Birmingham City Council v SS & Ors

These two conjoined appeals about entitlement to housing benefit are concerned with the power of a housing authority to determine that the net eligible rent payable in respect of accommodation, and thus the amount of housing benefit payable in respect of that accommodation, be restricted. The appellant, Birmingham City Council, as housing authority, contends that the Upper Tribunal (Administrative Appeals Chamber) adopted an incorrect approach under regulation 13(3) of the Housing Benefit Regulations 2006 (S.I. 2006/213) (“the 2006 regulations”), in the form substituted by paragraph 5 of Schedule 3 to the Housing Benefit and Council Tax Benefit (Consequential Provisions) Regulations 2006 (S.I. 2006/217) (“the consequential provisions regulations 2006”), when considering, in the case of claimants in “exempt” accommodation, whether their rent was “unreasonably high by comparison with the rent payable in respect of suitable alternative accommodation elsewhere”. In particular, the city council submits that the Upper Tribunal erred in effectively applying, as criteria of comparability, the availability and amount of public funding which would have enabled the landlords in question to reduce the rents they charged.

With permission granted by Sales L.J., the city council appeals against two decisions of the Upper Tribunal (Upper Tribunal Judge Mark): an interim decision dated 30 August 2013 and a final decision dated 11 March 2014. By those decisions the Upper Tribunal allowed the appeals of the first respondents, SA and SS, from the decision of the First-tier Tribunal (Social Entitlement Chamber), dated 20 October 2011 – for which reasons were given on 23 February 2012. The First-tier Tribunal had dismissed their appeals against the city council’s decision under regulation 13(3)(b), as reconsidered on 24 August 2010, to restrict the eligible rent payable for the exempt accommodation provided to them by the intervener, a registered charity called Roshni, in a refuge for women who had been the victims of domestic violence. The second respondent is the Secretary of State for Work and Pensions.

SA and SS left Roshni’s refuge long ago. Neither of them has played any part in the appeals to this court. Nor has the Secretary of State, beyond indicating support for the appeals, and – in the Government Legal Department’s letter to the court dated 26 May 2016 – that it adopts the argument put forward by the city council. Roshni has intervened. In seeking permission to do so, it confirmed that it did not seek to be heard “on the facts or on the merits of the individual cases”, but only “on the issue of principle raised by the appeal …”. In view of the decision of this court in Wirral Metropolitan Borough Council v Salisbury Independent Living Ltd. [2012] EWCA Civ 84, it did not apply to be joined as a respondent in place of, or in addition to, SA and SS (paragraphs 13 and 17 of its application for permission to intervene dated 30 April 2015). By an order dated 17 March 2016, Beatson L.J. gave it permission to make oral submissions at the hearing of the appeals, which it has done. We were assured by Mr Jonathan Manning, who appeared for the city council, and by Mr Stephen Knafler Q.C., who appeared for Roshni, that the appeals are not academic. If the city council fails to establish that it was entitled to restrict the eligible rents in this case it will have to pay Roshni – which received the payments of housing benefit directly – the full rent claimed by SA and SS up to the date on which they left Roshni’s refuge.

Roshni charged SA and SS each a contractual rent of £257.87 per week, of which £242.17 was, in principle, eligible for housing benefit. Under regulation 13(3) the city council restricted the housing benefit payable in respect of these rents to £179.20 per week in each case, and the First-tier Tribunal upheld those decisions. In its interim decision on the subsequent appeals the Upper Tribunal allowed those appeals, set aside the First-tier Tribunal’s decision, and gave the parties an opportunity to lodge further evidence and submissions before coming to the decisions it would substitute for that of the First-tier Tribunal. In its final decision it determined that SA’s and SS’s “net eligible rents” were in each case £242.17 per week.

When granting permission to appeal, Sales L.J. observed that the appeal raises this issue: whether, under regulation 13(3)(b) it was “appropriate to take into account the degree of public subsidy received by the owner of the accommodation occupied and by the owner of any other accommodation said to be a suitable alternative; and in particular, how the [2006 regulations] should be applied for the purposes of comparison in relation to accommodation owned by a charity, by a registered social landlord and by a private landlord”. A “registered social landlord” would include, for example, a housing association – a non-profit making organisation providing affordable homes for rent or purchase to those in housing need. It should be noted that, in England, registered social landlords are now known as “private registered providers of social housing” (see section 80 of the Housing and Regeneration Act 2008).

Mr Manning acknowledged that the four grounds of appeal in the appellant’s notice could be refined to two main questions. First, did the Upper Tribunal fail to follow the approach to the comparison of rents required by regulation 13(3), adopting instead an approach that involved an investigation of the income and costs of the individual landlords it had found to be the appropriate comparators (grounds 1 and 3 in the appellant’s notice)? And second, is support for the Upper Tribunal’s approach to be found in the judgment of Collins J. in R. v Coventry City Council, ex parte Morgan (also known as R. v Coventry City Council, ex parte Waite), 7 July 1995 Queen’s Bench Division (Crown Office List) (unreported)? The city council maintains that that case was wrongly decided because the court excluded certain parts of the residential rental sector from the comparison of rents under regulation 13(3) on the ground that this exercise must be undertaken on a “like with like” basis (grounds 2 and 4).

Housing benefit is “a means tested benefit provided under section 130 of the Social Security Contributions and Benefits Act 1992 and subordinate regulations”. Its “purpose is to help claimants with their rental costs”. There is “a prescribed mechanism for determining in each case the appropriate maximum housing benefit” (see the judgment of Lord Toulson in R. (on the application of Carmichael) v Secretary of State for Work and Pensions [2016] UKSC 58, at paragraph 5). As I have said, this appeal concerns regulation 13 in the form inserted into the 2006 regulations, by way of substitution, by the consequential provisions regulations 2006 (Schedule 3, paragraph 5) so as to make separate provision for the restriction of benefit for occupiers of “exempt” accommodation (supported housing) to whom the normal local housing allowance and “maximum rent” rules do not apply. Accordingly, references to regulation 13 (or to any of its paragraphs) in this judgment are to the substituted form of the regulation. The same applies to regulation 12.

In R. v Housing Benefit Review Board for East Devon District Council, ex parte Gibson (1993) 25 H.L.R. 487, a case concerning the restrictions on “unreasonable payments” in regulation 11 of the Housing Benefit (General) Regulations 1987 (S.I. 1987/1971) (“the 1987 regulations”), prior to its amendment by the Housing Benefit (General) Amendment Regulations 1991 (S.I. 1991/235) (“the 1991 regulations”), Sir Thomas Bingham M.R., as he then was, described the general effect of the statutory scheme in this way (at p.490):

“The effect of the scheme is that a person may be entitled to housing benefit if he or she is liable to make payments in respect of a dwelling in Great Britain which he or she occupies as a home. If the applicant has no income, or if the applicant’s income does not exceed a certain level, then there will be an entitlement to an amount up to a maximum of housing benefit in that particular case. The housing benefit may arise in any one week to 100 per cent. of the recipient’s eligible rent as defined, calculated on a weekly basis, less deductions. In general terms, a person’s eligible rent is what the applicant is liable to pay.”

He continued:

“It is evident that the scheme is directed to two possible abuses. The first is that a recipient of housing benefit may live in an unnecessarily luxurious and therefore expensive dwelling house. The second is that a landlord may charge more than the market rate to a recipient of housing benefit, both those abuses stemming from the fact that the prudence which activates the ordinary citizen may cease to do so if there is a belief that public funds will pay whatever the recipient of housing benefit may choose to require.

It is with those possible abuses in mind that the relevant regulations with which this appeal is concerned have to be read … .”

