Ward & Ors, R (on the application of) v The London Borough of Hillingdon & Ors

Both Ms Ward and Mr McDonagh are Irish Travellers. Ms Ward is 29 years' old and has three young children. She is a single parent. She cannot read or write. Having spent a significant part of her early life travelling, she says that she now wishes to settle, at least while her children are young and in school. Her children's educational opportunities are important to her.

Ms Ward was living with relatives on a caravan site in Hillingdon when she was asked to leave and became homeless in 2015. She applied to Hillingdon for homelessness assistance. She was housed in a bed and breakfast hotel and then moved to another, before she was provided with accommodation under Part 7 of the 1996 Act. She says that the accommodation was in very poor repair and an independent environmental health officer has confirmed that it is unfit for human habitation.

She was not placed on the housing register until her solicitors wrote on 29 January 2018 requiring Hillingdon to do so. In response Hillingdon notified her that she had now been placed on the register, but that because she had not lived in the borough for ten years she was placed in band D.

Mr McDonagh is 61 years' old. He lives with his wife adult son and daughters. All four are disabled and require care because of their disabilities. He too is disabled.

Mr McDonagh receives a Carer's Allowance in recognition of the more than 35 hours per week care he provides for his daughter. All three children receive Employment Support Allowance and Disability Living Allowance.

Mr McDonagh has lived in Hillingdon since 2011. He became homeless in 2012 and was accommodated by Hillingdon at a series of addresses until he was made to leave temporary accommodation on 1 March 2017. Between March 2017 and February 2018, save for occasional nights with friends or family, he and his children slept in his van until they were provided in February 2018 with accommodation under Hillingdon’s severe weather protocol. They continue to live in temporary accommodation.

By letter dated 8 June 2018 Hillingdon informed Mr McDonagh that the hardship panel, to whom his housing register application had been referred, had decided that he did not meet the criteria for social housing. He was informed that if he disagreed with the decision he could request a review. Because of a change of factual circumstances Mr McDonagh was later allowed onto the register and placed in band D.

In both these cases Ms Polly Neate, the chief executive of Shelter, gave evidence about the difficulties faced by Irish Travellers. She drew particular attention to the effect of homelessness on children; and explained that changing schools has a long term impact on children’s lives and futures. Ms Natalie Stables, the head of Ethnic Minority and Traveller Achievement Services in Salford also gave evidence. She explained that educational outcomes for Gypsy and Roma children (“GRT children”) are very poor; and are the worst out of all ethnic groups. The reasons are complex. They include a high level of adult illiteracy. There was some anecdotal evidence to

Mr Gullu is a Kurdish refugee of Turkish nationality. In Turkey he was tortured and persecuted on the ground of his ethnicity and beliefs. He came to the UK in July 2011 and applied for asylum. In November 2013 the Secretary of State (“the SSHD”) granted him indefinite leave to remain. While Mr Gullu’s asylum application was being processed he was destitute. The SSHD therefore provided him with accommodation. Normally accommodation for asylum seekers is provided outside London. But because Mr Gullu was receiving treatment for the psychological effects of torture, he was accommodated in Hillingdon. His entitlement to that accommodation came to an end when he was granted leave to remain. Thus he became homeless. On 18 December 2013 he applied to Hillingdon for housing assistance. Hillingdon initially refused his application and refused to register him on the housing register because he had not lived in Hillingdon for 10 years.

Thereafter, Hillingdon accepted that it owed Mr Gullu the main housing duty under section 193 (2) of the 1996 Act. But it maintained its refusal to register him on the housing register because he did not satisfy the 10-year residence requirement. Hillingdon confirmed on 10 February 2017 that he could register on the scheme because he has a reasonable preference. He was placed in Band D.

Ms Chloe Morgan, from the Refugee Council, gave evidence about the housing of asylum seekers. She said that where asylum seekers are housed during the pendency of their applications, they have no choice where to live. They are simply allocated housing: usually out of London and the south-east.

Moreover, accommodation providers move asylum seekers from place to place. Providers are permitted to move asylum seekers twice a year without their consent; but the Home Office Select Committee recorded instances of asylum seekers being moved more frequently. She concluded that the fact that asylum seekers are regularly moved around between different accommodation providers makes it difficult for them to build up local connections in a particular area. Turning to Hillingdon’s allocation policy her evidence was that the 10-year residence requirement will disproportionately affect asylum seekers and therefore refugees. As she explained:

“Refugees are very unlikely to have been resident in any Local Authority borough for 10 years.”

