‘if it had not occurred’ requires me (as I explain later) to consider the impact of the error on the decision-making process in the Decision. As I have explained, it would not necessarily have been irrational to infer Ms K was not ‘required’ to perform her driving duties under C1.44(1); so her role was not ‘genuine’ under Ground (z). But there are three obstacles to ‘immateriality’/s.31(2A) ‘repairing’ Ground (z) like this:

Mr Yarrow’s main submission on s.31(2A) SCA is not to ‘repair’ the Decision’s error in Ground (z) itself, but that ‘if the conduct complained of had not occurred’ in the sense of even without any reliance on Ground (z), it is ‘highly likely’ the Decision would have revoked the Claimant’s sponsor licence under Annex C2 Grounds (a) and (b). Mr Yarrow’s Skeleton Argument encapsulates it this way:

I have already found the factual findings in the Decision were rational. I must now consider whether they justified revocation under Annex C2 Grounds (a)/(b).

I repeat key parts of the Sponsor Guidance, starting with those revocation grounds:

The latter plainly does not require breach of actual sponsor duties (c.f. R(St Andrews College at [65]), simply lack of processes necessary to ensure full compliance with them: R(Tendercare) at [113]-[120]. The relevant duties on sponsors are at C1.38:

As Mr Gajjar placed significant reliance on it, I also repeat C7.26:

I need not repeat all the principles on revocation summarised in R(Prestwick) at [19]-[24] I quoted at paragraph 25 of this judgment, although obviously I bear it all fully in mind. But this observation by Tomlinson LJ in R(Raj Knoll) at [28] (not cited to me directly but extensively cited in R(Prestwick) ) is instructive:

Turning to the constituent elements of the Decision in this case, whilst I have upheld Ground 3 relating to the finding that Ms K’s role was not ‘genuine’ under Ground (z), I have also found that the Decision at paras.10-11 rationally concluded Ms K was recruited for a position that she was unable to carry out, as she had told the Claimant in her application form. I have also found that whilst the conclusion in para 12 of the Decision that this meant Ms K’s role was ‘not genuine’ was irrational, it was entirely justified for it to conclude in the alternative that Ms K’s recruitment was a ‘severe failure in recruitment practice’. Indeed, turning to the Guidance, this was also a clear and obvious breach of sponsor duties in C1.38, as Ms K was recruited when she was not appropriately qualified, or at least able, to fulfil her role. Therefore, the conclusion that Annex C2 Ground (a) was made out is indisputable. Moreover, since the Claimant’s recruitment process with Ms K breached sponsor duties, the Defendant was also plainly rationally entitled to be ‘not satisfied the Claimant was using a process necessary to fully comply with its sponsor duties’ and so Ground (b) was also proved. Therefore, I only uphold Ground 3 on ‘Ground (z)’.

Similarly, with Ground 2, I concluded the Decision rationally found recruitment of Mr F and the other worker were failures of recruitment practice even though the Defendant later withdrew this allegation about the other worker once it had a proper explanation. However, at the time of the Decision, the Defendant was entitled to conclude that the Claimant had employed two workers whom it did not have proper grounds to believe would meet immigration requirements, or were qualified to do the job, in breach of sponsor duties under C1.38 so satisfying Annex C2 Ground (a). Moreover, taken with Ms K, these were other quite different instances of poor recruitment practice reinforcing that Ground (b) was met. However, even focussing on Mr F alone and ignoring the (entirely rational, if understandably incorrect) finding with the other, the Decision was entitled to find Mr F’s recruitment satisfied Grounds (a) and (b) and as it did not apply Ground (z) to him. Therefore, I dismiss Ground 2.