In R. (on the application of Mehanne) v City of Westminster Housing Benefit Review Board [2001] 1 W.L.R. 539, Lord Bingham of Cornhill said this about the statutory regime for housing benefit (in paragraph 5 of his speech, at p.543A-B):

“… First, it is directed to the humane objective of assisting those of modest means to provide themselves with a roof over their heads. This is, after all, one of the most basic of human needs, and it is not surprisingly accepted as a proper object of public expenditure. But, secondly, such expenditure must be directed to meeting real needs. Thus expenditure may be restricted if a claimant is housed more expensively than necessary, whether because his accommodation is unnecessarily large, or because he is paying more than the market rate for the area in which he lives, or because he could be housed adequately but more economically in similar accommodation in an accessible but less expensive quarter. Special consideration is, however, given to those who are elderly, or unable to work, or who are responsible for a child or young person living with them.”

Paragraph 4(1)(b)(i) of Schedule 3, “Transitional and Savings Provisions”, to the consequential provisions regulations provided that the “eligible rent” of a person “who is liable to make payments in respect of a dwelling occupied by him as his home, which is exempt accommodation” was to be determined in accordance with “regulations 12 (rent) and 13 (maximum rent) of the Housing Benefit Regulations …”. Paragraph 4(10) of Schedule 3 to the consequential provisions regulations defined “exempt accommodation” as including accommodation “… (b) provided by … a housing association, a registered charity or voluntary organisation where that body or a person acting on its behalf also provides the claimant with care, support or supervision”. The accommodation with which we are concerned in this appeal was “exempt accommodation”.

Regulations 12 and 13 of the 2006 regulations preserved the use of the previous method for ascertaining the eligible rent in awards of housing benefit. The previous rules were created in response to concerns that determinations made by rent officers discharging their functions under the Housing Act 1996 might be harmful to those in various forms of supported housing. They retained the previous system of benefit calculation in which local authorities decided restrictions on the eligible rent of those in accommodation provided by housing associations, registered charities, voluntary organisations or authorities, where the landlord is also providing care, support or supervision. When the accommodation is unreasonably large for the claimant and his or her household, or the rent unreasonably high, any restriction on the eligible rent results from a comparison of the rent payable for the claimant’s accommodation with rents payable for suitable alternative accommodation elsewhere. Where the claimant is elderly, incapable of work or has children, his or her eligible rent may be restricted only if suitable alternative accommodation is actually available and it is reasonable to expect him or her to move.

Regulation 13 of the 2006 regulations, so far as is relevant here it provides:

“13.

Restrictions on unreasonable payments

…

In ex parte Gibson the Court of Appeal was concerned in particular with regulation 11(2) and (3) of the 1987 regulations. Regulation 11(2)(c) of the 1987 regulations differed from regulation 13(3)(b) of the 2006 regulations in that it conferred on the authority a discretion to decide whether a claimant’s eligible rent should be reduced, rather than making the reduction itself mandatory and conferring a discretion merely as to the amount. Regulation 11(3) was in similar terms to regulation 13(4) of the 2006 regulations. Sir Thomas Bingham M.R. said this about regulation 11(2) and (3) (on pp.493 and 494):

“It is, as I have already suggested, plain that the procedure is designed to protect the public purse. But it is fair, I think, to infer that the procedure is not designed to produce homelessness, which would be the result if a beneficiary’s rent were restricted, so that he could not afford to stay where he was but was unable to find any other accommodation to which he could be expected to move at the level of rent payable.

The effect of the regulations, therefore, under regulation 11(2)(c) is to confer a discretion as to whether the rent should be restricted but, in a case which falls within 11(3), to deny the local authority or the Review Board a discretion to restrict the rent, unless it is in effect satisfied that there is alternative accommodation suitable for the reasonable needs of the beneficiary and his family which is available on the market at the level of the rent allowed. Regulation 11(3) is clearly designed to make special provision for families which include elderly persons of the advanced age of 60, disabled persons and children and young persons. It is, I think, implicit that the restriction of the rent is recognised as being possibly such as to oblige a recipient of housing benefit to move, but certainly in an 11(3) case such a recipient is not to be obliged to do so unless it appears that there is somewhere that he and his family can move at the rent allowed.”

He accepted two submissions made on behalf of the Secretary of State:

“… First of all, [counsel for the Secretary of State] … has emphasised the Secretary of State’s anxiety that the scheme should be efficaciously operated so as to ensure that public funds are not wastefully dissipated. The court recognises that as an entirely proper interest and recognises that, in any decision that it may give, it must be very careful not to interpret this subordinate legislation in a way which would frustrate the intentions of the legislature or the Secretary of State’s intention to which I have referred.

Secondly, it has repeatedly said that it is not part of the local authority’s function and no part of the Review Board’s function to identify specific property available for a recipient’s occupation. Speaking for myself, I unreservedly accept that. … It is, in my judgment, quite sufficient if an active market is shown to exist in houses of the appropriate type in an appropriate place at the level of rent to which rent is restricted. There must, however, be evidence at least of that in a case falling within paragraph 11(3); otherwise the recipient, if he had to move, would have nowhere to go. … .”

Evans L.J. (on pp.501 and 502) said he “would be inclined to interpret paragraph (2)(c) as requiring that there shall be … a relevant active market in property of the relevant description, type or class …”, and that “in a paragraph (2)(c) case, if satisfied that there is an ascertainable market rent and an active market, then the authority is entitled to leave the applicants to take their chance of finding accommodation in that market”.

In R. v Waltham Forest London Borough Council, ex parte Holder (1997) 29 H.L.R. 71 Brooke J., as he then was, had to consider the meaning of the provisions relating to “suitable alternative accommodation” in regulation 11(6)(a) of the 1987 regulations – which was in similar terms to regulation 13(9) of the 2006 regulations. He observed (on p.78) that the “very concept must exclude the possibility that the authority can properly rely on a comparison with alternative accommodation in a completely different part of the country where rents are generally lower”. The two matters identified in paragraph (6)(a) as those which the authority must take into account were “the nature of the alternative accommodation and the facilities provided, having regard to the age and state of health of all the persons to whom paragraph (7) applies (the claimant, any members of his family and so on)” and “the terms of letting”. He went on to say (on pp.80 and 81):

“In my judgment on the proper construction of Regulation 11 as amended an authority when making the necessary comparison with suitable alternative accommodation must indeed satisfy itself that there is a relevant active market in property of the relevant description, type or class, but in making this judgment for the purposes of Regulation 11(2)(c) it need not concern itself with any financial considerations (other than the question of rent) which may make parts of that market wholly or relatively inaccessible to applicants like these. If satisfied that there is, on this basis, an ascertainable market rent and an active market, then as Evans L.J. said … [in ex parte Gibson] the authority is entitled to leave the applicants to take their chance of finding accommodation in that market.”