In R (Ahmad) v Newham LBC [2009] UKHL 14, [2019] PTSR 632 the House of Lords sounded a warning against over-interference in the policy choices made by a housing authority in framing its allocation policy. Lord Neuberger said at [46]:

“… as a general proposition, it is undesirable for the courts to get involved in questions of how priorities are accorded in housing allocation policies. Of course, there will be cases where the court has a duty to interfere, for instance if a policy does not comply with statutory requirements, or if it is plainly irrational. However, it seems unlikely that the legislature can have intended that judges should embark on the exercise of telling authorities how to decide on priorities as between applicants in need of rehousing, save in relatively rare and extreme circumstances. Housing allocation policy is a difficult exercise which requires not only social and political sensitivity and judgment, but also local expertise and knowledge.”

He added at [62]:

“Knowledge of the circumstances of applicants generally, long term strategy considerations, expertise, political and social awareness, and local knowledge all have a part to play when it comes to formulating and implementing a housing allocation scheme. With information essentially consisting of the scheme itself, the circumstances of the particular applicant and a few statistics (of questionable mutual consistency), the court should be very slow indeed to second guess Newham.”

Despite these warnings the court must, of course, be satisfied that a housing allocation policy does not unlawfully discriminate, as both Lady Hale and Lord Neuberger made clear at [14] and [37] respectively. Protection against unlawful discrimination, even in an area of social and economic policy, falls within the constitutional responsibility of the courts. Even in the area of welfare benefits, where the court would normally defer to the considered decision of the policy maker, if that decision results in unjustified discrimination, then it is the duty of the court to say so: R (SG) v Secretary of State for Work and Pensions [2015] UKSC 16, [2015] 1 WLR 1449 at [160].

In addition, paragraph 16 of the Secretary of State’s statutory guidance expressly requires housing authorities to comply with the Equality Act 2010.

In R (H) v Ealing LBC [2017] EWCA Civ 1127, [2018] PTSR 541 this court considered the correct approach to an allegation that a particular part of a housing allocation policy amounted to indirect discrimination. The policy in question set aside a small but not insignificant proportion of lettings for “working households” (“WHPS”) and “model tenants”. One of the arguments advanced on behalf of Ealing was that, in considering the question whether a particular part of the allocation policy (a “PCP”) amounted to indirect discrimination it was necessary to consider the whole of the policy “in the round”. The policy, considered in the round, contained a number of “safety valves” which meant that, overall, the disadvantaged groups were not disadvantaged by the particular parts of the policy under attack. This court had “no hesitation” is rejecting that argument. Sir Terence Etherton MR said:

“[58]. In the light of that acceptance on the basis of those statistics, and the concession by Ealing on this appeal that each of the two priority schemes is a PCP, it inevitably follows that the WHPS gives rise to indirect discrimination within section 19(2) of the EA 2010. The wording of section 19(2) is precisely applicable, namely that participation in the WHPS is open to those who are not women, disabled people or elderly, and people who do have any of those characteristics will be disadvantaged in comparison because they are less likely to be in work than those who do not have those characteristics, and the claimants here are within those protected groups and would be disadvantaged.

[59] In short, it is contradictory of Ealing to concede, on the one hand, that for the purposes of section 19(2) of the EA 2010 the WHPS is a PCP, and, on the other hand, to seek to rely on Ealing's housing policy as a whole to rebut the PCP's discriminatory impact on the relevant protected groups. What this highlights is that the matters on which Ealing relies, the socalled safety valves, are matters which properly are relevant to justification under section 19(2)(d) of the EA 2010 rather than the existence of indirect discrimination under section 19(2)(a)– (c) of the EA 2010.”

In the light of that, it is necessary to consider paragraph 2.2.4 of Hillingdon’s policy separately. The question is whether that is a PCP that amounts to indirect discrimination.

Supperstone J did not discuss this question, which was conceded by Hillingdon (see para [31] (iv) of the judgment); but concentrated on whether Hillingdon was able to justify its policy. Hillingdon was also prepared to make the same concession before Mostyn J. But he did not accept it; and held that the policy did not discriminate against Mr Gullu. Before us Mr Rutledge QC, on behalf of Hillingdon, repeated the concession; and said that he was not able to support Mostyn J’s reasoning on this question. Nevertheless, I ought to explain why I consider that this concession was rightly made; and why Mostyn J was wrong to reject it.