Clearly, if Ms K and Mr F had been only two ‘isolated failures’ as Mr Gajjar said, there might have been real debate about whether the Claimant’s recruitment processes were a threat to immigration control. However, as I found at paragraph 64 of this judgment, the Decision actually reached a fresh calculation, based on the Claimant’s own staff list, that turnover was actually over 60%. But even assuming it was ‘only’ 40%, that is 40% higher than the 28.3% average for the care sector. I found the conclusions in the Decision on turnover were entirely rational. Ms K and Mr F were only examples of a wider problem with the Claimant’s recruitment, most clearly evidenced by its staff turnover. As I also found, the Decision was entitled to conclude that high staff turnover raised concerns over the Claimant’s recruitment practice generally (not just with Mr F or Ms K) which the Defendant ‘considered a threat to immigration control’. So, the Decision did not simply have concerns about that threat, it concluded that the Claimant’s recruitment practice actually posed a threat to immigration control. As in R(St Andrew’s College) at [29], the sponsorship regime is based on trust and there is no need for UKVI to wait until there has been a breach of immigration control before acting. Indeed, in this case, the Defendant was entitled to conclude that with such lax and inadequate recruitment processes, migrants who would not meet immigration criteria (such as but not limited to Mr F) or could not actually do the job (such as but not limited to Ms K) were being recruited. Therefore, Annex C2 Ground (b) was plainly satisfied by the high staff turnover, as the Defendant was rationally ‘not satisfied you are using a process or procedure necessary to fully comply with sponsor duties’. Moreover, while not explicitly spelt out in the Decision, it implicitly also found breach of sponsor duties and Ground (a) as Mr Yarrow said due to sponsor’s duties in C1.3 to ‘play their part in ensuring the immigration system is not abused’. So, I also dismiss Ground 1.

Therefore, I turn back to s.31(2A) SCA on the basis that the only relevant ‘conduct’ I must treat as having ‘not occurred’ in the Decision is its findings on Ground (z). I have already set out s.31(2A) SCA, but for ease I repeat it, along with s.31(2B):

The overall purpose of s.31(2A)/(2B) was encapsulated by Coulson LJ in R(Gathercole) v Suffolk CC [2021] PTSR 359 at [38] (quoted in R(Treal) at [91]):

Both Counsel cited R(TTT) v Michaela CST [2024] PTSR 1627 at [269], where Linden J summarised principles relating to the application of s.31(2A) drawn up in an earlier case (citations and (viii) with that quote from R(Gathercole) omitted, although I also flag up queries about principles (iii) and (iv) which I discuss below):

However, this summary (especially points (iii) and (iv) as noted above) was recently questioned in R(Bradbury) v Brecon Beacons Park [2025] 4 WLR 58 (CA) by Lewis LJ at [73]-[75] (not cited to me but cited in R(J’s Supermarkets) at [104]). So, I have also cited R(Gathercole) (where s.31(2A) applied to a technical non-compliance with the Public Sector Equality Duty in a planning decision which made no difference to it). In R(Bradbury) itself, Lewis LJ held whilst a planning decision was flawed due to the absence of adequate wildlife assessments, s.31(2A) applied since those assessments would have led to the same conditions actually imposed anyway. Lewis LJ explained the operation of s.31(2A) at [70], [71] and [74]:

R(Bradbury) makes it clear that s.31(2A) has a narrower scope than principles (iii) and (iv) in R(TTT) might suggest. The distinction Lewis LJ drew at [71]/[74] is between wrongly ‘predicting what the public body would have done if the error had not been made’ and correctly ‘focussing on the impact of the error on the decision-making process to ascertain whether it is highly likely the decision the public body took would not have been substantially different if the error had not occurred’. Therefore, I must not substitute my own decision for that of the Defendant, nor predict what counterfactual decision it might have made without the error. I must decide what impact that actual error had on the actual decision and whether its outcome would have been the same or not substantially different without that error.