In ex parte Morgan, Collins J. considered whether, under regulation 11(2)(c) of the 1987 regulations, it was appropriate to compare a private sector landlord’s rent to rents being paid for similar “public sector accommodation” under a secure tenancy. He said (on pp.8 and 9):

“[In] order to compare the level of rent, it seems to me that [the Housing Benefit Review Board] have to bear in mind that they are concerned with whether the rent being paid by someone in the private sector is an unreasonably high rent. It is no part of the scheme to try to compel private sector tenants to move to council accommodation. … If, in deciding whether a rent is unreasonably high for a tenant in the private sector it is appropriate for the Council to have regard to the levels of public sector rent, then there is a potential unfairness to the private sector tenants because the usual result of taking account of what I call public sector tenancies is to produce a lower figure than is appropriate for private sector lettings. … [The] element of subsidy does mean that, on the whole, such rents are at a lower level. It seems to me that if [counsel for the Housing Benefit Review Board] is correct, it would be open to a Board to have regard only to public sector accommodation, because if they were satisfied that there was public sector accommodation which was … suitable alternative accommodation, then why should they not regard the rent being paid in the private sector as unreasonably high in comparison with that? [Counsel] says they would not do that. That would not be a correct way of going about it. But it seems to me that the difficulty with her submission is that there is nothing in the language of the regulation, if she be right, to prevent them doing that. Quite apart from that, it seems to me that even taking the public sector accommodation into account has the tendency to depress the amounts in question. I think that [counsel for the applicant] is right when he says that the purpose of this is to compare like with like, and to answer the question whether this landlord was charging an unreasonable rent for this tenant. That is what has to be considered. It seems to me that that is also consistent with the reference to the determination by a rent officer. It will be remembered that [regulation] 11(2)(c) says:

“Whether by reference to a determination by a rent officer or otherwise the rent payable is unreasonably high”.”

The First-tier Tribunal was “satisfied that the rent payable for the Roshni accommodation was unreasonably high by reference to the rent in respect of the alternative accommodation identified by [the city council]”, which it “accepted was suitable alternative accommodation” (paragraph 10 of its decision). It was “satisfied that at the level to which [the city council] had restricted the rent, £179.20 per week, there was a range of accommodation which was available” (also paragraph 10). It found that “for neither claimant was there any need to take into account a prospect of retaining employment as neither was in employment, nor was either required to be available for work as a condition of entitlement to benefit”; and that “[for] one of the claimants, the education of her child was not a relevant consideration because [the child] was only a matter of months old, and though the other claimant had a child [of] school age, the disruption to the child’s education had already arisen by means of the move to Roshni” (paragraph 11). It was therefore satisfied that the city council had “discharged the burden of showing that the [claimants’] rent was unreasonably higher than that of suitable alternative accommodation, and … had also shown that it was reasonable to expect the claimants to move” (paragraph 12).

In its interim decision ([2013] UKUT 418 (AAC) (30 August 2013)) the Upper Tribunal acknowledged that because both SA and SS had young children, this was a case in which regulation 13(4) would be engaged if the rent in question was found to be “unreasonably high” under regulation 13(3) (paragraph 8 of the interim decision), which would mean that no deduction could then be made unless suitable cheaper alternative accommodation was available.

Under the heading “The proper interpretation of Regulation 13”, the Upper Tribunal said (in paragraph 28) that the decision in ex parte Morgan “is authority for the proposition that one must, so far as practicable, compare like with like, and where there is relevant alternative private accommodation, it is the rent for that accommodation which must normally be the principal point of reference in considering the rent charged by a private landlord”. It went on to distinguish the case before it from ex parte Morgan:

“29.

Collins J was not, however, considering the position where there is no relevant unsubsidised accommodation with which a comparison can be made. Further, his comments about subsidy appear to me to be equally appropriate where the subsidy is a private charitable one just as where it is a public subsidy. An unsubsidised private landlord and his tenant are not to be penalised just because there happens to be a charity with a large endowment offering sub-market rents to a limited number of tenants.”

But it rejected the notion that “suitable alternative accommodation” in a case such as this must necessarily exclude subsidized accommodation:

“30.

On the other hand, it does not appear to me that “suitable alternative accommodation” can never include subsidised accommodation. That would have two results that would at least be highly unfortunate. Firstly, it would mean that in such a case there would be no statutory control over the rent which a private landlord might charge because there would be nothing relevant for the purpose of regulation 13(3)(b) compared to which the rent in question would be unreasonably high.

It continued:

“32.

I can see no reason why “suitable alternative accommodation” cannot include both subsidised and unsubsidised accommodation. No distinction is drawn in the use of the words to prevent the words from having their normal meaning. The effect of the decision in [ex parte Morgan] is that comparison must be like with like so far as reasonably possible, which means that in the normal case, involving private unsubsidised lettings, comparison should be within the unsubsidised private sector.

Under the heading “Comparison of Roshni with other subsidised accommodation”, it said this:

“35.

It is apparent from Roshni’s accounts that the loss of public funding meant that it could not operate once it lost its public funding without substantially increasing rents. It also appears, so far as I can presently tell, that the loss of public funding was the result of cut backs and not of any failings on its part, and that the new rental level was set to enable it to survive. I also note from the accounts for the year ending 31 March 2011 that it was 98% full during that year, indicating considerable demand for its facilities. The five comparables between them accommodate up to 83 women who have been victims of violence, although not all are in bedsitters and only two of them, with a maximum capacity for 17 women appear to concentrate on those from South Asia. Roshni’s 9 places therefore appears to represent a 50% increase on what would otherwise be available in that respect and an 11% increase on total availability.

…”.

Of the five “comparators” put forward, the “most obvious … in terms of both size and funding issues” was “comparator 3, [Gilgal,] a charity whose rent was used to determine the level to which Roshni’s rent should be reduced” (paragraph 36). Roshni’s accounts appeared to show “a far leaner operation than that of comparator 3” (paragraph 37). The accounts of two of the other comparators were “not as easy to compare with Roshni as comparator 3 as both appear to be larger bodies with significant other interests and funding”. A fourth appeared to be “an offshoot of a housing association …” and was “not a charity”. The other was “a very large and well known charity with substantial funding” (paragraph 38). The Upper Tribunal was “unclear on what basis it can be said that the rent charged by Roshni was unreasonably high compared with these comparators given the level of funding they received and that Roshni was doing no more than seeking to charge a rent which would enable it to cover its costs which appear modest when compared with comparator 3, there being no easy way of comparing them with those of the other comparators” (paragraph 39). If the city council wanted to maintain its contention that the rent charged by Roshni was unreasonably high by comparison with the rents charged by these comparators, it could now provide further evidence and submissions, and Roshni could respond (paragraph 40). In the final paragraph of its interim decision (paragraph 41) the Upper Tribunal said:

“I have not considered in this decision how, if it were to be decided that the rent was unreasonably high, a council should exercise its discretion to reduce it. If, however, the council does contend still that the rent was unreasonably high, then both the claimant and the council (and the Secretary of State if he wishes) should make submissions as to how I should exercise that discretion bearing in mind both the findings in this decision and the guidance of the Court of Appeal and House of Lords in [Mehanne].”

In its final decision ([2014] UKUT 137 (AAC) (11 March 2014)), the Upper Tribunal adhered to the approach it had adopted in the interim decision, rejecting the city council’s and the Secretary of State’s submissions challenging its correctness. It said (in paragraph 3) that Roshni was “a charity operating a women’s refuge for women from the South Asian continent who were experiencing domestic violence or parental conflict”. Before the financial year 2010/11, it had “received unrestricted public funding from Supporting People which it used to subsidise rentals for its rooms”. In the year ending on 31 March 2010 it had “received [£118,330] funding in this way (a reduction of more than [£30,000] in funding compared with the previous year)”. It had received “£64,054 in respect of accommodation”. It had “lost that funding, through no fault of its own at the end of that financial year”. The Upper Tribunal then referred (in paragraph 4) to the review Roshni had commissioned from “a local consultancy in housing and support”, which advised that “a contractual rent should be charged on £257.87 per week per room, … producing an eligible rent of £242.17”. This was “significantly in excess of the rent charged in the previous financial year by Roshni, which the council had accepted as reasonable” and “also, in total, significantly in excess of the total eligible rents charged by 5 comparators relied on by the council, which ranged between £140.75 and £210.16”. All those comparators, however, had “received public funding which they could use to subsidise rents”. There were “no comparators which were unsubsidised”. The Upper Tribunal referred (in paragraph 6) to the submissions made on behalf of the Secretary of State questioning the approach indicated in paragraph 33 of the interim decision:

“The Secretary of State has sought in further submissions to question whether this test is the correct one. He says that it is a highly subjective test as it would require a local authority to assess the extent to which the landlord has set the rent to compensate for a lack of subsidy, and that it would be difficult for a local authority to do this on the ground, as it may require it to make potentially complex theoretical judgments about what level of rent might have been charged by a comparator landlord had it not been in receipt of subsidy. It is also said that it would probably also require a wider consideration of the relevant organisation’s business model and projections to establish the level of rental income which a particular landlord required in order to remain operational, and thus … the level which they would … theoretically have been obliged to push rents up to had they not been in receipt of subsidy.”