Mostyn J dealt with the question of discrimination as follows:

“[15] The claimant is a recent arrival in Hillingdon who is a refugee. He is discriminated against in favour of long-term residents not because he is a refugee but because he is a shortterm resident. Nobody is suggesting that discrimination on that basis is to be impugned. Indeed, as I have pointed out, it has been expressly authorised by Parliament and strongly encouraged by the government.

[16]

The correct analogue is therefore another short-term resident who is not a refugee. That analogue might be a recent arrival from another part of the UK or a recent arrival from the EEA exercising treaty rights. The same treatment is meted out to the claimant and the analogue – both are denied priority by virtue of the 10-year residency rule. The claimant's case can only get off the ground if he can show that his circumstances and those of the analogue are materially different: that they are unalike cases. If he can show that they are unalike then the defendant has to justify the same treatment being applied to both.

[17]

But are they unalike? Mr Burton says the circumstances of a refugee and those of a voluntary migrant from Yorkshire or France are different because the refugee has no choice but to apply in Hillingdon whereas the analogue comes to Hillingdon by choice. Further, the refugee may be more vulnerable as a result of the persecution he has suffered which has resulted in the award of refugee status. All of this is true, but so what? The reason that each has started the 10-year journey may be different but that is immaterial to the process of starting the clock and counting the days, which is all that the measure stipulates.”

Mr Burton, for Mr Gullu, supported by Mr Squires for the Equality and Human Rights Commission, submits that Mostyn J did not distinguish between direct discrimination and indirect discrimination. It is not suggested that the 10-year residence requirement amounts to direct discrimination. But the whole basis of indirect discrimination is that what appears to be a neutral provision is substantially more difficult for a person with a protected characteristic to comply with than a person without that characteristic.

In the present case the protected characteristic is that of “race” as defined. Since “race” includes national origins, and a racial group may be made up of two or more distinct sub-groups, non-UK nationals are a class of people sharing a racial characteristic. It is also possible, of course, to define Mr Gullu’s protected characteristic differently. As well as being a non-UK national, he is also a nonUK/EEA national; a Turkish national, and a person of Kurdish ethnicity. Each of these could be described as a protected characteristic.

Mostyn J’s first error was to compare Mr Gullu with an individual analogue (“a voluntary migrant from Yorkshire or France”). If the case were one of direct discrimination, that would be the correct approach. But in a case of indirect discrimination, it is not. In the case of indirect discrimination, the correct comparison is between groups rather than individuals. Lady Hale explained this in Essop at [41]:

“… all the workers affected by the PCP in question should be considered. Then the comparison can be made between the impact of the PCP on the group with the relevant protected characteristic and its impact upon the group without it. This makes sense. It also matches the language of section 19(2)(b) which requires that “it”—ie the PCP in question—puts or would put persons with whom B shares the characteristic at a particular disadvantage compared with persons with whom B does not share it. There is no warrant for including only some of the persons affected by the PCP for comparison purposes. In general, therefore, identifying the PCP will also identify the pool for comparison.”

As Lady Hale also emphasised at [27]:

“… there is no requirement that the PCP in question put every member of the group sharing the particular protected

characteristic at a disadvantage.”

The groups, for comparative purposes, consist on the one hand of those who share the relevant protected characteristic (“the protected group”); and on the other hand those who do not (“the comparator group”). In Ms Ward’s case the relevant characteristic is being an Irish Traveller. In Mr Gullu’s case, it is being a non-UK national. So the question in Ms Ward’s case is: are Irish Travellers put at a disadvantage in satisfying the 10-year residence requirement as compared with persons who are not Irish Travellers? In her case it was common ground that they were. The question in Mr Gullu’s case is: are non-UK nationals put at a disadvantage in satisfying the 10-year residence requirement as compared with persons who are UK nationals? Mostyn J’s second error was to select as an analogue a person who could not satisfy the 10-year residence requirement (because they were a recently arrived voluntary migrant). Compounding this error was his concentration on some only of the comparator group. What he ought to have done was to have considered the comparator group (UK nationals) as a whole. The fact that some members of the comparator group are also disadvantaged by the PCP does not negate indirect discrimination, if a higher proportion of the protected group suffer that disadvantage.