R(Bradbury) was applied to sponsor revocation in R(J’s Supermarkets) at [110]-[115]. Having quoted Lewis LJ in R(Bradbury), DHCJ Atkinson KC considered two different s.31(2A) arguments. His conclusions on both were obiter, as he had already decided UKVI lawfully decided that roles were not ‘genuine’ under Annex C1 Ground (z) for deliberate misstatement of duties to facilitate entry into the UK. However, he considered at [110]-[112] that if Ground (z) revocation had been flawed, it was highly likely the rationale for it in the decision would have led to revocation under Ground (s), even though it was not explicitly referred to. In short, the revocation decision would have been the same, but with a different ‘label’. However, DHCJ Atkinson KC said he would not have been able to find s.31(2A) applied to the alternative finding in the decision of Annex C2 Grounds (a) and (b), for reasons which he – again understandably briefly - explained at [114]-[115]:

Nevertheless, especially in the context of it being a second unrequired fall-back argument, I do not read DHCJ Atkinson KC as saying if an Annex C1 ground in a revocation decision is flawed, s.31(2A) cannot apply to an alternative finding of revocation on an Annex C2 ground. Indeed, in that case, Annex C2 Grounds (a) and (b) were not direct alternatives to Ground (z), but comparatively minor reporting and information breaches, so it is unsurprising that s.31(2A) did not apply to them.

Unlike R(J’s Supermarkets), the Decision in this case did revoke under Annex C2 as an express alternative to it under Annex C1. Para.21 of the Decision stated:

I underline that to show that revocation was based not only on the flawed Ground (z), but the justified and rational Grounds (a) and (b). Mr Yarrow briefly argued this meant any error on Ground (z) was ‘immaterial’ at Common Law. But I cannot find the error on Ground (z) was so inconsequential as to be ‘immaterial’. Indeed, as I have explained, I also cannot find under s.31(2A) that it is highly likely if the error on Ground (z) had not been made, a different lawful route for the Defendant to conclude Ground (z) was proven would have led to the same revocation decision on the basis of Ground (z).

However, since the Decision to revoke the Claimant’s licence was explicitly framed in the alternative in para.21, when ‘evaluating the significance of the error on the decision-making process’ as Lewis LJ put it in R(Bradbury), I can conclude there was an express alternative - and I have found, entirely rational - reason for the revocation of the Claimant’s licence explicitly stated in the Decision itself. In my judgement, Grounds (a) and (b) were totally ‘untainted’ by the error in Ground (z) (especially on high staff turnover which was found a threat to immigration control and Mr F which did not relate to Ground (z) at all, but even with Ms K which did given the alternative finding at para.12). Subject to one point, in my judgement, I could therefore be satisfied that it is ‘highly likely’ that without Ground (z) the actual Decision would still have revoked the licence.

That one point is that raised in R(J’s Supermarkets) at [114]-[115] quoted above: that Annex C2 grounds do not inexorably lead to revocation, unlike Annex C1 grounds (subject to the unusual exercise of residual discretion outside the Guidance analysed in R(Tendercare) and accepted in R(Prestwick)). The distinction within the Guidance is clear from the difference between C10.4 and C10.5 – 10.6:

I would also reiterate that C10.7 states that for an Annex C3 ground (which does not arise here), UKVI ‘may revoke’ the licence but will ‘generally’ not do so.

Had the Decision explicitly stated that even if Ground (z) did not apply, Grounds (a) and (b) did still apply and by themselves justified revocation of the Claimant’s licence and there were no ‘exceptional circumstances’ not to do so, the application of s.31(2A) would have been entirely straightforward. As discussed in argument, this is rather like judges making ‘alternative findings’ in their judgments. However, the Decision did not explicitly address ‘exceptional circumstances’. Instead Paragraph 20 of the Decision stated:

It then quoted C10.4 which (as just quoted above) makes clear Annex C1 grounds are ‘mandatory’. Therefore, rather than saying that the Annex C2 grounds (a) and (b) in the Decision justified revocation rather than downgrading and there were no ‘exceptional circumstances’, the Decision said that downgrading was unavailable because there was an Annex C1 ground. On the other hand, the Decision did not say that only downgrading not revocation was justified by the Annex C2 grounds. Indeed, that would be inconsistent with para.21, which said each of Grounds (z), (a) and (b) ’individually, is a necessary basis for revocation’. Therefore, para.20 does not seem to me to prevent the application of s.31(2A), but rather to be part of the ‘conduct complained of’ (i.e. the legal error) in relation to it. So, I must consider whether the impact on the Decision as a whole of the Ground (z) error (including its impact in relation to the discretion not to revoke under Annex C2) means that without that Ground (z) error, it is or is not highly likely that the Decision would still have revoked the Claimant’s sponsor’s licence under Grounds (a) and (b).