The Upper Tribunal’s conclusions are in paragraphs 8 to 14 of its final decision:

“8.

… [The] test contemplated by me in [paragraph 33 of the interim decision] is very unlikely to involve the sort of complexity that is envisaged by the Secretary of State or the detailed analysis of the landlord’s accounts to which the council has subjected Roshni’s accounts in these cases. In the normal case, there will be a simple question as to the size of the subsidies in the comparables and their effect on the rental level. If without the subsidies the comparable rental levels would equal or exceed, or be very close to, the rents charged by the landlord, then it would normally follow that the rent charged by the landlord would not be unreasonably high by comparison with the comparators. The excess would simply compensate for the absence of subsidy and that excess would not be unreasonable. It would only be where the excess was more than minimally greater than the subsidised element of the comparable rents that the landlord might be expected to justify the difference and the local authority would have to consider the explanation. While there may be possible justifications for the extra rent that would mean that it was not unreasonably high by comparison, they are unlikely to depend on a detailed analysis of the landlord’s accounts, any more than the reasonableness of any other rent is determined, for example, by a rent officer by reference to such accounts. What may be relevant could include the reasonable provision of additional services or the omission of reasonable provisions by a comparator, leading to an unduly low rental. For example, the comparator may be letting rooms in a building due for demolition within a short period, so that there is no need to allow for the cost of major repairs in calculating the rent.

…

10.

In the present case, the closest comparator in terms of size and other factors is comparator 3, on which the council has relied in its latest submissions. Its accounts … show that it received £173,501 from Supporting People and £79,772 from accommodation charges. In other words only 31.5 per cent of its accommodation costs came from rental charges and the remaining 68.5 per cent came from relevant subsidies. It is plain that if the subsidy had been lost and it had been necessary for it to recoup all of its accommodation costs from rent, the rent would have needed to be in the order of £568 instead of £179. Even the smaller percentage subsidy of 58 per cent to which I came by a different calculation in my previous decision would mean a rental increase from £179 to £426. I find it impossible to see how Roshni’s unsubsidised rent of £242.17 can be regarded as unreasonably high in those circumstances. I note that it is suggested that other subsidies obtained in the course of the year by Roshni from other sources should be taken into account. However, they can only be taken into account if they can be used by Roshni to subsidise rents. Roshni’s evidence, which has not been seriously challenged and which I accept, is that most of those funds, which in any event fall far short of the lost grant from Supporting People, could only be used for other charitable purposes relating to children projects.

…

12.

As I have already indicated, it does not appear to me to be necessary or appropriate to engage on a detailed examination of Roshni’s and the comparators’ accounts. The council in this case has sought to do so, pointing out, for example, that Roshni’s support costs in their accounts are £21,815 more expensive than those of comparator 3. That is correct but it fails to differentiate between costs related to accommodation and those related to children projects. I note that comparator 3’s accounts show that it had a grant from the council of £16,972 for such projects, while Roshni’s accounts show not only a similar grant from the council but also two further grants totalling nearly £50,000, suggesting that Roshni will have spent far more on such projects than the comparator. The excess in relation to children projects could easily account for the £20,815 difference in support costs. Other criticisms could also be levelled at the council’s analysis of the accounts but for the reasons I have given they are not relevant to the outcome of this appeal.

13.

Given the demand for Roshni’s accommodation (it was 98 per cent full in the year in question), that it filled an important need beyond that provided by the funded charities, and that its loss of funding was the result of cutbacks and not of any failings on its part, I am satisfied that there is nothing remotely unreasonably high in the rents charged by it by comparison with the rent payable in respect of suitable alternative accommodation.

14.

I therefore find that the claimants’ eligible rents are the full amounts claimed by them.”

In response to the Upper Tribunal’s invitation in paragraph 41 of its interim decision, submissions on the discretion in regulation 13(3) had been made both for the city council and for Roshni. The city council had contended that a restriction of the eligible rent to £179.20 per week was appropriate; Roshni that if any restriction was justified – contrary to its case that there should be none – the appropriate eligible rent would be £210.16 per week. The Upper Tribunal did not express any conclusion on those submissions.

For the city council, Mr Manning submitted that the intention behind regulation 13(3)(b) is clear. The comparative nature of the exercise is essential. The local authority must undertake a comparison of rent levels. It must decide whether the rent payable by the claimant for his dwelling is “unreasonably high by comparison with the rent payable in respect of suitable alternative accommodation elsewhere”. It does not have to establish whether the landlord might be able to charge a lower rent – perhaps because it is operating inefficiently or uncompetitively having regard to its costs. The purpose of the provisions in regulation 13(3) and (4), Mr Manning contended, is twofold: first, to enable local authorities to protect the public purse by not, in effect, endorsing very high rents being charged by landlords when the “going rate” is much lower (regulation 13(3)); and second, to protect claimants by only permitting comparisons to be made with “suitable alternative accommodation” and requiring, in the case of the most vulnerable claimants, that such accommodation must actually be “available” to the claimant in question (regulation 13(4)). Refuges such as Roshni’s have not been exempted from eligible rent restriction under regulation 13(3). But the protection for vulnerable claimants under regulation 13(4) acts against the loss of such accommodation from the market.

Mr Manning submitted that the approach taken by the Upper Tribunal in this case was plainly wrong. If right, it would require a local authority to investigate the costs affecting the levels of rent landlords were charging, the availability to them of outside funding, and how good they were at running their business. In a particular case the authority might have to consider, for example, why a landlord had lost funding – either because of its own wrongdoing or for some other reason – or whether it needed to charge the rents it did to cover its reasonable operating costs. The legislature cannot have intended an exercise so complex and uncertain. The task of identifying what effect particular income streams might have on rents is not as easy as the Upper Tribunal seems to have thought. Subsidy may be available to one landlord and not to another for various reasons. And the impact of subsidy on rents will also differ widely. A local authority could not realistically be expected to explore such matters in determining whether a rent is comparatively “unreasonably high” under regulation 13(3)(b) – and it is not required to do so.

The practical difficulties of the Upper Tribunal’s approach are well illustrated, said Mr Manning, in paragraph 10 of its final decision. The Upper Tribunal made a comparison only with the rents charged by one other landlord – “comparator 3”. It seems to have assumed that all of the funding available to “comparator 3” was in fact being used to reduce the rent it was charging, while accepting that none of the funding potentially available to Roshni could have been used in this way. Roshni, it should be remembered, had chosen to stay outside the regulatory regime for registered providers of social housing, deliberately foregoing the funding which might have enabled it to charge lower rents. Without subsidy, it found, “comparator 3” would have been charging a rent of either £568 or £426, depending on the method of calculation. Even if this had been a finding the Upper Tribunal was entitled to make – which it was not – the existence of these two very different figures shows the kind of problem that can arise in a hypothetical exercise like this. It was not for the Upper Tribunal to speculate about what “comparator 3” might do if outside funding were no longer available to it.