If, then, one asks: does a 10-year residence requirement disadvantage non-UK nationals more than UK nationals, the answer must be “yes”. A long line of cases has accepted that a residence requirement disadvantages non-UK nationals. In Orphanos v Queen Mary College [1985] AC 761 a residence requirement in order to qualify for reduced tuition fees indirectly discriminated against a Cypriot national (before Cyprus joined the EU). Lord Fraser said:

“I have no doubt that the proportion of persons of non-British and non-E.E.C. nationality who wish to attend the college and who can comply with the requirement of having ordinarily resided in the E.E.C. area for three years immediately before 1 September 1982 is substantially smaller than the proportion of persons not of that group (i.e., persons who were British or E.E.C. nationals) who wish to attend the college and who can comply with it. That seems obvious and causes no difficulty.”

In R (Winder) v Sandwell MBC [2014] EWHC 2617 (Admin), [2015] PTSR 34 a residence requirement in order to benefit from a council tax reduction scheme indirectly discriminated against non-UK nationals. It is notable that in that case, the claimants were in fact all UK nationals. Hickinbottom J accepted the submission that:

“… the residence requirement is indirectly discriminatory because it is liable to affect a larger proportion of foreign (including EU) nationals than British nationals, because the former are inherently less likely to have lived their lives (and, in particular, the last two years) in Sandwell.”

He did not require statistical evidence because the potential indirect discrimination was “obvious”.

It is true that a person in one of the reasonable preference groups is permitted to join the register; but those who are homeless and cannot satisfy the 10-year residence requirement are placed in band D rather than in any higher band. That reduces their chances of being allocated accommodation; and in my judgment is a relevant disadvantage.

Mostyn J referred to the decision of Supperstone J in TW, but distinguished it at [18]:

“Travellers are a nomadic people. It is in their blood and is their fundamental tradition. Therefore, as a matter of probability it is surely much more likely that an Irish traveller will not complete the 10-year journey than his or her analogue. The traveller and the analogue are unalike cases which should be treated differently.”

That was not a valid ground for distinguishing the two protected groups. In Essop, Lady Hale said at [24]:

“… in none of the various definitions of indirect discrimination, is there any express requirement for an explanation of the reasons why a particular PCP puts one group at a disadvantage when compared with others. Thus there was no requirement in the 1975 Act that the claimant had to show why the proportion of women who could comply with the requirement was smaller than the proportion of men. It was enough that it was. There is no requirement in the Equality Act 2010 that the claimant show why the PCP puts one group sharing a particular protected characteristic at a particular disadvantage when compared with others. It is enough that it does.”

She repeated at [33]:

“In order to succeed in an indirect discrimination claim, it is not necessary to establish the reason for the particular disadvantage to which the group is put.”

It follows, in my judgment, that it was not legitimate to distinguish the decision of Supperstone J on the ground that the reason why Irish Travellers were put at a disadvantage by the PCP was known.

I therefore would hold that Mostyn J was wrong in finding that the PCP did not amount to indirect discrimination.

Section 149 of the Equality Act 2010 creates the public sector equality duty (“the PSED”). Although duties under the PSED were raised as specific grounds of appeal only in Mr Gullu’s appeal, I consider that performance of the PSED has a bearing on the approach both to the question of justification of indirect discrimination; and also to the alleged breach of section 11 of the Children Act 2004 in the formulation of the housing allocation policy.

The relevant parts of section 149 provide:

“(1)

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

(a)

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

(b)

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

(3)

Having due regard to the need to advance equality of opportunity between persons who share a relevant protected characteristic and persons who do not share it involves having due regard, in particular, to the need to—

(a)

remove or minimise disadvantages suffered by persons who share a relevant protected characteristic that are connected to that characteristic; …

(6)

Compliance with the duties in this section may involve treating some persons more favourably than others; but that is not to be taken as permitting conduct that would otherwise be prohibited by or under this Act.”

Compliance with the PSED requires the decision maker to be informed about what protected groups should be considered. That will involve a duty of inquiry, so that the decision-maker is properly informed before making a decision: Bracking v Secretary of State for Work and Pensions [2013] EWCA Civ 1345 at [26] (8).

