Inclusion Housing Community Interest Company v Regulator of Social Housing

Neil Brown, Inclusion’s CEO since July 2015, explains in his evidence that Inclusion is a health and social care landlord providing specialist supported housing (‘SSH’) for adults with physical disabilities and/or mental health needs. The accommodation has a range of different specifications depending upon the specific client group. This includes accommodation with specific adaptations, such as wet rooms, adapted kitchens, provision for alarm call, communal areas, door entry access and provision for future adaptations. There is a growing demand for this type of accommodation from local authorities which must meet statutory obligations to house vulnerable adults.

Mr Brown explains that, traditionally, regulated providers own the properties that they then let to tenants. In order to develop their businesses these providers secure funding, bonds and loans against these assets. They generally fund purchases with large loans against which they must make regular repayments. They are often dependent upon grants from central government to subsidise growth. Inclusion’s business, by contrast, operates on a ‘lease-backed model’ under which the properties are owned by investment funds, which lease them to Inclusion as head-lessor, with Inclusion then letting the properties to individual tenants.

In the majority of cases, Inclusion works with a developer to provide new SSH units, either new-build or by redevelopment of existing property. At the same time, Inclusion obtains a ‘nominations agreement’ (‘NA’) from a care provider or care commissioning body. Care providers are predominantly local authorities, but may also be private sector bodies contracted by local authorities. Care commissioning bodies are generally NHS entities. Under the NA, the care provider or care commissioning body agrees with Inclusion that it will nominate residents to rent the property from Inclusion once it is ready. In many cases, the counterparty to the NA commits itself over a substantial period to paying the rent for any voids and to covering any shortfall in rent.

Mr Brown says that Inclusion’s business model has a number of advantages over the traditional model. In particular, he explains that it promotes the development of new SSH, of which there is a shortage nationally. Moreover, unlike the traditional model, no public subsidy is required. Bringing newly developed SSH into the market has a series of knockon social benefits. It promotes independent living by those who would otherwise have to live in care homes. It reduces pressure on the public purse. Because the accommodation is often new and purpose-built, it results in high customer satisfaction. It also offers tenants security of tenure.

Inclusion has been a registered provider since June 2011. It began expanding substantially in 2015. At that time, Inclusion’s board commissioned consultants Campbell Tickell to report on the current and future viability of the business. Campbell

Campbell Tickell undertook an analysis of Inclusion’s tenancies. It noted that the majority of leases were ‘long dated’, with 41% exceeding 30 years. This, it was noted, was ‘a very long-term obligation and relies entirely on the sustainability of the income stream and in particular that (i) the existing housing benefit regime will continue for existing tenants and (ii) on re-letting, that the housing benefit regime and sponsorship of the local authorities will apply to new tenants’. The majority of the leases required

Campbell Tickell drew a series of conclusions which included the following:

Campbell Tickell recommended that Inclusion should do a number of things, including ‘re-negotiate leases with a view of de-risking the transactions’ and ‘[a]t the earliest opportunity to engage with the Homes and Communities Agency’, the then regulator.

Inclusion’s board met to discuss the Campbell Tickell report on 2 July 2015. David Williams of Campbell Tickell is recorded as advising that ‘if [Inclusion] replicated past negotiations with partners in the future, there would be problems for [Inclusion] because the model is flawed’ and that Inclusion should ‘in all cases make sure that [it] does not repeat the same lease terms as in previous years’. This, it was said, ‘should be done before any further contracts and leases are entered into’. An action plan was quickly devised, which referred – among other things – to the need to ‘review terms of leases and compile change strategy’.

Campbell Tickell’s advice was immediately to share the contents of their report with the then regulator, the HCA. There was a meeting on 17 July 2015. On 21 July 2015, the chairman of Inclusion’s board, Pete Ottowell, wrote to the HCA’s Senior Regulation Manager setting out Inclusion’s specific responses to the Campbell Tickell report.

Because the process of re-negotiating leases necessarily depended upon the co-operation of counterparties, it was not possible to give specific assurances as to the outcome of the process and no such assurances were given. A tabular document was produced and shared with the HCA, entitled ‘Inclusion High-Level Project Plan for HCA to be supported by detailed plans in all areas’. In the row dealing with leases, the comment was: ‘Review terms of leases and compile change strategy’. Next to this was an update: ‘Meeting Ward Hadaway 9/7/15. The drafting of standard lease underway to enable discussions with partners.’ Again, no assurance could be or was given as to the precise upshot of these discussions.

Ward Hadaway reviewed Inclusion’s standard lease and void agreement and proposed amendments to both. Among the amendments proposed was:

On 3 August 2015, the HCA wrote to Mr Ottowell indicating that it had considered Inclusion’s compliance with its regulatory obligations in light of the issues Inclusion had brought to its attention. As to the financial viability standard, it said this:

As to governance, the HCA indicated that, in the light of Inclusion’s proposals, it was satisfied that the Standard was met. Under the heading ‘Conclusion’, the HCA said this:

For most providers, IDAs take place every three or four years. However, their frequency is linked to the regulator’s assessment of the relative risk profile of the provider. When an IDA is begun, the regulator produces a document setting out its proposed scope. In this case, the scoping document was appended to a letter dated 27 February 2018. As part of the IDA, staff from the regulator attended Inclusion’s board meeting on 17 April 2018 and a fieldwork day was scheduled for 10 May 2018 to enable staff from the regulator to verify and substantiate information received from Inclusion. On 23 May 2018, the regulator decided that Inclusion should be placed on its list of providers with ‘gradings under review’ (‘GUR’). The purpose of this list is to alert stakeholders to the possibility that the provider may be moving towards non-compliance.