Mr Knafler supported the approach and conclusions of the Upper Tribunal. Regulation 13(3)(b), he accepted, envisages an active market with vacancies within a particular geographical area. But implicit in it, he submitted, there is a requirement for a comparison between rents in the same market – and a recognition that one cannot properly compare rents being charged for accommodation in the private sector with rents in the public sector, where subsidies are available and rents are restricted. The concept of a rent being “unreasonably high” allows the decision-maker a wide margin for judgment. Unlike regulation 13(9)(a) and (b), regulation 13(3) does not prescribe any statutorily relevant factors. It does not prevent the decision-maker from looking behind the rent being charged by a particular landlord to discover why it was as high or as low as it was. To preclude an exploration of the reasons why one rent is higher than another would be to strip the word “unreasonably” of any real meaning. As Mr Knafler put it, the Upper Tribunal is entitled to undertake – as it did here – a “multifactorial” assessment – subject only to the principles of Wednesbury review (see paragraphs 34 and 35 of the judgment of Laws L.J. in R. (on the application of Khatun) v Newham London Borough Council [2005] Q.B. 37, at pp.55 and 56). An appellate court should be slow to interfere with such a decision – the decision of a specialist tribunal within the sphere of its own expertise (see paragraphs 45 to 47 of the judgment of Lord Wilson in Mathieson v Secretary of State for Work and Pensions [2015] 1 W.L.R. 3250, at pp.3268 and 3269).

In this case, Mr Knafler submitted, the Upper Tribunal made no error of law. The matters it took into account were well within the range of factors bearing on the comparison of rents required under regulation 13(3)(b), and it was free to give them the weight it did. In the light of the findings it made on the evidence before it – in particular, the findings in paragraphs 32 and 35 of its interim decision and paragraphs 3, 4, 10, 12 and 13 of its final decision – it was entitled to conclude that the rents being charged by Roshni were not “unreasonably high”. Roshni had lost subsidy through no fault of its own. The Upper Tribunal found the rents being charged for accommodation at other refuges artificially low. It looked at the question of reasonableness objectively, in the round, and from the perspective of the public purse. It concluded, in effect, that Roshni was providing good value for public money. In fact, said Mr Knafler, Roshni was costing the public purse less than would have been so if it had been in receipt of subsidy.

But, said Mr Knafler, the Upper Tribunal could have reached the same conclusion simply by excluding from the comparison it undertook the comparators receiving subsidy and subject to rental constraint. This approach was, in fact, indicated to it in the reply submitted on behalf of SA and SS, dated 23 May 2013, to the city council’s submissions of 11 March 2013 and the Secretary of State’s of 4 February 2013. Mr Knafler relied, in particular, on paragraphs 10 and 12 of that reply. It was submitted there that “the inclusion of [registered social landlord] providers as comparators must produce a lower figure for the purposes of sub-paragraph (3)(b) in a case involving supported accommodation provided by a registered charity operating in the private sector” (paragraph 10). And “[the] fact that [registered social landlord] providers may not receive a direct subsidy towards their rent does not mitigate the obvious unfairness of carrying out a comparison which is not on a like-with-like basis or address the argument that such an approach is contrary to the legislative purpose as identified by Collins J in ex p Morgan” (also paragraph 10). Because “[registered social landlord] providers receive grants that are publicly funded and they are thereby to an extent subsidised”, their inclusion in the comparative exercise under regulation 13(3)(b) would prevent the comparison being, as it must be, “a realistic comparison of the rents for the type of accommodation at issue on a like-with-like basis” (paragraph 12(i)). And “… [registered social landlord] rents are … restrained, by the Regulator’s controls, to sub-market levels” (paragraph 12(iv)). Mr Knafler adopted those submissions. With the substitution of the words “not from completely different markets” for the words “like-with-like”, he said, they would have appeared in a respondent’s notice if Roshni had filed one.

Elegant and persuasive as Mr Knafler’s submissions were, I cannot accept them. In my view the argument presented by Mr Manning is essentially correct. It seems to reflect the true meaning of the provisions with which we are concerned. As Mr Manning submitted, there is a danger here of over-complication. The concept whose meaning is in dispute in this appeal – the concept, in regulation 13(3)(b), of the rent payable for the claimant’s dwelling being “unreasonably high by comparison with the rent payable in respect of suitable alternative accommodation elsewhere” – envisages, I think, a relatively straightforward procedure which is neither unduly burdensome nor unduly complex. On its face, it requires a simple comparison between rents.

One must start with the language of the relevant provisions themselves, and, in the normal way, construe the words used by the legislature in accordance with their natural and ordinary meaning, having regard to their statutory context.

The basic purposes of the regime for housing benefit were identified by Lord Bingham in Mehanne, in the passage of his speech that I have set out in paragraph 9 above. And the intention underlying the provisions for “unreasonable payments” in regulation 13 of the 2006 regulations was identified by Sir Thomas Bingham M.R., as he then was, in ex parte Gibson – put simply, the protection of the public purse from “unreasonable payments” of housing benefit and the protection of claimants from homelessness.

With this in mind, there are two things I would say about the drafting of regulation 13 and its role in the legislative scheme.

First, the provisions of regulation 13(3) and (4) do not refer at all to the landlord and its circumstances. The focus throughout is on the claimant – under paragraph (3)(a), the size of his dwelling, the size of his household, and his requirements for accommodation; under paragraph (3)(b), the rent payable for his dwelling, and the rent payable for alternative accommodation suitable for him; under the provision for discretion as to the amount of reduction in eligible rent in paragraph (3), the cost to him of suitable accommodation elsewhere; under paragraph (4), the availability to him of suitable alternative accommodation, and his ability to move from his present accommodation. The same may be said of the provisions relating to “suitable alternative accommodation” and the availability of such accommodation, in regulation 13(9) and (10) – under paragraphs (9)(a) and (10), the age and health of the claimant and members of his family, and the security of tenure he enjoys; and under paragraph (9)(b), his job and the education of his children.

Secondly, the questions which these provisions pose for a local authority are expressed, mostly, in terms of what is “reasonable” or “unreasonable”: under regulation 13 as whole, whether payments would be “unreasonable”; under paragraph (3)(a), whether the dwelling in question is larger than is “reasonably required”; under paragraph (3)(b), whether the rent is, comparatively, “unreasonably high”; under paragraph (4), whether it is “reasonable to expect” the claimant to move from his present accommodation; under regulation 13(9)(a), whether the accommodation in question will be occupied on terms affording security of tenure “reasonably equivalent” to that presently enjoyed by the claimant. All of these matters call for the exercise of judgment in the making of the relevant decision, and the touchstone for that exercise of judgment is reasonableness.

Critical in this appeal is the concept of the rent payable for the claimant’s dwelling being “unreasonably high by comparison with the rent payable in respect of suitable alternative accommodation elsewhere”. The practical question here is whether, in the view of the local authority concerned, a particular rent is high enough to engage the requirement that the claimant’s eligible rent is reduced to the extent the authority considers appropriate.

This is explicitly a comparative exercise. The comparison is specifically limited to accommodation which is both “suitable” for the claimant and “alternative” to the accommodation he is occupying, in the particular respects indicated in regulation 13(9). As regulation 13(9) indicates, the concept of “suitable” accommodation will be principally a matter of the physical nature of the accommodation itself and the facilities provided in it, and the security of tenure available. Other considerations, such as geographical location, are not excluded. But as Brooke J. said in ex parte Holder (on pp.80 and 81), the authority is not concerned here with “financial considerations (other than the question of rent) which may make parts of that market wholly or relatively inaccessible” to the claimant.