Mr Rutledge submitted that in formulating the policy, it is not incumbent on the policy maker to assess the potential indirect discriminatory effect on every conceivable group that shares a protected characteristic. I consider that there is considerable force in that submission. It could not credibly be suggested, for example, that in formulating its housing allocation policy Hillingdon ought to have considered its potential impact on Native Americans or Australian aborigines without some specific cause to do so. It may well be the case that, in formulating a policy, a policy maker has conscientiously attempted to assess the potential indirect discriminatory effect on groups sharing protected characteristics; but has reasonably and in all good faith overlooked a particular such group. It would need to be shown that the policy maker had given adequate thought, conducting such inquiries as might be necessary,

Mr Gullu relies heavily on the fact that the equality impact assessments did not consider, or did not consider adequately, refugees and non-UK nationals respectively. So far as non-UK nationals are concerned, consideration was limited to the half sentence that I have quoted, and the recognition that recent arrivals in the country would be disadvantaged.

I would not hold that Hillingdon were in breach of the PSED in carrying out the initial equality impact assessment in 2013. At that stage it had not been shown that there was any reason for Hillingdon specifically to have considered non-UK nationals or refugees. But by the time of the 2016 assessment Mr Gullu had made his challenge in court. In the light of that challenge, I consider that Hillingdon ought at least to have considered the position of non-UK nationals. But they did not. I would therefore hold that Mr Gullu has established a breach of the PSED.

This question is common to both appeals. It is important, however, to distinguish between two different questions. One question is whether the policy, taken as a whole, contains sufficient safeguards to eliminate the indirect discrimination resulting from the impugned PCP. Another, and different, question is whether any indirect discrimination is justified. Hillingdon’s argument to some extent conflated these two questions.

The burden lies on the policy maker to justify the impugned PCP. It is not a legal requirement of justification that the reasons put forward in defence of the PCP must have been present to the mind of the policy maker at the time when the PCP was introduced. It is open to a policy maker to advance an ex post facto justification: R (Elias) v Secretary of State for Defence [2006] EWCA Civ 1293, [2006] 1 WLR 3213 at [129]; Seldon v Clarkson Wright & Jakes [2012] UKSC 16, [2012] ICR 716 at [59] and [76]. However, in the case of an ex post facto justification, the court will not have had the benefit of the considered decision of the policy maker. Thus, as Lord Kerr pointed out in Re Brewster [2017] UKSC 8, [2017] 1 WLR 519 at [52]:

“Obviously, if reasons are proffered in defence of a decision which were not present to the mind of the decision-maker at the time that it was made, this will call for greater scrutiny than would be appropriate if they could be shown to have influenced the decision-maker when the particular scheme was devised. Even retrospective judgments, however, if made within the sphere of expertise of the decision-maker, are worthy of respect, provided that they are made bona fide.”

In TW Supperstone J held at [50] that Hillingdon’s attempt to justify the indirect discrimination failed because of the “paucity and inadequacy of their evidence”. He said at [52] that in order to justify indirect discrimination “the authority must address the possible impacts” of the PCP. At [53] he said:

“The only consideration given to the impact of the residence requirement on “race” was in relation to BME or “white” residents. However, the adverse impacts on racial groups defined by “colour” were not the only ones that required consideration. I agree with Mr Wise and Mr Squires that a residence requirement, especially one as long as ten years, is almost certain to have a significant and adverse impact on Irish Travellers, yet the position of Irish Travellers does not appear to have been considered at all when the council conducted their equality impact assessment. No consideration has therefore been given as to whether the council has struck the correct balance between disadvantage to Irish Travellers and the aims of the residence requirement.”

Mr Rutledge says that this does not matter. Justification is concerned with the effect of the PCP in question; not with the process by which the PCP was devised. He also submits that Supperstone J was wrong in so far as he held that justification could only succeed on the basis of material actually considered by the policy maker before introducing the PCP in issue.

I do not think that it is entirely clear whether Supperstone J was, as a matter of law, confining the material on which Hillingdon could rely by way of justification to material that had been considered before the introduction of the PCP. What he said at [50] and [52] seem to me to be neutral on the question. Paragraph [53] does suggest that the judge moved from his statement that Irish Travellers were not considered at the time of the equality impact assessment to the conclusion (“therefore”) that no consideration “has” been given to their position. If that is what he meant, then I would agree with Mr Rutledge that it would have mis-stated the law.