The reasons for Inclusion’s GUR listing are recorded in the regulator’s Reactive Engagement Decision (‘RED’) log and in notes of internal meetings. In essence, they were these. As at 31 March 2018, the number of units leased and managed by Inclusion had grown from 433 units (at 31 March 2015) to 1548 units. Of those, 1435 units had long-term leases of at least 20 years with no break clauses. The leases made Inclusion responsible for the full repair, management and insurance costs of the units as well as dilapidations at the point the leases were handed back. All the leases were subject to annual increases. 78% of them had annual increases in excess of CPI. The regulator considered the impact of a one-off housing benefit reduction, the position on voids and whether Inclusion was sustainable without growth. The regulator considered that there was a material risk of non-compliance with the Standard. The decision log records:

On 24 May 2018, staff from the regulator telephoned Neil Brown of Inclusion and explained that Inclusion was being listed as GUR. The listing was made public on 29 May 2018.

On 8 June 2018, there was a meeting between staff from the regulator and staff from Inclusion. This was summarised in a note. The relevant parts of the summary are as follows:

Mr Ottowell emailed after the meeting noting that it was ‘helpful to receive clarification from you in more detail on your areas of concern’ and confirming Inclusion’s intention to work with the regulator over the coming weeks to provide the further reassurance required.

On 19 June 2018, the regulator wrote to Inclusion recording the ‘key issues from the regulator’s perspective that had led to inclusion being placed on the GUR’. These were said to include:

That prompted a further email from Mr Ottowell to the regulator on 20 June 2018 in an attempt to provide ‘further reassurance’, attaching various documents, including a risk appetite document, a risk mitigation stress testing table, a risk register, a risk management framework and a contingency plan.

On 22 June 2018, Inclusion’s solicitors, Capsticks, wrote to the regulator noting that fairness required that Inclusion have a full and complete understanding of the regulator’s concerns about its compliance with the Standard before making representations so that it could address those concerns on a fully informed basis. The regulator was therefore asked to identify the elements of §§2.4, 2.5(b) and 2.6 with which it considered Inclusion was non-compliant and to explain its reasons.

The regulator wrote back on 29 June 2018 setting out some of the history of the engagement between Inclusion and the regulator. It said:

Capsticks responded on 2 July 2018, noting that the regulator’s position appeared to be that Inclusion should attempt to piece together the regulator’s concerns from a combination of its general policy documentation, two narrative paragraphs in its letter of 22 June 2018 and various emails letters, telephone conversations and meetings over the previous two months. Capsticks then invited the regulator to provide a single, clear, transparent and comprehensive statement of its concerns. They pointed out that the regulator’s letter of 22 June 2018 referenced as relevant to its concerns three paragraphs of the Standard (§§ 1.1, 1.2 and 2.2) which had never previously been raised.

The regulator responded on 11^th July 2018. It was not accepted that there had been a failure to provide sufficient information to enable Inclusion to make representations. It was the regulator’s expectation that, as a registered provider, inclusion would be fully aware of the requirements placed upon it by virtue of its registration. The regulator required providers to comply with the published standards and to be able to demonstrate such compliance. The reasons why the regulator was not satisfied as to Inclusion’s compliance with the financial viability and governance standards had been made clear throughout the IDA process. However, ‘for the avoidance of any possible doubt’, the ‘particular areas of regulatory concern’ were identified. These included, under the heading ‘identification and mitigation of risk’ the following:

Later in July 2018, Messrs Brown and Ottowell provided a detailed set of representations on behalf of Inclusion. These representations ran to some 121 paragraphs and were accompanied by 16 appendices. At §4 it was said that:

The representations contained a detailed description of Inclusion’s business model. It was explained that lease agreements were generally for periods of 20 to 30 years. Under these agreements, Inclusion would pay rent, which increased each year by reference to inflation. Its older lease agreements were linked to increases in RPI, but its more recent ones were linked to increases in CPI (which were in practice more modest) (§16.4). In the vast majority of cases, at the same time as entering into an agreement for lease with a developer, Inclusion would enter into a NA with a care provider or commissioning body. These NAs were reasonably long-term: typically 5-10 years in the case of agreements with care providers and up to 20 years in the case of agreements with local authorities. About 80% of Inclusion’s units in management were managed under NAs where the care body had a duty to cover rental income and service charge losses after a specific time until the empty unit had been re-tenanted. About 70% of Inclusion’s units in management were managed under NAs according to which the care body was obliged to reimburse Inclusion in the event of a rent shortfall. As to the 7% of Inclusion’s units not covered by an NA with a care body, Inclusion maintained a void self-insurance arrangement (§16.6). Furthermore, Inclusion would enter into a direct occupancy agreement with individual tenants. Many of these were long-term. The tenants were required to pay rent to Inclusion at a level calculated on a cost recovery basis. This covered Inclusion’s obligations under the main lease agreement, its cost of repairing the properties, a 10% void allowance (the self-insurance arrangement), and 8% sinking fund fee for major property repairs and 15% management fees (or calculated as a percentage of the index linked rent that Inclusion was obliged to pay under the main lease agreement): §17.

The representations pointed out, at §22, that the effect of Inclusion’s business model was to bring into the social housing sector new housing suitable for some of the most needy and vulnerable in society. The units owned and managed by Inclusion where new social housing units that would not otherwise be available in the sector. It was noted that providers operating according to the traditional model (under which the units were owned by the provider, normally subject to a mortgage) had struggled to provide new housing of this sort. This, it was said, was ‘not without risk, but the risks are modest, particularly when set against the substantial social benefit achieved’.

At §§30-32, the representations noted that Inclusion’s operating surplus had increased considerably since 2015/16, both in real terms and as a percentage of operating income, and was projected to increase still further in the coming years. At §41, reference was

At §61, the representations addressed possible future changes to housing benefit rules. It was accepted that this represented a ‘key risk’, but the risk was mitigated in a number of respects. First, all Inclusion’s housing was SSH. Under the Rent Standard Guidance, issued by the government under the Welfare Reform and Work Act 2016, SSH was exempt from the rent caps imposed in respect of other accommodation. As a result, the rent reductions applicable to other types of accommodation could not apply to Inclusion until April 2020 at the earliest. Secondly, in November 2017, the government had confirmed that the SSH (exempt rents) funding would remain in place beyond April 2020. Thirdly, Inclusion would have considerable warning in relation to any change to that state of affairs. This was because, given the importance of any policy change to the sector, the government would be legally obliged to provide substantial notice before implementing any such change and because in any event the occupancy agreements provided for annual rent reviews.