In ex parte Gibson, when considering the provision in the 1987 regulations corresponding to regulation 13(3)(b), this court recognized that there should be, as Sir Thomas Bingham M.R. described it, “an active market … in houses of the appropriate type in an appropriate place at the level of rent to which the rent is restricted”, and, in the words of Evans L.J., “an ascertainable market rent and an active market …”. That emphasis on a “market”, to inform the comparison of rents under regulation 13(3)(b) is, I think, important. Regulation 13(3)(b) demands a realistic – rather than merely theoretical – evaluation of rental levels for accommodation of the same kind in the relevant area. This will require there to be reliable evidence of a “market rent” with which to compare the rent being paid by the claimant for the accommodation he is in.

Regulation 13(3)(b) does not compel the authority to enquire into the circumstances of the landlord in considering whether the rent in question is comparatively “unreasonably high”. The comparison it requires is, specifically, a comparison between the rent “payable” for the claimant’s dwelling and the rent “payable” for suitable alternative accommodation elsewhere. The subsequent reference to the authority finding the claimant’s rent “unreasonably high” must be understood accordingly. Once it has established that the rent is comparatively unreasonably high for him to pay, it is then obliged, subject to the provisions of paragraphs (4) and (7), to treat his eligible rent as reduced, exercising its discretion as to the amount of the reduction. It exercises that discretion by having regard, in particular, to the cost of suitable accommodation elsewhere – that is, the amount of rent the claimant would have to pay for such accommodation. These provisions are directed to the circumstances of the claimant, the rent he pays for the accommodation he occupies, and the rent he would have to pay for the alternative accommodation which is suitable for him. They go with the general tenor of the provisions in paragraphs (3), (4) and (9), all of which, as I have said, concentrate not on the landlord and his circumstances but on the claimant as tenant, his household, his accommodation and the rent paid by him.

It seems to me therefore that the concept of a rent being “unreasonably” high in regulation 13(3) is that the rent is unreasonably high for the claimant to have to pay, rather than unreasonably high for a particular landlord to charge. An “unreasonably high” rent under regulation 13(3) will be a rent higher than the rent one could reasonably be expected to pay in what Evans L.J. in ex parte Gibson described as the “relevant active market in property of the relevant description, type or class” – the “ascertainable market rent”. I see no justification for reading into this concept artificial constraints on the relevant market in which that other accommodation exists, or on the market rent for such accommodation. If there is a “relevant active market” and an “ascertainable market rent” for suitable, alternative accommodation – supported by evidence of the range of rents being paid for accommodation of the appropriate type, occupied with the appropriate security of tenure, in the appropriate general location – the comparison between rents must be objective, realistic and complete.

To manipulate the comparison of rents by excluding, for example, all but the levels of rent payable for suitable alternative accommodation which happens to be provided by a landlord in the private sector without the benefit of subsidy, or some other external source of funding, will render the exercise subjective, unrealistic and partial. It shifts the comparison from one between rental levels for accommodation of sufficiently similar type and tenure in the relevant area to one between rents only in one part of the market, and then, within that part of the market, to rents charged by landlords who happen to be funded in a particular way. This is to distort the market, and to produce a false and unreliable comparison. The same may be said of an approach that modifies the comparison of rents by making hypothetical adjustments for the individual circumstances of a particular landlord. Again, this is liable to produce a flawed comparison by reducing or eliminating real disparities between rents for relevant accommodation in the relevant market. Discovering the reasons why a particular landlord came to lose or to forego public funding, exploring the possibility of its being able to charge a lower rent if it were to operate in a different way, or speculating about its circumstances changing in the future, may not always be an impossible or even a difficult task. Such questions may not, in every case, require detailed scrutiny of a landlord’s accounts, or close investigation of the way it runs its business. Here, in the Upper Tribunal’s view, they did not. But they bring to the exercise of comparing rents under regulation 13(3)(b) a degree of artificiality which, in my view, is not merely unnecessary but also inappropriate.

I think this understanding of regulation 13(3) sits well with the legislative intent. It recognizes the aim of protecting the public purse from excessive payments of housing benefit in cases where rents are, comparatively, “unreasonably high”. And it does not offend the purpose of protecting claimants from homelessness.

In a case where a claimant was paying a rent that was, comparatively, “unreasonably high” – except where paragraph (4) applied and suitable alternative accommodation was not available to him – the authority would be able to reduce his eligible rent by an “appropriate” amount. Claimants within one of the special categories in paragraph (4), as SA and SS were, would be protected against any deduction in their eligible rent, in two ways: first, by the requirement that suitable “cheaper” alternative accommodation is, in fact, “available”; second, that it is reasonable to expect him to move from the accommodation he is in. In exercising the discretion in paragraph (3), the authority is not obliged to reduce the eligible rent by the full amount of the difference between the rent the claimant was paying for the accommodation he was in and the rent he would pay for suitable alternative accommodation elsewhere. The requirement is simply that, in judging what an “appropriate” reduction would be, it must have regard, in particular, to the cost of the suitable alternative accommodation. In Mehanne the House of Lords emphasized the breadth of the authority’s discretion (see paragraphs 8 to 13 in the speech of Lord Bingham, at p.544C to p.546C; and paragraphs 19 to 27 in the speech of Lord Hope of Craighead, at p.547B to p.549F). Relevant considerations could include any personal or financial circumstances of the claimant bearing on his “housing situation” (paragraph 13 in Lord Bingham’s speech).

It follows, in my view, that the Upper Tribunal fell into error in the approach it took in the appeals of SA and SS. Its comparison of rental levels under regulation 13(3)(b) was not a true and full comparison between the rent payable by SA and SS for their accommodation in Roshni’s refuge and the rent payable for other accommodation of comparable type and tenure in the “relevant active market”. It effectively applied as criteria of comparability the availability and amount of public funding, or subsidy, which would enable the landlords in question to reduce the rents they charged. In undertaking the comparison, it adjusted upwards the real rent being charged by the only landlord whose accommodation it ultimately relied upon as comparable – “comparator 3” – to arrive at a range of hypothetical rents, on different assumptions of its own as to the percentage proportion of the landlord’s “accommodation costs” recouped from, respectively, subsidy and rent. This was not, in my view, an exercise consistent with the requirements of regulation 13(3).

The Upper Tribunal was clearly anxious to reflect in its decision the fact that Roshni was a charity, providing – without the benefit of public funding – accommodation for women who had suffered domestic violence, and that, without subsidy, charities of this kind could not afford to operate refuges, leaving “victims of violence either homeless or at risk at the homes they wished to leave” (paragraph 33 of the interim decision and paragraph 3 of the final decision). It was aware of the very high demand for Roshni’s accommodation. It found that Roshni was meeting a need which “funded charities” were not able to meet or meet in full. And it was satisfied that Roshni had lost public funding as “the result of cutbacks and not of any failings on its part” (paragraph 35 of the interim decision and paragraph 13 of the final decision). All of these considerations it saw as relevant to the comparison of rents under regulation 13(3)(b). I do not think it was right to do so. Where concerns such as these go to the risk of vulnerable claimants finding themselves without accommodation suitable for them, the legislative scheme allows for them – in the discretion given to the authority under regulation 13(3), and in the provisions in regulation 13(4) and (9)(b). Where they go to the difficulties of private sector landlords – charities among them – providing accommodation in refuges without the advantage of public subsidy, the legislature has not identified them as relevant to the comparison of rents under regulation 13(3)(b). And in my view they do not bear on that exercise.