Mr Gajjar submitted that as in R(J’s Supermarkets), the Decision did not consider whether any ‘exceptional circumstances’ would justify not revoking and did not (even without using that phrase) implicitly consider mitigating circumstances. He pointed out the Decision at para 20 had relied on a legally flawed mandatory ground. While s.31(2A) might have applied if there had been a free-standing decision on Annex C2 Grounds (a) and (b) explicitly rejecting ‘exceptional circumstances’ and ‘downgrading’, there was not. So, I could not be satisfied it was ‘highly likely’ revocation would still have occurred if the error on Ground (z) had not occurred. Mr Yarrow retorted whilst ‘exceptional circumstances’ had not been explicitly referred to in the Decision, s.31(2A) did apply. I draw his submissions into three points: (i) no ‘exceptional circumstances’ had been raised by the Claimant; (ii) there was nothing that could amount to ‘exceptional circumstances’; or (iii) even if there was, there was sufficient consideration in the Decision, in the context of the Minded to Find letter, to have effectively ruled out ‘exceptional circumstances’ or downgrading (mentioned, albeit briefly), without the Decision explicitly saying so. He said these three points could be reflected in different routes to s.31(2A) applying.

Mr Yarrow’s first two points raise the meaning of ‘exceptional circumstances’ as a matter of objective interpretation of language of the Guidance in context (Tesco), ‘sensibly, purposefully and holistically not rigidly and myopically’ (St Andrews). I will not suggest the sort of ‘checklist’ deprecated in R(Bradbury), but there are five points from the cases which are relevant to ‘exceptional circumstances’ in this case:

Here, ‘exceptional circumstances’ under C10.6 were simply never invoked by the Claimant’s previous solicitors in either of their representations before the Decision. As I pointed out at paras 11 and 12 of this judgment, the Minded to Find letter invited the Claimant to ‘alleviate its concerns’ and if it was not already obvious from that, as made clear by C9.11 of the Guidance, this was an opportunity for the Claimant to put forward any ‘mitigating arguments’ (including ‘exceptional circumstances’ given alleged breaches of sponsor duties). But the Response letter barely put forward any ‘mitigating arguments’ at all, let alone ‘exceptional circumstances’. Far from ‘alleviating concerns’, it was antagonistic. Certainly, it raised nothing that could have either justified or mitigated the identified failings (including high staff turnover, Mr F and Ms K), or given reassurance as to the Claimant’s compliance more generally. It is true Mr F and Ms K were only two specific examples of administrative error out of 139 workers for the Claimant. But more importantly there was very high staff turnover. Rather than alleviating the Defendant’s legitimate concerns about that, the Response letter suggested those concerns were irrational and its stance generally ‘tantamount to an abuse of power’. The Response letter failed to refer to impact of revocation on the Claimant let alone anyone else. Whilst the chasing email mentioned staff resignations, this was a complaint about the delay in making a decision, not the consequences of revocation.

Bearing those points in mind in this case, I accept Mr Yarrow’s submissions that s.31(2A) SCA applies and it is ‘highly likely’ that had the error on Ground (z) not occurred, the Decision would still have been to revoke the Claimant’s sponsor’s licence perfectly lawfully and rationally under Grounds (a) and (b) alone. Indeed, I make clear there are no fewer than five independent routes to that destination.

The first route to s.31(2A) applying is the Decision explicitly stated that revocation was justified under Grounds (a) and (b) irrespective of Ground (z) and had no need to mention ‘exceptional circumstances’ as none were raised or present: (R(Raj Knoll). Para 21 stated explicit alternative revocation grounds and para 19 said (my italics):