Under the heading ‘The Overarching Issue’, the representations said this:

At §84, the representations addressed what was termed the ‘Armageddon clause’ issue. This, it was explained, was a reference to the regulator’s suggestion that Inclusion should renegotiate its lease agreements so as to include a break clause entitling Inclusion to terminate its lease early if there were changes to the benefits system that made its model no longer financially viable. As to that, the following was said:

The regulator’s analysis of Inclusion’s representations is set out in a detailed tabular working paper created on 18 July 2018 and modified on 20 July 2018. A further internal meeting was scheduled for 14 August 2018. At that meeting, it was noted that Inclusion had provided a ‘more coherent explanation of its approach to risk management and is looking to improve and develop its governance arrangements and strengthen the board’. However, there were ‘some fundamental and material risks that remain’. The regulator did not consider that it could conclude its investigation yet. The representations were considered further during August 2018, as demonstrated by emails passing between members of the regulator’s staff. Meanwhile, Brenda Kirby, a former Chair of Inclusion’s Risk and Audit Committee, resigned and (on 16 August 2018) contacted the regulator raising concerns about Inclusion. There was contact between the regulator and another former board member, Jane Duncan, in October 2018. The regulator also decided, unusually, to arrange meetings with three organisations which had provided references for Inclusion. These meetings took place in October and November 2018. There was then a further meeting between Inclusion and the regulator on 19 November 2018.

The preparation of the final RED log took place in December 2018 and January 2019. This was considered at meetings on 10 and 15 January 2019. The regulator acknowledged that Inclusion had taken action to improve its governance arrangements following its GUR listing, but concluded that this action did not address the underlying issues identified by Inclusion’s own stress testing that, should the risk scenarios tested crystallise, Inclusion would need to reach agreement with third parties to amend existing contractual terms for leased properties in order to maintain its ongoing financial viability. It was noted that Neil Brown himself had accepted that there was no guarantee that the investment funds would agree to such amended terms. If no such agreement could be reached, the likely outcome would be insolvency. Inclusions plans for further growth were also considered. It was noted that this would be achieved by entering into more long-term leased properties with no break clauses and that this would increase the risk to the business. The conclusion was that Inclusion’s risk management and mitigation measures were not commensurate with its current and future risk profile. It was therefore non-compliant with §§2.2, 2.4 and 2.5 of the Standard and should be graded G3 in relation to governance.

As to financial viability, the regulator decided that Inclusion’s board had been unable to demonstrate that the risks to its viability, or the financial implications if risks crystallised, could be effectively managed or mitigated over the life of these contracts. The decision was taken to grade Inclusion as V3.

On 28 January 2019, the regulator wrote to Mr Ottowell to ‘confirm its judgement that Inclusion is non-compliant with both the governance and financial viability requirements of the Governance and Financial Viability standard’. It enclosed a draft RJ, invited Inclusion to comment on its factual accuracy and indicated its intention to publish the RJ on its website on 5 February 2019. It said that Inclusion would now be subject to a period of intensive regulatory engagement, which would last until the regulator was satisfied that Inclusion had returned to compliance.

On 31 January 2019, Inclusion wrote to the regulator noting its view that the draft RJ ‘doesn’t explain how it has reached its findings’. Inclusion asked the regulator to provide ‘full reasons and clear factual evidence to support its judgement so that we are able fully to understand and consider our position’. Whilst the offer of intensive regulatory engagement was ‘encouraging’, there was nothing in the very general concerns raised in the RJ which gave Inclusion any idea what types of step the regulator would be looking for Inclusion to take so as to become compliant. A factual error concerning Inclusion’s growth forecasts was also pointed out.

On 1 February 2019, the regulator decided to undertake an internal review of the RJ and wrote to Inclusion in these terms:

This was completed on 8 February 2019. It concluded that the decision-making process had followed the regulator’s agreed procedures. Inclusion attempted to stave off publication of the RJ by preparing and communicating to the regulator a proposed voluntary undertaking demonstrating how it would enhance compliance. On 12 February 2019, the regulator sent Inclusion a copy of the finalised RJ, which it said it would be publishing the next day. After some correspondence between solicitors, the final RJ was published by the regulator on its website on 15 February 2019.

I can deal with the arguments on delay and detriment to good administration relatively briefly.

CPR r. 54.5(1) requires that a claim for judicial review be filed (a) promptly and (b) in any event not later than 3 months after the grounds to make the claim first arose.

Ordinarily, the grant of permission precludes a defendant from contending that the claim should be dismissed on the ground of delay (although it does not preclude a submission that relief should be refused under s. 31(6) of the 1981 Act): R v Criminal Injuries Compensation Board ex p. A [1999] 2 AC 330, 341 (Lord Slynn). Here, however, Sir

In a case where the single judge considering permission on the papers wishes to leave the question of delay open to the substantive hearing, I would respectfully suggest that the better course is to say so expressly and then to order a ‘rolled-up’ hearing, rather than to

Ms Carss-Frisk accepts that, following R (Burkett) v Hammersmith and Fulham LBC [2002] 1 WLR 1593, the grounds to make a claim arise on the date when a decision having legal effect is made. She says that the internal decisions logs show that the decision here was taken on 15 January 2019 by Jonathan Walters, Deputy Director of Strategy and Performance, under the regulator’s scheme of delegation. She submits that this was the date on which time started to run, even though the decision was not communicated until 28 January 2019. In any event, she notes that these proceedings were not even brought within three months of that date. The internal review starting on 1 February 2019, Ms Carss-Frisk submits, was carried out for quality assurance purposes and did not involve the re-taking of the decision.