Mr Manning said the Upper Tribunal clearly relied on the passage in the judgment of Collins J. in ex parte Morgan where he accepted that the purpose of comparing rents under regulation 11(2)(c) of the 1987 regulations was “to compare like with like, and to answer the question whether this landlord was charging an unreasonable rent for this tenant”. That case, Mr Manning submitted, was wrongly decided, because the judge misconstrued regulation 11(2)(c). But anyway, as the Upper Tribunal acknowledged, the question there was different from the one raised by the appeals of SA and SS. It was, essentially, whether a rent being paid by a claimant for accommodation provided by a landlord in the “private sector” could properly be compared with rents charged in the “public sector”. In its interim decision the Upper Tribunal seems to have recognized the danger in departing from the statutory words. However, the approach it took was to ask itself, first, whether a “like with like” comparison was “reasonably possible” – and, having found it was not, to compare the rent charged by Roshni, an unsubsidized landlord, with the rent charged by a subsidized landlord in – as Mr Manning put it – “the same charitable sector of the housing rental market” (comparator 3). This was not, in fact, a “like with like” comparison of the kind contemplated by Collins J.. It was, ostensibly, a comparison between rents charged by landlords in the same part of the housing market, though differing in their status as providers of housing, their regulation and their funding. It was, said Mr Manning, an “intermediate approach”, consistent neither with regulation 13(3)(b) nor with the reasoning of Collins J. in ex parte Morgan.

Mr Knafler did not accept that ex parte Morgan was wrongly decided, or that it was of no relevance to the issue in this appeal. He aligned his submissions on the approach required under regulation 13(3)(b) with the “like with like” principle, but advocated a comparison between rents which were, as he put it, “not from completely different markets”.

As I understand the “like with like” approach indicated by Collins J. in ex parte Morgan, in any case where the claimant was in accommodation provided to him by a “private sector” landlord it would preclude from the comparison of rents under regulation 13(3)(b) rents payable for “suitable alternative accommodation elsewhere” if they were “public sector” rents. For reasons I have already given, I think it is hard to see how such an approach can be reconciled with the true construction of the wording of paragraph (3)(b). Taken to its logical conclusion – as the Upper Tribunal seems to have appreciated in this case – a “like with like” approach will tend to exclude rents that would otherwise find their place in a true and full comparison of rental levels for relevant accommodation in the relevant market, perhaps to the point at which no comparison is possible at all. It would thus defeat the purpose of paragraph (3)(b). But I do not think we have to go so far in this case as to hold that ex parte Morgan was wrongly decided. That is not a conclusion necessary to the determination of this appeal, because, as Mr Manning submitted, the course taken by the Upper Tribunal here was clearly a departure from the “like with like” approach. It was, as he submitted, an “intermediate approach”. And in my view, as I have said, that approach was incorrect.

I can see the good sense of the observations made by Collins J. about the legislative scheme as a whole. They have some resonance in a case such as this. As Collins J. said, it was no part of the legislative scheme for housing benefit to compel private sector tenants to move to accommodation provided in the public sector, if they could. Tenants eligible for housing benefit would likely be those in greatest need, least able to afford to pay for the accommodation provided to them, and most at risk of having to leave that accommodation and look for accommodation elsewhere if their housing benefit was reduced below the level of the rent they were paying. These observations seem especially powerful in a case where the claimants are the victims of domestic violence and the accommodation provided to them is in a refuge whose landlord is a charity without public funding. No doubt they will carry still more weight where there is a shortage of such accommodation in the relevant area. In this case, not surprisingly, they did – as is plain in paragraph 33 of the Upper Tribunal’s interim decision and paragraph 13 of the final decision.

I agree with Collins J. that the legislative scheme is not intended to force tenants in private sector accommodation to move to accommodation provided in the public sector. But to avoid that mischief it is not necessary to exclude public sector accommodation from the comparison of rents under regulation 13(3)(b). It can be avoided through the discretion in regulation 13(3) as to the amount of the reduction in “eligible rent” if the claimant’s rent is found to be “unreasonably high”. The range of factors which can bear on the exercise of that discretion is wide.

As I have said, although submissions on the discretion in regulation 13(3) were made to the Upper Tribunal in response to the invitation in paragraph 41 of its interim decision, it did not find it necessary to deal with those submissions in its final decision (see paragraphs 21, 22 and 23 above). Neither Mr Manning nor Mr Knafler addressed us on this aspect of the case at the hearing, but, at our request, they have made further written submissions in the light of the decisions of the Court of Appeal and the House of Lords in Mehanne.

At the relevant time in Mehanne (1 January 1996), the provision on discretion in regulation 11(2)(c) of the 1987 regulations, as amended by the 1991 regulations, was in materially similar terms to that in regulation 13(3) of the 2006 regulations; it stated that “… where it appears to the authority that the dwelling is larger than is reasonably required or that the rent is unreasonably high, the authority shall, subject to paragraphs (3) to (4), treat the claimant’s eligible rent as reduced by such amount as it considers appropriate having regard in particular to the cost of suitable alternative accommodation elsewhere and the claimant’s maximum housing benefit shall be calculated by reference to the … eligible rent as so reduced”. The sole issue before the House of Lords was whether, under that provision, it was open to the authority (or the review board) to “take account of the claimant’s personal circumstances (so far as relevant to his housing situation) when considering the amount by which it is appropriate to treat his eligible rent as reduced” (see paragraph 1 of Lord Bingham’s speech, at p.540G-H). As Lord Bingham noted (in paragraph 7, at p.543D), “[there] was much common ground between the parties on the construction of regulation 11”, including this (at p.544A-C):

“It was agreed that in a case to which paragraph (2) applied the local authority was not required to determine the claimant’s eligible rent by simply reducing his excess rent to his alternative rent. The paragraph could with the utmost simplicity have been drafted so as to have that effect, but such was not the course adopted. The obligation on the local authority was to “treat the claimant’s eligible rent … as reduced by such amount as it considers appropriate having regard in particular to the cost of suitable alternative accommodation elsewhere”. Thus the duty of the local authority, following ascertainment of the excess rent and the alternative rent, was not to perform the merely mechanical task of treating the former as reduced to the level of the latter. Instead, the local authority was to exercise a judgment, or, as it is often called, a discretion. The local authority was to treat the excess rent as reduced by some amount, but not to a level below the alternative rent.”

The claimant in that case was a refugee, who, with his wife, had claimed asylum in the United Kingdom. The House of Lords accepted, as had the Court of Appeal, that the factors relevant to the exercise of the discretion in that case would include the pregnancy of the claimant’s wife, his difficulty in finding other accommodation, and the consequences of his having to move from the accommodation he was in as a result of the reduction in the eligible rent.

In the Court of Appeal Mummery L.J. said ([2000] 1 W.L.R. 16, at p.24A-F):

“(1)

The substitution of “shall” for “may” did not deprive the review board of a discretion as to the amount of a reduction in the eligible rent. The amount of the reduction is not automatic or mandatory: it is by such amount “as it considers appropriate”. That is the language of discretion rather than of an obligation always to make a reduction in the eligible rent by the full amount of the difference between the rent and the cost of suitable alternative accommodation.

(2)

The use of the expression “in particular” does not limit the range of circumstances to be considered in determining the amount of reduction in the eligible rent or exclude all consideration of factors other than the factor particularly identified (i.e. the cost of suitable alternative accommodation). That factor is singled out for special mention and is thereby given the status of a mandatory consideration which carries the most weight in making a decision on the amount of any reduction in the eligible rent.

(3)

Other factors may be taken into account in determining the amount of the reduction, so long as they are reasonably relevant to that decision. It is for the decision-making body to decide how much importance to attach, or how much weight to give, to each of the other factors.