When analysing ‘the significance of the error on the decision-making process’, here the Decision itself expressed the conclusion to revoke as for failure to comply with sponsor duties (i.e. Annex C2 Ground (a)), without even relying on the flawed Ground (z). So, in this case at least, it is not impermissibly ‘counterfactual’ but entirely legitimate to test ‘the impact of the Ground (z) error in the decision-making process’ by applying a proverbial ‘blue pencil’ through the erroneous parts of the Decision (especially paras.12, 20 and 21 along with the heading of ‘Non-Genuine Vacancies’ and quotations of C1.44, 1.46 and Ground (z)). Doing so makes it clear from what remains that it is not only ‘highly likely’ but explicit in the Decision itself that without the Ground (z) error, the Claimant’s licence would still have been revoked permissibly under Grounds (a) and (b). Moreover no ‘exceptional circumstances’ were raised, perhaps because none were present. Moreover, despite his skill, Mr Gajjar was unable to identify any ‘exceptional circumstances’ which could have applied here.

If I am wrong about that, the second route for s.31(2A) to bite is that Decision explicitly and rationally found revocation was justified under Grounds (a) and (b) (at paras.19 and 21 and applying the ‘blue-pencil’ just discussed) for reasons fundamentally inconsistent with any ‘exceptional circumstances’ that could have been raised. As I have explained, the primary ‘steer’ in C10.6 on the meaning of that undefined concept is that it includes the adequacy of the sponsor’s general processes, procedures and reporting not limited to actual breaches, but it is narrower than the unrestricted discretion in Annex C3: R(Tendercare). Whilst there was no negative ‘compliance history’ or under-reporting, by the Decision itself the Defendant had assessed the staff turnover as 60%: even higher than the 40% previously suspected. Either way, the Decision rationally concluded the Claimant’s recruitment processes posed a threat to immigration control. This was a ‘more serious breach’ under C7.26: not just a ‘serious threat to immigration control’, but a ‘significant or systematic failing’. This manifested itself in two specific examples of serious failings in the recruitment process: Mr F and Ms K. Those were not ‘isolated examples’ but instances of the underlying problem. If they had been isolated instances, it is possible that with a clean compliance history, there might have been arguable ‘exceptional circumstances’ to avoid revocation, but the rational finding about high staff turnover threatening immigration control was simply logically inconsistent with any such ‘exceptional circumstances’ to avoid revocation. The Minded to Find letter did fairly ‘look for evidence’ from the Claimant relating to its compliance history and procedures. However, the Response Letter failed to put forward any significant circumstances relating to that issue, the reasons for the breaches, or the impact of revocation, even if the chasing email mentioned resignations. Even the evidence of Mrs Chipatiko, well after the Decision, simply outlined the normal consequences for a care business, workers and clients of revocation of sponsorship: unfortunately, there was nothing ‘exceptional’ at all about any of it. Therefore, I am entirely satisfied it is highly likely the Claimant’s licence would still have been revoked in the Decision if it had not made the Ground (z) error.

If I am wrong about that, the third route for s.31(2A) to apply also relies on the clear alternative finding of revocation under Grounds (a) and (b) separate from Ground (z) but assumes (which I do not accept) that it would be impermissible to ‘double-count’ the high turnover as part of Grounds (a) and (b) as well as negativing any ‘exceptional circumstances’. In that case, there are independent and rational conclusions on Grounds (a) and (b) in relation to Mr F and Ms K alone, whilst the reasoning in relation to high staff turnover is inconsistent with any ‘exceptional circumstances’ as just explained. Whilst this is not how the Decision was structured, it is no more than using the existing rational reasoning in the Decision and re-labelling parts of it, just as DHCJ Atkinson KC in R(J’s Supermarkets) said that if he had found Ground (z) had been flawed, s.31(2A) would have justified the revocation under Annex C1 Ground (s). Likewise, I am entirely satisfied it is highly likely the Claimant’s licence would still have been revoked in the Decision if it had not made the Ground (z) error.