In the end, however, I do not think that it matters. The material that Mr Rutledge relied on in defending the PCP was, in effect, the material contained in the 2013 equality impact assessment, the officers report to Cabinet (both of which preceded the introduction of the PCP) plus one paragraph in the witness statement of Ms Murphy; and the provisions of the policy itself. That was the position both before the judge, as the judge recorded at [42]; and before us. So in my judgment the judge was right to concentrate on that material.

As I have said, I do not consider that compliance with the PSED requires a policy maker to consider, in advance of formulating a policy, its potential impact on every conceivable protected group. There must be some trigger for considering a particular group. But that is not a complete answer. It may well be the case that, in formulating a policy, a policy maker has conscientiously attempted to assess the potential indirect discriminatory effect on a number of protected groups; but has reasonably and in all good faith overlooked a particular such group. In my judgment, it is incumbent on the policy maker once confronted with the omission, to justify the discrimination as regards that particular group. If a policy amounts to indirect discrimination against group E, I do not consider that it is an answer for the policy maker to say that it has considered groups A, B, C and D.

Mr Rutledge relied on a number of features of the scheme as amounting to what he called “safety valves”. What he meant by “the scheme” was not only Hillingdon’s allocation policy, but also the duties owed to the homeless under Part 7 of the Housing Act 1996. He pointed in particular, to the possibility of a higher banding being granted on account of hardship, the possibility of a direct offer outside the choice based letting scheme, and the continuing duty to house the unintentionally homeless in priority need under section 193 (2).

Supperstone J rejected this argument, which he dealt with shortly at [56]:

“As for the “safety valves”, I agree with Mr Wise and Mr Squires that they are not likely to benefit Irish Travellers any more than those with other ethnic origins. There is no evidence that the council gave consideration to whether they were likely to advantage persons who are Irish Travellers over other individuals with an equal need for housing.”

Mostyn J, on the other hand, did place reliance on other features of the policy. As he put it at [34]:

“It can be seen that the 10-year residence rule is not by any means the only gateway onto the register, nor is it, once you have got there, the only gateway to an uplift. There are other ways in, or up.”

This court considered a similar question in R (H) v Ealing. In that case, the evidence was that the impugned policy which disadvantaged certain protected groups was counterbalanced by other parts of the policy which conferred “significant priority on those groups”: see [76]. It was in those circumstances that the court was able to conclude that, taken overall, the policy did not disadvantage the protected groups.

As I have said, the key principle is that the goal is equality of outcome. If a PCP results in a relative disadvantage as regards one protected group, any measure relied on as a “safety valve” must overcome that relative disadvantage. Put simply, if the scales are tilted in one direction, adding an equal weight to each side of the scales does not eliminate the tilt. Mostyn J was wrong to hold otherwise.

In addition, there is no evidence that the “safety valves” within the allocation policy have actually operated to eliminate the disadvantage to the two protected groups in issue on these appeals. Indeed, in Mr McDonagh’s case his application to the hardship panel failed.

In my judgment, Supperstone J was correct to reject reliance on the “safety valves” as eliminating the indirect discrimination; and Mostyn J was wrong not to.

It follows, in my judgment, that taken as a whole, Hillingdon’s housing allocation policy does indirectly discriminate against the two protected groups in issue on these appeals.

I do not consider that this conclusion is affected by Mr Rutledge’s reliance on accommodation under Part 7 for those owed a duty under section 193 (2). Accommodation under Part 7 is likely to attract lesser security of tenure than accommodation under Part 6; is likely to be less affordable; and is likely to be out of borough, thus undermining further a person’s ability to satisfy the 10-year residency requirement. I also accept Mr Burton’s submission that discrimination in distributing commodity A is not offset by distribution of commodity B.

That is not necessarily the end of the case. It is possible, in some circumstances, that indirect discrimination is justified. It is common ground that the correct approach to justification of indirect discrimination is to follow the structure described by Lord Reed in Bank Mellat v HM Treasury (No 2) [2013] UKSC 39, [2014] AC 700 at [74]:

“… it is necessary to determine (1) whether the objective of the measure is sufficiently important to justify the limitation of a protected right, (2) whether the measure is rationally connected to the objective, (3) whether a less intrusive measure could have been used without unacceptably compromising the achievement of the objective, and (4) whether, balancing the severity of the measure's effects on the rights of the persons to whom it applies against the importance of the objective, to the extent that the measure will contribute to its achievement, the former outweighs the latter … In essence, the question at step four is whether the impact of the rights infringement is disproportionate to the likely benefit of the impugned measure.”