There has been a debate in the literature about whether the date on which ‘grounds to make the claim first arose’ is the date on which the decision was made or the date on which it is communicated. The Court of Appeal held that it was the former in R v Department of Transport ex p. Presvac Engineering Ltd (1992) 4 Admin LR 121. It has been argued that this may fall to be modified in the light of Lord Steyn’s principle in R (Anufrijeva) v Secretary of State for the Home Department [2004] 1 AC 604, [26], that an administrative decision does not have the character of a legal determination until it has been notified to the person it concerns. It is therefore suggested, applying the principles in Burkett, that it is only upon communication that time starts to run: see Auburn, Moffett & Sharland, Judicial Review: Principles and Procedure, §26.35. It is not necessary or appropriate to enter into this debate here, because I am satisfied that, even if there was a legally effective decision on 15 January 2019, there was on the facts a further decision not to alter it following the internal review.

A decision-maker is under no obligation to reconsider a final decision once it has been communicated. If, having received Inclusion’s letter of 31 January 2019, the regulator had written back to say ‘We have made our decision and will not be reconsidering it’, time would run from the date of the earlier decision. A claimant cannot in general start time running again by writing a letter asking the decision-maker to reconsider and then treating the refusal to reconsider as a new decision. But where a decision-maker, in response to a request to reconsider, chooses to conduct an internal review – and, as here, tells the requester that it is holding off publishing its final decision while it gives ‘serious consideration’ to the points made – the position is different. The matter can be tested by asking what would have happened if, having received the regulator’s email of 1 February 2019, Inclusion had issued a judicial review claim. The regulator would, surely, have been entitled to respond that the claim was premature because there was an internal review underway. The internal review could presumably have resulted in a different outcome. The internal review was, on the regulator’s evidence, not complete until 8 February 2019 and not communicated until 12 February 2019, when the regulator sent Inclusion the final version of the RJ. In these circumstances, the final RJ was a separate, challengeable decision; and time to challenge it began to run at the earliest on 8 February 2019. The claim form was filed within 3 months of that date.

I have considered the regulator’s submission that there was still ‘undue delay’ in waiting until close to the end of the 3-month period. The regulator relies in particular on the fact that the RJ required Inclusion to engage with the regulator. That being so, the regulator says that there was a particular onus on Inclusion to act promptly because, as must have been clear to it, any subsequent discussions would have to be based on the conclusions in the RJ. Such discussions as have taken place will have been wasted if relief is now granted.

In a case such as the present, it is obvious that it will be necessary for discussions to take place between the regulator and the registered provider after the publication of an RJ indicating non-compliance. These discussions will be likely to continue even if the provider is seeking to challenge the RJ in proceedings for judicial review. Unless and until a claim for judicial review succeeds, it will be necessary for the discussions to proceed on the basis of the conclusion set out in the RJ. That is what has happened here. In this case, the regulator has been aware since 31 January 2019 that Inclusion considered its reasons to be deficient. Inclusion’s letter before claim was sent on 1 April 2019. I do not accept that the discussions that were held before and after that date would have been materially different if Inclusion had intimated this claim a few weeks beforehand. This is not a case in which I would refuse relief pursuant to s. 31(6) of the 1981 Act, if relief were otherwise appropriate.

Mr Stilitz submitted that the substantive ‘reasons’ given in the RJ were in reality no more than a statement of the regulator’s conclusions. In the skeleton argument, he and Ms Slarks pointed out that there was no explanation of the regulator’s view on (i) whether and how Inclusion’s ‘worst case scenario’ was more concerning than that of a registered provider adhering to the ‘traditional model’ (in which the registered provider borrows money to purchase properties); (ii) whether and how Inclusion was at a higher risk of its worst case scenario eventuating than a ‘traditional model’ provider; or (iii) whether and why the regulator considered each of Inclusion’s mitigation strategies (NAs, large and increasing cash reserves, void insurance, void self-insurance) to be inadequate.

In oral submissions, Mr Stilitz elaborated on these criticisms. There was no reasoning as to why the levels of risk involved, when set against the protective measures put in place, were so imbalanced as to justify a non-compliant rating. No explanation was given as to why it was unacceptable to rely on renegotiation of existing agreements in a ‘disaster scenario’ and there was no proper appreciation of Inclusion’s case that in such a scenario there would be powerful incentives operating on counterparties to renegotiate. The RJ contained no criticism at all of Inclusion’s governing systems and structures and there is no attempt to analyse the improvements put in place since 2015. It was not explained why Inclusion’s ‘ambitious plans to expand its operating model’ were relevant to the assessment of risk. There was no attempt to engage in a detailed or careful way with the representations made by Inclusion in July 2018 or with information submitted since. Mr Stilitz said that these failures were of particular significance given that the result of the non-compliant rating was that Inclusion would have to work with the regulator to improve. It could only do that if it understood properly why the regulator had reached its conclusions.

In the course of his oral submissions, Mr Stilitz drew attention to a number of key points that he said were not addressed. First, Inclusion’s business model had the effect of bringing to the market new units of high-quality social housing that would otherwise not form part of the stock available. Second, while it was true that the business model depended upon the continued availability of housing benefit for tenants requiring SSH, it would be politically and legally difficult to withdraw support for this particular group of disabled and vulnerable individuals, or at least to do so in a rapid or radical way. Third, there was increasing demand for SSH and there was currently a nationwide shortage. Fourth, whilst it is true that Inclusion, as a subordinate landlord, took on a degree of commercial risk, it had adopted sophisticated means of mitigating that risk. Since 2015, the number of NAs and their coverage had increased. Fifth, it was not unrealistic to posit renegotiation of leases as a potential mitigation measure in the ‘disaster scenario’ in which several risks eventuate, because the investment funds would be powerfully incentivised to renegotiate given that the accommodation is largely adapted for the needs of disabled and vulnerable tenants. This meant that the funds would have limited alternatives to accepting modified terms.