If I am wrong about that as well, the fourth route to s.31(2A) applying is that even if (which I do not accept) the Decision was insufficiently clear that revocation was justified under Grounds (a) and (b) and/or there were no ‘exceptional circumstances’, that was a conclusion entirely rationally open to the decision-maker and consistent with the Decision, for reasons entirely untainted by its error under Ground (z). Given that, I am satisfied that it is highly likely the Claimant’s licence would still have been revoked in the Decision if it had not made the Ground (z) error, because:

The final route to s.31(2A) applying is that I am satisfied that it is highly likely that the Claimant’s licence would still have been revoked in the Decision if it had not made the Ground (z) error under Grounds (a) and (b) because that is what the Minded to Find letter had said would happen if its concerns were not alleviated, which they were not. For the reasons just explained, the underlying reasoning in the Decision on Grounds (a) and (b) was unimpeachable and unaffected by the Ground (z) error and was also inconsistent with ‘exceptional circumstances’. Grounds (a) and (b) established in the Decision were essentially those stated in the Minded to Find letter: high staff turnover (albeit then 40% not 60%) failed visa applications (albeit then four) and Ms K’s recruitment. The Minded to Find letter made clear all of those were alleged to be breaches of sponsor duties and explicitly added this:

Whilst that did not mention ‘exceptional circumstances’, the Response Letter was plainly the chance for any such ‘mitigating arguments’ to be presented (see C9.11 and R(Prestwick) at [114]). Notably, as Mr Yarrow pointed out, the Minded to Find letter did not allege Ground (z). Therefore, even if (which I do not accept) the reasoning in the Decision on revocation itself at paras.12, 20 and 21 (rather than the rational three underlying factual conclusions and on Grounds (a) and (b)) was wholly flawed, the Minded to Find letter had rationally and fairly made clear it would revoke under those grounds unless the Claimant alleviated its concerns. Far from doing so, the Response Letter aggravated the Defendant’s concerns and ironically prompted the legal error on Ground (z) in the Decision. Therefore, I am entirely satisfied it is highly likely the Claimant’s licence would still have been revoked in the Decision if it had not made the Ground (z) error that had been totally absent from the Minded to Find letter.

Therefore, for any one of these different reasons (which I make clear, especially in the present case would each individually justify s.31(2A) applying), I dismiss Grounds 1 and 2 of the challenge and whilst I would have upheld Ground 3, I refuse relief under s.31(2A) SCA. As in R(Gathercole), whilst there was a flaw in the Decision on Ground (z), quashing the Decision which is highly likely to have been effectively the same without reliance on Ground (z) at all, would be a waste of time and public money and indeed inflexible and unrealistic. It was not suggested by Mr Gajjar there would be ‘exceptional public interest’ under s.31(2B) so as not to invoke s.31(2A) and there is not. Indeed, since there was no departure in the Decision from the standard 12 months’ cooling off period in C10.16 Guidance, by the time of this judgment, the Claimant is able immediately re-apply for a sponsor licence, having also ‘cleared its name’ of the Annex C1 revocation for ‘non-genuine vacancies’. However, as made clear at C10.18, the Claimant’s application must ‘demonstrate it has addressed any reasons why the previous licence was revoked’ before it reapplies. I offer no comment either way on the prospects of a re-application, which is a matter for the Defendant.

By agreement after the draft judgment, I directed the parties to make written submissions on permission to appeal and costs. Whilst I provided for the Claimant to make both initial submissions and to reply to the Defendant’s submissions, the Claimant did not file any reply. Nor did it object to summary assessment, which for a one-day hearing and a modest claim for costs is plainly proportionate. CPR 44.2 provides:

I am dismissing the claim, so the Defendant is unquestionably ‘the successful party’, whose costs should in principle be paid by the Claimant. Mr Gajjar did not dispute that but did make two points on costs. Firstly, that the Defendant was late in filing and serving his statement of costs which can be ‘taken into account’ by the Court on costs under CPR 44 PD 44 para.9.6 and Simpson v MGN [2015] EWHC 126 (QB). However, as Warby J (as he then was) said in Simpson, it may not be just or proportionate to refuse an order and costs may be reduced (they were reduced by 10% in that case where it had caused delay, but alongside another issue with a costs budget). In this case, the Defendant’s delay in serving a costs statement was technical and had no impact at all, as I reserved judgment. Nevertheless, I shall ‘take it into account’ under para 9.6. More importantly, under CPR 44.2(4) and (5), whilst the Defendant was the ‘successful party’, it did not succeed on all issues. In particular, whilst I agreed with the Defendant that Ground (z) was not limited to ‘dishonesty or other reprehensible conduct’, I found the conclusion on Ground (z) in the Decision was irrational and not saved by s.31(2A) SCA. However, I dismissed the claim because I held that s.31(2A) SCA applied more generally because the Defendant also revoked under Grounds (a) and (b) which were rational. Whilst R(M) v Croydon LBC [2012] EWCA Civ 595 was cited, that was a case of a settlement. Yet as Mr Gajjar said and the cases on CPR 44.2(6)(a) cited in the White Book confirm, the Courts should be ready to reduce costs by a percentage to reflect the failure of the overall ‘successful party’ on particular issues and I agree with Mr Gajjar that a proportionate costs order under CPR 44.2(6)(a) is appropriate. However, as Mr Yarrow observed, in Multiplex v Cleveland Bridge [2009] 1 Costs LR 155, Jackson J (as he then was) said at [71(viii)]:

Here the Defendant’s costs schedule is £21,712.70 (less than a quarter of the Claimant’s costs schedule). Understandably, it does not break down the proportion of costs on particular issues, but given the overall modesty of the schedule, most of the costs expended are likely to have been common costs. Nevertheless, there should be some deduction by a proportion to reflect that the Defendant did not succeed on all issues and that its costs schedule was (slightly) late. Mr Gajjar did not suggest a percentage reduction and nor did Mr Yarrow. Bearing in mind the Defendant won the argument of principle on interpretation of Ground (z), but lost on its application in the Decision (and won on Grounds (a) and (b) and on the case generally), the appropriate reduction should be modest. Doing the best I can on a ‘broad-brush’ basis, I will make a proportional costs order making a deduction of 15%. On an overall figure of £21,712.70, following a deduction of 15%, given six hours preparing for a one-day hearing is on the high-side for a Defendant, I will also round-down to an overall figure for costs (including VAT) of £18,000. Bearing in mind the Claimant is a small business and the Defendant is a Government department, I will order payment within 3 months. Since the Claimant did not file a reply, I will give it 7 days to apply to vary my order on costs or the timescale for payment and will extend the overall time to appeal accordingly. But if no application is made or I reject the application to vary, my order will stand.

Mr Gajjar sought permission to appeal on three bases: (i) that my conclusion in paragraph [55(iv)] that Ground (z) did not necessarily entail a finding of dishonesty or reprehensible conduct; (ii) that I erred at paragraph [72-73] in finding that Ms K’s role including a duty to drive; and (iii) that I erred at paragraph [75] about the application of s.31(2A) SCA. I refuse permission to appeal, not least because all these criticisms relate to Ground (z) and ‘non-genuine roles’ on which the Claimant actually won on the facts, as I have just explained. I do not say that these challenges are academic, but they are not targeted at the point on which the Claimant actually lost the case: s.31(2A) and Grounds (a) and (b) of the Sponsor Guidance. Moreover Mr Gajjar’s points (ii) and (iii) are also fact-specific and of no wider application, as well as being not arguable in my view for the reasons I have given. I recognise that (i) is of wider application to other cases, but there is no compelling reason to grant permission in this case when it was not determinative of the result; and moreover, I considered that interpretation issue in detail and even on careful reflection do not consider my interpretation of Ground 1(z) of the Sponsor Guidance is arguably wrong. In any event, given the Court of Appeal has given a judgment so recently in R(Prestwick), I consider it is much better placed than me to consider whether to grant permission to appeal on that basis. Likewise, whilst the key basis of my decision on s.31(2A) SCA and Grounds (a) and (b) is not challenged in the application for permission to appeal, I do not consider those decisions give rise either to a compelling reason to grant permission as it is case-specific; and since I have reached that conclusion by five alternative routes, I do not consider a challenge is arguable, but again that is a matter for the Court of Appeal. For those reasons, I dismiss the claim, refuse permission to appeal and order the Claimant to pay the Defendant £18,000 in costs within 3 months. I am particularly grateful to Mr Gajjar and Mr Yarrow for their skilful submissions and assistance.