In some cases it may well be the case that there is indirect discrimination which can readily be justified. Thus, one exception to the 10-year residence requirement is people fleeing violence or harassment; and Hillingdon also accepts applicants under the West London Domestic Violence protocol. These exceptions may benefit women more than men; but no doubt that element of indirect discrimination is justifiable. In Essop at [29] Lady Hale gave other examples; such as fitness levels for firefighters or police officers.

Mostyn J dealt with the question of justification as follows at [38]:

“I turn to the four-limbed test: i) Is the objective of the rule, whether the qualification itself, or the uplift, sufficiently important to justify the limitation of a protected right? For these purposes a limitation of a protected right is assumed. In my judgment the answer to this question is plainly yes. The rule is obviously highly important and is an expression of national and local democratic processes. The actual limitation is, as I have explained, minimal and requires no more than that the claimant is treated the same as any other recent arrival.

ii)

Is the measure is rationally connected to that objective? The answer to this is plainly yes.

iii)

Could a less intrusive measure not have been used without unacceptably compromising the achievement of the objective? In my judgment to water down the rule for refugees to say five years would be quite wrong and arguably unlawful positive discrimination in their favour. The alternative ways in or up, set out above, entirely negate any merit which this argument might otherwise have.

iv)

When balancing the severity of the measure's effects on the rights of the persons to whom it applies against the importance of the objective, to the extent that the measure will contribute to its achievement, does the former outweigh the latter? The answer to this is plainly no. The latter greatly outweighs the former.”

He thus concluded at [39] that “the scheme is not manifestly without a reasonable foundation.”

There are a number of flaws in this reasoning. First, Mostyn J’s starting point was the individual analogue whom he had discussed in considering whether there was discrimination at all. For the reasons I have given, that was wrong. Second, he was wrong to say that the alternative ways in or up “entirely negate” the discrimination. For the reasons I have given, they do not. Third, he was wrong to say that giving preference to refugees would arguably amount to unlawful discrimination. Since refugees as such do not share a protected characteristic for the purposes of the Equality Act 2010, it would not. Fourth, Hillingdon led no evidence about the scale of the disadvantage that would be suffered by the protected group of non-UK nationals. It follows that an attempt to “balance” the severity of the effect against the importance of the objective is an impossible task. Put simply, there is nothing to put on one side of the scale. Moreover, Hillingdon did not explain how much (if at all) the relaxation of the 10-year residence qualification would compromise the objective of the scheme to reward long term residents of the borough. In this connection, it is relevant to note that the statutory guidance recommends a residence qualification period of at least two years, rather than the 10 that Hillingdon has chosen. Although it is fair to say that the officers’ first assessment did say that alternative lengths of residence qualification could be considered, there is no evidence that they actually were. Fifth, it is not for the court to “search around” for a justification that the policy maker has not advanced: R (Elias) v Secretary of State for Defence at [131]. Sixth, although the “manifestly without reasonable foundation” test applies to the first and second stages (and possibly to the third), it does not apply to the fourth: R (A) v Secretary of State for Health [2017] UKSC 41, [2017] 1 WLR 2492 at [33].

In my judgment, Hillingdon has not attempted to justify the indirect discrimination, in the sense of acknowledging that there is discrimination but then explaining why that discrimination is justified. Rather, its efforts have been directed towards showing (unsuccessfully in my judgment) that, taking the scheme as a whole, the negative effect on the relevant protected groups has been overcome. The nearest that Mr Rutledge came to advancing a positive case on justification was to submit that if the 10-year residence requirement were to be reduced, more people would qualify for inclusion on the housing register. That would have the knock-on effect that those who were on the register had a reduced chance of actually being housed. That, no doubt is

Accordingly, I would hold that the indirect discrimination has not yet been justified.

Mr Rutledge submitted that there was nothing that Hillingdon could lawfully do to mitigate any discrimination. It could not introduce a different residence qualification for Irish Travellers or non-UK nationals without being guilty of unlawful direct discrimination.