(4)

As to the other factors relied upon by the claimant in this case, the review board may take account of the following relevant factors which were all present during the relevant benefit period before the review board: the pregnancy of the claimant’s wife, which was known to the review board; the difficulty of the claimant in finding other accommodation and the consequences of having to move from this accommodation as a result of the reduction in the eligible rent (e.g. duty of the authority to rehouse the homeless). In my view, these are not irrelevant individual personal or financial circumstances of the claimant. They are all reasonably relevant to the housing situation of the claimant rather than, for example, to his inability to pay the rent.

(5)

The question of the amount of the reduction should therefore be reconsidered by the review board in the light of this judgment.”

Lord Bingham endorsed that analysis (in paragraph 13, on p.546B-C):

“Criticism was made of the expression “reasonably relevant” but it is obvious that Mummery [L.J.] meant “reasonably (or properly) regarded as relevant” and that is the correct test. In the absence of clear language I would be very reluctant to conclude that the local authority or the review board were precluded from considering matters which could affect the mind of a reasonable and fair-minded person when deciding on the level to which a claimant’s eligible rent should be treated as reduced, and the very general language of paragraph (2) gives no ground for inferring that such a constraint was intended.”

In his speech Lord Hope also rejected “too narrow a view of regulation 11(2)” (paragraph 24, p.548D). In the same vein as Lord Bingham, he said (in paragraph 26, on p. 549A-D):

“I would also attach importance to the fact that the risk of hardship due to homelessness is inherent in the nature of the exercise which the authority is required by regulation 11(2) to carry out. In order to avoid unreasonable demands on the public purse, the authority is required to make a reduction from the eligible rent if the dwelling which the claimant occupies is larger than is reasonably required or the rent is unreasonably high by comparison with that payable for suitable accommodation elsewhere. The claimant who is subject to this reduction is then faced with a choice. Either he must move to suitable accommodation which is available elsewhere for letting at a rent which will be covered by the amount of his housing benefit, or he must accept the fact that the housing benefit which he receives will fall short of what he needs to pay the rent. Some claimants will be able to make this choice. Others may not. For them, this may be a choice that cannot be made at all. Suitable alternative accommodation elsewhere may not be available in their area. A move to suitable accommodation in another area may bring with it loss of employment or other forms of very real hardship. And the margin between the rent that has to be paid and the housing benefit may be such that claimants will find themselves at risk of being evicted because they are genuinely unable to pay the rent when it falls due.”

(and in paragraph 27, on p.549E-F):

“The authority has a statutory duty to rehouse claimants who become homeless. There are likely to be cases where all that is needed to prevent homelessness is a reduction in the gap between the eligible rent and the alternative rent fixed by the housing benefit officer. Reducing the gap may be enough to enable claimants to meet their obligations and remain in the accommodation which they occupy. An interpretation of regulation 11(2) which would disable an authority from reducing the gap between the eligible rent and the housing benefit in cases where to do this would prevent homelessness would seem to be contrary to good sense. It would also, I think, be contrary to the overall purpose of the Regulations. I would hold that there is nothing in the wording of regulation 11(2) which drives one to the conclusion that the authority is so restricted in the discretion which it can exercise.”

Mr Knafler submitted that even if the rents charged by Roshni were unreasonably high in the “artificial sense advocated by [the city council]”, the Upper Tribunal’s conclusion in paragraph 10 of its final decision – to the effect that the rents were intrinsically reasonable as a matter of fact – would still be sound. Roshni was not, in fact, costing the public purse more than “the comparator refuges”. Its rents left it “just at the survival level”. If it did not survive, Mr Knafler submitted, “some victims of domestic violence would be homeless and without succour” (see paragraphs 33, 35, 36 and 39 of the interim decision and paragraphs 6 and 13 of the final decision). Inherent in the provisions of regulation 13 of the 2006 regulations (as it was in regulation 11 of the 1987 regulations) is the principle that reasonableness must prevail, both when an authority is considering whether the rent is unreasonably high and in the exercise of its residual discretion as to the amount of any reduction in the eligible rent – when there is “no limit to the range of relevant considerations”. As Lord Bingham said in paragraph 13 of his speech in Mehanne, the ultimate question is what a “reasonable and fair-minded person” would decide. Cast in that role – having set aside the decision of the First-tier Tribunal under section 12(2)(a) of the Tribunals, Courts and Enforcement Act 2007 and re-making it under section 12(2)(b)(ii) – the Upper Tribunal would clearly not have concluded that the eligible rent should be reduced in this case. And, Mr Knafler suggested, we are “just as well placed” to decide that question as the Upper Tribunal had been.

Mr Manning submitted that the point decided in Mehanne does not touch the issues we have to decide. This appeal is concerned only with the question of whether the Upper Tribunal erred in law in concluding that the rents were unreasonably high by comparison with the rent for suitable alternative accommodation elsewhere – and not at all with the question of how the discretion in regulation 13(3) should be exercised if the rents were unreasonably high. The decision in Mehanne is not authority for the proposition that a landlord’s circumstances are relevant to the exercise of the discretion, and lends no support to the argument that they are relevant to the prior question of whether a claimant’s rent is unreasonably high “by comparison with the market rate”. It emphasizes the need to protect the public purse from “unreasonable demands” upon it, and the importance of balancing that need against the circumstances of claimants, including those who are vulnerable (see, for example, paragraph 5 of the speech of Lord Bingham and paragraph 26 of Lord Hope’s). In this case, as the First-tier Tribunal accepted, the city council had restricted the eligible rents by a reasonable amount to a reasonable level (see paragraphs 10 to 12 of the First-tier Tribunal’s decision). This was not a live issue before the Upper Tribunal. It is not live in this appeal. And Roshni has no standing to raise it now.

There is, I think, a good deal of force in Mr Knafler’s submissions here. In conferring on the authority a discretion as to the appropriate amount of the reduction in a claimant’s eligible rent, the legislative scheme enshrines the principle that the eligible rent need not, and in some cases certainly should not, be reduced by the full difference between the rent he or she is actually paying and the rent payable for suitable alternative accommodation elsewhere. Where a reduction in the eligible rent by the whole of that difference, or a large proportion of it, could result in homelessness or hardship there may be a very strong case for tempering the reduction, perhaps even for limiting it to a nominal amount. In circumstances of the kind we are concerned with here, the case for doing that may be especially strong.

But in my view Mr Manning is right to submit that it would not be appropriate for us, in deciding this appeal, to attempt to exercise the discretion in regulation 13(3) ourselves. That, however, is not because the question of the appropriate amount of any reduction in the eligible rent is beyond the scope of the city council’s grounds of appeal. Nor is it because Roshni has no standing to raise the question itself. It is simply because we are not able to evaluate – in accordance with the lawful approach we have indicated – the evidence and argument bearing on the question of whether the rents payable by SA and SS for accommodation in Roshni’s refuge were “unreasonably high …”, and because, contrary to Mr Knafler’s submission, we could not then go on to determine by how much the eligible rent should be reduced. Those are quintessentially matters of judgment for the Upper Tribunal. If we allow the city council’s appeal to this court, the appeals of SA and SS should, I think, be sent back to the Upper Tribunal to be determined again in the light of our judgments. It will then be for the Upper Tribunal to decide afresh, taking the correct approach under regulation 13(3)(b), whether the rents in this case were “unreasonably high …”. If that is what the Upper Tribunal concludes, it would then have to go on to consider, under the discretion in regulation 13(3), the amount by which the eligible rent should be reduced. And in doing that it would have well in mind what was said by Lord Bingham and Lord Hope in Mehanne.

For the reasons I have given, I would allow these appeals.

I agree.

I also agree.