Mr Stilitz invited me to be ‘highly sceptical’ of the ‘embellished reasoning’ provided by the regulator ex post facto in the course of proceedings. He relies on R v Westminster City council ex p. Ermakov [1996] 2 All ER 302 and R (Nash) v Chelsea College of Art and Design [2001] EWHC 538 (Admin).

There is no dispute that the regulator was obliged to give reasons in the RJ. Section 92K(5) of the 2008 Act requires it to act in a way that is ‘transparent and accountable’.

The question whether reasons are adequate is context-specific. In South Buckinghamshire District Council v Porter (No. 2) [2004] 1 WLR 1953 (a planning case), Lord Brown said this at [36]:

So far as ex post facto reasons are concerned, the authorities draw a distinction between evidence elucidating those originally given and evidence contradicting the reasons originally given or providing wholly new reasons: Ermakov, pp. 325-6. Evidence of the former kind may be admissible; evidence of the latter kind is generally not. Furthermore, reasons proffered after the commencement of proceedings must be treated especially carefully, because there is a natural tendency to seek to defend and bolster a decision that is under challenge: Nash, [34(e)]. The evidence contained in the regulator’s witness statements is certainly not inconsistent with those given in the RJ. I regard it for the most part as elucidatory. In any event, the need for caution that applies when considering ex post facto reasons does not apply to the reasons contained in the RED logs or other records of meetings prior to the decision under challenge. They are a contemporaneous record of the regulator’s reasons and may, in my judgment, properly be taken into account to the extent that they are not inconsistent with what was said in the RJ.

Even focussing purely on the reasons given in the RJ, it is a key part of the context which informs the extent of the duty to give reasons in this case that the RJ was the culmination of a long process of engagement between Inclusion and the regulator. During the process, the regulator had outlined its concerns in meetings, letters, emails and telephone calls. Inclusion had been able to submit detailed representations. Aside from the reasons challenge, there was no suggestion that the process was unfair. Those challenging regulatory decisions often complain of a breach of natural justice on the basis that the regulator has failed to identify its concerns sufficiently clearly to enable meaningful representations to be made. No such complaint was made here. The fact that there had been a fair process of engagement prior to the RJ does not absolve the regulator of its duty to give properly intelligible reasons. But it does form part of the backdrop against which the intelligibility of the reasons ultimately given are to be judged. As in the South Buckinghamshire case, the reasons here were addressed to a party (Inclusion) that must be taken to be well aware of the issues involved.

With that in mind, I turn to the reasons given in the RJ. They may be paraphrased as follows (the paragraph numbers are those inserted by me in the excerpts cited at [61] above):

It is true that these reasons do not engage in detail with the risk mitigation measures described in Inclusion’s representations. They do, however, explain in broad terms why the regulator considered those mitigation measures to be insufficient to demonstrate compliance. Inclusion’s business model involves entering into long-term, index-linked commitments without break clauses. This was a feature which Campbell Tickell had identified in 2015 as problematic and which its David Williams had described at Inclusion’s board meeting on 2 July 2015 as a ‘flawed’ model. Inclusion had responded by trying to renegotiate the leases to insert break clauses, as Ward Hadaway had also recommended. This had not been successful. Against that, the level of protection for Inclusion’s long-term costs (which was principally through NAs with care providers and local authorities) was not sufficient to satisfy the regulator that the risks were adequately mitigated. (It was well known to both the regulator and Inclusion that, of the new units taken on since 2015, the majority had NAs with terms of 10 years or less and many had NAs with terms of 5 years or less.) If the risks did crystallise (for example because government policy on housing benefit changed), Inclusion would have to renegotiate its leases. This would be dependent on the agreement of third-party commercial entities and so could not be guaranteed. (Mr Stilitz criticises the word ‘goodwill’ because it does not reflect the fact that counterparties would be incentivised to renegotiate, but that reads too much into the language used. In context, the regulator was saying no more than that investment funds which had so far refused Inclusion’s invitation to renegotiate terms could not be guaranteed to do so in a ‘disaster scenario’; they would be likely to do whatever they thought was in their commercial interest.) If the agreement of these counterparties was not forthcoming, Inclusion’s tenants would be exposed. Inclusion’s growth aspirations (which involve entering into more long-term leases) compound the problem. (Inclusion had made clear that expansion had increased its cash reserves both in raw terms and as a percentage of turnover, but it had also very substantially increased the number of leases on 20-25 year terms, with no break clauses, thereby substantially increasing its overall exposure.)

Some of these findings are criticised by Mr Stilitz as irrational. I shall consider these criticisms under ground 2. But the fact that they can be made at all seems to me to undermine the suggestion that the reasons given left Inclusion in the dark about why it had been graded non-compliant. Insofar as complaint is made that the regulator failed to explain why Inclusion’s ‘worst case scenario’ was more concerning than that of a provider operating on the ‘traditional model’, the regulator was not obliged to categorise providers as operating on the ‘traditional’ or ‘rent’ models. This dichotomy is part of Inclusion’s own description of the way it operates, but the regulator does not, and does not have to, categorise providers in this way. It takes the view that there are many different kinds of business models and considers each one on its own terms. In any event, the regulator’s duty was to explain adequately why it reached the conclusions it did about Inclusion, not to explain why its business model was more concerning or risky than that of other providers. Against the background of engagement between the regulator and Inclusion over the course of the IDA, the reasons given were in my judgment intelligible and adequate. They enabled Inclusion to understand why the conclusion expressed in the RJ had been reached. Ground 1 is not made out.