It is not, of course, for the court to rewrite the policy, as Ahmad makes abundantly clear. But the position is not necessarily as bleak as Mr Rutledge suggests. Refugees, as such, do not share a protected characteristic for the purposes of the Equality Act. But unlike economic migrants, asylum seekers have no choice about where they live during the pendency of their applications. That, as Ms Morgan’s evidence shows, is a substantial impediment to building a local connection. Mostyn J did not refer to this evidence at all. In addition, a refugee is vulnerable, during the ten-year journey, to being moved against his will; with the consequence that his period of residence in Hillingdon may be interrupted without his having any choice in the matter.

It would therefore be possible, without unlawfully discriminating against a protected group, to introduce a special rule for refugees; since they are not a protected group. That would mitigate the disadvantage suffered by the most affected members of the protected group of non-UK nationals. Any residual indirect discrimination against non-UK nationals might then be capable of justification.

In addition, although Hillingdon may not directly discriminate against a protected group, they may lawfully indirectly discriminate against such a group if the discrimination is justified. It would, for example, be possible to introduce a different residence requirement for persons who have had no fixed abode. Such a provision is likely to amount to indirect discrimination in favour of Irish Travellers and other ethnic groups with a similar lifestyle. But it is possible that such indirect discrimination could also be justified. There are, no doubt, other possibilities; but they are for Hillingdon to consider.

Mr Wise QC argued that a failure to consider the position of Irish Travellers with children amounted to a breach of Hillingdon’s duty under section 11 (2) of the

It is important to stress that we are concerned in this case with the formulation of policy. As the judge rightly recognised, there were consultations with Hillingdon’s Children’s Services department before the policy was formulated. The policy contains numerous provisions dealing with children. Thus, for example, children away from home studying are exempt from the 10-year residence requirement. Children under 18 may be granted a form of tenancy in certain circumstances. The number and age of children are taken into account in assessment of housing need. There is special provision for care leavers and for foster parents and adoptive parents. Indeed, Hillingdon considered that the policy overall was so child-friendly that they introduced a counterbalancing priority for childless couples.

The judge relied on two cases which concerned individual decision making. In such cases the decision-maker will have to focus on safeguarding and promoting the welfare of the particular child or children likely to be affected by the decision. But I do not consider that that is the case at the stage of formulating policy.

It is of course easy, with hindsight, to see that the particular position of Irish Travellers with children under 18 had not been considered. But just as I consider that it is not a breach of the PSED to fail to deal with every conceivable protected group, so I consider that it is not a breach of section 11 (2) to fail to consider every situation in which children might find themselves. We do not know how many Irish Travellers with children under 18 live or wish to live in Hillingdon. Of the two claimants in these appeals one has children under the age of 18; but the other does not.

It may well be that if Ms Ward were to apply to the hardship panel, section 11 (2) of the Children Act 2004 would require Hillingdon to consider the impact on her children’s education of refusing to move her up from band D. But that has not happened, and it is not the way that the case was put.

In respectful disagreement with the judge, I would hold that there has been no breach of section 11 (2) of the Children Act 2004. Supperstone J declared that the relevant parts of the policy under attack amounted to a breach of duty under section 11 (2). I would discharge that declaration. But I would not rule out the possibility that, now that the position of Irish Travellers with children under the age of 18 has been drawn

Mr Burton had an additional argument based on articles 8 and 14 of the European Convention on Human Rights. The argument was that Mr Gullu had a particular status for the purposes of article 14 (namely that of a refugee); and that the provision of social housing was “within the ambit” of article 8 which requires respect for family and private life.

The ambit of article 8 is not an easy question. It was the subject of conflicting obiter observations in R (H) v Ealing. Any further consideration of the question in this case would also be obiter; and is best left to a case in which it matters.

I would allow Mr Gullu’s appeal. In Ms Ward’s case Supperstone J declared that the relevant parts of the policy “unlawfully discriminate” against Irish ravellers.

However, both Mr Burton and Mr Wise accepted that although Hillingdon had not yet justified the discrimination, it was still open to them to do so if better evidence were to be provided. In those circumstances, I consider that the declaration is too prescriptively phrased. I would prefer to follow the form of declaration made by the Supreme Court in R (Coll) v Secretary of State for Justice [2017] UKSC 40, [2017] 1 WLR 2093 at [45]. That would result in a declaration that the impugned provisions of the policy constitute indirect discrimination against Irish Travellers and non-UK nationals which is unlawful unless justified; and that Hillingdon has not yet shown such justification.

Subject to that change of wording, I would dismiss Hillingdon’s appeal on the discrimination point; but allow it on the Children Act point.