The foregoing analysis has concentrated on the reasons given in the RJ itself, understood by reference to the process of engagement between Inclusion and the regulator over the course of the IDA. Because I have found that the reasons actually given were adequate, it is not necessary to consider the submission that, even if they had not been, s. 31(2A) of the 1981 Act would preclude the grant of relief, given that adequate reasons have now been supplied in response to these proceedings.

Mr Stilitz criticises the regulator for reaching conclusions that were ‘fundamentally illogical’ and for failing to take into account relevant considerations. He relies in addition on Matadeen v Pointu [1999] 1 AC 98 for the proposition that ‘treating like cases alike is a general axiom of rational behaviour’ and says that the regulator acted unlawfully while not awarding non-compliant ratings to other providers whose business models were materially more risky.

As to risk, Mr Stilitz identifies the ‘crux of [the regulator’s] objection to Inclusion’s business model’ as follows: ‘in the absolute worst case scenario – if everything or almost everything that could go wrong for an RP did go wrong – Inclusion would seek to renegotiate its contracts with third parties and, if that failed, enter into insolvency arrangements’. In concluding that Inclusion’s strategy in this scenario was reliant on third party consent, Mr Stilitz says that the regulator failed to consider Inclusion’s primary mitigation strategies, namely (i) specialising in well-resourced SSH, (ii) increasing diversification of properties and partners, (iii) risk-sharing arrangements with care bodies, funds, developers and care providers, (iv) insurance, (v) maintaining a substantial buffer by way of cash reserves and (vi) maintaining rigorous and focussed governance processes.

Mr Stilitz submits that the regulator did not reject Inclusion’s evidence that any RP – whatever its business model – might in principle find itself in the disaster scenario and would then be dependent on renegotiating its financial commitments; that in any event the disaster scenario is unlikely to occur given the mitigation strategies; and that even in the disaster scenario, Inclusion’s cash reserves would enable it to continue to trade for at least a year, and maybe as much as two, while it wound down its business. If any risk of insolvency was enough to render a provider non-compliant, no provider could ever be compliant. The RJ discloses that no assessment was undertaken of the likelihood of the ‘worst case scenario’ eventuating.

As to governance, Mr Stilitz says that the RJ made no criticism of Inclusion’s governance processes, but jumps straight from the conclusion on risk to the conclusion that there must have been a flaw in Inclusion’s governance. As to financial viability, he points out that there was no criticism of Inclusion’s reserves and that the regulator has apparently jumped illogically from its conclusion about risk to the conclusion that Inclusion was non-compliant for financial viability. As to growth, Mr Stilitz notes that the regulator did not even comment on Inclusion’s evidence that its growth had decreased risk across the business by increasing the levels of its surplus both in real and proportionate terms and by diversifying its portfolio. The regulator made no findings that Inclusion’s growth to date or projected growth were over-ambitious. Instead, it adopted the position that any risk is ‘seemingly by definition’ compounded by the fact that Inclusion intends to grow. The regulator’s anti-growth stance amounts to a breach of the duty imposed by s. 108 of the Deregulation Act 2015 (‘the 2015 Act’) and an unjustified and unexplained departure from the pro-growth policy outlined in Regulating the Standards.

It is well established that, when entertaining a rationality challenge by way of judicial review to a decision that involves a judgment, the court is not the primary decisionmaker. This point has been given special emphasis in challenges to specialist regulators: see e.g. R v Director General of Telecommunications [1999] ECC 314, [26] (Lightman J); Fraser v NICE [2009] EWHC 452 (Admin), [47]-[48] (Simon J). However, it is also important not to erect so unrealistically high a hurdle as to render success in an irrationality challenge effectively impossible. As Sedley J said in R v Parliamentary Commissioner for Administration ex p. Balchin [1996] EWHC 152 (Admin):

Was there that sort of error of reasoning here? The regulator focussed on what might happen in the scenario where certain risks eventuated. But it did not describe this as an

There were other risks too. One was the risk that one or more care provider counterparties would default. The regulator pointed out that one such counterparty was Lifeways. Inclusion had itself identified the risk of Lifeways failing as a matter that had to be addressed in its contingency planning. Mr Stilitz points out that Inclusion had set out a detailed plan to reduce and mitigate this risk. The fact remained, however, that – as at October 2019, when Inclusion filed its evidence in these proceedings – Lifeways was the care provider for around 50% of its properties. The regulator was, in my judgment, entitled to take this into account, alongside the risk of a change in government policy.

As to mitigation strategies, the regulator understood that Inclusion specialised in SSH, but it was entitled to regard the prospect of a change in housing benefit or rent regulation policy, even one affecting those requiring SSH, as realistic. It understood that Inclusion was trying to diversify its range of partners (though its exposure to Lifeways remained substantial). But this diversification involved entering into more long-term leases, with no break clauses, thereby substantially increasing Inclusion’s liabilities. As to the risk sharing arrangements, the principal vehicles were the NAs. Mr Brown’s third witness statement for Inclusion establishes that, of the 1,870 units taken into management since the end of the calendar year 2015, every one is covered by a NA. However, it also makes clear that only 370 of these are covered by a NA of the same length as the lease (20 years), with void cover for that entire period. A further 617 are covered by void agreements of 10 years or more. That, of course, leaves just under half which benefit from lesser or no protection. The regulator was also aware of the fact that Inclusion had void insurance, though this covered only 17% of properties. The substantial cash reserves had the potential to create a buffer. But the purpose of this buffer was to provide time for renegotiation of leases. Ultimately, if one of the foreseeable risks (such as a change in housing benefit or rent regulation policy) eventuated, Inclusion would be stuck with its obligations to pay rent to the head landlord, with index-linked annual increases, under its long leases; the tenants (or many of them) would have no way of paying the rent; and the cash reserves would buy time for between 1 and 2 years. During that period, Inclusion would have to seek to renegotiate the leases, which would require the agreement of the commercial counterparties. As noted above, the use of the word ‘goodwill’ did not reflect any misunderstanding of the commercial incentives that might operate on counterparties during any such negotiation. It was merely a way of saying that all would depend on what those counterparties considered to be in their commercial interests (which could not be predicted).

I do not read the RJ, or the regulator’s evidence in these proceedings, as saying that any risk of insolvency would be enough to render a provider non-compliant. The regulator had to make a judgment about the likelihood of the risks eventuating and about the adequacy of the plan if they did eventuate. It judged that the likelihood was significant and that the plan was inadequate. The regulator’s view was that, if one of the adverse scenarios eventuated, the range of options open to a provider such as Inclusion were considerably more limited than the range of options that would be open to a provider which owned properties subject to loans. A provider of the latter kind would be able to respond to changing market conditions by switching funding structures or, ultimately, by raising capital through selling a portion of its assets. Insofar as the regulator reached a view that a model which involved the provider holding no assets but incurring increasing numbers of long, index-linked liabilities with no break clauses was inherently riskier, it was in my judgment entitled to reach that view.

As to the conclusion on financial viability, the regulator was aware that Inclusion’s reserves had increased, but it properly took into account the fact that Inclusion’s liabilities had also been increasing substantially. Inclusion’s total exposure to contractual lease payments was some £431m as at March 2019, having increased by 30% since the previous year. That was plainly relevant to Inclusion’s compliance with the financial viability standard. The conclusion that Inclusion was non-compliant with the financial viability standard involved no error of logic or approach. It resulted from a judgment, made by a specialist regulator, which was properly open to it on the evidence it had.

As to growth, it is true that Inclusion had demonstrated that its surpluses had grown (both in raw and in percentage terms) with its business. But, as noted above, so had its liabilities. Inclusion’s plan was to continue to enter into long-term leases with no break clauses in respect of some 350 new units per year. If it was rational to conclude that these liabilities gave rise to a substantial risk, not adequately addressed by mitigation measures (as I have concluded it was), it was rational to conclude that adding more units on precisely the same terms would ‘compound these exposures’. On the evidence, there was no illogical assumption that growth is inherently likely to increase risks. The conclusion was that growth on the same model already identified as risky would increase risk. In my judgment, that conclusion involved no error of logic or law.

Section 108 of the Deregulation Act 2015 requires regulators to whom the duty is applied to have regard to the desirability of promoting economic growth and, in that regard, to consider the importance of exercising their regulatory functions in a way which ensures that (a) regulatory action is taken only when it is needed, and (b) any action taken is proportionate. In her evidence for these proceedings, Ms MacGregor said that the duty in s. 108 of the 2015 Act did not apply to the regulator, because although its predecessor had been specified as an entity to which the duty applied (by virtue of the Economic Growth (Regulatory Functions) Order 2017 (‘the 2017 Order’: SI 2017/267)), the regulator had not been so specified upon its establishment in 2018. But the 2017 Order applies the duty to ‘the regulatory functions specified in the schedule’ (see art. 2(1)), which specifies ‘all regulatory functions exercisable by’ the listed entities, including the HCA. So, on a natural reading, the duty attaches to the function, not the entity, and it therefore applies to the functions now exercisable by the regulator insofar as they were at the time of the 2017 Order exercisable by the HCA.

Interesting though this is, in my judgment, it does not matter. Although there is no reference in the RJ to the desirability of promoting economic growth (something which also appears in general terms in Regulating the Standards), the main purpose of that duty is, as indicated by s. 108(2), to remind regulators that excessive regulation can stifle economic growth and to require them to bear in mind the need to ensure that regulatory action is limited to the minimum necessary. The regulator plainly thought that the regulatory action taken here was necessary in the public interest. It has explained its reasons for that view in its extensive evidence. Section 108 was not intended to preclude regulatory action where the regulator considered it necessary in the public interest. If the failure specifically to mention the duty imposed by s. 108 was an error of law, it was not the kind of error which could be regarded as material. To put it in terms of s. 31(2A) of the 1981 Act, in the light of all the evidence before me it is not only highly likely but practically certain that, even if the duty had been specifically drawn to the regulator’s attention, the result would have been the same.

As to governance, the regulator accepts that its conclusion was not premised on any identified criticism of Inclusion’s governance arrangements or structures. Initially, I could see some force in the submission that the ‘governance’ part of the Standard was concerned with process and structure rather than outcome. However, §1.1 of the Standard provides that governance arrangements ‘shall ensure’ that registered providers do certain things, including ‘have an effective risk management and internal control framework’. The regulator is in my judgment entitled to conclude that a provider whose business model is judged to involve too much risk has failed to ensure an effective risk management framework, particularly where, as here, the flaws were identified very clearly in 2015 and, in the regulator’s view, have not been sufficiently addressed since.

As to the allegation of unequal treatment between Inclusion and other providers graded as compliant (about which Mr Stilitz said very little in oral argument), the Supreme Court has made clear in R (Gallaher Group plc) v Competition and Markets Authority [2019] AC 96 that ‘unequal treatment’ is a ground for review if and only if it involves drawing irrational distinctions. In a context such as this, where each judgment is multi-factorial and fact-specific, a comparison between outcomes in different cases will rarely be an auspicious basis for a rationality challenge. In this case, the regulator has attached the RJs for the comparator cases and has briefly explained the key differences. It is plain that the circumstances of the purported comparators are materially dissimilar. There is no basis for the suggestion that the distinctions drawn were irrational.

Ground 2 is, therefore, not made out.

Under this ground, Mr Stilitz relies on §10 of the Code, which is set out at [16] above. He says that the regulator’s finding that Inclusion was non-compliant ‘on the basis of an extremely unlikely, managed risk’ and that this finding constituted an unreasoned departure from §10 of the Code.

This point adds nothing to ground 2. The regulator did not find that the risks here were ‘extremely unlikely’. On the contrary, it described the risks with which it was concerned as ‘reasonable risks associated with economic and policy cycles’. For the reasons I have given, that conclusion was rationally open to it, as was the conclusion that Inclusion was not managing the risks effectively. There was nothing in the RJ that could properly be characterised as a departure from the general terms of §10 of the Code.

The complaint here is that, even on the regulator’s findings, Inclusion met the criteria for grading as V2, rather than V3 in Regulating the Standards (see the summary and excerpts set out in [12]-[13] above).

This point also adds nothing to Ground 2. §4.9 of Regulating the Standards lists some of the characteristics that V2 providers ‘can often share’. But the decision whether to grade a particular provider at V2 or V3 ultimately depends on the exercise by the regulator of a judgment. The judgment is necessarily qualitative. It depends on whether the regulator’s ‘concerns’, which will be present in both V2 and V3 cases, can properly be described as ‘issues of serious regulatory concern’. The question in each case is whether, on the evidence, there is a proper, logical and lawful basis for the qualitative judgment reached. If (as I have concluded) there is, the illustrative indicia in §4.9 are unlikely to provide an independent basis for impugning the regulator’s judgment. They do not do so here.

Under this ground, Mr Stilitz relied on s. 92K(5) of the 2008 Act. In his skeleton argument, his submission as to why the duty imposed by that provision was breached was commendably succinct:

At the hearing, I raised an issue concerning the proper approach of the court to assessing compliance with s. 92K(5). On its face it appears to impose on the regulator a duty to exercise its functions in a way that (so far as possible) ‘is’ (rather than, for example, ‘is, in its view’) proportionate. In some contexts where the proportionality of action by a public authority is in issue, the authorities require that the court determines the question of proportionality for itself (giving whatever weight is appropriate in the circumstances to the view of the public authority): see e.g. Lumsdon v Legal Services Board [2015] AC 697 in the context of the proportionality of derogations from the fundamental freedoms guaranteed by EU law and R (SB) v Governors of Denbigh High School [2007] 1 AC 100 in the context of the proportionality of interferences with Convention rights. At my invitation, the parties filed brief, but helpful, supplemental written submissions on the question whether the same approach was required in this case.

For his part, Mr Stilitz submitted that the language of s. 92K(5) indicated that Parliament had indeed intended the court to decide for itself whether a particular exercise of functions was ‘(so far as possible) proportionate’. He relied upon the approach in Bank Mellat v HM Treasury (No. 2) [2014] AC 700, which he said concerned a similar statutory obligation to act in a way that was proportionate. Moreover, decisions to which the obligation in s. 92K(5) applies, like the decisions at issue in Bank Mellat, were likely to engage property rights under Article 1 of Protocol 1 to the ECHR (‘A1P1’).

Ms Carss-Frisk said that it was not appropriate to import the approach applicable in EU and Convention cases to a statutory obligation which (i) applies to every function of the regulator’s; (ii) is qualified by the words ‘so far as possible’ and (iii) involves no fundamental rights or EU rights. The proper approach is to apply a rationality standard of review, with the precise intensity of review varying with context, in accordance with Kennedy v Charity Commission [2015] AC 455.

My conclusions on the proper approach are as follows:

In the light of the analysis above, I start by noting (for the reasons I have given under ground 2) that the regulator acted rationally in awarding Inclusion gradings of G3 and V3. The regulator’s analysis of the proportionality of these judgments can be seen from a note of the internal meeting on 15 January 2019:

Although not determinative, the regulator’s conclusion that its actions were proportionate – based as it was on a rational and lawful analysis of its ‘serious concerns’ – is entitled to considerable weight. One matter not expressly considered by the regulator, but in my judgment plainly relevant to the overall assessment of proportionality, was the effect of the G3/V3 grading. In awarding such a grading, the regulator was not stopping Inclusion from trading. It was merely saying – publicly – that there were ‘serious issues of regulatory concern’ and that, ‘in agreement with [the regulator], the provider is working to improve its position’. I accept that this is likely to have, and in this case has had, a practical effect on Inclusion’s relations with certain third parties, but a G3/V3 grading is not a ‘striking off’. It is intended to encourage changes that will lead to compliance. Looking at the matter as a whole, I consider that in awarding the grading in this case, the regulator exercised its functions in a way that was (so far as possible) proportionate. Ground 5 is not, therefore, made out.

During the hearing, in the light of Mr Stilitz’s emphasis on the social benefits of Inclusion’s model, I asked whether it was part of Inclusion’s case that the regulator had failed to comply with its duty under s. 92K to perform its functions with a view to achieving so far as possible the economic regulation objective, which by s. 92K(2)(b) includes supporting ‘the provision of social housing sufficient to meet reasonable demands (including by encouraging and promoting private investment in social housing)’. Mr Stilitz said that this was part of his case, relying on §§10, 15 and 92 of his Statement of Facts and Grounds. Ms Carss-Frisk objected to this point being raised, saying that on a proper analysis of those paragraphs, the point was not pleaded and noting that, if it had, been it could and would have been addressed in evidence.

I have considered the paragraphs in the Statement of Facts and Grounds identified by Mr Stilitz. They advert to the fact that Inclusion’s model enables it to provide social housing which would not otherwise be available and point out that this constitutes a social benefit. They do not, however, include any pleaded allegation that the regulator erred by failing to give proper consideration to this fact, contrary to its duty under s. 92K. An allegation that the regulator had left out of account a statutorily mandatory consideration would, in my view, have to be pleaded. This is not just arid procedural pedantry; it is a matter of fairness. Here, the regulator has answered the pleaded allegations fully. This point was not among them. It would not be fair to entertain a ground of challenge not advanced at any stage before the hearing (including in the skeleton argument) to which responsive evidence could in principle have been relevant: see R (Talpada) v Secretary of State for the Home Department [2018] EWCA Civ 841, at [69] (Singh LJ). To the extent that Mr Stilitz’s submission implicitly included an application to amend his Statement of Facts and Grounds, I refuse it.